diff options
Diffstat (limited to 'fs/gfs2')
45 files changed, 9488 insertions, 6760 deletions
diff --git a/fs/gfs2/Kconfig b/fs/gfs2/Kconfig index c465ae066c6..90c6a8faaec 100644 --- a/fs/gfs2/Kconfig +++ b/fs/gfs2/Kconfig @@ -1,10 +1,6 @@ config GFS2_FS tristate "GFS2 file system support" depends on (64BIT || LBDAF) - select DLM if GFS2_FS_LOCKING_DLM - select CONFIGFS_FS if GFS2_FS_LOCKING_DLM - select SYSFS if GFS2_FS_LOCKING_DLM - select IP_SCTP if DLM_SCTP select FS_POSIX_ACL select CRC32 select QUOTACTL @@ -24,12 +20,12 @@ config GFS2_FS be found here: http://sources.redhat.com/cluster The "nolock" lock module is now built in to GFS2 by default. If - you want to use the DLM, be sure to enable HOTPLUG and IPv4/6 - networking. + you want to use the DLM, be sure to enable IPv4/6 networking. config GFS2_FS_LOCKING_DLM bool "GFS2 DLM locking" - depends on (GFS2_FS!=n) && NET && INET && (IPV6 || IPV6=n) && HOTPLUG + depends on (GFS2_FS!=n) && NET && INET && (IPV6 || IPV6=n) && \ + CONFIGFS_FS && SYSFS && (DLM=y || DLM=GFS2_FS) help Multiple node locking module for GFS2 diff --git a/fs/gfs2/Makefile b/fs/gfs2/Makefile index 21f7e46da4c..86128202384 100644 --- a/fs/gfs2/Makefile +++ b/fs/gfs2/Makefile @@ -1,9 +1,9 @@ -EXTRA_CFLAGS := -I$(src) +ccflags-y := -I$(src) obj-$(CONFIG_GFS2_FS) += gfs2.o gfs2-y := acl.o bmap.o dir.o xattr.o glock.o \ - glops.o inode.o log.o lops.o main.o meta_io.o \ + glops.o log.o lops.o main.o meta_io.o \ aops.o dentry.o export.o file.o \ - ops_fstype.o ops_inode.o quota.o \ + ops_fstype.o inode.o quota.o \ recovery.o rgrp.o super.o sys.o trans.o util.o gfs2-$(CONFIG_GFS2_FS_LOCKING_DLM) += lock_dlm.o diff --git a/fs/gfs2/acl.c b/fs/gfs2/acl.c index 7118f1a780a..3088e2a38e3 100644 --- a/fs/gfs2/acl.c +++ b/fs/gfs2/acl.c @@ -38,8 +38,9 @@ static const char *gfs2_acl_name(int type) return NULL; } -static struct posix_acl *gfs2_acl_get(struct gfs2_inode *ip, int type) +struct posix_acl *gfs2_get_acl(struct inode *inode, int type) { + struct gfs2_inode *ip = GFS2_I(inode); struct posix_acl *acl; const char *name; char *data; @@ -48,10 +49,6 @@ static struct posix_acl *gfs2_acl_get(struct gfs2_inode *ip, int type) if (!ip->i_eattr) return NULL; - acl = get_cached_acl(&ip->i_inode, type); - if (acl != ACL_NOT_CACHED) - return acl; - name = gfs2_acl_name(type); if (name == NULL) return ERR_PTR(-EINVAL); @@ -62,57 +59,12 @@ static struct posix_acl *gfs2_acl_get(struct gfs2_inode *ip, int type) if (len == 0) return NULL; - acl = posix_acl_from_xattr(data, len); + acl = posix_acl_from_xattr(&init_user_ns, data, len); kfree(data); return acl; } -/** - * gfs2_check_acl - Check an ACL to see if we're allowed to do something - * @inode: the file we want to do something to - * @mask: what we want to do - * - * Returns: errno - */ - -int gfs2_check_acl(struct inode *inode, int mask, unsigned int flags) -{ - struct posix_acl *acl; - int error; - - if (flags & IPERM_FLAG_RCU) - return -ECHILD; - - acl = gfs2_acl_get(GFS2_I(inode), ACL_TYPE_ACCESS); - if (IS_ERR(acl)) - return PTR_ERR(acl); - - if (acl) { - error = posix_acl_permission(inode, acl, mask); - posix_acl_release(acl); - return error; - } - - return -EAGAIN; -} - -static int gfs2_set_mode(struct inode *inode, mode_t mode) -{ - int error = 0; - - if (mode != inode->i_mode) { - struct iattr iattr; - - iattr.ia_valid = ATTR_MODE; - iattr.ia_mode = mode; - - error = gfs2_setattr_simple(GFS2_I(inode), &iattr); - } - - return error; -} - -static int gfs2_acl_set(struct inode *inode, int type, struct posix_acl *acl) +int gfs2_set_acl(struct inode *inode, struct posix_acl *acl, int type) { int error; int len; @@ -120,232 +72,50 @@ static int gfs2_acl_set(struct inode *inode, int type, struct posix_acl *acl) const char *name = gfs2_acl_name(type); BUG_ON(name == NULL); - len = posix_acl_to_xattr(acl, NULL, 0); - if (len == 0) - return 0; - data = kmalloc(len, GFP_NOFS); - if (data == NULL) - return -ENOMEM; - error = posix_acl_to_xattr(acl, data, len); - if (error < 0) - goto out; - error = __gfs2_xattr_set(inode, name, data, len, 0, GFS2_EATYPE_SYS); - if (!error) - set_cached_acl(inode, type, acl); -out: - kfree(data); - return error; -} - -int gfs2_acl_create(struct gfs2_inode *dip, struct inode *inode) -{ - struct gfs2_sbd *sdp = GFS2_SB(&dip->i_inode); - struct posix_acl *acl, *clone; - mode_t mode = inode->i_mode; - int error = 0; - - if (!sdp->sd_args.ar_posix_acl) - return 0; - if (S_ISLNK(inode->i_mode)) - return 0; - - acl = gfs2_acl_get(dip, ACL_TYPE_DEFAULT); - if (IS_ERR(acl)) - return PTR_ERR(acl); - if (!acl) { - mode &= ~current_umask(); - if (mode != inode->i_mode) - error = gfs2_set_mode(inode, mode); - return error; - } - - if (S_ISDIR(inode->i_mode)) { - error = gfs2_acl_set(inode, ACL_TYPE_DEFAULT, acl); - if (error) - goto out; - } - - clone = posix_acl_clone(acl, GFP_NOFS); - error = -ENOMEM; - if (!clone) - goto out; - posix_acl_release(acl); - acl = clone; - error = posix_acl_create_masq(acl, &mode); - if (error < 0) - goto out; - if (error == 0) - goto munge; + if (acl->a_count > GFS2_ACL_MAX_ENTRIES(GFS2_SB(inode))) + return -E2BIG; - error = gfs2_acl_set(inode, ACL_TYPE_ACCESS, acl); - if (error) - goto out; -munge: - error = gfs2_set_mode(inode, mode); -out: - posix_acl_release(acl); - return error; -} + if (type == ACL_TYPE_ACCESS) { + umode_t mode = inode->i_mode; -int gfs2_acl_chmod(struct gfs2_inode *ip, struct iattr *attr) -{ - struct posix_acl *acl, *clone; - char *data; - unsigned int len; - int error; + error = posix_acl_equiv_mode(acl, &mode); + if (error < 0) + return error; - acl = gfs2_acl_get(ip, ACL_TYPE_ACCESS); - if (IS_ERR(acl)) - return PTR_ERR(acl); - if (!acl) - return gfs2_setattr_simple(ip, attr); + if (error == 0) + acl = NULL; - clone = posix_acl_clone(acl, GFP_NOFS); - error = -ENOMEM; - if (!clone) - goto out; - posix_acl_release(acl); - acl = clone; + if (mode != inode->i_mode) { + inode->i_mode = mode; + mark_inode_dirty(inode); + } + } - error = posix_acl_chmod_masq(acl, attr->ia_mode); - if (!error) { - len = posix_acl_to_xattr(acl, NULL, 0); + if (acl) { + len = posix_acl_to_xattr(&init_user_ns, acl, NULL, 0); + if (len == 0) + return 0; data = kmalloc(len, GFP_NOFS); - error = -ENOMEM; if (data == NULL) + return -ENOMEM; + error = posix_acl_to_xattr(&init_user_ns, acl, data, len); + if (error < 0) goto out; - posix_acl_to_xattr(acl, data, len); - error = gfs2_xattr_acl_chmod(ip, attr, data); - kfree(data); - set_cached_acl(&ip->i_inode, ACL_TYPE_ACCESS, acl); + } else { + data = NULL; + len = 0; } -out: - posix_acl_release(acl); - return error; -} - -static int gfs2_acl_type(const char *name) -{ - if (strcmp(name, GFS2_POSIX_ACL_ACCESS) == 0) - return ACL_TYPE_ACCESS; - if (strcmp(name, GFS2_POSIX_ACL_DEFAULT) == 0) - return ACL_TYPE_DEFAULT; - return -EINVAL; -} - -static int gfs2_xattr_system_get(struct dentry *dentry, const char *name, - void *buffer, size_t size, int xtype) -{ - struct inode *inode = dentry->d_inode; - struct gfs2_sbd *sdp = GFS2_SB(inode); - struct posix_acl *acl; - int type; - int error; - - if (!sdp->sd_args.ar_posix_acl) - return -EOPNOTSUPP; - - type = gfs2_acl_type(name); - if (type < 0) - return type; - - acl = gfs2_acl_get(GFS2_I(inode), type); - if (IS_ERR(acl)) - return PTR_ERR(acl); - if (acl == NULL) - return -ENODATA; - - error = posix_acl_to_xattr(acl, buffer, size); - posix_acl_release(acl); - - return error; -} - -static int gfs2_xattr_system_set(struct dentry *dentry, const char *name, - const void *value, size_t size, int flags, - int xtype) -{ - struct inode *inode = dentry->d_inode; - struct gfs2_sbd *sdp = GFS2_SB(inode); - struct posix_acl *acl = NULL; - int error = 0, type; - - if (!sdp->sd_args.ar_posix_acl) - return -EOPNOTSUPP; - - type = gfs2_acl_type(name); - if (type < 0) - return type; - if (flags & XATTR_CREATE) - return -EINVAL; - if (type == ACL_TYPE_DEFAULT && !S_ISDIR(inode->i_mode)) - return value ? -EACCES : 0; - if ((current_fsuid() != inode->i_uid) && !capable(CAP_FOWNER)) - return -EPERM; - if (S_ISLNK(inode->i_mode)) - return -EOPNOTSUPP; - - if (!value) - goto set_acl; - - acl = posix_acl_from_xattr(value, size); - if (!acl) { - /* - * acl_set_file(3) may request that we set default ACLs with - * zero length -- defend (gracefully) against that here. - */ - goto out; - } - if (IS_ERR(acl)) { - error = PTR_ERR(acl); - goto out; - } - - error = posix_acl_valid(acl); + error = __gfs2_xattr_set(inode, name, data, len, 0, GFS2_EATYPE_SYS); if (error) - goto out_release; - - error = -EINVAL; - if (acl->a_count > GFS2_ACL_MAX_ENTRIES) - goto out_release; - - if (type == ACL_TYPE_ACCESS) { - mode_t mode = inode->i_mode; - error = posix_acl_equiv_mode(acl, &mode); - - if (error <= 0) { - posix_acl_release(acl); - acl = NULL; - - if (error < 0) - return error; - } - - error = gfs2_set_mode(inode, mode); - if (error) - goto out_release; - } + goto out; -set_acl: - error = __gfs2_xattr_set(inode, name, value, size, 0, GFS2_EATYPE_SYS); - if (!error) { - if (acl) - set_cached_acl(inode, type, acl); - else - forget_cached_acl(inode, type); - } -out_release: - posix_acl_release(acl); + if (acl) + set_cached_acl(inode, type, acl); + else + forget_cached_acl(inode, type); out: + kfree(data); return error; } - -const struct xattr_handler gfs2_xattr_system_handler = { - .prefix = XATTR_SYSTEM_PREFIX, - .flags = GFS2_EATYPE_SYS, - .get = gfs2_xattr_system_get, - .set = gfs2_xattr_system_set, -}; - diff --git a/fs/gfs2/acl.h b/fs/gfs2/acl.h index a93907c8159..2d65ec4cd4b 100644 --- a/fs/gfs2/acl.h +++ b/fs/gfs2/acl.h @@ -14,11 +14,9 @@ #define GFS2_POSIX_ACL_ACCESS "posix_acl_access" #define GFS2_POSIX_ACL_DEFAULT "posix_acl_default" -#define GFS2_ACL_MAX_ENTRIES 25 +#define GFS2_ACL_MAX_ENTRIES(sdp) ((300 << (sdp)->sd_sb.sb_bsize_shift) >> 12) -extern int gfs2_check_acl(struct inode *inode, int mask, unsigned int); -extern int gfs2_acl_create(struct gfs2_inode *dip, struct inode *inode); -extern int gfs2_acl_chmod(struct gfs2_inode *ip, struct iattr *attr); -extern const struct xattr_handler gfs2_xattr_system_handler; +extern struct posix_acl *gfs2_get_acl(struct inode *inode, int type); +extern int gfs2_set_acl(struct inode *inode, struct posix_acl *acl, int type); #endif /* __ACL_DOT_H__ */ diff --git a/fs/gfs2/aops.c b/fs/gfs2/aops.c index 4f36f8832b9..805b37fed63 100644 --- a/fs/gfs2/aops.c +++ b/fs/gfs2/aops.c @@ -20,6 +20,8 @@ #include <linux/swap.h> #include <linux/gfs2_ondisk.h> #include <linux/backing-dev.h> +#include <linux/aio.h> +#include <trace/events/writeback.h> #include "gfs2.h" #include "incore.h" @@ -36,8 +38,8 @@ #include "glops.h" -void gfs2_page_add_databufs(struct gfs2_inode *ip, struct page *page, - unsigned int from, unsigned int to) +static void gfs2_page_add_databufs(struct gfs2_inode *ip, struct page *page, + unsigned int from, unsigned int to) { struct buffer_head *head = page_buffers(page); unsigned int bsize = head->b_size; @@ -51,7 +53,7 @@ void gfs2_page_add_databufs(struct gfs2_inode *ip, struct page *page, continue; if (gfs2_is_jdata(ip)) set_buffer_uptodate(bh); - gfs2_trans_add_bh(ip->i_gl, bh, 0); + gfs2_trans_add_data(ip->i_gl, bh); } } @@ -109,7 +111,7 @@ static int gfs2_writepage_common(struct page *page, /* Is the page fully outside i_size? (truncate in progress) */ offset = i_size & (PAGE_CACHE_SIZE-1); if (page->index > end_index || (page->index == end_index && !offset)) { - page->mapping->a_ops->invalidatepage(page, 0); + page->mapping->a_ops->invalidatepage(page, 0, PAGE_CACHE_SIZE); goto out; } return 1; @@ -121,14 +123,13 @@ out: } /** - * gfs2_writeback_writepage - Write page for writeback mappings + * gfs2_writepage - Write page for writeback mappings * @page: The page * @wbc: The writeback control * */ -static int gfs2_writeback_writepage(struct page *page, - struct writeback_control *wbc) +static int gfs2_writepage(struct page *page, struct writeback_control *wbc) { int ret; @@ -140,32 +141,6 @@ static int gfs2_writeback_writepage(struct page *page, } /** - * gfs2_ordered_writepage - Write page for ordered data files - * @page: The page to write - * @wbc: The writeback control - * - */ - -static int gfs2_ordered_writepage(struct page *page, - struct writeback_control *wbc) -{ - struct inode *inode = page->mapping->host; - struct gfs2_inode *ip = GFS2_I(inode); - int ret; - - ret = gfs2_writepage_common(page, wbc); - if (ret <= 0) - return ret; - - if (!page_has_buffers(page)) { - create_empty_buffers(page, inode->i_sb->s_blocksize, - (1 << BH_Dirty)|(1 << BH_Uptodate)); - } - gfs2_page_add_databufs(ip, page, 0, inode->i_sb->s_blocksize-1); - return block_write_full_page(page, gfs2_get_block_noalloc, wbc); -} - -/** * __gfs2_jdata_writepage - The core of jdata writepage * @page: The page to write * @wbc: The writeback control @@ -230,16 +205,14 @@ out_ignore: } /** - * gfs2_writeback_writepages - Write a bunch of dirty pages back to disk + * gfs2_writepages - Write a bunch of dirty pages back to disk * @mapping: The mapping to write * @wbc: Write-back control * - * For the data=writeback case we can already ignore buffer heads - * and write whole extents at once. This is a big reduction in the - * number of I/O requests we send and the bmap calls we make in this case. + * Used for both ordered and writeback modes. */ -static int gfs2_writeback_writepages(struct address_space *mapping, - struct writeback_control *wbc) +static int gfs2_writepages(struct address_space *mapping, + struct writeback_control *wbc) { return mpage_writepages(mapping, wbc, gfs2_get_block_noalloc); } @@ -258,13 +231,11 @@ static int gfs2_writeback_writepages(struct address_space *mapping, static int gfs2_write_jdata_pagevec(struct address_space *mapping, struct writeback_control *wbc, struct pagevec *pvec, - int nr_pages, pgoff_t end) + int nr_pages, pgoff_t end, + pgoff_t *done_index) { struct inode *inode = mapping->host; struct gfs2_sbd *sdp = GFS2_SB(inode); - loff_t i_size = i_size_read(inode); - pgoff_t end_index = i_size >> PAGE_CACHE_SHIFT; - unsigned offset = i_size & (PAGE_CACHE_SIZE-1); unsigned nrblocks = nr_pages * (PAGE_CACHE_SIZE/inode->i_sb->s_blocksize); int i; int ret; @@ -276,39 +247,83 @@ static int gfs2_write_jdata_pagevec(struct address_space *mapping, for(i = 0; i < nr_pages; i++) { struct page *page = pvec->pages[i]; + /* + * At this point, the page may be truncated or + * invalidated (changing page->mapping to NULL), or + * even swizzled back from swapper_space to tmpfs file + * mapping. However, page->index will not change + * because we have a reference on the page. + */ + if (page->index > end) { + /* + * can't be range_cyclic (1st pass) because + * end == -1 in that case. + */ + ret = 1; + break; + } + + *done_index = page->index; + lock_page(page); if (unlikely(page->mapping != mapping)) { +continue_unlock: unlock_page(page); continue; } - if (!wbc->range_cyclic && page->index > end) { - ret = 1; - unlock_page(page); - continue; + if (!PageDirty(page)) { + /* someone wrote it for us */ + goto continue_unlock; } - if (wbc->sync_mode != WB_SYNC_NONE) - wait_on_page_writeback(page); - - if (PageWriteback(page) || - !clear_page_dirty_for_io(page)) { - unlock_page(page); - continue; + if (PageWriteback(page)) { + if (wbc->sync_mode != WB_SYNC_NONE) + wait_on_page_writeback(page); + else + goto continue_unlock; } - /* Is the page fully outside i_size? (truncate in progress) */ - if (page->index > end_index || (page->index == end_index && !offset)) { - page->mapping->a_ops->invalidatepage(page, 0); - unlock_page(page); - continue; - } + BUG_ON(PageWriteback(page)); + if (!clear_page_dirty_for_io(page)) + goto continue_unlock; + + trace_wbc_writepage(wbc, mapping->backing_dev_info); ret = __gfs2_jdata_writepage(page, wbc); + if (unlikely(ret)) { + if (ret == AOP_WRITEPAGE_ACTIVATE) { + unlock_page(page); + ret = 0; + } else { + + /* + * done_index is set past this page, + * so media errors will not choke + * background writeout for the entire + * file. This has consequences for + * range_cyclic semantics (ie. it may + * not be suitable for data integrity + * writeout). + */ + *done_index = page->index + 1; + ret = 1; + break; + } + } - if (ret || (--(wbc->nr_to_write) <= 0)) + /* + * We stop writing back only if we are not doing + * integrity sync. In case of integrity sync we have to + * keep going until we have written all the pages + * we tagged for writeback prior to entering this loop. + */ + if (--wbc->nr_to_write <= 0 && wbc->sync_mode == WB_SYNC_NONE) { ret = 1; + break; + } + } gfs2_trans_end(sdp); return ret; @@ -333,51 +348,69 @@ static int gfs2_write_cache_jdata(struct address_space *mapping, int done = 0; struct pagevec pvec; int nr_pages; + pgoff_t uninitialized_var(writeback_index); pgoff_t index; pgoff_t end; - int scanned = 0; + pgoff_t done_index; + int cycled; int range_whole = 0; + int tag; pagevec_init(&pvec, 0); if (wbc->range_cyclic) { - index = mapping->writeback_index; /* Start from prev offset */ + writeback_index = mapping->writeback_index; /* prev offset */ + index = writeback_index; + if (index == 0) + cycled = 1; + else + cycled = 0; end = -1; } else { index = wbc->range_start >> PAGE_CACHE_SHIFT; end = wbc->range_end >> PAGE_CACHE_SHIFT; if (wbc->range_start == 0 && wbc->range_end == LLONG_MAX) range_whole = 1; - scanned = 1; + cycled = 1; /* ignore range_cyclic tests */ } + if (wbc->sync_mode == WB_SYNC_ALL || wbc->tagged_writepages) + tag = PAGECACHE_TAG_TOWRITE; + else + tag = PAGECACHE_TAG_DIRTY; retry: - while (!done && (index <= end) && - (nr_pages = pagevec_lookup_tag(&pvec, mapping, &index, - PAGECACHE_TAG_DIRTY, - min(end - index, (pgoff_t)PAGEVEC_SIZE-1) + 1))) { - scanned = 1; - ret = gfs2_write_jdata_pagevec(mapping, wbc, &pvec, nr_pages, end); + if (wbc->sync_mode == WB_SYNC_ALL || wbc->tagged_writepages) + tag_pages_for_writeback(mapping, index, end); + done_index = index; + while (!done && (index <= end)) { + nr_pages = pagevec_lookup_tag(&pvec, mapping, &index, tag, + min(end - index, (pgoff_t)PAGEVEC_SIZE-1) + 1); + if (nr_pages == 0) + break; + + ret = gfs2_write_jdata_pagevec(mapping, wbc, &pvec, nr_pages, end, &done_index); if (ret) done = 1; if (ret > 0) ret = 0; - pagevec_release(&pvec); cond_resched(); } - if (!scanned && !done) { + if (!cycled && !done) { /* + * range_cyclic: * We hit the last page and there is more work to be done: wrap * back to the start of the file */ - scanned = 1; + cycled = 1; index = 0; + end = writeback_index - 1; goto retry; } if (wbc->range_cyclic || (range_whole && wbc->nr_to_write > 0)) - mapping->writeback_index = index; + mapping->writeback_index = done_index; + return ret; } @@ -398,7 +431,7 @@ static int gfs2_jdata_writepages(struct address_space *mapping, ret = gfs2_write_cache_jdata(mapping, wbc); if (ret == 0 && wbc->sync_mode == WB_SYNC_ALL) { - gfs2_log_flush(sdp, ip->i_gl); + gfs2_log_flush(sdp, ip->i_gl, NORMAL_FLUSH); ret = gfs2_write_cache_jdata(mapping, wbc); } return ret; @@ -434,12 +467,12 @@ static int stuffed_readpage(struct gfs2_inode *ip, struct page *page) if (error) return error; - kaddr = kmap_atomic(page, KM_USER0); + kaddr = kmap_atomic(page); if (dsize > (dibh->b_size - sizeof(struct gfs2_dinode))) dsize = (dibh->b_size - sizeof(struct gfs2_dinode)); memcpy(kaddr, dibh->b_data + sizeof(struct gfs2_dinode), dsize); memset(kaddr + dsize, 0, PAGE_CACHE_SIZE - dsize); - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); flush_dcache_page(page); brelse(dibh); SetPageUptodate(page); @@ -517,15 +550,14 @@ out: /** * gfs2_internal_read - read an internal file * @ip: The gfs2 inode - * @ra_state: The readahead state (or NULL for no readahead) * @buf: The buffer to fill * @pos: The file position * @size: The amount to read * */ -int gfs2_internal_read(struct gfs2_inode *ip, struct file_ra_state *ra_state, - char *buf, loff_t *pos, unsigned size) +int gfs2_internal_read(struct gfs2_inode *ip, char *buf, loff_t *pos, + unsigned size) { struct address_space *mapping = ip->i_inode.i_mapping; unsigned long index = *pos / PAGE_CACHE_SIZE; @@ -542,10 +574,9 @@ int gfs2_internal_read(struct gfs2_inode *ip, struct file_ra_state *ra_state, page = read_cache_page(mapping, index, __gfs2_readpage, NULL); if (IS_ERR(page)) return PTR_ERR(page); - p = kmap_atomic(page, KM_USER0); + p = kmap_atomic(page); memcpy(buf + copied, p + offset, amt); - kunmap_atomic(p, KM_USER0); - mark_page_accessed(page); + kunmap_atomic(p); page_cache_release(page); copied += amt; index++; @@ -613,9 +644,9 @@ static int gfs2_write_begin(struct file *file, struct address_space *mapping, struct gfs2_sbd *sdp = GFS2_SB(mapping->host); struct gfs2_inode *m_ip = GFS2_I(sdp->sd_statfs_inode); unsigned int data_blocks = 0, ind_blocks = 0, rblocks; + unsigned requested = 0; int alloc_required; int error = 0; - struct gfs2_alloc *al = NULL; pgoff_t index = pos >> PAGE_CACHE_SHIFT; unsigned from = pos & (PAGE_CACHE_SIZE - 1); struct page *page; @@ -639,18 +670,14 @@ static int gfs2_write_begin(struct file *file, struct address_space *mapping, gfs2_write_calc_reserv(ip, len, &data_blocks, &ind_blocks); if (alloc_required) { - al = gfs2_alloc_get(ip); - if (!al) { - error = -ENOMEM; - goto out_unlock; - } - + struct gfs2_alloc_parms ap = { .aflags = 0, }; error = gfs2_quota_lock_check(ip); if (error) - goto out_alloc_put; + goto out_unlock; - al->al_requested = data_blocks + ind_blocks; - error = gfs2_inplace_reserve(ip); + requested = data_blocks + ind_blocks; + ap.target = requested; + error = gfs2_inplace_reserve(ip, &ap); if (error) goto out_qunlock; } @@ -663,7 +690,7 @@ static int gfs2_write_begin(struct file *file, struct address_space *mapping, if (&ip->i_inode == sdp->sd_rindex) rblocks += 2 * RES_STATFS; if (alloc_required) - rblocks += gfs2_rg_blocks(al); + rblocks += gfs2_rg_blocks(ip, requested); error = gfs2_trans_begin(sdp, rblocks, PAGE_CACHE_SIZE/sdp->sd_sb.sb_bsize); @@ -695,6 +722,7 @@ out: if (error == 0) return 0; + unlock_page(page); page_cache_release(page); gfs2_trans_end(sdp); @@ -709,8 +737,6 @@ out_trans_fail: gfs2_inplace_release(ip); out_qunlock: gfs2_quota_unlock(ip); -out_alloc_put: - gfs2_alloc_put(ip); } out_unlock: if (&ip->i_inode == sdp->sd_rindex) { @@ -786,14 +812,13 @@ static int gfs2_stuffed_write_end(struct inode *inode, struct buffer_head *dibh, u64 to = pos + copied; void *kaddr; unsigned char *buf = dibh->b_data + sizeof(struct gfs2_dinode); - struct gfs2_dinode *di = (struct gfs2_dinode *)dibh->b_data; BUG_ON((pos + len) > (dibh->b_size - sizeof(struct gfs2_dinode))); - kaddr = kmap_atomic(page, KM_USER0); + kaddr = kmap_atomic(page); memcpy(buf + pos, kaddr + pos, copied); memset(kaddr + pos + copied, 0, len - copied); flush_dcache_page(page); - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); if (!PageUptodate(page)) SetPageUptodate(page); @@ -803,13 +828,12 @@ static int gfs2_stuffed_write_end(struct inode *inode, struct buffer_head *dibh, if (copied) { if (inode->i_size < to) i_size_write(inode, to); - gfs2_dinode_out(ip, di); mark_inode_dirty(inode); } if (inode == sdp->sd_rindex) { adjust_fs_space(inode); - ip->i_gh.gh_flags |= GL_NOCACHE; + sdp->sd_rindex_uptodate = 0; } brelse(dibh); @@ -849,10 +873,11 @@ static int gfs2_write_end(struct file *file, struct address_space *mapping, struct gfs2_sbd *sdp = GFS2_SB(inode); struct gfs2_inode *m_ip = GFS2_I(sdp->sd_statfs_inode); struct buffer_head *dibh; - struct gfs2_alloc *al = ip->i_alloc; unsigned int from = pos & (PAGE_CACHE_SIZE - 1); unsigned int to = from + len; int ret; + struct gfs2_trans *tr = current->journal_info; + BUG_ON(!tr); BUG_ON(gfs2_glock_is_locked_by_me(ip->i_gl) == NULL); @@ -863,8 +888,6 @@ static int gfs2_write_end(struct file *file, struct address_space *mapping, goto failed; } - gfs2_trans_add_bh(ip->i_gl, dibh, 1); - if (gfs2_is_stuffed(ip)) return gfs2_stuffed_write_end(inode, dibh, pos, len, copied, page); @@ -872,24 +895,23 @@ static int gfs2_write_end(struct file *file, struct address_space *mapping, gfs2_page_add_databufs(ip, page, from, to); ret = generic_write_end(file, mapping, pos, len, copied, page, fsdata); - if (ret > 0) { - gfs2_dinode_out(ip, dibh->b_data); - mark_inode_dirty(inode); - } + if (tr->tr_num_buf_new) + __mark_inode_dirty(inode, I_DIRTY_DATASYNC); + else + gfs2_trans_add_meta(ip->i_gl, dibh); + if (inode == sdp->sd_rindex) { adjust_fs_space(inode); - ip->i_gh.gh_flags |= GL_NOCACHE; + sdp->sd_rindex_uptodate = 0; } brelse(dibh); - gfs2_trans_end(sdp); failed: - if (al) { - gfs2_inplace_release(ip); + gfs2_trans_end(sdp); + gfs2_inplace_release(ip); + if (ip->i_res->rs_qa_qd_num) gfs2_quota_unlock(ip); - gfs2_alloc_put(ip); - } if (inode == sdp->sd_rindex) { gfs2_glock_dq(&m_ip->i_gh); gfs2_holder_uninit(&m_ip->i_gh); @@ -948,8 +970,8 @@ static void gfs2_discard(struct gfs2_sbd *sdp, struct buffer_head *bh) clear_buffer_dirty(bh); bd = bh->b_private; if (bd) { - if (!list_empty(&bd->bd_le.le_list) && !buffer_pinned(bh)) - list_del_init(&bd->bd_le.le_list); + if (!list_empty(&bd->bd_list) && !buffer_pinned(bh)) + list_del_init(&bd->bd_list); else gfs2_remove_from_journal(bh, current->journal_info, 0); } @@ -961,27 +983,33 @@ static void gfs2_discard(struct gfs2_sbd *sdp, struct buffer_head *bh) unlock_buffer(bh); } -static void gfs2_invalidatepage(struct page *page, unsigned long offset) +static void gfs2_invalidatepage(struct page *page, unsigned int offset, + unsigned int length) { struct gfs2_sbd *sdp = GFS2_SB(page->mapping->host); + unsigned int stop = offset + length; + int partial_page = (offset || length < PAGE_CACHE_SIZE); struct buffer_head *bh, *head; unsigned long pos = 0; BUG_ON(!PageLocked(page)); - if (offset == 0) + if (!partial_page) ClearPageChecked(page); if (!page_has_buffers(page)) goto out; bh = head = page_buffers(page); do { + if (pos + bh->b_size > stop) + return; + if (offset <= pos) gfs2_discard(sdp, bh); pos += bh->b_size; bh = bh->b_this_page; } while (bh != head); out: - if (offset == 0) + if (!partial_page) try_to_release_page(page, 0); } @@ -1012,11 +1040,11 @@ static int gfs2_ok_for_dio(struct gfs2_inode *ip, int rw, loff_t offset) static ssize_t gfs2_direct_IO(int rw, struct kiocb *iocb, - const struct iovec *iov, loff_t offset, - unsigned long nr_segs) + struct iov_iter *iter, loff_t offset) { struct file *file = iocb->ki_filp; struct inode *inode = file->f_mapping->host; + struct address_space *mapping = inode->i_mapping; struct gfs2_inode *ip = GFS2_I(inode); struct gfs2_holder gh; int rv; @@ -1037,11 +1065,41 @@ static ssize_t gfs2_direct_IO(int rw, struct kiocb *iocb, if (rv != 1) goto out; /* dio not valid, fall back to buffered i/o */ - rv = __blockdev_direct_IO(rw, iocb, inode, inode->i_sb->s_bdev, iov, - offset, nr_segs, gfs2_get_block_direct, - NULL, NULL, 0); + /* + * Now since we are holding a deferred (CW) lock at this point, you + * might be wondering why this is ever needed. There is a case however + * where we've granted a deferred local lock against a cached exclusive + * glock. That is ok provided all granted local locks are deferred, but + * it also means that it is possible to encounter pages which are + * cached and possibly also mapped. So here we check for that and sort + * them out ahead of the dio. The glock state machine will take care of + * everything else. + * + * If in fact the cached glock state (gl->gl_state) is deferred (CW) in + * the first place, mapping->nr_pages will always be zero. + */ + if (mapping->nrpages) { + loff_t lstart = offset & (PAGE_CACHE_SIZE - 1); + loff_t len = iov_iter_count(iter); + loff_t end = PAGE_ALIGN(offset + len) - 1; + + rv = 0; + if (len == 0) + goto out; + if (test_and_clear_bit(GIF_SW_PAGED, &ip->i_flags)) + unmap_shared_mapping_range(ip->i_inode.i_mapping, offset, len); + rv = filemap_write_and_wait_range(mapping, lstart, end); + if (rv) + goto out; + if (rw == WRITE) + truncate_inode_pages_range(mapping, lstart, end); + } + + rv = __blockdev_direct_IO(rw, iocb, inode, inode->i_sb->s_bdev, + iter, offset, + gfs2_get_block_direct, NULL, NULL, 0); out: - gfs2_glock_dq_m(1, &gh); + gfs2_glock_dq(&gh); gfs2_holder_uninit(&gh); return rv; } @@ -1068,55 +1126,52 @@ int gfs2_releasepage(struct page *page, gfp_t gfp_mask) return 0; gfs2_log_lock(sdp); + spin_lock(&sdp->sd_ail_lock); head = bh = page_buffers(page); do { if (atomic_read(&bh->b_count)) goto cannot_release; bd = bh->b_private; - if (bd && bd->bd_ail) + if (bd && bd->bd_tr) goto cannot_release; - gfs2_assert_warn(sdp, !buffer_pinned(bh)); - gfs2_assert_warn(sdp, !buffer_dirty(bh)); + if (buffer_pinned(bh) || buffer_dirty(bh)) + goto not_possible; bh = bh->b_this_page; } while(bh != head); - gfs2_log_unlock(sdp); + spin_unlock(&sdp->sd_ail_lock); head = bh = page_buffers(page); do { - gfs2_log_lock(sdp); bd = bh->b_private; if (bd) { gfs2_assert_warn(sdp, bd->bd_bh == bh); - gfs2_assert_warn(sdp, list_empty(&bd->bd_list_tr)); - if (!list_empty(&bd->bd_le.le_list)) { - if (!buffer_pinned(bh)) - list_del_init(&bd->bd_le.le_list); - else - bd = NULL; - } - if (bd) - bd->bd_bh = NULL; + if (!list_empty(&bd->bd_list)) + list_del_init(&bd->bd_list); + bd->bd_bh = NULL; bh->b_private = NULL; - } - gfs2_log_unlock(sdp); - if (bd) kmem_cache_free(gfs2_bufdata_cachep, bd); + } bh = bh->b_this_page; } while (bh != head); + gfs2_log_unlock(sdp); return try_to_free_buffers(page); + +not_possible: /* Should never happen */ + WARN_ON(buffer_dirty(bh)); + WARN_ON(buffer_pinned(bh)); cannot_release: + spin_unlock(&sdp->sd_ail_lock); gfs2_log_unlock(sdp); return 0; } static const struct address_space_operations gfs2_writeback_aops = { - .writepage = gfs2_writeback_writepage, - .writepages = gfs2_writeback_writepages, + .writepage = gfs2_writepage, + .writepages = gfs2_writepages, .readpage = gfs2_readpage, .readpages = gfs2_readpages, - .sync_page = block_sync_page, .write_begin = gfs2_write_begin, .write_end = gfs2_write_end, .bmap = gfs2_bmap, @@ -1129,10 +1184,10 @@ static const struct address_space_operations gfs2_writeback_aops = { }; static const struct address_space_operations gfs2_ordered_aops = { - .writepage = gfs2_ordered_writepage, + .writepage = gfs2_writepage, + .writepages = gfs2_writepages, .readpage = gfs2_readpage, .readpages = gfs2_readpages, - .sync_page = block_sync_page, .write_begin = gfs2_write_begin, .write_end = gfs2_write_end, .set_page_dirty = gfs2_set_page_dirty, @@ -1150,7 +1205,6 @@ static const struct address_space_operations gfs2_jdata_aops = { .writepages = gfs2_jdata_writepages, .readpage = gfs2_readpage, .readpages = gfs2_readpages, - .sync_page = block_sync_page, .write_begin = gfs2_write_begin, .write_end = gfs2_write_end, .set_page_dirty = gfs2_set_page_dirty, diff --git a/fs/gfs2/bmap.c b/fs/gfs2/bmap.c index 3c4039d5eef..e6ee5b6e8d9 100644 --- a/fs/gfs2/bmap.c +++ b/fs/gfs2/bmap.c @@ -10,6 +10,7 @@ #include <linux/spinlock.h> #include <linux/completion.h> #include <linux/buffer_head.h> +#include <linux/blkdev.h> #include <linux/gfs2_ondisk.h> #include <linux/crc32.h> @@ -21,6 +22,8 @@ #include "meta_io.h" #include "quota.h" #include "rgrp.h" +#include "log.h" +#include "super.h" #include "trans.h" #include "dir.h" #include "util.h" @@ -35,11 +38,6 @@ struct metapath { __u16 mp_list[GFS2_MAX_META_HEIGHT]; }; -typedef int (*block_call_t) (struct gfs2_inode *ip, struct buffer_head *dibh, - struct buffer_head *bh, __be64 *top, - __be64 *bottom, unsigned int height, - void *data); - struct strip_mine { int sm_first; unsigned int sm_height; @@ -63,7 +61,7 @@ static int gfs2_unstuffer_page(struct gfs2_inode *ip, struct buffer_head *dibh, int release = 0; if (!page || page->index) { - page = grab_cache_page(inode->i_mapping, 0); + page = find_or_create_page(inode->i_mapping, 0, GFP_NOFS); if (!page) return -ENOMEM; release = 1; @@ -96,7 +94,7 @@ static int gfs2_unstuffer_page(struct gfs2_inode *ip, struct buffer_head *dibh, if (!gfs2_is_jdata(ip)) mark_buffer_dirty(bh); if (!gfs2_is_writeback(ip)) - gfs2_trans_add_bh(ip->i_gl, bh, 0); + gfs2_trans_add_data(ip->i_gl, bh); if (release) { unlock_page(page); @@ -136,7 +134,7 @@ int gfs2_unstuff_dinode(struct gfs2_inode *ip, struct page *page) and write it out to disk */ unsigned int n = 1; - error = gfs2_alloc_block(ip, &block, &n); + error = gfs2_alloc_blocks(ip, &block, &n, 0, NULL); if (error) goto out_brelse; if (isdir) { @@ -156,7 +154,7 @@ int gfs2_unstuff_dinode(struct gfs2_inode *ip, struct page *page) /* Set up the pointer to the new block */ - gfs2_trans_add_bh(ip->i_gl, dibh, 1); + gfs2_trans_add_meta(ip->i_gl, dibh); di = (struct gfs2_dinode *)dibh->b_data; gfs2_buffer_clear_tail(dibh, sizeof(struct gfs2_dinode)); @@ -272,6 +270,30 @@ static inline __be64 *metapointer(unsigned int height, const struct metapath *mp return ((__be64 *)(bh->b_data + head_size)) + mp->mp_list[height]; } +static void gfs2_metapath_ra(struct gfs2_glock *gl, + const struct buffer_head *bh, const __be64 *pos) +{ + struct buffer_head *rabh; + const __be64 *endp = (const __be64 *)(bh->b_data + bh->b_size); + const __be64 *t; + + for (t = pos; t < endp; t++) { + if (!*t) + continue; + + rabh = gfs2_getbuf(gl, be64_to_cpu(*t), CREATE); + if (trylock_buffer(rabh)) { + if (!buffer_uptodate(rabh)) { + rabh->b_end_io = end_buffer_read_sync; + submit_bh(READA | REQ_META, rabh); + continue; + } + unlock_buffer(rabh); + } + brelse(rabh); + } +} + /** * lookup_metapath - Walk the metadata tree to a specific point * @ip: The inode @@ -303,7 +325,7 @@ static int lookup_metapath(struct gfs2_inode *ip, struct metapath *mp) if (!dblock) return x + 1; - ret = gfs2_meta_indirect_buffer(ip, x+1, dblock, 0, &mp->mp_bh[x+1]); + ret = gfs2_meta_indirect_buffer(ip, x+1, dblock, &mp->mp_bh[x+1]); if (ret) return ret; } @@ -384,7 +406,7 @@ static inline __be64 *gfs2_indirect_init(struct metapath *mp, BUG_ON(i < 1); BUG_ON(mp->mp_bh[i] != NULL); mp->mp_bh[i] = gfs2_meta_new(gl, bn); - gfs2_trans_add_bh(gl, mp->mp_bh[i], 1); + gfs2_trans_add_meta(gl, mp->mp_bh[i]); gfs2_metatype_set(mp->mp_bh[i], GFS2_METATYPE_IN, GFS2_FORMAT_IN); gfs2_buffer_clear_tail(mp->mp_bh[i], sizeof(struct gfs2_meta_header)); ptr += offset; @@ -431,12 +453,14 @@ static int gfs2_bmap_alloc(struct inode *inode, const sector_t lblock, { struct gfs2_inode *ip = GFS2_I(inode); struct gfs2_sbd *sdp = GFS2_SB(inode); + struct super_block *sb = sdp->sd_vfs; struct buffer_head *dibh = mp->mp_bh[0]; u64 bn, dblock = 0; unsigned n, i, blks, alloced = 0, iblks = 0, branch_start = 0; unsigned dblks = 0; unsigned ptrs_per_blk; const unsigned end_of_metadata = height - 1; + int ret; int eob = 0; enum alloc_state state; __be64 *ptr; @@ -445,7 +469,7 @@ static int gfs2_bmap_alloc(struct inode *inode, const sector_t lblock, BUG_ON(sheight < 1); BUG_ON(dibh == NULL); - gfs2_trans_add_bh(ip->i_gl, dibh, 1); + gfs2_trans_add_meta(ip->i_gl, dibh); if (height == sheight) { struct buffer_head *bh; @@ -480,7 +504,7 @@ static int gfs2_bmap_alloc(struct inode *inode, const sector_t lblock, do { int error; n = blks - alloced; - error = gfs2_alloc_block(ip, &bn, &n); + error = gfs2_alloc_blocks(ip, &bn, &n, 0, NULL); if (error) return error; alloced += n; @@ -521,7 +545,7 @@ static int gfs2_bmap_alloc(struct inode *inode, const sector_t lblock, /* Branching from existing tree */ case ALLOC_GROW_DEPTH: if (i > 1 && i < height) - gfs2_trans_add_bh(ip->i_gl, mp->mp_bh[i-1], 1); + gfs2_trans_add_meta(ip->i_gl, mp->mp_bh[i-1]); for (; i < height && n > 0; i++, n--) gfs2_indirect_init(mp, ip->i_gl, i, mp->mp_list[i-1], bn++); @@ -533,12 +557,21 @@ static int gfs2_bmap_alloc(struct inode *inode, const sector_t lblock, case ALLOC_DATA: BUG_ON(n > dblks); BUG_ON(mp->mp_bh[end_of_metadata] == NULL); - gfs2_trans_add_bh(ip->i_gl, mp->mp_bh[end_of_metadata], 1); + gfs2_trans_add_meta(ip->i_gl, mp->mp_bh[end_of_metadata]); dblks = n; ptr = metapointer(end_of_metadata, mp); dblock = bn; while (n-- > 0) *ptr++ = cpu_to_be64(bn++); + if (buffer_zeronew(bh_map)) { + ret = sb_issue_zeroout(sb, dblock, dblks, + GFP_NOFS); + if (ret) { + fs_err(sdp, + "Failed to zero data buffers\n"); + clear_buffer_zeronew(bh_map); + } + } break; } } while ((state != ALLOC_DATA) || !dblock); @@ -667,76 +700,6 @@ int gfs2_extent_map(struct inode *inode, u64 lblock, int *new, u64 *dblock, unsi } /** - * recursive_scan - recursively scan through the end of a file - * @ip: the inode - * @dibh: the dinode buffer - * @mp: the path through the metadata to the point to start - * @height: the height the recursion is at - * @block: the indirect block to look at - * @first: 1 if this is the first block - * @bc: the call to make for each piece of metadata - * @data: data opaque to this function to pass to @bc - * - * When this is first called @height and @block should be zero and - * @first should be 1. - * - * Returns: errno - */ - -static int recursive_scan(struct gfs2_inode *ip, struct buffer_head *dibh, - struct metapath *mp, unsigned int height, - u64 block, int first, block_call_t bc, - void *data) -{ - struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); - struct buffer_head *bh = NULL; - __be64 *top, *bottom; - u64 bn; - int error; - int mh_size = sizeof(struct gfs2_meta_header); - - if (!height) { - error = gfs2_meta_inode_buffer(ip, &bh); - if (error) - return error; - dibh = bh; - - top = (__be64 *)(bh->b_data + sizeof(struct gfs2_dinode)) + mp->mp_list[0]; - bottom = (__be64 *)(bh->b_data + sizeof(struct gfs2_dinode)) + sdp->sd_diptrs; - } else { - error = gfs2_meta_indirect_buffer(ip, height, block, 0, &bh); - if (error) - return error; - - top = (__be64 *)(bh->b_data + mh_size) + - (first ? mp->mp_list[height] : 0); - - bottom = (__be64 *)(bh->b_data + mh_size) + sdp->sd_inptrs; - } - - error = bc(ip, dibh, bh, top, bottom, height, data); - if (error) - goto out; - - if (height < ip->i_height - 1) - for (; top < bottom; top++, first = 0) { - if (!*top) - continue; - - bn = be64_to_cpu(*top); - - error = recursive_scan(ip, dibh, mp, height + 1, bn, - first, bc, data); - if (error) - break; - } - -out: - brelse(bh); - return error; -} - -/** * do_strip - Look for a layer a particular layer of the file and strip it off * @ip: the inode * @dibh: the dinode buffer @@ -744,26 +707,29 @@ out: * @top: The first pointer in the buffer * @bottom: One more than the last pointer * @height: the height this buffer is at - * @data: a pointer to a struct strip_mine + * @sm: a pointer to a struct strip_mine * * Returns: errno */ static int do_strip(struct gfs2_inode *ip, struct buffer_head *dibh, struct buffer_head *bh, __be64 *top, __be64 *bottom, - unsigned int height, void *data) + unsigned int height, struct strip_mine *sm) { - struct strip_mine *sm = data; struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); struct gfs2_rgrp_list rlist; u64 bn, bstart; - u32 blen; + u32 blen, btotal; __be64 *p; unsigned int rg_blocks = 0; int metadata; unsigned int revokes = 0; int x; - int error = 0; + int error; + + error = gfs2_rindex_update(sdp); + if (error) + return error; if (!*top) sm->sm_first = 0; @@ -779,14 +745,8 @@ static int do_strip(struct gfs2_inode *ip, struct buffer_head *dibh, metadata = (height != ip->i_height - 1); if (metadata) revokes = (height) ? sdp->sd_inptrs : sdp->sd_diptrs; - - if (ip != GFS2_I(sdp->sd_rindex)) - error = gfs2_rindex_hold(sdp, &ip->i_alloc->al_ri_gh); - else if (!sdp->sd_rgrps) - error = gfs2_ri_update(ip); - - if (error) - return error; + else if (ip->i_depth) + revokes = sdp->sd_inptrs; memset(&rlist, 0, sizeof(struct gfs2_rgrp_list)); bstart = 0; @@ -802,7 +762,7 @@ static int do_strip(struct gfs2_inode *ip, struct buffer_head *dibh, blen++; else { if (bstart) - gfs2_rlist_add(sdp, &rlist, bstart); + gfs2_rlist_add(ip, &rlist, bstart); bstart = bn; blen = 1; @@ -810,7 +770,7 @@ static int do_strip(struct gfs2_inode *ip, struct buffer_head *dibh, } if (bstart) - gfs2_rlist_add(sdp, &rlist, bstart); + gfs2_rlist_add(ip, &rlist, bstart); else goto out; /* Nothing to do */ @@ -826,6 +786,9 @@ static int do_strip(struct gfs2_inode *ip, struct buffer_head *dibh, if (error) goto out_rlist; + if (gfs2_rs_active(ip->i_res)) /* needs to be done with the rgrp glock held */ + gfs2_rs_deltree(ip->i_res); + error = gfs2_trans_begin(sdp, rg_blocks + RES_DINODE + RES_INDIRECT + RES_STATFS + RES_QUOTA, revokes); @@ -834,11 +797,12 @@ static int do_strip(struct gfs2_inode *ip, struct buffer_head *dibh, down_write(&ip->i_rw_mutex); - gfs2_trans_add_bh(ip->i_gl, dibh, 1); - gfs2_trans_add_bh(ip->i_gl, bh, 1); + gfs2_trans_add_meta(ip->i_gl, dibh); + gfs2_trans_add_meta(ip->i_gl, bh); bstart = 0; blen = 0; + btotal = 0; for (p = top; p < bottom; p++) { if (!*p) @@ -850,10 +814,8 @@ static int do_strip(struct gfs2_inode *ip, struct buffer_head *dibh, blen++; else { if (bstart) { - if (metadata) - gfs2_free_meta(ip, bstart, blen); - else - gfs2_free_data(ip, bstart, blen); + __gfs2_free_blocks(ip, bstart, blen, metadata); + btotal += blen; } bstart = bn; @@ -864,12 +826,14 @@ static int do_strip(struct gfs2_inode *ip, struct buffer_head *dibh, gfs2_add_inode_blocks(&ip->i_inode, -1); } if (bstart) { - if (metadata) - gfs2_free_meta(ip, bstart, blen); - else - gfs2_free_data(ip, bstart, blen); + __gfs2_free_blocks(ip, bstart, blen, metadata); + btotal += blen; } + gfs2_statfs_change(sdp, 0, +btotal, 0); + gfs2_quota_change(ip, -(s64)btotal, ip->i_inode.i_uid, + ip->i_inode.i_gid); + ip->i_inode.i_mtime = ip->i_inode.i_ctime = CURRENT_TIME; gfs2_dinode_out(ip, dibh->b_data); @@ -883,12 +847,82 @@ out_rg_gunlock: out_rlist: gfs2_rlist_free(&rlist); out: - if (ip != GFS2_I(sdp->sd_rindex)) - gfs2_glock_dq_uninit(&ip->i_alloc->al_ri_gh); return error; } /** + * recursive_scan - recursively scan through the end of a file + * @ip: the inode + * @dibh: the dinode buffer + * @mp: the path through the metadata to the point to start + * @height: the height the recursion is at + * @block: the indirect block to look at + * @first: 1 if this is the first block + * @sm: data opaque to this function to pass to @bc + * + * When this is first called @height and @block should be zero and + * @first should be 1. + * + * Returns: errno + */ + +static int recursive_scan(struct gfs2_inode *ip, struct buffer_head *dibh, + struct metapath *mp, unsigned int height, + u64 block, int first, struct strip_mine *sm) +{ + struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); + struct buffer_head *bh = NULL; + __be64 *top, *bottom; + u64 bn; + int error; + int mh_size = sizeof(struct gfs2_meta_header); + + if (!height) { + error = gfs2_meta_inode_buffer(ip, &bh); + if (error) + return error; + dibh = bh; + + top = (__be64 *)(bh->b_data + sizeof(struct gfs2_dinode)) + mp->mp_list[0]; + bottom = (__be64 *)(bh->b_data + sizeof(struct gfs2_dinode)) + sdp->sd_diptrs; + } else { + error = gfs2_meta_indirect_buffer(ip, height, block, &bh); + if (error) + return error; + + top = (__be64 *)(bh->b_data + mh_size) + + (first ? mp->mp_list[height] : 0); + + bottom = (__be64 *)(bh->b_data + mh_size) + sdp->sd_inptrs; + } + + error = do_strip(ip, dibh, bh, top, bottom, height, sm); + if (error) + goto out; + + if (height < ip->i_height - 1) { + + gfs2_metapath_ra(ip->i_gl, bh, top); + + for (; top < bottom; top++, first = 0) { + if (!*top) + continue; + + bn = be64_to_cpu(*top); + + error = recursive_scan(ip, dibh, mp, height + 1, bn, + first, sm); + if (error) + break; + } + } +out: + brelse(bh); + return error; +} + + +/** * gfs2_block_truncate_page - Deal with zeroing out data for truncate * * This is partly borrowed from ext3. @@ -904,7 +938,7 @@ static int gfs2_block_truncate_page(struct address_space *mapping, loff_t from) struct page *page; int err; - page = grab_cache_page(mapping, index); + page = find_or_create_page(mapping, index, GFP_NOFS); if (!page) return 0; @@ -948,7 +982,7 @@ static int gfs2_block_truncate_page(struct address_space *mapping, loff_t from) } if (!gfs2_is_writeback(ip)) - gfs2_trans_add_bh(ip->i_gl, bh, 0); + gfs2_trans_add_data(ip->i_gl, bh); zero_user(page, offset, length); mark_buffer_dirty(bh); @@ -958,6 +992,41 @@ unlock: return err; } +#define GFS2_JTRUNC_REVOKES 8192 + +/** + * gfs2_journaled_truncate - Wrapper for truncate_pagecache for jdata files + * @inode: The inode being truncated + * @oldsize: The original (larger) size + * @newsize: The new smaller size + * + * With jdata files, we have to journal a revoke for each block which is + * truncated. As a result, we need to split this into separate transactions + * if the number of pages being truncated gets too large. + */ + +static int gfs2_journaled_truncate(struct inode *inode, u64 oldsize, u64 newsize) +{ + struct gfs2_sbd *sdp = GFS2_SB(inode); + u64 max_chunk = GFS2_JTRUNC_REVOKES * sdp->sd_vfs->s_blocksize; + u64 chunk; + int error; + + while (oldsize != newsize) { + chunk = oldsize - newsize; + if (chunk > max_chunk) + chunk = max_chunk; + truncate_pagecache(inode, oldsize - chunk); + oldsize -= chunk; + gfs2_trans_end(sdp); + error = gfs2_trans_begin(sdp, RES_DINODE, GFS2_JTRUNC_REVOKES); + if (error) + return error; + } + + return 0; +} + static int trunc_start(struct inode *inode, u64 oldsize, u64 newsize) { struct gfs2_inode *ip = GFS2_I(inode); @@ -967,8 +1036,10 @@ static int trunc_start(struct inode *inode, u64 oldsize, u64 newsize) int journaled = gfs2_is_jdata(ip); int error; - error = gfs2_trans_begin(sdp, - RES_DINODE + (journaled ? RES_JDATA : 0), 0); + if (journaled) + error = gfs2_trans_begin(sdp, RES_DINODE + RES_JDATA, GFS2_JTRUNC_REVOKES); + else + error = gfs2_trans_begin(sdp, RES_DINODE, 0); if (error) return error; @@ -976,7 +1047,7 @@ static int trunc_start(struct inode *inode, u64 oldsize, u64 newsize) if (error) goto out; - gfs2_trans_add_bh(ip->i_gl, dibh, 1); + gfs2_trans_add_meta(ip->i_gl, dibh); if (gfs2_is_stuffed(ip)) { gfs2_buffer_clear_tail(dibh, sizeof(struct gfs2_dinode) + newsize); @@ -993,7 +1064,16 @@ static int trunc_start(struct inode *inode, u64 oldsize, u64 newsize) ip->i_inode.i_mtime = ip->i_inode.i_ctime = CURRENT_TIME; gfs2_dinode_out(ip, dibh->b_data); - truncate_pagecache(inode, oldsize, newsize); + if (journaled) + error = gfs2_journaled_truncate(inode, oldsize, newsize); + else + truncate_pagecache(inode, newsize); + + if (error) { + brelse(dibh); + return error; + } + out_brelse: brelse(dibh); out: @@ -1015,27 +1095,26 @@ static int trunc_dealloc(struct gfs2_inode *ip, u64 size) lblock = (size - 1) >> sdp->sd_sb.sb_bsize_shift; find_metapath(sdp, lblock, &mp, ip->i_height); - if (!gfs2_alloc_get(ip)) - return -ENOMEM; + error = gfs2_rindex_update(sdp); + if (error) + return error; - error = gfs2_quota_hold(ip, NO_QUOTA_CHANGE, NO_QUOTA_CHANGE); + error = gfs2_quota_hold(ip, NO_UID_QUOTA_CHANGE, NO_GID_QUOTA_CHANGE); if (error) - goto out; + return error; while (height--) { struct strip_mine sm; sm.sm_first = !!size; sm.sm_height = height; - error = recursive_scan(ip, NULL, &mp, 0, 0, 1, do_strip, &sm); + error = recursive_scan(ip, NULL, &mp, 0, 0, 1, &sm); if (error) break; } gfs2_quota_unhold(ip); -out: - gfs2_alloc_put(ip); return error; } @@ -1059,11 +1138,12 @@ static int trunc_end(struct gfs2_inode *ip) ip->i_height = 0; ip->i_goal = ip->i_no_addr; gfs2_buffer_clear_tail(dibh, sizeof(struct gfs2_dinode)); + gfs2_ordered_del_inode(ip); } ip->i_inode.i_mtime = ip->i_inode.i_ctime = CURRENT_TIME; ip->i_diskflags &= ~GFS2_DIF_TRUNC_IN_PROG; - gfs2_trans_add_bh(ip->i_gl, dibh, 1); + gfs2_trans_add_meta(ip->i_gl, dibh); gfs2_dinode_out(ip, dibh->b_data); brelse(dibh); @@ -1126,7 +1206,7 @@ void gfs2_trim_blocks(struct inode *inode) * earlier versions of GFS2 have a bug in the stuffed file reading * code which will result in a buffer overrun if the size is larger * than the max stuffed file size. In order to prevent this from - * occuring, such files are unstuffed, but in other cases we can + * occurring, such files are unstuffed, but in other cases we can * just update the inode size directly. * * Returns: 0 on success, or -ve on error @@ -1136,31 +1216,30 @@ static int do_grow(struct inode *inode, u64 size) { struct gfs2_inode *ip = GFS2_I(inode); struct gfs2_sbd *sdp = GFS2_SB(inode); + struct gfs2_alloc_parms ap = { .target = 1, }; struct buffer_head *dibh; - struct gfs2_alloc *al = NULL; int error; + int unstuff = 0; if (gfs2_is_stuffed(ip) && (size > (sdp->sd_sb.sb_bsize - sizeof(struct gfs2_dinode)))) { - al = gfs2_alloc_get(ip); - if (al == NULL) - return -ENOMEM; - error = gfs2_quota_lock_check(ip); if (error) - goto do_grow_alloc_put; + return error; - al->al_requested = 1; - error = gfs2_inplace_reserve(ip); + error = gfs2_inplace_reserve(ip, &ap); if (error) goto do_grow_qunlock; + unstuff = 1; } - error = gfs2_trans_begin(sdp, RES_DINODE + RES_STATFS + RES_RG_BIT, 0); + error = gfs2_trans_begin(sdp, RES_DINODE + RES_STATFS + RES_RG_BIT + + (sdp->sd_args.ar_quota == GFS2_QUOTA_OFF ? + 0 : RES_QUOTA), 0); if (error) goto do_grow_release; - if (al) { + if (unstuff) { error = gfs2_unstuff_dinode(ip, NULL); if (error) goto do_end_trans; @@ -1172,19 +1251,17 @@ static int do_grow(struct inode *inode, u64 size) i_size_write(inode, size); ip->i_inode.i_mtime = ip->i_inode.i_ctime = CURRENT_TIME; - gfs2_trans_add_bh(ip->i_gl, dibh, 1); + gfs2_trans_add_meta(ip->i_gl, dibh); gfs2_dinode_out(ip, dibh->b_data); brelse(dibh); do_end_trans: gfs2_trans_end(sdp); do_grow_release: - if (al) { + if (unstuff) { gfs2_inplace_release(ip); do_grow_qunlock: gfs2_quota_unlock(ip); -do_grow_alloc_put: - gfs2_alloc_put(ip); } return error; } @@ -1203,6 +1280,7 @@ do_grow_alloc_put: int gfs2_setattr_size(struct inode *inode, u64 newsize) { + struct gfs2_inode *ip = GFS2_I(inode); int ret; u64 oldsize; @@ -1212,11 +1290,27 @@ int gfs2_setattr_size(struct inode *inode, u64 newsize) if (ret) return ret; + ret = get_write_access(inode); + if (ret) + return ret; + + inode_dio_wait(inode); + + ret = gfs2_rs_alloc(ip); + if (ret) + goto out; + oldsize = inode->i_size; - if (newsize >= oldsize) - return do_grow(inode, newsize); + if (newsize >= oldsize) { + ret = do_grow(inode, newsize); + goto out; + } - return do_shrink(inode, oldsize, newsize); + gfs2_rs_deltree(ip->i_res); + ret = do_shrink(inode, oldsize, newsize); +out: + put_write_access(inode); + return ret; } int gfs2_truncatei_resume(struct gfs2_inode *ip) @@ -1234,6 +1328,121 @@ int gfs2_file_dealloc(struct gfs2_inode *ip) } /** + * gfs2_free_journal_extents - Free cached journal bmap info + * @jd: The journal + * + */ + +void gfs2_free_journal_extents(struct gfs2_jdesc *jd) +{ + struct gfs2_journal_extent *jext; + + while(!list_empty(&jd->extent_list)) { + jext = list_entry(jd->extent_list.next, struct gfs2_journal_extent, list); + list_del(&jext->list); + kfree(jext); + } +} + +/** + * gfs2_add_jextent - Add or merge a new extent to extent cache + * @jd: The journal descriptor + * @lblock: The logical block at start of new extent + * @dblock: The physical block at start of new extent + * @blocks: Size of extent in fs blocks + * + * Returns: 0 on success or -ENOMEM + */ + +static int gfs2_add_jextent(struct gfs2_jdesc *jd, u64 lblock, u64 dblock, u64 blocks) +{ + struct gfs2_journal_extent *jext; + + if (!list_empty(&jd->extent_list)) { + jext = list_entry(jd->extent_list.prev, struct gfs2_journal_extent, list); + if ((jext->dblock + jext->blocks) == dblock) { + jext->blocks += blocks; + return 0; + } + } + + jext = kzalloc(sizeof(struct gfs2_journal_extent), GFP_NOFS); + if (jext == NULL) + return -ENOMEM; + jext->dblock = dblock; + jext->lblock = lblock; + jext->blocks = blocks; + list_add_tail(&jext->list, &jd->extent_list); + jd->nr_extents++; + return 0; +} + +/** + * gfs2_map_journal_extents - Cache journal bmap info + * @sdp: The super block + * @jd: The journal to map + * + * Create a reusable "extent" mapping from all logical + * blocks to all physical blocks for the given journal. This will save + * us time when writing journal blocks. Most journals will have only one + * extent that maps all their logical blocks. That's because gfs2.mkfs + * arranges the journal blocks sequentially to maximize performance. + * So the extent would map the first block for the entire file length. + * However, gfs2_jadd can happen while file activity is happening, so + * those journals may not be sequential. Less likely is the case where + * the users created their own journals by mounting the metafs and + * laying it out. But it's still possible. These journals might have + * several extents. + * + * Returns: 0 on success, or error on failure + */ + +int gfs2_map_journal_extents(struct gfs2_sbd *sdp, struct gfs2_jdesc *jd) +{ + u64 lblock = 0; + u64 lblock_stop; + struct gfs2_inode *ip = GFS2_I(jd->jd_inode); + struct buffer_head bh; + unsigned int shift = sdp->sd_sb.sb_bsize_shift; + u64 size; + int rc; + + lblock_stop = i_size_read(jd->jd_inode) >> shift; + size = (lblock_stop - lblock) << shift; + jd->nr_extents = 0; + WARN_ON(!list_empty(&jd->extent_list)); + + do { + bh.b_state = 0; + bh.b_blocknr = 0; + bh.b_size = size; + rc = gfs2_block_map(jd->jd_inode, lblock, &bh, 0); + if (rc || !buffer_mapped(&bh)) + goto fail; + rc = gfs2_add_jextent(jd, lblock, bh.b_blocknr, bh.b_size >> shift); + if (rc) + goto fail; + size -= bh.b_size; + lblock += (bh.b_size >> ip->i_inode.i_blkbits); + } while(size > 0); + + fs_info(sdp, "journal %d mapped with %u extents\n", jd->jd_jid, + jd->nr_extents); + return 0; + +fail: + fs_warn(sdp, "error %d mapping journal %u at offset %llu (extent %u)\n", + rc, jd->jd_jid, + (unsigned long long)(i_size_read(jd->jd_inode) - size), + jd->nr_extents); + fs_warn(sdp, "bmap=%d lblock=%llu block=%llu, state=0x%08lx, size=%llu\n", + rc, (unsigned long long)lblock, (unsigned long long)bh.b_blocknr, + bh.b_state, (unsigned long long)bh.b_size); + gfs2_free_journal_extents(jd); + return rc; +} + +/** * gfs2_write_alloc_required - figure out if a write will require an allocation * @ip: the file being written to * @offset: the offset to write to diff --git a/fs/gfs2/bmap.h b/fs/gfs2/bmap.h index 42fea03e2bd..81ded5e2aaa 100644 --- a/fs/gfs2/bmap.h +++ b/fs/gfs2/bmap.h @@ -55,5 +55,7 @@ extern int gfs2_truncatei_resume(struct gfs2_inode *ip); extern int gfs2_file_dealloc(struct gfs2_inode *ip); extern int gfs2_write_alloc_required(struct gfs2_inode *ip, u64 offset, unsigned int len); +extern int gfs2_map_journal_extents(struct gfs2_sbd *sdp, struct gfs2_jdesc *jd); +extern void gfs2_free_journal_extents(struct gfs2_jdesc *jd); #endif /* __BMAP_DOT_H__ */ diff --git a/fs/gfs2/dentry.c b/fs/gfs2/dentry.c index 4a456338b87..d3a5d4e29ba 100644 --- a/fs/gfs2/dentry.c +++ b/fs/gfs2/dentry.c @@ -25,7 +25,7 @@ /** * gfs2_drevalidate - Check directory lookup consistency * @dentry: the mapping to check - * @nd: + * @flags: lookup flags * * Check to make sure the lookup necessary to arrive at this inode from its * parent is still good. @@ -33,7 +33,7 @@ * Returns: 1 if the dentry is ok, 0 if it isn't */ -static int gfs2_drevalidate(struct dentry *dentry, struct nameidata *nd) +static int gfs2_drevalidate(struct dentry *dentry, unsigned int flags) { struct dentry *parent; struct gfs2_sbd *sdp; @@ -44,7 +44,7 @@ static int gfs2_drevalidate(struct dentry *dentry, struct nameidata *nd) int error; int had_lock = 0; - if (nd->flags & LOOKUP_RCU) + if (flags & LOOKUP_RCU) return -ECHILD; parent = dget_parent(dentry); @@ -93,12 +93,9 @@ invalid_gunlock: if (!had_lock) gfs2_glock_dq_uninit(&d_gh); invalid: - if (inode && S_ISDIR(inode->i_mode)) { - if (have_submounts(dentry)) - goto valid; - shrink_dcache_parent(dentry); - } - d_drop(dentry); + if (check_submounts_and_drop(dentry) != 0) + goto valid; + dput(parent); return 0; @@ -109,8 +106,7 @@ fail: return 0; } -static int gfs2_dhash(const struct dentry *dentry, const struct inode *inode, - struct qstr *str) +static int gfs2_dhash(const struct dentry *dentry, struct qstr *str) { str->hash = gfs2_disk_hash(str->name, str->len); return 0; diff --git a/fs/gfs2/dir.c b/fs/gfs2/dir.c index 5c356d09c32..1a349f9a968 100644 --- a/fs/gfs2/dir.c +++ b/fs/gfs2/dir.c @@ -53,6 +53,8 @@ * but never before the maximum hash table size has been reached. */ +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + #include <linux/slab.h> #include <linux/spinlock.h> #include <linux/buffer_head.h> @@ -76,25 +78,24 @@ #define IS_LEAF 1 /* Hashed (leaf) directory */ #define IS_DINODE 2 /* Linear (stuffed dinode block) directory */ +#define MAX_RA_BLOCKS 32 /* max read-ahead blocks */ + #define gfs2_disk_hash2offset(h) (((u64)(h)) >> 1) #define gfs2_dir_offset2hash(p) ((u32)(((u64)(p)) << 1)) struct qstr gfs2_qdot __read_mostly; struct qstr gfs2_qdotdot __read_mostly; -typedef int (*leaf_call_t) (struct gfs2_inode *dip, u32 index, u32 len, - u64 leaf_no, void *data); typedef int (*gfs2_dscan_t)(const struct gfs2_dirent *dent, const struct qstr *name, void *opaque); - int gfs2_dir_get_new_buffer(struct gfs2_inode *ip, u64 block, struct buffer_head **bhp) { struct buffer_head *bh; bh = gfs2_meta_new(ip->i_gl, block); - gfs2_trans_add_bh(ip->i_gl, bh, 1); + gfs2_trans_add_meta(ip->i_gl, bh); gfs2_metatype_set(bh, GFS2_METATYPE_JD, GFS2_FORMAT_JD); gfs2_buffer_clear_tail(bh, sizeof(struct gfs2_meta_header)); *bhp = bh; @@ -128,7 +129,7 @@ static int gfs2_dir_write_stuffed(struct gfs2_inode *ip, const char *buf, if (error) return error; - gfs2_trans_add_bh(ip->i_gl, dibh, 1); + gfs2_trans_add_meta(ip->i_gl, dibh); memcpy(dibh->b_data + offset + sizeof(struct gfs2_dinode), buf, size); if (ip->i_inode.i_size < offset + size) i_size_write(&ip->i_inode, offset + size); @@ -210,7 +211,7 @@ static int gfs2_dir_write_data(struct gfs2_inode *ip, const char *buf, if (error) goto fail; - gfs2_trans_add_bh(ip->i_gl, bh, 1); + gfs2_trans_add_meta(ip->i_gl, bh); memcpy(bh->b_data + o, buf, amount); brelse(bh); @@ -232,7 +233,7 @@ out: i_size_write(&ip->i_inode, offset + copied); ip->i_inode.i_mtime = ip->i_inode.i_ctime = CURRENT_TIME; - gfs2_trans_add_bh(ip->i_gl, dibh, 1); + gfs2_trans_add_meta(ip->i_gl, dibh); gfs2_dinode_out(ip, dibh->b_data); brelse(dibh); @@ -243,16 +244,15 @@ fail: return error; } -static int gfs2_dir_read_stuffed(struct gfs2_inode *ip, char *buf, - u64 offset, unsigned int size) +static int gfs2_dir_read_stuffed(struct gfs2_inode *ip, __be64 *buf, + unsigned int size) { struct buffer_head *dibh; int error; error = gfs2_meta_inode_buffer(ip, &dibh); if (!error) { - offset += sizeof(struct gfs2_dinode); - memcpy(buf, dibh->b_data + offset, size); + memcpy(buf, dibh->b_data + sizeof(struct gfs2_dinode), size); brelse(dibh); } @@ -264,13 +264,12 @@ static int gfs2_dir_read_stuffed(struct gfs2_inode *ip, char *buf, * gfs2_dir_read_data - Read a data from a directory inode * @ip: The GFS2 Inode * @buf: The buffer to place result into - * @offset: File offset to begin jdata_readng from * @size: Amount of data to transfer * * Returns: The amount of data actually copied or the error */ -static int gfs2_dir_read_data(struct gfs2_inode *ip, char *buf, u64 offset, - unsigned int size, unsigned ra) +static int gfs2_dir_read_data(struct gfs2_inode *ip, __be64 *buf, + unsigned int size) { struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); u64 lblock, dblock; @@ -278,24 +277,14 @@ static int gfs2_dir_read_data(struct gfs2_inode *ip, char *buf, u64 offset, unsigned int o; int copied = 0; int error = 0; - u64 disksize = i_size_read(&ip->i_inode); - - if (offset >= disksize) - return 0; - - if (offset + size > disksize) - size = disksize - offset; - - if (!size) - return 0; if (gfs2_is_stuffed(ip)) - return gfs2_dir_read_stuffed(ip, buf, offset, size); + return gfs2_dir_read_stuffed(ip, buf, size); if (gfs2_assert_warn(sdp, gfs2_is_jdata(ip))) return -EINVAL; - lblock = offset; + lblock = 0; o = do_div(lblock, sdp->sd_jbsize) + sizeof(struct gfs2_meta_header); while (copied < size) { @@ -314,8 +303,6 @@ static int gfs2_dir_read_data(struct gfs2_inode *ip, char *buf, u64 offset, if (error || !dblock) goto fail; BUG_ON(extlen < 1); - if (!ra) - extlen = 1; bh = gfs2_meta_ra(ip->i_gl, dblock, extlen); } else { error = gfs2_meta_read(ip->i_gl, dblock, DIO_WAIT, &bh); @@ -331,7 +318,7 @@ static int gfs2_dir_read_data(struct gfs2_inode *ip, char *buf, u64 offset, extlen--; memcpy(buf, bh->b_data + o, amount); brelse(bh); - buf += amount; + buf += (amount/sizeof(__be64)); copied += amount; lblock++; o = sizeof(struct gfs2_meta_header); @@ -342,6 +329,79 @@ fail: return (copied) ? copied : error; } +/** + * gfs2_dir_get_hash_table - Get pointer to the dir hash table + * @ip: The inode in question + * + * Returns: The hash table or an error + */ + +static __be64 *gfs2_dir_get_hash_table(struct gfs2_inode *ip) +{ + struct inode *inode = &ip->i_inode; + int ret; + u32 hsize; + __be64 *hc; + + BUG_ON(!(ip->i_diskflags & GFS2_DIF_EXHASH)); + + hc = ip->i_hash_cache; + if (hc) + return hc; + + hsize = 1 << ip->i_depth; + hsize *= sizeof(__be64); + if (hsize != i_size_read(&ip->i_inode)) { + gfs2_consist_inode(ip); + return ERR_PTR(-EIO); + } + + hc = kmalloc(hsize, GFP_NOFS | __GFP_NOWARN); + if (hc == NULL) + hc = __vmalloc(hsize, GFP_NOFS, PAGE_KERNEL); + + if (hc == NULL) + return ERR_PTR(-ENOMEM); + + ret = gfs2_dir_read_data(ip, hc, hsize); + if (ret < 0) { + if (is_vmalloc_addr(hc)) + vfree(hc); + else + kfree(hc); + return ERR_PTR(ret); + } + + spin_lock(&inode->i_lock); + if (ip->i_hash_cache) { + if (is_vmalloc_addr(hc)) + vfree(hc); + else + kfree(hc); + } else { + ip->i_hash_cache = hc; + } + spin_unlock(&inode->i_lock); + + return ip->i_hash_cache; +} + +/** + * gfs2_dir_hash_inval - Invalidate dir hash + * @ip: The directory inode + * + * Must be called with an exclusive glock, or during glock invalidation. + */ +void gfs2_dir_hash_inval(struct gfs2_inode *ip) +{ + __be64 *hc = ip->i_hash_cache; + ip->i_hash_cache = NULL; + if (is_vmalloc_addr(hc)) + vfree(hc); + else + kfree(hc); +} + static inline int gfs2_dirent_sentinel(const struct gfs2_dirent *dent) { return dent->de_inum.no_addr == 0 || dent->de_inum.no_formal_ino == 0; @@ -449,8 +509,8 @@ static int gfs2_check_dirent(struct gfs2_dirent *dent, unsigned int offset, goto error; return 0; error: - printk(KERN_WARNING "gfs2_check_dirent: %s (%s)\n", msg, - first ? "first in block" : "not first in block"); + pr_warn("%s: %s (%s)\n", + __func__, msg, first ? "first in block" : "not first in block"); return -EIO; } @@ -473,8 +533,7 @@ static int gfs2_dirent_offset(const void *buf) } return offset; wrong_type: - printk(KERN_WARNING "gfs2_scan_dirent: wrong block type %u\n", - be32_to_cpu(h->mh_type)); + pr_warn("%s: wrong block type %u\n", __func__, be32_to_cpu(h->mh_type)); return -1; } @@ -601,7 +660,7 @@ static void dirent_del(struct gfs2_inode *dip, struct buffer_head *bh, return; } - gfs2_trans_add_bh(dip->i_gl, bh, 1); + gfs2_trans_add_meta(dip->i_gl, bh); /* If there is no prev entry, this is the first entry in the block. The de_rec_len is already as big as it needs to be. Just zero @@ -644,7 +703,7 @@ static struct gfs2_dirent *gfs2_init_dirent(struct inode *inode, offset = GFS2_DIRENT_SIZE(be16_to_cpu(dent->de_name_len)); totlen = be16_to_cpu(dent->de_rec_len); BUG_ON(offset + name->len > totlen); - gfs2_trans_add_bh(ip->i_gl, bh, 1); + gfs2_trans_add_meta(ip->i_gl, bh); ndent = (struct gfs2_dirent *)((char *)dent + offset); dent->de_rec_len = cpu_to_be16(offset); gfs2_qstr2dirent(name, totlen - offset, ndent); @@ -670,7 +729,7 @@ static int get_leaf(struct gfs2_inode *dip, u64 leaf_no, error = gfs2_meta_read(dip->i_gl, leaf_no, DIO_WAIT, bhp); if (!error && gfs2_metatype_check(GFS2_SB(&dip->i_inode), *bhp, GFS2_METATYPE_LF)) { - /* printk(KERN_INFO "block num=%llu\n", leaf_no); */ + /* pr_info("block num=%llu\n", leaf_no); */ error = -EIO; } @@ -689,17 +748,12 @@ static int get_leaf(struct gfs2_inode *dip, u64 leaf_no, static int get_leaf_nr(struct gfs2_inode *dip, u32 index, u64 *leaf_out) { - __be64 leaf_no; - int error; - - error = gfs2_dir_read_data(dip, (char *)&leaf_no, - index * sizeof(__be64), - sizeof(__be64), 0); - if (error != sizeof(u64)) - return (error < 0) ? error : -EIO; - - *leaf_out = be64_to_cpu(leaf_no); + __be64 *hash; + hash = gfs2_dir_get_hash_table(dip); + if (IS_ERR(hash)) + return PTR_ERR(hash); + *leaf_out = be64_to_cpu(*(hash + index)); return 0; } @@ -780,9 +834,10 @@ static struct gfs2_leaf *new_leaf(struct inode *inode, struct buffer_head **pbh, struct buffer_head *bh; struct gfs2_leaf *leaf; struct gfs2_dirent *dent; - struct qstr name = { .name = "", .len = 0, .hash = 0 }; + struct qstr name = { .name = "" }; + struct timespec tv = CURRENT_TIME; - error = gfs2_alloc_block(ip, &bn, &n); + error = gfs2_alloc_blocks(ip, &bn, &n, 0, NULL); if (error) return NULL; bh = gfs2_meta_new(ip->i_gl, bn); @@ -790,14 +845,18 @@ static struct gfs2_leaf *new_leaf(struct inode *inode, struct buffer_head **pbh, return NULL; gfs2_trans_add_unrevoke(GFS2_SB(inode), bn, 1); - gfs2_trans_add_bh(ip->i_gl, bh, 1); + gfs2_trans_add_meta(ip->i_gl, bh); gfs2_metatype_set(bh, GFS2_METATYPE_LF, GFS2_FORMAT_LF); leaf = (struct gfs2_leaf *)bh->b_data; leaf->lf_depth = cpu_to_be16(depth); leaf->lf_entries = 0; leaf->lf_dirent_format = cpu_to_be32(GFS2_FORMAT_DE); leaf->lf_next = 0; - memset(leaf->lf_reserved, 0, sizeof(leaf->lf_reserved)); + leaf->lf_inode = cpu_to_be64(ip->i_no_addr); + leaf->lf_dist = cpu_to_be32(1); + leaf->lf_nsec = cpu_to_be32(tv.tv_nsec); + leaf->lf_sec = cpu_to_be64(tv.tv_sec); + memset(leaf->lf_reserved2, 0, sizeof(leaf->lf_reserved2)); dent = (struct gfs2_dirent *)(leaf+1); gfs2_qstr2dirent(&name, bh->b_size - sizeof(struct gfs2_leaf), dent); *pbh = bh; @@ -875,7 +934,7 @@ static int dir_make_exhash(struct inode *inode) /* We're done with the new leaf block, now setup the new hash table. */ - gfs2_trans_add_bh(dip->i_gl, dibh, 1); + gfs2_trans_add_meta(dip->i_gl, dibh); gfs2_buffer_clear_tail(dibh, sizeof(struct gfs2_dinode)); lp = (__be64 *)(dibh->b_data + sizeof(struct gfs2_dinode)); @@ -935,7 +994,7 @@ static int dir_split_leaf(struct inode *inode, const struct qstr *name) return 1; /* can't split */ } - gfs2_trans_add_bh(dip->i_gl, obh, 1); + gfs2_trans_add_meta(dip->i_gl, obh); nleaf = new_leaf(inode, &nbh, be16_to_cpu(oleaf->lf_depth) + 1); if (!nleaf) { @@ -948,7 +1007,8 @@ static int dir_split_leaf(struct inode *inode, const struct qstr *name) len = 1 << (dip->i_depth - be16_to_cpu(oleaf->lf_depth)); half_len = len >> 1; if (!half_len) { - printk(KERN_WARNING "i_depth %u lf_depth %u index %u\n", dip->i_depth, be16_to_cpu(oleaf->lf_depth), index); + pr_warn("i_depth %u lf_depth %u index %u\n", + dip->i_depth, be16_to_cpu(oleaf->lf_depth), index); gfs2_consist_inode(dip); error = -EIO; goto fail_brelse; @@ -969,6 +1029,8 @@ static int dir_split_leaf(struct inode *inode, const struct qstr *name) for (x = 0; x < half_len; x++) lp[x] = cpu_to_be64(bn); + gfs2_dir_hash_inval(dip); + error = gfs2_dir_write_data(dip, (char *)lp, start * sizeof(u64), half_len * sizeof(u64)); if (error != half_len * sizeof(u64)) { @@ -1026,7 +1088,7 @@ static int dir_split_leaf(struct inode *inode, const struct qstr *name) error = gfs2_meta_inode_buffer(dip, &dibh); if (!gfs2_assert_withdraw(GFS2_SB(&dip->i_inode), !error)) { - gfs2_trans_add_bh(dip->i_gl, dibh, 1); + gfs2_trans_add_meta(dip->i_gl, dibh); gfs2_add_inode_blocks(&dip->i_inode, 1); gfs2_dinode_out(dip, dibh->b_data); brelse(dibh); @@ -1055,70 +1117,61 @@ fail_brelse: static int dir_double_exhash(struct gfs2_inode *dip) { - struct gfs2_sbd *sdp = GFS2_SB(&dip->i_inode); struct buffer_head *dibh; u32 hsize; - u64 *buf; - u64 *from, *to; - u64 block; - u64 disksize = i_size_read(&dip->i_inode); + u32 hsize_bytes; + __be64 *hc; + __be64 *hc2, *h; int x; int error = 0; hsize = 1 << dip->i_depth; - if (hsize * sizeof(u64) != disksize) { - gfs2_consist_inode(dip); - return -EIO; - } - - /* Allocate both the "from" and "to" buffers in one big chunk */ + hsize_bytes = hsize * sizeof(__be64); - buf = kcalloc(3, sdp->sd_hash_bsize, GFP_NOFS); - if (!buf) - return -ENOMEM; + hc = gfs2_dir_get_hash_table(dip); + if (IS_ERR(hc)) + return PTR_ERR(hc); - for (block = disksize >> sdp->sd_hash_bsize_shift; block--;) { - error = gfs2_dir_read_data(dip, (char *)buf, - block * sdp->sd_hash_bsize, - sdp->sd_hash_bsize, 1); - if (error != sdp->sd_hash_bsize) { - if (error >= 0) - error = -EIO; - goto fail; - } + hc2 = kmalloc(hsize_bytes * 2, GFP_NOFS | __GFP_NOWARN); + if (hc2 == NULL) + hc2 = __vmalloc(hsize_bytes * 2, GFP_NOFS, PAGE_KERNEL); - from = buf; - to = (u64 *)((char *)buf + sdp->sd_hash_bsize); + if (!hc2) + return -ENOMEM; - for (x = sdp->sd_hash_ptrs; x--; from++) { - *to++ = *from; /* No endianess worries */ - *to++ = *from; - } + h = hc2; + error = gfs2_meta_inode_buffer(dip, &dibh); + if (error) + goto out_kfree; - error = gfs2_dir_write_data(dip, - (char *)buf + sdp->sd_hash_bsize, - block * sdp->sd_sb.sb_bsize, - sdp->sd_sb.sb_bsize); - if (error != sdp->sd_sb.sb_bsize) { - if (error >= 0) - error = -EIO; - goto fail; - } + for (x = 0; x < hsize; x++) { + *h++ = *hc; + *h++ = *hc; + hc++; } - kfree(buf); - - error = gfs2_meta_inode_buffer(dip, &dibh); - if (!gfs2_assert_withdraw(sdp, !error)) { - dip->i_depth++; - gfs2_dinode_out(dip, dibh->b_data); - brelse(dibh); - } + error = gfs2_dir_write_data(dip, (char *)hc2, 0, hsize_bytes * 2); + if (error != (hsize_bytes * 2)) + goto fail; - return error; + gfs2_dir_hash_inval(dip); + dip->i_hash_cache = hc2; + dip->i_depth++; + gfs2_dinode_out(dip, dibh->b_data); + brelse(dibh); + return 0; fail: - kfree(buf); + /* Replace original hash table & size */ + gfs2_dir_write_data(dip, (char *)hc, 0, hsize_bytes); + i_size_write(&dip->i_inode, hsize_bytes); + gfs2_dinode_out(dip, dibh->b_data); + brelse(dibh); +out_kfree: + if (is_vmalloc_addr(hc2)) + vfree(hc2); + else + kfree(hc2); return error; } @@ -1167,9 +1220,7 @@ static int compare_dents(const void *a, const void *b) /** * do_filldir_main - read out directory entries * @dip: The GFS2 inode - * @offset: The offset in the file to read from - * @opaque: opaque data to pass to filldir - * @filldir: The function to pass entries to + * @ctx: what to feed the entries to * @darr: an array of struct gfs2_dirent pointers to read * @entries: the number of entries in darr * @copied: pointer to int that's non-zero if a entry has been copied out @@ -1179,11 +1230,10 @@ static int compare_dents(const void *a, const void *b) * the possibility that they will fall into different readdir buffers or * that someone will want to seek to that location. * - * Returns: errno, >0 on exception from filldir + * Returns: errno, >0 if the actor tells you to stop */ -static int do_filldir_main(struct gfs2_inode *dip, u64 *offset, - void *opaque, filldir_t filldir, +static int do_filldir_main(struct gfs2_inode *dip, struct dir_context *ctx, const struct gfs2_dirent **darr, u32 entries, int *copied) { @@ -1191,7 +1241,6 @@ static int do_filldir_main(struct gfs2_inode *dip, u64 *offset, u64 off, off_next; unsigned int x, y; int run = 0; - int error = 0; sort(darr, entries, sizeof(struct gfs2_dirent *), compare_dents, NULL); @@ -1208,9 +1257,9 @@ static int do_filldir_main(struct gfs2_inode *dip, u64 *offset, off_next = be32_to_cpu(dent_next->de_hash); off_next = gfs2_disk_hash2offset(off_next); - if (off < *offset) + if (off < ctx->pos) continue; - *offset = off; + ctx->pos = off; if (off_next == off) { if (*copied && !run) @@ -1219,26 +1268,25 @@ static int do_filldir_main(struct gfs2_inode *dip, u64 *offset, } else run = 0; } else { - if (off < *offset) + if (off < ctx->pos) continue; - *offset = off; + ctx->pos = off; } - error = filldir(opaque, (const char *)(dent + 1), + if (!dir_emit(ctx, (const char *)(dent + 1), be16_to_cpu(dent->de_name_len), - off, be64_to_cpu(dent->de_inum.no_addr), - be16_to_cpu(dent->de_type)); - if (error) + be64_to_cpu(dent->de_inum.no_addr), + be16_to_cpu(dent->de_type))) return 1; *copied = 1; } - /* Increment the *offset by one, so the next time we come into the + /* Increment the ctx->pos by one, so the next time we come into the do_filldir fxn, we get the next entry instead of the last one in the current leaf */ - (*offset)++; + ctx->pos++; return 0; } @@ -1262,8 +1310,8 @@ static void gfs2_free_sort_buffer(void *ptr) kfree(ptr); } -static int gfs2_dir_read_leaf(struct inode *inode, u64 *offset, void *opaque, - filldir_t filldir, int *copied, unsigned *depth, +static int gfs2_dir_read_leaf(struct inode *inode, struct dir_context *ctx, + int *copied, unsigned *depth, u64 leaf_no) { struct gfs2_inode *ip = GFS2_I(inode); @@ -1341,8 +1389,7 @@ static int gfs2_dir_read_leaf(struct inode *inode, u64 *offset, void *opaque, } while(lfn); BUG_ON(entries2 != entries); - error = do_filldir_main(ip, offset, opaque, filldir, darr, - entries, copied); + error = do_filldir_main(ip, ctx, darr, entries, copied); out_free: for(i = 0; i < leaf; i++) brelse(larr[i]); @@ -1352,22 +1399,65 @@ out: } /** + * gfs2_dir_readahead - Issue read-ahead requests for leaf blocks. + * + * Note: we can't calculate each index like dir_e_read can because we don't + * have the leaf, and therefore we don't have the depth, and therefore we + * don't have the length. So we have to just read enough ahead to make up + * for the loss of information. + */ +static void gfs2_dir_readahead(struct inode *inode, unsigned hsize, u32 index, + struct file_ra_state *f_ra) +{ + struct gfs2_inode *ip = GFS2_I(inode); + struct gfs2_glock *gl = ip->i_gl; + struct buffer_head *bh; + u64 blocknr = 0, last; + unsigned count; + + /* First check if we've already read-ahead for the whole range. */ + if (index + MAX_RA_BLOCKS < f_ra->start) + return; + + f_ra->start = max((pgoff_t)index, f_ra->start); + for (count = 0; count < MAX_RA_BLOCKS; count++) { + if (f_ra->start >= hsize) /* if exceeded the hash table */ + break; + + last = blocknr; + blocknr = be64_to_cpu(ip->i_hash_cache[f_ra->start]); + f_ra->start++; + if (blocknr == last) + continue; + + bh = gfs2_getbuf(gl, blocknr, 1); + if (trylock_buffer(bh)) { + if (buffer_uptodate(bh)) { + unlock_buffer(bh); + brelse(bh); + continue; + } + bh->b_end_io = end_buffer_read_sync; + submit_bh(READA | REQ_META, bh); + continue; + } + brelse(bh); + } +} + +/** * dir_e_read - Reads the entries from a directory into a filldir buffer * @dip: dinode pointer - * @offset: the hash of the last entry read shifted to the right once - * @opaque: buffer for the filldir function to fill - * @filldir: points to the filldir function to use + * @ctx: actor to feed the entries to * * Returns: errno */ -static int dir_e_read(struct inode *inode, u64 *offset, void *opaque, - filldir_t filldir) +static int dir_e_read(struct inode *inode, struct dir_context *ctx, + struct file_ra_state *f_ra) { struct gfs2_inode *dip = GFS2_I(inode); - struct gfs2_sbd *sdp = GFS2_SB(inode); u32 hsize, len = 0; - u32 ht_offset, lp_offset, ht_offset_cur = -1; u32 hash, index; __be64 *lp; int copied = 0; @@ -1375,37 +1465,21 @@ static int dir_e_read(struct inode *inode, u64 *offset, void *opaque, unsigned depth = 0; hsize = 1 << dip->i_depth; - if (hsize * sizeof(u64) != i_size_read(inode)) { - gfs2_consist_inode(dip); - return -EIO; - } - - hash = gfs2_dir_offset2hash(*offset); + hash = gfs2_dir_offset2hash(ctx->pos); index = hash >> (32 - dip->i_depth); - lp = kmalloc(sdp->sd_hash_bsize, GFP_NOFS); - if (!lp) - return -ENOMEM; + if (dip->i_hash_cache == NULL) + f_ra->start = 0; + lp = gfs2_dir_get_hash_table(dip); + if (IS_ERR(lp)) + return PTR_ERR(lp); - while (index < hsize) { - lp_offset = index & (sdp->sd_hash_ptrs - 1); - ht_offset = index - lp_offset; - - if (ht_offset_cur != ht_offset) { - error = gfs2_dir_read_data(dip, (char *)lp, - ht_offset * sizeof(__be64), - sdp->sd_hash_bsize, 1); - if (error != sdp->sd_hash_bsize) { - if (error >= 0) - error = -EIO; - goto out; - } - ht_offset_cur = ht_offset; - } + gfs2_dir_readahead(inode, hsize, index, f_ra); - error = gfs2_dir_read_leaf(inode, offset, opaque, filldir, + while (index < hsize) { + error = gfs2_dir_read_leaf(inode, ctx, &copied, &depth, - be64_to_cpu(lp[lp_offset])); + be64_to_cpu(lp[index])); if (error) break; @@ -1413,15 +1487,13 @@ static int dir_e_read(struct inode *inode, u64 *offset, void *opaque, index = (index & ~(len - 1)) + len; } -out: - kfree(lp); if (error > 0) error = 0; return error; } -int gfs2_dir_read(struct inode *inode, u64 *offset, void *opaque, - filldir_t filldir) +int gfs2_dir_read(struct inode *inode, struct dir_context *ctx, + struct file_ra_state *f_ra) { struct gfs2_inode *dip = GFS2_I(inode); struct gfs2_sbd *sdp = GFS2_SB(inode); @@ -1435,7 +1507,7 @@ int gfs2_dir_read(struct inode *inode, u64 *offset, void *opaque, return 0; if (dip->i_diskflags & GFS2_DIF_EXHASH) - return dir_e_read(inode, offset, opaque, filldir); + return dir_e_read(inode, ctx, f_ra); if (!gfs2_is_stuffed(dip)) { gfs2_consist_inode(dip); @@ -1467,7 +1539,7 @@ int gfs2_dir_read(struct inode *inode, u64 *offset, void *opaque, error = -EIO; goto out; } - error = do_filldir_main(dip, offset, opaque, filldir, darr, + error = do_filldir_main(dip, ctx, darr, dip->i_entries, &copied); out: kfree(darr); @@ -1483,9 +1555,9 @@ out: /** * gfs2_dir_search - Search a directory - * @dip: The GFS2 inode - * @filename: - * @inode: + * @dip: The GFS2 dir inode + * @name: The name we are looking up + * @fail_on_exist: Fail if the name exists rather than looking it up * * This routine searches a directory for a file or another directory. * Assumes a glock is held on dip. @@ -1493,22 +1565,25 @@ out: * Returns: errno */ -struct inode *gfs2_dir_search(struct inode *dir, const struct qstr *name) +struct inode *gfs2_dir_search(struct inode *dir, const struct qstr *name, + bool fail_on_exist) { struct buffer_head *bh; struct gfs2_dirent *dent; - struct inode *inode; + u64 addr, formal_ino; + u16 dtype; dent = gfs2_dirent_search(dir, name, gfs2_dirent_find, &bh); if (dent) { if (IS_ERR(dent)) return ERR_CAST(dent); - inode = gfs2_inode_lookup(dir->i_sb, - be16_to_cpu(dent->de_type), - be64_to_cpu(dent->de_inum.no_addr), - be64_to_cpu(dent->de_inum.no_formal_ino)); + dtype = be16_to_cpu(dent->de_type); + addr = be64_to_cpu(dent->de_inum.no_addr); + formal_ino = be64_to_cpu(dent->de_inum.no_formal_ino); brelse(bh); - return inode; + if (fail_on_exist) + return ERR_PTR(-EEXIST); + return gfs2_inode_lookup(dir->i_sb, dtype, addr, formal_ino, 0); } return ERR_PTR(-ENOENT); } @@ -1544,11 +1619,31 @@ out: return ret; } +/** + * dir_new_leaf - Add a new leaf onto hash chain + * @inode: The directory + * @name: The name we are adding + * + * This adds a new dir leaf onto an existing leaf when there is not + * enough space to add a new dir entry. This is a last resort after + * we've expanded the hash table to max size and also split existing + * leaf blocks, so it will only occur for very large directories. + * + * The dist parameter is set to 1 for leaf blocks directly attached + * to the hash table, 2 for one layer of indirection, 3 for two layers + * etc. We are thus able to tell the difference between an old leaf + * with dist set to zero (i.e. "don't know") and a new one where we + * set this information for debug/fsck purposes. + * + * Returns: 0 on success, or -ve on error + */ + static int dir_new_leaf(struct inode *inode, const struct qstr *name) { struct buffer_head *bh, *obh; struct gfs2_inode *ip = GFS2_I(inode); struct gfs2_leaf *leaf, *oleaf; + u32 dist = 1; int error; u32 index; u64 bn; @@ -1558,6 +1653,7 @@ static int dir_new_leaf(struct inode *inode, const struct qstr *name) if (error) return error; do { + dist++; oleaf = (struct gfs2_leaf *)obh->b_data; bn = be64_to_cpu(oleaf->lf_next); if (!bn) @@ -1568,13 +1664,14 @@ static int dir_new_leaf(struct inode *inode, const struct qstr *name) return error; } while(1); - gfs2_trans_add_bh(ip->i_gl, obh, 1); + gfs2_trans_add_meta(ip->i_gl, obh); leaf = new_leaf(inode, &bh, be16_to_cpu(oleaf->lf_depth)); if (!leaf) { brelse(obh); return -ENOSPC; } + leaf->lf_dist = cpu_to_be32(dist); oleaf->lf_next = cpu_to_be64(bh->b_blocknr); brelse(bh); brelse(obh); @@ -1582,54 +1679,74 @@ static int dir_new_leaf(struct inode *inode, const struct qstr *name) error = gfs2_meta_inode_buffer(ip, &bh); if (error) return error; - gfs2_trans_add_bh(ip->i_gl, bh, 1); + gfs2_trans_add_meta(ip->i_gl, bh); gfs2_add_inode_blocks(&ip->i_inode, 1); gfs2_dinode_out(ip, bh->b_data); brelse(bh); return 0; } +static u16 gfs2_inode_ra_len(const struct gfs2_inode *ip) +{ + u64 where = ip->i_no_addr + 1; + if (ip->i_eattr == where) + return 1; + return 0; +} + /** * gfs2_dir_add - Add new filename into directory - * @dip: The GFS2 inode - * @filename: The new name - * @inode: The inode number of the entry - * @type: The type of the entry + * @inode: The directory inode + * @name: The new name + * @nip: The GFS2 inode to be linked in to the directory + * @da: The directory addition info + * + * If the call to gfs2_diradd_alloc_required resulted in there being + * no need to allocate any new directory blocks, then it will contain + * a pointer to the directory entry and the bh in which it resides. We + * can use that without having to repeat the search. If there was no + * free space, then we must now create more space. * * Returns: 0 on success, error code on failure */ int gfs2_dir_add(struct inode *inode, const struct qstr *name, - const struct gfs2_inode *nip, unsigned type) + const struct gfs2_inode *nip, struct gfs2_diradd *da) { struct gfs2_inode *ip = GFS2_I(inode); - struct buffer_head *bh; - struct gfs2_dirent *dent; + struct buffer_head *bh = da->bh; + struct gfs2_dirent *dent = da->dent; + struct timespec tv; struct gfs2_leaf *leaf; int error; while(1) { - dent = gfs2_dirent_search(inode, name, gfs2_dirent_find_space, - &bh); + if (da->bh == NULL) { + dent = gfs2_dirent_search(inode, name, + gfs2_dirent_find_space, &bh); + } if (dent) { if (IS_ERR(dent)) return PTR_ERR(dent); dent = gfs2_init_dirent(inode, dent, name, bh); gfs2_inum_out(nip, dent); - dent->de_type = cpu_to_be16(type); + dent->de_type = cpu_to_be16(IF2DT(nip->i_inode.i_mode)); + dent->de_rahead = cpu_to_be16(gfs2_inode_ra_len(nip)); + tv = CURRENT_TIME; if (ip->i_diskflags & GFS2_DIF_EXHASH) { leaf = (struct gfs2_leaf *)bh->b_data; be16_add_cpu(&leaf->lf_entries, 1); + leaf->lf_nsec = cpu_to_be32(tv.tv_nsec); + leaf->lf_sec = cpu_to_be64(tv.tv_sec); } + da->dent = NULL; + da->bh = NULL; brelse(bh); - error = gfs2_meta_inode_buffer(ip, &bh); - if (error) - break; - gfs2_trans_add_bh(ip->i_gl, bh, 1); ip->i_entries++; - ip->i_inode.i_mtime = ip->i_inode.i_ctime = CURRENT_TIME; - gfs2_dinode_out(ip, bh->b_data); - brelse(bh); + ip->i_inode.i_mtime = ip->i_inode.i_ctime = tv; + if (S_ISDIR(nip->i_inode.i_mode)) + inc_nlink(&ip->i_inode); + mark_inode_dirty(inode); error = 0; break; } @@ -1672,11 +1789,12 @@ int gfs2_dir_add(struct inode *inode, const struct qstr *name, * Returns: 0 on success, error code on failure */ -int gfs2_dir_del(struct gfs2_inode *dip, const struct qstr *name) +int gfs2_dir_del(struct gfs2_inode *dip, const struct dentry *dentry) { + const struct qstr *name = &dentry->d_name; struct gfs2_dirent *dent, *prev = NULL; struct buffer_head *bh; - int error; + struct timespec tv = CURRENT_TIME; /* Returns _either_ the entry (if its first in block) or the previous entry otherwise */ @@ -1702,23 +1820,20 @@ int gfs2_dir_del(struct gfs2_inode *dip, const struct qstr *name) if (!entries) gfs2_consist_inode(dip); leaf->lf_entries = cpu_to_be16(--entries); + leaf->lf_nsec = cpu_to_be32(tv.tv_nsec); + leaf->lf_sec = cpu_to_be64(tv.tv_sec); } brelse(bh); - error = gfs2_meta_inode_buffer(dip, &bh); - if (error) - return error; - if (!dip->i_entries) gfs2_consist_inode(dip); - gfs2_trans_add_bh(dip->i_gl, bh, 1); dip->i_entries--; - dip->i_inode.i_mtime = dip->i_inode.i_ctime = CURRENT_TIME; - gfs2_dinode_out(dip, bh->b_data); - brelse(bh); + dip->i_inode.i_mtime = dip->i_inode.i_ctime = tv; + if (S_ISDIR(dentry->d_inode->i_mode)) + drop_nlink(&dip->i_inode); mark_inode_dirty(&dip->i_inode); - return error; + return 0; } /** @@ -1749,7 +1864,7 @@ int gfs2_dir_mvino(struct gfs2_inode *dip, const struct qstr *filename, if (IS_ERR(dent)) return PTR_ERR(dent); - gfs2_trans_add_bh(dip->i_gl, bh, 1); + gfs2_trans_add_meta(dip->i_gl, bh); gfs2_inum_out(nip, dent); dent->de_type = cpu_to_be16(new_type); @@ -1758,7 +1873,7 @@ int gfs2_dir_mvino(struct gfs2_inode *dip, const struct qstr *filename, error = gfs2_meta_inode_buffer(dip, &bh); if (error) return error; - gfs2_trans_add_bh(dip->i_gl, bh, 1); + gfs2_trans_add_meta(dip->i_gl, bh); } dip->i_inode.i_mtime = dip->i_inode.i_ctime = CURRENT_TIME; @@ -1768,94 +1883,20 @@ int gfs2_dir_mvino(struct gfs2_inode *dip, const struct qstr *filename, } /** - * foreach_leaf - call a function for each leaf in a directory - * @dip: the directory - * @lc: the function to call for each each - * @data: private data to pass to it - * - * Returns: errno - */ - -static int foreach_leaf(struct gfs2_inode *dip, leaf_call_t lc, void *data) -{ - struct gfs2_sbd *sdp = GFS2_SB(&dip->i_inode); - struct buffer_head *bh; - struct gfs2_leaf *leaf; - u32 hsize, len; - u32 ht_offset, lp_offset, ht_offset_cur = -1; - u32 index = 0; - __be64 *lp; - u64 leaf_no; - int error = 0; - - hsize = 1 << dip->i_depth; - if (hsize * sizeof(u64) != i_size_read(&dip->i_inode)) { - gfs2_consist_inode(dip); - return -EIO; - } - - lp = kmalloc(sdp->sd_hash_bsize, GFP_NOFS); - if (!lp) - return -ENOMEM; - - while (index < hsize) { - lp_offset = index & (sdp->sd_hash_ptrs - 1); - ht_offset = index - lp_offset; - - if (ht_offset_cur != ht_offset) { - error = gfs2_dir_read_data(dip, (char *)lp, - ht_offset * sizeof(__be64), - sdp->sd_hash_bsize, 1); - if (error != sdp->sd_hash_bsize) { - if (error >= 0) - error = -EIO; - goto out; - } - ht_offset_cur = ht_offset; - } - - leaf_no = be64_to_cpu(lp[lp_offset]); - if (leaf_no) { - error = get_leaf(dip, leaf_no, &bh); - if (error) - goto out; - leaf = (struct gfs2_leaf *)bh->b_data; - len = 1 << (dip->i_depth - be16_to_cpu(leaf->lf_depth)); - brelse(bh); - - error = lc(dip, index, len, leaf_no, data); - if (error) - goto out; - - index = (index & ~(len - 1)) + len; - } else - index++; - } - - if (index != hsize) { - gfs2_consist_inode(dip); - error = -EIO; - } - -out: - kfree(lp); - - return error; -} - -/** * leaf_dealloc - Deallocate a directory leaf * @dip: the directory * @index: the hash table offset in the directory * @len: the number of pointers to this leaf * @leaf_no: the leaf number - * @data: not used + * @leaf_bh: buffer_head for the starting leaf + * last_dealloc: 1 if this is the final dealloc for the leaf, else 0 * * Returns: errno */ static int leaf_dealloc(struct gfs2_inode *dip, u32 index, u32 len, - u64 leaf_no, void *data) + u64 leaf_no, struct buffer_head *leaf_bh, + int last_dealloc) { struct gfs2_sbd *sdp = GFS2_SB(&dip->i_inode); struct gfs2_leaf *tmp_leaf; @@ -1867,36 +1908,37 @@ static int leaf_dealloc(struct gfs2_inode *dip, u32 index, u32 len, unsigned int x, size = len * sizeof(u64); int error; + error = gfs2_rindex_update(sdp); + if (error) + return error; + memset(&rlist, 0, sizeof(struct gfs2_rgrp_list)); - ht = kzalloc(size, GFP_NOFS); + ht = kzalloc(size, GFP_NOFS | __GFP_NOWARN); + if (ht == NULL) + ht = vzalloc(size); if (!ht) return -ENOMEM; - if (!gfs2_alloc_get(dip)) { - error = -ENOMEM; - goto out; - } - - error = gfs2_quota_hold(dip, NO_QUOTA_CHANGE, NO_QUOTA_CHANGE); - if (error) - goto out_put; - - error = gfs2_rindex_hold(sdp, &dip->i_alloc->al_ri_gh); + error = gfs2_quota_hold(dip, NO_UID_QUOTA_CHANGE, NO_GID_QUOTA_CHANGE); if (error) - goto out_qs; + goto out; /* Count the number of leaves */ + bh = leaf_bh; for (blk = leaf_no; blk; blk = nblk) { - error = get_leaf(dip, blk, &bh); - if (error) - goto out_rlist; + if (blk != leaf_no) { + error = get_leaf(dip, blk, &bh); + if (error) + goto out_rlist; + } tmp_leaf = (struct gfs2_leaf *)bh->b_data; nblk = be64_to_cpu(tmp_leaf->lf_next); - brelse(bh); + if (blk != leaf_no) + brelse(bh); - gfs2_rlist_add(sdp, &rlist, blk); + gfs2_rlist_add(dip, &rlist, blk); l_blocks++; } @@ -1918,13 +1960,18 @@ static int leaf_dealloc(struct gfs2_inode *dip, u32 index, u32 len, if (error) goto out_rg_gunlock; + bh = leaf_bh; + for (blk = leaf_no; blk; blk = nblk) { - error = get_leaf(dip, blk, &bh); - if (error) - goto out_end_trans; + if (blk != leaf_no) { + error = get_leaf(dip, blk, &bh); + if (error) + goto out_end_trans; + } tmp_leaf = (struct gfs2_leaf *)bh->b_data; nblk = be64_to_cpu(tmp_leaf->lf_next); - brelse(bh); + if (blk != leaf_no) + brelse(bh); gfs2_free_meta(dip, blk, 1); gfs2_add_inode_blocks(&dip->i_inode, -1); @@ -1941,7 +1988,11 @@ static int leaf_dealloc(struct gfs2_inode *dip, u32 index, u32 len, if (error) goto out_end_trans; - gfs2_trans_add_bh(dip->i_gl, dibh, 1); + gfs2_trans_add_meta(dip->i_gl, dibh); + /* On the last dealloc, make this a regular file in case we crash. + (We don't want to free these blocks a second time.) */ + if (last_dealloc) + dip->i_inode.i_mode = S_IFREG; gfs2_dinode_out(dip, dibh->b_data); brelse(dibh); @@ -1951,13 +2002,12 @@ out_rg_gunlock: gfs2_glock_dq_m(rlist.rl_rgrps, rlist.rl_ghs); out_rlist: gfs2_rlist_free(&rlist); - gfs2_glock_dq_uninit(&dip->i_alloc->al_ri_gh); -out_qs: gfs2_quota_unhold(dip); -out_put: - gfs2_alloc_put(dip); out: - kfree(ht); + if (is_vmalloc_addr(ht)) + vfree(ht); + else + kfree(ht); return error; } @@ -1973,31 +2023,47 @@ out: int gfs2_dir_exhash_dealloc(struct gfs2_inode *dip) { - struct gfs2_sbd *sdp = GFS2_SB(&dip->i_inode); struct buffer_head *bh; - int error; + struct gfs2_leaf *leaf; + u32 hsize, len; + u32 index = 0, next_index; + __be64 *lp; + u64 leaf_no; + int error = 0, last; - /* Dealloc on-disk leaves to FREEMETA state */ - error = foreach_leaf(dip, leaf_dealloc, NULL); - if (error) - return error; + hsize = 1 << dip->i_depth; - /* Make this a regular file in case we crash. - (We don't want to free these blocks a second time.) */ + lp = gfs2_dir_get_hash_table(dip); + if (IS_ERR(lp)) + return PTR_ERR(lp); - error = gfs2_trans_begin(sdp, RES_DINODE, 0); - if (error) - return error; + while (index < hsize) { + leaf_no = be64_to_cpu(lp[index]); + if (leaf_no) { + error = get_leaf(dip, leaf_no, &bh); + if (error) + goto out; + leaf = (struct gfs2_leaf *)bh->b_data; + len = 1 << (dip->i_depth - be16_to_cpu(leaf->lf_depth)); - error = gfs2_meta_inode_buffer(dip, &bh); - if (!error) { - gfs2_trans_add_bh(dip->i_gl, bh, 1); - ((struct gfs2_dinode *)bh->b_data)->di_mode = - cpu_to_be32(S_IFREG); - brelse(bh); + next_index = (index & ~(len - 1)) + len; + last = ((next_index >= hsize) ? 1 : 0); + error = leaf_dealloc(dip, index, len, leaf_no, bh, + last); + brelse(bh); + if (error) + goto out; + index = next_index; + } else + index++; } - gfs2_trans_end(sdp); + if (index != hsize) { + gfs2_consist_inode(dip); + error = -EIO; + } + +out: return error; } @@ -2006,22 +2072,36 @@ int gfs2_dir_exhash_dealloc(struct gfs2_inode *dip) * gfs2_diradd_alloc_required - find if adding entry will require an allocation * @ip: the file being written to * @filname: the filename that's going to be added + * @da: The structure to return dir alloc info * - * Returns: 1 if alloc required, 0 if not, -ve on error + * Returns: 0 if ok, -ve on error */ -int gfs2_diradd_alloc_required(struct inode *inode, const struct qstr *name) +int gfs2_diradd_alloc_required(struct inode *inode, const struct qstr *name, + struct gfs2_diradd *da) { + struct gfs2_inode *ip = GFS2_I(inode); + struct gfs2_sbd *sdp = GFS2_SB(inode); + const unsigned int extra = sizeof(struct gfs2_dinode) - sizeof(struct gfs2_leaf); struct gfs2_dirent *dent; struct buffer_head *bh; + da->nr_blocks = 0; + da->bh = NULL; + da->dent = NULL; + dent = gfs2_dirent_search(inode, name, gfs2_dirent_find_space, &bh); if (!dent) { - return 1; + da->nr_blocks = sdp->sd_max_dirres; + if (!(ip->i_diskflags & GFS2_DIF_EXHASH) && + (GFS2_DIRENT_SIZE(name->len) < extra)) + da->nr_blocks = 1; + return 0; } if (IS_ERR(dent)) return PTR_ERR(dent); - brelse(bh); + da->bh = bh; + da->dent = dent; return 0; } diff --git a/fs/gfs2/dir.h b/fs/gfs2/dir.h index a98f644bd3d..126c65dda02 100644 --- a/fs/gfs2/dir.h +++ b/fs/gfs2/dir.h @@ -16,25 +16,42 @@ struct inode; struct gfs2_inode; struct gfs2_inum; +struct buffer_head; +struct gfs2_dirent; + +struct gfs2_diradd { + unsigned nr_blocks; + struct gfs2_dirent *dent; + struct buffer_head *bh; +}; extern struct inode *gfs2_dir_search(struct inode *dir, - const struct qstr *filename); + const struct qstr *filename, + bool fail_on_exist); extern int gfs2_dir_check(struct inode *dir, const struct qstr *filename, const struct gfs2_inode *ip); extern int gfs2_dir_add(struct inode *inode, const struct qstr *filename, - const struct gfs2_inode *ip, unsigned int type); -extern int gfs2_dir_del(struct gfs2_inode *dip, const struct qstr *filename); -extern int gfs2_dir_read(struct inode *inode, u64 *offset, void *opaque, - filldir_t filldir); + const struct gfs2_inode *ip, struct gfs2_diradd *da); +static inline void gfs2_dir_no_add(struct gfs2_diradd *da) +{ + if (da->bh) + brelse(da->bh); + da->bh = NULL; +} +extern int gfs2_dir_del(struct gfs2_inode *dip, const struct dentry *dentry); +extern int gfs2_dir_read(struct inode *inode, struct dir_context *ctx, + struct file_ra_state *f_ra); extern int gfs2_dir_mvino(struct gfs2_inode *dip, const struct qstr *filename, const struct gfs2_inode *nip, unsigned int new_type); extern int gfs2_dir_exhash_dealloc(struct gfs2_inode *dip); extern int gfs2_diradd_alloc_required(struct inode *dir, - const struct qstr *filename); + const struct qstr *filename, + struct gfs2_diradd *da); extern int gfs2_dir_get_new_buffer(struct gfs2_inode *ip, u64 block, struct buffer_head **bhp); +extern void gfs2_dir_hash_inval(struct gfs2_inode *ip); static inline u32 gfs2_disk_hash(const char *data, int len) { diff --git a/fs/gfs2/export.c b/fs/gfs2/export.c index 9023db8184f..8b9b3775e2e 100644 --- a/fs/gfs2/export.c +++ b/fs/gfs2/export.c @@ -28,17 +28,20 @@ #define GFS2_LARGE_FH_SIZE 8 #define GFS2_OLD_FH_SIZE 10 -static int gfs2_encode_fh(struct dentry *dentry, __u32 *p, int *len, - int connectable) +static int gfs2_encode_fh(struct inode *inode, __u32 *p, int *len, + struct inode *parent) { __be32 *fh = (__force __be32 *)p; - struct inode *inode = dentry->d_inode; struct super_block *sb = inode->i_sb; struct gfs2_inode *ip = GFS2_I(inode); - if (*len < GFS2_SMALL_FH_SIZE || - (connectable && *len < GFS2_LARGE_FH_SIZE)) - return 255; + if (parent && (*len < GFS2_LARGE_FH_SIZE)) { + *len = GFS2_LARGE_FH_SIZE; + return FILEID_INVALID; + } else if (*len < GFS2_SMALL_FH_SIZE) { + *len = GFS2_SMALL_FH_SIZE; + return FILEID_INVALID; + } fh[0] = cpu_to_be32(ip->i_no_formal_ino >> 32); fh[1] = cpu_to_be32(ip->i_no_formal_ino & 0xFFFFFFFF); @@ -46,14 +49,10 @@ static int gfs2_encode_fh(struct dentry *dentry, __u32 *p, int *len, fh[3] = cpu_to_be32(ip->i_no_addr & 0xFFFFFFFF); *len = GFS2_SMALL_FH_SIZE; - if (!connectable || inode == sb->s_root->d_inode) + if (!parent || inode == sb->s_root->d_inode) return *len; - spin_lock(&dentry->d_lock); - inode = dentry->d_parent->d_inode; - ip = GFS2_I(inode); - igrab(inode); - spin_unlock(&dentry->d_lock); + ip = GFS2_I(parent); fh[4] = cpu_to_be32(ip->i_no_formal_ino >> 32); fh[5] = cpu_to_be32(ip->i_no_formal_ino & 0xFFFFFFFF); @@ -61,12 +60,11 @@ static int gfs2_encode_fh(struct dentry *dentry, __u32 *p, int *len, fh[7] = cpu_to_be32(ip->i_no_addr & 0xFFFFFFFF); *len = GFS2_LARGE_FH_SIZE; - iput(inode); - return *len; } struct get_name_filldir { + struct dir_context ctx; struct gfs2_inum_host inum; char *name; }; @@ -91,10 +89,13 @@ static int gfs2_get_name(struct dentry *parent, char *name, struct inode *dir = parent->d_inode; struct inode *inode = child->d_inode; struct gfs2_inode *dip, *ip; - struct get_name_filldir gnfd; + struct get_name_filldir gnfd = { + .ctx.actor = get_name_filldir, + .name = name + }; struct gfs2_holder gh; - u64 offset = 0; int error; + struct file_ra_state f_ra = { .start = 0 }; if (!dir) return -EINVAL; @@ -108,13 +109,12 @@ static int gfs2_get_name(struct dentry *parent, char *name, *name = 0; gnfd.inum.no_addr = ip->i_no_addr; gnfd.inum.no_formal_ino = ip->i_no_formal_ino; - gnfd.name = name; error = gfs2_glock_nq_init(dip->i_gl, LM_ST_SHARED, 0, &gh); if (error) return error; - error = gfs2_dir_read(dir, &offset, &gnfd, get_name_filldir); + error = gfs2_dir_read(dir, &gnfd.ctx, &f_ra); gfs2_glock_dq_uninit(&gh); @@ -135,7 +135,7 @@ static struct dentry *gfs2_get_dentry(struct super_block *sb, struct gfs2_sbd *sdp = sb->s_fs_info; struct inode *inode; - inode = gfs2_ilookup(sb, inum->no_addr); + inode = gfs2_ilookup(sb, inum->no_addr, 0); if (inode) { if (GFS2_I(inode)->i_no_formal_ino != inum->no_formal_ino) { iput(inode); @@ -163,6 +163,8 @@ static struct dentry *gfs2_fh_to_dentry(struct super_block *sb, struct fid *fid, case GFS2_SMALL_FH_SIZE: case GFS2_LARGE_FH_SIZE: case GFS2_OLD_FH_SIZE: + if (fh_len < GFS2_SMALL_FH_SIZE) + return NULL; this.no_formal_ino = ((u64)be32_to_cpu(fh[0])) << 32; this.no_formal_ino |= be32_to_cpu(fh[1]); this.no_addr = ((u64)be32_to_cpu(fh[2])) << 32; @@ -182,6 +184,8 @@ static struct dentry *gfs2_fh_to_parent(struct super_block *sb, struct fid *fid, switch (fh_type) { case GFS2_LARGE_FH_SIZE: case GFS2_OLD_FH_SIZE: + if (fh_len < GFS2_LARGE_FH_SIZE) + return NULL; parent.no_formal_ino = ((u64)be32_to_cpu(fh[4])) << 32; parent.no_formal_ino |= be32_to_cpu(fh[5]); parent.no_addr = ((u64)be32_to_cpu(fh[6])) << 32; diff --git a/fs/gfs2/file.c b/fs/gfs2/file.c index 7cfdcb91336..26b3f952e6b 100644 --- a/fs/gfs2/file.c +++ b/fs/gfs2/file.c @@ -18,7 +18,6 @@ #include <linux/mount.h> #include <linux/fs.h> #include <linux/gfs2_ondisk.h> -#include <linux/ext2_fs.h> #include <linux/falloc.h> #include <linux/swap.h> #include <linux/crc32.h> @@ -26,6 +25,7 @@ #include <asm/uaccess.h> #include <linux/dlm.h> #include <linux/dlm_plock.h> +#include <linux/aio.h> #include "gfs2.h" #include "incore.h" @@ -45,7 +45,7 @@ * gfs2_llseek - seek to a location in a file * @file: the file * @offset: the offset - * @origin: Where to seek from (SEEK_SET, SEEK_CUR, or SEEK_END) + * @whence: Where to seek from (SEEK_SET, SEEK_CUR, or SEEK_END) * * SEEK_END requires the glock for the file because it references the * file's size. @@ -53,55 +53,57 @@ * Returns: The new offset, or errno */ -static loff_t gfs2_llseek(struct file *file, loff_t offset, int origin) +static loff_t gfs2_llseek(struct file *file, loff_t offset, int whence) { struct gfs2_inode *ip = GFS2_I(file->f_mapping->host); struct gfs2_holder i_gh; loff_t error; - if (origin == 2) { + switch (whence) { + case SEEK_END: /* These reference inode->i_size */ + case SEEK_DATA: + case SEEK_HOLE: error = gfs2_glock_nq_init(ip->i_gl, LM_ST_SHARED, LM_FLAG_ANY, &i_gh); if (!error) { - error = generic_file_llseek_unlocked(file, offset, origin); + error = generic_file_llseek(file, offset, whence); gfs2_glock_dq_uninit(&i_gh); } - } else - error = generic_file_llseek_unlocked(file, offset, origin); + break; + case SEEK_CUR: + case SEEK_SET: + error = generic_file_llseek(file, offset, whence); + break; + default: + error = -EINVAL; + } return error; } /** - * gfs2_readdir - Read directory entries from a directory + * gfs2_readdir - Iterator for a directory * @file: The directory to read from - * @dirent: Buffer for dirents - * @filldir: Function used to do the copying + * @ctx: What to feed directory entries to * * Returns: errno */ -static int gfs2_readdir(struct file *file, void *dirent, filldir_t filldir) +static int gfs2_readdir(struct file *file, struct dir_context *ctx) { struct inode *dir = file->f_mapping->host; struct gfs2_inode *dip = GFS2_I(dir); struct gfs2_holder d_gh; - u64 offset = file->f_pos; int error; - gfs2_holder_init(dip->i_gl, LM_ST_SHARED, 0, &d_gh); - error = gfs2_glock_nq(&d_gh); - if (error) { - gfs2_holder_uninit(&d_gh); + error = gfs2_glock_nq_init(dip->i_gl, LM_ST_SHARED, 0, &d_gh); + if (error) return error; - } - error = gfs2_dir_read(dir, &offset, dirent, filldir); + error = gfs2_dir_read(dir, ctx, &file->f_ra); gfs2_glock_dq_uninit(&d_gh); - file->f_pos = offset; - return error; } @@ -134,6 +136,7 @@ static const u32 fsflags_to_gfs2[32] = { [7] = GFS2_DIF_NOATIME, [12] = GFS2_DIF_EXHASH, [14] = GFS2_DIF_INHERIT_JDATA, + [17] = GFS2_DIF_TOPDIR, }; static const u32 gfs2_to_fsflags[32] = { @@ -142,12 +145,13 @@ static const u32 gfs2_to_fsflags[32] = { [gfs2fl_AppendOnly] = FS_APPEND_FL, [gfs2fl_NoAtime] = FS_NOATIME_FL, [gfs2fl_ExHash] = FS_INDEX_FL, + [gfs2fl_TopLevel] = FS_TOPDIR_FL, [gfs2fl_InheritJdata] = FS_JOURNAL_DATA_FL, }; static int gfs2_get_flags(struct file *filp, u32 __user *ptr) { - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); struct gfs2_inode *ip = GFS2_I(inode); struct gfs2_holder gh; int error; @@ -174,7 +178,9 @@ void gfs2_set_inode_flags(struct inode *inode) struct gfs2_inode *ip = GFS2_I(inode); unsigned int flags = inode->i_flags; - flags &= ~(S_SYNC|S_APPEND|S_IMMUTABLE|S_NOATIME|S_DIRSYNC); + flags &= ~(S_SYNC|S_APPEND|S_IMMUTABLE|S_NOATIME|S_DIRSYNC|S_NOSEC); + if ((ip->i_eattr == 0) && !is_sxid(inode->i_mode)) + inode->i_flags |= S_NOSEC; if (ip->i_diskflags & GFS2_DIF_IMMUTABLE) flags |= S_IMMUTABLE; if (ip->i_diskflags & GFS2_DIF_APPENDONLY) @@ -193,18 +199,19 @@ void gfs2_set_inode_flags(struct inode *inode) GFS2_DIF_NOATIME| \ GFS2_DIF_SYNC| \ GFS2_DIF_SYSTEM| \ + GFS2_DIF_TOPDIR| \ GFS2_DIF_INHERIT_JDATA) /** - * gfs2_set_flags - set flags on an inode - * @inode: The inode - * @flags: The flags to set + * do_gfs2_set_flags - set flags on an inode + * @filp: file pointer + * @reqflags: The flags to set * @mask: Indicates which flags are valid * */ static int do_gfs2_set_flags(struct file *filp, u32 reqflags, u32 mask) { - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); struct gfs2_inode *ip = GFS2_I(inode); struct gfs2_sbd *sdp = GFS2_SB(inode); struct buffer_head *bh; @@ -212,7 +219,7 @@ static int do_gfs2_set_flags(struct file *filp, u32 reqflags, u32 mask) int error; u32 new_flags, flags; - error = mnt_want_write(filp->f_path.mnt); + error = mnt_want_write_file(filp); if (error) return error; @@ -221,7 +228,7 @@ static int do_gfs2_set_flags(struct file *filp, u32 reqflags, u32 mask) goto out_drop_write; error = -EACCES; - if (!is_owner_or_cap(inode)) + if (!inode_owner_or_capable(inode)) goto out; error = 0; @@ -243,13 +250,13 @@ static int do_gfs2_set_flags(struct file *filp, u32 reqflags, u32 mask) !capable(CAP_LINUX_IMMUTABLE)) goto out; if (!IS_IMMUTABLE(inode)) { - error = gfs2_permission(inode, MAY_WRITE, 0); + error = gfs2_permission(inode, MAY_WRITE); if (error) goto out; } if ((flags ^ new_flags) & GFS2_DIF_JDATA) { if (flags & GFS2_DIF_JDATA) - gfs2_log_flush(sdp, ip->i_gl); + gfs2_log_flush(sdp, ip->i_gl, NORMAL_FLUSH); error = filemap_fdatawrite(inode->i_mapping); if (error) goto out; @@ -263,7 +270,7 @@ static int do_gfs2_set_flags(struct file *filp, u32 reqflags, u32 mask) error = gfs2_meta_inode_buffer(ip, &bh); if (error) goto out_trans_end; - gfs2_trans_add_bh(ip->i_gl, bh, 1); + gfs2_trans_add_meta(ip->i_gl, bh); ip->i_diskflags = new_flags; gfs2_dinode_out(ip, bh->b_data); brelse(bh); @@ -274,13 +281,13 @@ out_trans_end: out: gfs2_glock_dq_uninit(&gh); out_drop_write: - mnt_drop_write(filp->f_path.mnt); + mnt_drop_write_file(filp); return error; } static int gfs2_set_flags(struct file *filp, u32 __user *ptr) { - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); u32 fsflags, gfsflags; if (get_user(fsflags, ptr)) @@ -288,6 +295,7 @@ static int gfs2_set_flags(struct file *filp, u32 __user *ptr) gfsflags = fsflags_cvt(fsflags_to_gfs2, fsflags); if (!S_ISDIR(inode->i_mode)) { + gfsflags &= ~GFS2_DIF_TOPDIR; if (gfsflags & GFS2_DIF_INHERIT_JDATA) gfsflags ^= (GFS2_DIF_JDATA | GFS2_DIF_INHERIT_JDATA); return do_gfs2_set_flags(filp, gfsflags, ~0); @@ -302,11 +310,36 @@ static long gfs2_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) return gfs2_get_flags(filp, (u32 __user *)arg); case FS_IOC_SETFLAGS: return gfs2_set_flags(filp, (u32 __user *)arg); + case FITRIM: + return gfs2_fitrim(filp, (void __user *)arg); } return -ENOTTY; } /** + * gfs2_size_hint - Give a hint to the size of a write request + * @filep: The struct file + * @offset: The file offset of the write + * @size: The length of the write + * + * When we are about to do a write, this function records the total + * write size in order to provide a suitable hint to the lower layers + * about how many blocks will be required. + * + */ + +static void gfs2_size_hint(struct file *filep, loff_t offset, size_t size) +{ + struct inode *inode = file_inode(filep); + struct gfs2_sbd *sdp = GFS2_SB(inode); + struct gfs2_inode *ip = GFS2_I(inode); + size_t blks = (size + sdp->sd_sb.sb_bsize - 1) >> sdp->sd_sb.sb_bsize_shift; + int hint = min_t(size_t, INT_MAX, blks); + + atomic_set(&ip->i_res->rs_sizehint, hint); +} + +/** * gfs2_allocate_page_backing - Use bmap to allocate blocks * @page: The (locked) page to allocate backing for * @@ -338,7 +371,7 @@ static int gfs2_allocate_page_backing(struct page *page) /** * gfs2_page_mkwrite - Make a shared, mmap()ed, page writable * @vma: The virtual memory area - * @page: The page which is about to become writable + * @vmf: The virtual memory fault containing the page to become writable * * When the page becomes writable, we need to ensure that we have * blocks allocated on disk to back that page. @@ -347,37 +380,59 @@ static int gfs2_allocate_page_backing(struct page *page) static int gfs2_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf) { struct page *page = vmf->page; - struct inode *inode = vma->vm_file->f_path.dentry->d_inode; + struct inode *inode = file_inode(vma->vm_file); struct gfs2_inode *ip = GFS2_I(inode); struct gfs2_sbd *sdp = GFS2_SB(inode); + struct gfs2_alloc_parms ap = { .aflags = 0, }; unsigned long last_index; u64 pos = page->index << PAGE_CACHE_SHIFT; unsigned int data_blocks, ind_blocks, rblocks; struct gfs2_holder gh; - struct gfs2_alloc *al; + loff_t size; int ret; + sb_start_pagefault(inode->i_sb); + + /* Update file times before taking page lock */ + file_update_time(vma->vm_file); + + ret = get_write_access(inode); + if (ret) + goto out; + + ret = gfs2_rs_alloc(ip); + if (ret) + goto out_write_access; + + gfs2_size_hint(vma->vm_file, pos, PAGE_CACHE_SIZE); + gfs2_holder_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, &gh); ret = gfs2_glock_nq(&gh); if (ret) - goto out; + goto out_uninit; set_bit(GLF_DIRTY, &ip->i_gl->gl_flags); set_bit(GIF_SW_PAGED, &ip->i_flags); - if (!gfs2_write_alloc_required(ip, pos, PAGE_CACHE_SIZE)) + if (!gfs2_write_alloc_required(ip, pos, PAGE_CACHE_SIZE)) { + lock_page(page); + if (!PageUptodate(page) || page->mapping != inode->i_mapping) { + ret = -EAGAIN; + unlock_page(page); + } goto out_unlock; - ret = -ENOMEM; - al = gfs2_alloc_get(ip); - if (al == NULL) + } + + ret = gfs2_rindex_update(sdp); + if (ret) goto out_unlock; ret = gfs2_quota_lock_check(ip); if (ret) - goto out_alloc_put; + goto out_unlock; gfs2_write_calc_reserv(ip, PAGE_CACHE_SIZE, &data_blocks, &ind_blocks); - al->al_requested = data_blocks + ind_blocks; - ret = gfs2_inplace_reserve(ip); + ap.target = data_blocks + ind_blocks; + ret = gfs2_inplace_reserve(ip, &ap); if (ret) goto out_quota_unlock; @@ -386,7 +441,7 @@ static int gfs2_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf) rblocks += data_blocks ? data_blocks : 1; if (ind_blocks || data_blocks) { rblocks += RES_STATFS + RES_QUOTA; - rblocks += gfs2_rg_blocks(al); + rblocks += gfs2_rg_blocks(ip, data_blocks + ind_blocks); } ret = gfs2_trans_begin(sdp, rblocks, 0); if (ret) @@ -394,42 +449,54 @@ static int gfs2_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf) lock_page(page); ret = -EINVAL; - last_index = ip->i_inode.i_size >> PAGE_CACHE_SHIFT; - if (page->index > last_index) - goto out_unlock_page; + size = i_size_read(inode); + last_index = (size - 1) >> PAGE_CACHE_SHIFT; + /* Check page index against inode size */ + if (size == 0 || (page->index > last_index)) + goto out_trans_end; + + ret = -EAGAIN; + /* If truncated, we must retry the operation, we may have raced + * with the glock demotion code. + */ + if (!PageUptodate(page) || page->mapping != inode->i_mapping) + goto out_trans_end; + + /* Unstuff, if required, and allocate backing blocks for page */ ret = 0; - if (!PageUptodate(page) || page->mapping != ip->i_inode.i_mapping) - goto out_unlock_page; - if (gfs2_is_stuffed(ip)) { + if (gfs2_is_stuffed(ip)) ret = gfs2_unstuff_dinode(ip, page); - if (ret) - goto out_unlock_page; - } - ret = gfs2_allocate_page_backing(page); + if (ret == 0) + ret = gfs2_allocate_page_backing(page); -out_unlock_page: - unlock_page(page); +out_trans_end: + if (ret) + unlock_page(page); gfs2_trans_end(sdp); out_trans_fail: gfs2_inplace_release(ip); out_quota_unlock: gfs2_quota_unlock(ip); -out_alloc_put: - gfs2_alloc_put(ip); out_unlock: gfs2_glock_dq(&gh); -out: +out_uninit: gfs2_holder_uninit(&gh); - if (ret == -ENOMEM) - ret = VM_FAULT_OOM; - else if (ret) - ret = VM_FAULT_SIGBUS; - return ret; + if (ret == 0) { + set_page_dirty(page); + wait_for_stable_page(page); + } +out_write_access: + put_write_access(inode); +out: + sb_end_pagefault(inode->i_sb); + return block_page_mkwrite_return(ret); } static const struct vm_operations_struct gfs2_vm_ops = { .fault = filemap_fault, + .map_pages = filemap_map_pages, .page_mkwrite = gfs2_page_mkwrite, + .remap_pages = generic_file_remap_pages, }; /** @@ -448,38 +515,49 @@ static int gfs2_mmap(struct file *file, struct vm_area_struct *vma) { struct gfs2_inode *ip = GFS2_I(file->f_mapping->host); - if (!(file->f_flags & O_NOATIME)) { + if (!(file->f_flags & O_NOATIME) && + !IS_NOATIME(&ip->i_inode)) { struct gfs2_holder i_gh; int error; - gfs2_holder_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, &i_gh); - error = gfs2_glock_nq(&i_gh); + error = gfs2_glock_nq_init(ip->i_gl, LM_ST_SHARED, LM_FLAG_ANY, + &i_gh); + if (error) + return error; + /* grab lock to update inode */ + gfs2_glock_dq_uninit(&i_gh); file_accessed(file); - if (error == 0) - gfs2_glock_dq_uninit(&i_gh); } vma->vm_ops = &gfs2_vm_ops; - vma->vm_flags |= VM_CAN_NONLINEAR; return 0; } /** - * gfs2_open - open a file - * @inode: the inode to open - * @file: the struct file for this opening + * gfs2_open_common - This is common to open and atomic_open + * @inode: The inode being opened + * @file: The file being opened * - * Returns: errno + * This maybe called under a glock or not depending upon how it has + * been called. We must always be called under a glock for regular + * files, however. For other file types, it does not matter whether + * we hold the glock or not. + * + * Returns: Error code or 0 for success */ -static int gfs2_open(struct inode *inode, struct file *file) +int gfs2_open_common(struct inode *inode, struct file *file) { - struct gfs2_inode *ip = GFS2_I(inode); - struct gfs2_holder i_gh; struct gfs2_file *fp; - int error; + int ret; - fp = kzalloc(sizeof(struct gfs2_file), GFP_KERNEL); + if (S_ISREG(inode->i_mode)) { + ret = generic_file_open(inode, file); + if (ret) + return ret; + } + + fp = kzalloc(sizeof(struct gfs2_file), GFP_NOFS); if (!fp) return -ENOMEM; @@ -487,99 +565,126 @@ static int gfs2_open(struct inode *inode, struct file *file) gfs2_assert_warn(GFS2_SB(inode), !file->private_data); file->private_data = fp; + return 0; +} + +/** + * gfs2_open - open a file + * @inode: the inode to open + * @file: the struct file for this opening + * + * After atomic_open, this function is only used for opening files + * which are already cached. We must still get the glock for regular + * files to ensure that we have the file size uptodate for the large + * file check which is in the common code. That is only an issue for + * regular files though. + * + * Returns: errno + */ + +static int gfs2_open(struct inode *inode, struct file *file) +{ + struct gfs2_inode *ip = GFS2_I(inode); + struct gfs2_holder i_gh; + int error; + bool need_unlock = false; if (S_ISREG(ip->i_inode.i_mode)) { error = gfs2_glock_nq_init(ip->i_gl, LM_ST_SHARED, LM_FLAG_ANY, &i_gh); if (error) - goto fail; + return error; + need_unlock = true; + } - if (!(file->f_flags & O_LARGEFILE) && - i_size_read(inode) > MAX_NON_LFS) { - error = -EOVERFLOW; - goto fail_gunlock; - } + error = gfs2_open_common(inode, file); + if (need_unlock) gfs2_glock_dq_uninit(&i_gh); - } - return 0; - -fail_gunlock: - gfs2_glock_dq_uninit(&i_gh); -fail: - file->private_data = NULL; - kfree(fp); return error; } /** - * gfs2_close - called to close a struct file + * gfs2_release - called to close a struct file * @inode: the inode the struct file belongs to * @file: the struct file being closed * * Returns: errno */ -static int gfs2_close(struct inode *inode, struct file *file) +static int gfs2_release(struct inode *inode, struct file *file) { - struct gfs2_sbd *sdp = inode->i_sb->s_fs_info; - struct gfs2_file *fp; + struct gfs2_inode *ip = GFS2_I(inode); - fp = file->private_data; + kfree(file->private_data); file->private_data = NULL; - if (gfs2_assert_warn(sdp, fp)) - return -EIO; - - kfree(fp); + if (!(file->f_mode & FMODE_WRITE)) + return 0; + gfs2_rs_delete(ip, &inode->i_writecount); return 0; } /** * gfs2_fsync - sync the dirty data for a file (across the cluster) - * @file: the file that points to the dentry (we ignore this) - * @dentry: the dentry that points to the inode to sync + * @file: the file that points to the dentry + * @start: the start position in the file to sync + * @end: the end position in the file to sync + * @datasync: set if we can ignore timestamp changes * - * The VFS will flush "normal" data for us. We only need to worry - * about metadata here. For journaled data, we just do a log flush - * as we can't avoid it. Otherwise we can just bale out if datasync - * is set. For stuffed inodes we must flush the log in order to - * ensure that all data is on disk. + * We split the data flushing here so that we don't wait for the data + * until after we've also sent the metadata to disk. Note that for + * data=ordered, we will write & wait for the data at the log flush + * stage anyway, so this is unlikely to make much of a difference + * except in the data=writeback case. * - * The call to write_inode_now() is there to write back metadata and - * the inode itself. It does also try and write the data, but thats - * (hopefully) a no-op due to the VFS having already called filemap_fdatawrite() - * for us. + * If the fdatawrite fails due to any reason except -EIO, we will + * continue the remainder of the fsync, although we'll still report + * the error at the end. This is to match filemap_write_and_wait_range() + * behaviour. * * Returns: errno */ -static int gfs2_fsync(struct file *file, int datasync) +static int gfs2_fsync(struct file *file, loff_t start, loff_t end, + int datasync) { - struct inode *inode = file->f_mapping->host; - int sync_state = inode->i_state & (I_DIRTY_SYNC|I_DIRTY_DATASYNC); - int ret = 0; + struct address_space *mapping = file->f_mapping; + struct inode *inode = mapping->host; + int sync_state = inode->i_state & I_DIRTY; + struct gfs2_inode *ip = GFS2_I(inode); + int ret = 0, ret1 = 0; - if (gfs2_is_jdata(GFS2_I(inode))) { - gfs2_log_flush(GFS2_SB(inode), GFS2_I(inode)->i_gl); - return 0; + if (mapping->nrpages) { + ret1 = filemap_fdatawrite_range(mapping, start, end); + if (ret1 == -EIO) + return ret1; } - if (sync_state != 0) { - if (!datasync) - ret = write_inode_now(inode, 0); + if (!gfs2_is_jdata(ip)) + sync_state &= ~I_DIRTY_PAGES; + if (datasync) + sync_state &= ~I_DIRTY_SYNC; - if (gfs2_is_stuffed(GFS2_I(inode))) - gfs2_log_flush(GFS2_SB(inode), GFS2_I(inode)->i_gl); + if (sync_state) { + ret = sync_inode_metadata(inode, 1); + if (ret) + return ret; + if (gfs2_is_jdata(ip)) + filemap_write_and_wait(mapping); + gfs2_ail_flush(ip->i_gl, 1); } - return ret; + if (mapping->nrpages) + ret = filemap_fdatawait_range(mapping, start, end); + + return ret ? ret : ret1; } /** - * gfs2_file_aio_write - Perform a write to a file + * gfs2_file_write_iter - Perform a write to a file * @iocb: The io context * @iov: The data to write * @nr_segs: Number of @iov segments @@ -592,16 +697,20 @@ static int gfs2_fsync(struct file *file, int datasync) * */ -static ssize_t gfs2_file_aio_write(struct kiocb *iocb, const struct iovec *iov, - unsigned long nr_segs, loff_t pos) +static ssize_t gfs2_file_write_iter(struct kiocb *iocb, struct iov_iter *from) { struct file *file = iocb->ki_filp; + struct gfs2_inode *ip = GFS2_I(file_inode(file)); + int ret; + + ret = gfs2_rs_alloc(ip); + if (ret) + return ret; + + gfs2_size_hint(file, iocb->ki_pos, iov_iter_count(from)); if (file->f_flags & O_APPEND) { - struct dentry *dentry = file->f_dentry; - struct gfs2_inode *ip = GFS2_I(dentry->d_inode); struct gfs2_holder gh; - int ret; ret = gfs2_glock_nq_init(ip->i_gl, LM_ST_SHARED, 0, &gh); if (ret) @@ -609,72 +718,7 @@ static ssize_t gfs2_file_aio_write(struct kiocb *iocb, const struct iovec *iov, gfs2_glock_dq_uninit(&gh); } - return generic_file_aio_write(iocb, iov, nr_segs, pos); -} - -static void empty_write_end(struct page *page, unsigned from, - unsigned to) -{ - struct gfs2_inode *ip = GFS2_I(page->mapping->host); - - page_zero_new_buffers(page, from, to); - flush_dcache_page(page); - mark_page_accessed(page); - - if (!gfs2_is_writeback(ip)) - gfs2_page_add_databufs(ip, page, from, to); - - block_commit_write(page, from, to); -} - -static int write_empty_blocks(struct page *page, unsigned from, unsigned to) -{ - unsigned start, end, next; - struct buffer_head *bh, *head; - int error; - - if (!page_has_buffers(page)) { - error = __block_write_begin(page, from, to - from, gfs2_block_map); - if (unlikely(error)) - return error; - - empty_write_end(page, from, to); - return 0; - } - - bh = head = page_buffers(page); - next = end = 0; - while (next < from) { - next += bh->b_size; - bh = bh->b_this_page; - } - start = next; - do { - next += bh->b_size; - if (buffer_mapped(bh)) { - if (end) { - error = __block_write_begin(page, start, end - start, - gfs2_block_map); - if (unlikely(error)) - return error; - empty_write_end(page, start, end); - end = 0; - } - start = next; - } - else - end = next; - bh = bh->b_this_page; - } while (next < to); - - if (end) { - error = __block_write_begin(page, start, end - start, gfs2_block_map); - if (unlikely(error)) - return error; - empty_write_end(page, start, end); - } - - return 0; + return generic_file_write_iter(iocb, from); } static int fallocate_chunk(struct inode *inode, loff_t offset, loff_t len, @@ -683,22 +727,15 @@ static int fallocate_chunk(struct inode *inode, loff_t offset, loff_t len, struct gfs2_inode *ip = GFS2_I(inode); struct buffer_head *dibh; int error; - u64 start = offset >> PAGE_CACHE_SHIFT; - unsigned int start_offset = offset & ~PAGE_CACHE_MASK; - u64 end = (offset + len - 1) >> PAGE_CACHE_SHIFT; - pgoff_t curr; - struct page *page; - unsigned int end_offset = (offset + len) & ~PAGE_CACHE_MASK; - unsigned int from, to; - - if (!end_offset) - end_offset = PAGE_CACHE_SIZE; + loff_t size = len; + unsigned int nr_blks; + sector_t lblock = offset >> inode->i_blkbits; error = gfs2_meta_inode_buffer(ip, &dibh); if (unlikely(error)) - goto out; + return error; - gfs2_trans_add_bh(ip->i_gl, dibh, 1); + gfs2_trans_add_meta(ip->i_gl, dibh); if (gfs2_is_stuffed(ip)) { error = gfs2_unstuff_dinode(ip, NULL); @@ -706,40 +743,31 @@ static int fallocate_chunk(struct inode *inode, loff_t offset, loff_t len, goto out; } - curr = start; - offset = start << PAGE_CACHE_SHIFT; - from = start_offset; - to = PAGE_CACHE_SIZE; - while (curr <= end) { - page = grab_cache_page_write_begin(inode->i_mapping, curr, - AOP_FLAG_NOFS); - if (unlikely(!page)) { - error = -ENOMEM; - goto out; - } + while (len) { + struct buffer_head bh_map = { .b_state = 0, .b_blocknr = 0 }; + bh_map.b_size = len; + set_buffer_zeronew(&bh_map); - if (curr == end) - to = end_offset; - error = write_empty_blocks(page, from, to); - if (!error && offset + to > inode->i_size && - !(mode & FALLOC_FL_KEEP_SIZE)) { - i_size_write(inode, offset + to); - } - unlock_page(page); - page_cache_release(page); - if (error) + error = gfs2_block_map(inode, lblock, &bh_map, 1); + if (unlikely(error)) goto out; - curr++; - offset += PAGE_CACHE_SIZE; - from = 0; + len -= bh_map.b_size; + nr_blks = bh_map.b_size >> inode->i_blkbits; + lblock += nr_blks; + if (!buffer_new(&bh_map)) + continue; + if (unlikely(!buffer_zeronew(&bh_map))) { + error = -EIO; + goto out; + } } + if (offset + size > inode->i_size && !(mode & FALLOC_FL_KEEP_SIZE)) + i_size_write(inode, offset + size); - gfs2_dinode_out(ip, dibh->b_data); mark_inode_dirty(inode); - brelse(dibh); - out: + brelse(dibh); return error; } @@ -747,7 +775,7 @@ static void calc_max_reserv(struct gfs2_inode *ip, loff_t max, loff_t *len, unsigned int *data_blocks, unsigned int *ind_blocks) { const struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); - unsigned int max_blocks = ip->i_alloc->al_rgd->rd_free_clone; + unsigned int max_blocks = ip->i_rgd->rd_free_clone; unsigned int tmp, max_data = max_blocks - 3 * (sdp->sd_max_height - 1); for (tmp = max_data; tmp > sdp->sd_diptrs;) { @@ -770,67 +798,82 @@ static void calc_max_reserv(struct gfs2_inode *ip, loff_t max, loff_t *len, static long gfs2_fallocate(struct file *file, int mode, loff_t offset, loff_t len) { - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); struct gfs2_sbd *sdp = GFS2_SB(inode); struct gfs2_inode *ip = GFS2_I(inode); + struct gfs2_alloc_parms ap = { .aflags = 0, }; unsigned int data_blocks = 0, ind_blocks = 0, rblocks; loff_t bytes, max_bytes; - struct gfs2_alloc *al; int error; + const loff_t pos = offset; + const loff_t count = len; + loff_t bsize_mask = ~((loff_t)sdp->sd_sb.sb_bsize - 1); loff_t next = (offset + len - 1) >> sdp->sd_sb.sb_bsize_shift; + loff_t max_chunk_size = UINT_MAX & bsize_mask; + struct gfs2_holder gh; + next = (next + 1) << sdp->sd_sb.sb_bsize_shift; /* We only support the FALLOC_FL_KEEP_SIZE mode */ if (mode & ~FALLOC_FL_KEEP_SIZE) return -EOPNOTSUPP; - offset = (offset >> sdp->sd_sb.sb_bsize_shift) << - sdp->sd_sb.sb_bsize_shift; + offset &= bsize_mask; len = next - offset; bytes = sdp->sd_max_rg_data * sdp->sd_sb.sb_bsize / 2; if (!bytes) bytes = UINT_MAX; + bytes &= bsize_mask; + if (bytes == 0) + bytes = sdp->sd_sb.sb_bsize; + + error = gfs2_rs_alloc(ip); + if (error) + return error; - gfs2_holder_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, &ip->i_gh); - error = gfs2_glock_nq(&ip->i_gh); + mutex_lock(&inode->i_mutex); + + gfs2_holder_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, &gh); + error = gfs2_glock_nq(&gh); if (unlikely(error)) goto out_uninit; - if (!gfs2_write_alloc_required(ip, offset, len)) - goto out_unlock; + gfs2_size_hint(file, offset, len); while (len > 0) { if (len < bytes) bytes = len; - al = gfs2_alloc_get(ip); - if (!al) { - error = -ENOMEM; - goto out_unlock; + if (!gfs2_write_alloc_required(ip, offset, bytes)) { + len -= bytes; + offset += bytes; + continue; } - error = gfs2_quota_lock_check(ip); if (error) - goto out_alloc_put; + goto out_unlock; retry: gfs2_write_calc_reserv(ip, bytes, &data_blocks, &ind_blocks); - al->al_requested = data_blocks + ind_blocks; - error = gfs2_inplace_reserve(ip); + ap.target = data_blocks + ind_blocks; + error = gfs2_inplace_reserve(ip, &ap); if (error) { if (error == -ENOSPC && bytes > sdp->sd_sb.sb_bsize) { bytes >>= 1; + bytes &= bsize_mask; + if (bytes == 0) + bytes = sdp->sd_sb.sb_bsize; goto retry; } goto out_qunlock; } max_bytes = bytes; - calc_max_reserv(ip, len, &max_bytes, &data_blocks, &ind_blocks); - al->al_requested = data_blocks + ind_blocks; + calc_max_reserv(ip, (len > max_chunk_size)? max_chunk_size: len, + &max_bytes, &data_blocks, &ind_blocks); rblocks = RES_DINODE + ind_blocks + RES_STATFS + RES_QUOTA + - RES_RG_HDR + gfs2_rg_blocks(al); + RES_RG_HDR + gfs2_rg_blocks(ip, data_blocks + ind_blocks); if (gfs2_is_jdata(ip)) rblocks += data_blocks ? data_blocks : 1; @@ -849,20 +892,21 @@ retry: offset += max_bytes; gfs2_inplace_release(ip); gfs2_quota_unlock(ip); - gfs2_alloc_put(ip); } + + if (error == 0) + error = generic_write_sync(file, pos, count); goto out_unlock; out_trans_fail: gfs2_inplace_release(ip); out_qunlock: gfs2_quota_unlock(ip); -out_alloc_put: - gfs2_alloc_put(ip); out_unlock: - gfs2_glock_dq(&ip->i_gh); + gfs2_glock_dq(&gh); out_uninit: - gfs2_holder_uninit(&ip->i_gh); + gfs2_holder_uninit(&gh); + mutex_unlock(&inode->i_mutex); return error; } @@ -878,7 +922,7 @@ out_uninit: * cluster; until we do, disable leases (by just returning -EINVAL), * unless the administrator has requested purely local locking. * - * Locking: called under lock_flocks + * Locking: called under i_lock * * Returns: errno */ @@ -913,8 +957,11 @@ static int gfs2_lock(struct file *file, int cmd, struct file_lock *fl) cmd = F_SETLK; fl->fl_type = F_UNLCK; } - if (unlikely(test_bit(SDF_SHUTDOWN, &sdp->sd_flags))) + if (unlikely(test_bit(SDF_SHUTDOWN, &sdp->sd_flags))) { + if (fl->fl_type == F_UNLCK) + posix_lock_file_wait(file, fl); return -EIO; + } if (IS_GETLK(cmd)) return dlm_posix_get(ls->ls_dlm, ip->i_no_addr, file, fl); else if (fl->fl_type == F_UNLCK) @@ -927,14 +974,14 @@ static int do_flock(struct file *file, int cmd, struct file_lock *fl) { struct gfs2_file *fp = file->private_data; struct gfs2_holder *fl_gh = &fp->f_fl_gh; - struct gfs2_inode *ip = GFS2_I(file->f_path.dentry->d_inode); + struct gfs2_inode *ip = GFS2_I(file_inode(file)); struct gfs2_glock *gl; unsigned int state; int flags; int error = 0; state = (fl->fl_type == F_WRLCK) ? LM_ST_EXCLUSIVE : LM_ST_SHARED; - flags = (IS_SETLKW(cmd) ? 0 : LM_FLAG_TRY) | GL_EXACT | GL_NOCACHE; + flags = (IS_SETLKW(cmd) ? 0 : LM_FLAG_TRY) | GL_EXACT; mutex_lock(&fp->f_fl_mutex); @@ -944,7 +991,7 @@ static int do_flock(struct file *file, int cmd, struct file_lock *fl) goto out; flock_lock_file_wait(file, &(struct file_lock){.fl_type = F_UNLCK}); - gfs2_glock_dq_wait(fl_gh); + gfs2_glock_dq(fl_gh); gfs2_holder_reinit(state, flags, fl_gh); } else { error = gfs2_glock_get(GFS2_SB(&ip->i_inode), ip->i_no_addr, @@ -976,8 +1023,10 @@ static void do_unflock(struct file *file, struct file_lock *fl) mutex_lock(&fp->f_fl_mutex); flock_lock_file_wait(file, fl); - if (fl_gh->gh_gl) - gfs2_glock_dq_uninit(fl_gh); + if (fl_gh->gh_gl) { + gfs2_glock_dq_wait(fl_gh); + gfs2_holder_uninit(fl_gh); + } mutex_unlock(&fp->f_fl_mutex); } @@ -1007,28 +1056,28 @@ static int gfs2_flock(struct file *file, int cmd, struct file_lock *fl) const struct file_operations gfs2_file_fops = { .llseek = gfs2_llseek, - .read = do_sync_read, - .aio_read = generic_file_aio_read, - .write = do_sync_write, - .aio_write = gfs2_file_aio_write, + .read = new_sync_read, + .read_iter = generic_file_read_iter, + .write = new_sync_write, + .write_iter = gfs2_file_write_iter, .unlocked_ioctl = gfs2_ioctl, .mmap = gfs2_mmap, .open = gfs2_open, - .release = gfs2_close, + .release = gfs2_release, .fsync = gfs2_fsync, .lock = gfs2_lock, .flock = gfs2_flock, .splice_read = generic_file_splice_read, - .splice_write = generic_file_splice_write, + .splice_write = iter_file_splice_write, .setlease = gfs2_setlease, .fallocate = gfs2_fallocate, }; const struct file_operations gfs2_dir_fops = { - .readdir = gfs2_readdir, + .iterate = gfs2_readdir, .unlocked_ioctl = gfs2_ioctl, .open = gfs2_open, - .release = gfs2_close, + .release = gfs2_release, .fsync = gfs2_fsync, .lock = gfs2_lock, .flock = gfs2_flock, @@ -1039,26 +1088,26 @@ const struct file_operations gfs2_dir_fops = { const struct file_operations gfs2_file_fops_nolock = { .llseek = gfs2_llseek, - .read = do_sync_read, - .aio_read = generic_file_aio_read, - .write = do_sync_write, - .aio_write = gfs2_file_aio_write, + .read = new_sync_read, + .read_iter = generic_file_read_iter, + .write = new_sync_write, + .write_iter = gfs2_file_write_iter, .unlocked_ioctl = gfs2_ioctl, .mmap = gfs2_mmap, .open = gfs2_open, - .release = gfs2_close, + .release = gfs2_release, .fsync = gfs2_fsync, .splice_read = generic_file_splice_read, - .splice_write = generic_file_splice_write, + .splice_write = iter_file_splice_write, .setlease = generic_setlease, .fallocate = gfs2_fallocate, }; const struct file_operations gfs2_dir_fops_nolock = { - .readdir = gfs2_readdir, + .iterate = gfs2_readdir, .unlocked_ioctl = gfs2_ioctl, .open = gfs2_open, - .release = gfs2_close, + .release = gfs2_release, .fsync = gfs2_fsync, .llseek = default_llseek, }; diff --git a/fs/gfs2/glock.c b/fs/gfs2/glock.c index 7cd9a5a68d5..ee4e04fe60f 100644 --- a/fs/gfs2/glock.c +++ b/fs/gfs2/glock.c @@ -7,6 +7,8 @@ * of the GNU General Public License version 2. */ +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + #include <linux/sched.h> #include <linux/slab.h> #include <linux/spinlock.h> @@ -26,6 +28,12 @@ #include <linux/freezer.h> #include <linux/workqueue.h> #include <linux/jiffies.h> +#include <linux/rcupdate.h> +#include <linux/rculist_bl.h> +#include <linux/bit_spinlock.h> +#include <linux/percpu.h> +#include <linux/list_sort.h> +#include <linux/lockref.h> #include "gfs2.h" #include "incore.h" @@ -41,22 +49,16 @@ #define CREATE_TRACE_POINTS #include "trace_gfs2.h" -struct gfs2_gl_hash_bucket { - struct hlist_head hb_list; -}; - struct gfs2_glock_iter { - int hash; /* hash bucket index */ - struct gfs2_sbd *sdp; /* incore superblock */ - struct gfs2_glock *gl; /* current glock struct */ - char string[512]; /* scratch space */ + int hash; /* hash bucket index */ + unsigned nhash; /* Index within current bucket */ + struct gfs2_sbd *sdp; /* incore superblock */ + struct gfs2_glock *gl; /* current glock struct */ + loff_t last_pos; /* last position */ }; typedef void (*glock_examiner) (struct gfs2_glock * gl); -static int gfs2_dump_lockstate(struct gfs2_sbd *sdp); -static int __dump_glock(struct seq_file *seq, const struct gfs2_glock *gl); -#define GLOCK_BUG_ON(gl,x) do { if (unlikely(x)) { __dump_glock(NULL, gl); BUG(); } } while(0) static void do_xmote(struct gfs2_glock *gl, struct gfs2_holder *gh, unsigned int target); static struct dentry *gfs2_root; @@ -70,57 +72,9 @@ static DEFINE_SPINLOCK(lru_lock); #define GFS2_GL_HASH_SIZE (1 << GFS2_GL_HASH_SHIFT) #define GFS2_GL_HASH_MASK (GFS2_GL_HASH_SIZE - 1) -static struct gfs2_gl_hash_bucket gl_hash_table[GFS2_GL_HASH_SIZE]; +static struct hlist_bl_head gl_hash_table[GFS2_GL_HASH_SIZE]; static struct dentry *gfs2_root; -/* - * Despite what you might think, the numbers below are not arbitrary :-) - * They are taken from the ipv4 routing hash code, which is well tested - * and thus should be nearly optimal. Later on we might tweek the numbers - * but for now this should be fine. - * - * The reason for putting the locks in a separate array from the list heads - * is that we can have fewer locks than list heads and save memory. We use - * the same hash function for both, but with a different hash mask. - */ -#if defined(CONFIG_SMP) || defined(CONFIG_DEBUG_SPINLOCK) || \ - defined(CONFIG_PROVE_LOCKING) - -#ifdef CONFIG_LOCKDEP -# define GL_HASH_LOCK_SZ 256 -#else -# if NR_CPUS >= 32 -# define GL_HASH_LOCK_SZ 4096 -# elif NR_CPUS >= 16 -# define GL_HASH_LOCK_SZ 2048 -# elif NR_CPUS >= 8 -# define GL_HASH_LOCK_SZ 1024 -# elif NR_CPUS >= 4 -# define GL_HASH_LOCK_SZ 512 -# else -# define GL_HASH_LOCK_SZ 256 -# endif -#endif - -/* We never want more locks than chains */ -#if GFS2_GL_HASH_SIZE < GL_HASH_LOCK_SZ -# undef GL_HASH_LOCK_SZ -# define GL_HASH_LOCK_SZ GFS2_GL_HASH_SIZE -#endif - -static rwlock_t gl_hash_locks[GL_HASH_LOCK_SZ]; - -static inline rwlock_t *gl_lock_addr(unsigned int x) -{ - return &gl_hash_locks[x & (GL_HASH_LOCK_SZ-1)]; -} -#else /* not SMP, so no spinlocks required */ -static inline rwlock_t *gl_lock_addr(unsigned int x) -{ - return NULL; -} -#endif - /** * gl_hash() - Turn glock number into hash bucket number * @lock: The glock number @@ -141,25 +95,35 @@ static unsigned int gl_hash(const struct gfs2_sbd *sdp, return h; } -/** - * glock_free() - Perform a few checks and then release struct gfs2_glock - * @gl: The glock to release - * - * Also calls lock module to release its internal structure for this glock. - * - */ +static inline void spin_lock_bucket(unsigned int hash) +{ + hlist_bl_lock(&gl_hash_table[hash]); +} + +static inline void spin_unlock_bucket(unsigned int hash) +{ + hlist_bl_unlock(&gl_hash_table[hash]); +} + +static void gfs2_glock_dealloc(struct rcu_head *rcu) +{ + struct gfs2_glock *gl = container_of(rcu, struct gfs2_glock, gl_rcu); + + if (gl->gl_ops->go_flags & GLOF_ASPACE) { + kmem_cache_free(gfs2_glock_aspace_cachep, gl); + } else { + kfree(gl->gl_lksb.sb_lvbptr); + kmem_cache_free(gfs2_glock_cachep, gl); + } +} -static void glock_free(struct gfs2_glock *gl) +void gfs2_glock_free(struct gfs2_glock *gl) { struct gfs2_sbd *sdp = gl->gl_sbd; - struct address_space *mapping = gfs2_glock2aspace(gl); - struct kmem_cache *cachep = gfs2_glock_cachep; - GLOCK_BUG_ON(gl, mapping && mapping->nrpages); - trace_gfs2_glock_put(gl); - if (mapping) - cachep = gfs2_glock_aspace_cachep; - sdp->sd_lockstruct.ls_ops->lm_put_lock(cachep, gl); + call_rcu(&gl->gl_rcu, gfs2_glock_dealloc); + if (atomic_dec_and_test(&sdp->sd_glock_disposal)) + wake_up(&sdp->sd_glock_wait); } /** @@ -168,10 +132,10 @@ static void glock_free(struct gfs2_glock *gl) * */ -void gfs2_glock_hold(struct gfs2_glock *gl) +static void gfs2_glock_hold(struct gfs2_glock *gl) { - GLOCK_BUG_ON(gl, atomic_read(&gl->gl_ref) == 0); - atomic_inc(&gl->gl_ref); + GLOCK_BUG_ON(gl, __lockref_is_dead(&gl->gl_lockref)); + lockref_get(&gl->gl_lockref); } /** @@ -194,40 +158,35 @@ static int demote_ok(const struct gfs2_glock *gl) return 1; } -/** - * gfs2_glock_schedule_for_reclaim - Add a glock to the reclaim list - * @gl: the glock - * - */ -static void gfs2_glock_schedule_for_reclaim(struct gfs2_glock *gl) +void gfs2_glock_add_to_lru(struct gfs2_glock *gl) { - int may_reclaim; - may_reclaim = (demote_ok(gl) && - (atomic_read(&gl->gl_ref) == 1 || - (gl->gl_name.ln_type == LM_TYPE_INODE && - atomic_read(&gl->gl_ref) <= 2))); spin_lock(&lru_lock); - if (list_empty(&gl->gl_lru) && may_reclaim) { - list_add_tail(&gl->gl_lru, &lru_list); + + if (!list_empty(&gl->gl_lru)) + list_del_init(&gl->gl_lru); + else atomic_inc(&lru_count); - } + + list_add_tail(&gl->gl_lru, &lru_list); + set_bit(GLF_LRU, &gl->gl_flags); spin_unlock(&lru_lock); } -/** - * gfs2_glock_put_nolock() - Decrement reference count on glock - * @gl: The glock to put - * - * This function should only be used if the caller has its own reference - * to the glock, in addition to the one it is dropping. - */ +static void __gfs2_glock_remove_from_lru(struct gfs2_glock *gl) +{ + if (!list_empty(&gl->gl_lru)) { + list_del_init(&gl->gl_lru); + atomic_dec(&lru_count); + clear_bit(GLF_LRU, &gl->gl_flags); + } +} -void gfs2_glock_put_nolock(struct gfs2_glock *gl) +static void gfs2_glock_remove_from_lru(struct gfs2_glock *gl) { - if (atomic_dec_and_test(&gl->gl_ref)) - GLOCK_BUG_ON(gl, 1); - gfs2_glock_schedule_for_reclaim(gl); + spin_lock(&lru_lock); + __gfs2_glock_remove_from_lru(gl); + spin_unlock(&lru_lock); } /** @@ -236,30 +195,27 @@ void gfs2_glock_put_nolock(struct gfs2_glock *gl) * */ -int gfs2_glock_put(struct gfs2_glock *gl) +void gfs2_glock_put(struct gfs2_glock *gl) { - int rv = 0; + struct gfs2_sbd *sdp = gl->gl_sbd; + struct address_space *mapping = gfs2_glock2aspace(gl); - write_lock(gl_lock_addr(gl->gl_hash)); - if (atomic_dec_and_lock(&gl->gl_ref, &lru_lock)) { - hlist_del(&gl->gl_list); - if (!list_empty(&gl->gl_lru)) { - list_del_init(&gl->gl_lru); - atomic_dec(&lru_count); - } - spin_unlock(&lru_lock); - write_unlock(gl_lock_addr(gl->gl_hash)); - GLOCK_BUG_ON(gl, !list_empty(&gl->gl_holders)); - glock_free(gl); - rv = 1; - goto out; - } - spin_lock(&gl->gl_spin); - gfs2_glock_schedule_for_reclaim(gl); - spin_unlock(&gl->gl_spin); - write_unlock(gl_lock_addr(gl->gl_hash)); -out: - return rv; + if (lockref_put_or_lock(&gl->gl_lockref)) + return; + + lockref_mark_dead(&gl->gl_lockref); + + spin_lock(&lru_lock); + __gfs2_glock_remove_from_lru(gl); + spin_unlock(&lru_lock); + spin_unlock(&gl->gl_lockref.lock); + spin_lock_bucket(gl->gl_hash); + hlist_bl_del_rcu(&gl->gl_list); + spin_unlock_bucket(gl->gl_hash); + GLOCK_BUG_ON(gl, !list_empty(&gl->gl_holders)); + GLOCK_BUG_ON(gl, mapping && mapping->nrpages); + trace_gfs2_glock_put(gl); + sdp->sd_lockstruct.ls_ops->lm_put_lock(gl); } /** @@ -275,17 +231,15 @@ static struct gfs2_glock *search_bucket(unsigned int hash, const struct lm_lockname *name) { struct gfs2_glock *gl; - struct hlist_node *h; + struct hlist_bl_node *h; - hlist_for_each_entry(gl, h, &gl_hash_table[hash].hb_list, gl_list) { + hlist_bl_for_each_entry_rcu(gl, h, &gl_hash_table[hash], gl_list) { if (!lm_name_equal(&gl->gl_name, name)) continue; if (gl->gl_sbd != sdp) continue; - - atomic_inc(&gl->gl_ref); - - return gl; + if (lockref_get_not_dead(&gl->gl_lockref)) + return gl; } return NULL; @@ -323,7 +277,7 @@ static inline int may_grant(const struct gfs2_glock *gl, const struct gfs2_holde static void gfs2_holder_wake(struct gfs2_holder *gh) { clear_bit(HIF_WAIT, &gh->gh_iflags); - smp_mb__after_clear_bit(); + smp_mb__after_atomic(); wake_up_bit(&gh->gh_iflags, HIF_WAIT); } @@ -436,14 +390,19 @@ static void state_change(struct gfs2_glock *gl, unsigned int new_state) held2 = (new_state != LM_ST_UNLOCKED); if (held1 != held2) { + GLOCK_BUG_ON(gl, __lockref_is_dead(&gl->gl_lockref)); if (held2) - gfs2_glock_hold(gl); + gl->gl_lockref.count++; else - gfs2_glock_put_nolock(gl); + gl->gl_lockref.count--; } if (held1 && held2 && list_empty(&gl->gl_holders)) clear_bit(GLF_QUEUED, &gl->gl_flags); + if (new_state != gl->gl_target) + /* shorten our minimum hold time */ + gl->gl_hold_time = max(gl->gl_hold_time - GL_GLOCK_HOLD_DECR, + GL_GLOCK_MIN_HOLD); gl->gl_state = new_state; gl->gl_tchange = jiffies; } @@ -452,7 +411,7 @@ static void gfs2_demote_wake(struct gfs2_glock *gl) { gl->gl_demote_state = LM_ST_EXCLUSIVE; clear_bit(GLF_DEMOTE, &gl->gl_flags); - smp_mb__after_clear_bit(); + smp_mb__after_atomic(); wake_up_bit(&gl->gl_flags, GLF_DEMOTE); } @@ -511,7 +470,7 @@ retry: do_xmote(gl, gh, LM_ST_UNLOCKED); break; default: /* Everything else */ - printk(KERN_ERR "GFS2: wanted %u got %u\n", gl->gl_target, state); + pr_err("wanted %u got %u\n", gl->gl_target, state); GLOCK_BUG_ON(gl, 1); } spin_unlock(&gl->gl_spin); @@ -568,23 +527,26 @@ __acquires(&gl->gl_spin) do_error(gl, 0); /* Fail queued try locks */ } gl->gl_req = target; + set_bit(GLF_BLOCKING, &gl->gl_flags); + if ((gl->gl_req == LM_ST_UNLOCKED) || + (gl->gl_state == LM_ST_EXCLUSIVE) || + (lck_flags & (LM_FLAG_TRY|LM_FLAG_TRY_1CB))) + clear_bit(GLF_BLOCKING, &gl->gl_flags); spin_unlock(&gl->gl_spin); - if (glops->go_xmote_th) - glops->go_xmote_th(gl); + if (glops->go_sync) + glops->go_sync(gl); if (test_bit(GLF_INVALIDATE_IN_PROGRESS, &gl->gl_flags)) glops->go_inval(gl, target == LM_ST_DEFERRED ? 0 : DIO_METADATA); clear_bit(GLF_INVALIDATE_IN_PROGRESS, &gl->gl_flags); gfs2_glock_hold(gl); - if (target != LM_ST_UNLOCKED && (gl->gl_state == LM_ST_SHARED || - gl->gl_state == LM_ST_DEFERRED) && - !(lck_flags & (LM_FLAG_TRY | LM_FLAG_TRY_1CB))) - lck_flags |= LM_FLAG_TRY_1CB; - if (sdp->sd_lockstruct.ls_ops->lm_lock) { /* lock_dlm */ ret = sdp->sd_lockstruct.ls_ops->lm_lock(gl, target, lck_flags); - GLOCK_BUG_ON(gl, ret); + if (ret) { + pr_err("lm_lock ret %d\n", ret); + GLOCK_BUG_ON(gl, 1); + } } else { /* lock_nolock */ finish_xmote(gl, target); if (queue_delayed_work(glock_workqueue, &gl->gl_work, 0) == 0) @@ -658,15 +620,15 @@ out: out_sched: clear_bit(GLF_LOCK, &gl->gl_flags); - smp_mb__after_clear_bit(); - gfs2_glock_hold(gl); + smp_mb__after_atomic(); + gl->gl_lockref.count++; if (queue_delayed_work(glock_workqueue, &gl->gl_work, 0) == 0) - gfs2_glock_put_nolock(gl); + gl->gl_lockref.count--; return; out_unlock: clear_bit(GLF_LOCK, &gl->gl_flags); - smp_mb__after_clear_bit(); + smp_mb__after_atomic(); return; } @@ -682,7 +644,7 @@ static void delete_work_func(struct work_struct *work) /* Note: Unsafe to dereference ip as we don't hold right refs/locks */ if (ip) - inode = gfs2_ilookup(sdp->sd_vfs, no_addr); + inode = gfs2_ilookup(sdp->sd_vfs, no_addr, 1); else inode = gfs2_lookup_by_inum(sdp, no_addr, NULL, GFS2_BLKST_UNLINKED); if (inode && !IS_ERR(inode)) { @@ -703,20 +665,30 @@ static void glock_work_func(struct work_struct *work) drop_ref = 1; } spin_lock(&gl->gl_spin); - if (test_and_clear_bit(GLF_PENDING_DEMOTE, &gl->gl_flags) && + if (test_bit(GLF_PENDING_DEMOTE, &gl->gl_flags) && gl->gl_state != LM_ST_UNLOCKED && gl->gl_demote_state != LM_ST_EXCLUSIVE) { unsigned long holdtime, now = jiffies; - holdtime = gl->gl_tchange + gl->gl_ops->go_min_hold_time; + + holdtime = gl->gl_tchange + gl->gl_hold_time; if (time_before(now, holdtime)) delay = holdtime - now; - set_bit(delay ? GLF_PENDING_DEMOTE : GLF_DEMOTE, &gl->gl_flags); + + if (!delay) { + clear_bit(GLF_PENDING_DEMOTE, &gl->gl_flags); + set_bit(GLF_DEMOTE, &gl->gl_flags); + } } run_queue(gl, 0); spin_unlock(&gl->gl_spin); - if (!delay || - queue_delayed_work(glock_workqueue, &gl->gl_work, delay) == 0) + if (!delay) gfs2_glock_put(gl); + else { + if (gl->gl_name.ln_type != LM_TYPE_INODE) + delay = 0; + if (queue_delayed_work(glock_workqueue, &gl->gl_work, delay) == 0) + gfs2_glock_put(gl); + } if (drop_ref) gfs2_glock_put(gl); } @@ -743,10 +715,11 @@ int gfs2_glock_get(struct gfs2_sbd *sdp, u64 number, struct gfs2_glock *gl, *tmp; unsigned int hash = gl_hash(sdp, &name); struct address_space *mapping; + struct kmem_cache *cachep; - read_lock(gl_lock_addr(hash)); + rcu_read_lock(); gl = search_bucket(hash, sdp, &name); - read_unlock(gl_lock_addr(hash)); + rcu_read_unlock(); *glp = gl; if (gl) @@ -755,27 +728,43 @@ int gfs2_glock_get(struct gfs2_sbd *sdp, u64 number, return -ENOENT; if (glops->go_flags & GLOF_ASPACE) - gl = kmem_cache_alloc(gfs2_glock_aspace_cachep, GFP_KERNEL); + cachep = gfs2_glock_aspace_cachep; else - gl = kmem_cache_alloc(gfs2_glock_cachep, GFP_KERNEL); + cachep = gfs2_glock_cachep; + gl = kmem_cache_alloc(cachep, GFP_NOFS); if (!gl) return -ENOMEM; + memset(&gl->gl_lksb, 0, sizeof(struct dlm_lksb)); + + if (glops->go_flags & GLOF_LVB) { + gl->gl_lksb.sb_lvbptr = kzalloc(GFS2_MIN_LVB_SIZE, GFP_NOFS); + if (!gl->gl_lksb.sb_lvbptr) { + kmem_cache_free(cachep, gl); + return -ENOMEM; + } + } + atomic_inc(&sdp->sd_glock_disposal); + gl->gl_sbd = sdp; gl->gl_flags = 0; gl->gl_name = name; - atomic_set(&gl->gl_ref, 1); + gl->gl_lockref.count = 1; gl->gl_state = LM_ST_UNLOCKED; gl->gl_target = LM_ST_UNLOCKED; gl->gl_demote_state = LM_ST_EXCLUSIVE; gl->gl_hash = hash; gl->gl_ops = glops; - snprintf(gl->gl_strname, GDLM_STRNAME_BYTES, "%8x%16llx", name.ln_type, (unsigned long long)number); - memset(&gl->gl_lksb, 0, sizeof(struct dlm_lksb)); - gl->gl_lksb.sb_lvbptr = gl->gl_lvb; + gl->gl_dstamp = ktime_set(0, 0); + preempt_disable(); + /* We use the global stats to estimate the initial per-glock stats */ + gl->gl_stats = this_cpu_ptr(sdp->sd_lkstats)->lkstats[glops->go_type]; + preempt_enable(); + gl->gl_stats.stats[GFS2_LKS_DCOUNT] = 0; + gl->gl_stats.stats[GFS2_LKS_QCOUNT] = 0; gl->gl_tchange = jiffies; gl->gl_object = NULL; - gl->gl_sbd = sdp; + gl->gl_hold_time = GL_GLOCK_DFT_HOLD; INIT_DELAYED_WORK(&gl->gl_work, glock_work_func); INIT_WORK(&gl->gl_delete, delete_work_func); @@ -785,20 +774,22 @@ int gfs2_glock_get(struct gfs2_sbd *sdp, u64 number, mapping->host = s->s_bdev->bd_inode; mapping->flags = 0; mapping_set_gfp_mask(mapping, GFP_NOFS); - mapping->assoc_mapping = NULL; + mapping->private_data = NULL; mapping->backing_dev_info = s->s_bdi; mapping->writeback_index = 0; } - write_lock(gl_lock_addr(hash)); + spin_lock_bucket(hash); tmp = search_bucket(hash, sdp, &name); if (tmp) { - write_unlock(gl_lock_addr(hash)); - glock_free(gl); + spin_unlock_bucket(hash); + kfree(gl->gl_lksb.sb_lvbptr); + kmem_cache_free(cachep, gl); + atomic_dec(&sdp->sd_glock_disposal); gl = tmp; } else { - hlist_add_head(&gl->gl_list, &gl_hash_table[hash].hb_list); - write_unlock(gl_lock_addr(hash)); + hlist_bl_add_head_rcu(&gl->gl_list, &gl_hash_table[hash]); + spin_unlock_bucket(hash); } *glp = gl; @@ -885,16 +876,25 @@ static int gfs2_glock_demote_wait(void *word) return 0; } -static void wait_on_holder(struct gfs2_holder *gh) -{ - might_sleep(); - wait_on_bit(&gh->gh_iflags, HIF_WAIT, gfs2_glock_holder_wait, TASK_UNINTERRUPTIBLE); -} +/** + * gfs2_glock_wait - wait on a glock acquisition + * @gh: the glock holder + * + * Returns: 0 on success + */ -static void wait_on_demote(struct gfs2_glock *gl) +int gfs2_glock_wait(struct gfs2_holder *gh) { + unsigned long time1 = jiffies; + might_sleep(); - wait_on_bit(&gl->gl_flags, GLF_DEMOTE, gfs2_glock_demote_wait, TASK_UNINTERRUPTIBLE); + wait_on_bit(&gh->gh_iflags, HIF_WAIT, gfs2_glock_holder_wait, TASK_UNINTERRUPTIBLE); + if (time_after(jiffies, time1 + HZ)) /* have we waited > a second? */ + /* Lengthen the minimum hold time. */ + gh->gh_gl->gl_hold_time = min(gh->gh_gl->gl_hold_time + + GL_GLOCK_HOLD_INCR, + GL_GLOCK_MAX_HOLD); + return gh->gh_error; } /** @@ -907,7 +907,7 @@ static void wait_on_demote(struct gfs2_glock *gl) */ static void handle_callback(struct gfs2_glock *gl, unsigned int state, - unsigned long delay) + unsigned long delay, bool remote) { int bit = delay ? GLF_PENDING_DEMOTE : GLF_DEMOTE; @@ -920,21 +920,8 @@ static void handle_callback(struct gfs2_glock *gl, unsigned int state, gl->gl_demote_state = LM_ST_UNLOCKED; } if (gl->gl_ops->go_callback) - gl->gl_ops->go_callback(gl); - trace_gfs2_demote_rq(gl); -} - -/** - * gfs2_glock_wait - wait on a glock acquisition - * @gh: the glock holder - * - * Returns: 0 on success - */ - -int gfs2_glock_wait(struct gfs2_holder *gh) -{ - wait_on_holder(gh); - return gh->gh_error; + gl->gl_ops->go_callback(gl, remote); + trace_gfs2_demote_rq(gl, remote); } void gfs2_print_dbg(struct seq_file *seq, const char *fmt, ...) @@ -945,14 +932,12 @@ void gfs2_print_dbg(struct seq_file *seq, const char *fmt, ...) va_start(args, fmt); if (seq) { - struct gfs2_glock_iter *gi = seq->private; - vsprintf(gi->string, fmt, args); - seq_printf(seq, gi->string); + seq_vprintf(seq, fmt, args); } else { vaf.fmt = fmt; vaf.va = &args; - printk(KERN_ERR " %pV", &vaf); + pr_err("%pV", &vaf); } va_end(args); @@ -976,7 +961,7 @@ __acquires(&gl->gl_spin) struct gfs2_sbd *sdp = gl->gl_sbd; struct list_head *insert_pt = NULL; struct gfs2_holder *gh2; - int try_lock = 0; + int try_futile = 0; BUG_ON(gh->gh_owner_pid == NULL); if (test_and_set_bit(HIF_WAIT, &gh->gh_iflags)) @@ -984,7 +969,7 @@ __acquires(&gl->gl_spin) if (gh->gh_flags & (LM_FLAG_TRY | LM_FLAG_TRY_1CB)) { if (test_bit(GLF_LOCK, &gl->gl_flags)) - try_lock = 1; + try_futile = !may_grant(gl, gh); if (test_bit(GLF_INVALIDATE_IN_PROGRESS, &gl->gl_flags)) goto fail; } @@ -993,9 +978,8 @@ __acquires(&gl->gl_spin) if (unlikely(gh2->gh_owner_pid == gh->gh_owner_pid && (gh->gh_gl->gl_ops->go_type != LM_TYPE_FLOCK))) goto trap_recursive; - if (try_lock && - !(gh2->gh_flags & (LM_FLAG_TRY | LM_FLAG_TRY_1CB)) && - !may_grant(gl, gh)) { + if (try_futile && + !(gh2->gh_flags & (LM_FLAG_TRY | LM_FLAG_TRY_1CB))) { fail: gh->gh_error = GLR_TRYFAILED; gfs2_holder_wake(gh); @@ -1007,13 +991,15 @@ fail: insert_pt = &gh2->gh_list; } set_bit(GLF_QUEUED, &gl->gl_flags); + trace_gfs2_glock_queue(gh, 1); + gfs2_glstats_inc(gl, GFS2_LKS_QCOUNT); + gfs2_sbstats_inc(gl, GFS2_LKS_QCOUNT); if (likely(insert_pt == NULL)) { list_add_tail(&gh->gh_list, &gl->gl_holders); if (unlikely(gh->gh_flags & LM_FLAG_PRIORITY)) goto do_cancel; return; } - trace_gfs2_glock_queue(gh, 1); list_add_tail(&gh->gh_list, insert_pt); do_cancel: gh = list_entry(gl->gl_holders.next, struct gfs2_holder, gh_list); @@ -1026,15 +1012,15 @@ do_cancel: return; trap_recursive: - print_symbol(KERN_ERR "original: %s\n", gh2->gh_ip); - printk(KERN_ERR "pid: %d\n", pid_nr(gh2->gh_owner_pid)); - printk(KERN_ERR "lock type: %d req lock state : %d\n", + pr_err("original: %pSR\n", (void *)gh2->gh_ip); + pr_err("pid: %d\n", pid_nr(gh2->gh_owner_pid)); + pr_err("lock type: %d req lock state : %d\n", gh2->gh_gl->gl_name.ln_type, gh2->gh_state); - print_symbol(KERN_ERR "new: %s\n", gh->gh_ip); - printk(KERN_ERR "pid: %d\n", pid_nr(gh->gh_owner_pid)); - printk(KERN_ERR "lock type: %d req lock state : %d\n", + pr_err("new: %pSR\n", (void *)gh->gh_ip); + pr_err("pid: %d\n", pid_nr(gh->gh_owner_pid)); + pr_err("lock type: %d req lock state : %d\n", gh->gh_gl->gl_name.ln_type, gh->gh_state); - __dump_glock(NULL, gl); + gfs2_dump_glock(NULL, gl); BUG(); } @@ -1056,11 +1042,18 @@ int gfs2_glock_nq(struct gfs2_holder *gh) if (unlikely(test_bit(SDF_SHUTDOWN, &sdp->sd_flags))) return -EIO; + if (test_bit(GLF_LRU, &gl->gl_flags)) + gfs2_glock_remove_from_lru(gl); + spin_lock(&gl->gl_spin); add_to_queue(gh); - if ((LM_FLAG_NOEXP & gh->gh_flags) && - test_and_clear_bit(GLF_FROZEN, &gl->gl_flags)) + if (unlikely((LM_FLAG_NOEXP & gh->gh_flags) && + test_and_clear_bit(GLF_FROZEN, &gl->gl_flags))) { set_bit(GLF_REPLY_PENDING, &gl->gl_flags); + gl->gl_lockref.count++; + if (queue_delayed_work(glock_workqueue, &gl->gl_work, 0) == 0) + gl->gl_lockref.count--; + } run_queue(gl, 1); spin_unlock(&gl->gl_spin); @@ -1097,7 +1090,7 @@ void gfs2_glock_dq(struct gfs2_holder *gh) spin_lock(&gl->gl_spin); if (gh->gh_flags & GL_NOCACHE) - handle_callback(gl, LM_ST_UNLOCKED, 0); + handle_callback(gl, LM_ST_UNLOCKED, 0, false); list_del_init(&gh->gh_list); if (find_first_holder(gl) == NULL) { @@ -1113,6 +1106,9 @@ void gfs2_glock_dq(struct gfs2_holder *gh) !test_bit(GLF_DEMOTE, &gl->gl_flags)) fast_path = 1; } + if (!test_bit(GLF_LFLUSH, &gl->gl_flags) && demote_ok(gl)) + gfs2_glock_add_to_lru(gl); + trace_gfs2_glock_queue(gh, 0); spin_unlock(&gl->gl_spin); if (likely(fast_path)) @@ -1120,8 +1116,9 @@ void gfs2_glock_dq(struct gfs2_holder *gh) gfs2_glock_hold(gl); if (test_bit(GLF_PENDING_DEMOTE, &gl->gl_flags) && - !test_bit(GLF_DEMOTE, &gl->gl_flags)) - delay = gl->gl_ops->go_min_hold_time; + !test_bit(GLF_DEMOTE, &gl->gl_flags) && + gl->gl_name.ln_type == LM_TYPE_INODE) + delay = gl->gl_hold_time; if (queue_delayed_work(glock_workqueue, &gl->gl_work, delay) == 0) gfs2_glock_put(gl); } @@ -1130,7 +1127,8 @@ void gfs2_glock_dq_wait(struct gfs2_holder *gh) { struct gfs2_glock *gl = gh->gh_gl; gfs2_glock_dq(gh); - wait_on_demote(gl); + might_sleep(); + wait_on_bit(&gl->gl_flags, GLF_DEMOTE, gfs2_glock_demote_wait, TASK_UNINTERRUPTIBLE); } /** @@ -1151,7 +1149,7 @@ void gfs2_glock_dq_uninit(struct gfs2_holder *gh) * @number: the lock number * @glops: the glock operations for the type of glock * @state: the state to acquire the glock in - * @flags: modifier flags for the aquisition + * @flags: modifier flags for the acquisition * @gh: the struct gfs2_holder * * Returns: errno @@ -1276,25 +1274,8 @@ int gfs2_glock_nq_m(unsigned int num_gh, struct gfs2_holder *ghs) void gfs2_glock_dq_m(unsigned int num_gh, struct gfs2_holder *ghs) { - unsigned int x; - - for (x = 0; x < num_gh; x++) - gfs2_glock_dq(&ghs[x]); -} - -/** - * gfs2_glock_dq_uninit_m - release multiple glocks - * @num_gh: the number of structures - * @ghs: an array of struct gfs2_holder structures - * - */ - -void gfs2_glock_dq_uninit_m(unsigned int num_gh, struct gfs2_holder *ghs) -{ - unsigned int x; - - for (x = 0; x < num_gh; x++) - gfs2_glock_dq_uninit(&ghs[x]); + while (num_gh--) + gfs2_glock_dq(&ghs[num_gh]); } void gfs2_glock_cb(struct gfs2_glock *gl, unsigned int state) @@ -1304,16 +1285,17 @@ void gfs2_glock_cb(struct gfs2_glock *gl, unsigned int state) unsigned long now = jiffies; gfs2_glock_hold(gl); - holdtime = gl->gl_tchange + gl->gl_ops->go_min_hold_time; - if (test_bit(GLF_QUEUED, &gl->gl_flags)) { + holdtime = gl->gl_tchange + gl->gl_hold_time; + if (test_bit(GLF_QUEUED, &gl->gl_flags) && + gl->gl_name.ln_type == LM_TYPE_INODE) { if (time_before(now, holdtime)) delay = holdtime - now; if (test_bit(GLF_REPLY_PENDING, &gl->gl_flags)) - delay = gl->gl_ops->go_min_hold_time; + delay = gl->gl_hold_time; } spin_lock(&gl->gl_spin); - handle_callback(gl, state, delay); + handle_callback(gl, state, delay, true); spin_unlock(&gl->gl_spin); if (queue_delayed_work(glock_workqueue, &gl->gl_work, delay) == 0) gfs2_glock_put(gl); @@ -1365,7 +1347,7 @@ void gfs2_glock_complete(struct gfs2_glock *gl, int ret) spin_lock(&gl->gl_spin); gl->gl_reply = ret; - if (unlikely(test_bit(DFL_BLOCK_LOCKS, &ls->ls_flags))) { + if (unlikely(test_bit(DFL_BLOCK_LOCKS, &ls->ls_recover_flags))) { if (gfs2_should_freeze(gl)) { set_bit(GLF_FROZEN, &gl->gl_flags); spin_unlock(&gl->gl_spin); @@ -1373,65 +1355,132 @@ void gfs2_glock_complete(struct gfs2_glock *gl, int ret) } } - spin_unlock(&gl->gl_spin); + gl->gl_lockref.count++; set_bit(GLF_REPLY_PENDING, &gl->gl_flags); - smp_wmb(); - gfs2_glock_hold(gl); + spin_unlock(&gl->gl_spin); + if (queue_delayed_work(glock_workqueue, &gl->gl_work, 0) == 0) gfs2_glock_put(gl); } +static int glock_cmp(void *priv, struct list_head *a, struct list_head *b) +{ + struct gfs2_glock *gla, *glb; + + gla = list_entry(a, struct gfs2_glock, gl_lru); + glb = list_entry(b, struct gfs2_glock, gl_lru); + + if (gla->gl_name.ln_number > glb->gl_name.ln_number) + return 1; + if (gla->gl_name.ln_number < glb->gl_name.ln_number) + return -1; -static int gfs2_shrink_glock_memory(struct shrinker *shrink, int nr, gfp_t gfp_mask) + return 0; +} + +/** + * gfs2_dispose_glock_lru - Demote a list of glocks + * @list: The list to dispose of + * + * Disposing of glocks may involve disk accesses, so that here we sort + * the glocks by number (i.e. disk location of the inodes) so that if + * there are any such accesses, they'll be sent in order (mostly). + * + * Must be called under the lru_lock, but may drop and retake this + * lock. While the lru_lock is dropped, entries may vanish from the + * list, but no new entries will appear on the list (since it is + * private) + */ + +static void gfs2_dispose_glock_lru(struct list_head *list) +__releases(&lru_lock) +__acquires(&lru_lock) { struct gfs2_glock *gl; - int may_demote; - int nr_skipped = 0; - LIST_HEAD(skipped); - if (nr == 0) - goto out; + list_sort(NULL, list, glock_cmp); - if (!(gfp_mask & __GFP_FS)) - return -1; + while(!list_empty(list)) { + gl = list_entry(list->next, struct gfs2_glock, gl_lru); + list_del_init(&gl->gl_lru); + if (!spin_trylock(&gl->gl_spin)) { +add_back_to_lru: + list_add(&gl->gl_lru, &lru_list); + atomic_inc(&lru_count); + continue; + } + if (test_and_set_bit(GLF_LOCK, &gl->gl_flags)) { + spin_unlock(&gl->gl_spin); + goto add_back_to_lru; + } + clear_bit(GLF_LRU, &gl->gl_flags); + gl->gl_lockref.count++; + if (demote_ok(gl)) + handle_callback(gl, LM_ST_UNLOCKED, 0, false); + WARN_ON(!test_and_clear_bit(GLF_LOCK, &gl->gl_flags)); + if (queue_delayed_work(glock_workqueue, &gl->gl_work, 0) == 0) + gl->gl_lockref.count--; + spin_unlock(&gl->gl_spin); + cond_resched_lock(&lru_lock); + } +} + +/** + * gfs2_scan_glock_lru - Scan the LRU looking for locks to demote + * @nr: The number of entries to scan + * + * This function selects the entries on the LRU which are able to + * be demoted, and then kicks off the process by calling + * gfs2_dispose_glock_lru() above. + */ + +static long gfs2_scan_glock_lru(int nr) +{ + struct gfs2_glock *gl; + LIST_HEAD(skipped); + LIST_HEAD(dispose); + long freed = 0; spin_lock(&lru_lock); - while(nr && !list_empty(&lru_list)) { + while ((nr-- >= 0) && !list_empty(&lru_list)) { gl = list_entry(lru_list.next, struct gfs2_glock, gl_lru); - list_del_init(&gl->gl_lru); - atomic_dec(&lru_count); /* Test for being demotable */ - if (!test_and_set_bit(GLF_LOCK, &gl->gl_flags)) { - gfs2_glock_hold(gl); - spin_unlock(&lru_lock); - spin_lock(&gl->gl_spin); - may_demote = demote_ok(gl); - if (may_demote) { - handle_callback(gl, LM_ST_UNLOCKED, 0); - nr--; - } - clear_bit(GLF_LOCK, &gl->gl_flags); - smp_mb__after_clear_bit(); - if (queue_delayed_work(glock_workqueue, &gl->gl_work, 0) == 0) - gfs2_glock_put_nolock(gl); - spin_unlock(&gl->gl_spin); - spin_lock(&lru_lock); + if (!test_bit(GLF_LOCK, &gl->gl_flags)) { + list_move(&gl->gl_lru, &dispose); + atomic_dec(&lru_count); + freed++; continue; } - nr_skipped++; - list_add(&gl->gl_lru, &skipped); + + list_move(&gl->gl_lru, &skipped); } list_splice(&skipped, &lru_list); - atomic_add(nr_skipped, &lru_count); + if (!list_empty(&dispose)) + gfs2_dispose_glock_lru(&dispose); spin_unlock(&lru_lock); -out: - return (atomic_read(&lru_count) / 100) * sysctl_vfs_cache_pressure; + + return freed; +} + +static unsigned long gfs2_glock_shrink_scan(struct shrinker *shrink, + struct shrink_control *sc) +{ + if (!(sc->gfp_mask & __GFP_FS)) + return SHRINK_STOP; + return gfs2_scan_glock_lru(sc->nr_to_scan); +} + +static unsigned long gfs2_glock_shrink_count(struct shrinker *shrink, + struct shrink_control *sc) +{ + return vfs_pressure_ratio(atomic_read(&lru_count)); } static struct shrinker glock_shrinker = { - .shrink = gfs2_shrink_glock_memory, .seeks = DEFAULT_SEEKS, + .count_objects = gfs2_glock_shrink_count, + .scan_objects = gfs2_glock_shrink_scan, }; /** @@ -1440,42 +1489,30 @@ static struct shrinker glock_shrinker = { * @sdp: the filesystem * @bucket: the bucket * - * Returns: 1 if the bucket has entries */ -static int examine_bucket(glock_examiner examiner, struct gfs2_sbd *sdp, +static void examine_bucket(glock_examiner examiner, const struct gfs2_sbd *sdp, unsigned int hash) { - struct gfs2_glock *gl, *prev = NULL; - int has_entries = 0; - struct hlist_head *head = &gl_hash_table[hash].hb_list; + struct gfs2_glock *gl; + struct hlist_bl_head *head = &gl_hash_table[hash]; + struct hlist_bl_node *pos; - read_lock(gl_lock_addr(hash)); - /* Can't use hlist_for_each_entry - don't want prefetch here */ - if (hlist_empty(head)) - goto out; - gl = list_entry(head->first, struct gfs2_glock, gl_list); - while(1) { - if (!sdp || gl->gl_sbd == sdp) { - gfs2_glock_hold(gl); - read_unlock(gl_lock_addr(hash)); - if (prev) - gfs2_glock_put(prev); - prev = gl; + rcu_read_lock(); + hlist_bl_for_each_entry_rcu(gl, pos, head, gl_list) { + if ((gl->gl_sbd == sdp) && lockref_get_not_dead(&gl->gl_lockref)) examiner(gl); - has_entries = 1; - read_lock(gl_lock_addr(hash)); - } - if (gl->gl_list.next == NULL) - break; - gl = list_entry(gl->gl_list.next, struct gfs2_glock, gl_list); } -out: - read_unlock(gl_lock_addr(hash)); - if (prev) - gfs2_glock_put(prev); + rcu_read_unlock(); cond_resched(); - return has_entries; +} + +static void glock_hash_walk(glock_examiner examiner, const struct gfs2_sbd *sdp) +{ + unsigned x; + + for (x = 0; x < GFS2_GL_HASH_SIZE; x++) + examine_bucket(examiner, sdp, x); } @@ -1483,18 +1520,17 @@ out: * thaw_glock - thaw out a glock which has an unprocessed reply waiting * @gl: The glock to thaw * - * N.B. When we freeze a glock, we leave a ref to the glock outstanding, - * so this has to result in the ref count being dropped by one. */ static void thaw_glock(struct gfs2_glock *gl) { if (!test_and_clear_bit(GLF_FROZEN, &gl->gl_flags)) - return; + goto out; set_bit(GLF_REPLY_PENDING, &gl->gl_flags); - gfs2_glock_hold(gl); - if (queue_delayed_work(glock_workqueue, &gl->gl_work, 0) == 0) + if (queue_delayed_work(glock_workqueue, &gl->gl_work, 0) == 0) { +out: gfs2_glock_put(gl); + } } /** @@ -1505,18 +1541,12 @@ static void thaw_glock(struct gfs2_glock *gl) static void clear_glock(struct gfs2_glock *gl) { - spin_lock(&lru_lock); - if (!list_empty(&gl->gl_lru)) { - list_del_init(&gl->gl_lru); - atomic_dec(&lru_count); - } - spin_unlock(&lru_lock); + gfs2_glock_remove_from_lru(gl); spin_lock(&gl->gl_spin); if (gl->gl_state != LM_ST_UNLOCKED) - handle_callback(gl, LM_ST_UNLOCKED, 0); + handle_callback(gl, LM_ST_UNLOCKED, 0, false); spin_unlock(&gl->gl_spin); - gfs2_glock_hold(gl); if (queue_delayed_work(glock_workqueue, &gl->gl_work, 0) == 0) gfs2_glock_put(gl); } @@ -1529,10 +1559,19 @@ static void clear_glock(struct gfs2_glock *gl) void gfs2_glock_thaw(struct gfs2_sbd *sdp) { - unsigned x; + glock_hash_walk(thaw_glock, sdp); +} - for (x = 0; x < GFS2_GL_HASH_SIZE; x++) - examine_bucket(thaw_glock, sdp, x); +static void dump_glock(struct seq_file *seq, struct gfs2_glock *gl) +{ + spin_lock(&gl->gl_spin); + gfs2_dump_glock(seq, gl); + spin_unlock(&gl->gl_spin); +} + +static void dump_glock_func(struct gfs2_glock *gl) +{ + dump_glock(NULL, gl); } /** @@ -1545,13 +1584,12 @@ void gfs2_glock_thaw(struct gfs2_sbd *sdp) void gfs2_gl_hash_clear(struct gfs2_sbd *sdp) { - unsigned int x; - - for (x = 0; x < GFS2_GL_HASH_SIZE; x++) - examine_bucket(clear_glock, sdp, x); + set_bit(SDF_SKIP_DLM_UNLOCK, &sdp->sd_flags); + flush_workqueue(glock_workqueue); + glock_hash_walk(clear_glock, sdp); flush_workqueue(glock_workqueue); wait_event(sdp->sd_glock_wait, atomic_read(&sdp->sd_glock_disposal) == 0); - gfs2_dump_lockstate(sdp); + glock_hash_walk(dump_glock_func, sdp); } void gfs2_glock_finish_truncate(struct gfs2_inode *ip) @@ -1617,14 +1655,14 @@ static const char *hflags2str(char *buf, unsigned flags, unsigned long iflags) * @seq: the seq_file struct * @gh: the glock holder * - * Returns: 0 on success, -ENOBUFS when we run out of space */ -static int dump_holder(struct seq_file *seq, const struct gfs2_holder *gh) +static void dump_holder(struct seq_file *seq, const struct gfs2_holder *gh) { struct task_struct *gh_owner = NULL; char flags_buf[32]; + rcu_read_lock(); if (gh->gh_owner_pid) gh_owner = pid_task(gh->gh_owner_pid, PIDTYPE_PID); gfs2_print_dbg(seq, " H: s:%s f:%s e:%d p:%ld [%s] %pS\n", @@ -1634,12 +1672,14 @@ static int dump_holder(struct seq_file *seq, const struct gfs2_holder *gh) gh->gh_owner_pid ? (long)pid_nr(gh->gh_owner_pid) : -1, gh_owner ? gh_owner->comm : "(ended)", (void *)gh->gh_ip); - return 0; + rcu_read_unlock(); } -static const char *gflags2str(char *buf, const unsigned long *gflags) +static const char *gflags2str(char *buf, const struct gfs2_glock *gl) { + const unsigned long *gflags = &gl->gl_flags; char *p = buf; + if (test_bit(GLF_LOCK, gflags)) *p++ = 'l'; if (test_bit(GLF_DEMOTE, gflags)) @@ -1662,12 +1702,18 @@ static const char *gflags2str(char *buf, const unsigned long *gflags) *p++ = 'F'; if (test_bit(GLF_QUEUED, gflags)) *p++ = 'q'; + if (test_bit(GLF_LRU, gflags)) + *p++ = 'L'; + if (gl->gl_object) + *p++ = 'o'; + if (test_bit(GLF_BLOCKING, gflags)) + *p++ = 'b'; *p = 0; return buf; } /** - * __dump_glock - print information about a glock + * gfs2_dump_glock - print information about a glock * @seq: The seq_file struct * @gl: the glock * @@ -1681,113 +1727,127 @@ static const char *gflags2str(char *buf, const unsigned long *gflags) * example. The field's are n = number (id of the object), f = flags, * t = type, s = state, r = refcount, e = error, p = pid. * - * Returns: 0 on success, -ENOBUFS when we run out of space */ -static int __dump_glock(struct seq_file *seq, const struct gfs2_glock *gl) +void gfs2_dump_glock(struct seq_file *seq, const struct gfs2_glock *gl) { const struct gfs2_glock_operations *glops = gl->gl_ops; unsigned long long dtime; const struct gfs2_holder *gh; char gflags_buf[32]; - int error = 0; dtime = jiffies - gl->gl_demote_time; dtime *= 1000000/HZ; /* demote time in uSec */ if (!test_bit(GLF_DEMOTE, &gl->gl_flags)) dtime = 0; - gfs2_print_dbg(seq, "G: s:%s n:%u/%llx f:%s t:%s d:%s/%llu a:%d r:%d\n", + gfs2_print_dbg(seq, "G: s:%s n:%u/%llx f:%s t:%s d:%s/%llu a:%d v:%d r:%d m:%ld\n", state2str(gl->gl_state), gl->gl_name.ln_type, (unsigned long long)gl->gl_name.ln_number, - gflags2str(gflags_buf, &gl->gl_flags), + gflags2str(gflags_buf, gl), state2str(gl->gl_target), state2str(gl->gl_demote_state), dtime, atomic_read(&gl->gl_ail_count), - atomic_read(&gl->gl_ref)); + atomic_read(&gl->gl_revokes), + (int)gl->gl_lockref.count, gl->gl_hold_time); + + list_for_each_entry(gh, &gl->gl_holders, gh_list) + dump_holder(seq, gh); - list_for_each_entry(gh, &gl->gl_holders, gh_list) { - error = dump_holder(seq, gh); - if (error) - goto out; - } if (gl->gl_state != LM_ST_UNLOCKED && glops->go_dump) - error = glops->go_dump(seq, gl); -out: - return error; + glops->go_dump(seq, gl); } -static int dump_glock(struct seq_file *seq, struct gfs2_glock *gl) +static int gfs2_glstats_seq_show(struct seq_file *seq, void *iter_ptr) { - int ret; - spin_lock(&gl->gl_spin); - ret = __dump_glock(seq, gl); - spin_unlock(&gl->gl_spin); - return ret; -} + struct gfs2_glock *gl = iter_ptr; -/** - * gfs2_dump_lockstate - print out the current lockstate - * @sdp: the filesystem - * @ub: the buffer to copy the information into - * - * If @ub is NULL, dump the lockstate to the console. - * - */ + seq_printf(seq, "G: n:%u/%llx rtt:%lld/%lld rttb:%lld/%lld irt:%lld/%lld dcnt: %lld qcnt: %lld\n", + gl->gl_name.ln_type, + (unsigned long long)gl->gl_name.ln_number, + (long long)gl->gl_stats.stats[GFS2_LKS_SRTT], + (long long)gl->gl_stats.stats[GFS2_LKS_SRTTVAR], + (long long)gl->gl_stats.stats[GFS2_LKS_SRTTB], + (long long)gl->gl_stats.stats[GFS2_LKS_SRTTVARB], + (long long)gl->gl_stats.stats[GFS2_LKS_SIRT], + (long long)gl->gl_stats.stats[GFS2_LKS_SIRTVAR], + (long long)gl->gl_stats.stats[GFS2_LKS_DCOUNT], + (long long)gl->gl_stats.stats[GFS2_LKS_QCOUNT]); + return 0; +} -static int gfs2_dump_lockstate(struct gfs2_sbd *sdp) -{ - struct gfs2_glock *gl; - struct hlist_node *h; - unsigned int x; - int error = 0; +static const char *gfs2_gltype[] = { + "type", + "reserved", + "nondisk", + "inode", + "rgrp", + "meta", + "iopen", + "flock", + "plock", + "quota", + "journal", +}; - for (x = 0; x < GFS2_GL_HASH_SIZE; x++) { +static const char *gfs2_stype[] = { + [GFS2_LKS_SRTT] = "srtt", + [GFS2_LKS_SRTTVAR] = "srttvar", + [GFS2_LKS_SRTTB] = "srttb", + [GFS2_LKS_SRTTVARB] = "srttvarb", + [GFS2_LKS_SIRT] = "sirt", + [GFS2_LKS_SIRTVAR] = "sirtvar", + [GFS2_LKS_DCOUNT] = "dlm", + [GFS2_LKS_QCOUNT] = "queue", +}; - read_lock(gl_lock_addr(x)); +#define GFS2_NR_SBSTATS (ARRAY_SIZE(gfs2_gltype) * ARRAY_SIZE(gfs2_stype)) - hlist_for_each_entry(gl, h, &gl_hash_table[x].hb_list, gl_list) { - if (gl->gl_sbd != sdp) - continue; +static int gfs2_sbstats_seq_show(struct seq_file *seq, void *iter_ptr) +{ + struct gfs2_glock_iter *gi = seq->private; + struct gfs2_sbd *sdp = gi->sdp; + unsigned index = gi->hash >> 3; + unsigned subindex = gi->hash & 0x07; + s64 value; + int i; - error = dump_glock(NULL, gl); - if (error) - break; - } + if (index == 0 && subindex != 0) + return 0; - read_unlock(gl_lock_addr(x)); + seq_printf(seq, "%-10s %8s:", gfs2_gltype[index], + (index == 0) ? "cpu": gfs2_stype[subindex]); - if (error) - break; + for_each_possible_cpu(i) { + const struct gfs2_pcpu_lkstats *lkstats = per_cpu_ptr(sdp->sd_lkstats, i); + if (index == 0) { + value = i; + } else { + value = lkstats->lkstats[index - 1].stats[subindex]; + } + seq_printf(seq, " %15lld", (long long)value); } - - - return error; + seq_putc(seq, '\n'); + return 0; } - int __init gfs2_glock_init(void) { unsigned i; for(i = 0; i < GFS2_GL_HASH_SIZE; i++) { - INIT_HLIST_HEAD(&gl_hash_table[i].hb_list); - } -#ifdef GL_HASH_LOCK_SZ - for(i = 0; i < GL_HASH_LOCK_SZ; i++) { - rwlock_init(&gl_hash_locks[i]); + INIT_HLIST_BL_HEAD(&gl_hash_table[i]); } -#endif glock_workqueue = alloc_workqueue("glock_workqueue", WQ_MEM_RECLAIM | WQ_HIGHPRI | WQ_FREEZABLE, 0); - if (IS_ERR(glock_workqueue)) - return PTR_ERR(glock_workqueue); + if (!glock_workqueue) + return -ENOMEM; gfs2_delete_workqueue = alloc_workqueue("delete_workqueue", WQ_MEM_RECLAIM | WQ_FREEZABLE, 0); - if (IS_ERR(gfs2_delete_workqueue)) { + if (!gfs2_delete_workqueue) { destroy_workqueue(glock_workqueue); - return PTR_ERR(gfs2_delete_workqueue); + return -ENOMEM; } register_shrinker(&glock_shrinker); @@ -1802,64 +1862,70 @@ void gfs2_glock_exit(void) destroy_workqueue(gfs2_delete_workqueue); } +static inline struct gfs2_glock *glock_hash_chain(unsigned hash) +{ + return hlist_bl_entry(hlist_bl_first_rcu(&gl_hash_table[hash]), + struct gfs2_glock, gl_list); +} + +static inline struct gfs2_glock *glock_hash_next(struct gfs2_glock *gl) +{ + return hlist_bl_entry(rcu_dereference(gl->gl_list.next), + struct gfs2_glock, gl_list); +} + static int gfs2_glock_iter_next(struct gfs2_glock_iter *gi) { struct gfs2_glock *gl; -restart: - read_lock(gl_lock_addr(gi->hash)); - gl = gi->gl; - if (gl) { - gi->gl = hlist_entry(gl->gl_list.next, - struct gfs2_glock, gl_list); - } else { - gi->gl = hlist_entry(gl_hash_table[gi->hash].hb_list.first, - struct gfs2_glock, gl_list); - } - if (gi->gl) - gfs2_glock_hold(gi->gl); - read_unlock(gl_lock_addr(gi->hash)); - if (gl) - gfs2_glock_put(gl); - while (gi->gl == NULL) { - gi->hash++; - if (gi->hash >= GFS2_GL_HASH_SIZE) - return 1; - read_lock(gl_lock_addr(gi->hash)); - gi->gl = hlist_entry(gl_hash_table[gi->hash].hb_list.first, - struct gfs2_glock, gl_list); - if (gi->gl) - gfs2_glock_hold(gi->gl); - read_unlock(gl_lock_addr(gi->hash)); - } - - if (gi->sdp != gi->gl->gl_sbd) - goto restart; + do { + gl = gi->gl; + if (gl) { + gi->gl = glock_hash_next(gl); + gi->nhash++; + } else { + if (gi->hash >= GFS2_GL_HASH_SIZE) { + rcu_read_unlock(); + return 1; + } + gi->gl = glock_hash_chain(gi->hash); + gi->nhash = 0; + } + while (gi->gl == NULL) { + gi->hash++; + if (gi->hash >= GFS2_GL_HASH_SIZE) { + rcu_read_unlock(); + return 1; + } + gi->gl = glock_hash_chain(gi->hash); + gi->nhash = 0; + } + /* Skip entries for other sb and dead entries */ + } while (gi->sdp != gi->gl->gl_sbd || + __lockref_is_dead(&gi->gl->gl_lockref)); return 0; } -static void gfs2_glock_iter_free(struct gfs2_glock_iter *gi) -{ - if (gi->gl) - gfs2_glock_put(gi->gl); - gi->gl = NULL; -} - static void *gfs2_glock_seq_start(struct seq_file *seq, loff_t *pos) { struct gfs2_glock_iter *gi = seq->private; loff_t n = *pos; - gi->hash = 0; + if (gi->last_pos <= *pos) + n = gi->nhash + (*pos - gi->last_pos); + else + gi->hash = 0; + + gi->nhash = 0; + rcu_read_lock(); do { - if (gfs2_glock_iter_next(gi)) { - gfs2_glock_iter_free(gi); + if (gfs2_glock_iter_next(gi)) return NULL; - } } while (n--); + gi->last_pos = *pos; return gi->gl; } @@ -1869,11 +1935,9 @@ static void *gfs2_glock_seq_next(struct seq_file *seq, void *iter_ptr, struct gfs2_glock_iter *gi = seq->private; (*pos)++; - - if (gfs2_glock_iter_next(gi)) { - gfs2_glock_iter_free(gi); + gi->last_pos = *pos; + if (gfs2_glock_iter_next(gi)) return NULL; - } return gi->gl; } @@ -1881,12 +1945,45 @@ static void *gfs2_glock_seq_next(struct seq_file *seq, void *iter_ptr, static void gfs2_glock_seq_stop(struct seq_file *seq, void *iter_ptr) { struct gfs2_glock_iter *gi = seq->private; - gfs2_glock_iter_free(gi); + + if (gi->gl) + rcu_read_unlock(); + gi->gl = NULL; } static int gfs2_glock_seq_show(struct seq_file *seq, void *iter_ptr) { - return dump_glock(seq, iter_ptr); + dump_glock(seq, iter_ptr); + return 0; +} + +static void *gfs2_sbstats_seq_start(struct seq_file *seq, loff_t *pos) +{ + struct gfs2_glock_iter *gi = seq->private; + + gi->hash = *pos; + if (*pos >= GFS2_NR_SBSTATS) + return NULL; + preempt_disable(); + return SEQ_START_TOKEN; +} + +static void *gfs2_sbstats_seq_next(struct seq_file *seq, void *iter_ptr, + loff_t *pos) +{ + struct gfs2_glock_iter *gi = seq->private; + (*pos)++; + gi->hash++; + if (gi->hash >= GFS2_NR_SBSTATS) { + preempt_enable(); + return NULL; + } + return SEQ_START_TOKEN; +} + +static void gfs2_sbstats_seq_stop(struct seq_file *seq, void *iter_ptr) +{ + preempt_enable(); } static const struct seq_operations gfs2_glock_seq_ops = { @@ -1896,7 +1993,23 @@ static const struct seq_operations gfs2_glock_seq_ops = { .show = gfs2_glock_seq_show, }; -static int gfs2_debugfs_open(struct inode *inode, struct file *file) +static const struct seq_operations gfs2_glstats_seq_ops = { + .start = gfs2_glock_seq_start, + .next = gfs2_glock_seq_next, + .stop = gfs2_glock_seq_stop, + .show = gfs2_glstats_seq_show, +}; + +static const struct seq_operations gfs2_sbstats_seq_ops = { + .start = gfs2_sbstats_seq_start, + .next = gfs2_sbstats_seq_next, + .stop = gfs2_sbstats_seq_stop, + .show = gfs2_sbstats_seq_show, +}; + +#define GFS2_SEQ_GOODSIZE min(PAGE_SIZE << PAGE_ALLOC_COSTLY_ORDER, 65536UL) + +static int gfs2_glocks_open(struct inode *inode, struct file *file) { int ret = seq_open_private(file, &gfs2_glock_seq_ops, sizeof(struct gfs2_glock_iter)); @@ -1904,13 +2017,59 @@ static int gfs2_debugfs_open(struct inode *inode, struct file *file) struct seq_file *seq = file->private_data; struct gfs2_glock_iter *gi = seq->private; gi->sdp = inode->i_private; + seq->buf = kmalloc(GFS2_SEQ_GOODSIZE, GFP_KERNEL | __GFP_NOWARN); + if (seq->buf) + seq->size = GFS2_SEQ_GOODSIZE; } return ret; } -static const struct file_operations gfs2_debug_fops = { +static int gfs2_glstats_open(struct inode *inode, struct file *file) +{ + int ret = seq_open_private(file, &gfs2_glstats_seq_ops, + sizeof(struct gfs2_glock_iter)); + if (ret == 0) { + struct seq_file *seq = file->private_data; + struct gfs2_glock_iter *gi = seq->private; + gi->sdp = inode->i_private; + seq->buf = kmalloc(GFS2_SEQ_GOODSIZE, GFP_KERNEL | __GFP_NOWARN); + if (seq->buf) + seq->size = GFS2_SEQ_GOODSIZE; + } + return ret; +} + +static int gfs2_sbstats_open(struct inode *inode, struct file *file) +{ + int ret = seq_open_private(file, &gfs2_sbstats_seq_ops, + sizeof(struct gfs2_glock_iter)); + if (ret == 0) { + struct seq_file *seq = file->private_data; + struct gfs2_glock_iter *gi = seq->private; + gi->sdp = inode->i_private; + } + return ret; +} + +static const struct file_operations gfs2_glocks_fops = { + .owner = THIS_MODULE, + .open = gfs2_glocks_open, + .read = seq_read, + .llseek = seq_lseek, + .release = seq_release_private, +}; + +static const struct file_operations gfs2_glstats_fops = { .owner = THIS_MODULE, - .open = gfs2_debugfs_open, + .open = gfs2_glstats_open, + .read = seq_read, + .llseek = seq_lseek, + .release = seq_release_private, +}; + +static const struct file_operations gfs2_sbstats_fops = { + .owner = THIS_MODULE, + .open = gfs2_sbstats_open, .read = seq_read, .llseek = seq_lseek, .release = seq_release_private, @@ -1924,20 +2083,45 @@ int gfs2_create_debugfs_file(struct gfs2_sbd *sdp) sdp->debugfs_dentry_glocks = debugfs_create_file("glocks", S_IFREG | S_IRUGO, sdp->debugfs_dir, sdp, - &gfs2_debug_fops); + &gfs2_glocks_fops); if (!sdp->debugfs_dentry_glocks) - return -ENOMEM; + goto fail; + + sdp->debugfs_dentry_glstats = debugfs_create_file("glstats", + S_IFREG | S_IRUGO, + sdp->debugfs_dir, sdp, + &gfs2_glstats_fops); + if (!sdp->debugfs_dentry_glstats) + goto fail; + + sdp->debugfs_dentry_sbstats = debugfs_create_file("sbstats", + S_IFREG | S_IRUGO, + sdp->debugfs_dir, sdp, + &gfs2_sbstats_fops); + if (!sdp->debugfs_dentry_sbstats) + goto fail; return 0; +fail: + gfs2_delete_debugfs_file(sdp); + return -ENOMEM; } void gfs2_delete_debugfs_file(struct gfs2_sbd *sdp) { - if (sdp && sdp->debugfs_dir) { + if (sdp->debugfs_dir) { if (sdp->debugfs_dentry_glocks) { debugfs_remove(sdp->debugfs_dentry_glocks); sdp->debugfs_dentry_glocks = NULL; } + if (sdp->debugfs_dentry_glstats) { + debugfs_remove(sdp->debugfs_dentry_glstats); + sdp->debugfs_dentry_glstats = NULL; + } + if (sdp->debugfs_dentry_sbstats) { + debugfs_remove(sdp->debugfs_dentry_sbstats); + sdp->debugfs_dentry_sbstats = NULL; + } debugfs_remove(sdp->debugfs_dir); sdp->debugfs_dir = NULL; } diff --git a/fs/gfs2/glock.h b/fs/gfs2/glock.h index 691851ceb61..32572f71f02 100644 --- a/fs/gfs2/glock.h +++ b/fs/gfs2/glock.h @@ -113,12 +113,21 @@ enum { #define GLR_TRYFAILED 13 +#define GL_GLOCK_MAX_HOLD (long)(HZ / 5) +#define GL_GLOCK_DFT_HOLD (long)(HZ / 5) +#define GL_GLOCK_MIN_HOLD (long)(10) +#define GL_GLOCK_HOLD_INCR (long)(HZ / 20) +#define GL_GLOCK_HOLD_DECR (long)(HZ / 40) + struct lm_lockops { const char *lm_proto_name; - int (*lm_mount) (struct gfs2_sbd *sdp, const char *fsname); - void (*lm_unmount) (struct gfs2_sbd *sdp); + int (*lm_mount) (struct gfs2_sbd *sdp, const char *table); + void (*lm_first_done) (struct gfs2_sbd *sdp); + void (*lm_recovery_result) (struct gfs2_sbd *sdp, unsigned int jid, + unsigned int result); + void (*lm_unmount) (struct gfs2_sbd *sdp); void (*lm_withdraw) (struct gfs2_sbd *sdp); - void (*lm_put_lock) (struct kmem_cache *cachep, struct gfs2_glock *gl); + void (*lm_put_lock) (struct gfs2_glock *gl); int (*lm_lock) (struct gfs2_glock *gl, unsigned int req_state, unsigned int flags); void (*lm_cancel) (struct gfs2_glock *gl); @@ -169,33 +178,30 @@ static inline struct address_space *gfs2_glock2aspace(struct gfs2_glock *gl) return NULL; } -int gfs2_glock_get(struct gfs2_sbd *sdp, - u64 number, const struct gfs2_glock_operations *glops, - int create, struct gfs2_glock **glp); -void gfs2_glock_hold(struct gfs2_glock *gl); -void gfs2_glock_put_nolock(struct gfs2_glock *gl); -int gfs2_glock_put(struct gfs2_glock *gl); -void gfs2_holder_init(struct gfs2_glock *gl, unsigned int state, unsigned flags, - struct gfs2_holder *gh); -void gfs2_holder_reinit(unsigned int state, unsigned flags, - struct gfs2_holder *gh); -void gfs2_holder_uninit(struct gfs2_holder *gh); -int gfs2_glock_nq(struct gfs2_holder *gh); -int gfs2_glock_poll(struct gfs2_holder *gh); -int gfs2_glock_wait(struct gfs2_holder *gh); -void gfs2_glock_dq(struct gfs2_holder *gh); -void gfs2_glock_dq_wait(struct gfs2_holder *gh); - -void gfs2_glock_dq_uninit(struct gfs2_holder *gh); -int gfs2_glock_nq_num(struct gfs2_sbd *sdp, - u64 number, const struct gfs2_glock_operations *glops, - unsigned int state, int flags, struct gfs2_holder *gh); - -int gfs2_glock_nq_m(unsigned int num_gh, struct gfs2_holder *ghs); -void gfs2_glock_dq_m(unsigned int num_gh, struct gfs2_holder *ghs); -void gfs2_glock_dq_uninit_m(unsigned int num_gh, struct gfs2_holder *ghs); - -__attribute__ ((format(printf, 2, 3))) +extern int gfs2_glock_get(struct gfs2_sbd *sdp, u64 number, + const struct gfs2_glock_operations *glops, + int create, struct gfs2_glock **glp); +extern void gfs2_glock_put(struct gfs2_glock *gl); +extern void gfs2_holder_init(struct gfs2_glock *gl, unsigned int state, + unsigned flags, struct gfs2_holder *gh); +extern void gfs2_holder_reinit(unsigned int state, unsigned flags, + struct gfs2_holder *gh); +extern void gfs2_holder_uninit(struct gfs2_holder *gh); +extern int gfs2_glock_nq(struct gfs2_holder *gh); +extern int gfs2_glock_poll(struct gfs2_holder *gh); +extern int gfs2_glock_wait(struct gfs2_holder *gh); +extern void gfs2_glock_dq(struct gfs2_holder *gh); +extern void gfs2_glock_dq_wait(struct gfs2_holder *gh); +extern void gfs2_glock_dq_uninit(struct gfs2_holder *gh); +extern int gfs2_glock_nq_num(struct gfs2_sbd *sdp, u64 number, + const struct gfs2_glock_operations *glops, + unsigned int state, int flags, + struct gfs2_holder *gh); +extern int gfs2_glock_nq_m(unsigned int num_gh, struct gfs2_holder *ghs); +extern void gfs2_glock_dq_m(unsigned int num_gh, struct gfs2_holder *ghs); +extern void gfs2_dump_glock(struct seq_file *seq, const struct gfs2_glock *gl); +#define GLOCK_BUG_ON(gl,x) do { if (unlikely(x)) { gfs2_dump_glock(NULL, gl); BUG(); } } while(0) +extern __printf(2, 3) void gfs2_print_dbg(struct seq_file *seq, const char *fmt, ...); /** @@ -223,25 +229,21 @@ static inline int gfs2_glock_nq_init(struct gfs2_glock *gl, return error; } -/* Lock Value Block functions */ - -int gfs2_lvb_hold(struct gfs2_glock *gl); -void gfs2_lvb_unhold(struct gfs2_glock *gl); - -void gfs2_glock_cb(struct gfs2_glock *gl, unsigned int state); -void gfs2_glock_complete(struct gfs2_glock *gl, int ret); -void gfs2_reclaim_glock(struct gfs2_sbd *sdp); -void gfs2_gl_hash_clear(struct gfs2_sbd *sdp); -void gfs2_glock_finish_truncate(struct gfs2_inode *ip); -void gfs2_glock_thaw(struct gfs2_sbd *sdp); - -int __init gfs2_glock_init(void); -void gfs2_glock_exit(void); - -int gfs2_create_debugfs_file(struct gfs2_sbd *sdp); -void gfs2_delete_debugfs_file(struct gfs2_sbd *sdp); -int gfs2_register_debugfs(void); -void gfs2_unregister_debugfs(void); +extern void gfs2_glock_cb(struct gfs2_glock *gl, unsigned int state); +extern void gfs2_glock_complete(struct gfs2_glock *gl, int ret); +extern void gfs2_gl_hash_clear(struct gfs2_sbd *sdp); +extern void gfs2_glock_finish_truncate(struct gfs2_inode *ip); +extern void gfs2_glock_thaw(struct gfs2_sbd *sdp); +extern void gfs2_glock_add_to_lru(struct gfs2_glock *gl); +extern void gfs2_glock_free(struct gfs2_glock *gl); + +extern int __init gfs2_glock_init(void); +extern void gfs2_glock_exit(void); + +extern int gfs2_create_debugfs_file(struct gfs2_sbd *sdp); +extern void gfs2_delete_debugfs_file(struct gfs2_sbd *sdp); +extern int gfs2_register_debugfs(void); +extern void gfs2_unregister_debugfs(void); extern const struct lm_lockops gfs2_dlm_ops; diff --git a/fs/gfs2/glops.c b/fs/gfs2/glops.c index 263561bf1a5..2ffc67dce87 100644 --- a/fs/gfs2/glops.c +++ b/fs/gfs2/glops.c @@ -26,53 +26,107 @@ #include "rgrp.h" #include "util.h" #include "trans.h" +#include "dir.h" + +static void gfs2_ail_error(struct gfs2_glock *gl, const struct buffer_head *bh) +{ + fs_err(gl->gl_sbd, "AIL buffer %p: blocknr %llu state 0x%08lx mapping %p page state 0x%lx\n", + bh, (unsigned long long)bh->b_blocknr, bh->b_state, + bh->b_page->mapping, bh->b_page->flags); + fs_err(gl->gl_sbd, "AIL glock %u:%llu mapping %p\n", + gl->gl_name.ln_type, gl->gl_name.ln_number, + gfs2_glock2aspace(gl)); + gfs2_lm_withdraw(gl->gl_sbd, "AIL error\n"); +} /** - * ail_empty_gl - remove all buffers for a given lock from the AIL + * __gfs2_ail_flush - remove all buffers for a given lock from the AIL * @gl: the glock + * @fsync: set when called from fsync (not all buffers will be clean) * * None of the buffers should be dirty, locked, or pinned. */ -static void gfs2_ail_empty_gl(struct gfs2_glock *gl) +static void __gfs2_ail_flush(struct gfs2_glock *gl, bool fsync, + unsigned int nr_revokes) { struct gfs2_sbd *sdp = gl->gl_sbd; struct list_head *head = &gl->gl_ail_list; - struct gfs2_bufdata *bd; + struct gfs2_bufdata *bd, *tmp; struct buffer_head *bh; + const unsigned long b_state = (1UL << BH_Dirty)|(1UL << BH_Pinned)|(1UL << BH_Lock); + + gfs2_log_lock(sdp); + spin_lock(&sdp->sd_ail_lock); + list_for_each_entry_safe_reverse(bd, tmp, head, bd_ail_gl_list) { + if (nr_revokes == 0) + break; + bh = bd->bd_bh; + if (bh->b_state & b_state) { + if (fsync) + continue; + gfs2_ail_error(gl, bh); + } + gfs2_trans_add_revoke(sdp, bd); + nr_revokes--; + } + GLOCK_BUG_ON(gl, !fsync && atomic_read(&gl->gl_ail_count)); + spin_unlock(&sdp->sd_ail_lock); + gfs2_log_unlock(sdp); +} + + +static void gfs2_ail_empty_gl(struct gfs2_glock *gl) +{ + struct gfs2_sbd *sdp = gl->gl_sbd; struct gfs2_trans tr; memset(&tr, 0, sizeof(tr)); + INIT_LIST_HEAD(&tr.tr_buf); + INIT_LIST_HEAD(&tr.tr_databuf); tr.tr_revokes = atomic_read(&gl->gl_ail_count); if (!tr.tr_revokes) return; - /* A shortened, inline version of gfs2_trans_begin() */ + /* A shortened, inline version of gfs2_trans_begin() + * tr->alloced is not set since the transaction structure is + * on the stack */ tr.tr_reserved = 1 + gfs2_struct2blk(sdp, tr.tr_revokes, sizeof(u64)); tr.tr_ip = (unsigned long)__builtin_return_address(0); - INIT_LIST_HEAD(&tr.tr_list_buf); - gfs2_log_reserve(sdp, tr.tr_reserved); - BUG_ON(current->journal_info); + sb_start_intwrite(sdp->sd_vfs); + if (gfs2_log_reserve(sdp, tr.tr_reserved) < 0) { + sb_end_intwrite(sdp->sd_vfs); + return; + } + WARN_ON_ONCE(current->journal_info); current->journal_info = &tr; - gfs2_log_lock(sdp); - while (!list_empty(head)) { - bd = list_entry(head->next, struct gfs2_bufdata, - bd_ail_gl_list); - bh = bd->bd_bh; - gfs2_remove_from_ail(bd); - bd->bd_bh = NULL; - bh->b_private = NULL; - bd->bd_blkno = bh->b_blocknr; - gfs2_assert_withdraw(sdp, !buffer_busy(bh)); - gfs2_trans_add_revoke(sdp, bd); - } - gfs2_assert_withdraw(sdp, !atomic_read(&gl->gl_ail_count)); - gfs2_log_unlock(sdp); + __gfs2_ail_flush(gl, 0, tr.tr_revokes); + + gfs2_trans_end(sdp); + gfs2_log_flush(sdp, NULL, NORMAL_FLUSH); +} + +void gfs2_ail_flush(struct gfs2_glock *gl, bool fsync) +{ + struct gfs2_sbd *sdp = gl->gl_sbd; + unsigned int revokes = atomic_read(&gl->gl_ail_count); + unsigned int max_revokes = (sdp->sd_sb.sb_bsize - sizeof(struct gfs2_log_descriptor)) / sizeof(u64); + int ret; + + if (!revokes) + return; + while (revokes > max_revokes) + max_revokes += (sdp->sd_sb.sb_bsize - sizeof(struct gfs2_meta_header)) / sizeof(u64); + + ret = gfs2_trans_begin(sdp, 0, max_revokes); + if (ret) + return; + __gfs2_ail_flush(gl, fsync, max_revokes); gfs2_trans_end(sdp); - gfs2_log_flush(sdp, NULL); + gfs2_log_flush(sdp, NULL, NORMAL_FLUSH); } /** @@ -86,18 +140,26 @@ static void gfs2_ail_empty_gl(struct gfs2_glock *gl) static void rgrp_go_sync(struct gfs2_glock *gl) { - struct address_space *metamapping = gfs2_glock2aspace(gl); + struct gfs2_sbd *sdp = gl->gl_sbd; + struct address_space *mapping = &sdp->sd_aspace; + struct gfs2_rgrpd *rgd; int error; if (!test_and_clear_bit(GLF_DIRTY, &gl->gl_flags)) return; - BUG_ON(gl->gl_state != LM_ST_EXCLUSIVE); + GLOCK_BUG_ON(gl, gl->gl_state != LM_ST_EXCLUSIVE); - gfs2_log_flush(gl->gl_sbd, gl); - filemap_fdatawrite(metamapping); - error = filemap_fdatawait(metamapping); - mapping_set_error(metamapping, error); + gfs2_log_flush(sdp, gl, NORMAL_FLUSH); + filemap_fdatawrite_range(mapping, gl->gl_vm.start, gl->gl_vm.end); + error = filemap_fdatawait_range(mapping, gl->gl_vm.start, gl->gl_vm.end); + mapping_set_error(mapping, error); gfs2_ail_empty_gl(gl); + + spin_lock(&gl->gl_spin); + rgd = gl->gl_object; + if (rgd) + gfs2_free_clones(rgd); + spin_unlock(&gl->gl_spin); } /** @@ -112,11 +174,12 @@ static void rgrp_go_sync(struct gfs2_glock *gl) static void rgrp_go_inval(struct gfs2_glock *gl, int flags) { - struct address_space *mapping = gfs2_glock2aspace(gl); + struct gfs2_sbd *sdp = gl->gl_sbd; + struct address_space *mapping = &sdp->sd_aspace; - BUG_ON(!(flags & DIO_METADATA)); - gfs2_assert_withdraw(gl->gl_sbd, !atomic_read(&gl->gl_ail_count)); - truncate_inode_pages(mapping, 0); + WARN_ON_ONCE(!(flags & DIO_METADATA)); + gfs2_assert_withdraw(sdp, !atomic_read(&gl->gl_ail_count)); + truncate_inode_pages_range(mapping, gl->gl_vm.start, gl->gl_vm.end); if (gl->gl_object) { struct gfs2_rgrpd *rgd = (struct gfs2_rgrpd *)gl->gl_object; @@ -138,14 +201,17 @@ static void inode_go_sync(struct gfs2_glock *gl) if (ip && !S_ISREG(ip->i_inode.i_mode)) ip = NULL; - if (ip && test_and_clear_bit(GIF_SW_PAGED, &ip->i_flags)) - unmap_shared_mapping_range(ip->i_inode.i_mapping, 0, 0); + if (ip) { + if (test_and_clear_bit(GIF_SW_PAGED, &ip->i_flags)) + unmap_shared_mapping_range(ip->i_inode.i_mapping, 0, 0); + inode_dio_wait(&ip->i_inode); + } if (!test_and_clear_bit(GLF_DIRTY, &gl->gl_flags)) return; - BUG_ON(gl->gl_state != LM_ST_EXCLUSIVE); + GLOCK_BUG_ON(gl, gl->gl_state != LM_ST_EXCLUSIVE); - gfs2_log_flush(gl->gl_sbd, gl); + gfs2_log_flush(gl->gl_sbd, gl, NORMAL_FLUSH); filemap_fdatawrite(metamapping); if (ip) { struct address_space *mapping = ip->i_inode.i_mapping; @@ -160,7 +226,7 @@ static void inode_go_sync(struct gfs2_glock *gl) * Writeback of the data mapping may cause the dirty flag to be set * so we have to clear it again here. */ - smp_mb__before_clear_bit(); + smp_mb__before_atomic(); clear_bit(GLF_DIRTY, &gl->gl_flags); } @@ -168,8 +234,8 @@ static void inode_go_sync(struct gfs2_glock *gl) * inode_go_inval - prepare a inode glock to be released * @gl: the glock * @flags: - * - * Normally we invlidate everything, but if we are moving into + * + * Normally we invalidate everything, but if we are moving into * LM_ST_DEFERRED from LM_ST_SHARED or LM_ST_EXCLUSIVE then we * can keep hold of the metadata, since it won't have changed. * @@ -187,11 +253,14 @@ static void inode_go_inval(struct gfs2_glock *gl, int flags) if (ip) { set_bit(GIF_INVALID, &ip->i_flags); forget_all_cached_acls(&ip->i_inode); + gfs2_dir_hash_inval(ip); } } - if (ip == GFS2_I(gl->gl_sbd->sd_rindex)) + if (ip == GFS2_I(gl->gl_sbd->sd_rindex)) { + gfs2_log_flush(gl->gl_sbd, NULL, NORMAL_FLUSH); gl->gl_sbd->sd_rindex_uptodate = 0; + } if (ip && S_ISREG(ip->i_inode.i_mode)) truncate_inode_pages(ip->i_inode.i_mapping, 0); } @@ -206,12 +275,130 @@ static void inode_go_inval(struct gfs2_glock *gl, int flags) static int inode_go_demote_ok(const struct gfs2_glock *gl) { struct gfs2_sbd *sdp = gl->gl_sbd; + struct gfs2_holder *gh; + if (sdp->sd_jindex == gl->gl_object || sdp->sd_rindex == gl->gl_object) return 0; + + if (!list_empty(&gl->gl_holders)) { + gh = list_entry(gl->gl_holders.next, struct gfs2_holder, gh_list); + if (gh->gh_list.next != &gl->gl_holders) + return 0; + } + return 1; } /** + * gfs2_set_nlink - Set the inode's link count based on on-disk info + * @inode: The inode in question + * @nlink: The link count + * + * If the link count has hit zero, it must never be raised, whatever the + * on-disk inode might say. When new struct inodes are created the link + * count is set to 1, so that we can safely use this test even when reading + * in on disk information for the first time. + */ + +static void gfs2_set_nlink(struct inode *inode, u32 nlink) +{ + /* + * We will need to review setting the nlink count here in the + * light of the forthcoming ro bind mount work. This is a reminder + * to do that. + */ + if ((inode->i_nlink != nlink) && (inode->i_nlink != 0)) { + if (nlink == 0) + clear_nlink(inode); + else + set_nlink(inode, nlink); + } +} + +static int gfs2_dinode_in(struct gfs2_inode *ip, const void *buf) +{ + const struct gfs2_dinode *str = buf; + struct timespec atime; + u16 height, depth; + + if (unlikely(ip->i_no_addr != be64_to_cpu(str->di_num.no_addr))) + goto corrupt; + ip->i_no_formal_ino = be64_to_cpu(str->di_num.no_formal_ino); + ip->i_inode.i_mode = be32_to_cpu(str->di_mode); + ip->i_inode.i_rdev = 0; + switch (ip->i_inode.i_mode & S_IFMT) { + case S_IFBLK: + case S_IFCHR: + ip->i_inode.i_rdev = MKDEV(be32_to_cpu(str->di_major), + be32_to_cpu(str->di_minor)); + break; + }; + + i_uid_write(&ip->i_inode, be32_to_cpu(str->di_uid)); + i_gid_write(&ip->i_inode, be32_to_cpu(str->di_gid)); + gfs2_set_nlink(&ip->i_inode, be32_to_cpu(str->di_nlink)); + i_size_write(&ip->i_inode, be64_to_cpu(str->di_size)); + gfs2_set_inode_blocks(&ip->i_inode, be64_to_cpu(str->di_blocks)); + atime.tv_sec = be64_to_cpu(str->di_atime); + atime.tv_nsec = be32_to_cpu(str->di_atime_nsec); + if (timespec_compare(&ip->i_inode.i_atime, &atime) < 0) + ip->i_inode.i_atime = atime; + ip->i_inode.i_mtime.tv_sec = be64_to_cpu(str->di_mtime); + ip->i_inode.i_mtime.tv_nsec = be32_to_cpu(str->di_mtime_nsec); + ip->i_inode.i_ctime.tv_sec = be64_to_cpu(str->di_ctime); + ip->i_inode.i_ctime.tv_nsec = be32_to_cpu(str->di_ctime_nsec); + + ip->i_goal = be64_to_cpu(str->di_goal_meta); + ip->i_generation = be64_to_cpu(str->di_generation); + + ip->i_diskflags = be32_to_cpu(str->di_flags); + ip->i_eattr = be64_to_cpu(str->di_eattr); + /* i_diskflags and i_eattr must be set before gfs2_set_inode_flags() */ + gfs2_set_inode_flags(&ip->i_inode); + height = be16_to_cpu(str->di_height); + if (unlikely(height > GFS2_MAX_META_HEIGHT)) + goto corrupt; + ip->i_height = (u8)height; + + depth = be16_to_cpu(str->di_depth); + if (unlikely(depth > GFS2_DIR_MAX_DEPTH)) + goto corrupt; + ip->i_depth = (u8)depth; + ip->i_entries = be32_to_cpu(str->di_entries); + + if (S_ISREG(ip->i_inode.i_mode)) + gfs2_set_aops(&ip->i_inode); + + return 0; +corrupt: + gfs2_consist_inode(ip); + return -EIO; +} + +/** + * gfs2_inode_refresh - Refresh the incore copy of the dinode + * @ip: The GFS2 inode + * + * Returns: errno + */ + +int gfs2_inode_refresh(struct gfs2_inode *ip) +{ + struct buffer_head *dibh; + int error; + + error = gfs2_meta_inode_buffer(ip, &dibh); + if (error) + return error; + + error = gfs2_dinode_in(ip, dibh->b_data); + brelse(dibh); + clear_bit(GIF_INVALID, &ip->i_flags); + + return error; +} + +/** * inode_go_lock - operation done after an inode lock is locked by a process * @gl: the glock * @flags: @@ -235,6 +422,9 @@ static int inode_go_lock(struct gfs2_holder *gh) return error; } + if (gh->gh_state != LM_ST_DEFERRED) + inode_dio_wait(&ip->i_inode); + if ((ip->i_diskflags & GFS2_DIF_TRUNC_IN_PROG) && (gl->gl_state == LM_ST_EXCLUSIVE) && (gh->gh_state == LM_ST_EXCLUSIVE)) { @@ -254,89 +444,55 @@ static int inode_go_lock(struct gfs2_holder *gh) * @seq: The iterator * @ip: the inode * - * Returns: 0 on success, -ENOBUFS when we run out of space */ -static int inode_go_dump(struct seq_file *seq, const struct gfs2_glock *gl) +static void inode_go_dump(struct seq_file *seq, const struct gfs2_glock *gl) { const struct gfs2_inode *ip = gl->gl_object; if (ip == NULL) - return 0; + return; gfs2_print_dbg(seq, " I: n:%llu/%llu t:%u f:0x%02lx d:0x%08x s:%llu\n", (unsigned long long)ip->i_no_formal_ino, (unsigned long long)ip->i_no_addr, IF2DT(ip->i_inode.i_mode), ip->i_flags, (unsigned int)ip->i_diskflags, (unsigned long long)i_size_read(&ip->i_inode)); - return 0; } /** - * rgrp_go_demote_ok - Check to see if it's ok to unlock a RG's glock - * @gl: the glock - * - * Returns: 1 if it's ok - */ - -static int rgrp_go_demote_ok(const struct gfs2_glock *gl) -{ - const struct address_space *mapping = (const struct address_space *)(gl + 1); - return !mapping->nrpages; -} - -/** - * rgrp_go_lock - operation done after an rgrp lock is locked by - * a first holder on this node. - * @gl: the glock - * @flags: - * - * Returns: errno - */ - -static int rgrp_go_lock(struct gfs2_holder *gh) -{ - return gfs2_rgrp_bh_get(gh->gh_gl->gl_object); -} - -/** - * rgrp_go_unlock - operation done before an rgrp lock is unlocked by - * a last holder on this node. - * @gl: the glock - * @flags: - * - */ - -static void rgrp_go_unlock(struct gfs2_holder *gh) -{ - gfs2_rgrp_bh_put(gh->gh_gl->gl_object); -} - -/** - * trans_go_sync - promote/demote the transaction glock + * freeze_go_sync - promote/demote the freeze glock * @gl: the glock * @state: the requested state * @flags: * */ -static void trans_go_sync(struct gfs2_glock *gl) +static void freeze_go_sync(struct gfs2_glock *gl) { struct gfs2_sbd *sdp = gl->gl_sbd; + DEFINE_WAIT(wait); - if (gl->gl_state != LM_ST_UNLOCKED && + if (gl->gl_state == LM_ST_SHARED && test_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags)) { - gfs2_meta_syncfs(sdp); - gfs2_log_shutdown(sdp); + atomic_set(&sdp->sd_log_freeze, 1); + wake_up(&sdp->sd_logd_waitq); + do { + prepare_to_wait(&sdp->sd_log_frozen_wait, &wait, + TASK_UNINTERRUPTIBLE); + if (atomic_read(&sdp->sd_log_freeze)) + io_schedule(); + } while(atomic_read(&sdp->sd_log_freeze)); + finish_wait(&sdp->sd_log_frozen_wait, &wait); } } /** - * trans_go_xmote_bh - After promoting/demoting the transaction glock + * freeze_go_xmote_bh - After promoting/demoting the freeze glock * @gl: the glock * */ -static int trans_go_xmote_bh(struct gfs2_glock *gl, struct gfs2_holder *gh) +static int freeze_go_xmote_bh(struct gfs2_glock *gl, struct gfs2_holder *gh) { struct gfs2_sbd *sdp = gl->gl_sbd; struct gfs2_inode *ip = GFS2_I(sdp->sd_jdesc->jd_inode); @@ -369,7 +525,7 @@ static int trans_go_xmote_bh(struct gfs2_glock *gl, struct gfs2_holder *gh) * Always returns 0 */ -static int trans_go_demote_ok(const struct gfs2_glock *gl) +static int freeze_go_demote_ok(const struct gfs2_glock *gl) { return 0; } @@ -380,15 +536,19 @@ static int trans_go_demote_ok(const struct gfs2_glock *gl) * * gl_spin lock is held while calling this */ -static void iopen_go_callback(struct gfs2_glock *gl) +static void iopen_go_callback(struct gfs2_glock *gl, bool remote) { struct gfs2_inode *ip = (struct gfs2_inode *)gl->gl_object; + struct gfs2_sbd *sdp = gl->gl_sbd; + + if (!remote || (sdp->sd_vfs->s_flags & MS_RDONLY)) + return; if (gl->gl_demote_state == LM_ST_UNLOCKED && gl->gl_state == LM_ST_SHARED && ip) { - gfs2_glock_hold(gl); + gl->gl_lockref.count++; if (queue_work(gfs2_delete_workqueue, &gl->gl_delete) == 0) - gfs2_glock_put_nolock(gl); + gl->gl_lockref.count--; } } @@ -397,32 +557,29 @@ const struct gfs2_glock_operations gfs2_meta_glops = { }; const struct gfs2_glock_operations gfs2_inode_glops = { - .go_xmote_th = inode_go_sync, + .go_sync = inode_go_sync, .go_inval = inode_go_inval, .go_demote_ok = inode_go_demote_ok, .go_lock = inode_go_lock, .go_dump = inode_go_dump, .go_type = LM_TYPE_INODE, - .go_min_hold_time = HZ / 5, .go_flags = GLOF_ASPACE, }; const struct gfs2_glock_operations gfs2_rgrp_glops = { - .go_xmote_th = rgrp_go_sync, + .go_sync = rgrp_go_sync, .go_inval = rgrp_go_inval, - .go_demote_ok = rgrp_go_demote_ok, - .go_lock = rgrp_go_lock, - .go_unlock = rgrp_go_unlock, + .go_lock = gfs2_rgrp_go_lock, + .go_unlock = gfs2_rgrp_go_unlock, .go_dump = gfs2_rgrp_dump, .go_type = LM_TYPE_RGRP, - .go_min_hold_time = HZ / 5, - .go_flags = GLOF_ASPACE, + .go_flags = GLOF_LVB, }; -const struct gfs2_glock_operations gfs2_trans_glops = { - .go_xmote_th = trans_go_sync, - .go_xmote_bh = trans_go_xmote_bh, - .go_demote_ok = trans_go_demote_ok, +const struct gfs2_glock_operations gfs2_freeze_glops = { + .go_sync = freeze_go_sync, + .go_xmote_bh = freeze_go_xmote_bh, + .go_demote_ok = freeze_go_demote_ok, .go_type = LM_TYPE_NONDISK, }; @@ -441,6 +598,7 @@ const struct gfs2_glock_operations gfs2_nondisk_glops = { const struct gfs2_glock_operations gfs2_quota_glops = { .go_type = LM_TYPE_QUOTA, + .go_flags = GLOF_LVB, }; const struct gfs2_glock_operations gfs2_journal_glops = { diff --git a/fs/gfs2/glops.h b/fs/gfs2/glops.h index b3aa2e3210f..7455d2629bc 100644 --- a/fs/gfs2/glops.h +++ b/fs/gfs2/glops.h @@ -15,7 +15,7 @@ extern const struct gfs2_glock_operations gfs2_meta_glops; extern const struct gfs2_glock_operations gfs2_inode_glops; extern const struct gfs2_glock_operations gfs2_rgrp_glops; -extern const struct gfs2_glock_operations gfs2_trans_glops; +extern const struct gfs2_glock_operations gfs2_freeze_glops; extern const struct gfs2_glock_operations gfs2_iopen_glops; extern const struct gfs2_glock_operations gfs2_flock_glops; extern const struct gfs2_glock_operations gfs2_nondisk_glops; @@ -23,4 +23,6 @@ extern const struct gfs2_glock_operations gfs2_quota_glops; extern const struct gfs2_glock_operations gfs2_journal_glops; extern const struct gfs2_glock_operations *gfs2_glops_list[]; +extern void gfs2_ail_flush(struct gfs2_glock *gl, bool fsync); + #endif /* __GLOPS_DOT_H__ */ diff --git a/fs/gfs2/incore.h b/fs/gfs2/incore.h index a79790c0627..67d310c9ada 100644 --- a/fs/gfs2/incore.h +++ b/fs/gfs2/incore.h @@ -15,18 +15,23 @@ #include <linux/workqueue.h> #include <linux/dlm.h> #include <linux/buffer_head.h> +#include <linux/rcupdate.h> +#include <linux/rculist_bl.h> +#include <linux/completion.h> +#include <linux/rbtree.h> +#include <linux/ktime.h> +#include <linux/percpu.h> +#include <linux/lockref.h> #define DIO_WAIT 0x00000010 #define DIO_METADATA 0x00000020 -#define DIO_ALL 0x00000100 struct gfs2_log_operations; -struct gfs2_log_element; +struct gfs2_bufdata; struct gfs2_holder; struct gfs2_glock; struct gfs2_quota_data; struct gfs2_trans; -struct gfs2_ail; struct gfs2_jdesc; struct gfs2_sbd; struct lm_lockops; @@ -47,9 +52,8 @@ struct gfs2_log_header_host { */ struct gfs2_log_operations { - void (*lo_add) (struct gfs2_sbd *sdp, struct gfs2_log_element *le); - void (*lo_before_commit) (struct gfs2_sbd *sdp); - void (*lo_after_commit) (struct gfs2_sbd *sdp, struct gfs2_ail *ai); + void (*lo_before_commit) (struct gfs2_sbd *sdp, struct gfs2_trans *tr); + void (*lo_after_commit) (struct gfs2_sbd *sdp, struct gfs2_trans *tr); void (*lo_before_scan) (struct gfs2_jdesc *jd, struct gfs2_log_header_host *head, int pass); int (*lo_scan_elements) (struct gfs2_jdesc *jd, unsigned int start, @@ -59,11 +63,6 @@ struct gfs2_log_operations { const char *lo_name; }; -struct gfs2_log_element { - struct list_head le_list; - const struct gfs2_log_operations *le_ops; -}; - #define GBF_FULL 1 struct gfs2_bitmap { @@ -73,11 +72,11 @@ struct gfs2_bitmap { u32 bi_offset; u32 bi_start; u32 bi_len; + u32 bi_blocks; }; struct gfs2_rgrpd { - struct list_head rd_list; /* Link with superblock */ - struct list_head rd_list_mru; + struct rb_node rd_node; /* Link with superblock */ struct gfs2_glock *rd_gl; /* Glock for this rgrp */ u64 rd_addr; /* grp block disk address */ u64 rd_data0; /* first data location */ @@ -85,46 +84,70 @@ struct gfs2_rgrpd { u32 rd_data; /* num of data blocks in rgrp */ u32 rd_bitbytes; /* number of bytes in data bitmaps */ u32 rd_free; + u32 rd_reserved; /* number of blocks reserved */ u32 rd_free_clone; u32 rd_dinodes; u64 rd_igeneration; struct gfs2_bitmap *rd_bits; - struct mutex rd_mutex; - struct gfs2_log_element rd_le; struct gfs2_sbd *rd_sbd; - unsigned int rd_bh_count; + struct gfs2_rgrp_lvb *rd_rgl; u32 rd_last_alloc; u32 rd_flags; + u32 rd_extfail_pt; /* extent failure point */ #define GFS2_RDF_CHECK 0x10000000 /* check for unlinked inodes */ #define GFS2_RDF_UPTODATE 0x20000000 /* rg is up to date */ #define GFS2_RDF_ERROR 0x40000000 /* error in rg */ #define GFS2_RDF_MASK 0xf0000000 /* mask for internal flags */ + spinlock_t rd_rsspin; /* protects reservation related vars */ + struct rb_root rd_rstree; /* multi-block reservation tree */ +}; + +struct gfs2_rbm { + struct gfs2_rgrpd *rgd; + u32 offset; /* The offset is bitmap relative */ + int bii; /* Bitmap index */ }; +static inline struct gfs2_bitmap *rbm_bi(const struct gfs2_rbm *rbm) +{ + return rbm->rgd->rd_bits + rbm->bii; +} + +static inline u64 gfs2_rbm_to_block(const struct gfs2_rbm *rbm) +{ + return rbm->rgd->rd_data0 + (rbm_bi(rbm)->bi_start * GFS2_NBBY) + + rbm->offset; +} + +static inline bool gfs2_rbm_eq(const struct gfs2_rbm *rbm1, + const struct gfs2_rbm *rbm2) +{ + return (rbm1->rgd == rbm2->rgd) && (rbm1->bii == rbm2->bii) && + (rbm1->offset == rbm2->offset); +} + enum gfs2_state_bits { BH_Pinned = BH_PrivateStart, BH_Escaped = BH_PrivateStart + 1, + BH_Zeronew = BH_PrivateStart + 2, }; BUFFER_FNS(Pinned, pinned) TAS_BUFFER_FNS(Pinned, pinned) BUFFER_FNS(Escaped, escaped) TAS_BUFFER_FNS(Escaped, escaped) +BUFFER_FNS(Zeronew, zeronew) +TAS_BUFFER_FNS(Zeronew, zeronew) struct gfs2_bufdata { struct buffer_head *bd_bh; struct gfs2_glock *bd_gl; + u64 bd_blkno; - union { - struct list_head list_tr; - u64 blkno; - } u; -#define bd_list_tr u.list_tr -#define bd_blkno u.blkno + struct list_head bd_list; + const struct gfs2_log_operations *bd_ops; - struct gfs2_log_element bd_le; - - struct gfs2_ail *bd_ail; + struct gfs2_trans *bd_tr; struct list_head bd_ail_st_list; struct list_head bd_ail_gl_list; }; @@ -137,8 +160,45 @@ struct gfs2_bufdata { #define GDLM_STRNAME_BYTES 25 #define GDLM_LVB_SIZE 32 +/* + * ls_recover_flags: + * + * DFL_BLOCK_LOCKS: dlm is in recovery and will grant locks that had been + * held by failed nodes whose journals need recovery. Those locks should + * only be used for journal recovery until the journal recovery is done. + * This is set by the dlm recover_prep callback and cleared by the + * gfs2_control thread when journal recovery is complete. To avoid + * races between recover_prep setting and gfs2_control clearing, recover_spin + * is held while changing this bit and reading/writing recover_block + * and recover_start. + * + * DFL_NO_DLM_OPS: dlm lockspace ops/callbacks are not being used. + * + * DFL_FIRST_MOUNT: this node is the first to mount this fs and is doing + * recovery of all journals before allowing other nodes to mount the fs. + * This is cleared when FIRST_MOUNT_DONE is set. + * + * DFL_FIRST_MOUNT_DONE: this node was the first mounter, and has finished + * recovery of all journals, and now allows other nodes to mount the fs. + * + * DFL_MOUNT_DONE: gdlm_mount has completed successfully and cleared + * BLOCK_LOCKS for the first time. The gfs2_control thread should now + * control clearing BLOCK_LOCKS for further recoveries. + * + * DFL_UNMOUNT: gdlm_unmount sets to keep sdp off gfs2_control_wq. + * + * DFL_DLM_RECOVERY: set while dlm is in recovery, between recover_prep() + * and recover_done(), i.e. set while recover_block == recover_start. + */ + enum { DFL_BLOCK_LOCKS = 0, + DFL_NO_DLM_OPS = 1, + DFL_FIRST_MOUNT = 2, + DFL_FIRST_MOUNT_DONE = 3, + DFL_MOUNT_DONE = 4, + DFL_UNMOUNT = 5, + DFL_DLM_RECOVERY = 6, }; struct lm_lockname { @@ -152,18 +212,34 @@ struct lm_lockname { struct gfs2_glock_operations { - void (*go_xmote_th) (struct gfs2_glock *gl); + void (*go_sync) (struct gfs2_glock *gl); int (*go_xmote_bh) (struct gfs2_glock *gl, struct gfs2_holder *gh); void (*go_inval) (struct gfs2_glock *gl, int flags); int (*go_demote_ok) (const struct gfs2_glock *gl); int (*go_lock) (struct gfs2_holder *gh); void (*go_unlock) (struct gfs2_holder *gh); - int (*go_dump)(struct seq_file *seq, const struct gfs2_glock *gl); - void (*go_callback) (struct gfs2_glock *gl); + void (*go_dump)(struct seq_file *seq, const struct gfs2_glock *gl); + void (*go_callback)(struct gfs2_glock *gl, bool remote); const int go_type; - const unsigned long go_min_hold_time; const unsigned long go_flags; #define GLOF_ASPACE 1 +#define GLOF_LVB 2 +}; + +enum { + GFS2_LKS_SRTT = 0, /* Non blocking smoothed round trip time */ + GFS2_LKS_SRTTVAR = 1, /* Non blocking smoothed variance */ + GFS2_LKS_SRTTB = 2, /* Blocking smoothed round trip time */ + GFS2_LKS_SRTTVARB = 3, /* Blocking smoothed variance */ + GFS2_LKS_SIRT = 4, /* Smoothed Inter-request time */ + GFS2_LKS_SIRTVAR = 5, /* Smoothed Inter-request variance */ + GFS2_LKS_DCOUNT = 6, /* Count of dlm requests */ + GFS2_LKS_QCOUNT = 7, /* Count of gfs2_holder queues */ + GFS2_NR_LKSTATS +}; + +struct gfs2_lkstats { + s64 stats[GFS2_NR_LKSTATS]; }; enum { @@ -186,6 +262,45 @@ struct gfs2_holder { unsigned long gh_ip; }; +/* Resource group multi-block reservation, in order of appearance: + + Step 1. Function prepares to write, allocates a mb, sets the size hint. + Step 2. User calls inplace_reserve to target an rgrp, sets the rgrp info + Step 3. Function get_local_rgrp locks the rgrp, determines which bits to use + Step 4. Bits are assigned from the rgrp based on either the reservation + or wherever it can. +*/ + +struct gfs2_blkreserv { + /* components used during write (step 1): */ + atomic_t rs_sizehint; /* hint of the write size */ + + struct gfs2_holder rs_rgd_gh; /* Filled in by get_local_rgrp */ + struct rb_node rs_node; /* link to other block reservations */ + struct gfs2_rbm rs_rbm; /* Start of reservation */ + u32 rs_free; /* how many blocks are still free */ + u64 rs_inum; /* Inode number for reservation */ + + /* ancillary quota stuff */ + struct gfs2_quota_data *rs_qa_qd[2 * MAXQUOTAS]; + struct gfs2_holder rs_qa_qd_ghs[2 * MAXQUOTAS]; + unsigned int rs_qa_qd_num; +}; + +/* + * Allocation parameters + * @target: The number of blocks we'd ideally like to allocate + * @aflags: The flags (e.g. Orlov flag) + * + * The intent is to gradually expand this structure over time in + * order to give more information, e.g. alignment, min extent size + * to the allocation code. + */ +struct gfs2_alloc_parms { + u32 target; + u32 aflags; +}; + enum { GLF_LOCK = 1, GLF_DEMOTE = 3, @@ -198,15 +313,19 @@ enum { GLF_INITIAL = 10, GLF_FROZEN = 11, GLF_QUEUED = 12, + GLF_LRU = 13, + GLF_OBJECT = 14, /* Used only for tracing */ + GLF_BLOCKING = 15, }; struct gfs2_glock { - struct hlist_node gl_list; + struct hlist_bl_node gl_list; + struct gfs2_sbd *gl_sbd; unsigned long gl_flags; /* GLF_... */ struct lm_lockname gl_name; - atomic_t gl_ref; - spinlock_t gl_spin; + struct lockref gl_lockref; +#define gl_spin gl_lockref.lock /* State fields protected by gl_spin */ unsigned int gl_state:2, /* Current state */ @@ -217,54 +336,44 @@ struct gfs2_glock { unsigned int gl_hash; unsigned long gl_demote_time; /* time of first demote request */ + long gl_hold_time; struct list_head gl_holders; const struct gfs2_glock_operations *gl_ops; - char gl_strname[GDLM_STRNAME_BYTES]; + ktime_t gl_dstamp; + struct gfs2_lkstats gl_stats; struct dlm_lksb gl_lksb; - char gl_lvb[32]; unsigned long gl_tchange; void *gl_object; struct list_head gl_lru; - - struct gfs2_sbd *gl_sbd; - struct list_head gl_ail_list; atomic_t gl_ail_count; + atomic_t gl_revokes; struct delayed_work gl_work; - struct work_struct gl_delete; + union { + /* For inode and iopen glocks only */ + struct work_struct gl_delete; + /* For rgrp glocks only */ + struct { + loff_t start; + loff_t end; + } gl_vm; + }; + struct rcu_head gl_rcu; }; #define GFS2_MIN_LVB_SIZE 32 /* Min size of LVB that gfs2 supports */ -struct gfs2_alloc { - /* Quota stuff */ - - struct gfs2_quota_data *al_qd[2*MAXQUOTAS]; - struct gfs2_holder al_qd_ghs[2*MAXQUOTAS]; - unsigned int al_qd_num; - - u32 al_requested; /* Filled in by caller of gfs2_inplace_reserve() */ - u32 al_alloced; /* Filled in by gfs2_alloc_*() */ - - /* Filled in by gfs2_inplace_reserve() */ - - unsigned int al_line; - char *al_file; - struct gfs2_holder al_ri_gh; - struct gfs2_holder al_rgd_gh; - struct gfs2_rgrpd *al_rgd; - -}; - enum { GIF_INVALID = 0, GIF_QD_LOCKED = 1, + GIF_ALLOC_FAILED = 2, GIF_SW_PAGED = 3, + GIF_ORDERED = 4, + GIF_FREE_VFS_INODE = 5, }; - struct gfs2_inode { struct inode i_inode; u64 i_no_addr; @@ -275,10 +384,13 @@ struct gfs2_inode { struct gfs2_glock *i_gl; /* Move into i_gh? */ struct gfs2_holder i_iopen_gh; struct gfs2_holder i_gh; /* for prepare/commit_write only */ - struct gfs2_alloc *i_alloc; + struct gfs2_blkreserv *i_res; /* rgrp multi-block reservation */ + struct gfs2_rgrpd *i_rgd; u64 i_goal; /* goal block for allocations */ struct rw_semaphore i_rw_mutex; + struct list_head i_ordered; struct list_head i_trunc_list; + __be64 *i_hash_cache; u32 i_entries; u32 i_diskflags; u8 i_height; @@ -311,18 +423,20 @@ struct gfs2_revoke_replay { }; enum { - QDF_USER = 0, QDF_CHANGE = 1, QDF_LOCKED = 2, + QDF_REFRESH = 3, }; struct gfs2_quota_data { + struct hlist_bl_node qd_hlist; struct list_head qd_list; - struct list_head qd_reclaim; - - atomic_t qd_count; + struct kqid qd_id; + struct gfs2_sbd *qd_sbd; + struct lockref qd_lockref; + struct list_head qd_lru; + unsigned qd_hash; - u32 qd_id; unsigned long qd_flags; /* QDF_... */ s64 qd_change; @@ -340,6 +454,7 @@ struct gfs2_quota_data { u64 qd_sync_gen; unsigned long qd_last_warn; + struct rcu_head qd_rcu; }; struct gfs2_trans { @@ -348,34 +463,28 @@ struct gfs2_trans { unsigned int tr_blocks; unsigned int tr_revokes; unsigned int tr_reserved; + unsigned int tr_touched:1; + unsigned int tr_attached:1; + unsigned int tr_alloced:1; - struct gfs2_holder tr_t_gh; - - int tr_touched; - - unsigned int tr_num_buf; unsigned int tr_num_buf_new; unsigned int tr_num_databuf_new; unsigned int tr_num_buf_rm; unsigned int tr_num_databuf_rm; - struct list_head tr_list_buf; - unsigned int tr_num_revoke; unsigned int tr_num_revoke_rm; -}; - -struct gfs2_ail { - struct list_head ai_list; - unsigned int ai_first; - struct list_head ai_ail1_list; - struct list_head ai_ail2_list; + struct list_head tr_list; + struct list_head tr_databuf; + struct list_head tr_buf; - u64 ai_sync_gen; + unsigned int tr_first; + struct list_head tr_ail1_list; + struct list_head tr_ail2_list; }; struct gfs2_journal_extent { - struct list_head extent_list; + struct list_head list; unsigned int lblock; /* First logical block */ u64 dblock; /* First disk block */ @@ -385,12 +494,23 @@ struct gfs2_journal_extent { struct gfs2_jdesc { struct list_head jd_list; struct list_head extent_list; + unsigned int nr_extents; struct work_struct jd_work; struct inode *jd_inode; unsigned long jd_flags; #define JDF_RECOVERY 1 unsigned int jd_jid; unsigned int jd_blocks; + int jd_recover_error; + /* Replay stuff */ + + unsigned int jd_found_blocks; + unsigned int jd_found_revokes; + unsigned int jd_replayed_blocks; + + struct list_head jd_revoke_list; + unsigned int jd_replay_tail; + }; struct gfs2_statfs_change_host { @@ -429,6 +549,7 @@ struct gfs2_args { unsigned int ar_discard:1; /* discard requests */ unsigned int ar_errors:2; /* errors=withdraw | panic */ unsigned int ar_nobarrier:1; /* do not send barriers */ + unsigned int ar_rgrplvb:1; /* use lvbs for rgrp info */ int ar_commit; /* Commit interval */ int ar_statfs_quantum; /* The fast statfs interval */ int ar_quota_quantum; /* The quota interval */ @@ -440,7 +561,6 @@ struct gfs2_tune { unsigned int gt_logd_secs; - unsigned int gt_quota_simul_sync; /* Max quotavals to sync at once */ unsigned int gt_quota_warn_period; /* Secs between quota warn msgs */ unsigned int gt_quota_scale_num; /* Numerator */ unsigned int gt_quota_scale_den; /* Denominator */ @@ -460,6 +580,8 @@ enum { SDF_NORECOVERY = 4, SDF_DEMOTE = 5, SDF_NOJOURNALID = 6, + SDF_RORECOVERY = 7, /* read only recovery */ + SDF_SKIP_DLM_UNLOCK = 8, }; #define GFS2_FSNAME_LEN 256 @@ -484,7 +606,6 @@ struct gfs2_sb_host { char sb_lockproto[GFS2_LOCKNAME_LEN]; char sb_locktable[GFS2_LOCKNAME_LEN]; - u8 sb_uuid[16]; }; /* @@ -499,18 +620,36 @@ struct gfs2_sb_host { struct lm_lockstruct { int ls_jid; unsigned int ls_first; - unsigned int ls_first_done; - unsigned int ls_nodir; const struct lm_lockops *ls_ops; - unsigned long ls_flags; dlm_lockspace_t *ls_dlm; - int ls_recover_jid_done; - int ls_recover_jid_status; + int ls_recover_jid_done; /* These two are deprecated, */ + int ls_recover_jid_status; /* used previously by gfs_controld */ + + struct dlm_lksb ls_mounted_lksb; /* mounted_lock */ + struct dlm_lksb ls_control_lksb; /* control_lock */ + char ls_control_lvb[GDLM_LVB_SIZE]; /* control_lock lvb */ + struct completion ls_sync_wait; /* {control,mounted}_{lock,unlock} */ + char *ls_lvb_bits; + + spinlock_t ls_recover_spin; /* protects following fields */ + unsigned long ls_recover_flags; /* DFL_ */ + uint32_t ls_recover_mount; /* gen in first recover_done cb */ + uint32_t ls_recover_start; /* gen in last recover_done cb */ + uint32_t ls_recover_block; /* copy recover_start in last recover_prep */ + uint32_t ls_recover_size; /* size of recover_submit, recover_result */ + uint32_t *ls_recover_submit; /* gen in last recover_slot cb per jid */ + uint32_t *ls_recover_result; /* result of last jid recovery */ +}; + +struct gfs2_pcpu_lkstats { + /* One struct for each glock type */ + struct gfs2_lkstats lkstats[10]; }; struct gfs2_sbd { struct super_block *sd_vfs; + struct gfs2_pcpu_lkstats __percpu *sd_lkstats; struct kobject sd_kobj; unsigned long sd_flags; /* SDF_... */ struct gfs2_sb_host sd_sb; @@ -526,6 +665,7 @@ struct gfs2_sbd { u32 sd_hash_bsize_shift; u32 sd_hash_ptrs; /* Number of pointers in a hash block */ u32 sd_qc_per_block; + u32 sd_blocks_per_bitmap; u32 sd_max_dirres; /* Max blocks needed to add a directory entry */ u32 sd_max_height; /* Max height of a file's metadata tree */ u64 sd_heightsize[GFS2_MAX_META_HEIGHT + 1]; @@ -540,9 +680,12 @@ struct gfs2_sbd { struct lm_lockstruct sd_lockstruct; struct gfs2_holder sd_live_gh; struct gfs2_glock *sd_rename_gl; - struct gfs2_glock *sd_trans_gl; + struct gfs2_glock *sd_freeze_gl; wait_queue_head_t sd_glock_wait; atomic_t sd_glock_disposal; + struct completion sd_locking_init; + struct completion sd_wdack; + struct delayed_work sd_control_work; /* Inode Stuff */ @@ -567,10 +710,7 @@ struct gfs2_sbd { int sd_rindex_uptodate; spinlock_t sd_rindex_spin; - struct mutex sd_rindex_mutex; - struct list_head sd_rindex_list; - struct list_head sd_rindex_mru_list; - struct gfs2_rgrpd *sd_rindex_forward; + struct rb_root sd_rindex_tree; unsigned int sd_rgrps; unsigned int sd_max_rg_data; @@ -588,6 +728,8 @@ struct gfs2_sbd { struct gfs2_holder sd_sc_gh; struct gfs2_holder sd_qc_gh; + struct completion sd_journal_ready; + /* Daemon stuff */ struct task_struct *sd_logd_process; @@ -598,36 +740,33 @@ struct gfs2_sbd { struct list_head sd_quota_list; atomic_t sd_quota_count; struct mutex sd_quota_mutex; + struct mutex sd_quota_sync_mutex; wait_queue_head_t sd_quota_wait; struct list_head sd_trunc_list; spinlock_t sd_trunc_lock; unsigned int sd_quota_slots; - unsigned int sd_quota_chunks; - unsigned char **sd_quota_bitmap; + unsigned long *sd_quota_bitmap; + spinlock_t sd_bitmap_lock; u64 sd_quota_sync_gen; /* Log stuff */ + struct address_space sd_aspace; + spinlock_t sd_log_lock; + struct gfs2_trans *sd_log_tr; unsigned int sd_log_blks_reserved; - unsigned int sd_log_commited_buf; - unsigned int sd_log_commited_databuf; int sd_log_commited_revoke; atomic_t sd_log_pinned; - unsigned int sd_log_num_buf; unsigned int sd_log_num_revoke; - unsigned int sd_log_num_rg; - unsigned int sd_log_num_databuf; - struct list_head sd_log_le_buf; struct list_head sd_log_le_revoke; - struct list_head sd_log_le_rg; - struct list_head sd_log_le_databuf; struct list_head sd_log_le_ordered; + spinlock_t sd_ordered_lock; atomic_t sd_log_thresh1; atomic_t sd_log_thresh2; @@ -642,29 +781,25 @@ struct gfs2_sbd { struct rw_semaphore sd_log_flush_lock; atomic_t sd_log_in_flight; + struct bio *sd_log_bio; wait_queue_head_t sd_log_flush_wait; + int sd_log_error; unsigned int sd_log_flush_head; u64 sd_log_flush_wrapped; + spinlock_t sd_ail_lock; struct list_head sd_ail1_list; struct list_head sd_ail2_list; - u64 sd_ail_sync_gen; - - /* Replay stuff */ - - struct list_head sd_revoke_list; - unsigned int sd_replay_tail; - - unsigned int sd_found_blocks; - unsigned int sd_found_revokes; - unsigned int sd_replayed_blocks; /* For quiescing the filesystem */ - struct gfs2_holder sd_freeze_gh; - struct mutex sd_freeze_lock; - unsigned int sd_freeze_count; + struct gfs2_holder sd_freeze_root_gh; + struct gfs2_holder sd_thaw_gh; + atomic_t sd_log_freeze; + atomic_t sd_frozen_root; + wait_queue_head_t sd_frozen_root_wait; + wait_queue_head_t sd_log_frozen_wait; char sd_fsname[GFS2_FSNAME_LEN]; char sd_table_name[GFS2_FSNAME_LEN]; @@ -674,8 +809,23 @@ struct gfs2_sbd { unsigned long sd_last_warning; struct dentry *debugfs_dir; /* debugfs directory */ - struct dentry *debugfs_dentry_glocks; /* for debugfs */ + struct dentry *debugfs_dentry_glocks; + struct dentry *debugfs_dentry_glstats; + struct dentry *debugfs_dentry_sbstats; }; +static inline void gfs2_glstats_inc(struct gfs2_glock *gl, int which) +{ + gl->gl_stats.stats[which]++; +} + +static inline void gfs2_sbstats_inc(const struct gfs2_glock *gl, int which) +{ + const struct gfs2_sbd *sdp = gl->gl_sbd; + preempt_disable(); + this_cpu_ptr(sdp->sd_lkstats)->lkstats[gl->gl_name.ln_type].stats[which]++; + preempt_enable(); +} + #endif /* __INCORE_DOT_H__ */ diff --git a/fs/gfs2/inode.c b/fs/gfs2/inode.c index 7aa7d4f8984..e62e5947788 100644 --- a/fs/gfs2/inode.c +++ b/fs/gfs2/inode.c @@ -1,23 +1,25 @@ /* * Copyright (C) Sistina Software, Inc. 1997-2003 All rights reserved. - * Copyright (C) 2004-2008 Red Hat, Inc. All rights reserved. + * Copyright (C) 2004-2011 Red Hat, Inc. All rights reserved. * * This copyrighted material is made available to anyone wishing to use, * modify, copy, or redistribute it subject to the terms and conditions * of the GNU General Public License version 2. */ -#include <linux/sched.h> #include <linux/slab.h> #include <linux/spinlock.h> #include <linux/completion.h> #include <linux/buffer_head.h> +#include <linux/namei.h> +#include <linux/mm.h> +#include <linux/xattr.h> #include <linux/posix_acl.h> -#include <linux/sort.h> #include <linux/gfs2_ondisk.h> #include <linux/crc32.h> +#include <linux/fiemap.h> #include <linux/security.h> -#include <linux/time.h> +#include <asm/uaccess.h> #include "gfs2.h" #include "incore.h" @@ -26,51 +28,70 @@ #include "dir.h" #include "xattr.h" #include "glock.h" -#include "glops.h" #include "inode.h" -#include "log.h" #include "meta_io.h" #include "quota.h" #include "rgrp.h" #include "trans.h" #include "util.h" +#include "super.h" +#include "glops.h" -struct gfs2_inum_range_host { - u64 ir_start; - u64 ir_length; +struct gfs2_skip_data { + u64 no_addr; + int skipped; + int non_block; }; static int iget_test(struct inode *inode, void *opaque) { struct gfs2_inode *ip = GFS2_I(inode); - u64 *no_addr = opaque; + struct gfs2_skip_data *data = opaque; - if (ip->i_no_addr == *no_addr) + if (ip->i_no_addr == data->no_addr) { + if (data->non_block && + inode->i_state & (I_FREEING|I_CLEAR|I_WILL_FREE)) { + data->skipped = 1; + return 0; + } return 1; - + } return 0; } static int iget_set(struct inode *inode, void *opaque) { struct gfs2_inode *ip = GFS2_I(inode); - u64 *no_addr = opaque; + struct gfs2_skip_data *data = opaque; - inode->i_ino = (unsigned long)*no_addr; - ip->i_no_addr = *no_addr; + if (data->skipped) + return -ENOENT; + inode->i_ino = (unsigned long)(data->no_addr); + ip->i_no_addr = data->no_addr; return 0; } -struct inode *gfs2_ilookup(struct super_block *sb, u64 no_addr) +struct inode *gfs2_ilookup(struct super_block *sb, u64 no_addr, int non_block) { unsigned long hash = (unsigned long)no_addr; - return ilookup5(sb, hash, iget_test, &no_addr); + struct gfs2_skip_data data; + + data.no_addr = no_addr; + data.skipped = 0; + data.non_block = non_block; + return ilookup5(sb, hash, iget_test, &data); } -static struct inode *gfs2_iget(struct super_block *sb, u64 no_addr) +static struct inode *gfs2_iget(struct super_block *sb, u64 no_addr, + int non_block) { + struct gfs2_skip_data data; unsigned long hash = (unsigned long)no_addr; - return iget5_locked(sb, hash, iget_test, iget_set, &no_addr); + + data.no_addr = no_addr; + data.skipped = 0; + data.non_block = non_block; + return iget5_locked(sb, hash, iget_test, iget_set, &data); } /** @@ -111,23 +132,24 @@ static void gfs2_set_iop(struct inode *inode) * @sb: The super block * @no_addr: The inode number * @type: The type of the inode + * non_block: Can we block on inodes that are being freed? * * Returns: A VFS inode, or an error */ struct inode *gfs2_inode_lookup(struct super_block *sb, unsigned int type, - u64 no_addr, u64 no_formal_ino) + u64 no_addr, u64 no_formal_ino, int non_block) { struct inode *inode; struct gfs2_inode *ip; struct gfs2_glock *io_gl = NULL; int error; - inode = gfs2_iget(sb, no_addr); + inode = gfs2_iget(sb, no_addr, non_block); ip = GFS2_I(inode); if (!inode) - return ERR_PTR(-ENOBUFS); + return ERR_PTR(-ENOMEM); if (inode->i_state & I_NEW) { struct gfs2_sbd *sdp = GFS2_SB(inode); @@ -167,6 +189,7 @@ struct inode *gfs2_inode_lookup(struct super_block *sb, unsigned int type, return inode; fail_refresh: + ip->i_iopen_gh.gh_flags |= GL_NOCACHE; ip->i_iopen_gh.gh_gl->gl_object = NULL; gfs2_glock_dq_uninit(&ip->i_iopen_gh); fail_iopen: @@ -185,11 +208,12 @@ struct inode *gfs2_lookup_by_inum(struct gfs2_sbd *sdp, u64 no_addr, { struct super_block *sb = sdp->sd_vfs; struct gfs2_holder i_gh; - struct inode *inode; + struct inode *inode = NULL; int error; + /* Must not read in block until block type is verified */ error = gfs2_glock_nq_num(sdp, no_addr, &gfs2_inode_glops, - LM_ST_SHARED, LM_FLAG_ANY, &i_gh); + LM_ST_EXCLUSIVE, GL_SKIP, &i_gh); if (error) return ERR_PTR(error); @@ -197,7 +221,7 @@ struct inode *gfs2_lookup_by_inum(struct gfs2_sbd *sdp, u64 no_addr, if (error) goto fail; - inode = gfs2_inode_lookup(sb, DT_UNKNOWN, no_addr, 0); + inode = gfs2_inode_lookup(sb, DT_UNKNOWN, no_addr, 0, 1); if (IS_ERR(inode)) goto fail; @@ -222,203 +246,6 @@ fail_iput: goto fail; } -static int gfs2_dinode_in(struct gfs2_inode *ip, const void *buf) -{ - const struct gfs2_dinode *str = buf; - struct timespec atime; - u16 height, depth; - - if (unlikely(ip->i_no_addr != be64_to_cpu(str->di_num.no_addr))) - goto corrupt; - ip->i_no_formal_ino = be64_to_cpu(str->di_num.no_formal_ino); - ip->i_inode.i_mode = be32_to_cpu(str->di_mode); - ip->i_inode.i_rdev = 0; - switch (ip->i_inode.i_mode & S_IFMT) { - case S_IFBLK: - case S_IFCHR: - ip->i_inode.i_rdev = MKDEV(be32_to_cpu(str->di_major), - be32_to_cpu(str->di_minor)); - break; - }; - - ip->i_inode.i_uid = be32_to_cpu(str->di_uid); - ip->i_inode.i_gid = be32_to_cpu(str->di_gid); - /* - * We will need to review setting the nlink count here in the - * light of the forthcoming ro bind mount work. This is a reminder - * to do that. - */ - ip->i_inode.i_nlink = be32_to_cpu(str->di_nlink); - i_size_write(&ip->i_inode, be64_to_cpu(str->di_size)); - gfs2_set_inode_blocks(&ip->i_inode, be64_to_cpu(str->di_blocks)); - atime.tv_sec = be64_to_cpu(str->di_atime); - atime.tv_nsec = be32_to_cpu(str->di_atime_nsec); - if (timespec_compare(&ip->i_inode.i_atime, &atime) < 0) - ip->i_inode.i_atime = atime; - ip->i_inode.i_mtime.tv_sec = be64_to_cpu(str->di_mtime); - ip->i_inode.i_mtime.tv_nsec = be32_to_cpu(str->di_mtime_nsec); - ip->i_inode.i_ctime.tv_sec = be64_to_cpu(str->di_ctime); - ip->i_inode.i_ctime.tv_nsec = be32_to_cpu(str->di_ctime_nsec); - - ip->i_goal = be64_to_cpu(str->di_goal_meta); - ip->i_generation = be64_to_cpu(str->di_generation); - - ip->i_diskflags = be32_to_cpu(str->di_flags); - gfs2_set_inode_flags(&ip->i_inode); - height = be16_to_cpu(str->di_height); - if (unlikely(height > GFS2_MAX_META_HEIGHT)) - goto corrupt; - ip->i_height = (u8)height; - - depth = be16_to_cpu(str->di_depth); - if (unlikely(depth > GFS2_DIR_MAX_DEPTH)) - goto corrupt; - ip->i_depth = (u8)depth; - ip->i_entries = be32_to_cpu(str->di_entries); - - ip->i_eattr = be64_to_cpu(str->di_eattr); - if (S_ISREG(ip->i_inode.i_mode)) - gfs2_set_aops(&ip->i_inode); - - return 0; -corrupt: - if (gfs2_consist_inode(ip)) - gfs2_dinode_print(ip); - return -EIO; -} - -/** - * gfs2_inode_refresh - Refresh the incore copy of the dinode - * @ip: The GFS2 inode - * - * Returns: errno - */ - -int gfs2_inode_refresh(struct gfs2_inode *ip) -{ - struct buffer_head *dibh; - int error; - - error = gfs2_meta_inode_buffer(ip, &dibh); - if (error) - return error; - - if (gfs2_metatype_check(GFS2_SB(&ip->i_inode), dibh, GFS2_METATYPE_DI)) { - brelse(dibh); - return -EIO; - } - - error = gfs2_dinode_in(ip, dibh->b_data); - brelse(dibh); - clear_bit(GIF_INVALID, &ip->i_flags); - - return error; -} - -int gfs2_dinode_dealloc(struct gfs2_inode *ip) -{ - struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); - struct gfs2_alloc *al; - struct gfs2_rgrpd *rgd; - int error; - - if (gfs2_get_inode_blocks(&ip->i_inode) != 1) { - if (gfs2_consist_inode(ip)) - gfs2_dinode_print(ip); - return -EIO; - } - - al = gfs2_alloc_get(ip); - if (!al) - return -ENOMEM; - - error = gfs2_quota_hold(ip, NO_QUOTA_CHANGE, NO_QUOTA_CHANGE); - if (error) - goto out; - - error = gfs2_rindex_hold(sdp, &al->al_ri_gh); - if (error) - goto out_qs; - - rgd = gfs2_blk2rgrpd(sdp, ip->i_no_addr); - if (!rgd) { - gfs2_consist_inode(ip); - error = -EIO; - goto out_rindex_relse; - } - - error = gfs2_glock_nq_init(rgd->rd_gl, LM_ST_EXCLUSIVE, 0, - &al->al_rgd_gh); - if (error) - goto out_rindex_relse; - - error = gfs2_trans_begin(sdp, RES_RG_BIT + RES_STATFS + RES_QUOTA, 1); - if (error) - goto out_rg_gunlock; - - set_bit(GLF_DIRTY, &ip->i_gl->gl_flags); - set_bit(GLF_LFLUSH, &ip->i_gl->gl_flags); - - gfs2_free_di(rgd, ip); - - gfs2_trans_end(sdp); - -out_rg_gunlock: - gfs2_glock_dq_uninit(&al->al_rgd_gh); -out_rindex_relse: - gfs2_glock_dq_uninit(&al->al_ri_gh); -out_qs: - gfs2_quota_unhold(ip); -out: - gfs2_alloc_put(ip); - return error; -} - -/** - * gfs2_change_nlink - Change nlink count on inode - * @ip: The GFS2 inode - * @diff: The change in the nlink count required - * - * Returns: errno - */ -int gfs2_change_nlink(struct gfs2_inode *ip, int diff) -{ - struct buffer_head *dibh; - u32 nlink; - int error; - - BUG_ON(diff != 1 && diff != -1); - nlink = ip->i_inode.i_nlink + diff; - - /* If we are reducing the nlink count, but the new value ends up being - bigger than the old one, we must have underflowed. */ - if (diff < 0 && nlink > ip->i_inode.i_nlink) { - if (gfs2_consist_inode(ip)) - gfs2_dinode_print(ip); - return -EIO; - } - - error = gfs2_meta_inode_buffer(ip, &dibh); - if (error) - return error; - - if (diff > 0) - inc_nlink(&ip->i_inode); - else - drop_nlink(&ip->i_inode); - - ip->i_inode.i_ctime = CURRENT_TIME; - - gfs2_trans_add_bh(ip->i_gl, dibh, 1); - gfs2_dinode_out(ip, dibh->b_data); - brelse(dibh); - mark_inode_dirty(&ip->i_inode); - - if (ip->i_inode.i_nlink == 0) - gfs2_unlink_di(&ip->i_inode); /* mark inode unlinked */ - - return error; -} struct inode *gfs2_lookup_simple(struct inode *dip, const char *name) { @@ -481,12 +308,12 @@ struct inode *gfs2_lookupi(struct inode *dir, const struct qstr *name, } if (!is_root) { - error = gfs2_permission(dir, MAY_EXEC, 0); + error = gfs2_permission(dir, MAY_EXEC); if (error) goto out; } - inode = gfs2_dir_search(dir, name); + inode = gfs2_dir_search(dir, name, false); if (IS_ERR(inode)) error = PTR_ERR(inode); out: @@ -507,28 +334,17 @@ out: */ static int create_ok(struct gfs2_inode *dip, const struct qstr *name, - unsigned int mode) + umode_t mode) { int error; - error = gfs2_permission(&dip->i_inode, MAY_WRITE | MAY_EXEC, 0); + error = gfs2_permission(&dip->i_inode, MAY_WRITE | MAY_EXEC); if (error) return error; /* Don't create entries in an unlinked directory */ if (!dip->i_inode.i_nlink) - return -EPERM; - - error = gfs2_dir_check(&dip->i_inode, name, NULL); - switch (error) { - case -ENOENT: - error = 0; - break; - case 0: - return -EEXIST; - default: - return error; - } + return -ENOENT; if (dip->i_entries == (u32)-1) return -EFBIG; @@ -538,195 +354,186 @@ static int create_ok(struct gfs2_inode *dip, const struct qstr *name, return 0; } -static void munge_mode_uid_gid(struct gfs2_inode *dip, unsigned int *mode, - unsigned int *uid, unsigned int *gid) +static void munge_mode_uid_gid(const struct gfs2_inode *dip, + struct inode *inode) { if (GFS2_SB(&dip->i_inode)->sd_args.ar_suiddir && - (dip->i_inode.i_mode & S_ISUID) && dip->i_inode.i_uid) { - if (S_ISDIR(*mode)) - *mode |= S_ISUID; - else if (dip->i_inode.i_uid != current_fsuid()) - *mode &= ~07111; - *uid = dip->i_inode.i_uid; + (dip->i_inode.i_mode & S_ISUID) && + !uid_eq(dip->i_inode.i_uid, GLOBAL_ROOT_UID)) { + if (S_ISDIR(inode->i_mode)) + inode->i_mode |= S_ISUID; + else if (!uid_eq(dip->i_inode.i_uid, current_fsuid())) + inode->i_mode &= ~07111; + inode->i_uid = dip->i_inode.i_uid; } else - *uid = current_fsuid(); + inode->i_uid = current_fsuid(); if (dip->i_inode.i_mode & S_ISGID) { - if (S_ISDIR(*mode)) - *mode |= S_ISGID; - *gid = dip->i_inode.i_gid; + if (S_ISDIR(inode->i_mode)) + inode->i_mode |= S_ISGID; + inode->i_gid = dip->i_inode.i_gid; } else - *gid = current_fsgid(); + inode->i_gid = current_fsgid(); } -static int alloc_dinode(struct gfs2_inode *dip, u64 *no_addr, u64 *generation) +static int alloc_dinode(struct gfs2_inode *ip, u32 flags, unsigned *dblocks) { - struct gfs2_sbd *sdp = GFS2_SB(&dip->i_inode); + struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); + struct gfs2_alloc_parms ap = { .target = *dblocks, .aflags = flags, }; int error; - if (gfs2_alloc_get(dip) == NULL) - return -ENOMEM; - - dip->i_alloc->al_requested = RES_DINODE; - error = gfs2_inplace_reserve(dip); + error = gfs2_quota_lock_check(ip); if (error) goto out; - error = gfs2_trans_begin(sdp, RES_RG_BIT + RES_STATFS, 0); + error = gfs2_inplace_reserve(ip, &ap); + if (error) + goto out_quota; + + error = gfs2_trans_begin(sdp, (*dblocks * RES_RG_BIT) + RES_STATFS + RES_QUOTA, 0); if (error) goto out_ipreserv; - error = gfs2_alloc_di(dip, no_addr, generation); + error = gfs2_alloc_blocks(ip, &ip->i_no_addr, dblocks, 1, &ip->i_generation); + ip->i_no_formal_ino = ip->i_generation; + ip->i_inode.i_ino = ip->i_no_addr; + ip->i_goal = ip->i_no_addr; gfs2_trans_end(sdp); out_ipreserv: - gfs2_inplace_release(dip); + gfs2_inplace_release(ip); +out_quota: + gfs2_quota_unlock(ip); out: - gfs2_alloc_put(dip); return error; } +static void gfs2_init_dir(struct buffer_head *dibh, + const struct gfs2_inode *parent) +{ + struct gfs2_dinode *di = (struct gfs2_dinode *)dibh->b_data; + struct gfs2_dirent *dent = (struct gfs2_dirent *)(di+1); + + gfs2_qstr2dirent(&gfs2_qdot, GFS2_DIRENT_SIZE(gfs2_qdot.len), dent); + dent->de_inum = di->di_num; /* already GFS2 endian */ + dent->de_type = cpu_to_be16(DT_DIR); + + dent = (struct gfs2_dirent *)((char*)dent + GFS2_DIRENT_SIZE(1)); + gfs2_qstr2dirent(&gfs2_qdotdot, dibh->b_size - GFS2_DIRENT_SIZE(1) - sizeof(struct gfs2_dinode), dent); + gfs2_inum_out(parent, dent); + dent->de_type = cpu_to_be16(DT_DIR); + +} + +/** + * gfs2_init_xattr - Initialise an xattr block for a new inode + * @ip: The inode in question + * + * This sets up an empty xattr block for a new inode, ready to + * take any ACLs, LSM xattrs, etc. + */ + +static void gfs2_init_xattr(struct gfs2_inode *ip) +{ + struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); + struct buffer_head *bh; + struct gfs2_ea_header *ea; + + bh = gfs2_meta_new(ip->i_gl, ip->i_eattr); + gfs2_trans_add_meta(ip->i_gl, bh); + gfs2_metatype_set(bh, GFS2_METATYPE_EA, GFS2_FORMAT_EA); + gfs2_buffer_clear_tail(bh, sizeof(struct gfs2_meta_header)); + + ea = GFS2_EA_BH2FIRST(bh); + ea->ea_rec_len = cpu_to_be32(sdp->sd_jbsize); + ea->ea_type = GFS2_EATYPE_UNUSED; + ea->ea_flags = GFS2_EAFLAG_LAST; + + brelse(bh); +} + /** * init_dinode - Fill in a new dinode structure - * @dip: the directory this inode is being created in - * @gl: The glock covering the new inode - * @inum: the inode number - * @mode: the file permissions - * @uid: - * @gid: + * @dip: The directory this inode is being created in + * @ip: The inode + * @symname: The symlink destination (if a symlink) + * @bhp: The buffer head (returned to caller) * */ -static void init_dinode(struct gfs2_inode *dip, struct gfs2_glock *gl, - const struct gfs2_inum_host *inum, unsigned int mode, - unsigned int uid, unsigned int gid, - const u64 *generation, dev_t dev, struct buffer_head **bhp) +static void init_dinode(struct gfs2_inode *dip, struct gfs2_inode *ip, + const char *symname) { - struct gfs2_sbd *sdp = GFS2_SB(&dip->i_inode); struct gfs2_dinode *di; struct buffer_head *dibh; - struct timespec tv = CURRENT_TIME; - dibh = gfs2_meta_new(gl, inum->no_addr); - gfs2_trans_add_bh(gl, dibh, 1); - gfs2_metatype_set(dibh, GFS2_METATYPE_DI, GFS2_FORMAT_DI); - gfs2_buffer_clear_tail(dibh, sizeof(struct gfs2_dinode)); + dibh = gfs2_meta_new(ip->i_gl, ip->i_no_addr); + gfs2_trans_add_meta(ip->i_gl, dibh); di = (struct gfs2_dinode *)dibh->b_data; + gfs2_dinode_out(ip, di); - di->di_num.no_formal_ino = cpu_to_be64(inum->no_formal_ino); - di->di_num.no_addr = cpu_to_be64(inum->no_addr); - di->di_mode = cpu_to_be32(mode); - di->di_uid = cpu_to_be32(uid); - di->di_gid = cpu_to_be32(gid); - di->di_nlink = 0; - di->di_size = 0; - di->di_blocks = cpu_to_be64(1); - di->di_atime = di->di_mtime = di->di_ctime = cpu_to_be64(tv.tv_sec); - di->di_major = cpu_to_be32(MAJOR(dev)); - di->di_minor = cpu_to_be32(MINOR(dev)); - di->di_goal_meta = di->di_goal_data = cpu_to_be64(inum->no_addr); - di->di_generation = cpu_to_be64(*generation); - di->di_flags = 0; - - if (S_ISREG(mode)) { - if ((dip->i_diskflags & GFS2_DIF_INHERIT_JDATA) || - gfs2_tune_get(sdp, gt_new_files_jdata)) - di->di_flags |= cpu_to_be32(GFS2_DIF_JDATA); - } else if (S_ISDIR(mode)) { - di->di_flags |= cpu_to_be32(dip->i_diskflags & - GFS2_DIF_INHERIT_JDATA); - } - + di->di_major = cpu_to_be32(MAJOR(ip->i_inode.i_rdev)); + di->di_minor = cpu_to_be32(MINOR(ip->i_inode.i_rdev)); di->__pad1 = 0; - di->di_payload_format = cpu_to_be32(S_ISDIR(mode) ? GFS2_FORMAT_DE : 0); - di->di_height = 0; di->__pad2 = 0; di->__pad3 = 0; - di->di_depth = 0; - di->di_entries = 0; memset(&di->__pad4, 0, sizeof(di->__pad4)); - di->di_eattr = 0; - di->di_atime_nsec = cpu_to_be32(tv.tv_nsec); - di->di_mtime_nsec = cpu_to_be32(tv.tv_nsec); - di->di_ctime_nsec = cpu_to_be32(tv.tv_nsec); memset(&di->di_reserved, 0, sizeof(di->di_reserved)); - - set_buffer_uptodate(dibh); - - *bhp = dibh; -} - -static int make_dinode(struct gfs2_inode *dip, struct gfs2_glock *gl, - unsigned int mode, const struct gfs2_inum_host *inum, - const u64 *generation, dev_t dev, struct buffer_head **bhp) -{ - struct gfs2_sbd *sdp = GFS2_SB(&dip->i_inode); - unsigned int uid, gid; - int error; - - munge_mode_uid_gid(dip, &mode, &uid, &gid); - if (!gfs2_alloc_get(dip)) - return -ENOMEM; - - error = gfs2_quota_lock(dip, uid, gid); - if (error) - goto out; + gfs2_buffer_clear_tail(dibh, sizeof(struct gfs2_dinode)); - error = gfs2_quota_check(dip, uid, gid); - if (error) - goto out_quota; + switch(ip->i_inode.i_mode & S_IFMT) { + case S_IFDIR: + gfs2_init_dir(dibh, dip); + break; + case S_IFLNK: + memcpy(dibh->b_data + sizeof(struct gfs2_dinode), symname, ip->i_inode.i_size); + break; + } - error = gfs2_trans_begin(sdp, RES_DINODE + RES_QUOTA, 0); - if (error) - goto out_quota; + set_buffer_uptodate(dibh); + brelse(dibh); +} - init_dinode(dip, gl, inum, mode, uid, gid, generation, dev, bhp); - gfs2_quota_change(dip, +1, uid, gid); - gfs2_trans_end(sdp); +/** + * gfs2_trans_da_blocks - Calculate number of blocks to link inode + * @dip: The directory we are linking into + * @da: The dir add information + * @nr_inodes: The number of inodes involved + * + * This calculate the number of blocks we need to reserve in a + * transaction to link @nr_inodes into a directory. In most cases + * @nr_inodes will be 2 (the directory plus the inode being linked in) + * but in case of rename, 4 may be required. + * + * Returns: Number of blocks + */ -out_quota: - gfs2_quota_unlock(dip); -out: - gfs2_alloc_put(dip); - return error; +static unsigned gfs2_trans_da_blks(const struct gfs2_inode *dip, + const struct gfs2_diradd *da, + unsigned nr_inodes) +{ + return da->nr_blocks + gfs2_rg_blocks(dip, da->nr_blocks) + + (nr_inodes * RES_DINODE) + RES_QUOTA + RES_STATFS; } static int link_dinode(struct gfs2_inode *dip, const struct qstr *name, - struct gfs2_inode *ip) + struct gfs2_inode *ip, struct gfs2_diradd *da) { struct gfs2_sbd *sdp = GFS2_SB(&dip->i_inode); - struct gfs2_alloc *al; - int alloc_required; - struct buffer_head *dibh; + struct gfs2_alloc_parms ap = { .target = da->nr_blocks, }; int error; - al = gfs2_alloc_get(dip); - if (!al) - return -ENOMEM; - - error = gfs2_quota_lock(dip, NO_QUOTA_CHANGE, NO_QUOTA_CHANGE); - if (error) - goto fail; - - error = alloc_required = gfs2_diradd_alloc_required(&dip->i_inode, name); - if (alloc_required < 0) - goto fail_quota_locks; - if (alloc_required) { - error = gfs2_quota_check(dip, dip->i_inode.i_uid, dip->i_inode.i_gid); + if (da->nr_blocks) { + error = gfs2_quota_lock_check(dip); if (error) goto fail_quota_locks; - al->al_requested = sdp->sd_max_dirres; - - error = gfs2_inplace_reserve(dip); + error = gfs2_inplace_reserve(dip, &ap); if (error) goto fail_quota_locks; - error = gfs2_trans_begin(sdp, sdp->sd_max_dirres + - al->al_rgd->rd_length + - 2 * RES_DINODE + - RES_STATFS + RES_QUOTA, 0); + error = gfs2_trans_begin(sdp, gfs2_trans_da_blks(dip, da, 2), 0); if (error) goto fail_ipreserv; } else { @@ -735,91 +542,78 @@ static int link_dinode(struct gfs2_inode *dip, const struct qstr *name, goto fail_quota_locks; } - error = gfs2_dir_add(&dip->i_inode, name, ip, IF2DT(ip->i_inode.i_mode)); + error = gfs2_dir_add(&dip->i_inode, name, ip, da); if (error) goto fail_end_trans; - error = gfs2_meta_inode_buffer(ip, &dibh); - if (error) - goto fail_end_trans; - ip->i_inode.i_nlink = 1; - gfs2_trans_add_bh(ip->i_gl, dibh, 1); - gfs2_dinode_out(ip, dibh->b_data); - brelse(dibh); - return 0; - fail_end_trans: gfs2_trans_end(sdp); - fail_ipreserv: - if (dip->i_alloc->al_rgd) - gfs2_inplace_release(dip); - + gfs2_inplace_release(dip); fail_quota_locks: gfs2_quota_unlock(dip); - -fail: - gfs2_alloc_put(dip); return error; } -static int gfs2_security_init(struct gfs2_inode *dip, struct gfs2_inode *ip) +static int gfs2_initxattrs(struct inode *inode, const struct xattr *xattr_array, + void *fs_info) { - int err; - size_t len; - void *value; - char *name; - - err = security_inode_init_security(&ip->i_inode, &dip->i_inode, - &name, &value, &len); - - if (err) { - if (err == -EOPNOTSUPP) - return 0; - return err; + const struct xattr *xattr; + int err = 0; + + for (xattr = xattr_array; xattr->name != NULL; xattr++) { + err = __gfs2_xattr_set(inode, xattr->name, xattr->value, + xattr->value_len, 0, + GFS2_EATYPE_SECURITY); + if (err < 0) + break; } - - err = __gfs2_xattr_set(&ip->i_inode, name, value, len, 0, - GFS2_EATYPE_SECURITY); - kfree(value); - kfree(name); - return err; } /** - * gfs2_createi - Create a new inode - * @ghs: An array of two holders - * @name: The name of the new file - * @mode: the permissions on the new inode - * - * @ghs[0] is an initialized holder for the directory - * @ghs[1] is the holder for the inode lock - * - * If the return value is not NULL, the glocks on both the directory and the new - * file are held. A transaction has been started and an inplace reservation - * is held, as well. + * gfs2_create_inode - Create a new inode + * @dir: The parent directory + * @dentry: The new dentry + * @file: If non-NULL, the file which is being opened + * @mode: The permissions on the new inode + * @dev: For device nodes, this is the device number + * @symname: For symlinks, this is the link destination + * @size: The initial size of the inode (ignored for directories) * - * Returns: An inode + * Returns: 0 on success, or error code */ -struct inode *gfs2_createi(struct gfs2_holder *ghs, const struct qstr *name, - unsigned int mode, dev_t dev) +static int gfs2_create_inode(struct inode *dir, struct dentry *dentry, + struct file *file, + umode_t mode, dev_t dev, const char *symname, + unsigned int size, int excl, int *opened) { + const struct qstr *name = &dentry->d_name; + struct posix_acl *default_acl, *acl; + struct gfs2_holder ghs[2]; struct inode *inode = NULL; - struct gfs2_inode *dip = ghs->gh_gl->gl_object; - struct inode *dir = &dip->i_inode; + struct gfs2_inode *dip = GFS2_I(dir), *ip; struct gfs2_sbd *sdp = GFS2_SB(&dip->i_inode); - struct gfs2_inum_host inum = { .no_addr = 0, .no_formal_ino = 0 }; - int error; - u64 generation; - struct buffer_head *bh = NULL; + struct gfs2_glock *io_gl; + struct dentry *d; + int error, free_vfs_inode = 0; + u32 aflags = 0; + unsigned blocks = 1; + struct gfs2_diradd da = { .bh = NULL, }; if (!name->len || name->len > GFS2_FNAMESIZE) - return ERR_PTR(-ENAMETOOLONG); + return -ENAMETOOLONG; + + error = gfs2_rs_alloc(dip); + if (error) + return error; - gfs2_holder_reinit(LM_ST_EXCLUSIVE, 0, ghs); - error = gfs2_glock_nq(ghs); + error = gfs2_rindex_update(sdp); + if (error) + return error; + + error = gfs2_glock_nq_init(dip->i_gl, LM_ST_EXCLUSIVE, 0, ghs); if (error) goto fail; @@ -827,72 +621,1036 @@ struct inode *gfs2_createi(struct gfs2_holder *ghs, const struct qstr *name, if (error) goto fail_gunlock; - error = alloc_dinode(dip, &inum.no_addr, &generation); - if (error) + inode = gfs2_dir_search(dir, &dentry->d_name, !S_ISREG(mode) || excl); + error = PTR_ERR(inode); + if (!IS_ERR(inode)) { + d = d_splice_alias(inode, dentry); + error = PTR_ERR(d); + if (IS_ERR(d)) + goto fail_gunlock; + error = 0; + if (file) { + if (S_ISREG(inode->i_mode)) { + WARN_ON(d != NULL); + error = finish_open(file, dentry, gfs2_open_common, opened); + } else { + error = finish_no_open(file, d); + } + } else { + dput(d); + } + gfs2_glock_dq_uninit(ghs); + return error; + } else if (error != -ENOENT) { + goto fail_gunlock; + } + + error = gfs2_diradd_alloc_required(dir, name, &da); + if (error < 0) goto fail_gunlock; - inum.no_formal_ino = generation; - error = gfs2_glock_nq_num(sdp, inum.no_addr, &gfs2_inode_glops, - LM_ST_EXCLUSIVE, GL_SKIP, ghs + 1); - if (error) + inode = new_inode(sdp->sd_vfs); + error = -ENOMEM; + if (!inode) goto fail_gunlock; - error = make_dinode(dip, ghs[1].gh_gl, mode, &inum, &generation, dev, &bh); + error = posix_acl_create(dir, &mode, &default_acl, &acl); if (error) - goto fail_gunlock2; + goto fail_free_vfs_inode; - inode = gfs2_inode_lookup(dir->i_sb, IF2DT(mode), inum.no_addr, - inum.no_formal_ino); - if (IS_ERR(inode)) - goto fail_gunlock2; + ip = GFS2_I(inode); + error = gfs2_rs_alloc(ip); + if (error) + goto fail_free_acls; + + inode->i_mode = mode; + set_nlink(inode, S_ISDIR(mode) ? 2 : 1); + inode->i_rdev = dev; + inode->i_size = size; + inode->i_atime = inode->i_mtime = inode->i_ctime = CURRENT_TIME; + gfs2_set_inode_blocks(inode, 1); + munge_mode_uid_gid(dip, inode); + ip->i_goal = dip->i_goal; + ip->i_diskflags = 0; + ip->i_eattr = 0; + ip->i_height = 0; + ip->i_depth = 0; + ip->i_entries = 0; + + switch(mode & S_IFMT) { + case S_IFREG: + if ((dip->i_diskflags & GFS2_DIF_INHERIT_JDATA) || + gfs2_tune_get(sdp, gt_new_files_jdata)) + ip->i_diskflags |= GFS2_DIF_JDATA; + gfs2_set_aops(inode); + break; + case S_IFDIR: + ip->i_diskflags |= (dip->i_diskflags & GFS2_DIF_INHERIT_JDATA); + ip->i_diskflags |= GFS2_DIF_JDATA; + ip->i_entries = 2; + break; + } + gfs2_set_inode_flags(inode); + + if ((GFS2_I(sdp->sd_root_dir->d_inode) == dip) || + (dip->i_diskflags & GFS2_DIF_TOPDIR)) + aflags |= GFS2_AF_ORLOV; - error = gfs2_inode_refresh(GFS2_I(inode)); + if (default_acl || acl) + blocks++; + + error = alloc_dinode(ip, aflags, &blocks); if (error) - goto fail_gunlock2; + goto fail_free_inode; + + gfs2_set_inode_blocks(inode, blocks); + + error = gfs2_glock_get(sdp, ip->i_no_addr, &gfs2_inode_glops, CREATE, &ip->i_gl); + if (error) + goto fail_free_inode; - error = gfs2_acl_create(dip, inode); + ip->i_gl->gl_object = ip; + error = gfs2_glock_nq_init(ip->i_gl, LM_ST_EXCLUSIVE, GL_SKIP, ghs + 1); + if (error) + goto fail_free_inode; + + error = gfs2_trans_begin(sdp, blocks, 0); if (error) goto fail_gunlock2; - error = gfs2_security_init(dip, GFS2_I(inode)); + if (blocks > 1) { + ip->i_eattr = ip->i_no_addr + 1; + gfs2_init_xattr(ip); + } + init_dinode(dip, ip, symname); + gfs2_trans_end(sdp); + + error = gfs2_glock_get(sdp, ip->i_no_addr, &gfs2_iopen_glops, CREATE, &io_gl); if (error) goto fail_gunlock2; - error = link_dinode(dip, name, GFS2_I(inode)); + error = gfs2_glock_nq_init(io_gl, LM_ST_SHARED, GL_EXACT, &ip->i_iopen_gh); if (error) goto fail_gunlock2; - if (bh) - brelse(bh); - return inode; + ip->i_iopen_gh.gh_gl->gl_object = ip; + gfs2_glock_put(io_gl); + gfs2_set_iop(inode); + insert_inode_hash(inode); + + if (default_acl) { + error = gfs2_set_acl(inode, default_acl, ACL_TYPE_DEFAULT); + posix_acl_release(default_acl); + } + if (acl) { + if (!error) + error = gfs2_set_acl(inode, acl, ACL_TYPE_ACCESS); + posix_acl_release(acl); + } + + if (error) + goto fail_gunlock3; + + error = security_inode_init_security(&ip->i_inode, &dip->i_inode, name, + &gfs2_initxattrs, NULL); + if (error) + goto fail_gunlock3; + + error = link_dinode(dip, name, ip, &da); + if (error) + goto fail_gunlock3; + + mark_inode_dirty(inode); + d_instantiate(dentry, inode); + if (file) { + *opened |= FILE_CREATED; + error = finish_open(file, dentry, gfs2_open_common, opened); + } + gfs2_glock_dq_uninit(ghs); + gfs2_glock_dq_uninit(ghs + 1); + return error; + +fail_gunlock3: + gfs2_glock_dq_uninit(ghs + 1); + if (ip->i_gl) + gfs2_glock_put(ip->i_gl); + goto fail_gunlock; fail_gunlock2: gfs2_glock_dq_uninit(ghs + 1); - if (inode && !IS_ERR(inode)) - iput(inode); +fail_free_inode: + if (ip->i_gl) + gfs2_glock_put(ip->i_gl); + gfs2_rs_delete(ip, NULL); +fail_free_acls: + if (default_acl) + posix_acl_release(default_acl); + if (acl) + posix_acl_release(acl); +fail_free_vfs_inode: + free_vfs_inode = 1; fail_gunlock: - gfs2_glock_dq(ghs); + gfs2_dir_no_add(&da); + gfs2_glock_dq_uninit(ghs); + if (inode && !IS_ERR(inode)) { + clear_nlink(inode); + if (!free_vfs_inode) + mark_inode_dirty(inode); + set_bit(free_vfs_inode ? GIF_FREE_VFS_INODE : GIF_ALLOC_FAILED, + &GFS2_I(inode)->i_flags); + iput(inode); + } fail: - if (bh) - brelse(bh); - return ERR_PTR(error); + return error; } -static int __gfs2_setattr_simple(struct gfs2_inode *ip, struct iattr *attr) +/** + * gfs2_create - Create a file + * @dir: The directory in which to create the file + * @dentry: The dentry of the new file + * @mode: The mode of the new file + * + * Returns: errno + */ + +static int gfs2_create(struct inode *dir, struct dentry *dentry, + umode_t mode, bool excl) { - struct inode *inode = &ip->i_inode; + return gfs2_create_inode(dir, dentry, NULL, S_IFREG | mode, 0, NULL, 0, excl, NULL); +} + +/** + * __gfs2_lookup - Look up a filename in a directory and return its inode + * @dir: The directory inode + * @dentry: The dentry of the new inode + * @file: File to be opened + * @opened: atomic_open flags + * + * + * Returns: errno + */ + +static struct dentry *__gfs2_lookup(struct inode *dir, struct dentry *dentry, + struct file *file, int *opened) +{ + struct inode *inode; + struct dentry *d; + struct gfs2_holder gh; + struct gfs2_glock *gl; + int error; + + inode = gfs2_lookupi(dir, &dentry->d_name, 0); + if (!inode) + return NULL; + if (IS_ERR(inode)) + return ERR_CAST(inode); + + gl = GFS2_I(inode)->i_gl; + error = gfs2_glock_nq_init(gl, LM_ST_SHARED, LM_FLAG_ANY, &gh); + if (error) { + iput(inode); + return ERR_PTR(error); + } + + d = d_splice_alias(inode, dentry); + if (IS_ERR(d)) { + iput(inode); + gfs2_glock_dq_uninit(&gh); + return d; + } + if (file && S_ISREG(inode->i_mode)) + error = finish_open(file, dentry, gfs2_open_common, opened); + + gfs2_glock_dq_uninit(&gh); + if (error) { + dput(d); + return ERR_PTR(error); + } + return d; +} + +static struct dentry *gfs2_lookup(struct inode *dir, struct dentry *dentry, + unsigned flags) +{ + return __gfs2_lookup(dir, dentry, NULL, NULL); +} + +/** + * gfs2_link - Link to a file + * @old_dentry: The inode to link + * @dir: Add link to this directory + * @dentry: The name of the link + * + * Link the inode in "old_dentry" into the directory "dir" with the + * name in "dentry". + * + * Returns: errno + */ + +static int gfs2_link(struct dentry *old_dentry, struct inode *dir, + struct dentry *dentry) +{ + struct gfs2_inode *dip = GFS2_I(dir); + struct gfs2_sbd *sdp = GFS2_SB(dir); + struct inode *inode = old_dentry->d_inode; + struct gfs2_inode *ip = GFS2_I(inode); + struct gfs2_holder ghs[2]; struct buffer_head *dibh; + struct gfs2_diradd da = { .bh = NULL, }; int error; + if (S_ISDIR(inode->i_mode)) + return -EPERM; + + error = gfs2_rs_alloc(dip); + if (error) + return error; + + gfs2_holder_init(dip->i_gl, LM_ST_EXCLUSIVE, 0, ghs); + gfs2_holder_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, ghs + 1); + + error = gfs2_glock_nq(ghs); /* parent */ + if (error) + goto out_parent; + + error = gfs2_glock_nq(ghs + 1); /* child */ + if (error) + goto out_child; + + error = -ENOENT; + if (inode->i_nlink == 0) + goto out_gunlock; + + error = gfs2_permission(dir, MAY_WRITE | MAY_EXEC); + if (error) + goto out_gunlock; + + error = gfs2_dir_check(dir, &dentry->d_name, NULL); + switch (error) { + case -ENOENT: + break; + case 0: + error = -EEXIST; + default: + goto out_gunlock; + } + + error = -EINVAL; + if (!dip->i_inode.i_nlink) + goto out_gunlock; + error = -EFBIG; + if (dip->i_entries == (u32)-1) + goto out_gunlock; + error = -EPERM; + if (IS_IMMUTABLE(inode) || IS_APPEND(inode)) + goto out_gunlock; + error = -EINVAL; + if (!ip->i_inode.i_nlink) + goto out_gunlock; + error = -EMLINK; + if (ip->i_inode.i_nlink == (u32)-1) + goto out_gunlock; + + error = gfs2_diradd_alloc_required(dir, &dentry->d_name, &da); + if (error < 0) + goto out_gunlock; + + if (da.nr_blocks) { + struct gfs2_alloc_parms ap = { .target = da.nr_blocks, }; + error = gfs2_quota_lock_check(dip); + if (error) + goto out_gunlock; + + error = gfs2_inplace_reserve(dip, &ap); + if (error) + goto out_gunlock_q; + + error = gfs2_trans_begin(sdp, gfs2_trans_da_blks(dip, &da, 2), 0); + if (error) + goto out_ipres; + } else { + error = gfs2_trans_begin(sdp, 2 * RES_DINODE + RES_LEAF, 0); + if (error) + goto out_ipres; + } + error = gfs2_meta_inode_buffer(ip, &dibh); if (error) + goto out_end_trans; + + error = gfs2_dir_add(dir, &dentry->d_name, ip, &da); + if (error) + goto out_brelse; + + gfs2_trans_add_meta(ip->i_gl, dibh); + inc_nlink(&ip->i_inode); + ip->i_inode.i_ctime = CURRENT_TIME; + ihold(inode); + d_instantiate(dentry, inode); + mark_inode_dirty(inode); + +out_brelse: + brelse(dibh); +out_end_trans: + gfs2_trans_end(sdp); +out_ipres: + if (da.nr_blocks) + gfs2_inplace_release(dip); +out_gunlock_q: + if (da.nr_blocks) + gfs2_quota_unlock(dip); +out_gunlock: + gfs2_dir_no_add(&da); + gfs2_glock_dq(ghs + 1); +out_child: + gfs2_glock_dq(ghs); +out_parent: + gfs2_holder_uninit(ghs); + gfs2_holder_uninit(ghs + 1); + return error; +} + +/* + * gfs2_unlink_ok - check to see that a inode is still in a directory + * @dip: the directory + * @name: the name of the file + * @ip: the inode + * + * Assumes that the lock on (at least) @dip is held. + * + * Returns: 0 if the parent/child relationship is correct, errno if it isn't + */ + +static int gfs2_unlink_ok(struct gfs2_inode *dip, const struct qstr *name, + const struct gfs2_inode *ip) +{ + int error; + + if (IS_IMMUTABLE(&ip->i_inode) || IS_APPEND(&ip->i_inode)) + return -EPERM; + + if ((dip->i_inode.i_mode & S_ISVTX) && + !uid_eq(dip->i_inode.i_uid, current_fsuid()) && + !uid_eq(ip->i_inode.i_uid, current_fsuid()) && !capable(CAP_FOWNER)) + return -EPERM; + + if (IS_APPEND(&dip->i_inode)) + return -EPERM; + + error = gfs2_permission(&dip->i_inode, MAY_WRITE | MAY_EXEC); + if (error) return error; - setattr_copy(inode, attr); + error = gfs2_dir_check(&dip->i_inode, name, ip); + if (error) + return error; + + return 0; +} + +/** + * gfs2_unlink_inode - Removes an inode from its parent dir and unlinks it + * @dip: The parent directory + * @name: The name of the entry in the parent directory + * @inode: The inode to be removed + * + * Called with all the locks and in a transaction. This will only be + * called for a directory after it has been checked to ensure it is empty. + * + * Returns: 0 on success, or an error + */ + +static int gfs2_unlink_inode(struct gfs2_inode *dip, + const struct dentry *dentry) +{ + struct inode *inode = dentry->d_inode; + struct gfs2_inode *ip = GFS2_I(inode); + int error; + + error = gfs2_dir_del(dip, dentry); + if (error) + return error; + + ip->i_entries = 0; + inode->i_ctime = CURRENT_TIME; + if (S_ISDIR(inode->i_mode)) + clear_nlink(inode); + else + drop_nlink(inode); mark_inode_dirty(inode); - gfs2_trans_add_bh(ip->i_gl, dibh, 1); - gfs2_dinode_out(ip, dibh->b_data); + if (inode->i_nlink == 0) + gfs2_unlink_di(inode); + return 0; +} + + +/** + * gfs2_unlink - Unlink an inode (this does rmdir as well) + * @dir: The inode of the directory containing the inode to unlink + * @dentry: The file itself + * + * This routine uses the type of the inode as a flag to figure out + * whether this is an unlink or an rmdir. + * + * Returns: errno + */ + +static int gfs2_unlink(struct inode *dir, struct dentry *dentry) +{ + struct gfs2_inode *dip = GFS2_I(dir); + struct gfs2_sbd *sdp = GFS2_SB(dir); + struct inode *inode = dentry->d_inode; + struct gfs2_inode *ip = GFS2_I(inode); + struct gfs2_holder ghs[3]; + struct gfs2_rgrpd *rgd; + int error; + + error = gfs2_rindex_update(sdp); + if (error) + return error; + + error = -EROFS; + + gfs2_holder_init(dip->i_gl, LM_ST_EXCLUSIVE, 0, ghs); + gfs2_holder_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, ghs + 1); + + rgd = gfs2_blk2rgrpd(sdp, ip->i_no_addr, 1); + if (!rgd) + goto out_inodes; + + gfs2_holder_init(rgd->rd_gl, LM_ST_EXCLUSIVE, 0, ghs + 2); + + + error = gfs2_glock_nq(ghs); /* parent */ + if (error) + goto out_parent; + + error = gfs2_glock_nq(ghs + 1); /* child */ + if (error) + goto out_child; + + error = -ENOENT; + if (inode->i_nlink == 0) + goto out_rgrp; + + if (S_ISDIR(inode->i_mode)) { + error = -ENOTEMPTY; + if (ip->i_entries > 2 || inode->i_nlink > 2) + goto out_rgrp; + } + + error = gfs2_glock_nq(ghs + 2); /* rgrp */ + if (error) + goto out_rgrp; + + error = gfs2_unlink_ok(dip, &dentry->d_name, ip); + if (error) + goto out_gunlock; + + error = gfs2_trans_begin(sdp, 2*RES_DINODE + 3*RES_LEAF + RES_RG_BIT, 0); + if (error) + goto out_end_trans; + + error = gfs2_unlink_inode(dip, dentry); + +out_end_trans: + gfs2_trans_end(sdp); +out_gunlock: + gfs2_glock_dq(ghs + 2); +out_rgrp: + gfs2_glock_dq(ghs + 1); +out_child: + gfs2_glock_dq(ghs); +out_parent: + gfs2_holder_uninit(ghs + 2); +out_inodes: + gfs2_holder_uninit(ghs + 1); + gfs2_holder_uninit(ghs); + return error; +} + +/** + * gfs2_symlink - Create a symlink + * @dir: The directory to create the symlink in + * @dentry: The dentry to put the symlink in + * @symname: The thing which the link points to + * + * Returns: errno + */ + +static int gfs2_symlink(struct inode *dir, struct dentry *dentry, + const char *symname) +{ + struct gfs2_sbd *sdp = GFS2_SB(dir); + unsigned int size; + + size = strlen(symname); + if (size > sdp->sd_sb.sb_bsize - sizeof(struct gfs2_dinode) - 1) + return -ENAMETOOLONG; + + return gfs2_create_inode(dir, dentry, NULL, S_IFLNK | S_IRWXUGO, 0, symname, size, 0, NULL); +} + +/** + * gfs2_mkdir - Make a directory + * @dir: The parent directory of the new one + * @dentry: The dentry of the new directory + * @mode: The mode of the new directory + * + * Returns: errno + */ + +static int gfs2_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode) +{ + struct gfs2_sbd *sdp = GFS2_SB(dir); + unsigned dsize = sdp->sd_sb.sb_bsize - sizeof(struct gfs2_dinode); + return gfs2_create_inode(dir, dentry, NULL, S_IFDIR | mode, 0, NULL, dsize, 0, NULL); +} + +/** + * gfs2_mknod - Make a special file + * @dir: The directory in which the special file will reside + * @dentry: The dentry of the special file + * @mode: The mode of the special file + * @dev: The device specification of the special file + * + */ + +static int gfs2_mknod(struct inode *dir, struct dentry *dentry, umode_t mode, + dev_t dev) +{ + return gfs2_create_inode(dir, dentry, NULL, mode, dev, NULL, 0, 0, NULL); +} + +/** + * gfs2_atomic_open - Atomically open a file + * @dir: The directory + * @dentry: The proposed new entry + * @file: The proposed new struct file + * @flags: open flags + * @mode: File mode + * @opened: Flag to say whether the file has been opened or not + * + * Returns: error code or 0 for success + */ + +static int gfs2_atomic_open(struct inode *dir, struct dentry *dentry, + struct file *file, unsigned flags, + umode_t mode, int *opened) +{ + struct dentry *d; + bool excl = !!(flags & O_EXCL); + + d = __gfs2_lookup(dir, dentry, file, opened); + if (IS_ERR(d)) + return PTR_ERR(d); + if (d != NULL) + dentry = d; + if (dentry->d_inode) { + if (!(*opened & FILE_OPENED)) { + if (d == NULL) + dget(dentry); + return finish_no_open(file, dentry); + } + dput(d); + return 0; + } + + BUG_ON(d != NULL); + if (!(flags & O_CREAT)) + return -ENOENT; + + return gfs2_create_inode(dir, dentry, file, S_IFREG | mode, 0, NULL, 0, excl, opened); +} + +/* + * gfs2_ok_to_move - check if it's ok to move a directory to another directory + * @this: move this + * @to: to here + * + * Follow @to back to the root and make sure we don't encounter @this + * Assumes we already hold the rename lock. + * + * Returns: errno + */ + +static int gfs2_ok_to_move(struct gfs2_inode *this, struct gfs2_inode *to) +{ + struct inode *dir = &to->i_inode; + struct super_block *sb = dir->i_sb; + struct inode *tmp; + int error = 0; + + igrab(dir); + + for (;;) { + if (dir == &this->i_inode) { + error = -EINVAL; + break; + } + if (dir == sb->s_root->d_inode) { + error = 0; + break; + } + + tmp = gfs2_lookupi(dir, &gfs2_qdotdot, 1); + if (!tmp) { + error = -ENOENT; + break; + } + if (IS_ERR(tmp)) { + error = PTR_ERR(tmp); + break; + } + + iput(dir); + dir = tmp; + } + + iput(dir); + + return error; +} + +/** + * gfs2_rename - Rename a file + * @odir: Parent directory of old file name + * @odentry: The old dentry of the file + * @ndir: Parent directory of new file name + * @ndentry: The new dentry of the file + * + * Returns: errno + */ + +static int gfs2_rename(struct inode *odir, struct dentry *odentry, + struct inode *ndir, struct dentry *ndentry) +{ + struct gfs2_inode *odip = GFS2_I(odir); + struct gfs2_inode *ndip = GFS2_I(ndir); + struct gfs2_inode *ip = GFS2_I(odentry->d_inode); + struct gfs2_inode *nip = NULL; + struct gfs2_sbd *sdp = GFS2_SB(odir); + struct gfs2_holder ghs[5], r_gh = { .gh_gl = NULL, }; + struct gfs2_rgrpd *nrgd; + unsigned int num_gh; + int dir_rename = 0; + struct gfs2_diradd da = { .nr_blocks = 0, }; + unsigned int x; + int error; + + if (ndentry->d_inode) { + nip = GFS2_I(ndentry->d_inode); + if (ip == nip) + return 0; + } + + error = gfs2_rindex_update(sdp); + if (error) + return error; + + error = gfs2_rs_alloc(ndip); + if (error) + return error; + + if (odip != ndip) { + error = gfs2_glock_nq_init(sdp->sd_rename_gl, LM_ST_EXCLUSIVE, + 0, &r_gh); + if (error) + goto out; + + if (S_ISDIR(ip->i_inode.i_mode)) { + dir_rename = 1; + /* don't move a dirctory into it's subdir */ + error = gfs2_ok_to_move(ip, ndip); + if (error) + goto out_gunlock_r; + } + } + + num_gh = 1; + gfs2_holder_init(odip->i_gl, LM_ST_EXCLUSIVE, 0, ghs); + if (odip != ndip) { + gfs2_holder_init(ndip->i_gl, LM_ST_EXCLUSIVE, 0, ghs + num_gh); + num_gh++; + } + gfs2_holder_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, ghs + num_gh); + num_gh++; + + if (nip) { + gfs2_holder_init(nip->i_gl, LM_ST_EXCLUSIVE, 0, ghs + num_gh); + num_gh++; + /* grab the resource lock for unlink flag twiddling + * this is the case of the target file already existing + * so we unlink before doing the rename + */ + nrgd = gfs2_blk2rgrpd(sdp, nip->i_no_addr, 1); + if (nrgd) + gfs2_holder_init(nrgd->rd_gl, LM_ST_EXCLUSIVE, 0, ghs + num_gh++); + } + + for (x = 0; x < num_gh; x++) { + error = gfs2_glock_nq(ghs + x); + if (error) + goto out_gunlock; + } + + error = -ENOENT; + if (ip->i_inode.i_nlink == 0) + goto out_gunlock; + + /* Check out the old directory */ + + error = gfs2_unlink_ok(odip, &odentry->d_name, ip); + if (error) + goto out_gunlock; + + /* Check out the new directory */ + + if (nip) { + error = gfs2_unlink_ok(ndip, &ndentry->d_name, nip); + if (error) + goto out_gunlock; + + if (nip->i_inode.i_nlink == 0) { + error = -EAGAIN; + goto out_gunlock; + } + + if (S_ISDIR(nip->i_inode.i_mode)) { + if (nip->i_entries < 2) { + gfs2_consist_inode(nip); + error = -EIO; + goto out_gunlock; + } + if (nip->i_entries > 2) { + error = -ENOTEMPTY; + goto out_gunlock; + } + } + } else { + error = gfs2_permission(ndir, MAY_WRITE | MAY_EXEC); + if (error) + goto out_gunlock; + + error = gfs2_dir_check(ndir, &ndentry->d_name, NULL); + switch (error) { + case -ENOENT: + error = 0; + break; + case 0: + error = -EEXIST; + default: + goto out_gunlock; + }; + + if (odip != ndip) { + if (!ndip->i_inode.i_nlink) { + error = -ENOENT; + goto out_gunlock; + } + if (ndip->i_entries == (u32)-1) { + error = -EFBIG; + goto out_gunlock; + } + if (S_ISDIR(ip->i_inode.i_mode) && + ndip->i_inode.i_nlink == (u32)-1) { + error = -EMLINK; + goto out_gunlock; + } + } + } + + /* Check out the dir to be renamed */ + + if (dir_rename) { + error = gfs2_permission(odentry->d_inode, MAY_WRITE); + if (error) + goto out_gunlock; + } + + if (nip == NULL) { + error = gfs2_diradd_alloc_required(ndir, &ndentry->d_name, &da); + if (error) + goto out_gunlock; + } + + if (da.nr_blocks) { + struct gfs2_alloc_parms ap = { .target = da.nr_blocks, }; + error = gfs2_quota_lock_check(ndip); + if (error) + goto out_gunlock; + + error = gfs2_inplace_reserve(ndip, &ap); + if (error) + goto out_gunlock_q; + + error = gfs2_trans_begin(sdp, gfs2_trans_da_blks(ndip, &da, 4) + + 4 * RES_LEAF + 4, 0); + if (error) + goto out_ipreserv; + } else { + error = gfs2_trans_begin(sdp, 4 * RES_DINODE + + 5 * RES_LEAF + 4, 0); + if (error) + goto out_gunlock; + } + + /* Remove the target file, if it exists */ + + if (nip) + error = gfs2_unlink_inode(ndip, ndentry); + + if (dir_rename) { + error = gfs2_dir_mvino(ip, &gfs2_qdotdot, ndip, DT_DIR); + if (error) + goto out_end_trans; + } else { + struct buffer_head *dibh; + error = gfs2_meta_inode_buffer(ip, &dibh); + if (error) + goto out_end_trans; + ip->i_inode.i_ctime = CURRENT_TIME; + gfs2_trans_add_meta(ip->i_gl, dibh); + gfs2_dinode_out(ip, dibh->b_data); + brelse(dibh); + } + + error = gfs2_dir_del(odip, odentry); + if (error) + goto out_end_trans; + + error = gfs2_dir_add(ndir, &ndentry->d_name, ip, &da); + if (error) + goto out_end_trans; + +out_end_trans: + gfs2_trans_end(sdp); +out_ipreserv: + if (da.nr_blocks) + gfs2_inplace_release(ndip); +out_gunlock_q: + if (da.nr_blocks) + gfs2_quota_unlock(ndip); +out_gunlock: + gfs2_dir_no_add(&da); + while (x--) { + gfs2_glock_dq(ghs + x); + gfs2_holder_uninit(ghs + x); + } +out_gunlock_r: + if (r_gh.gh_gl) + gfs2_glock_dq_uninit(&r_gh); +out: + return error; +} + +/** + * gfs2_follow_link - Follow a symbolic link + * @dentry: The dentry of the link + * @nd: Data that we pass to vfs_follow_link() + * + * This can handle symlinks of any size. + * + * Returns: 0 on success or error code + */ + +static void *gfs2_follow_link(struct dentry *dentry, struct nameidata *nd) +{ + struct gfs2_inode *ip = GFS2_I(dentry->d_inode); + struct gfs2_holder i_gh; + struct buffer_head *dibh; + unsigned int size; + char *buf; + int error; + + gfs2_holder_init(ip->i_gl, LM_ST_SHARED, 0, &i_gh); + error = gfs2_glock_nq(&i_gh); + if (error) { + gfs2_holder_uninit(&i_gh); + nd_set_link(nd, ERR_PTR(error)); + return NULL; + } + + size = (unsigned int)i_size_read(&ip->i_inode); + if (size == 0) { + gfs2_consist_inode(ip); + buf = ERR_PTR(-EIO); + goto out; + } + + error = gfs2_meta_inode_buffer(ip, &dibh); + if (error) { + buf = ERR_PTR(error); + goto out; + } + + buf = kzalloc(size + 1, GFP_NOFS); + if (!buf) + buf = ERR_PTR(-ENOMEM); + else + memcpy(buf, dibh->b_data + sizeof(struct gfs2_dinode), size); brelse(dibh); +out: + gfs2_glock_dq_uninit(&i_gh); + nd_set_link(nd, buf); + return NULL; +} + +/** + * gfs2_permission - + * @inode: The inode + * @mask: The mask to be tested + * @flags: Indicates whether this is an RCU path walk or not + * + * This may be called from the VFS directly, or from within GFS2 with the + * inode locked, so we look to see if the glock is already locked and only + * lock the glock if its not already been done. + * + * Returns: errno + */ + +int gfs2_permission(struct inode *inode, int mask) +{ + struct gfs2_inode *ip; + struct gfs2_holder i_gh; + struct gfs2_sbd *sdp = GFS2_SB(inode); + int error; + int unlock = 0; + int frozen_root = 0; + + + ip = GFS2_I(inode); + if (gfs2_glock_is_locked_by_me(ip->i_gl) == NULL) { + if (unlikely(gfs2_glock_is_held_excl(sdp->sd_freeze_gl) && + inode == sdp->sd_root_dir->d_inode && + atomic_inc_not_zero(&sdp->sd_frozen_root))) + frozen_root = 1; + else { + if (mask & MAY_NOT_BLOCK) + return -ECHILD; + error = gfs2_glock_nq_init(ip->i_gl, LM_ST_SHARED, LM_FLAG_ANY, &i_gh); + if (error) + return error; + unlock = 1; + } + } + + if ((mask & MAY_WRITE) && IS_IMMUTABLE(inode)) + error = -EACCES; + else + error = generic_permission(inode, mask); + if (unlock) + gfs2_glock_dq_uninit(&i_gh); + else if (frozen_root && atomic_dec_and_test(&sdp->sd_frozen_root)) + wake_up(&sdp->sd_frozen_root_wait); + + return error; +} + +static int __gfs2_setattr_simple(struct inode *inode, struct iattr *attr) +{ + setattr_copy(inode, attr); + mark_inode_dirty(inode); return 0; } @@ -901,81 +1659,339 @@ static int __gfs2_setattr_simple(struct gfs2_inode *ip, struct iattr *attr) * @ip: * @attr: * - * Called with a reference on the vnode. - * * Returns: errno */ -int gfs2_setattr_simple(struct gfs2_inode *ip, struct iattr *attr) +int gfs2_setattr_simple(struct inode *inode, struct iattr *attr) { int error; if (current->journal_info) - return __gfs2_setattr_simple(ip, attr); + return __gfs2_setattr_simple(inode, attr); - error = gfs2_trans_begin(GFS2_SB(&ip->i_inode), RES_DINODE, 0); + error = gfs2_trans_begin(GFS2_SB(inode), RES_DINODE, 0); if (error) return error; - error = __gfs2_setattr_simple(ip, attr); - gfs2_trans_end(GFS2_SB(&ip->i_inode)); + error = __gfs2_setattr_simple(inode, attr); + gfs2_trans_end(GFS2_SB(inode)); return error; } -void gfs2_dinode_out(const struct gfs2_inode *ip, void *buf) -{ - struct gfs2_dinode *str = buf; - - str->di_header.mh_magic = cpu_to_be32(GFS2_MAGIC); - str->di_header.mh_type = cpu_to_be32(GFS2_METATYPE_DI); - str->di_header.mh_format = cpu_to_be32(GFS2_FORMAT_DI); - str->di_num.no_addr = cpu_to_be64(ip->i_no_addr); - str->di_num.no_formal_ino = cpu_to_be64(ip->i_no_formal_ino); - str->di_mode = cpu_to_be32(ip->i_inode.i_mode); - str->di_uid = cpu_to_be32(ip->i_inode.i_uid); - str->di_gid = cpu_to_be32(ip->i_inode.i_gid); - str->di_nlink = cpu_to_be32(ip->i_inode.i_nlink); - str->di_size = cpu_to_be64(i_size_read(&ip->i_inode)); - str->di_blocks = cpu_to_be64(gfs2_get_inode_blocks(&ip->i_inode)); - str->di_atime = cpu_to_be64(ip->i_inode.i_atime.tv_sec); - str->di_mtime = cpu_to_be64(ip->i_inode.i_mtime.tv_sec); - str->di_ctime = cpu_to_be64(ip->i_inode.i_ctime.tv_sec); - - str->di_goal_meta = cpu_to_be64(ip->i_goal); - str->di_goal_data = cpu_to_be64(ip->i_goal); - str->di_generation = cpu_to_be64(ip->i_generation); - - str->di_flags = cpu_to_be32(ip->i_diskflags); - str->di_height = cpu_to_be16(ip->i_height); - str->di_payload_format = cpu_to_be32(S_ISDIR(ip->i_inode.i_mode) && - !(ip->i_diskflags & GFS2_DIF_EXHASH) ? - GFS2_FORMAT_DE : 0); - str->di_depth = cpu_to_be16(ip->i_depth); - str->di_entries = cpu_to_be32(ip->i_entries); - - str->di_eattr = cpu_to_be64(ip->i_eattr); - str->di_atime_nsec = cpu_to_be32(ip->i_inode.i_atime.tv_nsec); - str->di_mtime_nsec = cpu_to_be32(ip->i_inode.i_mtime.tv_nsec); - str->di_ctime_nsec = cpu_to_be32(ip->i_inode.i_ctime.tv_nsec); -} - -void gfs2_dinode_print(const struct gfs2_inode *ip) -{ - printk(KERN_INFO " no_formal_ino = %llu\n", - (unsigned long long)ip->i_no_formal_ino); - printk(KERN_INFO " no_addr = %llu\n", - (unsigned long long)ip->i_no_addr); - printk(KERN_INFO " i_size = %llu\n", - (unsigned long long)i_size_read(&ip->i_inode)); - printk(KERN_INFO " blocks = %llu\n", - (unsigned long long)gfs2_get_inode_blocks(&ip->i_inode)); - printk(KERN_INFO " i_goal = %llu\n", - (unsigned long long)ip->i_goal); - printk(KERN_INFO " i_diskflags = 0x%.8X\n", ip->i_diskflags); - printk(KERN_INFO " i_height = %u\n", ip->i_height); - printk(KERN_INFO " i_depth = %u\n", ip->i_depth); - printk(KERN_INFO " i_entries = %u\n", ip->i_entries); - printk(KERN_INFO " i_eattr = %llu\n", - (unsigned long long)ip->i_eattr); +static int setattr_chown(struct inode *inode, struct iattr *attr) +{ + struct gfs2_inode *ip = GFS2_I(inode); + struct gfs2_sbd *sdp = GFS2_SB(inode); + kuid_t ouid, nuid; + kgid_t ogid, ngid; + int error; + + ouid = inode->i_uid; + ogid = inode->i_gid; + nuid = attr->ia_uid; + ngid = attr->ia_gid; + + if (!(attr->ia_valid & ATTR_UID) || uid_eq(ouid, nuid)) + ouid = nuid = NO_UID_QUOTA_CHANGE; + if (!(attr->ia_valid & ATTR_GID) || gid_eq(ogid, ngid)) + ogid = ngid = NO_GID_QUOTA_CHANGE; + + error = get_write_access(inode); + if (error) + return error; + + error = gfs2_rs_alloc(ip); + if (error) + goto out; + + error = gfs2_rindex_update(sdp); + if (error) + goto out; + + error = gfs2_quota_lock(ip, nuid, ngid); + if (error) + goto out; + + if (!uid_eq(ouid, NO_UID_QUOTA_CHANGE) || + !gid_eq(ogid, NO_GID_QUOTA_CHANGE)) { + error = gfs2_quota_check(ip, nuid, ngid); + if (error) + goto out_gunlock_q; + } + + error = gfs2_trans_begin(sdp, RES_DINODE + 2 * RES_QUOTA, 0); + if (error) + goto out_gunlock_q; + + error = gfs2_setattr_simple(inode, attr); + if (error) + goto out_end_trans; + + if (!uid_eq(ouid, NO_UID_QUOTA_CHANGE) || + !gid_eq(ogid, NO_GID_QUOTA_CHANGE)) { + u64 blocks = gfs2_get_inode_blocks(&ip->i_inode); + gfs2_quota_change(ip, -blocks, ouid, ogid); + gfs2_quota_change(ip, blocks, nuid, ngid); + } + +out_end_trans: + gfs2_trans_end(sdp); +out_gunlock_q: + gfs2_quota_unlock(ip); +out: + put_write_access(inode); + return error; +} + +/** + * gfs2_setattr - Change attributes on an inode + * @dentry: The dentry which is changing + * @attr: The structure describing the change + * + * The VFS layer wants to change one or more of an inodes attributes. Write + * that change out to disk. + * + * Returns: errno + */ + +static int gfs2_setattr(struct dentry *dentry, struct iattr *attr) +{ + struct inode *inode = dentry->d_inode; + struct gfs2_inode *ip = GFS2_I(inode); + struct gfs2_holder i_gh; + int error; + + error = gfs2_rs_alloc(ip); + if (error) + return error; + + error = gfs2_glock_nq_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, &i_gh); + if (error) + return error; + + error = -EPERM; + if (IS_IMMUTABLE(inode) || IS_APPEND(inode)) + goto out; + + error = inode_change_ok(inode, attr); + if (error) + goto out; + + if (attr->ia_valid & ATTR_SIZE) + error = gfs2_setattr_size(inode, attr->ia_size); + else if (attr->ia_valid & (ATTR_UID | ATTR_GID)) + error = setattr_chown(inode, attr); + else { + error = gfs2_setattr_simple(inode, attr); + if (!error && attr->ia_valid & ATTR_MODE) + error = posix_acl_chmod(inode, inode->i_mode); + } + +out: + if (!error) + mark_inode_dirty(inode); + gfs2_glock_dq_uninit(&i_gh); + return error; } +/** + * gfs2_getattr - Read out an inode's attributes + * @mnt: The vfsmount the inode is being accessed from + * @dentry: The dentry to stat + * @stat: The inode's stats + * + * This may be called from the VFS directly, or from within GFS2 with the + * inode locked, so we look to see if the glock is already locked and only + * lock the glock if its not already been done. Note that its the NFS + * readdirplus operation which causes this to be called (from filldir) + * with the glock already held. + * + * Returns: errno + */ + +static int gfs2_getattr(struct vfsmount *mnt, struct dentry *dentry, + struct kstat *stat) +{ + struct inode *inode = dentry->d_inode; + struct gfs2_inode *ip = GFS2_I(inode); + struct gfs2_holder gh; + struct gfs2_sbd *sdp = GFS2_SB(inode); + int error; + int unlock = 0; + int frozen_root = 0; + + if (gfs2_glock_is_locked_by_me(ip->i_gl) == NULL) { + if (unlikely(gfs2_glock_is_held_excl(sdp->sd_freeze_gl) && + inode == sdp->sd_root_dir->d_inode && + atomic_inc_not_zero(&sdp->sd_frozen_root))) + frozen_root = 1; + else { + error = gfs2_glock_nq_init(ip->i_gl, LM_ST_SHARED, LM_FLAG_ANY, &gh); + if (error) + return error; + unlock = 1; + } + } + + generic_fillattr(inode, stat); + if (unlock) + gfs2_glock_dq_uninit(&gh); + else if (frozen_root && atomic_dec_and_test(&sdp->sd_frozen_root)) + wake_up(&sdp->sd_frozen_root_wait); + + return 0; +} + +static int gfs2_setxattr(struct dentry *dentry, const char *name, + const void *data, size_t size, int flags) +{ + struct inode *inode = dentry->d_inode; + struct gfs2_inode *ip = GFS2_I(inode); + struct gfs2_holder gh; + int ret; + + gfs2_holder_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, &gh); + ret = gfs2_glock_nq(&gh); + if (ret == 0) { + ret = gfs2_rs_alloc(ip); + if (ret == 0) + ret = generic_setxattr(dentry, name, data, size, flags); + gfs2_glock_dq(&gh); + } + gfs2_holder_uninit(&gh); + return ret; +} + +static ssize_t gfs2_getxattr(struct dentry *dentry, const char *name, + void *data, size_t size) +{ + struct inode *inode = dentry->d_inode; + struct gfs2_inode *ip = GFS2_I(inode); + struct gfs2_holder gh; + int ret; + + /* For selinux during lookup */ + if (gfs2_glock_is_locked_by_me(ip->i_gl)) + return generic_getxattr(dentry, name, data, size); + + gfs2_holder_init(ip->i_gl, LM_ST_SHARED, LM_FLAG_ANY, &gh); + ret = gfs2_glock_nq(&gh); + if (ret == 0) { + ret = generic_getxattr(dentry, name, data, size); + gfs2_glock_dq(&gh); + } + gfs2_holder_uninit(&gh); + return ret; +} + +static int gfs2_removexattr(struct dentry *dentry, const char *name) +{ + struct inode *inode = dentry->d_inode; + struct gfs2_inode *ip = GFS2_I(inode); + struct gfs2_holder gh; + int ret; + + gfs2_holder_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, &gh); + ret = gfs2_glock_nq(&gh); + if (ret == 0) { + ret = gfs2_rs_alloc(ip); + if (ret == 0) + ret = generic_removexattr(dentry, name); + gfs2_glock_dq(&gh); + } + gfs2_holder_uninit(&gh); + return ret; +} + +static int gfs2_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, + u64 start, u64 len) +{ + struct gfs2_inode *ip = GFS2_I(inode); + struct gfs2_holder gh; + int ret; + + ret = fiemap_check_flags(fieinfo, FIEMAP_FLAG_SYNC); + if (ret) + return ret; + + mutex_lock(&inode->i_mutex); + + ret = gfs2_glock_nq_init(ip->i_gl, LM_ST_SHARED, 0, &gh); + if (ret) + goto out; + + if (gfs2_is_stuffed(ip)) { + u64 phys = ip->i_no_addr << inode->i_blkbits; + u64 size = i_size_read(inode); + u32 flags = FIEMAP_EXTENT_LAST|FIEMAP_EXTENT_NOT_ALIGNED| + FIEMAP_EXTENT_DATA_INLINE; + phys += sizeof(struct gfs2_dinode); + phys += start; + if (start + len > size) + len = size - start; + if (start < size) + ret = fiemap_fill_next_extent(fieinfo, start, phys, + len, flags); + if (ret == 1) + ret = 0; + } else { + ret = __generic_block_fiemap(inode, fieinfo, start, len, + gfs2_block_map); + } + + gfs2_glock_dq_uninit(&gh); +out: + mutex_unlock(&inode->i_mutex); + return ret; +} + +const struct inode_operations gfs2_file_iops = { + .permission = gfs2_permission, + .setattr = gfs2_setattr, + .getattr = gfs2_getattr, + .setxattr = gfs2_setxattr, + .getxattr = gfs2_getxattr, + .listxattr = gfs2_listxattr, + .removexattr = gfs2_removexattr, + .fiemap = gfs2_fiemap, + .get_acl = gfs2_get_acl, + .set_acl = gfs2_set_acl, +}; + +const struct inode_operations gfs2_dir_iops = { + .create = gfs2_create, + .lookup = gfs2_lookup, + .link = gfs2_link, + .unlink = gfs2_unlink, + .symlink = gfs2_symlink, + .mkdir = gfs2_mkdir, + .rmdir = gfs2_unlink, + .mknod = gfs2_mknod, + .rename = gfs2_rename, + .permission = gfs2_permission, + .setattr = gfs2_setattr, + .getattr = gfs2_getattr, + .setxattr = gfs2_setxattr, + .getxattr = gfs2_getxattr, + .listxattr = gfs2_listxattr, + .removexattr = gfs2_removexattr, + .fiemap = gfs2_fiemap, + .get_acl = gfs2_get_acl, + .set_acl = gfs2_set_acl, + .atomic_open = gfs2_atomic_open, +}; + +const struct inode_operations gfs2_symlink_iops = { + .readlink = generic_readlink, + .follow_link = gfs2_follow_link, + .put_link = kfree_put_link, + .permission = gfs2_permission, + .setattr = gfs2_setattr, + .getattr = gfs2_getattr, + .setxattr = gfs2_setxattr, + .getxattr = gfs2_getxattr, + .listxattr = gfs2_listxattr, + .removexattr = gfs2_removexattr, + .fiemap = gfs2_fiemap, +}; + diff --git a/fs/gfs2/inode.h b/fs/gfs2/inode.h index 3e00a66e7cb..ba4d9492d42 100644 --- a/fs/gfs2/inode.h +++ b/fs/gfs2/inode.h @@ -17,10 +17,7 @@ extern int gfs2_releasepage(struct page *page, gfp_t gfp_mask); extern int gfs2_internal_read(struct gfs2_inode *ip, - struct file_ra_state *ra_state, char *buf, loff_t *pos, unsigned size); -extern void gfs2_page_add_databufs(struct gfs2_inode *ip, struct page *page, - unsigned int from, unsigned int to); extern void gfs2_set_aops(struct inode *inode); static inline int gfs2_is_stuffed(const struct gfs2_inode *ip) @@ -97,26 +94,22 @@ err: } extern struct inode *gfs2_inode_lookup(struct super_block *sb, unsigned type, - u64 no_addr, u64 no_formal_ino); + u64 no_addr, u64 no_formal_ino, + int non_block); extern struct inode *gfs2_lookup_by_inum(struct gfs2_sbd *sdp, u64 no_addr, u64 *no_formal_ino, unsigned int blktype); -extern struct inode *gfs2_ilookup(struct super_block *sb, u64 no_addr); +extern struct inode *gfs2_ilookup(struct super_block *sb, u64 no_addr, int nonblock); extern int gfs2_inode_refresh(struct gfs2_inode *ip); -extern int gfs2_dinode_dealloc(struct gfs2_inode *inode); -extern int gfs2_change_nlink(struct gfs2_inode *ip, int diff); extern struct inode *gfs2_lookupi(struct inode *dir, const struct qstr *name, int is_root); -extern struct inode *gfs2_createi(struct gfs2_holder *ghs, - const struct qstr *name, - unsigned int mode, dev_t dev); -extern int gfs2_permission(struct inode *inode, int mask, unsigned int flags); -extern int gfs2_setattr_simple(struct gfs2_inode *ip, struct iattr *attr); +extern int gfs2_permission(struct inode *inode, int mask); +extern int gfs2_setattr_simple(struct inode *inode, struct iattr *attr); extern struct inode *gfs2_lookup_simple(struct inode *dip, const char *name); extern void gfs2_dinode_out(const struct gfs2_inode *ip, void *buf); -extern void gfs2_dinode_print(const struct gfs2_inode *ip); +extern int gfs2_open_common(struct inode *inode, struct file *file); extern const struct inode_operations gfs2_file_iops; extern const struct inode_operations gfs2_dir_iops; diff --git a/fs/gfs2/lock_dlm.c b/fs/gfs2/lock_dlm.c index 6e493aee28f..4fafea1c9ec 100644 --- a/fs/gfs2/lock_dlm.c +++ b/fs/gfs2/lock_dlm.c @@ -1,42 +1,133 @@ /* * Copyright (C) Sistina Software, Inc. 1997-2003 All rights reserved. - * Copyright (C) 2004-2009 Red Hat, Inc. All rights reserved. + * Copyright 2004-2011 Red Hat, Inc. * * This copyrighted material is made available to anyone wishing to use, * modify, copy, or redistribute it subject to the terms and conditions * of the GNU General Public License version 2. */ +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + #include <linux/fs.h> #include <linux/dlm.h> #include <linux/slab.h> #include <linux/types.h> +#include <linux/delay.h> #include <linux/gfs2_ondisk.h> #include "incore.h" #include "glock.h" #include "util.h" +#include "sys.h" +#include "trace_gfs2.h" + +extern struct workqueue_struct *gfs2_control_wq; + +/** + * gfs2_update_stats - Update time based stats + * @mv: Pointer to mean/variance structure to update + * @sample: New data to include + * + * @delta is the difference between the current rtt sample and the + * running average srtt. We add 1/8 of that to the srtt in order to + * update the current srtt estimate. The varience estimate is a bit + * more complicated. We subtract the abs value of the @delta from + * the current variance estimate and add 1/4 of that to the running + * total. + * + * Note that the index points at the array entry containing the smoothed + * mean value, and the variance is always in the following entry + * + * Reference: TCP/IP Illustrated, vol 2, p. 831,832 + * All times are in units of integer nanoseconds. Unlike the TCP/IP case, + * they are not scaled fixed point. + */ + +static inline void gfs2_update_stats(struct gfs2_lkstats *s, unsigned index, + s64 sample) +{ + s64 delta = sample - s->stats[index]; + s->stats[index] += (delta >> 3); + index++; + s->stats[index] += ((abs64(delta) - s->stats[index]) >> 2); +} +/** + * gfs2_update_reply_times - Update locking statistics + * @gl: The glock to update + * + * This assumes that gl->gl_dstamp has been set earlier. + * + * The rtt (lock round trip time) is an estimate of the time + * taken to perform a dlm lock request. We update it on each + * reply from the dlm. + * + * The blocking flag is set on the glock for all dlm requests + * which may potentially block due to lock requests from other nodes. + * DLM requests where the current lock state is exclusive, the + * requested state is null (or unlocked) or where the TRY or + * TRY_1CB flags are set are classified as non-blocking. All + * other DLM requests are counted as (potentially) blocking. + */ +static inline void gfs2_update_reply_times(struct gfs2_glock *gl) +{ + struct gfs2_pcpu_lkstats *lks; + const unsigned gltype = gl->gl_name.ln_type; + unsigned index = test_bit(GLF_BLOCKING, &gl->gl_flags) ? + GFS2_LKS_SRTTB : GFS2_LKS_SRTT; + s64 rtt; + + preempt_disable(); + rtt = ktime_to_ns(ktime_sub(ktime_get_real(), gl->gl_dstamp)); + lks = this_cpu_ptr(gl->gl_sbd->sd_lkstats); + gfs2_update_stats(&gl->gl_stats, index, rtt); /* Local */ + gfs2_update_stats(&lks->lkstats[gltype], index, rtt); /* Global */ + preempt_enable(); + + trace_gfs2_glock_lock_time(gl, rtt); +} + +/** + * gfs2_update_request_times - Update locking statistics + * @gl: The glock to update + * + * The irt (lock inter-request times) measures the average time + * between requests to the dlm. It is updated immediately before + * each dlm call. + */ + +static inline void gfs2_update_request_times(struct gfs2_glock *gl) +{ + struct gfs2_pcpu_lkstats *lks; + const unsigned gltype = gl->gl_name.ln_type; + ktime_t dstamp; + s64 irt; + preempt_disable(); + dstamp = gl->gl_dstamp; + gl->gl_dstamp = ktime_get_real(); + irt = ktime_to_ns(ktime_sub(gl->gl_dstamp, dstamp)); + lks = this_cpu_ptr(gl->gl_sbd->sd_lkstats); + gfs2_update_stats(&gl->gl_stats, GFS2_LKS_SIRT, irt); /* Local */ + gfs2_update_stats(&lks->lkstats[gltype], GFS2_LKS_SIRT, irt); /* Global */ + preempt_enable(); +} + static void gdlm_ast(void *arg) { struct gfs2_glock *gl = arg; unsigned ret = gl->gl_state; - struct gfs2_sbd *sdp = gl->gl_sbd; + gfs2_update_reply_times(gl); BUG_ON(gl->gl_lksb.sb_flags & DLM_SBF_DEMOTED); - if (gl->gl_lksb.sb_flags & DLM_SBF_VALNOTVALID) - memset(gl->gl_lvb, 0, GDLM_LVB_SIZE); + if ((gl->gl_lksb.sb_flags & DLM_SBF_VALNOTVALID) && gl->gl_lksb.sb_lvbptr) + memset(gl->gl_lksb.sb_lvbptr, 0, GDLM_LVB_SIZE); switch (gl->gl_lksb.sb_status) { case -DLM_EUNLOCK: /* Unlocked, so glock can be freed */ - if (gl->gl_ops->go_flags & GLOF_ASPACE) - kmem_cache_free(gfs2_glock_aspace_cachep, gl); - else - kmem_cache_free(gfs2_glock_cachep, gl); - if (atomic_dec_and_test(&sdp->sd_glock_disposal)) - wake_up(&sdp->sd_glock_wait); + gfs2_glock_free(gl); return; case -DLM_ECANCEL: /* Cancel while getting lock */ ret |= LM_OUT_CANCELED; @@ -87,7 +178,7 @@ static void gdlm_bast(void *arg, int mode) gfs2_glock_cb(gl, LM_ST_SHARED); break; default: - printk(KERN_ERR "unknown bast mode %d", mode); + pr_err("unknown bast mode %d\n", mode); BUG(); } } @@ -106,16 +197,19 @@ static int make_mode(const unsigned int lmstate) case LM_ST_SHARED: return DLM_LOCK_PR; } - printk(KERN_ERR "unknown LM state %d", lmstate); + pr_err("unknown LM state %d\n", lmstate); BUG(); return -1; } -static u32 make_flags(const u32 lkid, const unsigned int gfs_flags, +static u32 make_flags(struct gfs2_glock *gl, const unsigned int gfs_flags, const int req) { u32 lkf = 0; + if (gl->gl_lksb.sb_lvbptr) + lkf |= DLM_LKF_VALBLK; + if (gfs_flags & LM_FLAG_TRY) lkf |= DLM_LKF_NOQUEUE; @@ -138,49 +232,85 @@ static u32 make_flags(const u32 lkid, const unsigned int gfs_flags, BUG(); } - if (lkid != 0) + if (gl->gl_lksb.sb_lkid != 0) { lkf |= DLM_LKF_CONVERT; - - lkf |= DLM_LKF_VALBLK; + if (test_bit(GLF_BLOCKING, &gl->gl_flags)) + lkf |= DLM_LKF_QUECVT; + } return lkf; } +static void gfs2_reverse_hex(char *c, u64 value) +{ + *c = '0'; + while (value) { + *c-- = hex_asc[value & 0x0f]; + value >>= 4; + } +} + static int gdlm_lock(struct gfs2_glock *gl, unsigned int req_state, unsigned int flags) { struct lm_lockstruct *ls = &gl->gl_sbd->sd_lockstruct; int req; u32 lkf; + char strname[GDLM_STRNAME_BYTES] = ""; req = make_mode(req_state); - lkf = make_flags(gl->gl_lksb.sb_lkid, flags, req); - + lkf = make_flags(gl, flags, req); + gfs2_glstats_inc(gl, GFS2_LKS_DCOUNT); + gfs2_sbstats_inc(gl, GFS2_LKS_DCOUNT); + if (gl->gl_lksb.sb_lkid) { + gfs2_update_request_times(gl); + } else { + memset(strname, ' ', GDLM_STRNAME_BYTES - 1); + strname[GDLM_STRNAME_BYTES - 1] = '\0'; + gfs2_reverse_hex(strname + 7, gl->gl_name.ln_type); + gfs2_reverse_hex(strname + 23, gl->gl_name.ln_number); + gl->gl_dstamp = ktime_get_real(); + } /* * Submit the actual lock request. */ - return dlm_lock(ls->ls_dlm, req, &gl->gl_lksb, lkf, gl->gl_strname, + return dlm_lock(ls->ls_dlm, req, &gl->gl_lksb, lkf, strname, GDLM_STRNAME_BYTES - 1, 0, gdlm_ast, gl, gdlm_bast); } -static void gdlm_put_lock(struct kmem_cache *cachep, struct gfs2_glock *gl) +static void gdlm_put_lock(struct gfs2_glock *gl) { struct gfs2_sbd *sdp = gl->gl_sbd; struct lm_lockstruct *ls = &sdp->sd_lockstruct; + int lvb_needs_unlock = 0; int error; if (gl->gl_lksb.sb_lkid == 0) { - kmem_cache_free(cachep, gl); - if (atomic_dec_and_test(&sdp->sd_glock_disposal)) - wake_up(&sdp->sd_glock_wait); + gfs2_glock_free(gl); + return; + } + + clear_bit(GLF_BLOCKING, &gl->gl_flags); + gfs2_glstats_inc(gl, GFS2_LKS_DCOUNT); + gfs2_sbstats_inc(gl, GFS2_LKS_DCOUNT); + gfs2_update_request_times(gl); + + /* don't want to skip dlm_unlock writing the lvb when lock is ex */ + + if (gl->gl_lksb.sb_lvbptr && (gl->gl_state == LM_ST_EXCLUSIVE)) + lvb_needs_unlock = 1; + + if (test_bit(SDF_SKIP_DLM_UNLOCK, &sdp->sd_flags) && + !lvb_needs_unlock) { + gfs2_glock_free(gl); return; } error = dlm_unlock(ls->ls_dlm, gl->gl_lksb.sb_lkid, DLM_LKF_VALBLK, NULL, gl); if (error) { - printk(KERN_ERR "gdlm_unlock %x,%llx err=%d\n", + pr_err("gdlm_unlock %x,%llx err=%d\n", gl->gl_name.ln_type, (unsigned long long)gl->gl_name.ln_number, error); return; @@ -193,34 +323,1001 @@ static void gdlm_cancel(struct gfs2_glock *gl) dlm_unlock(ls->ls_dlm, gl->gl_lksb.sb_lkid, DLM_LKF_CANCEL, NULL, gl); } -static int gdlm_mount(struct gfs2_sbd *sdp, const char *fsname) +/* + * dlm/gfs2 recovery coordination using dlm_recover callbacks + * + * 1. dlm_controld sees lockspace members change + * 2. dlm_controld blocks dlm-kernel locking activity + * 3. dlm_controld within dlm-kernel notifies gfs2 (recover_prep) + * 4. dlm_controld starts and finishes its own user level recovery + * 5. dlm_controld starts dlm-kernel dlm_recoverd to do kernel recovery + * 6. dlm_recoverd notifies gfs2 of failed nodes (recover_slot) + * 7. dlm_recoverd does its own lock recovery + * 8. dlm_recoverd unblocks dlm-kernel locking activity + * 9. dlm_recoverd notifies gfs2 when done (recover_done with new generation) + * 10. gfs2_control updates control_lock lvb with new generation and jid bits + * 11. gfs2_control enqueues journals for gfs2_recover to recover (maybe none) + * 12. gfs2_recover dequeues and recovers journals of failed nodes + * 13. gfs2_recover provides recovery results to gfs2_control (recovery_result) + * 14. gfs2_control updates control_lock lvb jid bits for recovered journals + * 15. gfs2_control unblocks normal locking when all journals are recovered + * + * - failures during recovery + * + * recover_prep() may set BLOCK_LOCKS (step 3) again before gfs2_control + * clears BLOCK_LOCKS (step 15), e.g. another node fails while still + * recovering for a prior failure. gfs2_control needs a way to detect + * this so it can leave BLOCK_LOCKS set in step 15. This is managed using + * the recover_block and recover_start values. + * + * recover_done() provides a new lockspace generation number each time it + * is called (step 9). This generation number is saved as recover_start. + * When recover_prep() is called, it sets BLOCK_LOCKS and sets + * recover_block = recover_start. So, while recover_block is equal to + * recover_start, BLOCK_LOCKS should remain set. (recover_spin must + * be held around the BLOCK_LOCKS/recover_block/recover_start logic.) + * + * - more specific gfs2 steps in sequence above + * + * 3. recover_prep sets BLOCK_LOCKS and sets recover_block = recover_start + * 6. recover_slot records any failed jids (maybe none) + * 9. recover_done sets recover_start = new generation number + * 10. gfs2_control sets control_lock lvb = new gen + bits for failed jids + * 12. gfs2_recover does journal recoveries for failed jids identified above + * 14. gfs2_control clears control_lock lvb bits for recovered jids + * 15. gfs2_control checks if recover_block == recover_start (step 3 occured + * again) then do nothing, otherwise if recover_start > recover_block + * then clear BLOCK_LOCKS. + * + * - parallel recovery steps across all nodes + * + * All nodes attempt to update the control_lock lvb with the new generation + * number and jid bits, but only the first to get the control_lock EX will + * do so; others will see that it's already done (lvb already contains new + * generation number.) + * + * . All nodes get the same recover_prep/recover_slot/recover_done callbacks + * . All nodes attempt to set control_lock lvb gen + bits for the new gen + * . One node gets control_lock first and writes the lvb, others see it's done + * . All nodes attempt to recover jids for which they see control_lock bits set + * . One node succeeds for a jid, and that one clears the jid bit in the lvb + * . All nodes will eventually see all lvb bits clear and unblock locks + * + * - is there a problem with clearing an lvb bit that should be set + * and missing a journal recovery? + * + * 1. jid fails + * 2. lvb bit set for step 1 + * 3. jid recovered for step 1 + * 4. jid taken again (new mount) + * 5. jid fails (for step 4) + * 6. lvb bit set for step 5 (will already be set) + * 7. lvb bit cleared for step 3 + * + * This is not a problem because the failure in step 5 does not + * require recovery, because the mount in step 4 could not have + * progressed far enough to unblock locks and access the fs. The + * control_mount() function waits for all recoveries to be complete + * for the latest lockspace generation before ever unblocking locks + * and returning. The mount in step 4 waits until the recovery in + * step 1 is done. + * + * - special case of first mounter: first node to mount the fs + * + * The first node to mount a gfs2 fs needs to check all the journals + * and recover any that need recovery before other nodes are allowed + * to mount the fs. (Others may begin mounting, but they must wait + * for the first mounter to be done before taking locks on the fs + * or accessing the fs.) This has two parts: + * + * 1. The mounted_lock tells a node it's the first to mount the fs. + * Each node holds the mounted_lock in PR while it's mounted. + * Each node tries to acquire the mounted_lock in EX when it mounts. + * If a node is granted the mounted_lock EX it means there are no + * other mounted nodes (no PR locks exist), and it is the first mounter. + * The mounted_lock is demoted to PR when first recovery is done, so + * others will fail to get an EX lock, but will get a PR lock. + * + * 2. The control_lock blocks others in control_mount() while the first + * mounter is doing first mount recovery of all journals. + * A mounting node needs to acquire control_lock in EX mode before + * it can proceed. The first mounter holds control_lock in EX while doing + * the first mount recovery, blocking mounts from other nodes, then demotes + * control_lock to NL when it's done (others_may_mount/first_done), + * allowing other nodes to continue mounting. + * + * first mounter: + * control_lock EX/NOQUEUE success + * mounted_lock EX/NOQUEUE success (no other PR, so no other mounters) + * set first=1 + * do first mounter recovery + * mounted_lock EX->PR + * control_lock EX->NL, write lvb generation + * + * other mounter: + * control_lock EX/NOQUEUE success (if fail -EAGAIN, retry) + * mounted_lock EX/NOQUEUE fail -EAGAIN (expected due to other mounters PR) + * mounted_lock PR/NOQUEUE success + * read lvb generation + * control_lock EX->NL + * set first=0 + * + * - mount during recovery + * + * If a node mounts while others are doing recovery (not first mounter), + * the mounting node will get its initial recover_done() callback without + * having seen any previous failures/callbacks. + * + * It must wait for all recoveries preceding its mount to be finished + * before it unblocks locks. It does this by repeating the "other mounter" + * steps above until the lvb generation number is >= its mount generation + * number (from initial recover_done) and all lvb bits are clear. + * + * - control_lock lvb format + * + * 4 bytes generation number: the latest dlm lockspace generation number + * from recover_done callback. Indicates the jid bitmap has been updated + * to reflect all slot failures through that generation. + * 4 bytes unused. + * GDLM_LVB_SIZE-8 bytes of jid bit map. If bit N is set, it indicates + * that jid N needs recovery. + */ + +#define JID_BITMAP_OFFSET 8 /* 4 byte generation number + 4 byte unused */ + +static void control_lvb_read(struct lm_lockstruct *ls, uint32_t *lvb_gen, + char *lvb_bits) +{ + __le32 gen; + memcpy(lvb_bits, ls->ls_control_lvb, GDLM_LVB_SIZE); + memcpy(&gen, lvb_bits, sizeof(__le32)); + *lvb_gen = le32_to_cpu(gen); +} + +static void control_lvb_write(struct lm_lockstruct *ls, uint32_t lvb_gen, + char *lvb_bits) +{ + __le32 gen; + memcpy(ls->ls_control_lvb, lvb_bits, GDLM_LVB_SIZE); + gen = cpu_to_le32(lvb_gen); + memcpy(ls->ls_control_lvb, &gen, sizeof(__le32)); +} + +static int all_jid_bits_clear(char *lvb) +{ + return !memchr_inv(lvb + JID_BITMAP_OFFSET, 0, + GDLM_LVB_SIZE - JID_BITMAP_OFFSET); +} + +static void sync_wait_cb(void *arg) +{ + struct lm_lockstruct *ls = arg; + complete(&ls->ls_sync_wait); +} + +static int sync_unlock(struct gfs2_sbd *sdp, struct dlm_lksb *lksb, char *name) { struct lm_lockstruct *ls = &sdp->sd_lockstruct; int error; - if (fsname == NULL) { - fs_info(sdp, "no fsname found\n"); - return -EINVAL; + error = dlm_unlock(ls->ls_dlm, lksb->sb_lkid, 0, lksb, ls); + if (error) { + fs_err(sdp, "%s lkid %x error %d\n", + name, lksb->sb_lkid, error); + return error; + } + + wait_for_completion(&ls->ls_sync_wait); + + if (lksb->sb_status != -DLM_EUNLOCK) { + fs_err(sdp, "%s lkid %x status %d\n", + name, lksb->sb_lkid, lksb->sb_status); + return -1; + } + return 0; +} + +static int sync_lock(struct gfs2_sbd *sdp, int mode, uint32_t flags, + unsigned int num, struct dlm_lksb *lksb, char *name) +{ + struct lm_lockstruct *ls = &sdp->sd_lockstruct; + char strname[GDLM_STRNAME_BYTES]; + int error, status; + + memset(strname, 0, GDLM_STRNAME_BYTES); + snprintf(strname, GDLM_STRNAME_BYTES, "%8x%16x", LM_TYPE_NONDISK, num); + + error = dlm_lock(ls->ls_dlm, mode, lksb, flags, + strname, GDLM_STRNAME_BYTES - 1, + 0, sync_wait_cb, ls, NULL); + if (error) { + fs_err(sdp, "%s lkid %x flags %x mode %d error %d\n", + name, lksb->sb_lkid, flags, mode, error); + return error; + } + + wait_for_completion(&ls->ls_sync_wait); + + status = lksb->sb_status; + + if (status && status != -EAGAIN) { + fs_err(sdp, "%s lkid %x flags %x mode %d status %d\n", + name, lksb->sb_lkid, flags, mode, status); + } + + return status; +} + +static int mounted_unlock(struct gfs2_sbd *sdp) +{ + struct lm_lockstruct *ls = &sdp->sd_lockstruct; + return sync_unlock(sdp, &ls->ls_mounted_lksb, "mounted_lock"); +} + +static int mounted_lock(struct gfs2_sbd *sdp, int mode, uint32_t flags) +{ + struct lm_lockstruct *ls = &sdp->sd_lockstruct; + return sync_lock(sdp, mode, flags, GFS2_MOUNTED_LOCK, + &ls->ls_mounted_lksb, "mounted_lock"); +} + +static int control_unlock(struct gfs2_sbd *sdp) +{ + struct lm_lockstruct *ls = &sdp->sd_lockstruct; + return sync_unlock(sdp, &ls->ls_control_lksb, "control_lock"); +} + +static int control_lock(struct gfs2_sbd *sdp, int mode, uint32_t flags) +{ + struct lm_lockstruct *ls = &sdp->sd_lockstruct; + return sync_lock(sdp, mode, flags, GFS2_CONTROL_LOCK, + &ls->ls_control_lksb, "control_lock"); +} + +static void gfs2_control_func(struct work_struct *work) +{ + struct gfs2_sbd *sdp = container_of(work, struct gfs2_sbd, sd_control_work.work); + struct lm_lockstruct *ls = &sdp->sd_lockstruct; + uint32_t block_gen, start_gen, lvb_gen, flags; + int recover_set = 0; + int write_lvb = 0; + int recover_size; + int i, error; + + spin_lock(&ls->ls_recover_spin); + /* + * No MOUNT_DONE means we're still mounting; control_mount() + * will set this flag, after which this thread will take over + * all further clearing of BLOCK_LOCKS. + * + * FIRST_MOUNT means this node is doing first mounter recovery, + * for which recovery control is handled by + * control_mount()/control_first_done(), not this thread. + */ + if (!test_bit(DFL_MOUNT_DONE, &ls->ls_recover_flags) || + test_bit(DFL_FIRST_MOUNT, &ls->ls_recover_flags)) { + spin_unlock(&ls->ls_recover_spin); + return; + } + block_gen = ls->ls_recover_block; + start_gen = ls->ls_recover_start; + spin_unlock(&ls->ls_recover_spin); + + /* + * Equal block_gen and start_gen implies we are between + * recover_prep and recover_done callbacks, which means + * dlm recovery is in progress and dlm locking is blocked. + * There's no point trying to do any work until recover_done. + */ + + if (block_gen == start_gen) + return; + + /* + * Propagate recover_submit[] and recover_result[] to lvb: + * dlm_recoverd adds to recover_submit[] jids needing recovery + * gfs2_recover adds to recover_result[] journal recovery results + * + * set lvb bit for jids in recover_submit[] if the lvb has not + * yet been updated for the generation of the failure + * + * clear lvb bit for jids in recover_result[] if the result of + * the journal recovery is SUCCESS + */ + + error = control_lock(sdp, DLM_LOCK_EX, DLM_LKF_CONVERT|DLM_LKF_VALBLK); + if (error) { + fs_err(sdp, "control lock EX error %d\n", error); + return; + } + + control_lvb_read(ls, &lvb_gen, ls->ls_lvb_bits); + + spin_lock(&ls->ls_recover_spin); + if (block_gen != ls->ls_recover_block || + start_gen != ls->ls_recover_start) { + fs_info(sdp, "recover generation %u block1 %u %u\n", + start_gen, block_gen, ls->ls_recover_block); + spin_unlock(&ls->ls_recover_spin); + control_lock(sdp, DLM_LOCK_NL, DLM_LKF_CONVERT); + return; + } + + recover_size = ls->ls_recover_size; + + if (lvb_gen <= start_gen) { + /* + * Clear lvb bits for jids we've successfully recovered. + * Because all nodes attempt to recover failed journals, + * a journal can be recovered multiple times successfully + * in succession. Only the first will really do recovery, + * the others find it clean, but still report a successful + * recovery. So, another node may have already recovered + * the jid and cleared the lvb bit for it. + */ + for (i = 0; i < recover_size; i++) { + if (ls->ls_recover_result[i] != LM_RD_SUCCESS) + continue; + + ls->ls_recover_result[i] = 0; + + if (!test_bit_le(i, ls->ls_lvb_bits + JID_BITMAP_OFFSET)) + continue; + + __clear_bit_le(i, ls->ls_lvb_bits + JID_BITMAP_OFFSET); + write_lvb = 1; + } + } + + if (lvb_gen == start_gen) { + /* + * Failed slots before start_gen are already set in lvb. + */ + for (i = 0; i < recover_size; i++) { + if (!ls->ls_recover_submit[i]) + continue; + if (ls->ls_recover_submit[i] < lvb_gen) + ls->ls_recover_submit[i] = 0; + } + } else if (lvb_gen < start_gen) { + /* + * Failed slots before start_gen are not yet set in lvb. + */ + for (i = 0; i < recover_size; i++) { + if (!ls->ls_recover_submit[i]) + continue; + if (ls->ls_recover_submit[i] < start_gen) { + ls->ls_recover_submit[i] = 0; + __set_bit_le(i, ls->ls_lvb_bits + JID_BITMAP_OFFSET); + } + } + /* even if there are no bits to set, we need to write the + latest generation to the lvb */ + write_lvb = 1; + } else { + /* + * we should be getting a recover_done() for lvb_gen soon + */ + } + spin_unlock(&ls->ls_recover_spin); + + if (write_lvb) { + control_lvb_write(ls, start_gen, ls->ls_lvb_bits); + flags = DLM_LKF_CONVERT | DLM_LKF_VALBLK; + } else { + flags = DLM_LKF_CONVERT; + } + + error = control_lock(sdp, DLM_LOCK_NL, flags); + if (error) { + fs_err(sdp, "control lock NL error %d\n", error); + return; + } + + /* + * Everyone will see jid bits set in the lvb, run gfs2_recover_set(), + * and clear a jid bit in the lvb if the recovery is a success. + * Eventually all journals will be recovered, all jid bits will + * be cleared in the lvb, and everyone will clear BLOCK_LOCKS. + */ + + for (i = 0; i < recover_size; i++) { + if (test_bit_le(i, ls->ls_lvb_bits + JID_BITMAP_OFFSET)) { + fs_info(sdp, "recover generation %u jid %d\n", + start_gen, i); + gfs2_recover_set(sdp, i); + recover_set++; + } + } + if (recover_set) + return; + + /* + * No more jid bits set in lvb, all recovery is done, unblock locks + * (unless a new recover_prep callback has occured blocking locks + * again while working above) + */ + + spin_lock(&ls->ls_recover_spin); + if (ls->ls_recover_block == block_gen && + ls->ls_recover_start == start_gen) { + clear_bit(DFL_BLOCK_LOCKS, &ls->ls_recover_flags); + spin_unlock(&ls->ls_recover_spin); + fs_info(sdp, "recover generation %u done\n", start_gen); + gfs2_glock_thaw(sdp); + } else { + fs_info(sdp, "recover generation %u block2 %u %u\n", + start_gen, block_gen, ls->ls_recover_block); + spin_unlock(&ls->ls_recover_spin); + } +} + +static int control_mount(struct gfs2_sbd *sdp) +{ + struct lm_lockstruct *ls = &sdp->sd_lockstruct; + uint32_t start_gen, block_gen, mount_gen, lvb_gen; + int mounted_mode; + int retries = 0; + int error; + + memset(&ls->ls_mounted_lksb, 0, sizeof(struct dlm_lksb)); + memset(&ls->ls_control_lksb, 0, sizeof(struct dlm_lksb)); + memset(&ls->ls_control_lvb, 0, GDLM_LVB_SIZE); + ls->ls_control_lksb.sb_lvbptr = ls->ls_control_lvb; + init_completion(&ls->ls_sync_wait); + + set_bit(DFL_BLOCK_LOCKS, &ls->ls_recover_flags); + + error = control_lock(sdp, DLM_LOCK_NL, DLM_LKF_VALBLK); + if (error) { + fs_err(sdp, "control_mount control_lock NL error %d\n", error); + return error; + } + + error = mounted_lock(sdp, DLM_LOCK_NL, 0); + if (error) { + fs_err(sdp, "control_mount mounted_lock NL error %d\n", error); + control_unlock(sdp); + return error; + } + mounted_mode = DLM_LOCK_NL; + +restart: + if (retries++ && signal_pending(current)) { + error = -EINTR; + goto fail; + } + + /* + * We always start with both locks in NL. control_lock is + * demoted to NL below so we don't need to do it here. + */ + + if (mounted_mode != DLM_LOCK_NL) { + error = mounted_lock(sdp, DLM_LOCK_NL, DLM_LKF_CONVERT); + if (error) + goto fail; + mounted_mode = DLM_LOCK_NL; + } + + /* + * Other nodes need to do some work in dlm recovery and gfs2_control + * before the recover_done and control_lock will be ready for us below. + * A delay here is not required but often avoids having to retry. + */ + + msleep_interruptible(500); + + /* + * Acquire control_lock in EX and mounted_lock in either EX or PR. + * control_lock lvb keeps track of any pending journal recoveries. + * mounted_lock indicates if any other nodes have the fs mounted. + */ + + error = control_lock(sdp, DLM_LOCK_EX, DLM_LKF_CONVERT|DLM_LKF_NOQUEUE|DLM_LKF_VALBLK); + if (error == -EAGAIN) { + goto restart; + } else if (error) { + fs_err(sdp, "control_mount control_lock EX error %d\n", error); + goto fail; + } + + error = mounted_lock(sdp, DLM_LOCK_EX, DLM_LKF_CONVERT|DLM_LKF_NOQUEUE); + if (!error) { + mounted_mode = DLM_LOCK_EX; + goto locks_done; + } else if (error != -EAGAIN) { + fs_err(sdp, "control_mount mounted_lock EX error %d\n", error); + goto fail; + } + + error = mounted_lock(sdp, DLM_LOCK_PR, DLM_LKF_CONVERT|DLM_LKF_NOQUEUE); + if (!error) { + mounted_mode = DLM_LOCK_PR; + goto locks_done; + } else { + /* not even -EAGAIN should happen here */ + fs_err(sdp, "control_mount mounted_lock PR error %d\n", error); + goto fail; + } + +locks_done: + /* + * If we got both locks above in EX, then we're the first mounter. + * If not, then we need to wait for the control_lock lvb to be + * updated by other mounted nodes to reflect our mount generation. + * + * In simple first mounter cases, first mounter will see zero lvb_gen, + * but in cases where all existing nodes leave/fail before mounting + * nodes finish control_mount, then all nodes will be mounting and + * lvb_gen will be non-zero. + */ + + control_lvb_read(ls, &lvb_gen, ls->ls_lvb_bits); + + if (lvb_gen == 0xFFFFFFFF) { + /* special value to force mount attempts to fail */ + fs_err(sdp, "control_mount control_lock disabled\n"); + error = -EINVAL; + goto fail; + } + + if (mounted_mode == DLM_LOCK_EX) { + /* first mounter, keep both EX while doing first recovery */ + spin_lock(&ls->ls_recover_spin); + clear_bit(DFL_BLOCK_LOCKS, &ls->ls_recover_flags); + set_bit(DFL_MOUNT_DONE, &ls->ls_recover_flags); + set_bit(DFL_FIRST_MOUNT, &ls->ls_recover_flags); + spin_unlock(&ls->ls_recover_spin); + fs_info(sdp, "first mounter control generation %u\n", lvb_gen); + return 0; + } + + error = control_lock(sdp, DLM_LOCK_NL, DLM_LKF_CONVERT); + if (error) + goto fail; + + /* + * We are not first mounter, now we need to wait for the control_lock + * lvb generation to be >= the generation from our first recover_done + * and all lvb bits to be clear (no pending journal recoveries.) + */ + + if (!all_jid_bits_clear(ls->ls_lvb_bits)) { + /* journals need recovery, wait until all are clear */ + fs_info(sdp, "control_mount wait for journal recovery\n"); + goto restart; + } + + spin_lock(&ls->ls_recover_spin); + block_gen = ls->ls_recover_block; + start_gen = ls->ls_recover_start; + mount_gen = ls->ls_recover_mount; + + if (lvb_gen < mount_gen) { + /* wait for mounted nodes to update control_lock lvb to our + generation, which might include new recovery bits set */ + fs_info(sdp, "control_mount wait1 block %u start %u mount %u " + "lvb %u flags %lx\n", block_gen, start_gen, mount_gen, + lvb_gen, ls->ls_recover_flags); + spin_unlock(&ls->ls_recover_spin); + goto restart; + } + + if (lvb_gen != start_gen) { + /* wait for mounted nodes to update control_lock lvb to the + latest recovery generation */ + fs_info(sdp, "control_mount wait2 block %u start %u mount %u " + "lvb %u flags %lx\n", block_gen, start_gen, mount_gen, + lvb_gen, ls->ls_recover_flags); + spin_unlock(&ls->ls_recover_spin); + goto restart; + } + + if (block_gen == start_gen) { + /* dlm recovery in progress, wait for it to finish */ + fs_info(sdp, "control_mount wait3 block %u start %u mount %u " + "lvb %u flags %lx\n", block_gen, start_gen, mount_gen, + lvb_gen, ls->ls_recover_flags); + spin_unlock(&ls->ls_recover_spin); + goto restart; + } + + clear_bit(DFL_BLOCK_LOCKS, &ls->ls_recover_flags); + set_bit(DFL_MOUNT_DONE, &ls->ls_recover_flags); + memset(ls->ls_recover_submit, 0, ls->ls_recover_size*sizeof(uint32_t)); + memset(ls->ls_recover_result, 0, ls->ls_recover_size*sizeof(uint32_t)); + spin_unlock(&ls->ls_recover_spin); + return 0; + +fail: + mounted_unlock(sdp); + control_unlock(sdp); + return error; +} + +static int dlm_recovery_wait(void *word) +{ + schedule(); + return 0; +} + +static int control_first_done(struct gfs2_sbd *sdp) +{ + struct lm_lockstruct *ls = &sdp->sd_lockstruct; + uint32_t start_gen, block_gen; + int error; + +restart: + spin_lock(&ls->ls_recover_spin); + start_gen = ls->ls_recover_start; + block_gen = ls->ls_recover_block; + + if (test_bit(DFL_BLOCK_LOCKS, &ls->ls_recover_flags) || + !test_bit(DFL_MOUNT_DONE, &ls->ls_recover_flags) || + !test_bit(DFL_FIRST_MOUNT, &ls->ls_recover_flags)) { + /* sanity check, should not happen */ + fs_err(sdp, "control_first_done start %u block %u flags %lx\n", + start_gen, block_gen, ls->ls_recover_flags); + spin_unlock(&ls->ls_recover_spin); + control_unlock(sdp); + return -1; + } + + if (start_gen == block_gen) { + /* + * Wait for the end of a dlm recovery cycle to switch from + * first mounter recovery. We can ignore any recover_slot + * callbacks between the recover_prep and next recover_done + * because we are still the first mounter and any failed nodes + * have not fully mounted, so they don't need recovery. + */ + spin_unlock(&ls->ls_recover_spin); + fs_info(sdp, "control_first_done wait gen %u\n", start_gen); + + wait_on_bit(&ls->ls_recover_flags, DFL_DLM_RECOVERY, + dlm_recovery_wait, TASK_UNINTERRUPTIBLE); + goto restart; + } + + clear_bit(DFL_FIRST_MOUNT, &ls->ls_recover_flags); + set_bit(DFL_FIRST_MOUNT_DONE, &ls->ls_recover_flags); + memset(ls->ls_recover_submit, 0, ls->ls_recover_size*sizeof(uint32_t)); + memset(ls->ls_recover_result, 0, ls->ls_recover_size*sizeof(uint32_t)); + spin_unlock(&ls->ls_recover_spin); + + memset(ls->ls_lvb_bits, 0, GDLM_LVB_SIZE); + control_lvb_write(ls, start_gen, ls->ls_lvb_bits); + + error = mounted_lock(sdp, DLM_LOCK_PR, DLM_LKF_CONVERT); + if (error) + fs_err(sdp, "control_first_done mounted PR error %d\n", error); + + error = control_lock(sdp, DLM_LOCK_NL, DLM_LKF_CONVERT|DLM_LKF_VALBLK); + if (error) + fs_err(sdp, "control_first_done control NL error %d\n", error); + + return error; +} + +/* + * Expand static jid arrays if necessary (by increments of RECOVER_SIZE_INC) + * to accomodate the largest slot number. (NB dlm slot numbers start at 1, + * gfs2 jids start at 0, so jid = slot - 1) + */ + +#define RECOVER_SIZE_INC 16 + +static int set_recover_size(struct gfs2_sbd *sdp, struct dlm_slot *slots, + int num_slots) +{ + struct lm_lockstruct *ls = &sdp->sd_lockstruct; + uint32_t *submit = NULL; + uint32_t *result = NULL; + uint32_t old_size, new_size; + int i, max_jid; + + if (!ls->ls_lvb_bits) { + ls->ls_lvb_bits = kzalloc(GDLM_LVB_SIZE, GFP_NOFS); + if (!ls->ls_lvb_bits) + return -ENOMEM; + } + + max_jid = 0; + for (i = 0; i < num_slots; i++) { + if (max_jid < slots[i].slot - 1) + max_jid = slots[i].slot - 1; + } + + old_size = ls->ls_recover_size; + + if (old_size >= max_jid + 1) + return 0; + + new_size = old_size + RECOVER_SIZE_INC; + + submit = kcalloc(new_size, sizeof(uint32_t), GFP_NOFS); + result = kcalloc(new_size, sizeof(uint32_t), GFP_NOFS); + if (!submit || !result) { + kfree(submit); + kfree(result); + return -ENOMEM; + } + + spin_lock(&ls->ls_recover_spin); + memcpy(submit, ls->ls_recover_submit, old_size * sizeof(uint32_t)); + memcpy(result, ls->ls_recover_result, old_size * sizeof(uint32_t)); + kfree(ls->ls_recover_submit); + kfree(ls->ls_recover_result); + ls->ls_recover_submit = submit; + ls->ls_recover_result = result; + ls->ls_recover_size = new_size; + spin_unlock(&ls->ls_recover_spin); + return 0; +} + +static void free_recover_size(struct lm_lockstruct *ls) +{ + kfree(ls->ls_lvb_bits); + kfree(ls->ls_recover_submit); + kfree(ls->ls_recover_result); + ls->ls_recover_submit = NULL; + ls->ls_recover_result = NULL; + ls->ls_recover_size = 0; +} + +/* dlm calls before it does lock recovery */ + +static void gdlm_recover_prep(void *arg) +{ + struct gfs2_sbd *sdp = arg; + struct lm_lockstruct *ls = &sdp->sd_lockstruct; + + spin_lock(&ls->ls_recover_spin); + ls->ls_recover_block = ls->ls_recover_start; + set_bit(DFL_DLM_RECOVERY, &ls->ls_recover_flags); + + if (!test_bit(DFL_MOUNT_DONE, &ls->ls_recover_flags) || + test_bit(DFL_FIRST_MOUNT, &ls->ls_recover_flags)) { + spin_unlock(&ls->ls_recover_spin); + return; + } + set_bit(DFL_BLOCK_LOCKS, &ls->ls_recover_flags); + spin_unlock(&ls->ls_recover_spin); +} + +/* dlm calls after recover_prep has been completed on all lockspace members; + identifies slot/jid of failed member */ + +static void gdlm_recover_slot(void *arg, struct dlm_slot *slot) +{ + struct gfs2_sbd *sdp = arg; + struct lm_lockstruct *ls = &sdp->sd_lockstruct; + int jid = slot->slot - 1; + + spin_lock(&ls->ls_recover_spin); + if (ls->ls_recover_size < jid + 1) { + fs_err(sdp, "recover_slot jid %d gen %u short size %d", + jid, ls->ls_recover_block, ls->ls_recover_size); + spin_unlock(&ls->ls_recover_spin); + return; + } + + if (ls->ls_recover_submit[jid]) { + fs_info(sdp, "recover_slot jid %d gen %u prev %u\n", + jid, ls->ls_recover_block, ls->ls_recover_submit[jid]); + } + ls->ls_recover_submit[jid] = ls->ls_recover_block; + spin_unlock(&ls->ls_recover_spin); +} + +/* dlm calls after recover_slot and after it completes lock recovery */ + +static void gdlm_recover_done(void *arg, struct dlm_slot *slots, int num_slots, + int our_slot, uint32_t generation) +{ + struct gfs2_sbd *sdp = arg; + struct lm_lockstruct *ls = &sdp->sd_lockstruct; + + /* ensure the ls jid arrays are large enough */ + set_recover_size(sdp, slots, num_slots); + + spin_lock(&ls->ls_recover_spin); + ls->ls_recover_start = generation; + + if (!ls->ls_recover_mount) { + ls->ls_recover_mount = generation; + ls->ls_jid = our_slot - 1; } - error = dlm_new_lockspace(fsname, strlen(fsname), &ls->ls_dlm, - DLM_LSFL_FS | DLM_LSFL_NEWEXCL | - (ls->ls_nodir ? DLM_LSFL_NODIR : 0), - GDLM_LVB_SIZE); + if (!test_bit(DFL_UNMOUNT, &ls->ls_recover_flags)) + queue_delayed_work(gfs2_control_wq, &sdp->sd_control_work, 0); + + clear_bit(DFL_DLM_RECOVERY, &ls->ls_recover_flags); + smp_mb__after_atomic(); + wake_up_bit(&ls->ls_recover_flags, DFL_DLM_RECOVERY); + spin_unlock(&ls->ls_recover_spin); +} + +/* gfs2_recover thread has a journal recovery result */ + +static void gdlm_recovery_result(struct gfs2_sbd *sdp, unsigned int jid, + unsigned int result) +{ + struct lm_lockstruct *ls = &sdp->sd_lockstruct; + + if (test_bit(DFL_NO_DLM_OPS, &ls->ls_recover_flags)) + return; + + /* don't care about the recovery of own journal during mount */ + if (jid == ls->ls_jid) + return; + + spin_lock(&ls->ls_recover_spin); + if (test_bit(DFL_FIRST_MOUNT, &ls->ls_recover_flags)) { + spin_unlock(&ls->ls_recover_spin); + return; + } + if (ls->ls_recover_size < jid + 1) { + fs_err(sdp, "recovery_result jid %d short size %d", + jid, ls->ls_recover_size); + spin_unlock(&ls->ls_recover_spin); + return; + } + + fs_info(sdp, "recover jid %d result %s\n", jid, + result == LM_RD_GAVEUP ? "busy" : "success"); + + ls->ls_recover_result[jid] = result; + + /* GAVEUP means another node is recovering the journal; delay our + next attempt to recover it, to give the other node a chance to + finish before trying again */ + + if (!test_bit(DFL_UNMOUNT, &ls->ls_recover_flags)) + queue_delayed_work(gfs2_control_wq, &sdp->sd_control_work, + result == LM_RD_GAVEUP ? HZ : 0); + spin_unlock(&ls->ls_recover_spin); +} + +const struct dlm_lockspace_ops gdlm_lockspace_ops = { + .recover_prep = gdlm_recover_prep, + .recover_slot = gdlm_recover_slot, + .recover_done = gdlm_recover_done, +}; + +static int gdlm_mount(struct gfs2_sbd *sdp, const char *table) +{ + struct lm_lockstruct *ls = &sdp->sd_lockstruct; + char cluster[GFS2_LOCKNAME_LEN]; + const char *fsname; + uint32_t flags; + int error, ops_result; + + /* + * initialize everything + */ + + INIT_DELAYED_WORK(&sdp->sd_control_work, gfs2_control_func); + spin_lock_init(&ls->ls_recover_spin); + ls->ls_recover_flags = 0; + ls->ls_recover_mount = 0; + ls->ls_recover_start = 0; + ls->ls_recover_block = 0; + ls->ls_recover_size = 0; + ls->ls_recover_submit = NULL; + ls->ls_recover_result = NULL; + ls->ls_lvb_bits = NULL; + + error = set_recover_size(sdp, NULL, 0); if (error) - printk(KERN_ERR "dlm_new_lockspace error %d", error); + goto fail; + + /* + * prepare dlm_new_lockspace args + */ + + fsname = strchr(table, ':'); + if (!fsname) { + fs_info(sdp, "no fsname found\n"); + error = -EINVAL; + goto fail_free; + } + memset(cluster, 0, sizeof(cluster)); + memcpy(cluster, table, strlen(table) - strlen(fsname)); + fsname++; + + flags = DLM_LSFL_FS | DLM_LSFL_NEWEXCL; + + /* + * create/join lockspace + */ + + error = dlm_new_lockspace(fsname, cluster, flags, GDLM_LVB_SIZE, + &gdlm_lockspace_ops, sdp, &ops_result, + &ls->ls_dlm); + if (error) { + fs_err(sdp, "dlm_new_lockspace error %d\n", error); + goto fail_free; + } + + if (ops_result < 0) { + /* + * dlm does not support ops callbacks, + * old dlm_controld/gfs_controld are used, try without ops. + */ + fs_info(sdp, "dlm lockspace ops not used\n"); + free_recover_size(ls); + set_bit(DFL_NO_DLM_OPS, &ls->ls_recover_flags); + return 0; + } + if (!test_bit(SDF_NOJOURNALID, &sdp->sd_flags)) { + fs_err(sdp, "dlm lockspace ops disallow jid preset\n"); + error = -EINVAL; + goto fail_release; + } + + /* + * control_mount() uses control_lock to determine first mounter, + * and for later mounts, waits for any recoveries to be cleared. + */ + + error = control_mount(sdp); + if (error) { + fs_err(sdp, "mount control error %d\n", error); + goto fail_release; + } + + ls->ls_first = !!test_bit(DFL_FIRST_MOUNT, &ls->ls_recover_flags); + clear_bit(SDF_NOJOURNALID, &sdp->sd_flags); + smp_mb__after_atomic(); + wake_up_bit(&sdp->sd_flags, SDF_NOJOURNALID); + return 0; + +fail_release: + dlm_release_lockspace(ls->ls_dlm, 2); +fail_free: + free_recover_size(ls); +fail: return error; } +static void gdlm_first_done(struct gfs2_sbd *sdp) +{ + struct lm_lockstruct *ls = &sdp->sd_lockstruct; + int error; + + if (test_bit(DFL_NO_DLM_OPS, &ls->ls_recover_flags)) + return; + + error = control_first_done(sdp); + if (error) + fs_err(sdp, "mount first_done error %d\n", error); +} + static void gdlm_unmount(struct gfs2_sbd *sdp) { struct lm_lockstruct *ls = &sdp->sd_lockstruct; + if (test_bit(DFL_NO_DLM_OPS, &ls->ls_recover_flags)) + goto release; + + /* wait for gfs2_control_wq to be done with this mount */ + + spin_lock(&ls->ls_recover_spin); + set_bit(DFL_UNMOUNT, &ls->ls_recover_flags); + spin_unlock(&ls->ls_recover_spin); + flush_delayed_work(&sdp->sd_control_work); + + /* mounted_lock and control_lock will be purged in dlm recovery */ +release: if (ls->ls_dlm) { dlm_release_lockspace(ls->ls_dlm, 2); ls->ls_dlm = NULL; } + + free_recover_size(ls); } static const match_table_t dlm_tokens = { @@ -234,6 +1331,8 @@ static const match_table_t dlm_tokens = { const struct lm_lockops gfs2_dlm_ops = { .lm_proto_name = "lock_dlm", .lm_mount = gdlm_mount, + .lm_first_done = gdlm_first_done, + .lm_recovery_result = gdlm_recovery_result, .lm_unmount = gdlm_unmount, .lm_put_lock = gdlm_put_lock, .lm_lock = gdlm_lock, diff --git a/fs/gfs2/log.c b/fs/gfs2/log.c index eb01f3575e1..3966fadbceb 100644 --- a/fs/gfs2/log.c +++ b/fs/gfs2/log.c @@ -18,6 +18,9 @@ #include <linux/kthread.h> #include <linux/freezer.h> #include <linux/bio.h> +#include <linux/blkdev.h> +#include <linux/writeback.h> +#include <linux/list_sort.h> #include "gfs2.h" #include "incore.h" @@ -30,8 +33,6 @@ #include "dir.h" #include "trace_gfs2.h" -#define PULL 1 - /** * gfs2_struct2blk - compute stuff * @sdp: the filesystem @@ -67,13 +68,13 @@ unsigned int gfs2_struct2blk(struct gfs2_sbd *sdp, unsigned int nstruct, * @mapping: The associated mapping (maybe NULL) * @bd: The gfs2_bufdata to remove * - * The log lock _must_ be held when calling this function + * The ail lock _must_ be held when calling this function * */ void gfs2_remove_from_ail(struct gfs2_bufdata *bd) { - bd->bd_ail = NULL; + bd->bd_tr = NULL; list_del_init(&bd->bd_ail_st_list); list_del_init(&bd->bd_ail_gl_list); atomic_dec(&bd->bd_gl->gl_ail_count); @@ -83,55 +84,100 @@ void gfs2_remove_from_ail(struct gfs2_bufdata *bd) /** * gfs2_ail1_start_one - Start I/O on a part of the AIL * @sdp: the filesystem - * @tr: the part of the AIL + * @wbc: The writeback control structure + * @ai: The ail structure * */ -static void gfs2_ail1_start_one(struct gfs2_sbd *sdp, struct gfs2_ail *ai) -__releases(&sdp->sd_log_lock) -__acquires(&sdp->sd_log_lock) +static int gfs2_ail1_start_one(struct gfs2_sbd *sdp, + struct writeback_control *wbc, + struct gfs2_trans *tr) +__releases(&sdp->sd_ail_lock) +__acquires(&sdp->sd_ail_lock) { + struct gfs2_glock *gl = NULL; + struct address_space *mapping; struct gfs2_bufdata *bd, *s; struct buffer_head *bh; - int retry; - do { - retry = 0; + list_for_each_entry_safe_reverse(bd, s, &tr->tr_ail1_list, bd_ail_st_list) { + bh = bd->bd_bh; - list_for_each_entry_safe_reverse(bd, s, &ai->ai_ail1_list, - bd_ail_st_list) { - bh = bd->bd_bh; + gfs2_assert(sdp, bd->bd_tr == tr); - gfs2_assert(sdp, bd->bd_ail == ai); + if (!buffer_busy(bh)) { + if (!buffer_uptodate(bh)) + gfs2_io_error_bh(sdp, bh); + list_move(&bd->bd_ail_st_list, &tr->tr_ail2_list); + continue; + } - if (!buffer_busy(bh)) { - if (!buffer_uptodate(bh)) - gfs2_io_error_bh(sdp, bh); - list_move(&bd->bd_ail_st_list, &ai->ai_ail2_list); - continue; - } + if (!buffer_dirty(bh)) + continue; + if (gl == bd->bd_gl) + continue; + gl = bd->bd_gl; + list_move(&bd->bd_ail_st_list, &tr->tr_ail1_list); + mapping = bh->b_page->mapping; + if (!mapping) + continue; + spin_unlock(&sdp->sd_ail_lock); + generic_writepages(mapping, wbc); + spin_lock(&sdp->sd_ail_lock); + if (wbc->nr_to_write <= 0) + break; + return 1; + } - if (!buffer_dirty(bh)) - continue; + return 0; +} - list_move(&bd->bd_ail_st_list, &ai->ai_ail1_list); - get_bh(bh); - gfs2_log_unlock(sdp); - lock_buffer(bh); - if (test_clear_buffer_dirty(bh)) { - bh->b_end_io = end_buffer_write_sync; - submit_bh(WRITE_SYNC_PLUG, bh); - } else { - unlock_buffer(bh); - brelse(bh); - } - gfs2_log_lock(sdp); +/** + * gfs2_ail1_flush - start writeback of some ail1 entries + * @sdp: The super block + * @wbc: The writeback control structure + * + * Writes back some ail1 entries, according to the limits in the + * writeback control structure + */ + +void gfs2_ail1_flush(struct gfs2_sbd *sdp, struct writeback_control *wbc) +{ + struct list_head *head = &sdp->sd_ail1_list; + struct gfs2_trans *tr; + struct blk_plug plug; - retry = 1; + trace_gfs2_ail_flush(sdp, wbc, 1); + blk_start_plug(&plug); + spin_lock(&sdp->sd_ail_lock); +restart: + list_for_each_entry_reverse(tr, head, tr_list) { + if (wbc->nr_to_write <= 0) break; - } - } while (retry); + if (gfs2_ail1_start_one(sdp, wbc, tr)) + goto restart; + } + spin_unlock(&sdp->sd_ail_lock); + blk_finish_plug(&plug); + trace_gfs2_ail_flush(sdp, wbc, 0); +} + +/** + * gfs2_ail1_start - start writeback of all ail1 entries + * @sdp: The superblock + */ + +static void gfs2_ail1_start(struct gfs2_sbd *sdp) +{ + struct writeback_control wbc = { + .sync_mode = WB_SYNC_NONE, + .nr_to_write = LONG_MAX, + .range_start = 0, + .range_end = LLONG_MAX, + }; + + return gfs2_ail1_flush(sdp, &wbc); } /** @@ -141,84 +187,72 @@ __acquires(&sdp->sd_log_lock) * */ -static int gfs2_ail1_empty_one(struct gfs2_sbd *sdp, struct gfs2_ail *ai, int flags) +static void gfs2_ail1_empty_one(struct gfs2_sbd *sdp, struct gfs2_trans *tr) { struct gfs2_bufdata *bd, *s; struct buffer_head *bh; - list_for_each_entry_safe_reverse(bd, s, &ai->ai_ail1_list, + list_for_each_entry_safe_reverse(bd, s, &tr->tr_ail1_list, bd_ail_st_list) { bh = bd->bd_bh; - - gfs2_assert(sdp, bd->bd_ail == ai); - - if (buffer_busy(bh)) { - if (flags & DIO_ALL) - continue; - else - break; - } - + gfs2_assert(sdp, bd->bd_tr == tr); + if (buffer_busy(bh)) + continue; if (!buffer_uptodate(bh)) gfs2_io_error_bh(sdp, bh); - - list_move(&bd->bd_ail_st_list, &ai->ai_ail2_list); + list_move(&bd->bd_ail_st_list, &tr->tr_ail2_list); } - return list_empty(&ai->ai_ail1_list); } -static void gfs2_ail1_start(struct gfs2_sbd *sdp) -{ - struct list_head *head; - u64 sync_gen; - struct gfs2_ail *ai; - int done = 0; - - gfs2_log_lock(sdp); - head = &sdp->sd_ail1_list; - if (list_empty(head)) { - gfs2_log_unlock(sdp); - return; - } - sync_gen = sdp->sd_ail_sync_gen++; - - while(!done) { - done = 1; - list_for_each_entry_reverse(ai, head, ai_list) { - if (ai->ai_sync_gen >= sync_gen) - continue; - ai->ai_sync_gen = sync_gen; - gfs2_ail1_start_one(sdp, ai); /* This may drop log lock */ - done = 0; - break; - } - } - - gfs2_log_unlock(sdp); -} +/** + * gfs2_ail1_empty - Try to empty the ail1 lists + * @sdp: The superblock + * + * Tries to empty the ail1 lists, starting with the oldest first + */ -static int gfs2_ail1_empty(struct gfs2_sbd *sdp, int flags) +static int gfs2_ail1_empty(struct gfs2_sbd *sdp) { - struct gfs2_ail *ai, *s; + struct gfs2_trans *tr, *s; + int oldest_tr = 1; int ret; - gfs2_log_lock(sdp); - - list_for_each_entry_safe_reverse(ai, s, &sdp->sd_ail1_list, ai_list) { - if (gfs2_ail1_empty_one(sdp, ai, flags)) - list_move(&ai->ai_list, &sdp->sd_ail2_list); - else if (!(flags & DIO_ALL)) - break; + spin_lock(&sdp->sd_ail_lock); + list_for_each_entry_safe_reverse(tr, s, &sdp->sd_ail1_list, tr_list) { + gfs2_ail1_empty_one(sdp, tr); + if (list_empty(&tr->tr_ail1_list) && oldest_tr) + list_move(&tr->tr_list, &sdp->sd_ail2_list); + else + oldest_tr = 0; } - ret = list_empty(&sdp->sd_ail1_list); - - gfs2_log_unlock(sdp); + spin_unlock(&sdp->sd_ail_lock); return ret; } +static void gfs2_ail1_wait(struct gfs2_sbd *sdp) +{ + struct gfs2_trans *tr; + struct gfs2_bufdata *bd; + struct buffer_head *bh; + + spin_lock(&sdp->sd_ail_lock); + list_for_each_entry_reverse(tr, &sdp->sd_ail1_list, tr_list) { + list_for_each_entry(bd, &tr->tr_ail1_list, bd_ail_st_list) { + bh = bd->bd_bh; + if (!buffer_locked(bh)) + continue; + get_bh(bh); + spin_unlock(&sdp->sd_ail_lock); + wait_on_buffer(bh); + brelse(bh); + return; + } + } + spin_unlock(&sdp->sd_ail_lock); +} /** * gfs2_ail2_empty_one - Check whether or not a trans in the AIL has been synced @@ -227,43 +261,60 @@ static int gfs2_ail1_empty(struct gfs2_sbd *sdp, int flags) * */ -static void gfs2_ail2_empty_one(struct gfs2_sbd *sdp, struct gfs2_ail *ai) +static void gfs2_ail2_empty_one(struct gfs2_sbd *sdp, struct gfs2_trans *tr) { - struct list_head *head = &ai->ai_ail2_list; + struct list_head *head = &tr->tr_ail2_list; struct gfs2_bufdata *bd; while (!list_empty(head)) { bd = list_entry(head->prev, struct gfs2_bufdata, bd_ail_st_list); - gfs2_assert(sdp, bd->bd_ail == ai); + gfs2_assert(sdp, bd->bd_tr == tr); gfs2_remove_from_ail(bd); } } static void ail2_empty(struct gfs2_sbd *sdp, unsigned int new_tail) { - struct gfs2_ail *ai, *safe; + struct gfs2_trans *tr, *safe; unsigned int old_tail = sdp->sd_log_tail; int wrap = (new_tail < old_tail); int a, b, rm; - gfs2_log_lock(sdp); + spin_lock(&sdp->sd_ail_lock); - list_for_each_entry_safe(ai, safe, &sdp->sd_ail2_list, ai_list) { - a = (old_tail <= ai->ai_first); - b = (ai->ai_first < new_tail); + list_for_each_entry_safe(tr, safe, &sdp->sd_ail2_list, tr_list) { + a = (old_tail <= tr->tr_first); + b = (tr->tr_first < new_tail); rm = (wrap) ? (a || b) : (a && b); if (!rm) continue; - gfs2_ail2_empty_one(sdp, ai); - list_del(&ai->ai_list); - gfs2_assert_warn(sdp, list_empty(&ai->ai_ail1_list)); - gfs2_assert_warn(sdp, list_empty(&ai->ai_ail2_list)); - kfree(ai); + gfs2_ail2_empty_one(sdp, tr); + list_del(&tr->tr_list); + gfs2_assert_warn(sdp, list_empty(&tr->tr_ail1_list)); + gfs2_assert_warn(sdp, list_empty(&tr->tr_ail2_list)); + kfree(tr); } - gfs2_log_unlock(sdp); + spin_unlock(&sdp->sd_ail_lock); +} + +/** + * gfs2_log_release - Release a given number of log blocks + * @sdp: The GFS2 superblock + * @blks: The number of blocks + * + */ + +void gfs2_log_release(struct gfs2_sbd *sdp, unsigned int blks) +{ + + atomic_add(blks, &sdp->sd_log_blks_free); + trace_gfs2_log_blocks(sdp, blks); + gfs2_assert_withdraw(sdp, atomic_read(&sdp->sd_log_blks_free) <= + sdp->sd_jdesc->jd_blocks); + up_read(&sdp->sd_log_flush_lock); } /** @@ -288,7 +339,7 @@ static void ail2_empty(struct gfs2_sbd *sdp, unsigned int new_tail) int gfs2_log_reserve(struct gfs2_sbd *sdp, unsigned int blks) { - unsigned reserved_blks = 6 * (4096 / sdp->sd_vfs->s_blocksize); + unsigned reserved_blks = 7 * (4096 / sdp->sd_vfs->s_blocksize); unsigned wanted = blks + reserved_blks; DEFINE_WAIT(wait); int did_wait = 0; @@ -324,20 +375,11 @@ retry: wake_up(&sdp->sd_log_waitq); down_read(&sdp->sd_log_flush_lock); - - return 0; -} - -static u64 log_bmap(struct gfs2_sbd *sdp, unsigned int lbn) -{ - struct gfs2_journal_extent *je; - - list_for_each_entry(je, &sdp->sd_jdesc->extent_list, extent_list) { - if (lbn >= je->lblock && lbn < je->lblock + je->blocks) - return je->dblock + lbn - je->lblock; + if (unlikely(!test_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags))) { + gfs2_log_release(sdp, blks); + return -EROFS; } - - return -1; + return 0; } /** @@ -392,24 +434,22 @@ static inline unsigned int log_distance(struct gfs2_sbd *sdp, unsigned int newer static unsigned int calc_reserved(struct gfs2_sbd *sdp) { unsigned int reserved = 0; - unsigned int mbuf_limit, metabufhdrs_needed; - unsigned int dbuf_limit, databufhdrs_needed; - unsigned int revokes = 0; - - mbuf_limit = buf_limit(sdp); - metabufhdrs_needed = (sdp->sd_log_commited_buf + - (mbuf_limit - 1)) / mbuf_limit; - dbuf_limit = databuf_limit(sdp); - databufhdrs_needed = (sdp->sd_log_commited_databuf + - (dbuf_limit - 1)) / dbuf_limit; + unsigned int mbuf; + unsigned int dbuf; + struct gfs2_trans *tr = sdp->sd_log_tr; + + if (tr) { + mbuf = tr->tr_num_buf_new - tr->tr_num_buf_rm; + dbuf = tr->tr_num_databuf_new - tr->tr_num_databuf_rm; + reserved = mbuf + dbuf; + /* Account for header blocks */ + reserved += DIV_ROUND_UP(mbuf, buf_limit(sdp)); + reserved += DIV_ROUND_UP(dbuf, databuf_limit(sdp)); + } if (sdp->sd_log_commited_revoke > 0) - revokes = gfs2_struct2blk(sdp, sdp->sd_log_commited_revoke, + reserved += gfs2_struct2blk(sdp, sdp->sd_log_commited_revoke, sizeof(u64)); - - reserved = sdp->sd_log_commited_buf + metabufhdrs_needed + - sdp->sd_log_commited_databuf + databufhdrs_needed + - revokes; /* One for the overall header */ if (reserved) reserved++; @@ -418,139 +458,184 @@ static unsigned int calc_reserved(struct gfs2_sbd *sdp) static unsigned int current_tail(struct gfs2_sbd *sdp) { - struct gfs2_ail *ai; + struct gfs2_trans *tr; unsigned int tail; - gfs2_log_lock(sdp); + spin_lock(&sdp->sd_ail_lock); if (list_empty(&sdp->sd_ail1_list)) { tail = sdp->sd_log_head; } else { - ai = list_entry(sdp->sd_ail1_list.prev, struct gfs2_ail, ai_list); - tail = ai->ai_first; + tr = list_entry(sdp->sd_ail1_list.prev, struct gfs2_trans, + tr_list); + tail = tr->tr_first; } - gfs2_log_unlock(sdp); + spin_unlock(&sdp->sd_ail_lock); return tail; } -void gfs2_log_incr_head(struct gfs2_sbd *sdp) +static void log_pull_tail(struct gfs2_sbd *sdp, unsigned int new_tail) { - if (sdp->sd_log_flush_head == sdp->sd_log_tail) - BUG_ON(sdp->sd_log_flush_head != sdp->sd_log_head); + unsigned int dist = log_distance(sdp, new_tail, sdp->sd_log_tail); - if (++sdp->sd_log_flush_head == sdp->sd_jdesc->jd_blocks) { - sdp->sd_log_flush_head = 0; - sdp->sd_log_flush_wrapped = 1; - } + ail2_empty(sdp, new_tail); + + atomic_add(dist, &sdp->sd_log_blks_free); + trace_gfs2_log_blocks(sdp, dist); + gfs2_assert_withdraw(sdp, atomic_read(&sdp->sd_log_blks_free) <= + sdp->sd_jdesc->jd_blocks); + + sdp->sd_log_tail = new_tail; } -/** - * gfs2_log_write_endio - End of I/O for a log buffer - * @bh: The buffer head - * @uptodate: I/O Status - * - */ -static void gfs2_log_write_endio(struct buffer_head *bh, int uptodate) +static void log_flush_wait(struct gfs2_sbd *sdp) { - struct gfs2_sbd *sdp = bh->b_private; - bh->b_private = NULL; + DEFINE_WAIT(wait); - end_buffer_write_sync(bh, uptodate); - if (atomic_dec_and_test(&sdp->sd_log_in_flight)) - wake_up(&sdp->sd_log_flush_wait); + if (atomic_read(&sdp->sd_log_in_flight)) { + do { + prepare_to_wait(&sdp->sd_log_flush_wait, &wait, + TASK_UNINTERRUPTIBLE); + if (atomic_read(&sdp->sd_log_in_flight)) + io_schedule(); + } while(atomic_read(&sdp->sd_log_in_flight)); + finish_wait(&sdp->sd_log_flush_wait, &wait); + } } -/** - * gfs2_log_get_buf - Get and initialize a buffer to use for log control data - * @sdp: The GFS2 superblock - * - * Returns: the buffer_head - */ - -struct buffer_head *gfs2_log_get_buf(struct gfs2_sbd *sdp) +static int ip_cmp(void *priv, struct list_head *a, struct list_head *b) { - u64 blkno = log_bmap(sdp, sdp->sd_log_flush_head); - struct buffer_head *bh; + struct gfs2_inode *ipa, *ipb; - bh = sb_getblk(sdp->sd_vfs, blkno); - lock_buffer(bh); - memset(bh->b_data, 0, bh->b_size); - set_buffer_uptodate(bh); - clear_buffer_dirty(bh); - gfs2_log_incr_head(sdp); - atomic_inc(&sdp->sd_log_in_flight); - bh->b_private = sdp; - bh->b_end_io = gfs2_log_write_endio; + ipa = list_entry(a, struct gfs2_inode, i_ordered); + ipb = list_entry(b, struct gfs2_inode, i_ordered); - return bh; + if (ipa->i_no_addr < ipb->i_no_addr) + return -1; + if (ipa->i_no_addr > ipb->i_no_addr) + return 1; + return 0; } -/** - * gfs2_fake_write_endio - - * @bh: The buffer head - * @uptodate: The I/O Status - * - */ - -static void gfs2_fake_write_endio(struct buffer_head *bh, int uptodate) +static void gfs2_ordered_write(struct gfs2_sbd *sdp) { - struct buffer_head *real_bh = bh->b_private; - struct gfs2_bufdata *bd = real_bh->b_private; - struct gfs2_sbd *sdp = bd->bd_gl->gl_sbd; + struct gfs2_inode *ip; + LIST_HEAD(written); - end_buffer_write_sync(bh, uptodate); - free_buffer_head(bh); - unlock_buffer(real_bh); - brelse(real_bh); - if (atomic_dec_and_test(&sdp->sd_log_in_flight)) - wake_up(&sdp->sd_log_flush_wait); + spin_lock(&sdp->sd_ordered_lock); + list_sort(NULL, &sdp->sd_log_le_ordered, &ip_cmp); + while (!list_empty(&sdp->sd_log_le_ordered)) { + ip = list_entry(sdp->sd_log_le_ordered.next, struct gfs2_inode, i_ordered); + list_move(&ip->i_ordered, &written); + if (ip->i_inode.i_mapping->nrpages == 0) + continue; + spin_unlock(&sdp->sd_ordered_lock); + filemap_fdatawrite(ip->i_inode.i_mapping); + spin_lock(&sdp->sd_ordered_lock); + } + list_splice(&written, &sdp->sd_log_le_ordered); + spin_unlock(&sdp->sd_ordered_lock); } -/** - * gfs2_log_fake_buf - Build a fake buffer head to write metadata buffer to log - * @sdp: the filesystem - * @data: the data the buffer_head should point to - * - * Returns: the log buffer descriptor - */ - -struct buffer_head *gfs2_log_fake_buf(struct gfs2_sbd *sdp, - struct buffer_head *real) +static void gfs2_ordered_wait(struct gfs2_sbd *sdp) { - u64 blkno = log_bmap(sdp, sdp->sd_log_flush_head); - struct buffer_head *bh; + struct gfs2_inode *ip; - bh = alloc_buffer_head(GFP_NOFS | __GFP_NOFAIL); - atomic_set(&bh->b_count, 1); - bh->b_state = (1 << BH_Mapped) | (1 << BH_Uptodate) | (1 << BH_Lock); - set_bh_page(bh, real->b_page, bh_offset(real)); - bh->b_blocknr = blkno; - bh->b_size = sdp->sd_sb.sb_bsize; - bh->b_bdev = sdp->sd_vfs->s_bdev; - bh->b_private = real; - bh->b_end_io = gfs2_fake_write_endio; + spin_lock(&sdp->sd_ordered_lock); + while (!list_empty(&sdp->sd_log_le_ordered)) { + ip = list_entry(sdp->sd_log_le_ordered.next, struct gfs2_inode, i_ordered); + list_del(&ip->i_ordered); + WARN_ON(!test_and_clear_bit(GIF_ORDERED, &ip->i_flags)); + if (ip->i_inode.i_mapping->nrpages == 0) + continue; + spin_unlock(&sdp->sd_ordered_lock); + filemap_fdatawait(ip->i_inode.i_mapping); + spin_lock(&sdp->sd_ordered_lock); + } + spin_unlock(&sdp->sd_ordered_lock); +} - gfs2_log_incr_head(sdp); - atomic_inc(&sdp->sd_log_in_flight); +void gfs2_ordered_del_inode(struct gfs2_inode *ip) +{ + struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); - return bh; + spin_lock(&sdp->sd_ordered_lock); + if (test_and_clear_bit(GIF_ORDERED, &ip->i_flags)) + list_del(&ip->i_ordered); + spin_unlock(&sdp->sd_ordered_lock); } -static void log_pull_tail(struct gfs2_sbd *sdp, unsigned int new_tail) +void gfs2_add_revoke(struct gfs2_sbd *sdp, struct gfs2_bufdata *bd) { - unsigned int dist = log_distance(sdp, new_tail, sdp->sd_log_tail); + struct buffer_head *bh = bd->bd_bh; + struct gfs2_glock *gl = bd->bd_gl; - ail2_empty(sdp, new_tail); - - atomic_add(dist, &sdp->sd_log_blks_free); - trace_gfs2_log_blocks(sdp, dist); - gfs2_assert_withdraw(sdp, atomic_read(&sdp->sd_log_blks_free) <= - sdp->sd_jdesc->jd_blocks); + bh->b_private = NULL; + bd->bd_blkno = bh->b_blocknr; + gfs2_remove_from_ail(bd); /* drops ref on bh */ + bd->bd_bh = NULL; + bd->bd_ops = &gfs2_revoke_lops; + sdp->sd_log_num_revoke++; + atomic_inc(&gl->gl_revokes); + set_bit(GLF_LFLUSH, &gl->gl_flags); + list_add(&bd->bd_list, &sdp->sd_log_le_revoke); +} + +void gfs2_write_revokes(struct gfs2_sbd *sdp) +{ + struct gfs2_trans *tr; + struct gfs2_bufdata *bd, *tmp; + int have_revokes = 0; + int max_revokes = (sdp->sd_sb.sb_bsize - sizeof(struct gfs2_log_descriptor)) / sizeof(u64); + + gfs2_ail1_empty(sdp); + spin_lock(&sdp->sd_ail_lock); + list_for_each_entry(tr, &sdp->sd_ail1_list, tr_list) { + list_for_each_entry(bd, &tr->tr_ail2_list, bd_ail_st_list) { + if (list_empty(&bd->bd_list)) { + have_revokes = 1; + goto done; + } + } + } +done: + spin_unlock(&sdp->sd_ail_lock); + if (have_revokes == 0) + return; + while (sdp->sd_log_num_revoke > max_revokes) + max_revokes += (sdp->sd_sb.sb_bsize - sizeof(struct gfs2_meta_header)) / sizeof(u64); + max_revokes -= sdp->sd_log_num_revoke; + if (!sdp->sd_log_num_revoke) { + atomic_dec(&sdp->sd_log_blks_free); + /* If no blocks have been reserved, we need to also + * reserve a block for the header */ + if (!sdp->sd_log_blks_reserved) + atomic_dec(&sdp->sd_log_blks_free); + } + gfs2_log_lock(sdp); + spin_lock(&sdp->sd_ail_lock); + list_for_each_entry(tr, &sdp->sd_ail1_list, tr_list) { + list_for_each_entry_safe(bd, tmp, &tr->tr_ail2_list, bd_ail_st_list) { + if (max_revokes == 0) + goto out_of_blocks; + if (!list_empty(&bd->bd_list)) + continue; + gfs2_add_revoke(sdp, bd); + max_revokes--; + } + } +out_of_blocks: + spin_unlock(&sdp->sd_ail_lock); + gfs2_log_unlock(sdp); - sdp->sd_log_tail = new_tail; + if (!sdp->sd_log_num_revoke) { + atomic_inc(&sdp->sd_log_blks_free); + if (!sdp->sd_log_blks_reserved) + atomic_inc(&sdp->sd_log_blks_free); + } } /** @@ -560,25 +645,18 @@ static void log_pull_tail(struct gfs2_sbd *sdp, unsigned int new_tail) * Returns: the initialized log buffer descriptor */ -static void log_write_header(struct gfs2_sbd *sdp, u32 flags, int pull) +static void log_write_header(struct gfs2_sbd *sdp, u32 flags) { - u64 blkno = log_bmap(sdp, sdp->sd_log_flush_head); - struct buffer_head *bh; struct gfs2_log_header *lh; unsigned int tail; u32 hash; + int rw = WRITE_FLUSH_FUA | REQ_META; + struct page *page = mempool_alloc(gfs2_page_pool, GFP_NOIO); + lh = page_address(page); + clear_page(lh); - bh = sb_getblk(sdp->sd_vfs, blkno); - lock_buffer(bh); - memset(bh->b_data, 0, bh->b_size); - set_buffer_uptodate(bh); - clear_buffer_dirty(bh); - - gfs2_ail1_empty(sdp, 0); tail = current_tail(sdp); - lh = (struct gfs2_log_header *)bh->b_data; - memset(lh, 0, sizeof(struct gfs2_log_header)); lh->lh_header.mh_magic = cpu_to_be32(GFS2_MAGIC); lh->lh_header.mh_type = cpu_to_be32(GFS2_METATYPE_LH); lh->lh_header.__pad0 = cpu_to_be64(0); @@ -588,96 +666,22 @@ static void log_write_header(struct gfs2_sbd *sdp, u32 flags, int pull) lh->lh_flags = cpu_to_be32(flags); lh->lh_tail = cpu_to_be32(tail); lh->lh_blkno = cpu_to_be32(sdp->sd_log_flush_head); - hash = gfs2_disk_hash(bh->b_data, sizeof(struct gfs2_log_header)); + hash = gfs2_disk_hash(page_address(page), sizeof(struct gfs2_log_header)); lh->lh_hash = cpu_to_be32(hash); - bh->b_end_io = end_buffer_write_sync; - get_bh(bh); - if (test_bit(SDF_NOBARRIERS, &sdp->sd_flags)) - submit_bh(WRITE_SYNC | REQ_META, bh); - else - submit_bh(WRITE_FLUSH_FUA | REQ_META, bh); - wait_on_buffer(bh); + if (test_bit(SDF_NOBARRIERS, &sdp->sd_flags)) { + gfs2_ordered_wait(sdp); + log_flush_wait(sdp); + rw = WRITE_SYNC | REQ_META | REQ_PRIO; + } - if (!buffer_uptodate(bh)) - gfs2_io_error_bh(sdp, bh); - brelse(bh); + sdp->sd_log_idle = (tail == sdp->sd_log_flush_head); + gfs2_log_write_page(sdp, page); + gfs2_log_flush_bio(sdp, rw); + log_flush_wait(sdp); if (sdp->sd_log_tail != tail) log_pull_tail(sdp, tail); - else - gfs2_assert_withdraw(sdp, !pull); - - sdp->sd_log_idle = (tail == sdp->sd_log_flush_head); - gfs2_log_incr_head(sdp); -} - -static void log_flush_commit(struct gfs2_sbd *sdp) -{ - DEFINE_WAIT(wait); - - if (atomic_read(&sdp->sd_log_in_flight)) { - do { - prepare_to_wait(&sdp->sd_log_flush_wait, &wait, - TASK_UNINTERRUPTIBLE); - if (atomic_read(&sdp->sd_log_in_flight)) - io_schedule(); - } while(atomic_read(&sdp->sd_log_in_flight)); - finish_wait(&sdp->sd_log_flush_wait, &wait); - } - - log_write_header(sdp, 0, 0); -} - -static void gfs2_ordered_write(struct gfs2_sbd *sdp) -{ - struct gfs2_bufdata *bd; - struct buffer_head *bh; - LIST_HEAD(written); - - gfs2_log_lock(sdp); - while (!list_empty(&sdp->sd_log_le_ordered)) { - bd = list_entry(sdp->sd_log_le_ordered.next, struct gfs2_bufdata, bd_le.le_list); - list_move(&bd->bd_le.le_list, &written); - bh = bd->bd_bh; - if (!buffer_dirty(bh)) - continue; - get_bh(bh); - gfs2_log_unlock(sdp); - lock_buffer(bh); - if (buffer_mapped(bh) && test_clear_buffer_dirty(bh)) { - bh->b_end_io = end_buffer_write_sync; - submit_bh(WRITE_SYNC_PLUG, bh); - } else { - unlock_buffer(bh); - brelse(bh); - } - gfs2_log_lock(sdp); - } - list_splice(&written, &sdp->sd_log_le_ordered); - gfs2_log_unlock(sdp); -} - -static void gfs2_ordered_wait(struct gfs2_sbd *sdp) -{ - struct gfs2_bufdata *bd; - struct buffer_head *bh; - - gfs2_log_lock(sdp); - while (!list_empty(&sdp->sd_log_le_ordered)) { - bd = list_entry(sdp->sd_log_le_ordered.prev, struct gfs2_bufdata, bd_le.le_list); - bh = bd->bd_bh; - if (buffer_locked(bh)) { - get_bh(bh); - gfs2_log_unlock(sdp); - wait_on_buffer(bh); - brelse(bh); - gfs2_log_lock(sdp); - continue; - } - list_del_init(&bd->bd_le.le_list); - } - gfs2_log_unlock(sdp); } /** @@ -687,9 +691,10 @@ static void gfs2_ordered_wait(struct gfs2_sbd *sdp) * */ -void gfs2_log_flush(struct gfs2_sbd *sdp, struct gfs2_glock *gl) +void gfs2_log_flush(struct gfs2_sbd *sdp, struct gfs2_glock *gl, + enum gfs2_flush_type type) { - struct gfs2_ail *ai; + struct gfs2_trans *tr; down_write(&sdp->sd_log_flush_lock); @@ -700,76 +705,129 @@ void gfs2_log_flush(struct gfs2_sbd *sdp, struct gfs2_glock *gl) } trace_gfs2_log_flush(sdp, 1); - ai = kzalloc(sizeof(struct gfs2_ail), GFP_NOFS | __GFP_NOFAIL); - INIT_LIST_HEAD(&ai->ai_ail1_list); - INIT_LIST_HEAD(&ai->ai_ail2_list); - - if (sdp->sd_log_num_buf != sdp->sd_log_commited_buf) { - printk(KERN_INFO "GFS2: log buf %u %u\n", sdp->sd_log_num_buf, - sdp->sd_log_commited_buf); - gfs2_assert_withdraw(sdp, 0); - } - if (sdp->sd_log_num_databuf != sdp->sd_log_commited_databuf) { - printk(KERN_INFO "GFS2: log databuf %u %u\n", - sdp->sd_log_num_databuf, sdp->sd_log_commited_databuf); - gfs2_assert_withdraw(sdp, 0); + sdp->sd_log_flush_head = sdp->sd_log_head; + sdp->sd_log_flush_wrapped = 0; + tr = sdp->sd_log_tr; + if (tr) { + sdp->sd_log_tr = NULL; + INIT_LIST_HEAD(&tr->tr_ail1_list); + INIT_LIST_HEAD(&tr->tr_ail2_list); + tr->tr_first = sdp->sd_log_flush_head; } + gfs2_assert_withdraw(sdp, sdp->sd_log_num_revoke == sdp->sd_log_commited_revoke); - sdp->sd_log_flush_head = sdp->sd_log_head; - sdp->sd_log_flush_wrapped = 0; - ai->ai_first = sdp->sd_log_flush_head; - gfs2_ordered_write(sdp); - lops_before_commit(sdp); - gfs2_ordered_wait(sdp); + lops_before_commit(sdp, tr); + gfs2_log_flush_bio(sdp, WRITE); - if (sdp->sd_log_head != sdp->sd_log_flush_head) - log_flush_commit(sdp); - else if (sdp->sd_log_tail != current_tail(sdp) && !sdp->sd_log_idle){ - gfs2_log_lock(sdp); + if (sdp->sd_log_head != sdp->sd_log_flush_head) { + log_flush_wait(sdp); + log_write_header(sdp, 0); + } else if (sdp->sd_log_tail != current_tail(sdp) && !sdp->sd_log_idle){ atomic_dec(&sdp->sd_log_blks_free); /* Adjust for unreserved buffer */ trace_gfs2_log_blocks(sdp, -1); - gfs2_log_unlock(sdp); - log_write_header(sdp, 0, PULL); + log_write_header(sdp, 0); } - lops_after_commit(sdp, ai); + lops_after_commit(sdp, tr); gfs2_log_lock(sdp); sdp->sd_log_head = sdp->sd_log_flush_head; sdp->sd_log_blks_reserved = 0; - sdp->sd_log_commited_buf = 0; - sdp->sd_log_commited_databuf = 0; sdp->sd_log_commited_revoke = 0; - if (!list_empty(&ai->ai_ail1_list)) { - list_add(&ai->ai_list, &sdp->sd_ail1_list); - ai = NULL; + spin_lock(&sdp->sd_ail_lock); + if (tr && !list_empty(&tr->tr_ail1_list)) { + list_add(&tr->tr_list, &sdp->sd_ail1_list); + tr = NULL; } + spin_unlock(&sdp->sd_ail_lock); gfs2_log_unlock(sdp); + + if (atomic_read(&sdp->sd_log_freeze)) + type = FREEZE_FLUSH; + if (type != NORMAL_FLUSH) { + if (!sdp->sd_log_idle) { + for (;;) { + gfs2_ail1_start(sdp); + gfs2_ail1_wait(sdp); + if (gfs2_ail1_empty(sdp)) + break; + } + atomic_dec(&sdp->sd_log_blks_free); /* Adjust for unreserved buffer */ + trace_gfs2_log_blocks(sdp, -1); + sdp->sd_log_flush_wrapped = 0; + log_write_header(sdp, 0); + sdp->sd_log_head = sdp->sd_log_flush_head; + } + if (type == SHUTDOWN_FLUSH || type == FREEZE_FLUSH) + gfs2_log_shutdown(sdp); + if (type == FREEZE_FLUSH) { + int error; + + atomic_set(&sdp->sd_log_freeze, 0); + wake_up(&sdp->sd_log_frozen_wait); + error = gfs2_glock_nq_init(sdp->sd_freeze_gl, + LM_ST_SHARED, 0, + &sdp->sd_thaw_gh); + if (error) { + printk(KERN_INFO "GFS2: couln't get freeze lock : %d\n", error); + gfs2_assert_withdraw(sdp, 0); + } + else + gfs2_glock_dq_uninit(&sdp->sd_thaw_gh); + } + } + trace_gfs2_log_flush(sdp, 0); up_write(&sdp->sd_log_flush_lock); - kfree(ai); + kfree(tr); +} + +/** + * gfs2_merge_trans - Merge a new transaction into a cached transaction + * @old: Original transaction to be expanded + * @new: New transaction to be merged + */ + +static void gfs2_merge_trans(struct gfs2_trans *old, struct gfs2_trans *new) +{ + WARN_ON_ONCE(old->tr_attached != 1); + + old->tr_num_buf_new += new->tr_num_buf_new; + old->tr_num_databuf_new += new->tr_num_databuf_new; + old->tr_num_buf_rm += new->tr_num_buf_rm; + old->tr_num_databuf_rm += new->tr_num_databuf_rm; + old->tr_num_revoke += new->tr_num_revoke; + old->tr_num_revoke_rm += new->tr_num_revoke_rm; + + list_splice_tail_init(&new->tr_databuf, &old->tr_databuf); + list_splice_tail_init(&new->tr_buf, &old->tr_buf); } static void log_refund(struct gfs2_sbd *sdp, struct gfs2_trans *tr) { unsigned int reserved; unsigned int unused; + unsigned int maxres; gfs2_log_lock(sdp); - sdp->sd_log_commited_buf += tr->tr_num_buf_new - tr->tr_num_buf_rm; - sdp->sd_log_commited_databuf += tr->tr_num_databuf_new - - tr->tr_num_databuf_rm; - gfs2_assert_withdraw(sdp, (((int)sdp->sd_log_commited_buf) >= 0) || - (((int)sdp->sd_log_commited_databuf) >= 0)); + if (sdp->sd_log_tr) { + gfs2_merge_trans(sdp->sd_log_tr, tr); + } else if (tr->tr_num_buf_new || tr->tr_num_databuf_new) { + gfs2_assert_withdraw(sdp, tr->tr_alloced); + sdp->sd_log_tr = tr; + tr->tr_attached = 1; + } + sdp->sd_log_commited_revoke += tr->tr_num_revoke - tr->tr_num_revoke_rm; reserved = calc_reserved(sdp); - gfs2_assert_withdraw(sdp, sdp->sd_log_blks_reserved + tr->tr_reserved >= reserved); - unused = sdp->sd_log_blks_reserved - reserved + tr->tr_reserved; + maxres = sdp->sd_log_blks_reserved + tr->tr_reserved; + gfs2_assert_withdraw(sdp, maxres >= reserved); + unused = maxres - reserved; atomic_add(unused, &sdp->sd_log_blks_free); trace_gfs2_log_blocks(sdp, unused); gfs2_assert_withdraw(sdp, atomic_read(&sdp->sd_log_blks_free) <= @@ -779,21 +837,6 @@ static void log_refund(struct gfs2_sbd *sdp, struct gfs2_trans *tr) gfs2_log_unlock(sdp); } -static void buf_lo_incore_commit(struct gfs2_sbd *sdp, struct gfs2_trans *tr) -{ - struct list_head *head = &tr->tr_list_buf; - struct gfs2_bufdata *bd; - - gfs2_log_lock(sdp); - while (!list_empty(head)) { - bd = list_entry(head->next, struct gfs2_bufdata, bd_list_tr); - list_del_init(&bd->bd_list_tr); - tr->tr_num_buf--; - } - gfs2_log_unlock(sdp); - gfs2_assert_warn(sdp, !tr->tr_num_buf); -} - /** * gfs2_log_commit - Commit a transaction to the log * @sdp: the filesystem @@ -812,9 +855,6 @@ static void buf_lo_incore_commit(struct gfs2_sbd *sdp, struct gfs2_trans *tr) void gfs2_log_commit(struct gfs2_sbd *sdp, struct gfs2_trans *tr) { log_refund(sdp, tr); - buf_lo_incore_commit(sdp, tr); - - up_read(&sdp->sd_log_flush_lock); if (atomic_read(&sdp->sd_log_pinned) > atomic_read(&sdp->sd_log_thresh1) || ((sdp->sd_jdesc->jd_blocks - atomic_read(&sdp->sd_log_blks_free)) > @@ -830,52 +870,25 @@ void gfs2_log_commit(struct gfs2_sbd *sdp, struct gfs2_trans *tr) void gfs2_log_shutdown(struct gfs2_sbd *sdp) { - down_write(&sdp->sd_log_flush_lock); - gfs2_assert_withdraw(sdp, !sdp->sd_log_blks_reserved); - gfs2_assert_withdraw(sdp, !sdp->sd_log_num_buf); gfs2_assert_withdraw(sdp, !sdp->sd_log_num_revoke); - gfs2_assert_withdraw(sdp, !sdp->sd_log_num_rg); - gfs2_assert_withdraw(sdp, !sdp->sd_log_num_databuf); gfs2_assert_withdraw(sdp, list_empty(&sdp->sd_ail1_list)); sdp->sd_log_flush_head = sdp->sd_log_head; sdp->sd_log_flush_wrapped = 0; - log_write_header(sdp, GFS2_LOG_HEAD_UNMOUNT, - (sdp->sd_log_tail == current_tail(sdp)) ? 0 : PULL); + log_write_header(sdp, GFS2_LOG_HEAD_UNMOUNT); - gfs2_assert_warn(sdp, atomic_read(&sdp->sd_log_blks_free) == sdp->sd_jdesc->jd_blocks); gfs2_assert_warn(sdp, sdp->sd_log_head == sdp->sd_log_tail); gfs2_assert_warn(sdp, list_empty(&sdp->sd_ail2_list)); sdp->sd_log_head = sdp->sd_log_flush_head; sdp->sd_log_tail = sdp->sd_log_head; - - up_write(&sdp->sd_log_flush_lock); -} - - -/** - * gfs2_meta_syncfs - sync all the buffers in a filesystem - * @sdp: the filesystem - * - */ - -void gfs2_meta_syncfs(struct gfs2_sbd *sdp) -{ - gfs2_log_flush(sdp, NULL); - for (;;) { - gfs2_ail1_start(sdp); - if (gfs2_ail1_empty(sdp, DIO_ALL)) - break; - msleep(10); - } } static inline int gfs2_jrnl_flush_reqd(struct gfs2_sbd *sdp) { - return (atomic_read(&sdp->sd_log_pinned) >= atomic_read(&sdp->sd_log_thresh1)); + return (atomic_read(&sdp->sd_log_pinned) >= atomic_read(&sdp->sd_log_thresh1) || atomic_read(&sdp->sd_log_freeze)); } static inline int gfs2_ail_flush_reqd(struct gfs2_sbd *sdp) @@ -897,29 +910,27 @@ int gfs2_logd(void *data) struct gfs2_sbd *sdp = data; unsigned long t = 1; DEFINE_WAIT(wait); - unsigned preflush; while (!kthread_should_stop()) { - preflush = atomic_read(&sdp->sd_log_pinned); if (gfs2_jrnl_flush_reqd(sdp) || t == 0) { - gfs2_ail1_empty(sdp, DIO_ALL); - gfs2_log_flush(sdp, NULL); - gfs2_ail1_empty(sdp, DIO_ALL); + gfs2_ail1_empty(sdp); + gfs2_log_flush(sdp, NULL, NORMAL_FLUSH); } if (gfs2_ail_flush_reqd(sdp)) { gfs2_ail1_start(sdp); - io_schedule(); - gfs2_ail1_empty(sdp, 0); - gfs2_log_flush(sdp, NULL); - gfs2_ail1_empty(sdp, DIO_ALL); + gfs2_ail1_wait(sdp); + gfs2_ail1_empty(sdp); + gfs2_log_flush(sdp, NULL, NORMAL_FLUSH); } - wake_up(&sdp->sd_log_waitq); + if (!gfs2_ail_flush_reqd(sdp)) + wake_up(&sdp->sd_log_waitq); + t = gfs2_tune_get(sdp, gt_logd_secs) * HZ; - if (freezing(current)) - refrigerator(); + + try_to_freeze(); do { prepare_to_wait(&sdp->sd_logd_waitq, &wait, diff --git a/fs/gfs2/log.h b/fs/gfs2/log.h index 0d007f92023..9499a604921 100644 --- a/fs/gfs2/log.h +++ b/fs/gfs2/log.h @@ -12,6 +12,7 @@ #include <linux/list.h> #include <linux/spinlock.h> +#include <linux/writeback.h> #include "incore.h" /** @@ -47,21 +48,38 @@ static inline void gfs2_log_pointers_init(struct gfs2_sbd *sdp, sdp->sd_log_head = sdp->sd_log_tail = value; } +static inline void gfs2_ordered_add_inode(struct gfs2_inode *ip) +{ + struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); + + if (!test_bit(GIF_ORDERED, &ip->i_flags)) { + spin_lock(&sdp->sd_ordered_lock); + if (!test_and_set_bit(GIF_ORDERED, &ip->i_flags)) + list_add(&ip->i_ordered, &sdp->sd_log_le_ordered); + spin_unlock(&sdp->sd_ordered_lock); + } +} +extern void gfs2_ordered_del_inode(struct gfs2_inode *ip); extern unsigned int gfs2_struct2blk(struct gfs2_sbd *sdp, unsigned int nstruct, unsigned int ssize); +extern void gfs2_log_release(struct gfs2_sbd *sdp, unsigned int blks); extern int gfs2_log_reserve(struct gfs2_sbd *sdp, unsigned int blks); -extern void gfs2_log_incr_head(struct gfs2_sbd *sdp); - -extern struct buffer_head *gfs2_log_get_buf(struct gfs2_sbd *sdp); -extern struct buffer_head *gfs2_log_fake_buf(struct gfs2_sbd *sdp, - struct buffer_head *real); -extern void gfs2_log_flush(struct gfs2_sbd *sdp, struct gfs2_glock *gl); +enum gfs2_flush_type { + NORMAL_FLUSH = 0, + SYNC_FLUSH, + SHUTDOWN_FLUSH, + FREEZE_FLUSH +}; +extern void gfs2_log_flush(struct gfs2_sbd *sdp, struct gfs2_glock *gl, + enum gfs2_flush_type type); extern void gfs2_log_commit(struct gfs2_sbd *sdp, struct gfs2_trans *trans); extern void gfs2_remove_from_ail(struct gfs2_bufdata *bd); +extern void gfs2_ail1_flush(struct gfs2_sbd *sdp, struct writeback_control *wbc); extern void gfs2_log_shutdown(struct gfs2_sbd *sdp); -extern void gfs2_meta_syncfs(struct gfs2_sbd *sdp); extern int gfs2_logd(void *data); +extern void gfs2_add_revoke(struct gfs2_sbd *sdp, struct gfs2_bufdata *bd); +extern void gfs2_write_revokes(struct gfs2_sbd *sdp); #endif /* __LOG_DOT_H__ */ diff --git a/fs/gfs2/lops.c b/fs/gfs2/lops.c index bf33f822058..2c1ae861dc9 100644 --- a/fs/gfs2/lops.c +++ b/fs/gfs2/lops.c @@ -12,9 +12,11 @@ #include <linux/spinlock.h> #include <linux/completion.h> #include <linux/buffer_head.h> +#include <linux/mempool.h> #include <linux/gfs2_ondisk.h> #include <linux/bio.h> #include <linux/fs.h> +#include <linux/list_sort.h> #include "gfs2.h" #include "incore.h" @@ -36,11 +38,11 @@ * * The log lock must be held when calling this function */ -static void gfs2_pin(struct gfs2_sbd *sdp, struct buffer_head *bh) +void gfs2_pin(struct gfs2_sbd *sdp, struct buffer_head *bh) { struct gfs2_bufdata *bd; - gfs2_assert_withdraw(sdp, test_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags)); + BUG_ON(!current->journal_info); clear_buffer_dirty(bh); if (test_set_buffer_pinned(bh)) @@ -51,37 +53,65 @@ static void gfs2_pin(struct gfs2_sbd *sdp, struct buffer_head *bh) /* If this buffer is in the AIL and it has already been written * to in-place disk block, remove it from the AIL. */ - if (bd->bd_ail) - list_move(&bd->bd_ail_st_list, &bd->bd_ail->ai_ail2_list); + spin_lock(&sdp->sd_ail_lock); + if (bd->bd_tr) + list_move(&bd->bd_ail_st_list, &bd->bd_tr->tr_ail2_list); + spin_unlock(&sdp->sd_ail_lock); get_bh(bh); atomic_inc(&sdp->sd_log_pinned); trace_gfs2_pin(bd, 1); } +static bool buffer_is_rgrp(const struct gfs2_bufdata *bd) +{ + return bd->bd_gl->gl_name.ln_type == LM_TYPE_RGRP; +} + +static void maybe_release_space(struct gfs2_bufdata *bd) +{ + struct gfs2_glock *gl = bd->bd_gl; + struct gfs2_sbd *sdp = gl->gl_sbd; + struct gfs2_rgrpd *rgd = gl->gl_object; + unsigned int index = bd->bd_bh->b_blocknr - gl->gl_name.ln_number; + struct gfs2_bitmap *bi = rgd->rd_bits + index; + + if (bi->bi_clone == NULL) + return; + if (sdp->sd_args.ar_discard) + gfs2_rgrp_send_discards(sdp, rgd->rd_data0, bd->bd_bh, bi, 1, NULL); + memcpy(bi->bi_clone + bi->bi_offset, + bd->bd_bh->b_data + bi->bi_offset, bi->bi_len); + clear_bit(GBF_FULL, &bi->bi_flags); + rgd->rd_free_clone = rgd->rd_free; + rgd->rd_extfail_pt = rgd->rd_free; +} + /** * gfs2_unpin - Unpin a buffer * @sdp: the filesystem the buffer belongs to * @bh: The buffer to unpin * @ai: + * @flags: The inode dirty flags * */ static void gfs2_unpin(struct gfs2_sbd *sdp, struct buffer_head *bh, - struct gfs2_ail *ai) + struct gfs2_trans *tr) { struct gfs2_bufdata *bd = bh->b_private; - gfs2_assert_withdraw(sdp, buffer_uptodate(bh)); - - if (!buffer_pinned(bh)) - gfs2_assert_withdraw(sdp, 0); + BUG_ON(!buffer_uptodate(bh)); + BUG_ON(!buffer_pinned(bh)); lock_buffer(bh); mark_buffer_dirty(bh); clear_buffer_pinned(bh); - gfs2_log_lock(sdp); - if (bd->bd_ail) { + if (buffer_is_rgrp(bd)) + maybe_release_space(bd); + + spin_lock(&sdp->sd_ail_lock); + if (bd->bd_tr) { list_del(&bd->bd_ail_st_list); brelse(bh); } else { @@ -89,128 +119,367 @@ static void gfs2_unpin(struct gfs2_sbd *sdp, struct buffer_head *bh, list_add(&bd->bd_ail_gl_list, &gl->gl_ail_list); atomic_inc(&gl->gl_ail_count); } - bd->bd_ail = ai; - list_add(&bd->bd_ail_st_list, &ai->ai_ail1_list); + bd->bd_tr = tr; + list_add(&bd->bd_ail_st_list, &tr->tr_ail1_list); + spin_unlock(&sdp->sd_ail_lock); + clear_bit(GLF_LFLUSH, &bd->bd_gl->gl_flags); trace_gfs2_pin(bd, 0); - gfs2_log_unlock(sdp); unlock_buffer(bh); atomic_dec(&sdp->sd_log_pinned); } +static void gfs2_log_incr_head(struct gfs2_sbd *sdp) +{ + BUG_ON((sdp->sd_log_flush_head == sdp->sd_log_tail) && + (sdp->sd_log_flush_head != sdp->sd_log_head)); + + if (++sdp->sd_log_flush_head == sdp->sd_jdesc->jd_blocks) { + sdp->sd_log_flush_head = 0; + sdp->sd_log_flush_wrapped = 1; + } +} + +static u64 gfs2_log_bmap(struct gfs2_sbd *sdp) +{ + unsigned int lbn = sdp->sd_log_flush_head; + struct gfs2_journal_extent *je; + u64 block; + + list_for_each_entry(je, &sdp->sd_jdesc->extent_list, list) { + if ((lbn >= je->lblock) && (lbn < (je->lblock + je->blocks))) { + block = je->dblock + lbn - je->lblock; + gfs2_log_incr_head(sdp); + return block; + } + } + + return -1; +} + +/** + * gfs2_end_log_write_bh - end log write of pagecache data with buffers + * @sdp: The superblock + * @bvec: The bio_vec + * @error: The i/o status + * + * This finds the relavent buffers and unlocks then and sets the + * error flag according to the status of the i/o request. This is + * used when the log is writing data which has an in-place version + * that is pinned in the pagecache. + */ + +static void gfs2_end_log_write_bh(struct gfs2_sbd *sdp, struct bio_vec *bvec, + int error) +{ + struct buffer_head *bh, *next; + struct page *page = bvec->bv_page; + unsigned size; + + bh = page_buffers(page); + size = bvec->bv_len; + while (bh_offset(bh) < bvec->bv_offset) + bh = bh->b_this_page; + do { + if (error) + set_buffer_write_io_error(bh); + unlock_buffer(bh); + next = bh->b_this_page; + size -= bh->b_size; + brelse(bh); + bh = next; + } while(bh && size); +} + +/** + * gfs2_end_log_write - end of i/o to the log + * @bio: The bio + * @error: Status of i/o request + * + * Each bio_vec contains either data from the pagecache or data + * relating to the log itself. Here we iterate over the bio_vec + * array, processing both kinds of data. + * + */ + +static void gfs2_end_log_write(struct bio *bio, int error) +{ + struct gfs2_sbd *sdp = bio->bi_private; + struct bio_vec *bvec; + struct page *page; + int i; + + if (error) { + sdp->sd_log_error = error; + fs_err(sdp, "Error %d writing to log\n", error); + } + + bio_for_each_segment_all(bvec, bio, i) { + page = bvec->bv_page; + if (page_has_buffers(page)) + gfs2_end_log_write_bh(sdp, bvec, error); + else + mempool_free(page, gfs2_page_pool); + } + + bio_put(bio); + if (atomic_dec_and_test(&sdp->sd_log_in_flight)) + wake_up(&sdp->sd_log_flush_wait); +} + +/** + * gfs2_log_flush_bio - Submit any pending log bio + * @sdp: The superblock + * @rw: The rw flags + * + * Submit any pending part-built or full bio to the block device. If + * there is no pending bio, then this is a no-op. + */ + +void gfs2_log_flush_bio(struct gfs2_sbd *sdp, int rw) +{ + if (sdp->sd_log_bio) { + atomic_inc(&sdp->sd_log_in_flight); + submit_bio(rw, sdp->sd_log_bio); + sdp->sd_log_bio = NULL; + } +} + +/** + * gfs2_log_alloc_bio - Allocate a new bio for log writing + * @sdp: The superblock + * @blkno: The next device block number we want to write to + * + * This should never be called when there is a cached bio in the + * super block. When it returns, there will be a cached bio in the + * super block which will have as many bio_vecs as the device is + * happy to handle. + * + * Returns: Newly allocated bio + */ + +static struct bio *gfs2_log_alloc_bio(struct gfs2_sbd *sdp, u64 blkno) +{ + struct super_block *sb = sdp->sd_vfs; + unsigned nrvecs = bio_get_nr_vecs(sb->s_bdev); + struct bio *bio; + + BUG_ON(sdp->sd_log_bio); + + while (1) { + bio = bio_alloc(GFP_NOIO, nrvecs); + if (likely(bio)) + break; + nrvecs = max(nrvecs/2, 1U); + } + + bio->bi_iter.bi_sector = blkno * (sb->s_blocksize >> 9); + bio->bi_bdev = sb->s_bdev; + bio->bi_end_io = gfs2_end_log_write; + bio->bi_private = sdp; + + sdp->sd_log_bio = bio; + + return bio; +} + +/** + * gfs2_log_get_bio - Get cached log bio, or allocate a new one + * @sdp: The superblock + * @blkno: The device block number we want to write to + * + * If there is a cached bio, then if the next block number is sequential + * with the previous one, return it, otherwise flush the bio to the + * device. If there is not a cached bio, or we just flushed it, then + * allocate a new one. + * + * Returns: The bio to use for log writes + */ -static inline struct gfs2_log_descriptor *bh_log_desc(struct buffer_head *bh) +static struct bio *gfs2_log_get_bio(struct gfs2_sbd *sdp, u64 blkno) { - return (struct gfs2_log_descriptor *)bh->b_data; + struct bio *bio = sdp->sd_log_bio; + u64 nblk; + + if (bio) { + nblk = bio_end_sector(bio); + nblk >>= sdp->sd_fsb2bb_shift; + if (blkno == nblk) + return bio; + gfs2_log_flush_bio(sdp, WRITE); + } + + return gfs2_log_alloc_bio(sdp, blkno); } -static inline __be64 *bh_log_ptr(struct buffer_head *bh) + +/** + * gfs2_log_write - write to log + * @sdp: the filesystem + * @page: the page to write + * @size: the size of the data to write + * @offset: the offset within the page + * + * Try and add the page segment to the current bio. If that fails, + * submit the current bio to the device and create a new one, and + * then add the page segment to that. + */ + +static void gfs2_log_write(struct gfs2_sbd *sdp, struct page *page, + unsigned size, unsigned offset) { - struct gfs2_log_descriptor *ld = bh_log_desc(bh); - return (__force __be64 *)(ld + 1); + u64 blkno = gfs2_log_bmap(sdp); + struct bio *bio; + int ret; + + bio = gfs2_log_get_bio(sdp, blkno); + ret = bio_add_page(bio, page, size, offset); + if (ret == 0) { + gfs2_log_flush_bio(sdp, WRITE); + bio = gfs2_log_alloc_bio(sdp, blkno); + ret = bio_add_page(bio, page, size, offset); + WARN_ON(ret == 0); + } } -static inline __be64 *bh_ptr_end(struct buffer_head *bh) +/** + * gfs2_log_write_bh - write a buffer's content to the log + * @sdp: The super block + * @bh: The buffer pointing to the in-place location + * + * This writes the content of the buffer to the next available location + * in the log. The buffer will be unlocked once the i/o to the log has + * completed. + */ + +static void gfs2_log_write_bh(struct gfs2_sbd *sdp, struct buffer_head *bh) { - return (__force __be64 *)(bh->b_data + bh->b_size); + gfs2_log_write(sdp, bh->b_page, bh->b_size, bh_offset(bh)); } +/** + * gfs2_log_write_page - write one block stored in a page, into the log + * @sdp: The superblock + * @page: The struct page + * + * This writes the first block-sized part of the page into the log. Note + * that the page must have been allocated from the gfs2_page_pool mempool + * and that after this has been called, ownership has been transferred and + * the page may be freed at any time. + */ -static struct buffer_head *gfs2_get_log_desc(struct gfs2_sbd *sdp, u32 ld_type) +void gfs2_log_write_page(struct gfs2_sbd *sdp, struct page *page) { - struct buffer_head *bh = gfs2_log_get_buf(sdp); - struct gfs2_log_descriptor *ld = bh_log_desc(bh); + struct super_block *sb = sdp->sd_vfs; + gfs2_log_write(sdp, page, sb->s_blocksize, 0); +} + +static struct page *gfs2_get_log_desc(struct gfs2_sbd *sdp, u32 ld_type, + u32 ld_length, u32 ld_data1) +{ + struct page *page = mempool_alloc(gfs2_page_pool, GFP_NOIO); + struct gfs2_log_descriptor *ld = page_address(page); + clear_page(ld); ld->ld_header.mh_magic = cpu_to_be32(GFS2_MAGIC); ld->ld_header.mh_type = cpu_to_be32(GFS2_METATYPE_LD); ld->ld_header.mh_format = cpu_to_be32(GFS2_FORMAT_LD); ld->ld_type = cpu_to_be32(ld_type); - ld->ld_length = 0; - ld->ld_data1 = 0; + ld->ld_length = cpu_to_be32(ld_length); + ld->ld_data1 = cpu_to_be32(ld_data1); ld->ld_data2 = 0; - memset(ld->ld_reserved, 0, sizeof(ld->ld_reserved)); - return bh; + return page; } -static void buf_lo_add(struct gfs2_sbd *sdp, struct gfs2_log_element *le) +static void gfs2_check_magic(struct buffer_head *bh) { - struct gfs2_bufdata *bd = container_of(le, struct gfs2_bufdata, bd_le); - struct gfs2_meta_header *mh; - struct gfs2_trans *tr; + void *kaddr; + __be32 *ptr; - lock_buffer(bd->bd_bh); - gfs2_log_lock(sdp); - if (!list_empty(&bd->bd_list_tr)) - goto out; - tr = current->journal_info; - tr->tr_touched = 1; - tr->tr_num_buf++; - list_add(&bd->bd_list_tr, &tr->tr_list_buf); - if (!list_empty(&le->le_list)) - goto out; - set_bit(GLF_LFLUSH, &bd->bd_gl->gl_flags); - set_bit(GLF_DIRTY, &bd->bd_gl->gl_flags); - gfs2_meta_check(sdp, bd->bd_bh); - gfs2_pin(sdp, bd->bd_bh); - mh = (struct gfs2_meta_header *)bd->bd_bh->b_data; - mh->__pad0 = cpu_to_be64(0); - mh->mh_jid = cpu_to_be32(sdp->sd_jdesc->jd_jid); - sdp->sd_log_num_buf++; - list_add(&le->le_list, &sdp->sd_log_le_buf); - tr->tr_num_buf_new++; -out: - gfs2_log_unlock(sdp); - unlock_buffer(bd->bd_bh); + clear_buffer_escaped(bh); + kaddr = kmap_atomic(bh->b_page); + ptr = kaddr + bh_offset(bh); + if (*ptr == cpu_to_be32(GFS2_MAGIC)) + set_buffer_escaped(bh); + kunmap_atomic(kaddr); } -static void buf_lo_before_commit(struct gfs2_sbd *sdp) +static int blocknr_cmp(void *priv, struct list_head *a, struct list_head *b) +{ + struct gfs2_bufdata *bda, *bdb; + + bda = list_entry(a, struct gfs2_bufdata, bd_list); + bdb = list_entry(b, struct gfs2_bufdata, bd_list); + + if (bda->bd_bh->b_blocknr < bdb->bd_bh->b_blocknr) + return -1; + if (bda->bd_bh->b_blocknr > bdb->bd_bh->b_blocknr) + return 1; + return 0; +} + +static void gfs2_before_commit(struct gfs2_sbd *sdp, unsigned int limit, + unsigned int total, struct list_head *blist, + bool is_databuf) { - struct buffer_head *bh; struct gfs2_log_descriptor *ld; struct gfs2_bufdata *bd1 = NULL, *bd2; - unsigned int total; - unsigned int limit; + struct page *page; unsigned int num; unsigned n; __be64 *ptr; - limit = buf_limit(sdp); - /* for 4k blocks, limit = 503 */ - gfs2_log_lock(sdp); - total = sdp->sd_log_num_buf; - bd1 = bd2 = list_prepare_entry(bd1, &sdp->sd_log_le_buf, bd_le.le_list); + list_sort(NULL, blist, blocknr_cmp); + bd1 = bd2 = list_prepare_entry(bd1, blist, bd_list); while(total) { num = total; if (total > limit) num = limit; gfs2_log_unlock(sdp); - bh = gfs2_get_log_desc(sdp, GFS2_LOG_DESC_METADATA); + page = gfs2_get_log_desc(sdp, + is_databuf ? GFS2_LOG_DESC_JDATA : + GFS2_LOG_DESC_METADATA, num + 1, num); + ld = page_address(page); gfs2_log_lock(sdp); - ld = bh_log_desc(bh); - ptr = bh_log_ptr(bh); - ld->ld_length = cpu_to_be32(num + 1); - ld->ld_data1 = cpu_to_be32(num); + ptr = (__be64 *)(ld + 1); n = 0; - list_for_each_entry_continue(bd1, &sdp->sd_log_le_buf, - bd_le.le_list) { + list_for_each_entry_continue(bd1, blist, bd_list) { *ptr++ = cpu_to_be64(bd1->bd_bh->b_blocknr); + if (is_databuf) { + gfs2_check_magic(bd1->bd_bh); + *ptr++ = cpu_to_be64(buffer_escaped(bd1->bd_bh) ? 1 : 0); + } if (++n >= num) break; } gfs2_log_unlock(sdp); - submit_bh(WRITE_SYNC_PLUG, bh); + gfs2_log_write_page(sdp, page); gfs2_log_lock(sdp); n = 0; - list_for_each_entry_continue(bd2, &sdp->sd_log_le_buf, - bd_le.le_list) { + list_for_each_entry_continue(bd2, blist, bd_list) { get_bh(bd2->bd_bh); gfs2_log_unlock(sdp); lock_buffer(bd2->bd_bh); - bh = gfs2_log_fake_buf(sdp, bd2->bd_bh); - submit_bh(WRITE_SYNC_PLUG, bh); + + if (buffer_escaped(bd2->bd_bh)) { + void *kaddr; + page = mempool_alloc(gfs2_page_pool, GFP_NOIO); + ptr = page_address(page); + kaddr = kmap_atomic(bd2->bd_bh->b_page); + memcpy(ptr, kaddr + bh_offset(bd2->bd_bh), + bd2->bd_bh->b_size); + kunmap_atomic(kaddr); + *(__be32 *)ptr = 0; + clear_buffer_escaped(bd2->bd_bh); + unlock_buffer(bd2->bd_bh); + brelse(bd2->bd_bh); + gfs2_log_write_page(sdp, page); + } else { + gfs2_log_write_bh(sdp, bd2->bd_bh); + } gfs2_log_lock(sdp); if (++n >= num) break; @@ -222,31 +491,40 @@ static void buf_lo_before_commit(struct gfs2_sbd *sdp) gfs2_log_unlock(sdp); } -static void buf_lo_after_commit(struct gfs2_sbd *sdp, struct gfs2_ail *ai) +static void buf_lo_before_commit(struct gfs2_sbd *sdp, struct gfs2_trans *tr) +{ + unsigned int limit = buf_limit(sdp); /* 503 for 4k blocks */ + unsigned int nbuf; + if (tr == NULL) + return; + nbuf = tr->tr_num_buf_new - tr->tr_num_buf_rm; + gfs2_before_commit(sdp, limit, nbuf, &tr->tr_buf, 0); +} + +static void buf_lo_after_commit(struct gfs2_sbd *sdp, struct gfs2_trans *tr) { - struct list_head *head = &sdp->sd_log_le_buf; + struct list_head *head; struct gfs2_bufdata *bd; - while (!list_empty(head)) { - bd = list_entry(head->next, struct gfs2_bufdata, bd_le.le_list); - list_del_init(&bd->bd_le.le_list); - sdp->sd_log_num_buf--; + if (tr == NULL) + return; - gfs2_unpin(sdp, bd->bd_bh, ai); + head = &tr->tr_buf; + while (!list_empty(head)) { + bd = list_entry(head->next, struct gfs2_bufdata, bd_list); + list_del_init(&bd->bd_list); + gfs2_unpin(sdp, bd->bd_bh, tr); } - gfs2_assert_warn(sdp, !sdp->sd_log_num_buf); } static void buf_lo_before_scan(struct gfs2_jdesc *jd, struct gfs2_log_header_host *head, int pass) { - struct gfs2_sbd *sdp = GFS2_SB(jd->jd_inode); - if (pass != 0) return; - sdp->sd_found_blocks = 0; - sdp->sd_replayed_blocks = 0; + jd->jd_found_blocks = 0; + jd->jd_replayed_blocks = 0; } static int buf_lo_scan_elements(struct gfs2_jdesc *jd, unsigned int start, @@ -269,9 +547,9 @@ static int buf_lo_scan_elements(struct gfs2_jdesc *jd, unsigned int start, for (; blks; gfs2_replay_incr_blk(sdp, &start), blks--) { blkno = be64_to_cpu(*ptr++); - sdp->sd_found_blocks++; + jd->jd_found_blocks++; - if (gfs2_revoke_check(sdp, blkno, start)) + if (gfs2_revoke_check(jd, blkno, start)) continue; error = gfs2_replay_read_block(jd, start, &bh_log); @@ -292,12 +570,34 @@ static int buf_lo_scan_elements(struct gfs2_jdesc *jd, unsigned int start, if (error) break; - sdp->sd_replayed_blocks++; + jd->jd_replayed_blocks++; } return error; } +/** + * gfs2_meta_sync - Sync all buffers associated with a glock + * @gl: The glock + * + */ + +static void gfs2_meta_sync(struct gfs2_glock *gl) +{ + struct address_space *mapping = gfs2_glock2aspace(gl); + struct gfs2_sbd *sdp = gl->gl_sbd; + int error; + + if (mapping == NULL) + mapping = &sdp->sd_aspace; + + filemap_fdatawrite(mapping); + error = filemap_fdatawait(mapping); + + if (error) + gfs2_io_error(gl->gl_sbd); +} + static void buf_lo_after_scan(struct gfs2_jdesc *jd, int error, int pass) { struct gfs2_inode *ip = GFS2_I(jd->jd_inode); @@ -313,75 +613,73 @@ static void buf_lo_after_scan(struct gfs2_jdesc *jd, int error, int pass) gfs2_meta_sync(ip->i_gl); fs_info(sdp, "jid=%u: Replayed %u of %u blocks\n", - jd->jd_jid, sdp->sd_replayed_blocks, sdp->sd_found_blocks); + jd->jd_jid, jd->jd_replayed_blocks, jd->jd_found_blocks); } -static void revoke_lo_add(struct gfs2_sbd *sdp, struct gfs2_log_element *le) +static void revoke_lo_before_commit(struct gfs2_sbd *sdp, struct gfs2_trans *tr) { - struct gfs2_trans *tr; - - tr = current->journal_info; - tr->tr_touched = 1; - tr->tr_num_revoke++; - sdp->sd_log_num_revoke++; - list_add(&le->le_list, &sdp->sd_log_le_revoke); -} - -static void revoke_lo_before_commit(struct gfs2_sbd *sdp) -{ - struct gfs2_log_descriptor *ld; struct gfs2_meta_header *mh; - struct buffer_head *bh; unsigned int offset; struct list_head *head = &sdp->sd_log_le_revoke; struct gfs2_bufdata *bd; + struct page *page; + unsigned int length; + gfs2_write_revokes(sdp); if (!sdp->sd_log_num_revoke) return; - bh = gfs2_get_log_desc(sdp, GFS2_LOG_DESC_REVOKE); - ld = bh_log_desc(bh); - ld->ld_length = cpu_to_be32(gfs2_struct2blk(sdp, sdp->sd_log_num_revoke, - sizeof(u64))); - ld->ld_data1 = cpu_to_be32(sdp->sd_log_num_revoke); + length = gfs2_struct2blk(sdp, sdp->sd_log_num_revoke, sizeof(u64)); + page = gfs2_get_log_desc(sdp, GFS2_LOG_DESC_REVOKE, length, sdp->sd_log_num_revoke); offset = sizeof(struct gfs2_log_descriptor); - while (!list_empty(head)) { - bd = list_entry(head->next, struct gfs2_bufdata, bd_le.le_list); - list_del_init(&bd->bd_le.le_list); + list_for_each_entry(bd, head, bd_list) { sdp->sd_log_num_revoke--; if (offset + sizeof(u64) > sdp->sd_sb.sb_bsize) { - submit_bh(WRITE_SYNC_PLUG, bh); - bh = gfs2_log_get_buf(sdp); - mh = (struct gfs2_meta_header *)bh->b_data; + gfs2_log_write_page(sdp, page); + page = mempool_alloc(gfs2_page_pool, GFP_NOIO); + mh = page_address(page); + clear_page(mh); mh->mh_magic = cpu_to_be32(GFS2_MAGIC); mh->mh_type = cpu_to_be32(GFS2_METATYPE_LB); mh->mh_format = cpu_to_be32(GFS2_FORMAT_LB); offset = sizeof(struct gfs2_meta_header); } - *(__be64 *)(bh->b_data + offset) = cpu_to_be64(bd->bd_blkno); - kmem_cache_free(gfs2_bufdata_cachep, bd); - + *(__be64 *)(page_address(page) + offset) = cpu_to_be64(bd->bd_blkno); offset += sizeof(u64); } gfs2_assert_withdraw(sdp, !sdp->sd_log_num_revoke); - submit_bh(WRITE_SYNC_PLUG, bh); + gfs2_log_write_page(sdp, page); +} + +static void revoke_lo_after_commit(struct gfs2_sbd *sdp, struct gfs2_trans *tr) +{ + struct list_head *head = &sdp->sd_log_le_revoke; + struct gfs2_bufdata *bd; + struct gfs2_glock *gl; + + while (!list_empty(head)) { + bd = list_entry(head->next, struct gfs2_bufdata, bd_list); + list_del_init(&bd->bd_list); + gl = bd->bd_gl; + atomic_dec(&gl->gl_revokes); + clear_bit(GLF_LFLUSH, &gl->gl_flags); + kmem_cache_free(gfs2_bufdata_cachep, bd); + } } static void revoke_lo_before_scan(struct gfs2_jdesc *jd, struct gfs2_log_header_host *head, int pass) { - struct gfs2_sbd *sdp = GFS2_SB(jd->jd_inode); - if (pass != 0) return; - sdp->sd_found_revokes = 0; - sdp->sd_replay_tail = head->lh_tail; + jd->jd_found_revokes = 0; + jd->jd_replay_tail = head->lh_tail; } static int revoke_lo_scan_elements(struct gfs2_jdesc *jd, unsigned int start, @@ -413,13 +711,13 @@ static int revoke_lo_scan_elements(struct gfs2_jdesc *jd, unsigned int start, while (offset + sizeof(u64) <= sdp->sd_sb.sb_bsize) { blkno = be64_to_cpu(*(__be64 *)(bh->b_data + offset)); - error = gfs2_revoke_add(sdp, blkno, start); + error = gfs2_revoke_add(jd, blkno, start); if (error < 0) { brelse(bh); return error; } else if (error) - sdp->sd_found_revokes++; + jd->jd_found_revokes++; if (!--revokes) break; @@ -439,170 +737,16 @@ static void revoke_lo_after_scan(struct gfs2_jdesc *jd, int error, int pass) struct gfs2_sbd *sdp = GFS2_SB(jd->jd_inode); if (error) { - gfs2_revoke_clean(sdp); + gfs2_revoke_clean(jd); return; } if (pass != 1) return; fs_info(sdp, "jid=%u: Found %u revoke tags\n", - jd->jd_jid, sdp->sd_found_revokes); - - gfs2_revoke_clean(sdp); -} - -static void rg_lo_add(struct gfs2_sbd *sdp, struct gfs2_log_element *le) -{ - struct gfs2_rgrpd *rgd; - struct gfs2_trans *tr = current->journal_info; + jd->jd_jid, jd->jd_found_revokes); - tr->tr_touched = 1; - - rgd = container_of(le, struct gfs2_rgrpd, rd_le); - - gfs2_log_lock(sdp); - if (!list_empty(&le->le_list)){ - gfs2_log_unlock(sdp); - return; - } - gfs2_rgrp_bh_hold(rgd); - sdp->sd_log_num_rg++; - list_add(&le->le_list, &sdp->sd_log_le_rg); - gfs2_log_unlock(sdp); -} - -static void rg_lo_after_commit(struct gfs2_sbd *sdp, struct gfs2_ail *ai) -{ - struct list_head *head = &sdp->sd_log_le_rg; - struct gfs2_rgrpd *rgd; - - while (!list_empty(head)) { - rgd = list_entry(head->next, struct gfs2_rgrpd, rd_le.le_list); - list_del_init(&rgd->rd_le.le_list); - sdp->sd_log_num_rg--; - - gfs2_rgrp_repolish_clones(rgd); - gfs2_rgrp_bh_put(rgd); - } - gfs2_assert_warn(sdp, !sdp->sd_log_num_rg); -} - -/** - * databuf_lo_add - Add a databuf to the transaction. - * - * This is used in two distinct cases: - * i) In ordered write mode - * We put the data buffer on a list so that we can ensure that its - * synced to disk at the right time - * ii) In journaled data mode - * We need to journal the data block in the same way as metadata in - * the functions above. The difference is that here we have a tag - * which is two __be64's being the block number (as per meta data) - * and a flag which says whether the data block needs escaping or - * not. This means we need a new log entry for each 251 or so data - * blocks, which isn't an enormous overhead but twice as much as - * for normal metadata blocks. - */ -static void databuf_lo_add(struct gfs2_sbd *sdp, struct gfs2_log_element *le) -{ - struct gfs2_bufdata *bd = container_of(le, struct gfs2_bufdata, bd_le); - struct gfs2_trans *tr = current->journal_info; - struct address_space *mapping = bd->bd_bh->b_page->mapping; - struct gfs2_inode *ip = GFS2_I(mapping->host); - - lock_buffer(bd->bd_bh); - gfs2_log_lock(sdp); - if (tr) { - if (!list_empty(&bd->bd_list_tr)) - goto out; - tr->tr_touched = 1; - if (gfs2_is_jdata(ip)) { - tr->tr_num_buf++; - list_add(&bd->bd_list_tr, &tr->tr_list_buf); - } - } - if (!list_empty(&le->le_list)) - goto out; - - set_bit(GLF_LFLUSH, &bd->bd_gl->gl_flags); - set_bit(GLF_DIRTY, &bd->bd_gl->gl_flags); - if (gfs2_is_jdata(ip)) { - gfs2_pin(sdp, bd->bd_bh); - tr->tr_num_databuf_new++; - sdp->sd_log_num_databuf++; - list_add_tail(&le->le_list, &sdp->sd_log_le_databuf); - } else { - list_add_tail(&le->le_list, &sdp->sd_log_le_ordered); - } -out: - gfs2_log_unlock(sdp); - unlock_buffer(bd->bd_bh); -} - -static void gfs2_check_magic(struct buffer_head *bh) -{ - void *kaddr; - __be32 *ptr; - - clear_buffer_escaped(bh); - kaddr = kmap_atomic(bh->b_page, KM_USER0); - ptr = kaddr + bh_offset(bh); - if (*ptr == cpu_to_be32(GFS2_MAGIC)) - set_buffer_escaped(bh); - kunmap_atomic(kaddr, KM_USER0); -} - -static void gfs2_write_blocks(struct gfs2_sbd *sdp, struct buffer_head *bh, - struct list_head *list, struct list_head *done, - unsigned int n) -{ - struct buffer_head *bh1; - struct gfs2_log_descriptor *ld; - struct gfs2_bufdata *bd; - __be64 *ptr; - - if (!bh) - return; - - ld = bh_log_desc(bh); - ld->ld_length = cpu_to_be32(n + 1); - ld->ld_data1 = cpu_to_be32(n); - - ptr = bh_log_ptr(bh); - - get_bh(bh); - submit_bh(WRITE_SYNC_PLUG, bh); - gfs2_log_lock(sdp); - while(!list_empty(list)) { - bd = list_entry(list->next, struct gfs2_bufdata, bd_le.le_list); - list_move_tail(&bd->bd_le.le_list, done); - get_bh(bd->bd_bh); - while (be64_to_cpu(*ptr) != bd->bd_bh->b_blocknr) { - gfs2_log_incr_head(sdp); - ptr += 2; - } - gfs2_log_unlock(sdp); - lock_buffer(bd->bd_bh); - if (buffer_escaped(bd->bd_bh)) { - void *kaddr; - bh1 = gfs2_log_get_buf(sdp); - kaddr = kmap_atomic(bd->bd_bh->b_page, KM_USER0); - memcpy(bh1->b_data, kaddr + bh_offset(bd->bd_bh), - bh1->b_size); - kunmap_atomic(kaddr, KM_USER0); - *(__be32 *)bh1->b_data = 0; - clear_buffer_escaped(bd->bd_bh); - unlock_buffer(bd->bd_bh); - brelse(bd->bd_bh); - } else { - bh1 = gfs2_log_fake_buf(sdp, bd->bd_bh); - } - submit_bh(WRITE_SYNC_PLUG, bh1); - gfs2_log_lock(sdp); - ptr += 2; - } - gfs2_log_unlock(sdp); - brelse(bh); + gfs2_revoke_clean(jd); } /** @@ -610,39 +754,14 @@ static void gfs2_write_blocks(struct gfs2_sbd *sdp, struct buffer_head *bh, * */ -static void databuf_lo_before_commit(struct gfs2_sbd *sdp) +static void databuf_lo_before_commit(struct gfs2_sbd *sdp, struct gfs2_trans *tr) { - struct gfs2_bufdata *bd = NULL; - struct buffer_head *bh = NULL; - unsigned int n = 0; - __be64 *ptr = NULL, *end = NULL; - LIST_HEAD(processed); - LIST_HEAD(in_progress); - - gfs2_log_lock(sdp); - while (!list_empty(&sdp->sd_log_le_databuf)) { - if (ptr == end) { - gfs2_log_unlock(sdp); - gfs2_write_blocks(sdp, bh, &in_progress, &processed, n); - n = 0; - bh = gfs2_get_log_desc(sdp, GFS2_LOG_DESC_JDATA); - ptr = bh_log_ptr(bh); - end = bh_ptr_end(bh) - 1; - gfs2_log_lock(sdp); - continue; - } - bd = list_entry(sdp->sd_log_le_databuf.next, struct gfs2_bufdata, bd_le.le_list); - list_move_tail(&bd->bd_le.le_list, &in_progress); - gfs2_check_magic(bd->bd_bh); - *ptr++ = cpu_to_be64(bd->bd_bh->b_blocknr); - *ptr++ = cpu_to_be64(buffer_escaped(bh) ? 1 : 0); - n++; - } - gfs2_log_unlock(sdp); - gfs2_write_blocks(sdp, bh, &in_progress, &processed, n); - gfs2_log_lock(sdp); - list_splice(&processed, &sdp->sd_log_le_databuf); - gfs2_log_unlock(sdp); + unsigned int limit = databuf_limit(sdp); + unsigned int nbuf; + if (tr == NULL) + return; + nbuf = tr->tr_num_databuf_new - tr->tr_num_databuf_rm; + gfs2_before_commit(sdp, limit, nbuf, &tr->tr_databuf, 1); } static int databuf_lo_scan_elements(struct gfs2_jdesc *jd, unsigned int start, @@ -666,9 +785,9 @@ static int databuf_lo_scan_elements(struct gfs2_jdesc *jd, unsigned int start, blkno = be64_to_cpu(*ptr++); esc = be64_to_cpu(*ptr++); - sdp->sd_found_blocks++; + jd->jd_found_blocks++; - if (gfs2_revoke_check(sdp, blkno, start)) + if (gfs2_revoke_check(jd, blkno, start)) continue; error = gfs2_replay_read_block(jd, start, &bh_log); @@ -687,10 +806,8 @@ static int databuf_lo_scan_elements(struct gfs2_jdesc *jd, unsigned int start, brelse(bh_log); brelse(bh_ip); - if (error) - break; - sdp->sd_replayed_blocks++; + jd->jd_replayed_blocks++; } return error; @@ -714,26 +831,27 @@ static void databuf_lo_after_scan(struct gfs2_jdesc *jd, int error, int pass) gfs2_meta_sync(ip->i_gl); fs_info(sdp, "jid=%u: Replayed %u of %u data blocks\n", - jd->jd_jid, sdp->sd_replayed_blocks, sdp->sd_found_blocks); + jd->jd_jid, jd->jd_replayed_blocks, jd->jd_found_blocks); } -static void databuf_lo_after_commit(struct gfs2_sbd *sdp, struct gfs2_ail *ai) +static void databuf_lo_after_commit(struct gfs2_sbd *sdp, struct gfs2_trans *tr) { - struct list_head *head = &sdp->sd_log_le_databuf; + struct list_head *head; struct gfs2_bufdata *bd; + if (tr == NULL) + return; + + head = &tr->tr_databuf; while (!list_empty(head)) { - bd = list_entry(head->next, struct gfs2_bufdata, bd_le.le_list); - list_del_init(&bd->bd_le.le_list); - sdp->sd_log_num_databuf--; - gfs2_unpin(sdp, bd->bd_bh, ai); + bd = list_entry(head->next, struct gfs2_bufdata, bd_list); + list_del_init(&bd->bd_list); + gfs2_unpin(sdp, bd->bd_bh, tr); } - gfs2_assert_warn(sdp, !sdp->sd_log_num_databuf); } const struct gfs2_log_operations gfs2_buf_lops = { - .lo_add = buf_lo_add, .lo_before_commit = buf_lo_before_commit, .lo_after_commit = buf_lo_after_commit, .lo_before_scan = buf_lo_before_scan, @@ -743,22 +861,15 @@ const struct gfs2_log_operations gfs2_buf_lops = { }; const struct gfs2_log_operations gfs2_revoke_lops = { - .lo_add = revoke_lo_add, .lo_before_commit = revoke_lo_before_commit, + .lo_after_commit = revoke_lo_after_commit, .lo_before_scan = revoke_lo_before_scan, .lo_scan_elements = revoke_lo_scan_elements, .lo_after_scan = revoke_lo_after_scan, .lo_name = "revoke", }; -const struct gfs2_log_operations gfs2_rg_lops = { - .lo_add = rg_lo_add, - .lo_after_commit = rg_lo_after_commit, - .lo_name = "rg", -}; - const struct gfs2_log_operations gfs2_databuf_lops = { - .lo_add = databuf_lo_add, .lo_before_commit = databuf_lo_before_commit, .lo_after_commit = databuf_lo_after_commit, .lo_scan_elements = databuf_lo_scan_elements, @@ -769,7 +880,6 @@ const struct gfs2_log_operations gfs2_databuf_lops = { const struct gfs2_log_operations *gfs2_log_ops[] = { &gfs2_databuf_lops, &gfs2_buf_lops, - &gfs2_rg_lops, &gfs2_revoke_lops, NULL, }; diff --git a/fs/gfs2/lops.h b/fs/gfs2/lops.h index 3c0b2737658..a65a7ba32ff 100644 --- a/fs/gfs2/lops.h +++ b/fs/gfs2/lops.h @@ -23,10 +23,12 @@ extern const struct gfs2_log_operations gfs2_glock_lops; extern const struct gfs2_log_operations gfs2_buf_lops; extern const struct gfs2_log_operations gfs2_revoke_lops; -extern const struct gfs2_log_operations gfs2_rg_lops; extern const struct gfs2_log_operations gfs2_databuf_lops; extern const struct gfs2_log_operations *gfs2_log_ops[]; +extern void gfs2_log_write_page(struct gfs2_sbd *sdp, struct page *page); +extern void gfs2_log_flush_bio(struct gfs2_sbd *sdp, int rw); +extern void gfs2_pin(struct gfs2_sbd *sdp, struct buffer_head *bh); static inline unsigned int buf_limit(struct gfs2_sbd *sdp) { @@ -44,33 +46,22 @@ static inline unsigned int databuf_limit(struct gfs2_sbd *sdp) return limit; } -static inline void lops_init_le(struct gfs2_log_element *le, - const struct gfs2_log_operations *lops) -{ - INIT_LIST_HEAD(&le->le_list); - le->le_ops = lops; -} - -static inline void lops_add(struct gfs2_sbd *sdp, struct gfs2_log_element *le) -{ - if (le->le_ops->lo_add) - le->le_ops->lo_add(sdp, le); -} - -static inline void lops_before_commit(struct gfs2_sbd *sdp) +static inline void lops_before_commit(struct gfs2_sbd *sdp, + struct gfs2_trans *tr) { int x; for (x = 0; gfs2_log_ops[x]; x++) if (gfs2_log_ops[x]->lo_before_commit) - gfs2_log_ops[x]->lo_before_commit(sdp); + gfs2_log_ops[x]->lo_before_commit(sdp, tr); } -static inline void lops_after_commit(struct gfs2_sbd *sdp, struct gfs2_ail *ai) +static inline void lops_after_commit(struct gfs2_sbd *sdp, + struct gfs2_trans *tr) { int x; for (x = 0; gfs2_log_ops[x]; x++) if (gfs2_log_ops[x]->lo_after_commit) - gfs2_log_ops[x]->lo_after_commit(sdp, ai); + gfs2_log_ops[x]->lo_after_commit(sdp, tr); } static inline void lops_before_scan(struct gfs2_jdesc *jd, diff --git a/fs/gfs2/main.c b/fs/gfs2/main.c index 72c31a315d9..82b6ac82965 100644 --- a/fs/gfs2/main.c +++ b/fs/gfs2/main.c @@ -7,6 +7,8 @@ * of the GNU General Public License version 2. */ +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + #include <linux/slab.h> #include <linux/spinlock.h> #include <linux/completion.h> @@ -14,7 +16,10 @@ #include <linux/module.h> #include <linux/init.h> #include <linux/gfs2_ondisk.h> -#include <asm/atomic.h> +#include <linux/rcupdate.h> +#include <linux/rculist_bl.h> +#include <linux/atomic.h> +#include <linux/mempool.h> #include "gfs2.h" #include "incore.h" @@ -26,10 +31,7 @@ #include "recovery.h" #include "dir.h" -static struct shrinker qd_shrinker = { - .shrink = gfs2_shrink_qd_memory, - .seeks = DEFAULT_SEEKS, -}; +struct workqueue_struct *gfs2_control_wq; static void gfs2_init_inode_once(void *foo) { @@ -38,19 +40,21 @@ static void gfs2_init_inode_once(void *foo) inode_init_once(&ip->i_inode); init_rwsem(&ip->i_rw_mutex); INIT_LIST_HEAD(&ip->i_trunc_list); - ip->i_alloc = NULL; + ip->i_res = NULL; + ip->i_hash_cache = NULL; } static void gfs2_init_glock_once(void *foo) { struct gfs2_glock *gl = foo; - INIT_HLIST_NODE(&gl->gl_list); + INIT_HLIST_BL_NODE(&gl->gl_list); spin_lock_init(&gl->gl_spin); INIT_LIST_HEAD(&gl->gl_holders); INIT_LIST_HEAD(&gl->gl_lru); INIT_LIST_HEAD(&gl->gl_ail_list); atomic_set(&gl->gl_ail_count, 0); + atomic_set(&gl->gl_revokes, 0); } static void gfs2_init_gl_aspace_once(void *foo) @@ -74,11 +78,16 @@ static int __init init_gfs2_fs(void) gfs2_str2qstr(&gfs2_qdot, "."); gfs2_str2qstr(&gfs2_qdotdot, ".."); + gfs2_quota_hash_init(); error = gfs2_sys_init(); if (error) return error; + error = list_lru_init(&gfs2_qd_lru); + if (error) + goto fail_lru; + error = gfs2_glock_init(); if (error) goto fail; @@ -125,7 +134,13 @@ static int __init init_gfs2_fs(void) if (!gfs2_quotad_cachep) goto fail; - register_shrinker(&qd_shrinker); + gfs2_rsrv_cachep = kmem_cache_create("gfs2_mblk", + sizeof(struct gfs2_blkreserv), + 0, 0, NULL); + if (!gfs2_rsrv_cachep) + goto fail; + + register_shrinker(&gfs2_qd_shrinker); error = register_filesystem(&gfs2_fs_type); if (error) @@ -141,20 +156,38 @@ static int __init init_gfs2_fs(void) if (!gfs_recovery_wq) goto fail_wq; + gfs2_control_wq = alloc_workqueue("gfs2_control", + WQ_UNBOUND | WQ_FREEZABLE, 0); + if (!gfs2_control_wq) + goto fail_recovery; + + gfs2_page_pool = mempool_create_page_pool(64, 0); + if (!gfs2_page_pool) + goto fail_control; + gfs2_register_debugfs(); - printk("GFS2 (built %s %s) installed\n", __DATE__, __TIME__); + pr_info("GFS2 installed\n"); return 0; +fail_control: + destroy_workqueue(gfs2_control_wq); +fail_recovery: + destroy_workqueue(gfs_recovery_wq); fail_wq: unregister_filesystem(&gfs2meta_fs_type); fail_unregister: unregister_filesystem(&gfs2_fs_type); fail: - unregister_shrinker(&qd_shrinker); + list_lru_destroy(&gfs2_qd_lru); +fail_lru: + unregister_shrinker(&gfs2_qd_shrinker); gfs2_glock_exit(); + if (gfs2_rsrv_cachep) + kmem_cache_destroy(gfs2_rsrv_cachep); + if (gfs2_quotad_cachep) kmem_cache_destroy(gfs2_quotad_cachep); @@ -184,13 +217,19 @@ fail: static void __exit exit_gfs2_fs(void) { - unregister_shrinker(&qd_shrinker); + unregister_shrinker(&gfs2_qd_shrinker); gfs2_glock_exit(); gfs2_unregister_debugfs(); unregister_filesystem(&gfs2_fs_type); unregister_filesystem(&gfs2meta_fs_type); destroy_workqueue(gfs_recovery_wq); + destroy_workqueue(gfs2_control_wq); + list_lru_destroy(&gfs2_qd_lru); + + rcu_barrier(); + mempool_destroy(gfs2_page_pool); + kmem_cache_destroy(gfs2_rsrv_cachep); kmem_cache_destroy(gfs2_quotad_cachep); kmem_cache_destroy(gfs2_rgrpd_cachep); kmem_cache_destroy(gfs2_bufdata_cachep); diff --git a/fs/gfs2/meta_io.c b/fs/gfs2/meta_io.c index 939739c7b3f..b984a6e190b 100644 --- a/fs/gfs2/meta_io.c +++ b/fs/gfs2/meta_io.c @@ -31,13 +31,14 @@ #include "rgrp.h" #include "trans.h" #include "util.h" +#include "trace_gfs2.h" static int gfs2_aspace_writepage(struct page *page, struct writeback_control *wbc) { struct buffer_head *bh, *head; int nr_underway = 0; - int write_op = REQ_META | - (wbc->sync_mode == WB_SYNC_ALL ? WRITE_SYNC_PLUG : WRITE); + int write_op = REQ_META | REQ_PRIO | + (wbc->sync_mode == WB_SYNC_ALL ? WRITE_SYNC : WRITE); BUG_ON(!PageLocked(page)); BUG_ON(!page_has_buffers(page)); @@ -51,7 +52,7 @@ static int gfs2_aspace_writepage(struct page *page, struct writeback_control *wb /* * If it's a fully non-blocking write attempt and we cannot * lock the buffer then redirty the page. Note that this can - * potentially cause a busy-wait loop from pdflush and kswapd + * potentially cause a busy-wait loop from flusher thread and kswapd * activity, but those code paths have their own higher-level * throttling. */ @@ -94,26 +95,12 @@ static int gfs2_aspace_writepage(struct page *page, struct writeback_control *wb const struct address_space_operations gfs2_meta_aops = { .writepage = gfs2_aspace_writepage, .releasepage = gfs2_releasepage, - .sync_page = block_sync_page, }; -/** - * gfs2_meta_sync - Sync all buffers associated with a glock - * @gl: The glock - * - */ - -void gfs2_meta_sync(struct gfs2_glock *gl) -{ - struct address_space *mapping = gfs2_glock2aspace(gl); - int error; - - filemap_fdatawrite(mapping); - error = filemap_fdatawait(mapping); - - if (error) - gfs2_io_error(gl->gl_sbd); -} +const struct address_space_operations gfs2_rgrp_aops = { + .writepage = gfs2_aspace_writepage, + .releasepage = gfs2_releasepage, +}; /** * gfs2_getbuf - Get a buffer with a given address space @@ -134,6 +121,9 @@ struct buffer_head *gfs2_getbuf(struct gfs2_glock *gl, u64 blkno, int create) unsigned long index; unsigned int bufnum; + if (mapping == NULL) + mapping = &sdp->sd_aspace; + shift = PAGE_CACHE_SHIFT - sdp->sd_sb.sb_bsize_shift; index = blkno >> shift; /* convert block to page */ bufnum = blkno - (index << shift); /* block buf index within page */ @@ -146,7 +136,8 @@ struct buffer_head *gfs2_getbuf(struct gfs2_glock *gl, u64 blkno, int create) yield(); } } else { - page = find_lock_page(mapping, index); + page = find_get_page_flags(mapping, index, + FGP_LOCK|FGP_ACCESSED); if (!page) return NULL; } @@ -163,7 +154,6 @@ struct buffer_head *gfs2_getbuf(struct gfs2_glock *gl, u64 blkno, int create) map_bh(bh, sdp->sd_vfs, blkno); unlock_page(page); - mark_page_accessed(page); page_cache_release(page); return bh; @@ -213,8 +203,10 @@ int gfs2_meta_read(struct gfs2_glock *gl, u64 blkno, int flags, struct gfs2_sbd *sdp = gl->gl_sbd; struct buffer_head *bh; - if (unlikely(test_bit(SDF_SHUTDOWN, &sdp->sd_flags))) + if (unlikely(test_bit(SDF_SHUTDOWN, &sdp->sd_flags))) { + *bhp = NULL; return -EIO; + } *bhp = bh = gfs2_getbuf(gl, blkno, CREATE); @@ -225,7 +217,7 @@ int gfs2_meta_read(struct gfs2_glock *gl, u64 blkno, int flags, } bh->b_end_io = end_buffer_read_sync; get_bh(bh); - submit_bh(READ_SYNC | REQ_META, bh); + submit_bh(READ_SYNC | REQ_META | REQ_PRIO, bh); if (!(flags & DIO_WAIT)) return 0; @@ -235,6 +227,7 @@ int gfs2_meta_read(struct gfs2_glock *gl, u64 blkno, int flags, if (tr && tr->tr_touched) gfs2_io_error_bh(sdp, bh); brelse(bh); + *bhp = NULL; return -EIO; } @@ -268,71 +261,34 @@ int gfs2_meta_wait(struct gfs2_sbd *sdp, struct buffer_head *bh) return 0; } -/** - * gfs2_attach_bufdata - attach a struct gfs2_bufdata structure to a buffer - * @gl: the glock the buffer belongs to - * @bh: The buffer to be attached to - * @meta: Flag to indicate whether its metadata or not - */ - -void gfs2_attach_bufdata(struct gfs2_glock *gl, struct buffer_head *bh, - int meta) -{ - struct gfs2_bufdata *bd; - - if (meta) - lock_page(bh->b_page); - - if (bh->b_private) { - if (meta) - unlock_page(bh->b_page); - return; - } - - bd = kmem_cache_zalloc(gfs2_bufdata_cachep, GFP_NOFS | __GFP_NOFAIL); - bd->bd_bh = bh; - bd->bd_gl = gl; - - INIT_LIST_HEAD(&bd->bd_list_tr); - if (meta) - lops_init_le(&bd->bd_le, &gfs2_buf_lops); - else - lops_init_le(&bd->bd_le, &gfs2_databuf_lops); - bh->b_private = bd; - - if (meta) - unlock_page(bh->b_page); -} - void gfs2_remove_from_journal(struct buffer_head *bh, struct gfs2_trans *tr, int meta) { struct address_space *mapping = bh->b_page->mapping; struct gfs2_sbd *sdp = gfs2_mapping2sbd(mapping); struct gfs2_bufdata *bd = bh->b_private; + int was_pinned = 0; if (test_clear_buffer_pinned(bh)) { + trace_gfs2_pin(bd, 0); atomic_dec(&sdp->sd_log_pinned); - list_del_init(&bd->bd_le.le_list); - if (meta) { - gfs2_assert_warn(sdp, sdp->sd_log_num_buf); - sdp->sd_log_num_buf--; + list_del_init(&bd->bd_list); + if (meta) tr->tr_num_buf_rm++; - } else { - gfs2_assert_warn(sdp, sdp->sd_log_num_databuf); - sdp->sd_log_num_databuf--; + else tr->tr_num_databuf_rm++; - } tr->tr_touched = 1; + was_pinned = 1; brelse(bh); } if (bd) { - if (bd->bd_ail) { - gfs2_remove_from_ail(bd); - bh->b_private = NULL; - bd->bd_bh = NULL; - bd->bd_blkno = bh->b_blocknr; + spin_lock(&sdp->sd_ail_lock); + if (bd->bd_tr) { gfs2_trans_add_revoke(sdp, bd); + } else if (was_pinned) { + bh->b_private = NULL; + kmem_cache_free(gfs2_bufdata_cachep, bd); } + spin_unlock(&sdp->sd_ail_lock); } clear_buffer_dirty(bh); clear_buffer_uptodate(bh); @@ -372,33 +328,24 @@ void gfs2_meta_wipe(struct gfs2_inode *ip, u64 bstart, u32 blen) * @ip: The GFS2 inode * @height: The level of this buf in the metadata (indir addr) tree (if any) * @num: The block number (device relative) of the buffer - * @new: Non-zero if we may create a new buffer * @bhp: the buffer is returned here * * Returns: errno */ int gfs2_meta_indirect_buffer(struct gfs2_inode *ip, int height, u64 num, - int new, struct buffer_head **bhp) + struct buffer_head **bhp) { struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); struct gfs2_glock *gl = ip->i_gl; struct buffer_head *bh; int ret = 0; + u32 mtype = height ? GFS2_METATYPE_IN : GFS2_METATYPE_DI; - if (new) { - BUG_ON(height == 0); - bh = gfs2_meta_new(gl, num); - gfs2_trans_add_bh(ip->i_gl, bh, 1); - gfs2_metatype_set(bh, GFS2_METATYPE_IN, GFS2_FORMAT_IN); - gfs2_buffer_clear_tail(bh, sizeof(struct gfs2_meta_header)); - } else { - u32 mtype = height ? GFS2_METATYPE_IN : GFS2_METATYPE_DI; - ret = gfs2_meta_read(gl, num, DIO_WAIT, &bh); - if (ret == 0 && gfs2_metatype_check(sdp, bh, mtype)) { - brelse(bh); - ret = -EIO; - } + ret = gfs2_meta_read(gl, num, DIO_WAIT, &bh); + if (ret == 0 && gfs2_metatype_check(sdp, bh, mtype)) { + brelse(bh); + ret = -EIO; } *bhp = bh; return ret; @@ -441,7 +388,7 @@ struct buffer_head *gfs2_meta_ra(struct gfs2_glock *gl, u64 dblock, u32 extlen) bh = gfs2_getbuf(gl, dblock, CREATE); if (!buffer_uptodate(bh) && !buffer_locked(bh)) - ll_rw_block(READA, 1, &bh); + ll_rw_block(READA | REQ_META, 1, &bh); brelse(bh); dblock++; extlen--; diff --git a/fs/gfs2/meta_io.h b/fs/gfs2/meta_io.h index 6a1d9ba1641..ac5d8027d33 100644 --- a/fs/gfs2/meta_io.h +++ b/fs/gfs2/meta_io.h @@ -38,47 +38,41 @@ static inline void gfs2_buffer_copy_tail(struct buffer_head *to_bh, } extern const struct address_space_operations gfs2_meta_aops; +extern const struct address_space_operations gfs2_rgrp_aops; static inline struct gfs2_sbd *gfs2_mapping2sbd(struct address_space *mapping) { struct inode *inode = mapping->host; if (mapping->a_ops == &gfs2_meta_aops) return (((struct gfs2_glock *)mapping) - 1)->gl_sbd; + else if (mapping->a_ops == &gfs2_rgrp_aops) + return container_of(mapping, struct gfs2_sbd, sd_aspace); else return inode->i_sb->s_fs_info; } -void gfs2_meta_sync(struct gfs2_glock *gl); - -struct buffer_head *gfs2_meta_new(struct gfs2_glock *gl, u64 blkno); -int gfs2_meta_read(struct gfs2_glock *gl, u64 blkno, - int flags, struct buffer_head **bhp); -int gfs2_meta_wait(struct gfs2_sbd *sdp, struct buffer_head *bh); -struct buffer_head *gfs2_getbuf(struct gfs2_glock *gl, u64 blkno, int create); - -void gfs2_attach_bufdata(struct gfs2_glock *gl, struct buffer_head *bh, - int meta); - -void gfs2_remove_from_journal(struct buffer_head *bh, struct gfs2_trans *tr, - int meta); - -void gfs2_meta_wipe(struct gfs2_inode *ip, u64 bstart, u32 blen); - -int gfs2_meta_indirect_buffer(struct gfs2_inode *ip, int height, u64 num, - int new, struct buffer_head **bhp); +extern struct buffer_head *gfs2_meta_new(struct gfs2_glock *gl, u64 blkno); +extern int gfs2_meta_read(struct gfs2_glock *gl, u64 blkno, int flags, + struct buffer_head **bhp); +extern int gfs2_meta_wait(struct gfs2_sbd *sdp, struct buffer_head *bh); +extern struct buffer_head *gfs2_getbuf(struct gfs2_glock *gl, u64 blkno, + int create); +extern void gfs2_remove_from_journal(struct buffer_head *bh, + struct gfs2_trans *tr, int meta); +extern void gfs2_meta_wipe(struct gfs2_inode *ip, u64 bstart, u32 blen); +extern int gfs2_meta_indirect_buffer(struct gfs2_inode *ip, int height, u64 num, + struct buffer_head **bhp); static inline int gfs2_meta_inode_buffer(struct gfs2_inode *ip, struct buffer_head **bhp) { - return gfs2_meta_indirect_buffer(ip, 0, ip->i_no_addr, 0, bhp); + return gfs2_meta_indirect_buffer(ip, 0, ip->i_no_addr, bhp); } struct buffer_head *gfs2_meta_ra(struct gfs2_glock *gl, u64 dblock, u32 extlen); #define buffer_busy(bh) \ ((bh)->b_state & ((1ul << BH_Dirty) | (1ul << BH_Lock) | (1ul << BH_Pinned))) -#define buffer_in_io(bh) \ -((bh)->b_state & ((1ul << BH_Dirty) | (1ul << BH_Lock))) #endif /* __DIO_DOT_H__ */ diff --git a/fs/gfs2/ops_fstype.c b/fs/gfs2/ops_fstype.c index 777927ce6f7..bc564c0d6d1 100644 --- a/fs/gfs2/ops_fstype.c +++ b/fs/gfs2/ops_fstype.c @@ -7,6 +7,8 @@ * of the GNU General Public License version 2. */ +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + #include <linux/sched.h> #include <linux/slab.h> #include <linux/spinlock.h> @@ -14,10 +16,13 @@ #include <linux/buffer_head.h> #include <linux/blkdev.h> #include <linux/kthread.h> +#include <linux/export.h> #include <linux/namei.h> #include <linux/mount.h> #include <linux/gfs2_ondisk.h> #include <linux/quotaops.h> +#include <linux/lockdep.h> +#include <linux/module.h> #include "gfs2.h" #include "incore.h" @@ -33,6 +38,7 @@ #include "log.h" #include "quota.h" #include "dir.h" +#include "meta_io.h" #include "trace_gfs2.h" #define DO 0 @@ -48,7 +54,6 @@ static void gfs2_tune_init(struct gfs2_tune *gt) { spin_lock_init(>->gt_spin); - gt->gt_quota_simul_sync = 64; gt->gt_quota_warn_period = 10; gt->gt_quota_scale_num = 1; gt->gt_quota_scale_den = 1; @@ -60,6 +65,7 @@ static void gfs2_tune_init(struct gfs2_tune *gt) static struct gfs2_sbd *init_sbd(struct super_block *sb) { struct gfs2_sbd *sdp; + struct address_space *mapping; sdp = kzalloc(sizeof(struct gfs2_sbd), GFP_KERNEL); if (!sdp) @@ -67,48 +73,67 @@ static struct gfs2_sbd *init_sbd(struct super_block *sb) sb->s_fs_info = sdp; sdp->sd_vfs = sb; + sdp->sd_lkstats = alloc_percpu(struct gfs2_pcpu_lkstats); + if (!sdp->sd_lkstats) { + kfree(sdp); + return NULL; + } + set_bit(SDF_NOJOURNALID, &sdp->sd_flags); gfs2_tune_init(&sdp->sd_tune); init_waitqueue_head(&sdp->sd_glock_wait); atomic_set(&sdp->sd_glock_disposal, 0); + init_completion(&sdp->sd_locking_init); + init_completion(&sdp->sd_wdack); spin_lock_init(&sdp->sd_statfs_spin); spin_lock_init(&sdp->sd_rindex_spin); - mutex_init(&sdp->sd_rindex_mutex); - INIT_LIST_HEAD(&sdp->sd_rindex_list); - INIT_LIST_HEAD(&sdp->sd_rindex_mru_list); + sdp->sd_rindex_tree.rb_node = NULL; INIT_LIST_HEAD(&sdp->sd_jindex_list); spin_lock_init(&sdp->sd_jindex_spin); mutex_init(&sdp->sd_jindex_mutex); + init_completion(&sdp->sd_journal_ready); INIT_LIST_HEAD(&sdp->sd_quota_list); mutex_init(&sdp->sd_quota_mutex); + mutex_init(&sdp->sd_quota_sync_mutex); init_waitqueue_head(&sdp->sd_quota_wait); INIT_LIST_HEAD(&sdp->sd_trunc_list); spin_lock_init(&sdp->sd_trunc_lock); + spin_lock_init(&sdp->sd_bitmap_lock); + + mapping = &sdp->sd_aspace; + + address_space_init_once(mapping); + mapping->a_ops = &gfs2_rgrp_aops; + mapping->host = sb->s_bdev->bd_inode; + mapping->flags = 0; + mapping_set_gfp_mask(mapping, GFP_NOFS); + mapping->private_data = NULL; + mapping->backing_dev_info = sb->s_bdi; + mapping->writeback_index = 0; spin_lock_init(&sdp->sd_log_lock); atomic_set(&sdp->sd_log_pinned, 0); - INIT_LIST_HEAD(&sdp->sd_log_le_buf); INIT_LIST_HEAD(&sdp->sd_log_le_revoke); - INIT_LIST_HEAD(&sdp->sd_log_le_rg); - INIT_LIST_HEAD(&sdp->sd_log_le_databuf); INIT_LIST_HEAD(&sdp->sd_log_le_ordered); + spin_lock_init(&sdp->sd_ordered_lock); init_waitqueue_head(&sdp->sd_log_waitq); init_waitqueue_head(&sdp->sd_logd_waitq); + spin_lock_init(&sdp->sd_ail_lock); INIT_LIST_HEAD(&sdp->sd_ail1_list); INIT_LIST_HEAD(&sdp->sd_ail2_list); init_rwsem(&sdp->sd_log_flush_lock); atomic_set(&sdp->sd_log_in_flight, 0); init_waitqueue_head(&sdp->sd_log_flush_wait); - - INIT_LIST_HEAD(&sdp->sd_revoke_list); - - mutex_init(&sdp->sd_freeze_lock); + init_waitqueue_head(&sdp->sd_log_frozen_wait); + atomic_set(&sdp->sd_log_freeze, 0); + atomic_set(&sdp->sd_frozen_root, 0); + init_waitqueue_head(&sdp->sd_frozen_root_wait); return sdp; } @@ -125,12 +150,14 @@ static struct gfs2_sbd *init_sbd(struct super_block *sb) * changed. */ -static int gfs2_check_sb(struct gfs2_sbd *sdp, struct gfs2_sb_host *sb, int silent) +static int gfs2_check_sb(struct gfs2_sbd *sdp, int silent) { + struct gfs2_sb_host *sb = &sdp->sd_sb; + if (sb->sb_magic != GFS2_MAGIC || sb->sb_type != GFS2_METATYPE_SB) { if (!silent) - printk(KERN_WARNING "GFS2: not a GFS2 filesystem\n"); + pr_warn("not a GFS2 filesystem\n"); return -EINVAL; } @@ -152,12 +179,14 @@ static void end_bio_io_page(struct bio *bio, int error) if (!error) SetPageUptodate(page); else - printk(KERN_WARNING "gfs2: error %d reading superblock\n", error); + pr_warn("error %d reading superblock\n", error); unlock_page(page); } -static void gfs2_sb_in(struct gfs2_sb_host *sb, const void *buf) +static void gfs2_sb_in(struct gfs2_sbd *sdp, const void *buf) { + struct gfs2_sb_host *sb = &sdp->sd_sb; + struct super_block *s = sdp->sd_vfs; const struct gfs2_sb *str = buf; sb->sb_magic = be32_to_cpu(str->sb_header.mh_magic); @@ -174,7 +203,7 @@ static void gfs2_sb_in(struct gfs2_sb_host *sb, const void *buf) memcpy(sb->sb_lockproto, str->sb_lockproto, GFS2_LOCKNAME_LEN); memcpy(sb->sb_locktable, str->sb_locktable, GFS2_LOCKNAME_LEN); - memcpy(sb->sb_uuid, str->sb_uuid, 16); + memcpy(s->s_uuid, str->sb_uuid, 16); } /** @@ -196,7 +225,7 @@ static void gfs2_sb_in(struct gfs2_sb_host *sb, const void *buf) * Returns: 0 on success or error */ -static int gfs2_read_super(struct gfs2_sbd *sdp, sector_t sector) +static int gfs2_read_super(struct gfs2_sbd *sdp, sector_t sector, int silent) { struct super_block *sb = sdp->sd_vfs; struct gfs2_sb *p; @@ -205,14 +234,14 @@ static int gfs2_read_super(struct gfs2_sbd *sdp, sector_t sector) page = alloc_page(GFP_NOFS); if (unlikely(!page)) - return -ENOBUFS; + return -ENOMEM; ClearPageUptodate(page); ClearPageDirty(page); lock_page(page); bio = bio_alloc(GFP_NOFS, 1); - bio->bi_sector = sector * (sb->s_blocksize >> 9); + bio->bi_iter.bi_sector = sector * (sb->s_blocksize >> 9); bio->bi_bdev = sb->s_bdev; bio_add_page(bio, page, PAGE_SIZE, 0); @@ -226,10 +255,10 @@ static int gfs2_read_super(struct gfs2_sbd *sdp, sector_t sector) return -EIO; } p = kmap(page); - gfs2_sb_in(&sdp->sd_sb, p); + gfs2_sb_in(sdp, p); kunmap(page); __free_page(page); - return 0; + return gfs2_check_sb(sdp, silent); } /** @@ -246,17 +275,13 @@ static int gfs2_read_sb(struct gfs2_sbd *sdp, int silent) unsigned int x; int error; - error = gfs2_read_super(sdp, GFS2_SB_ADDR >> sdp->sd_fsb2bb_shift); + error = gfs2_read_super(sdp, GFS2_SB_ADDR >> sdp->sd_fsb2bb_shift, silent); if (error) { if (!silent) fs_err(sdp, "can't read superblock\n"); return error; } - error = gfs2_check_sb(sdp, &sdp->sd_sb, silent); - if (error) - return error; - sdp->sd_fsb2bb_shift = sdp->sd_sb.sb_bsize_shift - GFS2_BASIC_BLOCK_SHIFT; sdp->sd_fsb2bb = 1 << sdp->sd_fsb2bb_shift; @@ -271,6 +296,9 @@ static int gfs2_read_sb(struct gfs2_sbd *sdp, int silent) sdp->sd_qc_per_block = (sdp->sd_sb.sb_bsize - sizeof(struct gfs2_meta_header)) / sizeof(struct gfs2_quota_change); + sdp->sd_blocks_per_bitmap = (sdp->sd_sb.sb_bsize - + sizeof(struct gfs2_meta_header)) + * GFS2_NBBY; /* not the rgrp bitmap, subsequent bitmaps only */ /* Compute maximum reservation required to add a entry to a directory */ @@ -339,14 +367,10 @@ static int init_names(struct gfs2_sbd *sdp, int silent) /* Try to autodetect */ if (!proto[0] || !table[0]) { - error = gfs2_read_super(sdp, GFS2_SB_ADDR >> sdp->sd_fsb2bb_shift); + error = gfs2_read_super(sdp, GFS2_SB_ADDR >> sdp->sd_fsb2bb_shift, silent); if (error) return error; - error = gfs2_check_sb(sdp, &sdp->sd_sb, silent); - if (error) - goto out; - if (!proto[0]) proto = sdp->sd_sb.sb_lockproto; if (!table[0]) @@ -363,7 +387,6 @@ static int init_names(struct gfs2_sbd *sdp, int silent) while ((table = strchr(table, '/'))) *table = '_'; -out: return error; } @@ -401,8 +424,8 @@ static int init_locking(struct gfs2_sbd *sdp, struct gfs2_holder *mount_gh, goto fail_live; } - error = gfs2_glock_get(sdp, GFS2_TRANS_LOCK, &gfs2_trans_glops, - CREATE, &sdp->sd_trans_gl); + error = gfs2_glock_get(sdp, GFS2_FREEZE_LOCK, &gfs2_freeze_glops, + CREATE, &sdp->sd_freeze_gl); if (error) { fs_err(sdp, "can't create transaction glock: %d\n", error); goto fail_rename; @@ -411,7 +434,7 @@ static int init_locking(struct gfs2_sbd *sdp, struct gfs2_holder *mount_gh, return 0; fail_trans: - gfs2_glock_put(sdp->sd_trans_gl); + gfs2_glock_put(sdp->sd_freeze_gl); fail_rename: gfs2_glock_put(sdp->sd_rename_gl); fail_live: @@ -429,15 +452,14 @@ static int gfs2_lookup_root(struct super_block *sb, struct dentry **dptr, struct dentry *dentry; struct inode *inode; - inode = gfs2_inode_lookup(sb, DT_DIR, no_addr, 0); + inode = gfs2_inode_lookup(sb, DT_DIR, no_addr, 0, 0); if (IS_ERR(inode)) { fs_err(sdp, "can't read in %s inode: %ld\n", name, PTR_ERR(inode)); return PTR_ERR(inode); } - dentry = d_alloc_root(inode); + dentry = d_make_root(inode); if (!dentry) { fs_err(sdp, "can't alloc %s dentry\n", name); - iput(inode); return -ENOMEM; } *dptr = dentry; @@ -500,73 +522,16 @@ out: return ret; } -/** - * map_journal_extents - create a reusable "extent" mapping from all logical - * blocks to all physical blocks for the given journal. This will save - * us time when writing journal blocks. Most journals will have only one - * extent that maps all their logical blocks. That's because gfs2.mkfs - * arranges the journal blocks sequentially to maximize performance. - * So the extent would map the first block for the entire file length. - * However, gfs2_jadd can happen while file activity is happening, so - * those journals may not be sequential. Less likely is the case where - * the users created their own journals by mounting the metafs and - * laying it out. But it's still possible. These journals might have - * several extents. - * - * TODO: This should be done in bigger chunks rather than one block at a time, - * but since it's only done at mount time, I'm not worried about the - * time it takes. - */ -static int map_journal_extents(struct gfs2_sbd *sdp) -{ - struct gfs2_jdesc *jd = sdp->sd_jdesc; - unsigned int lb; - u64 db, prev_db; /* logical block, disk block, prev disk block */ - struct gfs2_inode *ip = GFS2_I(jd->jd_inode); - struct gfs2_journal_extent *jext = NULL; - struct buffer_head bh; - int rc = 0; - - prev_db = 0; - - for (lb = 0; lb < i_size_read(jd->jd_inode) >> sdp->sd_sb.sb_bsize_shift; lb++) { - bh.b_state = 0; - bh.b_blocknr = 0; - bh.b_size = 1 << ip->i_inode.i_blkbits; - rc = gfs2_block_map(jd->jd_inode, lb, &bh, 0); - db = bh.b_blocknr; - if (rc || !db) { - printk(KERN_INFO "GFS2 journal mapping error %d: lb=" - "%u db=%llu\n", rc, lb, (unsigned long long)db); - break; - } - if (!prev_db || db != prev_db + 1) { - jext = kzalloc(sizeof(struct gfs2_journal_extent), - GFP_KERNEL); - if (!jext) { - printk(KERN_INFO "GFS2 error: out of memory " - "mapping journal extents.\n"); - rc = -ENOMEM; - break; - } - jext->dblock = db; - jext->lblock = lb; - jext->blocks = 1; - list_add_tail(&jext->extent_list, &jd->extent_list); - } else { - jext->blocks++; - } - prev_db = db; - } - return rc; -} - static void gfs2_others_may_mount(struct gfs2_sbd *sdp) { char *message = "FIRSTMOUNT=Done"; char *envp[] = { message, NULL }; - struct lm_lockstruct *ls = &sdp->sd_lockstruct; - ls->ls_first_done = 1; + + fs_info(sdp, "first mount done, others may mount\n"); + + if (sdp->sd_lockstruct.ls_ops->lm_first_done) + sdp->sd_lockstruct.ls_ops->lm_first_done(sdp); + kobject_uevent_env(&sdp->sd_kobj, KOBJ_CHANGE, envp); } @@ -615,6 +580,8 @@ static int gfs2_jindex_hold(struct gfs2_sbd *sdp, struct gfs2_holder *ji_gh) break; INIT_LIST_HEAD(&jd->extent_list); + INIT_LIST_HEAD(&jd->jd_revoke_list); + INIT_WORK(&jd->jd_work, gfs2_recover_func); jd->jd_inode = gfs2_lookupi(sdp->sd_jindex, &name, 1); if (!jd->jd_inode || IS_ERR(jd->jd_inode)) { @@ -637,6 +604,48 @@ static int gfs2_jindex_hold(struct gfs2_sbd *sdp, struct gfs2_holder *ji_gh) return error; } +/** + * check_journal_clean - Make sure a journal is clean for a spectator mount + * @sdp: The GFS2 superblock + * @jd: The journal descriptor + * + * Returns: 0 if the journal is clean or locked, else an error + */ +static int check_journal_clean(struct gfs2_sbd *sdp, struct gfs2_jdesc *jd) +{ + int error; + struct gfs2_holder j_gh; + struct gfs2_log_header_host head; + struct gfs2_inode *ip; + + ip = GFS2_I(jd->jd_inode); + error = gfs2_glock_nq_init(ip->i_gl, LM_ST_SHARED, LM_FLAG_NOEXP | + GL_EXACT | GL_NOCACHE, &j_gh); + if (error) { + fs_err(sdp, "Error locking journal for spectator mount.\n"); + return -EPERM; + } + error = gfs2_jdesc_check(jd); + if (error) { + fs_err(sdp, "Error checking journal for spectator mount.\n"); + goto out_unlock; + } + error = gfs2_find_jhead(jd, &head); + if (error) { + fs_err(sdp, "Error parsing journal for spectator mount.\n"); + goto out_unlock; + } + if (!(head.lh_flags & GFS2_LOG_HEAD_UNMOUNT)) { + error = -EPERM; + fs_err(sdp, "jid=%u: Journal is dirty, so the first mounter " + "must not be a spectator.\n", jd->jd_jid); + } + +out_unlock: + gfs2_glock_dq_uninit(&j_gh); + return error; +} + static int init_journal(struct gfs2_sbd *sdp, int undo) { struct inode *master = sdp->sd_master_dir->d_inode; @@ -655,7 +664,6 @@ static int init_journal(struct gfs2_sbd *sdp, int undo) fs_err(sdp, "can't lookup journal index: %d\n", error); return PTR_ERR(sdp->sd_jindex); } - ip = GFS2_I(sdp->sd_jindex); /* Load in the journal index special file */ @@ -717,15 +725,22 @@ static int init_journal(struct gfs2_sbd *sdp, int undo) atomic_set(&sdp->sd_log_thresh2, 4*sdp->sd_jdesc->jd_blocks/5); /* Map the extents for this journal's blocks */ - map_journal_extents(sdp); + gfs2_map_journal_extents(sdp, sdp->sd_jdesc); } trace_gfs2_log_blocks(sdp, atomic_read(&sdp->sd_log_blks_free)); if (sdp->sd_lockstruct.ls_first) { unsigned int x; for (x = 0; x < sdp->sd_journals; x++) { - error = gfs2_recover_journal(gfs2_jdesc_find(sdp, x), - true); + struct gfs2_jdesc *jd = gfs2_jdesc_find(sdp, x); + + if (sdp->sd_args.ar_spectator) { + error = check_journal_clean(sdp, jd); + if (error) + goto fail_jinode_gh; + continue; + } + error = gfs2_recover_journal(jd, true); if (error) { fs_err(sdp, "error recovering journal %u: %d\n", x, error); @@ -745,7 +760,15 @@ static int init_journal(struct gfs2_sbd *sdp, int undo) set_bit(SDF_JOURNAL_CHECKED, &sdp->sd_flags); gfs2_glock_dq_uninit(&ji_gh); jindex = 0; - + if (!sdp->sd_args.ar_spectator) { + error = gfs2_glock_nq_init(sdp->sd_freeze_gl, LM_ST_SHARED, 0, + &sdp->sd_thaw_gh); + if (error) { + fs_err(sdp, "can't acquire freeze glock: %d\n", error); + goto fail_jinode_gh; + } + } + gfs2_glock_dq_uninit(&sdp->sd_thaw_gh); return 0; fail_jinode_gh: @@ -763,17 +786,18 @@ fail: return error; } +static struct lock_class_key gfs2_quota_imutex_key; static int init_inodes(struct gfs2_sbd *sdp, int undo) { int error = 0; - struct gfs2_inode *ip; struct inode *master = sdp->sd_master_dir->d_inode; if (undo) goto fail_qinode; error = init_journal(sdp, undo); + complete_all(&sdp->sd_journal_ready); if (error) goto fail; @@ -792,7 +816,6 @@ static int init_inodes(struct gfs2_sbd *sdp, int undo) fs_err(sdp, "can't get resource index inode: %d\n", error); goto fail_statfs; } - ip = GFS2_I(sdp->sd_rindex); sdp->sd_rindex_uptodate = 0; /* Read in the quota inode */ @@ -802,6 +825,17 @@ static int init_inodes(struct gfs2_sbd *sdp, int undo) fs_err(sdp, "can't get quota file inode: %d\n", error); goto fail_rindex; } + /* + * i_mutex on quota files is special. Since this inode is hidden system + * file, we are safe to define locking ourselves. + */ + lockdep_set_class(&sdp->sd_quota_inode->i_mutex, + &gfs2_quota_imutex_key); + + error = gfs2_rindex_update(sdp); + if (error) + goto fail_qinode; + return 0; fail_qinode: @@ -889,56 +923,14 @@ fail: return error; } -static int init_threads(struct gfs2_sbd *sdp, int undo) -{ - struct task_struct *p; - int error = 0; - - if (undo) - goto fail_quotad; - - p = kthread_run(gfs2_logd, sdp, "gfs2_logd"); - error = IS_ERR(p); - if (error) { - fs_err(sdp, "can't start logd thread: %d\n", error); - return error; - } - sdp->sd_logd_process = p; - - p = kthread_run(gfs2_quotad, sdp, "gfs2_quotad"); - error = IS_ERR(p); - if (error) { - fs_err(sdp, "can't start quotad thread: %d\n", error); - goto fail; - } - sdp->sd_quotad_process = p; - - return 0; - - -fail_quotad: - kthread_stop(sdp->sd_quotad_process); -fail: - kthread_stop(sdp->sd_logd_process); - return error; -} - static const match_table_t nolock_tokens = { { Opt_jid, "jid=%d\n", }, { Opt_err, NULL }, }; -static void nolock_put_lock(struct kmem_cache *cachep, struct gfs2_glock *gl) -{ - struct gfs2_sbd *sdp = gl->gl_sbd; - kmem_cache_free(cachep, gl); - if (atomic_dec_and_test(&sdp->sd_glock_disposal)) - wake_up(&sdp->sd_glock_wait); -} - static const struct lm_lockops nolock_ops = { .lm_proto_name = "lock_nolock", - .lm_put_lock = nolock_put_lock, + .lm_put_lock = gfs2_glock_free, .lm_tokens = &nolock_tokens, }; @@ -958,7 +950,6 @@ static int gfs2_lm_mount(struct gfs2_sbd *sdp, int silent) struct gfs2_args *args = &sdp->sd_args; const char *proto = sdp->sd_proto_name; const char *table = sdp->sd_table_name; - const char *fsname; char *o, *options; int ret; @@ -970,7 +961,7 @@ static int gfs2_lm_mount(struct gfs2_sbd *sdp, int silent) lm = &gfs2_dlm_ops; #endif } else { - printk(KERN_INFO "GFS2: can't find protocol %s\n", proto); + pr_info("can't find protocol %s\n", proto); return -ENOENT; } @@ -996,6 +987,7 @@ static int gfs2_lm_mount(struct gfs2_sbd *sdp, int silent) ls->ls_jid = option; break; case Opt_id: + case Opt_nodir: /* Obsolete, but left for backward compat purposes */ break; case Opt_first: @@ -1004,12 +996,6 @@ static int gfs2_lm_mount(struct gfs2_sbd *sdp, int silent) goto hostdata_error; ls->ls_first = option; break; - case Opt_nodir: - ret = match_int(&tmp[0], &option); - if (ret || (option != 0 && option != 1)) - goto hostdata_error; - ls->ls_nodir = option; - break; case Opt_err: default: hostdata_error: @@ -1018,22 +1004,15 @@ hostdata_error: } } - if (sdp->sd_args.ar_spectator) - snprintf(sdp->sd_fsname, GFS2_FSNAME_LEN, "%s.s", table); - else - snprintf(sdp->sd_fsname, GFS2_FSNAME_LEN, "%s.%u", table, - sdp->sd_lockstruct.ls_jid); - - fsname = strchr(table, ':'); - if (fsname) - fsname++; if (lm->lm_mount == NULL) { fs_info(sdp, "Now mounting FS...\n"); + complete_all(&sdp->sd_locking_init); return 0; } - ret = lm->lm_mount(sdp, fsname); + ret = lm->lm_mount(sdp, table); if (ret == 0) fs_info(sdp, "Joined cluster. Now mounting FS...\n"); + complete_all(&sdp->sd_locking_init); return ret; } @@ -1089,20 +1068,21 @@ static int fill_super(struct super_block *sb, struct gfs2_args *args, int silent sdp = init_sbd(sb); if (!sdp) { - printk(KERN_WARNING "GFS2: can't alloc struct gfs2_sbd\n"); + pr_warn("can't alloc struct gfs2_sbd\n"); return -ENOMEM; } sdp->sd_args = *args; if (sdp->sd_args.ar_spectator) { sb->s_flags |= MS_RDONLY; - set_bit(SDF_NORECOVERY, &sdp->sd_flags); + set_bit(SDF_RORECOVERY, &sdp->sd_flags); } if (sdp->sd_args.ar_posix_acl) sb->s_flags |= MS_POSIXACL; if (sdp->sd_args.ar_nobarrier) set_bit(SDF_NOBARRIERS, &sdp->sd_flags); + sb->s_flags |= MS_NOSEC; sb->s_magic = GFS2_MAGIC; sb->s_op = &gfs2_super_ops; sb->s_d_op = &gfs2_dops; @@ -1126,25 +1106,39 @@ static int fill_super(struct super_block *sb, struct gfs2_args *args, int silent if (sdp->sd_args.ar_statfs_quantum) { sdp->sd_tune.gt_statfs_slow = 0; sdp->sd_tune.gt_statfs_quantum = sdp->sd_args.ar_statfs_quantum; - } - else { + } else { sdp->sd_tune.gt_statfs_slow = 1; sdp->sd_tune.gt_statfs_quantum = 30; } error = init_names(sdp, silent); - if (error) - goto fail; + if (error) { + /* In this case, we haven't initialized sysfs, so we have to + manually free the sdp. */ + free_percpu(sdp->sd_lkstats); + kfree(sdp); + sb->s_fs_info = NULL; + return error; + } - gfs2_create_debugfs_file(sdp); + snprintf(sdp->sd_fsname, GFS2_FSNAME_LEN, "%s", sdp->sd_table_name); error = gfs2_sys_fs_add(sdp); + /* + * If we hit an error here, gfs2_sys_fs_add will have called function + * kobject_put which causes the sysfs usage count to go to zero, which + * causes sysfs to call function gfs2_sbd_release, which frees sdp. + * Subsequent error paths here will call gfs2_sys_fs_del, which also + * kobject_put to free sdp. + */ if (error) - goto fail; + return error; + + gfs2_create_debugfs_file(sdp); error = gfs2_lm_mount(sdp, silent); if (error) - goto fail_sys; + goto fail_debug; error = init_locking(sdp, &mount_gh, DO); if (error) @@ -1172,6 +1166,13 @@ static int fill_super(struct super_block *sb, struct gfs2_args *args, int silent goto fail_sb; } + if (sdp->sd_args.ar_spectator) + snprintf(sdp->sd_fsname, GFS2_FSNAME_LEN, "%s.s", + sdp->sd_table_name); + else + snprintf(sdp->sd_fsname, GFS2_FSNAME_LEN, "%s.%u", + sdp->sd_table_name, sdp->sd_lockstruct.ls_jid); + error = init_inodes(sdp, DO); if (error) goto fail_sb; @@ -1186,15 +1187,11 @@ static int fill_super(struct super_block *sb, struct gfs2_args *args, int silent goto fail_per_node; } - error = init_threads(sdp, DO); - if (error) - goto fail_per_node; - if (!(sb->s_flags & MS_RDONLY)) { error = gfs2_make_fs_rw(sdp); if (error) { fs_err(sdp, "can't make FS RW: %d\n", error); - goto fail_threads; + goto fail_per_node; } } @@ -1202,8 +1199,6 @@ static int fill_super(struct super_block *sb, struct gfs2_args *args, int silent gfs2_online_uevent(sdp); return 0; -fail_threads: - init_threads(sdp, UNDO); fail_per_node: init_per_node(sdp, UNDO); fail_inodes: @@ -1219,13 +1214,15 @@ fail_sb: fail_locking: init_locking(sdp, &mount_gh, UNDO); fail_lm: + complete_all(&sdp->sd_journal_ready); gfs2_gl_hash_clear(sdp); gfs2_lm_unmount(sdp); -fail_sys: - gfs2_sys_fs_del(sdp); -fail: +fail_debug: gfs2_delete_debugfs_file(sdp); - kfree(sdp); + free_percpu(sdp->sd_lkstats); + /* gfs2_sys_fs_del must be the last thing we do, since it causes + * sysfs to call function gfs2_sbd_release, which frees sdp. */ + gfs2_sys_fs_del(sdp); sb->s_fs_info = NULL; return error; } @@ -1291,14 +1288,24 @@ static struct dentry *gfs2_mount(struct file_system_type *fs_type, int flags, error = -EBUSY; goto error_bdev; } - s = sget(fs_type, test_gfs2_super, set_gfs2_super, bdev); + s = sget(fs_type, test_gfs2_super, set_gfs2_super, flags, bdev); mutex_unlock(&bdev->bd_fsfreeze_mutex); error = PTR_ERR(s); if (IS_ERR(s)) goto error_bdev; - if (s->s_root) + if (s->s_root) { + /* + * s_umount nests inside bd_mutex during + * __invalidate_device(). blkdev_put() acquires + * bd_mutex and can't be called under s_umount. Drop + * s_umount temporarily. This is safe as we're + * holding an active reference. + */ + up_write(&s->s_umount); blkdev_put(bdev, mode); + down_write(&s->s_umount); + } memset(&args, 0, sizeof(args)); args.ar_quota = GFS2_QUOTA_DEFAULT; @@ -1310,7 +1317,7 @@ static struct dentry *gfs2_mount(struct file_system_type *fs_type, int flags, error = gfs2_mount_args(&args, data); if (error) { - printk(KERN_WARNING "GFS2: can't parse mount arguments\n"); + pr_warn("can't parse mount arguments\n"); goto error_super; } @@ -1321,7 +1328,6 @@ static struct dentry *gfs2_mount(struct file_system_type *fs_type, int flags, } else { char b[BDEVNAME_SIZE]; - s->s_flags = flags; s->s_mode = mode; strlcpy(s->s_id, bdevname(bdev, b), sizeof(s->s_id)); sb_set_blocksize(s, block_size(bdev)); @@ -1361,15 +1367,15 @@ static struct dentry *gfs2_mount_meta(struct file_system_type *fs_type, error = kern_path(dev_name, LOOKUP_FOLLOW, &path); if (error) { - printk(KERN_WARNING "GFS2: path_lookup on %s returned error %d\n", - dev_name, error); + pr_warn("path_lookup on %s returned error %d\n", + dev_name, error); return ERR_PTR(error); } - s = sget(&gfs2_fs_type, test_gfs2_super, set_meta_super, + s = sget(&gfs2_fs_type, test_gfs2_super, set_meta_super, flags, path.dentry->d_inode->i_sb->s_bdev); path_put(&path); if (IS_ERR(s)) { - printk(KERN_WARNING "GFS2: gfs2 mount does not exist\n"); + pr_warn("gfs2 mount does not exist\n"); return ERR_CAST(s); } if ((flags ^ s->s_flags) & MS_RDONLY) { @@ -1389,15 +1395,15 @@ static void gfs2_kill_sb(struct super_block *sb) return; } - gfs2_meta_syncfs(sdp); + gfs2_log_flush(sdp, NULL, SYNC_FLUSH); dput(sdp->sd_root_dir); dput(sdp->sd_master_dir); sdp->sd_root_dir = NULL; sdp->sd_master_dir = NULL; shrink_dcache_sb(sb); - kill_block_super(sb); gfs2_delete_debugfs_file(sdp); - kfree(sdp); + free_percpu(sdp->sd_lkstats); + kill_block_super(sb); } struct file_system_type gfs2_fs_type = { @@ -1407,6 +1413,7 @@ struct file_system_type gfs2_fs_type = { .kill_sb = gfs2_kill_sb, .owner = THIS_MODULE, }; +MODULE_ALIAS_FS("gfs2"); struct file_system_type gfs2meta_fs_type = { .name = "gfs2meta", @@ -1414,4 +1421,4 @@ struct file_system_type gfs2meta_fs_type = { .mount = gfs2_mount_meta, .owner = THIS_MODULE, }; - +MODULE_ALIAS_FS("gfs2meta"); diff --git a/fs/gfs2/ops_inode.c b/fs/gfs2/ops_inode.c deleted file mode 100644 index d8b26ac2e20..00000000000 --- a/fs/gfs2/ops_inode.c +++ /dev/null @@ -1,1344 +0,0 @@ -/* - * Copyright (C) Sistina Software, Inc. 1997-2003 All rights reserved. - * Copyright (C) 2004-2006 Red Hat, Inc. All rights reserved. - * - * This copyrighted material is made available to anyone wishing to use, - * modify, copy, or redistribute it subject to the terms and conditions - * of the GNU General Public License version 2. - */ - -#include <linux/slab.h> -#include <linux/spinlock.h> -#include <linux/completion.h> -#include <linux/buffer_head.h> -#include <linux/namei.h> -#include <linux/mm.h> -#include <linux/xattr.h> -#include <linux/posix_acl.h> -#include <linux/gfs2_ondisk.h> -#include <linux/crc32.h> -#include <linux/fiemap.h> -#include <asm/uaccess.h> - -#include "gfs2.h" -#include "incore.h" -#include "acl.h" -#include "bmap.h" -#include "dir.h" -#include "xattr.h" -#include "glock.h" -#include "inode.h" -#include "meta_io.h" -#include "quota.h" -#include "rgrp.h" -#include "trans.h" -#include "util.h" -#include "super.h" - -/** - * gfs2_create - Create a file - * @dir: The directory in which to create the file - * @dentry: The dentry of the new file - * @mode: The mode of the new file - * - * Returns: errno - */ - -static int gfs2_create(struct inode *dir, struct dentry *dentry, - int mode, struct nameidata *nd) -{ - struct gfs2_inode *dip = GFS2_I(dir); - struct gfs2_sbd *sdp = GFS2_SB(dir); - struct gfs2_holder ghs[2]; - struct inode *inode; - - gfs2_holder_init(dip->i_gl, 0, 0, ghs); - - for (;;) { - inode = gfs2_createi(ghs, &dentry->d_name, S_IFREG | mode, 0); - if (!IS_ERR(inode)) { - gfs2_trans_end(sdp); - if (dip->i_alloc->al_rgd) - gfs2_inplace_release(dip); - gfs2_quota_unlock(dip); - gfs2_alloc_put(dip); - gfs2_glock_dq_uninit_m(2, ghs); - mark_inode_dirty(inode); - break; - } else if (PTR_ERR(inode) != -EEXIST || - (nd && nd->flags & LOOKUP_EXCL)) { - gfs2_holder_uninit(ghs); - return PTR_ERR(inode); - } - - inode = gfs2_lookupi(dir, &dentry->d_name, 0); - if (inode) { - if (!IS_ERR(inode)) { - gfs2_holder_uninit(ghs); - break; - } else { - gfs2_holder_uninit(ghs); - return PTR_ERR(inode); - } - } - } - - d_instantiate(dentry, inode); - - return 0; -} - -/** - * gfs2_lookup - Look up a filename in a directory and return its inode - * @dir: The directory inode - * @dentry: The dentry of the new inode - * @nd: passed from Linux VFS, ignored by us - * - * Called by the VFS layer. Lock dir and call gfs2_lookupi() - * - * Returns: errno - */ - -static struct dentry *gfs2_lookup(struct inode *dir, struct dentry *dentry, - struct nameidata *nd) -{ - struct inode *inode = NULL; - - inode = gfs2_lookupi(dir, &dentry->d_name, 0); - if (inode && IS_ERR(inode)) - return ERR_CAST(inode); - - if (inode) { - struct gfs2_glock *gl = GFS2_I(inode)->i_gl; - struct gfs2_holder gh; - int error; - error = gfs2_glock_nq_init(gl, LM_ST_SHARED, LM_FLAG_ANY, &gh); - if (error) { - iput(inode); - return ERR_PTR(error); - } - gfs2_glock_dq_uninit(&gh); - return d_splice_alias(inode, dentry); - } - d_add(dentry, inode); - - return NULL; -} - -/** - * gfs2_link - Link to a file - * @old_dentry: The inode to link - * @dir: Add link to this directory - * @dentry: The name of the link - * - * Link the inode in "old_dentry" into the directory "dir" with the - * name in "dentry". - * - * Returns: errno - */ - -static int gfs2_link(struct dentry *old_dentry, struct inode *dir, - struct dentry *dentry) -{ - struct gfs2_inode *dip = GFS2_I(dir); - struct gfs2_sbd *sdp = GFS2_SB(dir); - struct inode *inode = old_dentry->d_inode; - struct gfs2_inode *ip = GFS2_I(inode); - struct gfs2_holder ghs[2]; - int alloc_required; - int error; - - if (S_ISDIR(inode->i_mode)) - return -EPERM; - - gfs2_holder_init(dip->i_gl, LM_ST_EXCLUSIVE, 0, ghs); - gfs2_holder_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, ghs + 1); - - error = gfs2_glock_nq(ghs); /* parent */ - if (error) - goto out_parent; - - error = gfs2_glock_nq(ghs + 1); /* child */ - if (error) - goto out_child; - - error = gfs2_permission(dir, MAY_WRITE | MAY_EXEC, 0); - if (error) - goto out_gunlock; - - error = gfs2_dir_check(dir, &dentry->d_name, NULL); - switch (error) { - case -ENOENT: - break; - case 0: - error = -EEXIST; - default: - goto out_gunlock; - } - - error = -EINVAL; - if (!dip->i_inode.i_nlink) - goto out_gunlock; - error = -EFBIG; - if (dip->i_entries == (u32)-1) - goto out_gunlock; - error = -EPERM; - if (IS_IMMUTABLE(inode) || IS_APPEND(inode)) - goto out_gunlock; - error = -EINVAL; - if (!ip->i_inode.i_nlink) - goto out_gunlock; - error = -EMLINK; - if (ip->i_inode.i_nlink == (u32)-1) - goto out_gunlock; - - alloc_required = error = gfs2_diradd_alloc_required(dir, &dentry->d_name); - if (error < 0) - goto out_gunlock; - error = 0; - - if (alloc_required) { - struct gfs2_alloc *al = gfs2_alloc_get(dip); - if (!al) { - error = -ENOMEM; - goto out_gunlock; - } - - error = gfs2_quota_lock_check(dip); - if (error) - goto out_alloc; - - al->al_requested = sdp->sd_max_dirres; - - error = gfs2_inplace_reserve(dip); - if (error) - goto out_gunlock_q; - - error = gfs2_trans_begin(sdp, sdp->sd_max_dirres + - gfs2_rg_blocks(al) + - 2 * RES_DINODE + RES_STATFS + - RES_QUOTA, 0); - if (error) - goto out_ipres; - } else { - error = gfs2_trans_begin(sdp, 2 * RES_DINODE + RES_LEAF, 0); - if (error) - goto out_ipres; - } - - error = gfs2_dir_add(dir, &dentry->d_name, ip, IF2DT(inode->i_mode)); - if (error) - goto out_end_trans; - - error = gfs2_change_nlink(ip, +1); - -out_end_trans: - gfs2_trans_end(sdp); -out_ipres: - if (alloc_required) - gfs2_inplace_release(dip); -out_gunlock_q: - if (alloc_required) - gfs2_quota_unlock(dip); -out_alloc: - if (alloc_required) - gfs2_alloc_put(dip); -out_gunlock: - gfs2_glock_dq(ghs + 1); -out_child: - gfs2_glock_dq(ghs); -out_parent: - gfs2_holder_uninit(ghs); - gfs2_holder_uninit(ghs + 1); - if (!error) { - ihold(inode); - d_instantiate(dentry, inode); - mark_inode_dirty(inode); - } - return error; -} - -/* - * gfs2_unlink_ok - check to see that a inode is still in a directory - * @dip: the directory - * @name: the name of the file - * @ip: the inode - * - * Assumes that the lock on (at least) @dip is held. - * - * Returns: 0 if the parent/child relationship is correct, errno if it isn't - */ - -static int gfs2_unlink_ok(struct gfs2_inode *dip, const struct qstr *name, - const struct gfs2_inode *ip) -{ - int error; - - if (IS_IMMUTABLE(&ip->i_inode) || IS_APPEND(&ip->i_inode)) - return -EPERM; - - if ((dip->i_inode.i_mode & S_ISVTX) && - dip->i_inode.i_uid != current_fsuid() && - ip->i_inode.i_uid != current_fsuid() && !capable(CAP_FOWNER)) - return -EPERM; - - if (IS_APPEND(&dip->i_inode)) - return -EPERM; - - error = gfs2_permission(&dip->i_inode, MAY_WRITE | MAY_EXEC, 0); - if (error) - return error; - - error = gfs2_dir_check(&dip->i_inode, name, ip); - if (error) - return error; - - return 0; -} - -/** - * gfs2_unlink - Unlink a file - * @dir: The inode of the directory containing the file to unlink - * @dentry: The file itself - * - * Unlink a file. Call gfs2_unlinki() - * - * Returns: errno - */ - -static int gfs2_unlink(struct inode *dir, struct dentry *dentry) -{ - struct gfs2_inode *dip = GFS2_I(dir); - struct gfs2_sbd *sdp = GFS2_SB(dir); - struct gfs2_inode *ip = GFS2_I(dentry->d_inode); - struct gfs2_holder ghs[3]; - struct gfs2_rgrpd *rgd; - struct gfs2_holder ri_gh; - int error; - - error = gfs2_rindex_hold(sdp, &ri_gh); - if (error) - return error; - - gfs2_holder_init(dip->i_gl, LM_ST_EXCLUSIVE, 0, ghs); - gfs2_holder_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, ghs + 1); - - rgd = gfs2_blk2rgrpd(sdp, ip->i_no_addr); - gfs2_holder_init(rgd->rd_gl, LM_ST_EXCLUSIVE, 0, ghs + 2); - - - error = gfs2_glock_nq(ghs); /* parent */ - if (error) - goto out_parent; - - error = gfs2_glock_nq(ghs + 1); /* child */ - if (error) - goto out_child; - - error = gfs2_glock_nq(ghs + 2); /* rgrp */ - if (error) - goto out_rgrp; - - error = gfs2_unlink_ok(dip, &dentry->d_name, ip); - if (error) - goto out_gunlock; - - error = gfs2_trans_begin(sdp, 2*RES_DINODE + RES_LEAF + RES_RG_BIT, 0); - if (error) - goto out_gunlock; - - error = gfs2_dir_del(dip, &dentry->d_name); - if (error) - goto out_end_trans; - - error = gfs2_change_nlink(ip, -1); - -out_end_trans: - gfs2_trans_end(sdp); -out_gunlock: - gfs2_glock_dq(ghs + 2); -out_rgrp: - gfs2_holder_uninit(ghs + 2); - gfs2_glock_dq(ghs + 1); -out_child: - gfs2_holder_uninit(ghs + 1); - gfs2_glock_dq(ghs); -out_parent: - gfs2_holder_uninit(ghs); - gfs2_glock_dq_uninit(&ri_gh); - return error; -} - -/** - * gfs2_symlink - Create a symlink - * @dir: The directory to create the symlink in - * @dentry: The dentry to put the symlink in - * @symname: The thing which the link points to - * - * Returns: errno - */ - -static int gfs2_symlink(struct inode *dir, struct dentry *dentry, - const char *symname) -{ - struct gfs2_inode *dip = GFS2_I(dir), *ip; - struct gfs2_sbd *sdp = GFS2_SB(dir); - struct gfs2_holder ghs[2]; - struct inode *inode; - struct buffer_head *dibh; - int size; - int error; - - /* Must be stuffed with a null terminator for gfs2_follow_link() */ - size = strlen(symname); - if (size > sdp->sd_sb.sb_bsize - sizeof(struct gfs2_dinode) - 1) - return -ENAMETOOLONG; - - gfs2_holder_init(dip->i_gl, 0, 0, ghs); - - inode = gfs2_createi(ghs, &dentry->d_name, S_IFLNK | S_IRWXUGO, 0); - if (IS_ERR(inode)) { - gfs2_holder_uninit(ghs); - return PTR_ERR(inode); - } - - ip = ghs[1].gh_gl->gl_object; - - i_size_write(inode, size); - - error = gfs2_meta_inode_buffer(ip, &dibh); - - if (!gfs2_assert_withdraw(sdp, !error)) { - gfs2_dinode_out(ip, dibh->b_data); - memcpy(dibh->b_data + sizeof(struct gfs2_dinode), symname, - size); - brelse(dibh); - } - - gfs2_trans_end(sdp); - if (dip->i_alloc->al_rgd) - gfs2_inplace_release(dip); - gfs2_quota_unlock(dip); - gfs2_alloc_put(dip); - - gfs2_glock_dq_uninit_m(2, ghs); - - d_instantiate(dentry, inode); - mark_inode_dirty(inode); - - return 0; -} - -/** - * gfs2_mkdir - Make a directory - * @dir: The parent directory of the new one - * @dentry: The dentry of the new directory - * @mode: The mode of the new directory - * - * Returns: errno - */ - -static int gfs2_mkdir(struct inode *dir, struct dentry *dentry, int mode) -{ - struct gfs2_inode *dip = GFS2_I(dir), *ip; - struct gfs2_sbd *sdp = GFS2_SB(dir); - struct gfs2_holder ghs[2]; - struct inode *inode; - struct buffer_head *dibh; - int error; - - gfs2_holder_init(dip->i_gl, 0, 0, ghs); - - inode = gfs2_createi(ghs, &dentry->d_name, S_IFDIR | mode, 0); - if (IS_ERR(inode)) { - gfs2_holder_uninit(ghs); - return PTR_ERR(inode); - } - - ip = ghs[1].gh_gl->gl_object; - - ip->i_inode.i_nlink = 2; - i_size_write(inode, sdp->sd_sb.sb_bsize - sizeof(struct gfs2_dinode)); - ip->i_diskflags |= GFS2_DIF_JDATA; - ip->i_entries = 2; - - error = gfs2_meta_inode_buffer(ip, &dibh); - - if (!gfs2_assert_withdraw(sdp, !error)) { - struct gfs2_dinode *di = (struct gfs2_dinode *)dibh->b_data; - struct gfs2_dirent *dent = (struct gfs2_dirent *)(di+1); - - gfs2_trans_add_bh(ip->i_gl, dibh, 1); - gfs2_qstr2dirent(&gfs2_qdot, GFS2_DIRENT_SIZE(gfs2_qdot.len), dent); - dent->de_inum = di->di_num; /* already GFS2 endian */ - dent->de_type = cpu_to_be16(DT_DIR); - di->di_entries = cpu_to_be32(1); - - dent = (struct gfs2_dirent *)((char*)dent + GFS2_DIRENT_SIZE(1)); - gfs2_qstr2dirent(&gfs2_qdotdot, dibh->b_size - GFS2_DIRENT_SIZE(1) - sizeof(struct gfs2_dinode), dent); - - gfs2_inum_out(dip, dent); - dent->de_type = cpu_to_be16(DT_DIR); - - gfs2_dinode_out(ip, di); - - brelse(dibh); - } - - error = gfs2_change_nlink(dip, +1); - gfs2_assert_withdraw(sdp, !error); /* dip already pinned */ - - gfs2_trans_end(sdp); - if (dip->i_alloc->al_rgd) - gfs2_inplace_release(dip); - gfs2_quota_unlock(dip); - gfs2_alloc_put(dip); - - gfs2_glock_dq_uninit_m(2, ghs); - - d_instantiate(dentry, inode); - mark_inode_dirty(inode); - - return 0; -} - -/** - * gfs2_rmdiri - Remove a directory - * @dip: The parent directory of the directory to be removed - * @name: The name of the directory to be removed - * @ip: The GFS2 inode of the directory to be removed - * - * Assumes Glocks on dip and ip are held - * - * Returns: errno - */ - -static int gfs2_rmdiri(struct gfs2_inode *dip, const struct qstr *name, - struct gfs2_inode *ip) -{ - int error; - - if (ip->i_entries != 2) { - if (gfs2_consist_inode(ip)) - gfs2_dinode_print(ip); - return -EIO; - } - - error = gfs2_dir_del(dip, name); - if (error) - return error; - - error = gfs2_change_nlink(dip, -1); - if (error) - return error; - - error = gfs2_dir_del(ip, &gfs2_qdot); - if (error) - return error; - - error = gfs2_dir_del(ip, &gfs2_qdotdot); - if (error) - return error; - - /* It looks odd, but it really should be done twice */ - error = gfs2_change_nlink(ip, -1); - if (error) - return error; - - error = gfs2_change_nlink(ip, -1); - if (error) - return error; - - return error; -} - -/** - * gfs2_rmdir - Remove a directory - * @dir: The parent directory of the directory to be removed - * @dentry: The dentry of the directory to remove - * - * Remove a directory. Call gfs2_rmdiri() - * - * Returns: errno - */ - -static int gfs2_rmdir(struct inode *dir, struct dentry *dentry) -{ - struct gfs2_inode *dip = GFS2_I(dir); - struct gfs2_sbd *sdp = GFS2_SB(dir); - struct gfs2_inode *ip = GFS2_I(dentry->d_inode); - struct gfs2_holder ghs[3]; - struct gfs2_rgrpd *rgd; - struct gfs2_holder ri_gh; - int error; - - error = gfs2_rindex_hold(sdp, &ri_gh); - if (error) - return error; - gfs2_holder_init(dip->i_gl, LM_ST_EXCLUSIVE, 0, ghs); - gfs2_holder_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, ghs + 1); - - rgd = gfs2_blk2rgrpd(sdp, ip->i_no_addr); - gfs2_holder_init(rgd->rd_gl, LM_ST_EXCLUSIVE, 0, ghs + 2); - - error = gfs2_glock_nq(ghs); /* parent */ - if (error) - goto out_parent; - - error = gfs2_glock_nq(ghs + 1); /* child */ - if (error) - goto out_child; - - error = gfs2_glock_nq(ghs + 2); /* rgrp */ - if (error) - goto out_rgrp; - - error = gfs2_unlink_ok(dip, &dentry->d_name, ip); - if (error) - goto out_gunlock; - - if (ip->i_entries < 2) { - if (gfs2_consist_inode(ip)) - gfs2_dinode_print(ip); - error = -EIO; - goto out_gunlock; - } - if (ip->i_entries > 2) { - error = -ENOTEMPTY; - goto out_gunlock; - } - - error = gfs2_trans_begin(sdp, 2 * RES_DINODE + 3 * RES_LEAF + RES_RG_BIT, 0); - if (error) - goto out_gunlock; - - error = gfs2_rmdiri(dip, &dentry->d_name, ip); - - gfs2_trans_end(sdp); - -out_gunlock: - gfs2_glock_dq(ghs + 2); -out_rgrp: - gfs2_holder_uninit(ghs + 2); - gfs2_glock_dq(ghs + 1); -out_child: - gfs2_holder_uninit(ghs + 1); - gfs2_glock_dq(ghs); -out_parent: - gfs2_holder_uninit(ghs); - gfs2_glock_dq_uninit(&ri_gh); - return error; -} - -/** - * gfs2_mknod - Make a special file - * @dir: The directory in which the special file will reside - * @dentry: The dentry of the special file - * @mode: The mode of the special file - * @rdev: The device specification of the special file - * - */ - -static int gfs2_mknod(struct inode *dir, struct dentry *dentry, int mode, - dev_t dev) -{ - struct gfs2_inode *dip = GFS2_I(dir); - struct gfs2_sbd *sdp = GFS2_SB(dir); - struct gfs2_holder ghs[2]; - struct inode *inode; - - gfs2_holder_init(dip->i_gl, 0, 0, ghs); - - inode = gfs2_createi(ghs, &dentry->d_name, mode, dev); - if (IS_ERR(inode)) { - gfs2_holder_uninit(ghs); - return PTR_ERR(inode); - } - - gfs2_trans_end(sdp); - if (dip->i_alloc->al_rgd) - gfs2_inplace_release(dip); - gfs2_quota_unlock(dip); - gfs2_alloc_put(dip); - - gfs2_glock_dq_uninit_m(2, ghs); - - d_instantiate(dentry, inode); - mark_inode_dirty(inode); - - return 0; -} - -/* - * gfs2_ok_to_move - check if it's ok to move a directory to another directory - * @this: move this - * @to: to here - * - * Follow @to back to the root and make sure we don't encounter @this - * Assumes we already hold the rename lock. - * - * Returns: errno - */ - -static int gfs2_ok_to_move(struct gfs2_inode *this, struct gfs2_inode *to) -{ - struct inode *dir = &to->i_inode; - struct super_block *sb = dir->i_sb; - struct inode *tmp; - int error = 0; - - igrab(dir); - - for (;;) { - if (dir == &this->i_inode) { - error = -EINVAL; - break; - } - if (dir == sb->s_root->d_inode) { - error = 0; - break; - } - - tmp = gfs2_lookupi(dir, &gfs2_qdotdot, 1); - if (IS_ERR(tmp)) { - error = PTR_ERR(tmp); - break; - } - - iput(dir); - dir = tmp; - } - - iput(dir); - - return error; -} - -/** - * gfs2_rename - Rename a file - * @odir: Parent directory of old file name - * @odentry: The old dentry of the file - * @ndir: Parent directory of new file name - * @ndentry: The new dentry of the file - * - * Returns: errno - */ - -static int gfs2_rename(struct inode *odir, struct dentry *odentry, - struct inode *ndir, struct dentry *ndentry) -{ - struct gfs2_inode *odip = GFS2_I(odir); - struct gfs2_inode *ndip = GFS2_I(ndir); - struct gfs2_inode *ip = GFS2_I(odentry->d_inode); - struct gfs2_inode *nip = NULL; - struct gfs2_sbd *sdp = GFS2_SB(odir); - struct gfs2_holder ghs[5], r_gh = { .gh_gl = NULL, }, ri_gh; - struct gfs2_rgrpd *nrgd; - unsigned int num_gh; - int dir_rename = 0; - int alloc_required = 0; - unsigned int x; - int error; - - if (ndentry->d_inode) { - nip = GFS2_I(ndentry->d_inode); - if (ip == nip) - return 0; - } - - error = gfs2_rindex_hold(sdp, &ri_gh); - if (error) - return error; - - if (odip != ndip) { - error = gfs2_glock_nq_init(sdp->sd_rename_gl, LM_ST_EXCLUSIVE, - 0, &r_gh); - if (error) - goto out; - - if (S_ISDIR(ip->i_inode.i_mode)) { - dir_rename = 1; - /* don't move a dirctory into it's subdir */ - error = gfs2_ok_to_move(ip, ndip); - if (error) - goto out_gunlock_r; - } - } - - num_gh = 1; - gfs2_holder_init(odip->i_gl, LM_ST_EXCLUSIVE, 0, ghs); - if (odip != ndip) { - gfs2_holder_init(ndip->i_gl, LM_ST_EXCLUSIVE, 0, ghs + num_gh); - num_gh++; - } - gfs2_holder_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, ghs + num_gh); - num_gh++; - - if (nip) { - gfs2_holder_init(nip->i_gl, LM_ST_EXCLUSIVE, 0, ghs + num_gh); - num_gh++; - /* grab the resource lock for unlink flag twiddling - * this is the case of the target file already existing - * so we unlink before doing the rename - */ - nrgd = gfs2_blk2rgrpd(sdp, nip->i_no_addr); - if (nrgd) - gfs2_holder_init(nrgd->rd_gl, LM_ST_EXCLUSIVE, 0, ghs + num_gh++); - } - - for (x = 0; x < num_gh; x++) { - error = gfs2_glock_nq(ghs + x); - if (error) - goto out_gunlock; - } - - /* Check out the old directory */ - - error = gfs2_unlink_ok(odip, &odentry->d_name, ip); - if (error) - goto out_gunlock; - - /* Check out the new directory */ - - if (nip) { - error = gfs2_unlink_ok(ndip, &ndentry->d_name, nip); - if (error) - goto out_gunlock; - - if (S_ISDIR(nip->i_inode.i_mode)) { - if (nip->i_entries < 2) { - if (gfs2_consist_inode(nip)) - gfs2_dinode_print(nip); - error = -EIO; - goto out_gunlock; - } - if (nip->i_entries > 2) { - error = -ENOTEMPTY; - goto out_gunlock; - } - } - } else { - error = gfs2_permission(ndir, MAY_WRITE | MAY_EXEC, 0); - if (error) - goto out_gunlock; - - error = gfs2_dir_check(ndir, &ndentry->d_name, NULL); - switch (error) { - case -ENOENT: - error = 0; - break; - case 0: - error = -EEXIST; - default: - goto out_gunlock; - }; - - if (odip != ndip) { - if (!ndip->i_inode.i_nlink) { - error = -EINVAL; - goto out_gunlock; - } - if (ndip->i_entries == (u32)-1) { - error = -EFBIG; - goto out_gunlock; - } - if (S_ISDIR(ip->i_inode.i_mode) && - ndip->i_inode.i_nlink == (u32)-1) { - error = -EMLINK; - goto out_gunlock; - } - } - } - - /* Check out the dir to be renamed */ - - if (dir_rename) { - error = gfs2_permission(odentry->d_inode, MAY_WRITE, 0); - if (error) - goto out_gunlock; - } - - if (nip == NULL) - alloc_required = gfs2_diradd_alloc_required(ndir, &ndentry->d_name); - error = alloc_required; - if (error < 0) - goto out_gunlock; - error = 0; - - if (alloc_required) { - struct gfs2_alloc *al = gfs2_alloc_get(ndip); - if (!al) { - error = -ENOMEM; - goto out_gunlock; - } - - error = gfs2_quota_lock_check(ndip); - if (error) - goto out_alloc; - - al->al_requested = sdp->sd_max_dirres; - - error = gfs2_inplace_reserve_ri(ndip); - if (error) - goto out_gunlock_q; - - error = gfs2_trans_begin(sdp, sdp->sd_max_dirres + - gfs2_rg_blocks(al) + - 4 * RES_DINODE + 4 * RES_LEAF + - RES_STATFS + RES_QUOTA + 4, 0); - if (error) - goto out_ipreserv; - } else { - error = gfs2_trans_begin(sdp, 4 * RES_DINODE + - 5 * RES_LEAF + 4, 0); - if (error) - goto out_gunlock; - } - - /* Remove the target file, if it exists */ - - if (nip) { - if (S_ISDIR(nip->i_inode.i_mode)) - error = gfs2_rmdiri(ndip, &ndentry->d_name, nip); - else { - error = gfs2_dir_del(ndip, &ndentry->d_name); - if (error) - goto out_end_trans; - error = gfs2_change_nlink(nip, -1); - } - if (error) - goto out_end_trans; - } - - if (dir_rename) { - error = gfs2_change_nlink(ndip, +1); - if (error) - goto out_end_trans; - error = gfs2_change_nlink(odip, -1); - if (error) - goto out_end_trans; - - error = gfs2_dir_mvino(ip, &gfs2_qdotdot, ndip, DT_DIR); - if (error) - goto out_end_trans; - } else { - struct buffer_head *dibh; - error = gfs2_meta_inode_buffer(ip, &dibh); - if (error) - goto out_end_trans; - ip->i_inode.i_ctime = CURRENT_TIME; - gfs2_trans_add_bh(ip->i_gl, dibh, 1); - gfs2_dinode_out(ip, dibh->b_data); - brelse(dibh); - } - - error = gfs2_dir_del(odip, &odentry->d_name); - if (error) - goto out_end_trans; - - error = gfs2_dir_add(ndir, &ndentry->d_name, ip, IF2DT(ip->i_inode.i_mode)); - if (error) - goto out_end_trans; - -out_end_trans: - gfs2_trans_end(sdp); -out_ipreserv: - if (alloc_required) - gfs2_inplace_release(ndip); -out_gunlock_q: - if (alloc_required) - gfs2_quota_unlock(ndip); -out_alloc: - if (alloc_required) - gfs2_alloc_put(ndip); -out_gunlock: - while (x--) { - gfs2_glock_dq(ghs + x); - gfs2_holder_uninit(ghs + x); - } -out_gunlock_r: - if (r_gh.gh_gl) - gfs2_glock_dq_uninit(&r_gh); -out: - gfs2_glock_dq_uninit(&ri_gh); - return error; -} - -/** - * gfs2_follow_link - Follow a symbolic link - * @dentry: The dentry of the link - * @nd: Data that we pass to vfs_follow_link() - * - * This can handle symlinks of any size. - * - * Returns: 0 on success or error code - */ - -static void *gfs2_follow_link(struct dentry *dentry, struct nameidata *nd) -{ - struct gfs2_inode *ip = GFS2_I(dentry->d_inode); - struct gfs2_holder i_gh; - struct buffer_head *dibh; - unsigned int x, size; - char *buf; - int error; - - gfs2_holder_init(ip->i_gl, LM_ST_SHARED, 0, &i_gh); - error = gfs2_glock_nq(&i_gh); - if (error) { - gfs2_holder_uninit(&i_gh); - nd_set_link(nd, ERR_PTR(error)); - return NULL; - } - - size = (unsigned int)i_size_read(&ip->i_inode); - if (size == 0) { - gfs2_consist_inode(ip); - buf = ERR_PTR(-EIO); - goto out; - } - - error = gfs2_meta_inode_buffer(ip, &dibh); - if (error) { - buf = ERR_PTR(error); - goto out; - } - - x = size + 1; - buf = kmalloc(x, GFP_NOFS); - if (!buf) - buf = ERR_PTR(-ENOMEM); - else - memcpy(buf, dibh->b_data + sizeof(struct gfs2_dinode), x); - brelse(dibh); -out: - gfs2_glock_dq_uninit(&i_gh); - nd_set_link(nd, buf); - return NULL; -} - -static void gfs2_put_link(struct dentry *dentry, struct nameidata *nd, void *p) -{ - char *s = nd_get_link(nd); - if (!IS_ERR(s)) - kfree(s); -} - -/** - * gfs2_permission - - * @inode: - * @mask: - * @nd: passed from Linux VFS, ignored by us - * - * This may be called from the VFS directly, or from within GFS2 with the - * inode locked, so we look to see if the glock is already locked and only - * lock the glock if its not already been done. - * - * Returns: errno - */ - -int gfs2_permission(struct inode *inode, int mask, unsigned int flags) -{ - struct gfs2_inode *ip; - struct gfs2_holder i_gh; - int error; - int unlock = 0; - - if (flags & IPERM_FLAG_RCU) - return -ECHILD; - - ip = GFS2_I(inode); - if (gfs2_glock_is_locked_by_me(ip->i_gl) == NULL) { - error = gfs2_glock_nq_init(ip->i_gl, LM_ST_SHARED, LM_FLAG_ANY, &i_gh); - if (error) - return error; - unlock = 1; - } - - if ((mask & MAY_WRITE) && IS_IMMUTABLE(inode)) - error = -EACCES; - else - error = generic_permission(inode, mask, flags, gfs2_check_acl); - if (unlock) - gfs2_glock_dq_uninit(&i_gh); - - return error; -} - -static int setattr_chown(struct inode *inode, struct iattr *attr) -{ - struct gfs2_inode *ip = GFS2_I(inode); - struct gfs2_sbd *sdp = GFS2_SB(inode); - u32 ouid, ogid, nuid, ngid; - int error; - - ouid = inode->i_uid; - ogid = inode->i_gid; - nuid = attr->ia_uid; - ngid = attr->ia_gid; - - if (!(attr->ia_valid & ATTR_UID) || ouid == nuid) - ouid = nuid = NO_QUOTA_CHANGE; - if (!(attr->ia_valid & ATTR_GID) || ogid == ngid) - ogid = ngid = NO_QUOTA_CHANGE; - - if (!gfs2_alloc_get(ip)) - return -ENOMEM; - - error = gfs2_quota_lock(ip, nuid, ngid); - if (error) - goto out_alloc; - - if (ouid != NO_QUOTA_CHANGE || ogid != NO_QUOTA_CHANGE) { - error = gfs2_quota_check(ip, nuid, ngid); - if (error) - goto out_gunlock_q; - } - - error = gfs2_trans_begin(sdp, RES_DINODE + 2 * RES_QUOTA, 0); - if (error) - goto out_gunlock_q; - - error = gfs2_setattr_simple(ip, attr); - if (error) - goto out_end_trans; - - if (ouid != NO_QUOTA_CHANGE || ogid != NO_QUOTA_CHANGE) { - u64 blocks = gfs2_get_inode_blocks(&ip->i_inode); - gfs2_quota_change(ip, -blocks, ouid, ogid); - gfs2_quota_change(ip, blocks, nuid, ngid); - } - -out_end_trans: - gfs2_trans_end(sdp); -out_gunlock_q: - gfs2_quota_unlock(ip); -out_alloc: - gfs2_alloc_put(ip); - return error; -} - -/** - * gfs2_setattr - Change attributes on an inode - * @dentry: The dentry which is changing - * @attr: The structure describing the change - * - * The VFS layer wants to change one or more of an inodes attributes. Write - * that change out to disk. - * - * Returns: errno - */ - -static int gfs2_setattr(struct dentry *dentry, struct iattr *attr) -{ - struct inode *inode = dentry->d_inode; - struct gfs2_inode *ip = GFS2_I(inode); - struct gfs2_holder i_gh; - int error; - - error = gfs2_glock_nq_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, &i_gh); - if (error) - return error; - - error = -EPERM; - if (IS_IMMUTABLE(inode) || IS_APPEND(inode)) - goto out; - - error = inode_change_ok(inode, attr); - if (error) - goto out; - - if (attr->ia_valid & ATTR_SIZE) - error = gfs2_setattr_size(inode, attr->ia_size); - else if (attr->ia_valid & (ATTR_UID | ATTR_GID)) - error = setattr_chown(inode, attr); - else if ((attr->ia_valid & ATTR_MODE) && IS_POSIXACL(inode)) - error = gfs2_acl_chmod(ip, attr); - else - error = gfs2_setattr_simple(ip, attr); - -out: - gfs2_glock_dq_uninit(&i_gh); - if (!error) - mark_inode_dirty(inode); - return error; -} - -/** - * gfs2_getattr - Read out an inode's attributes - * @mnt: The vfsmount the inode is being accessed from - * @dentry: The dentry to stat - * @stat: The inode's stats - * - * This may be called from the VFS directly, or from within GFS2 with the - * inode locked, so we look to see if the glock is already locked and only - * lock the glock if its not already been done. Note that its the NFS - * readdirplus operation which causes this to be called (from filldir) - * with the glock already held. - * - * Returns: errno - */ - -static int gfs2_getattr(struct vfsmount *mnt, struct dentry *dentry, - struct kstat *stat) -{ - struct inode *inode = dentry->d_inode; - struct gfs2_inode *ip = GFS2_I(inode); - struct gfs2_holder gh; - int error; - int unlock = 0; - - if (gfs2_glock_is_locked_by_me(ip->i_gl) == NULL) { - error = gfs2_glock_nq_init(ip->i_gl, LM_ST_SHARED, LM_FLAG_ANY, &gh); - if (error) - return error; - unlock = 1; - } - - generic_fillattr(inode, stat); - if (unlock) - gfs2_glock_dq_uninit(&gh); - - return 0; -} - -static int gfs2_setxattr(struct dentry *dentry, const char *name, - const void *data, size_t size, int flags) -{ - struct inode *inode = dentry->d_inode; - struct gfs2_inode *ip = GFS2_I(inode); - struct gfs2_holder gh; - int ret; - - gfs2_holder_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, &gh); - ret = gfs2_glock_nq(&gh); - if (ret == 0) { - ret = generic_setxattr(dentry, name, data, size, flags); - gfs2_glock_dq(&gh); - } - gfs2_holder_uninit(&gh); - return ret; -} - -static ssize_t gfs2_getxattr(struct dentry *dentry, const char *name, - void *data, size_t size) -{ - struct inode *inode = dentry->d_inode; - struct gfs2_inode *ip = GFS2_I(inode); - struct gfs2_holder gh; - int ret; - - gfs2_holder_init(ip->i_gl, LM_ST_SHARED, LM_FLAG_ANY, &gh); - ret = gfs2_glock_nq(&gh); - if (ret == 0) { - ret = generic_getxattr(dentry, name, data, size); - gfs2_glock_dq(&gh); - } - gfs2_holder_uninit(&gh); - return ret; -} - -static int gfs2_removexattr(struct dentry *dentry, const char *name) -{ - struct inode *inode = dentry->d_inode; - struct gfs2_inode *ip = GFS2_I(inode); - struct gfs2_holder gh; - int ret; - - gfs2_holder_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, &gh); - ret = gfs2_glock_nq(&gh); - if (ret == 0) { - ret = generic_removexattr(dentry, name); - gfs2_glock_dq(&gh); - } - gfs2_holder_uninit(&gh); - return ret; -} - -static int gfs2_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, - u64 start, u64 len) -{ - struct gfs2_inode *ip = GFS2_I(inode); - struct gfs2_holder gh; - int ret; - - ret = fiemap_check_flags(fieinfo, FIEMAP_FLAG_SYNC); - if (ret) - return ret; - - mutex_lock(&inode->i_mutex); - - ret = gfs2_glock_nq_init(ip->i_gl, LM_ST_SHARED, 0, &gh); - if (ret) - goto out; - - if (gfs2_is_stuffed(ip)) { - u64 phys = ip->i_no_addr << inode->i_blkbits; - u64 size = i_size_read(inode); - u32 flags = FIEMAP_EXTENT_LAST|FIEMAP_EXTENT_NOT_ALIGNED| - FIEMAP_EXTENT_DATA_INLINE; - phys += sizeof(struct gfs2_dinode); - phys += start; - if (start + len > size) - len = size - start; - if (start < size) - ret = fiemap_fill_next_extent(fieinfo, start, phys, - len, flags); - if (ret == 1) - ret = 0; - } else { - ret = __generic_block_fiemap(inode, fieinfo, start, len, - gfs2_block_map); - } - - gfs2_glock_dq_uninit(&gh); -out: - mutex_unlock(&inode->i_mutex); - return ret; -} - -const struct inode_operations gfs2_file_iops = { - .permission = gfs2_permission, - .setattr = gfs2_setattr, - .getattr = gfs2_getattr, - .setxattr = gfs2_setxattr, - .getxattr = gfs2_getxattr, - .listxattr = gfs2_listxattr, - .removexattr = gfs2_removexattr, - .fiemap = gfs2_fiemap, -}; - -const struct inode_operations gfs2_dir_iops = { - .create = gfs2_create, - .lookup = gfs2_lookup, - .link = gfs2_link, - .unlink = gfs2_unlink, - .symlink = gfs2_symlink, - .mkdir = gfs2_mkdir, - .rmdir = gfs2_rmdir, - .mknod = gfs2_mknod, - .rename = gfs2_rename, - .permission = gfs2_permission, - .setattr = gfs2_setattr, - .getattr = gfs2_getattr, - .setxattr = gfs2_setxattr, - .getxattr = gfs2_getxattr, - .listxattr = gfs2_listxattr, - .removexattr = gfs2_removexattr, - .fiemap = gfs2_fiemap, -}; - -const struct inode_operations gfs2_symlink_iops = { - .readlink = generic_readlink, - .follow_link = gfs2_follow_link, - .put_link = gfs2_put_link, - .permission = gfs2_permission, - .setattr = gfs2_setattr, - .getattr = gfs2_getattr, - .setxattr = gfs2_setxattr, - .getxattr = gfs2_getxattr, - .listxattr = gfs2_listxattr, - .removexattr = gfs2_removexattr, - .fiemap = gfs2_fiemap, -}; - diff --git a/fs/gfs2/quota.c b/fs/gfs2/quota.c index a689901963d..64b29f7f6b4 100644 --- a/fs/gfs2/quota.c +++ b/fs/gfs2/quota.c @@ -36,8 +36,11 @@ * the quota file, so it is not being constantly read. */ +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + #include <linux/sched.h> #include <linux/slab.h> +#include <linux/mm.h> #include <linux/spinlock.h> #include <linux/completion.h> #include <linux/buffer_head.h> @@ -49,6 +52,13 @@ #include <linux/freezer.h> #include <linux/quota.h> #include <linux/dqblk_xfs.h> +#include <linux/lockref.h> +#include <linux/list_lru.h> +#include <linux/rcupdate.h> +#include <linux/rculist_bl.h> +#include <linux/bit_spinlock.h> +#include <linux/jhash.h> +#include <linux/vmalloc.h> #include "gfs2.h" #include "incore.h" @@ -64,38 +74,63 @@ #include "inode.h" #include "util.h" -#define QUOTA_USER 1 -#define QUOTA_GROUP 0 +#define GFS2_QD_HASH_SHIFT 12 +#define GFS2_QD_HASH_SIZE (1 << GFS2_QD_HASH_SHIFT) +#define GFS2_QD_HASH_MASK (GFS2_QD_HASH_SIZE - 1) -struct gfs2_quota_change_host { - u64 qc_change; - u32 qc_flags; /* GFS2_QCF_... */ - u32 qc_id; -}; +/* Lock order: qd_lock -> bucket lock -> qd->lockref.lock -> lru lock */ +/* -> sd_bitmap_lock */ +static DEFINE_SPINLOCK(qd_lock); +struct list_lru gfs2_qd_lru; -static LIST_HEAD(qd_lru_list); -static atomic_t qd_lru_count = ATOMIC_INIT(0); -static DEFINE_SPINLOCK(qd_lru_lock); +static struct hlist_bl_head qd_hash_table[GFS2_QD_HASH_SIZE]; -int gfs2_shrink_qd_memory(struct shrinker *shrink, int nr, gfp_t gfp_mask) +static unsigned int gfs2_qd_hash(const struct gfs2_sbd *sdp, + const struct kqid qid) { - struct gfs2_quota_data *qd; - struct gfs2_sbd *sdp; + unsigned int h; - if (nr == 0) - goto out; + h = jhash(&sdp, sizeof(struct gfs2_sbd *), 0); + h = jhash(&qid, sizeof(struct kqid), h); - if (!(gfp_mask & __GFP_FS)) - return -1; + return h & GFS2_QD_HASH_MASK; +} + +static inline void spin_lock_bucket(unsigned int hash) +{ + hlist_bl_lock(&qd_hash_table[hash]); +} + +static inline void spin_unlock_bucket(unsigned int hash) +{ + hlist_bl_unlock(&qd_hash_table[hash]); +} - spin_lock(&qd_lru_lock); - while (nr && !list_empty(&qd_lru_list)) { - qd = list_entry(qd_lru_list.next, - struct gfs2_quota_data, qd_reclaim); +static void gfs2_qd_dealloc(struct rcu_head *rcu) +{ + struct gfs2_quota_data *qd = container_of(rcu, struct gfs2_quota_data, qd_rcu); + kmem_cache_free(gfs2_quotad_cachep, qd); +} + +static void gfs2_qd_dispose(struct list_head *list) +{ + struct gfs2_quota_data *qd; + struct gfs2_sbd *sdp; + + while (!list_empty(list)) { + qd = list_entry(list->next, struct gfs2_quota_data, qd_lru); sdp = qd->qd_gl->gl_sbd; + list_del(&qd->qd_lru); + /* Free from the filesystem-specific list */ + spin_lock(&qd_lock); list_del(&qd->qd_list); + spin_unlock(&qd_lock); + + spin_lock_bucket(qd->qd_hash); + hlist_bl_del_rcu(&qd->qd_hlist); + spin_unlock_bucket(qd->qd_hash); gfs2_assert_warn(sdp, !qd->qd_change); gfs2_assert_warn(sdp, !qd->qd_slot_count); @@ -105,194 +140,228 @@ int gfs2_shrink_qd_memory(struct shrinker *shrink, int nr, gfp_t gfp_mask) atomic_dec(&sdp->sd_quota_count); /* Delete it from the common reclaim list */ - list_del_init(&qd->qd_reclaim); - atomic_dec(&qd_lru_count); - spin_unlock(&qd_lru_lock); - kmem_cache_free(gfs2_quotad_cachep, qd); - spin_lock(&qd_lru_lock); - nr--; + call_rcu(&qd->qd_rcu, gfs2_qd_dealloc); } - spin_unlock(&qd_lru_lock); +} -out: - return (atomic_read(&qd_lru_count) * sysctl_vfs_cache_pressure) / 100; + +static enum lru_status gfs2_qd_isolate(struct list_head *item, spinlock_t *lock, void *arg) +{ + struct list_head *dispose = arg; + struct gfs2_quota_data *qd = list_entry(item, struct gfs2_quota_data, qd_lru); + + if (!spin_trylock(&qd->qd_lockref.lock)) + return LRU_SKIP; + + if (qd->qd_lockref.count == 0) { + lockref_mark_dead(&qd->qd_lockref); + list_move(&qd->qd_lru, dispose); + } + + spin_unlock(&qd->qd_lockref.lock); + return LRU_REMOVED; +} + +static unsigned long gfs2_qd_shrink_scan(struct shrinker *shrink, + struct shrink_control *sc) +{ + LIST_HEAD(dispose); + unsigned long freed; + + if (!(sc->gfp_mask & __GFP_FS)) + return SHRINK_STOP; + + freed = list_lru_walk_node(&gfs2_qd_lru, sc->nid, gfs2_qd_isolate, + &dispose, &sc->nr_to_scan); + + gfs2_qd_dispose(&dispose); + + return freed; +} + +static unsigned long gfs2_qd_shrink_count(struct shrinker *shrink, + struct shrink_control *sc) +{ + return vfs_pressure_ratio(list_lru_count_node(&gfs2_qd_lru, sc->nid)); +} + +struct shrinker gfs2_qd_shrinker = { + .count_objects = gfs2_qd_shrink_count, + .scan_objects = gfs2_qd_shrink_scan, + .seeks = DEFAULT_SEEKS, + .flags = SHRINKER_NUMA_AWARE, +}; + + +static u64 qd2index(struct gfs2_quota_data *qd) +{ + struct kqid qid = qd->qd_id; + return (2 * (u64)from_kqid(&init_user_ns, qid)) + + ((qid.type == USRQUOTA) ? 0 : 1); } static u64 qd2offset(struct gfs2_quota_data *qd) { u64 offset; - offset = 2 * (u64)qd->qd_id + !test_bit(QDF_USER, &qd->qd_flags); + offset = qd2index(qd); offset *= sizeof(struct gfs2_quota); return offset; } -static int qd_alloc(struct gfs2_sbd *sdp, int user, u32 id, - struct gfs2_quota_data **qdp) +static struct gfs2_quota_data *qd_alloc(unsigned hash, struct gfs2_sbd *sdp, struct kqid qid) { struct gfs2_quota_data *qd; int error; qd = kmem_cache_zalloc(gfs2_quotad_cachep, GFP_NOFS); if (!qd) - return -ENOMEM; + return NULL; - atomic_set(&qd->qd_count, 1); - qd->qd_id = id; - if (user) - set_bit(QDF_USER, &qd->qd_flags); + qd->qd_sbd = sdp; + qd->qd_lockref.count = 1; + spin_lock_init(&qd->qd_lockref.lock); + qd->qd_id = qid; qd->qd_slot = -1; - INIT_LIST_HEAD(&qd->qd_reclaim); + INIT_LIST_HEAD(&qd->qd_lru); + qd->qd_hash = hash; - error = gfs2_glock_get(sdp, 2 * (u64)id + !user, + error = gfs2_glock_get(sdp, qd2index(qd), &gfs2_quota_glops, CREATE, &qd->qd_gl); if (error) goto fail; - *qdp = qd; - - return 0; + return qd; fail: kmem_cache_free(gfs2_quotad_cachep, qd); - return error; + return NULL; } -static int qd_get(struct gfs2_sbd *sdp, int user, u32 id, - struct gfs2_quota_data **qdp) +static struct gfs2_quota_data *gfs2_qd_search_bucket(unsigned int hash, + const struct gfs2_sbd *sdp, + struct kqid qid) { - struct gfs2_quota_data *qd = NULL, *new_qd = NULL; - int error, found; - - *qdp = NULL; + struct gfs2_quota_data *qd; + struct hlist_bl_node *h; - for (;;) { - found = 0; - spin_lock(&qd_lru_lock); - list_for_each_entry(qd, &sdp->sd_quota_list, qd_list) { - if (qd->qd_id == id && - !test_bit(QDF_USER, &qd->qd_flags) == !user) { - if (!atomic_read(&qd->qd_count) && - !list_empty(&qd->qd_reclaim)) { - /* Remove it from reclaim list */ - list_del_init(&qd->qd_reclaim); - atomic_dec(&qd_lru_count); - } - atomic_inc(&qd->qd_count); - found = 1; - break; - } + hlist_bl_for_each_entry_rcu(qd, h, &qd_hash_table[hash], qd_hlist) { + if (!qid_eq(qd->qd_id, qid)) + continue; + if (qd->qd_sbd != sdp) + continue; + if (lockref_get_not_dead(&qd->qd_lockref)) { + list_lru_del(&gfs2_qd_lru, &qd->qd_lru); + return qd; } + } - if (!found) - qd = NULL; + return NULL; +} - if (!qd && new_qd) { - qd = new_qd; - list_add(&qd->qd_list, &sdp->sd_quota_list); - atomic_inc(&sdp->sd_quota_count); - new_qd = NULL; - } - spin_unlock(&qd_lru_lock); +static int qd_get(struct gfs2_sbd *sdp, struct kqid qid, + struct gfs2_quota_data **qdp) +{ + struct gfs2_quota_data *qd, *new_qd; + unsigned int hash = gfs2_qd_hash(sdp, qid); - if (qd) { - if (new_qd) { - gfs2_glock_put(new_qd->qd_gl); - kmem_cache_free(gfs2_quotad_cachep, new_qd); - } - *qdp = qd; - return 0; - } + rcu_read_lock(); + *qdp = qd = gfs2_qd_search_bucket(hash, sdp, qid); + rcu_read_unlock(); - error = qd_alloc(sdp, user, id, &new_qd); - if (error) - return error; + if (qd) + return 0; + + new_qd = qd_alloc(hash, sdp, qid); + if (!new_qd) + return -ENOMEM; + + spin_lock(&qd_lock); + spin_lock_bucket(hash); + *qdp = qd = gfs2_qd_search_bucket(hash, sdp, qid); + if (qd == NULL) { + *qdp = new_qd; + list_add(&new_qd->qd_list, &sdp->sd_quota_list); + hlist_bl_add_head_rcu(&new_qd->qd_hlist, &qd_hash_table[hash]); + atomic_inc(&sdp->sd_quota_count); + } + spin_unlock_bucket(hash); + spin_unlock(&qd_lock); + + if (qd) { + gfs2_glock_put(new_qd->qd_gl); + kmem_cache_free(gfs2_quotad_cachep, new_qd); } + + return 0; } + static void qd_hold(struct gfs2_quota_data *qd) { struct gfs2_sbd *sdp = qd->qd_gl->gl_sbd; - gfs2_assert(sdp, atomic_read(&qd->qd_count)); - atomic_inc(&qd->qd_count); + gfs2_assert(sdp, !__lockref_is_dead(&qd->qd_lockref)); + lockref_get(&qd->qd_lockref); } static void qd_put(struct gfs2_quota_data *qd) { - if (atomic_dec_and_lock(&qd->qd_count, &qd_lru_lock)) { - /* Add to the reclaim list */ - list_add_tail(&qd->qd_reclaim, &qd_lru_list); - atomic_inc(&qd_lru_count); - spin_unlock(&qd_lru_lock); - } + if (lockref_put_or_lock(&qd->qd_lockref)) + return; + + qd->qd_lockref.count = 0; + list_lru_add(&gfs2_qd_lru, &qd->qd_lru); + spin_unlock(&qd->qd_lockref.lock); + } static int slot_get(struct gfs2_quota_data *qd) { - struct gfs2_sbd *sdp = qd->qd_gl->gl_sbd; - unsigned int c, o = 0, b; - unsigned char byte = 0; + struct gfs2_sbd *sdp = qd->qd_sbd; + unsigned int bit; + int error = 0; - spin_lock(&qd_lru_lock); + spin_lock(&sdp->sd_bitmap_lock); + if (qd->qd_slot_count != 0) + goto out; - if (qd->qd_slot_count++) { - spin_unlock(&qd_lru_lock); - return 0; + error = -ENOSPC; + bit = find_first_zero_bit(sdp->sd_quota_bitmap, sdp->sd_quota_slots); + if (bit < sdp->sd_quota_slots) { + set_bit(bit, sdp->sd_quota_bitmap); + qd->qd_slot = bit; + error = 0; +out: + qd->qd_slot_count++; } + spin_unlock(&sdp->sd_bitmap_lock); - for (c = 0; c < sdp->sd_quota_chunks; c++) - for (o = 0; o < PAGE_SIZE; o++) { - byte = sdp->sd_quota_bitmap[c][o]; - if (byte != 0xFF) - goto found; - } - - goto fail; - -found: - for (b = 0; b < 8; b++) - if (!(byte & (1 << b))) - break; - qd->qd_slot = c * (8 * PAGE_SIZE) + o * 8 + b; - - if (qd->qd_slot >= sdp->sd_quota_slots) - goto fail; - - sdp->sd_quota_bitmap[c][o] |= 1 << b; - - spin_unlock(&qd_lru_lock); - - return 0; - -fail: - qd->qd_slot_count--; - spin_unlock(&qd_lru_lock); - return -ENOSPC; + return error; } static void slot_hold(struct gfs2_quota_data *qd) { - struct gfs2_sbd *sdp = qd->qd_gl->gl_sbd; + struct gfs2_sbd *sdp = qd->qd_sbd; - spin_lock(&qd_lru_lock); + spin_lock(&sdp->sd_bitmap_lock); gfs2_assert(sdp, qd->qd_slot_count); qd->qd_slot_count++; - spin_unlock(&qd_lru_lock); + spin_unlock(&sdp->sd_bitmap_lock); } static void slot_put(struct gfs2_quota_data *qd) { - struct gfs2_sbd *sdp = qd->qd_gl->gl_sbd; + struct gfs2_sbd *sdp = qd->qd_sbd; - spin_lock(&qd_lru_lock); + spin_lock(&sdp->sd_bitmap_lock); gfs2_assert(sdp, qd->qd_slot_count); if (!--qd->qd_slot_count) { - gfs2_icbit_munge(sdp, sdp->sd_quota_bitmap, qd->qd_slot, 0); + BUG_ON(!test_and_clear_bit(qd->qd_slot, sdp->sd_quota_bitmap)); qd->qd_slot = -1; } - spin_unlock(&qd_lru_lock); + spin_unlock(&sdp->sd_bitmap_lock); } static int bh_get(struct gfs2_quota_data *qd) @@ -356,6 +425,24 @@ static void bh_put(struct gfs2_quota_data *qd) mutex_unlock(&sdp->sd_quota_mutex); } +static int qd_check_sync(struct gfs2_sbd *sdp, struct gfs2_quota_data *qd, + u64 *sync_gen) +{ + if (test_bit(QDF_LOCKED, &qd->qd_flags) || + !test_bit(QDF_CHANGE, &qd->qd_flags) || + (sync_gen && (qd->qd_sync_gen >= *sync_gen))) + return 0; + + if (!lockref_get_not_dead(&qd->qd_lockref)) + return 0; + + list_move_tail(&qd->qd_list, &sdp->sd_quota_list); + set_bit(QDF_LOCKED, &qd->qd_flags); + qd->qd_change_sync = qd->qd_change; + slot_hold(qd); + return 1; +} + static int qd_fish(struct gfs2_sbd *sdp, struct gfs2_quota_data **qdp) { struct gfs2_quota_data *qd = NULL; @@ -367,31 +454,18 @@ static int qd_fish(struct gfs2_sbd *sdp, struct gfs2_quota_data **qdp) if (sdp->sd_vfs->s_flags & MS_RDONLY) return 0; - spin_lock(&qd_lru_lock); + spin_lock(&qd_lock); list_for_each_entry(qd, &sdp->sd_quota_list, qd_list) { - if (test_bit(QDF_LOCKED, &qd->qd_flags) || - !test_bit(QDF_CHANGE, &qd->qd_flags) || - qd->qd_sync_gen >= sdp->sd_quota_sync_gen) - continue; - - list_move_tail(&qd->qd_list, &sdp->sd_quota_list); - - set_bit(QDF_LOCKED, &qd->qd_flags); - gfs2_assert_warn(sdp, atomic_read(&qd->qd_count)); - atomic_inc(&qd->qd_count); - qd->qd_change_sync = qd->qd_change; - gfs2_assert_warn(sdp, qd->qd_slot_count); - qd->qd_slot_count++; - found = 1; - - break; + found = qd_check_sync(sdp, qd, &sdp->sd_quota_sync_gen); + if (found) + break; } if (!found) qd = NULL; - spin_unlock(&qd_lru_lock); + spin_unlock(&qd_lock); if (qd) { gfs2_assert_warn(sdp, qd->qd_change_sync); @@ -409,43 +483,6 @@ static int qd_fish(struct gfs2_sbd *sdp, struct gfs2_quota_data **qdp) return 0; } -static int qd_trylock(struct gfs2_quota_data *qd) -{ - struct gfs2_sbd *sdp = qd->qd_gl->gl_sbd; - - if (sdp->sd_vfs->s_flags & MS_RDONLY) - return 0; - - spin_lock(&qd_lru_lock); - - if (test_bit(QDF_LOCKED, &qd->qd_flags) || - !test_bit(QDF_CHANGE, &qd->qd_flags)) { - spin_unlock(&qd_lru_lock); - return 0; - } - - list_move_tail(&qd->qd_list, &sdp->sd_quota_list); - - set_bit(QDF_LOCKED, &qd->qd_flags); - gfs2_assert_warn(sdp, atomic_read(&qd->qd_count)); - atomic_inc(&qd->qd_count); - qd->qd_change_sync = qd->qd_change; - gfs2_assert_warn(sdp, qd->qd_slot_count); - qd->qd_slot_count++; - - spin_unlock(&qd_lru_lock); - - gfs2_assert_warn(sdp, qd->qd_change_sync); - if (bh_get(qd)) { - clear_bit(QDF_LOCKED, &qd->qd_flags); - slot_put(qd); - qd_put(qd); - return 0; - } - - return 1; -} - static void qd_unlock(struct gfs2_quota_data *qd) { gfs2_assert_warn(qd->qd_gl->gl_sbd, @@ -456,12 +493,12 @@ static void qd_unlock(struct gfs2_quota_data *qd) qd_put(qd); } -static int qdsb_get(struct gfs2_sbd *sdp, int user, u32 id, +static int qdsb_get(struct gfs2_sbd *sdp, struct kqid qid, struct gfs2_quota_data **qdp) { int error; - error = qd_get(sdp, user, id, qdp); + error = qd_get(sdp, qid, qdp); if (error) return error; @@ -489,45 +526,54 @@ static void qdsb_put(struct gfs2_quota_data *qd) qd_put(qd); } -int gfs2_quota_hold(struct gfs2_inode *ip, u32 uid, u32 gid) +int gfs2_quota_hold(struct gfs2_inode *ip, kuid_t uid, kgid_t gid) { struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); - struct gfs2_alloc *al = ip->i_alloc; - struct gfs2_quota_data **qd = al->al_qd; + struct gfs2_quota_data **qd; int error; - if (gfs2_assert_warn(sdp, !al->al_qd_num) || + if (ip->i_res == NULL) { + error = gfs2_rs_alloc(ip); + if (error) + return error; + } + + qd = ip->i_res->rs_qa_qd; + + if (gfs2_assert_warn(sdp, !ip->i_res->rs_qa_qd_num) || gfs2_assert_warn(sdp, !test_bit(GIF_QD_LOCKED, &ip->i_flags))) return -EIO; if (sdp->sd_args.ar_quota == GFS2_QUOTA_OFF) return 0; - error = qdsb_get(sdp, QUOTA_USER, ip->i_inode.i_uid, qd); + error = qdsb_get(sdp, make_kqid_uid(ip->i_inode.i_uid), qd); if (error) goto out; - al->al_qd_num++; + ip->i_res->rs_qa_qd_num++; qd++; - error = qdsb_get(sdp, QUOTA_GROUP, ip->i_inode.i_gid, qd); + error = qdsb_get(sdp, make_kqid_gid(ip->i_inode.i_gid), qd); if (error) goto out; - al->al_qd_num++; + ip->i_res->rs_qa_qd_num++; qd++; - if (uid != NO_QUOTA_CHANGE && uid != ip->i_inode.i_uid) { - error = qdsb_get(sdp, QUOTA_USER, uid, qd); + if (!uid_eq(uid, NO_UID_QUOTA_CHANGE) && + !uid_eq(uid, ip->i_inode.i_uid)) { + error = qdsb_get(sdp, make_kqid_uid(uid), qd); if (error) goto out; - al->al_qd_num++; + ip->i_res->rs_qa_qd_num++; qd++; } - if (gid != NO_QUOTA_CHANGE && gid != ip->i_inode.i_gid) { - error = qdsb_get(sdp, QUOTA_GROUP, gid, qd); + if (!gid_eq(gid, NO_GID_QUOTA_CHANGE) && + !gid_eq(gid, ip->i_inode.i_gid)) { + error = qdsb_get(sdp, make_kqid_gid(gid), qd); if (error) goto out; - al->al_qd_num++; + ip->i_res->rs_qa_qd_num++; qd++; } @@ -540,16 +586,17 @@ out: void gfs2_quota_unhold(struct gfs2_inode *ip) { struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); - struct gfs2_alloc *al = ip->i_alloc; unsigned int x; + if (ip->i_res == NULL) + return; gfs2_assert_warn(sdp, !test_bit(GIF_QD_LOCKED, &ip->i_flags)); - for (x = 0; x < al->al_qd_num; x++) { - qdsb_put(al->al_qd[x]); - al->al_qd[x] = NULL; + for (x = 0; x < ip->i_res->rs_qa_qd_num; x++) { + qdsb_put(ip->i_res->rs_qa_qd[x]); + ip->i_res->rs_qa_qd[x] = NULL; } - al->al_qd_num = 0; + ip->i_res->rs_qa_qd_num = 0; } static int sort_qd(const void *a, const void *b) @@ -557,18 +604,10 @@ static int sort_qd(const void *a, const void *b) const struct gfs2_quota_data *qd_a = *(const struct gfs2_quota_data **)a; const struct gfs2_quota_data *qd_b = *(const struct gfs2_quota_data **)b; - if (!test_bit(QDF_USER, &qd_a->qd_flags) != - !test_bit(QDF_USER, &qd_b->qd_flags)) { - if (test_bit(QDF_USER, &qd_a->qd_flags)) - return -1; - else - return 1; - } - if (qd_a->qd_id < qd_b->qd_id) + if (qid_lt(qd_a->qd_id, qd_b->qd_id)) return -1; - if (qd_a->qd_id > qd_b->qd_id) + if (qid_lt(qd_b->qd_id, qd_a->qd_id)) return 1; - return 0; } @@ -580,22 +619,22 @@ static void do_qc(struct gfs2_quota_data *qd, s64 change) s64 x; mutex_lock(&sdp->sd_quota_mutex); - gfs2_trans_add_bh(ip->i_gl, qd->qd_bh, 1); + gfs2_trans_add_meta(ip->i_gl, qd->qd_bh); if (!test_bit(QDF_CHANGE, &qd->qd_flags)) { qc->qc_change = 0; qc->qc_flags = 0; - if (test_bit(QDF_USER, &qd->qd_flags)) + if (qd->qd_id.type == USRQUOTA) qc->qc_flags = cpu_to_be32(GFS2_QCF_USER); - qc->qc_id = cpu_to_be32(qd->qd_id); + qc->qc_id = cpu_to_be32(from_kqid(&init_user_ns, qd->qd_id)); } x = be64_to_cpu(qc->qc_change) + change; qc->qc_change = cpu_to_be64(x); - spin_lock(&qd_lru_lock); + spin_lock(&qd_lock); qd->qd_change = x; - spin_unlock(&qd_lru_lock); + spin_unlock(&qd_lock); if (!x) { gfs2_assert_warn(sdp, test_bit(QDF_CHANGE, &qd->qd_flags)); @@ -636,47 +675,47 @@ static int gfs2_adjust_quota(struct gfs2_inode *ip, loff_t loc, unsigned long index = loc >> PAGE_CACHE_SHIFT; unsigned offset = loc & (PAGE_CACHE_SIZE - 1); unsigned blocksize, iblock, pos; - struct buffer_head *bh, *dibh; + struct buffer_head *bh; struct page *page; void *kaddr, *ptr; - struct gfs2_quota q, *qp; + struct gfs2_quota q; int err, nbytes; u64 size; - if (gfs2_is_stuffed(ip)) - gfs2_unstuff_dinode(ip, NULL); + if (gfs2_is_stuffed(ip)) { + err = gfs2_unstuff_dinode(ip, NULL); + if (err) + return err; + } memset(&q, 0, sizeof(struct gfs2_quota)); - err = gfs2_internal_read(ip, NULL, (char *)&q, &loc, sizeof(q)); + err = gfs2_internal_read(ip, (char *)&q, &loc, sizeof(q)); if (err < 0) return err; err = -EIO; - qp = &q; - qp->qu_value = be64_to_cpu(qp->qu_value); - qp->qu_value += change; - qp->qu_value = cpu_to_be64(qp->qu_value); - qd->qd_qb.qb_value = qp->qu_value; + be64_add_cpu(&q.qu_value, change); + qd->qd_qb.qb_value = q.qu_value; if (fdq) { if (fdq->d_fieldmask & FS_DQ_BSOFT) { - qp->qu_warn = cpu_to_be64(fdq->d_blk_softlimit >> sdp->sd_fsb2bb_shift); - qd->qd_qb.qb_warn = qp->qu_warn; + q.qu_warn = cpu_to_be64(fdq->d_blk_softlimit >> sdp->sd_fsb2bb_shift); + qd->qd_qb.qb_warn = q.qu_warn; } if (fdq->d_fieldmask & FS_DQ_BHARD) { - qp->qu_limit = cpu_to_be64(fdq->d_blk_hardlimit >> sdp->sd_fsb2bb_shift); - qd->qd_qb.qb_limit = qp->qu_limit; + q.qu_limit = cpu_to_be64(fdq->d_blk_hardlimit >> sdp->sd_fsb2bb_shift); + qd->qd_qb.qb_limit = q.qu_limit; } if (fdq->d_fieldmask & FS_DQ_BCOUNT) { - qp->qu_value = cpu_to_be64(fdq->d_bcount >> sdp->sd_fsb2bb_shift); - qd->qd_qb.qb_value = qp->qu_value; + q.qu_value = cpu_to_be64(fdq->d_bcount >> sdp->sd_fsb2bb_shift); + qd->qd_qb.qb_value = q.qu_value; } } /* Write the quota into the quota file on disk */ - ptr = qp; + ptr = &q; nbytes = sizeof(struct gfs2_quota); get_a_page: - page = grab_cache_page(mapping, index); + page = find_or_create_page(mapping, index, GFP_NOFS); if (!page) return -ENOMEM; @@ -707,20 +746,20 @@ get_a_page: set_buffer_uptodate(bh); if (!buffer_uptodate(bh)) { - ll_rw_block(READ_META, 1, &bh); + ll_rw_block(READ | REQ_META, 1, &bh); wait_on_buffer(bh); if (!buffer_uptodate(bh)) goto unlock_out; } - gfs2_trans_add_bh(ip->i_gl, bh, 0); + gfs2_trans_add_data(ip->i_gl, bh); - kaddr = kmap_atomic(page, KM_USER0); + kaddr = kmap_atomic(page); if (offset + sizeof(struct gfs2_quota) > PAGE_CACHE_SIZE) nbytes = PAGE_CACHE_SIZE - offset; memcpy(kaddr + offset, ptr, nbytes); flush_dcache_page(page); - kunmap_atomic(kaddr, KM_USER0); + kunmap_atomic(kaddr); unlock_page(page); page_cache_release(page); @@ -734,22 +773,14 @@ get_a_page: goto get_a_page; } - /* Update the disk inode timestamp and size (if extended) */ - err = gfs2_meta_inode_buffer(ip, &dibh); - if (err) - goto out; - size = loc + sizeof(struct gfs2_quota); if (size > inode->i_size) i_size_write(inode, size); inode->i_mtime = inode->i_atime = CURRENT_TIME; - gfs2_trans_add_bh(ip->i_gl, dibh, 1); - gfs2_dinode_out(ip, dibh->b_data); - brelse(dibh); mark_inode_dirty(inode); + set_bit(QDF_REFRESH, &qd->qd_flags); + return 0; -out: - return err; unlock_out: unlock_page(page); page_cache_release(page); @@ -760,15 +791,20 @@ static int do_sync(unsigned int num_qd, struct gfs2_quota_data **qda) { struct gfs2_sbd *sdp = (*qda)->qd_gl->gl_sbd; struct gfs2_inode *ip = GFS2_I(sdp->sd_quota_inode); + struct gfs2_alloc_parms ap = { .aflags = 0, }; unsigned int data_blocks, ind_blocks; struct gfs2_holder *ghs, i_gh; unsigned int qx, x; struct gfs2_quota_data *qd; + unsigned reserved; loff_t offset; unsigned int nalloc = 0, blocks; - struct gfs2_alloc *al = NULL; int error; + error = gfs2_rs_alloc(ip); + if (error) + return error; + gfs2_write_calc_reserv(ip, sizeof(struct gfs2_quota), &data_blocks, &ind_blocks); @@ -777,7 +813,7 @@ static int do_sync(unsigned int num_qd, struct gfs2_quota_data **qda) return -ENOMEM; sort(qda, num_qd, sizeof(struct gfs2_quota_data *), sort_qd, NULL); - mutex_lock_nested(&ip->i_inode.i_mutex, I_MUTEX_QUOTA); + mutex_lock(&ip->i_inode.i_mutex); for (qx = 0; qx < num_qd; qx++) { error = gfs2_glock_nq_init(qda[qx]->qd_gl, LM_ST_EXCLUSIVE, GL_NOCACHE, &ghs[qx]); @@ -796,31 +832,25 @@ static int do_sync(unsigned int num_qd, struct gfs2_quota_data **qda) nalloc++; } - al = gfs2_alloc_get(ip); - if (!al) { - error = -ENOMEM; - goto out_gunlock; - } /* * 1 blk for unstuffing inode if stuffed. We add this extra * block to the reservation unconditionally. If the inode * doesn't need unstuffing, the block will be released to the * rgrp since it won't be allocated during the transaction */ - al->al_requested = 1; /* +3 in the end for unstuffing block, inode size update block * and another block in case quota straddles page boundary and * two blocks need to be updated instead of 1 */ blocks = num_qd * data_blocks + RES_DINODE + num_qd + 3; - if (nalloc) - al->al_requested += nalloc * (data_blocks + ind_blocks); - error = gfs2_inplace_reserve(ip); + reserved = 1 + (nalloc * (data_blocks + ind_blocks)); + ap.target = reserved; + error = gfs2_inplace_reserve(ip, &ap); if (error) goto out_alloc; if (nalloc) - blocks += gfs2_rg_blocks(al) + nalloc * ind_blocks + RES_STATFS; + blocks += gfs2_rg_blocks(ip, reserved) + nalloc * ind_blocks + RES_STATFS; error = gfs2_trans_begin(sdp, blocks, 0); if (error) @@ -834,6 +864,7 @@ static int do_sync(unsigned int num_qd, struct gfs2_quota_data **qda) goto out_end_trans; do_qc(qd, -qd->qd_change_sync); + set_bit(QDF_REFRESH, &qd->qd_flags); } error = 0; @@ -843,15 +874,13 @@ out_end_trans: out_ipres: gfs2_inplace_release(ip); out_alloc: - gfs2_alloc_put(ip); -out_gunlock: gfs2_glock_dq_uninit(&i_gh); out: while (qx--) gfs2_glock_dq_uninit(&ghs[qx]); mutex_unlock(&ip->i_inode.i_mutex); kfree(ghs); - gfs2_log_flush(ip->i_gl->gl_sbd, ip->i_gl); + gfs2_log_flush(ip->i_gl->gl_sbd, ip->i_gl, NORMAL_FLUSH); return error; } @@ -865,11 +894,11 @@ static int update_qd(struct gfs2_sbd *sdp, struct gfs2_quota_data *qd) memset(&q, 0, sizeof(struct gfs2_quota)); pos = qd2offset(qd); - error = gfs2_internal_read(ip, NULL, (char *)&q, &pos, sizeof(q)); + error = gfs2_internal_read(ip, (char *)&q, &pos, sizeof(q)); if (error < 0) return error; - qlvb = (struct gfs2_quota_lvb *)qd->qd_gl->gl_lvb; + qlvb = (struct gfs2_quota_lvb *)qd->qd_gl->gl_lksb.sb_lvbptr; qlvb->qb_magic = cpu_to_be32(GFS2_MAGIC); qlvb->__pad = 0; qlvb->qb_limit = q.qu_limit; @@ -893,7 +922,7 @@ restart: if (error) return error; - qd->qd_qb = *(struct gfs2_quota_lvb *)qd->qd_gl->gl_lvb; + qd->qd_qb = *(struct gfs2_quota_lvb *)qd->qd_gl->gl_lksb.sb_lvbptr; if (force_refresh || qd->qd_qb.qb_magic != cpu_to_be32(GFS2_MAGIC)) { gfs2_glock_dq_uninit(q_gh); @@ -925,24 +954,30 @@ fail: return error; } -int gfs2_quota_lock(struct gfs2_inode *ip, u32 uid, u32 gid) +int gfs2_quota_lock(struct gfs2_inode *ip, kuid_t uid, kgid_t gid) { struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); - struct gfs2_alloc *al = ip->i_alloc; + struct gfs2_quota_data *qd; unsigned int x; int error = 0; - gfs2_quota_hold(ip, uid, gid); + error = gfs2_quota_hold(ip, uid, gid); + if (error) + return error; if (capable(CAP_SYS_RESOURCE) || sdp->sd_args.ar_quota != GFS2_QUOTA_ON) return 0; - sort(al->al_qd, al->al_qd_num, sizeof(struct gfs2_quota_data *), - sort_qd, NULL); + sort(ip->i_res->rs_qa_qd, ip->i_res->rs_qa_qd_num, + sizeof(struct gfs2_quota_data *), sort_qd, NULL); - for (x = 0; x < al->al_qd_num; x++) { - error = do_glock(al->al_qd[x], NO_FORCE, &al->al_qd_ghs[x]); + for (x = 0; x < ip->i_res->rs_qa_qd_num; x++) { + int force = NO_FORCE; + qd = ip->i_res->rs_qa_qd[x]; + if (test_and_clear_bit(QDF_REFRESH, &qd->qd_flags)) + force = FORCE; + error = do_glock(qd, force, &ip->i_res->rs_qa_qd_ghs[x]); if (error) break; } @@ -951,7 +986,7 @@ int gfs2_quota_lock(struct gfs2_inode *ip, u32 uid, u32 gid) set_bit(GIF_QD_LOCKED, &ip->i_flags); else { while (x--) - gfs2_glock_dq_uninit(&al->al_qd_ghs[x]); + gfs2_glock_dq_uninit(&ip->i_res->rs_qa_qd_ghs[x]); gfs2_quota_unhold(ip); } @@ -969,9 +1004,9 @@ static int need_sync(struct gfs2_quota_data *qd) if (!qd->qd_qb.qb_limit) return 0; - spin_lock(&qd_lru_lock); + spin_lock(&qd_lock); value = qd->qd_change; - spin_unlock(&qd_lru_lock); + spin_unlock(&qd_lock); spin_lock(>->gt_spin); num = gt->gt_quota_scale_num; @@ -996,25 +1031,42 @@ static int need_sync(struct gfs2_quota_data *qd) void gfs2_quota_unlock(struct gfs2_inode *ip) { - struct gfs2_alloc *al = ip->i_alloc; + struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); struct gfs2_quota_data *qda[4]; unsigned int count = 0; unsigned int x; + int found; if (!test_and_clear_bit(GIF_QD_LOCKED, &ip->i_flags)) goto out; - for (x = 0; x < al->al_qd_num; x++) { + for (x = 0; x < ip->i_res->rs_qa_qd_num; x++) { struct gfs2_quota_data *qd; int sync; - qd = al->al_qd[x]; + qd = ip->i_res->rs_qa_qd[x]; sync = need_sync(qd); - gfs2_glock_dq_uninit(&al->al_qd_ghs[x]); + gfs2_glock_dq_uninit(&ip->i_res->rs_qa_qd_ghs[x]); + if (!sync) + continue; - if (sync && qd_trylock(qd)) - qda[count++] = qd; + spin_lock(&qd_lock); + found = qd_check_sync(sdp, qd, NULL); + spin_unlock(&qd_lock); + + if (!found) + continue; + + gfs2_assert_warn(sdp, qd->qd_change_sync); + if (bh_get(qd)) { + clear_bit(QDF_LOCKED, &qd->qd_flags); + slot_put(qd); + qd_put(qd); + continue; + } + + qda[count++] = qd; } if (count) { @@ -1033,18 +1085,17 @@ static int print_message(struct gfs2_quota_data *qd, char *type) { struct gfs2_sbd *sdp = qd->qd_gl->gl_sbd; - printk(KERN_INFO "GFS2: fsid=%s: quota %s for %s %u\n", - sdp->sd_fsname, type, - (test_bit(QDF_USER, &qd->qd_flags)) ? "user" : "group", - qd->qd_id); + fs_info(sdp, "quota %s for %s %u\n", + type, + (qd->qd_id.type == USRQUOTA) ? "user" : "group", + from_kqid(&init_user_ns, qd->qd_id)); return 0; } -int gfs2_quota_check(struct gfs2_inode *ip, u32 uid, u32 gid) +int gfs2_quota_check(struct gfs2_inode *ip, kuid_t uid, kgid_t gid) { struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); - struct gfs2_alloc *al = ip->i_alloc; struct gfs2_quota_data *qd; s64 value; unsigned int x; @@ -1056,22 +1107,21 @@ int gfs2_quota_check(struct gfs2_inode *ip, u32 uid, u32 gid) if (sdp->sd_args.ar_quota != GFS2_QUOTA_ON) return 0; - for (x = 0; x < al->al_qd_num; x++) { - qd = al->al_qd[x]; + for (x = 0; x < ip->i_res->rs_qa_qd_num; x++) { + qd = ip->i_res->rs_qa_qd[x]; - if (!((qd->qd_id == uid && test_bit(QDF_USER, &qd->qd_flags)) || - (qd->qd_id == gid && !test_bit(QDF_USER, &qd->qd_flags)))) + if (!(qid_eq(qd->qd_id, make_kqid_uid(uid)) || + qid_eq(qd->qd_id, make_kqid_gid(gid)))) continue; value = (s64)be64_to_cpu(qd->qd_qb.qb_value); - spin_lock(&qd_lru_lock); + spin_lock(&qd_lock); value += qd->qd_change; - spin_unlock(&qd_lru_lock); + spin_unlock(&qd_lock); if (be64_to_cpu(qd->qd_qb.qb_limit) && (s64)be64_to_cpu(qd->qd_qb.qb_limit) < value) { print_message(qd, "exceeded"); - quota_send_warning(test_bit(QDF_USER, &qd->qd_flags) ? - USRQUOTA : GRPQUOTA, qd->qd_id, + quota_send_warning(qd->qd_id, sdp->sd_vfs->s_dev, QUOTA_NL_BHARDWARN); error = -EDQUOT; @@ -1081,8 +1131,7 @@ int gfs2_quota_check(struct gfs2_inode *ip, u32 uid, u32 gid) time_after_eq(jiffies, qd->qd_last_warn + gfs2_tune_get(sdp, gt_quota_warn_period) * HZ)) { - quota_send_warning(test_bit(QDF_USER, &qd->qd_flags) ? - USRQUOTA : GRPQUOTA, qd->qd_id, + quota_send_warning(qd->qd_id, sdp->sd_vfs->s_dev, QUOTA_NL_BSOFTWARN); error = print_message(qd, "warning"); qd->qd_last_warn = jiffies; @@ -1093,9 +1142,8 @@ int gfs2_quota_check(struct gfs2_inode *ip, u32 uid, u32 gid) } void gfs2_quota_change(struct gfs2_inode *ip, s64 change, - u32 uid, u32 gid) + kuid_t uid, kgid_t gid) { - struct gfs2_alloc *al = ip->i_alloc; struct gfs2_quota_data *qd; unsigned int x; @@ -1104,31 +1152,32 @@ void gfs2_quota_change(struct gfs2_inode *ip, s64 change, if (ip->i_diskflags & GFS2_DIF_SYSTEM) return; - for (x = 0; x < al->al_qd_num; x++) { - qd = al->al_qd[x]; + for (x = 0; x < ip->i_res->rs_qa_qd_num; x++) { + qd = ip->i_res->rs_qa_qd[x]; - if ((qd->qd_id == uid && test_bit(QDF_USER, &qd->qd_flags)) || - (qd->qd_id == gid && !test_bit(QDF_USER, &qd->qd_flags))) { + if (qid_eq(qd->qd_id, make_kqid_uid(uid)) || + qid_eq(qd->qd_id, make_kqid_gid(gid))) { do_qc(qd, change); } } } -int gfs2_quota_sync(struct super_block *sb, int type, int wait) +int gfs2_quota_sync(struct super_block *sb, int type) { struct gfs2_sbd *sdp = sb->s_fs_info; struct gfs2_quota_data **qda; - unsigned int max_qd = gfs2_tune_get(sdp, gt_quota_simul_sync); + unsigned int max_qd = PAGE_SIZE/sizeof(struct gfs2_holder); unsigned int num_qd; unsigned int x; int error = 0; - sdp->sd_quota_sync_gen++; - qda = kcalloc(max_qd, sizeof(struct gfs2_quota_data *), GFP_KERNEL); if (!qda) return -ENOMEM; + mutex_lock(&sdp->sd_quota_sync_mutex); + sdp->sd_quota_sync_gen++; + do { num_qd = 0; @@ -1153,23 +1202,19 @@ int gfs2_quota_sync(struct super_block *sb, int type, int wait) } } while (!error && num_qd == max_qd); + mutex_unlock(&sdp->sd_quota_sync_mutex); kfree(qda); return error; } -static int gfs2_quota_sync_timeo(struct super_block *sb, int type) -{ - return gfs2_quota_sync(sb, type, 0); -} - -int gfs2_quota_refresh(struct gfs2_sbd *sdp, int user, u32 id) +int gfs2_quota_refresh(struct gfs2_sbd *sdp, struct kqid qid) { struct gfs2_quota_data *qd; struct gfs2_holder q_gh; int error; - error = qd_get(sdp, user, id, &qd); + error = qd_get(sdp, qid, &qd); if (error) return error; @@ -1181,15 +1226,6 @@ int gfs2_quota_refresh(struct gfs2_sbd *sdp, int user, u32 id) return error; } -static void gfs2_quota_change_in(struct gfs2_quota_change_host *qc, const void *buf) -{ - const struct gfs2_quota_change *str = buf; - - qc->qc_change = be64_to_cpu(str->qc_change); - qc->qc_flags = be32_to_cpu(str->qc_flags); - qc->qc_id = be32_to_cpu(str->qc_id); -} - int gfs2_quota_init(struct gfs2_sbd *sdp) { struct gfs2_inode *ip = GFS2_I(sdp->sd_qc_inode); @@ -1197,6 +1233,8 @@ int gfs2_quota_init(struct gfs2_sbd *sdp) unsigned int blocks = size >> sdp->sd_sb.sb_bsize_shift; unsigned int x, slot = 0; unsigned int found = 0; + unsigned int hash; + unsigned int bm_size; u64 dblock; u32 extlen = 0; int error; @@ -1205,23 +1243,19 @@ int gfs2_quota_init(struct gfs2_sbd *sdp) return -EIO; sdp->sd_quota_slots = blocks * sdp->sd_qc_per_block; - sdp->sd_quota_chunks = DIV_ROUND_UP(sdp->sd_quota_slots, 8 * PAGE_SIZE); - + bm_size = DIV_ROUND_UP(sdp->sd_quota_slots, 8 * sizeof(unsigned long)); + bm_size *= sizeof(unsigned long); error = -ENOMEM; - - sdp->sd_quota_bitmap = kcalloc(sdp->sd_quota_chunks, - sizeof(unsigned char *), GFP_NOFS); + sdp->sd_quota_bitmap = kzalloc(bm_size, GFP_NOFS | __GFP_NOWARN); + if (sdp->sd_quota_bitmap == NULL) + sdp->sd_quota_bitmap = __vmalloc(bm_size, GFP_NOFS | + __GFP_ZERO, PAGE_KERNEL); if (!sdp->sd_quota_bitmap) return error; - for (x = 0; x < sdp->sd_quota_chunks; x++) { - sdp->sd_quota_bitmap[x] = kzalloc(PAGE_SIZE, GFP_NOFS); - if (!sdp->sd_quota_bitmap[x]) - goto fail; - } - for (x = 0; x < blocks; x++) { struct buffer_head *bh; + const struct gfs2_quota_change *qc; unsigned int y; if (!extlen) { @@ -1239,34 +1273,41 @@ int gfs2_quota_init(struct gfs2_sbd *sdp) goto fail; } + qc = (const struct gfs2_quota_change *)(bh->b_data + sizeof(struct gfs2_meta_header)); for (y = 0; y < sdp->sd_qc_per_block && slot < sdp->sd_quota_slots; y++, slot++) { - struct gfs2_quota_change_host qc; struct gfs2_quota_data *qd; - - gfs2_quota_change_in(&qc, bh->b_data + - sizeof(struct gfs2_meta_header) + - y * sizeof(struct gfs2_quota_change)); - if (!qc.qc_change) + s64 qc_change = be64_to_cpu(qc->qc_change); + u32 qc_flags = be32_to_cpu(qc->qc_flags); + enum quota_type qtype = (qc_flags & GFS2_QCF_USER) ? + USRQUOTA : GRPQUOTA; + struct kqid qc_id = make_kqid(&init_user_ns, qtype, + be32_to_cpu(qc->qc_id)); + qc++; + if (!qc_change) continue; - error = qd_alloc(sdp, (qc.qc_flags & GFS2_QCF_USER), - qc.qc_id, &qd); - if (error) { + hash = gfs2_qd_hash(sdp, qc_id); + qd = qd_alloc(hash, sdp, qc_id); + if (qd == NULL) { brelse(bh); goto fail; } set_bit(QDF_CHANGE, &qd->qd_flags); - qd->qd_change = qc.qc_change; + qd->qd_change = qc_change; qd->qd_slot = slot; qd->qd_slot_count = 1; - spin_lock(&qd_lru_lock); - gfs2_icbit_munge(sdp, sdp->sd_quota_bitmap, slot, 1); + spin_lock(&qd_lock); + BUG_ON(test_and_set_bit(slot, sdp->sd_quota_bitmap)); list_add(&qd->qd_list, &sdp->sd_quota_list); atomic_inc(&sdp->sd_quota_count); - spin_unlock(&qd_lru_lock); + spin_unlock(&qd_lock); + + spin_lock_bucket(hash); + hlist_bl_add_head_rcu(&qd->qd_hlist, &qd_hash_table[hash]); + spin_unlock_bucket(hash); found++; } @@ -1290,51 +1331,41 @@ void gfs2_quota_cleanup(struct gfs2_sbd *sdp) { struct list_head *head = &sdp->sd_quota_list; struct gfs2_quota_data *qd; - unsigned int x; - spin_lock(&qd_lru_lock); + spin_lock(&qd_lock); while (!list_empty(head)) { qd = list_entry(head->prev, struct gfs2_quota_data, qd_list); - if (atomic_read(&qd->qd_count) > 1 || - (atomic_read(&qd->qd_count) && - !test_bit(QDF_CHANGE, &qd->qd_flags))) { - list_move(&qd->qd_list, head); - spin_unlock(&qd_lru_lock); - schedule(); - spin_lock(&qd_lru_lock); - continue; - } - list_del(&qd->qd_list); + /* Also remove if this qd exists in the reclaim list */ - if (!list_empty(&qd->qd_reclaim)) { - list_del_init(&qd->qd_reclaim); - atomic_dec(&qd_lru_count); - } + list_lru_del(&gfs2_qd_lru, &qd->qd_lru); atomic_dec(&sdp->sd_quota_count); - spin_unlock(&qd_lru_lock); + spin_unlock(&qd_lock); + + spin_lock_bucket(qd->qd_hash); + hlist_bl_del_rcu(&qd->qd_hlist); + spin_unlock_bucket(qd->qd_hash); - if (!atomic_read(&qd->qd_count)) { - gfs2_assert_warn(sdp, !qd->qd_change); - gfs2_assert_warn(sdp, !qd->qd_slot_count); - } else - gfs2_assert_warn(sdp, qd->qd_slot_count == 1); + gfs2_assert_warn(sdp, !qd->qd_change); + gfs2_assert_warn(sdp, !qd->qd_slot_count); gfs2_assert_warn(sdp, !qd->qd_bh_count); gfs2_glock_put(qd->qd_gl); - kmem_cache_free(gfs2_quotad_cachep, qd); + call_rcu(&qd->qd_rcu, gfs2_qd_dealloc); - spin_lock(&qd_lru_lock); + spin_lock(&qd_lock); } - spin_unlock(&qd_lru_lock); + spin_unlock(&qd_lock); gfs2_assert_warn(sdp, !atomic_read(&sdp->sd_quota_count)); if (sdp->sd_quota_bitmap) { - for (x = 0; x < sdp->sd_quota_chunks; x++) - kfree(sdp->sd_quota_bitmap[x]); - kfree(sdp->sd_quota_bitmap); + if (is_vmalloc_addr(sdp->sd_quota_bitmap)) + vfree(sdp->sd_quota_bitmap); + else + kfree(sdp->sd_quota_bitmap); + sdp->sd_quota_bitmap = NULL; } } @@ -1417,14 +1448,14 @@ int gfs2_quotad(void *data) &tune->gt_statfs_quantum); /* Update quota file */ - quotad_check_timeo(sdp, "sync", gfs2_quota_sync_timeo, t, + quotad_check_timeo(sdp, "sync", gfs2_quota_sync, t, "ad_timeo, &tune->gt_quota_quantum); /* Check for & recover partially truncated inodes */ quotad_check_trunc_list(sdp); - if (freezing(current)) - refrigerator(); + try_to_freeze(); + t = min(quotad_timeo, statfs_timeo); prepare_to_wait(&sdp->sd_quota_wait, &wait, TASK_INTERRUPTIBLE); @@ -1466,11 +1497,11 @@ static int gfs2_quota_get_xstate(struct super_block *sb, } fqs->qs_uquota.qfs_nextents = 1; /* unsupported */ fqs->qs_gquota = fqs->qs_uquota; /* its the same inode in both cases */ - fqs->qs_incoredqs = atomic_read(&qd_lru_count); + fqs->qs_incoredqs = list_lru_count(&gfs2_qd_lru); return 0; } -static int gfs2_get_dqblk(struct super_block *sb, int type, qid_t id, +static int gfs2_get_dqblk(struct super_block *sb, struct kqid qid, struct fs_disk_quota *fdq) { struct gfs2_sbd *sdp = sb->s_fs_info; @@ -1484,24 +1515,21 @@ static int gfs2_get_dqblk(struct super_block *sb, int type, qid_t id, if (sdp->sd_args.ar_quota == GFS2_QUOTA_OFF) return -ESRCH; /* Crazy XFS error code */ - if (type == USRQUOTA) - type = QUOTA_USER; - else if (type == GRPQUOTA) - type = QUOTA_GROUP; - else + if ((qid.type != USRQUOTA) && + (qid.type != GRPQUOTA)) return -EINVAL; - error = qd_get(sdp, type, id, &qd); + error = qd_get(sdp, qid, &qd); if (error) return error; error = do_glock(qd, FORCE, &q_gh); if (error) goto out; - qlvb = (struct gfs2_quota_lvb *)qd->qd_gl->gl_lvb; + qlvb = (struct gfs2_quota_lvb *)qd->qd_gl->gl_lksb.sb_lvbptr; fdq->d_version = FS_DQUOT_VERSION; - fdq->d_flags = (type == QUOTA_USER) ? FS_USER_QUOTA : FS_GROUP_QUOTA; - fdq->d_id = id; + fdq->d_flags = (qid.type == USRQUOTA) ? FS_USER_QUOTA : FS_GROUP_QUOTA; + fdq->d_id = from_kqid_munged(current_user_ns(), qid); fdq->d_blk_hardlimit = be64_to_cpu(qlvb->qb_limit) << sdp->sd_fsb2bb_shift; fdq->d_blk_softlimit = be64_to_cpu(qlvb->qb_warn) << sdp->sd_fsb2bb_shift; fdq->d_bcount = be64_to_cpu(qlvb->qb_value) << sdp->sd_fsb2bb_shift; @@ -1515,7 +1543,7 @@ out: /* GFS2 only supports a subset of the XFS fields */ #define GFS2_FIELDMASK (FS_DQ_BSOFT|FS_DQ_BHARD|FS_DQ_BCOUNT) -static int gfs2_set_dqblk(struct super_block *sb, int type, qid_t id, +static int gfs2_set_dqblk(struct super_block *sb, struct kqid qid, struct fs_disk_quota *fdq) { struct gfs2_sbd *sdp = sb->s_fs_info; @@ -1525,41 +1553,31 @@ static int gfs2_set_dqblk(struct super_block *sb, int type, qid_t id, unsigned int data_blocks, ind_blocks; unsigned int blocks = 0; int alloc_required; - struct gfs2_alloc *al; loff_t offset; int error; if (sdp->sd_args.ar_quota == GFS2_QUOTA_OFF) return -ESRCH; /* Crazy XFS error code */ - switch(type) { - case USRQUOTA: - type = QUOTA_USER; - if (fdq->d_flags != FS_USER_QUOTA) - return -EINVAL; - break; - case GRPQUOTA: - type = QUOTA_GROUP; - if (fdq->d_flags != FS_GROUP_QUOTA) - return -EINVAL; - break; - default: + if ((qid.type != USRQUOTA) && + (qid.type != GRPQUOTA)) return -EINVAL; - } if (fdq->d_fieldmask & ~GFS2_FIELDMASK) return -EINVAL; - if (fdq->d_id != id) - return -EINVAL; - error = qd_get(sdp, type, id, &qd); + error = qd_get(sdp, qid, &qd); if (error) return error; + error = gfs2_rs_alloc(ip); + if (error) + goto out_put; + mutex_lock(&ip->i_inode.i_mutex); error = gfs2_glock_nq_init(qd->qd_gl, LM_ST_EXCLUSIVE, 0, &q_gh); if (error) - goto out_put; + goto out_unlockput; error = gfs2_glock_nq_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, &i_gh); if (error) goto out_q; @@ -1587,20 +1605,23 @@ static int gfs2_set_dqblk(struct super_block *sb, int type, qid_t id, offset = qd2offset(qd); alloc_required = gfs2_write_alloc_required(ip, offset, sizeof(struct gfs2_quota)); + if (gfs2_is_stuffed(ip)) + alloc_required = 1; if (alloc_required) { - al = gfs2_alloc_get(ip); - if (al == NULL) - goto out_i; + struct gfs2_alloc_parms ap = { .aflags = 0, }; gfs2_write_calc_reserv(ip, sizeof(struct gfs2_quota), &data_blocks, &ind_blocks); - blocks = al->al_requested = 1 + data_blocks + ind_blocks; - error = gfs2_inplace_reserve(ip); + blocks = 1 + data_blocks + ind_blocks; + ap.target = blocks; + error = gfs2_inplace_reserve(ip, &ap); if (error) - goto out_alloc; - blocks += gfs2_rg_blocks(al); + goto out_i; + blocks += gfs2_rg_blocks(ip, blocks); } - error = gfs2_trans_begin(sdp, blocks + RES_DINODE + 1, 0); + /* Some quotas span block boundaries and can update two blocks, + adding an extra block to the transaction to handle such quotas */ + error = gfs2_trans_begin(sdp, blocks + RES_DINODE + 2, 0); if (error) goto out_release; @@ -1609,17 +1630,15 @@ static int gfs2_set_dqblk(struct super_block *sb, int type, qid_t id, gfs2_trans_end(sdp); out_release: - if (alloc_required) { + if (alloc_required) gfs2_inplace_release(ip); -out_alloc: - gfs2_alloc_put(ip); - } out_i: gfs2_glock_dq_uninit(&i_gh); out_q: gfs2_glock_dq_uninit(&q_gh); -out_put: +out_unlockput: mutex_unlock(&ip->i_inode.i_mutex); +out_put: qd_put(qd); return error; } @@ -1630,3 +1649,11 @@ const struct quotactl_ops gfs2_quotactl_ops = { .get_dqblk = gfs2_get_dqblk, .set_dqblk = gfs2_set_dqblk, }; + +void __init gfs2_quota_hash_init(void) +{ + unsigned i; + + for(i = 0; i < GFS2_QD_HASH_SIZE; i++) + INIT_HLIST_BL_HEAD(&qd_hash_table[i]); +} diff --git a/fs/gfs2/quota.h b/fs/gfs2/quota.h index e7d236ca48b..55d506eb3c4 100644 --- a/fs/gfs2/quota.h +++ b/fs/gfs2/quota.h @@ -10,23 +10,26 @@ #ifndef __QUOTA_DOT_H__ #define __QUOTA_DOT_H__ +#include <linux/list_lru.h> + struct gfs2_inode; struct gfs2_sbd; -#define NO_QUOTA_CHANGE ((u32)-1) +#define NO_UID_QUOTA_CHANGE INVALID_UID +#define NO_GID_QUOTA_CHANGE INVALID_GID -extern int gfs2_quota_hold(struct gfs2_inode *ip, u32 uid, u32 gid); +extern int gfs2_quota_hold(struct gfs2_inode *ip, kuid_t uid, kgid_t gid); extern void gfs2_quota_unhold(struct gfs2_inode *ip); -extern int gfs2_quota_lock(struct gfs2_inode *ip, u32 uid, u32 gid); +extern int gfs2_quota_lock(struct gfs2_inode *ip, kuid_t uid, kgid_t gid); extern void gfs2_quota_unlock(struct gfs2_inode *ip); -extern int gfs2_quota_check(struct gfs2_inode *ip, u32 uid, u32 gid); +extern int gfs2_quota_check(struct gfs2_inode *ip, kuid_t uid, kgid_t gid); extern void gfs2_quota_change(struct gfs2_inode *ip, s64 change, - u32 uid, u32 gid); + kuid_t uid, kgid_t gid); -extern int gfs2_quota_sync(struct super_block *sb, int type, int wait); -extern int gfs2_quota_refresh(struct gfs2_sbd *sdp, int user, u32 id); +extern int gfs2_quota_sync(struct super_block *sb, int type); +extern int gfs2_quota_refresh(struct gfs2_sbd *sdp, struct kqid qid); extern int gfs2_quota_init(struct gfs2_sbd *sdp); extern void gfs2_quota_cleanup(struct gfs2_sbd *sdp); @@ -40,7 +43,7 @@ static inline int gfs2_quota_lock_check(struct gfs2_inode *ip) int ret; if (sdp->sd_args.ar_quota == GFS2_QUOTA_OFF) return 0; - ret = gfs2_quota_lock(ip, NO_QUOTA_CHANGE, NO_QUOTA_CHANGE); + ret = gfs2_quota_lock(ip, NO_UID_QUOTA_CHANGE, NO_GID_QUOTA_CHANGE); if (ret) return ret; if (sdp->sd_args.ar_quota != GFS2_QUOTA_ON) @@ -51,7 +54,9 @@ static inline int gfs2_quota_lock_check(struct gfs2_inode *ip) return ret; } -extern int gfs2_shrink_qd_memory(struct shrinker *shrink, int nr, gfp_t gfp_mask); extern const struct quotactl_ops gfs2_quotactl_ops; +extern struct shrinker gfs2_qd_shrinker; +extern struct list_lru gfs2_qd_lru; +extern void __init gfs2_quota_hash_init(void); #endif /* __QUOTA_DOT_H__ */ diff --git a/fs/gfs2/recovery.c b/fs/gfs2/recovery.c index f2a02edcac8..94555d4c569 100644 --- a/fs/gfs2/recovery.c +++ b/fs/gfs2/recovery.c @@ -52,9 +52,9 @@ int gfs2_replay_read_block(struct gfs2_jdesc *jd, unsigned int blk, return error; } -int gfs2_revoke_add(struct gfs2_sbd *sdp, u64 blkno, unsigned int where) +int gfs2_revoke_add(struct gfs2_jdesc *jd, u64 blkno, unsigned int where) { - struct list_head *head = &sdp->sd_revoke_list; + struct list_head *head = &jd->jd_revoke_list; struct gfs2_revoke_replay *rr; int found = 0; @@ -81,13 +81,13 @@ int gfs2_revoke_add(struct gfs2_sbd *sdp, u64 blkno, unsigned int where) return 1; } -int gfs2_revoke_check(struct gfs2_sbd *sdp, u64 blkno, unsigned int where) +int gfs2_revoke_check(struct gfs2_jdesc *jd, u64 blkno, unsigned int where) { struct gfs2_revoke_replay *rr; int wrap, a, b, revoke; int found = 0; - list_for_each_entry(rr, &sdp->sd_revoke_list, rr_list) { + list_for_each_entry(rr, &jd->jd_revoke_list, rr_list) { if (rr->rr_blkno == blkno) { found = 1; break; @@ -97,17 +97,17 @@ int gfs2_revoke_check(struct gfs2_sbd *sdp, u64 blkno, unsigned int where) if (!found) return 0; - wrap = (rr->rr_where < sdp->sd_replay_tail); - a = (sdp->sd_replay_tail < where); + wrap = (rr->rr_where < jd->jd_replay_tail); + a = (jd->jd_replay_tail < where); b = (where < rr->rr_where); revoke = (wrap) ? (a || b) : (a && b); return revoke; } -void gfs2_revoke_clean(struct gfs2_sbd *sdp) +void gfs2_revoke_clean(struct gfs2_jdesc *jd) { - struct list_head *head = &sdp->sd_revoke_list; + struct list_head *head = &jd->jd_revoke_list; struct gfs2_revoke_replay *rr; while (!list_empty(head)) { @@ -436,12 +436,16 @@ static void gfs2_recovery_done(struct gfs2_sbd *sdp, unsigned int jid, char env_status[20]; char *envp[] = { env_jid, env_status, NULL }; struct lm_lockstruct *ls = &sdp->sd_lockstruct; + ls->ls_recover_jid_done = jid; ls->ls_recover_jid_status = message; sprintf(env_jid, "JID=%d", jid); sprintf(env_status, "RECOVERY=%s", message == LM_RD_SUCCESS ? "Done" : "Failed"); kobject_uevent_env(&sdp->sd_kobj, KOBJ_CHANGE, envp); + + if (sdp->sd_lockstruct.ls_ops->lm_recovery_result) + sdp->sd_lockstruct.ls_ops->lm_recovery_result(sdp, jid, message); } void gfs2_recover_func(struct work_struct *work) @@ -450,7 +454,7 @@ void gfs2_recover_func(struct work_struct *work) struct gfs2_inode *ip = GFS2_I(jd->jd_inode); struct gfs2_sbd *sdp = GFS2_SB(jd->jd_inode); struct gfs2_log_header_host head; - struct gfs2_holder j_gh, ji_gh, t_gh; + struct gfs2_holder j_gh, ji_gh, thaw_gh; unsigned long t; int ro = 0; unsigned int pass; @@ -504,15 +508,17 @@ void gfs2_recover_func(struct work_struct *work) t = jiffies; - /* Acquire a shared hold on the transaction lock */ + /* Acquire a shared hold on the freeze lock */ - error = gfs2_glock_nq_init(sdp->sd_trans_gl, LM_ST_SHARED, - LM_FLAG_NOEXP | LM_FLAG_PRIORITY | - GL_NOCACHE, &t_gh); + error = gfs2_glock_nq_init(sdp->sd_freeze_gl, LM_ST_SHARED, + LM_FLAG_NOEXP | LM_FLAG_PRIORITY, + &thaw_gh); if (error) goto fail_gunlock_ji; - if (test_bit(SDF_JOURNAL_CHECKED, &sdp->sd_flags)) { + if (test_bit(SDF_RORECOVERY, &sdp->sd_flags)) { + ro = 1; + } else if (test_bit(SDF_JOURNAL_CHECKED, &sdp->sd_flags)) { if (!test_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags)) ro = 1; } else { @@ -532,7 +538,7 @@ void gfs2_recover_func(struct work_struct *work) fs_warn(sdp, "jid=%u: Can't replay: read-only block " "device\n", jd->jd_jid); error = -EROFS; - goto fail_gunlock_tr; + goto fail_gunlock_thaw; } fs_info(sdp, "jid=%u: Replaying journal...\n", jd->jd_jid); @@ -543,14 +549,14 @@ void gfs2_recover_func(struct work_struct *work) head.lh_blkno, pass); lops_after_scan(jd, error, pass); if (error) - goto fail_gunlock_tr; + goto fail_gunlock_thaw; } error = clean_journal(jd, &head); if (error) - goto fail_gunlock_tr; + goto fail_gunlock_thaw; - gfs2_glock_dq_uninit(&t_gh); + gfs2_glock_dq_uninit(&thaw_gh); t = DIV_ROUND_UP(jiffies - t, HZ); fs_info(sdp, "jid=%u: Journal replayed in %lus\n", jd->jd_jid, t); @@ -566,8 +572,8 @@ void gfs2_recover_func(struct work_struct *work) fs_info(sdp, "jid=%u: Done\n", jd->jd_jid); goto done; -fail_gunlock_tr: - gfs2_glock_dq_uninit(&t_gh); +fail_gunlock_thaw: + gfs2_glock_dq_uninit(&thaw_gh); fail_gunlock_ji: if (jlocked) { gfs2_glock_dq_uninit(&ji_gh); @@ -577,10 +583,11 @@ fail_gunlock_j: fs_info(sdp, "jid=%u: %s\n", jd->jd_jid, (error) ? "Failed" : "Done"); fail: + jd->jd_recover_error = error; gfs2_recovery_done(sdp, jd->jd_jid, LM_RD_GAVEUP); done: clear_bit(JDF_RECOVERY, &jd->jd_flags); - smp_mb__after_clear_bit(); + smp_mb__after_atomic(); wake_up_bit(&jd->jd_flags, JDF_RECOVERY); } @@ -605,6 +612,6 @@ int gfs2_recover_journal(struct gfs2_jdesc *jd, bool wait) wait_on_bit(&jd->jd_flags, JDF_RECOVERY, gfs2_recovery_wait, TASK_UNINTERRUPTIBLE); - return 0; + return wait ? jd->jd_recover_error : 0; } diff --git a/fs/gfs2/recovery.h b/fs/gfs2/recovery.h index 2226136c764..6142836cce9 100644 --- a/fs/gfs2/recovery.h +++ b/fs/gfs2/recovery.h @@ -23,9 +23,9 @@ static inline void gfs2_replay_incr_blk(struct gfs2_sbd *sdp, unsigned int *blk) extern int gfs2_replay_read_block(struct gfs2_jdesc *jd, unsigned int blk, struct buffer_head **bh); -extern int gfs2_revoke_add(struct gfs2_sbd *sdp, u64 blkno, unsigned int where); -extern int gfs2_revoke_check(struct gfs2_sbd *sdp, u64 blkno, unsigned int where); -extern void gfs2_revoke_clean(struct gfs2_sbd *sdp); +extern int gfs2_revoke_add(struct gfs2_jdesc *jd, u64 blkno, unsigned int where); +extern int gfs2_revoke_check(struct gfs2_jdesc *jd, u64 blkno, unsigned int where); +extern void gfs2_revoke_clean(struct gfs2_jdesc *jd); extern int gfs2_find_jhead(struct gfs2_jdesc *jd, struct gfs2_log_header_host *head); diff --git a/fs/gfs2/rgrp.c b/fs/gfs2/rgrp.c index 7293ea27020..f4cb9c0d6bb 100644 --- a/fs/gfs2/rgrp.c +++ b/fs/gfs2/rgrp.c @@ -7,6 +7,8 @@ * of the GNU General Public License version 2. */ +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + #include <linux/slab.h> #include <linux/spinlock.h> #include <linux/completion.h> @@ -15,6 +17,8 @@ #include <linux/gfs2_ondisk.h> #include <linux/prefetch.h> #include <linux/blkdev.h> +#include <linux/rbtree.h> +#include <linux/random.h> #include "gfs2.h" #include "incore.h" @@ -55,6 +59,11 @@ * 3 = Used (metadata) */ +struct gfs2_extent { + struct gfs2_rbm rbm; + u32 len; +}; + static const char valid_change[16] = { /* current */ /* n */ 0, 1, 1, 1, @@ -63,42 +72,49 @@ static const char valid_change[16] = { 1, 0, 0, 0 }; -static u32 rgblk_search(struct gfs2_rgrpd *rgd, u32 goal, - unsigned char old_state, unsigned char new_state, - unsigned int *n); +static int gfs2_rbm_find(struct gfs2_rbm *rbm, u8 state, u32 *minext, + const struct gfs2_inode *ip, bool nowrap, + const struct gfs2_alloc_parms *ap); + /** * gfs2_setbit - Set a bit in the bitmaps - * @buffer: the buffer that holds the bitmaps - * @buflen: the length (in bytes) of the buffer - * @block: the block to set + * @rbm: The position of the bit to set + * @do_clone: Also set the clone bitmap, if it exists * @new_state: the new state of the block * */ -static inline void gfs2_setbit(struct gfs2_rgrpd *rgd, unsigned char *buf1, - unsigned char *buf2, unsigned int offset, - unsigned int buflen, u32 block, +static inline void gfs2_setbit(const struct gfs2_rbm *rbm, bool do_clone, unsigned char new_state) { unsigned char *byte1, *byte2, *end, cur_state; - const unsigned int bit = (block % GFS2_NBBY) * GFS2_BIT_SIZE; + struct gfs2_bitmap *bi = rbm_bi(rbm); + unsigned int buflen = bi->bi_len; + const unsigned int bit = (rbm->offset % GFS2_NBBY) * GFS2_BIT_SIZE; - byte1 = buf1 + offset + (block / GFS2_NBBY); - end = buf1 + offset + buflen; + byte1 = bi->bi_bh->b_data + bi->bi_offset + (rbm->offset / GFS2_NBBY); + end = bi->bi_bh->b_data + bi->bi_offset + buflen; BUG_ON(byte1 >= end); cur_state = (*byte1 >> bit) & GFS2_BIT_MASK; if (unlikely(!valid_change[new_state * 4 + cur_state])) { - gfs2_consist_rgrpd(rgd); + pr_warn("buf_blk = 0x%x old_state=%d, new_state=%d\n", + rbm->offset, cur_state, new_state); + pr_warn("rgrp=0x%llx bi_start=0x%x\n", + (unsigned long long)rbm->rgd->rd_addr, bi->bi_start); + pr_warn("bi_offset=0x%x bi_len=0x%x\n", + bi->bi_offset, bi->bi_len); + dump_stack(); + gfs2_consist_rgrpd(rbm->rgd); return; } *byte1 ^= (cur_state ^ new_state) << bit; - if (buf2) { - byte2 = buf2 + offset + (block / GFS2_NBBY); + if (do_clone && bi->bi_clone) { + byte2 = bi->bi_clone + bi->bi_offset + (rbm->offset / GFS2_NBBY); cur_state = (*byte2 >> bit) & GFS2_BIT_MASK; *byte2 ^= (cur_state ^ new_state) << bit; } @@ -106,29 +122,22 @@ static inline void gfs2_setbit(struct gfs2_rgrpd *rgd, unsigned char *buf1, /** * gfs2_testbit - test a bit in the bitmaps - * @buffer: the buffer that holds the bitmaps - * @buflen: the length (in bytes) of the buffer - * @block: the block to read + * @rbm: The bit to test * + * Returns: The two bit block state of the requested bit */ -static inline unsigned char gfs2_testbit(struct gfs2_rgrpd *rgd, - const unsigned char *buffer, - unsigned int buflen, u32 block) +static inline u8 gfs2_testbit(const struct gfs2_rbm *rbm) { - const unsigned char *byte, *end; - unsigned char cur_state; + struct gfs2_bitmap *bi = rbm_bi(rbm); + const u8 *buffer = bi->bi_bh->b_data + bi->bi_offset; + const u8 *byte; unsigned int bit; - byte = buffer + (block / GFS2_NBBY); - bit = (block % GFS2_NBBY) * GFS2_BIT_SIZE; - end = buffer + buflen; - - gfs2_assert(rgd->rd_sbd, byte < end); + byte = buffer + (rbm->offset / GFS2_NBBY); + bit = (rbm->offset % GFS2_NBBY) * GFS2_BIT_SIZE; - cur_state = (*byte >> bit) & GFS2_BIT_MASK; - - return cur_state; + return (*byte >> bit) & GFS2_BIT_MASK; } /** @@ -165,9 +174,30 @@ static inline u64 gfs2_bit_search(const __le64 *ptr, u64 mask, u8 state) } /** + * rs_cmp - multi-block reservation range compare + * @blk: absolute file system block number of the new reservation + * @len: number of blocks in the new reservation + * @rs: existing reservation to compare against + * + * returns: 1 if the block range is beyond the reach of the reservation + * -1 if the block range is before the start of the reservation + * 0 if the block range overlaps with the reservation + */ +static inline int rs_cmp(u64 blk, u32 len, struct gfs2_blkreserv *rs) +{ + u64 startblk = gfs2_rbm_to_block(&rs->rs_rbm); + + if (blk >= startblk + rs->rs_free) + return 1; + if (blk + len - 1 < startblk) + return -1; + return 0; +} + +/** * gfs2_bitfit - Search an rgrp's bitmap buffer to find a bit-pair representing * a block in a given allocation state. - * @buffer: the buffer that holds the bitmaps + * @buf: the buffer that holds the bitmaps * @len: the length (in bytes) of the buffer * @goal: start search at this block's bit-pair (within @buffer) * @state: GFS2_BLKST_XXX the state of the block we're looking for. @@ -195,8 +225,6 @@ static u32 gfs2_bitfit(const u8 *buf, const unsigned int len, u64 mask = 0x5555555555555555ULL; u32 bit; - BUG_ON(state > 3); - /* Mask off bits we don't care about at the start of the search */ mask <<= spoint; tmp = gfs2_bit_search(ptr, mask, state); @@ -218,7 +246,164 @@ static u32 gfs2_bitfit(const u8 *buf, const unsigned int len, } /** + * gfs2_rbm_from_block - Set the rbm based upon rgd and block number + * @rbm: The rbm with rgd already set correctly + * @block: The block number (filesystem relative) + * + * This sets the bi and offset members of an rbm based on a + * resource group and a filesystem relative block number. The + * resource group must be set in the rbm on entry, the bi and + * offset members will be set by this function. + * + * Returns: 0 on success, or an error code + */ + +static int gfs2_rbm_from_block(struct gfs2_rbm *rbm, u64 block) +{ + u64 rblock = block - rbm->rgd->rd_data0; + + if (WARN_ON_ONCE(rblock > UINT_MAX)) + return -EINVAL; + if (block >= rbm->rgd->rd_data0 + rbm->rgd->rd_data) + return -E2BIG; + + rbm->bii = 0; + rbm->offset = (u32)(rblock); + /* Check if the block is within the first block */ + if (rbm->offset < rbm_bi(rbm)->bi_blocks) + return 0; + + /* Adjust for the size diff between gfs2_meta_header and gfs2_rgrp */ + rbm->offset += (sizeof(struct gfs2_rgrp) - + sizeof(struct gfs2_meta_header)) * GFS2_NBBY; + rbm->bii = rbm->offset / rbm->rgd->rd_sbd->sd_blocks_per_bitmap; + rbm->offset -= rbm->bii * rbm->rgd->rd_sbd->sd_blocks_per_bitmap; + return 0; +} + +/** + * gfs2_rbm_incr - increment an rbm structure + * @rbm: The rbm with rgd already set correctly + * + * This function takes an existing rbm structure and increments it to the next + * viable block offset. + * + * Returns: If incrementing the offset would cause the rbm to go past the + * end of the rgrp, true is returned, otherwise false. + * + */ + +static bool gfs2_rbm_incr(struct gfs2_rbm *rbm) +{ + if (rbm->offset + 1 < rbm_bi(rbm)->bi_blocks) { /* in the same bitmap */ + rbm->offset++; + return false; + } + if (rbm->bii == rbm->rgd->rd_length - 1) /* at the last bitmap */ + return true; + + rbm->offset = 0; + rbm->bii++; + return false; +} + +/** + * gfs2_unaligned_extlen - Look for free blocks which are not byte aligned + * @rbm: Position to search (value/result) + * @n_unaligned: Number of unaligned blocks to check + * @len: Decremented for each block found (terminate on zero) + * + * Returns: true if a non-free block is encountered + */ + +static bool gfs2_unaligned_extlen(struct gfs2_rbm *rbm, u32 n_unaligned, u32 *len) +{ + u32 n; + u8 res; + + for (n = 0; n < n_unaligned; n++) { + res = gfs2_testbit(rbm); + if (res != GFS2_BLKST_FREE) + return true; + (*len)--; + if (*len == 0) + return true; + if (gfs2_rbm_incr(rbm)) + return true; + } + + return false; +} + +/** + * gfs2_free_extlen - Return extent length of free blocks + * @rrbm: Starting position + * @len: Max length to check + * + * Starting at the block specified by the rbm, see how many free blocks + * there are, not reading more than len blocks ahead. This can be done + * using memchr_inv when the blocks are byte aligned, but has to be done + * on a block by block basis in case of unaligned blocks. Also this + * function can cope with bitmap boundaries (although it must stop on + * a resource group boundary) + * + * Returns: Number of free blocks in the extent + */ + +static u32 gfs2_free_extlen(const struct gfs2_rbm *rrbm, u32 len) +{ + struct gfs2_rbm rbm = *rrbm; + u32 n_unaligned = rbm.offset & 3; + u32 size = len; + u32 bytes; + u32 chunk_size; + u8 *ptr, *start, *end; + u64 block; + struct gfs2_bitmap *bi; + + if (n_unaligned && + gfs2_unaligned_extlen(&rbm, 4 - n_unaligned, &len)) + goto out; + + n_unaligned = len & 3; + /* Start is now byte aligned */ + while (len > 3) { + bi = rbm_bi(&rbm); + start = bi->bi_bh->b_data; + if (bi->bi_clone) + start = bi->bi_clone; + end = start + bi->bi_bh->b_size; + start += bi->bi_offset; + BUG_ON(rbm.offset & 3); + start += (rbm.offset / GFS2_NBBY); + bytes = min_t(u32, len / GFS2_NBBY, (end - start)); + ptr = memchr_inv(start, 0, bytes); + chunk_size = ((ptr == NULL) ? bytes : (ptr - start)); + chunk_size *= GFS2_NBBY; + BUG_ON(len < chunk_size); + len -= chunk_size; + block = gfs2_rbm_to_block(&rbm); + if (gfs2_rbm_from_block(&rbm, block + chunk_size)) { + n_unaligned = 0; + break; + } + if (ptr) { + n_unaligned = 3; + break; + } + n_unaligned = len & 3; + } + + /* Deal with any bits left over at the end */ + if (n_unaligned) + gfs2_unaligned_extlen(&rbm, n_unaligned, &len); +out: + return size - len; +} + +/** * gfs2_bitcount - count the number of bits in a certain state + * @rgd: the resource group descriptor * @buffer: the buffer that holds the bitmaps * @buflen: the length (in bytes) of the buffer * @state: the state of the block we're looking for @@ -252,7 +437,6 @@ static u32 gfs2_bitcount(struct gfs2_rgrpd *rgd, const u8 *buffer, /** * gfs2_rgrp_verify - Verify that a resource group is consistent - * @sdp: the filesystem * @rgd: the rgrp * */ @@ -310,25 +494,38 @@ static inline int rgrp_contains_block(struct gfs2_rgrpd *rgd, u64 block) /** * gfs2_blk2rgrpd - Find resource group for a given data/meta block number * @sdp: The GFS2 superblock - * @n: The data block number + * @blk: The data block number + * @exact: True if this needs to be an exact match * * Returns: The resource group, or NULL if not found */ -struct gfs2_rgrpd *gfs2_blk2rgrpd(struct gfs2_sbd *sdp, u64 blk) +struct gfs2_rgrpd *gfs2_blk2rgrpd(struct gfs2_sbd *sdp, u64 blk, bool exact) { - struct gfs2_rgrpd *rgd; + struct rb_node *n, *next; + struct gfs2_rgrpd *cur; spin_lock(&sdp->sd_rindex_spin); - - list_for_each_entry(rgd, &sdp->sd_rindex_mru_list, rd_list_mru) { - if (rgrp_contains_block(rgd, blk)) { - list_move(&rgd->rd_list_mru, &sdp->sd_rindex_mru_list); + n = sdp->sd_rindex_tree.rb_node; + while (n) { + cur = rb_entry(n, struct gfs2_rgrpd, rd_node); + next = NULL; + if (blk < cur->rd_addr) + next = n->rb_left; + else if (blk >= cur->rd_data0 + cur->rd_data) + next = n->rb_right; + if (next == NULL) { spin_unlock(&sdp->sd_rindex_spin); - return rgd; + if (exact) { + if (blk < cur->rd_addr) + return NULL; + if (blk >= cur->rd_data0 + cur->rd_data) + return NULL; + } + return cur; } + n = next; } - spin_unlock(&sdp->sd_rindex_spin); return NULL; @@ -343,66 +540,209 @@ struct gfs2_rgrpd *gfs2_blk2rgrpd(struct gfs2_sbd *sdp, u64 blk) struct gfs2_rgrpd *gfs2_rgrpd_get_first(struct gfs2_sbd *sdp) { - gfs2_assert(sdp, !list_empty(&sdp->sd_rindex_list)); - return list_entry(sdp->sd_rindex_list.next, struct gfs2_rgrpd, rd_list); + const struct rb_node *n; + struct gfs2_rgrpd *rgd; + + spin_lock(&sdp->sd_rindex_spin); + n = rb_first(&sdp->sd_rindex_tree); + rgd = rb_entry(n, struct gfs2_rgrpd, rd_node); + spin_unlock(&sdp->sd_rindex_spin); + + return rgd; } /** * gfs2_rgrpd_get_next - get the next RG - * @rgd: A RG + * @rgd: the resource group descriptor * * Returns: The next rgrp */ struct gfs2_rgrpd *gfs2_rgrpd_get_next(struct gfs2_rgrpd *rgd) { - if (rgd->rd_list.next == &rgd->rd_sbd->sd_rindex_list) + struct gfs2_sbd *sdp = rgd->rd_sbd; + const struct rb_node *n; + + spin_lock(&sdp->sd_rindex_spin); + n = rb_next(&rgd->rd_node); + if (n == NULL) + n = rb_first(&sdp->sd_rindex_tree); + + if (unlikely(&rgd->rd_node == n)) { + spin_unlock(&sdp->sd_rindex_spin); return NULL; - return list_entry(rgd->rd_list.next, struct gfs2_rgrpd, rd_list); + } + rgd = rb_entry(n, struct gfs2_rgrpd, rd_node); + spin_unlock(&sdp->sd_rindex_spin); + return rgd; +} + +void gfs2_free_clones(struct gfs2_rgrpd *rgd) +{ + int x; + + for (x = 0; x < rgd->rd_length; x++) { + struct gfs2_bitmap *bi = rgd->rd_bits + x; + kfree(bi->bi_clone); + bi->bi_clone = NULL; + } +} + +/** + * gfs2_rs_alloc - make sure we have a reservation assigned to the inode + * @ip: the inode for this reservation + */ +int gfs2_rs_alloc(struct gfs2_inode *ip) +{ + int error = 0; + + down_write(&ip->i_rw_mutex); + if (ip->i_res) + goto out; + + ip->i_res = kmem_cache_zalloc(gfs2_rsrv_cachep, GFP_NOFS); + if (!ip->i_res) { + error = -ENOMEM; + goto out; + } + + RB_CLEAR_NODE(&ip->i_res->rs_node); +out: + up_write(&ip->i_rw_mutex); + return error; } -static void clear_rgrpdi(struct gfs2_sbd *sdp) +static void dump_rs(struct seq_file *seq, const struct gfs2_blkreserv *rs) +{ + gfs2_print_dbg(seq, " B: n:%llu s:%llu b:%u f:%u\n", + (unsigned long long)rs->rs_inum, + (unsigned long long)gfs2_rbm_to_block(&rs->rs_rbm), + rs->rs_rbm.offset, rs->rs_free); +} + +/** + * __rs_deltree - remove a multi-block reservation from the rgd tree + * @rs: The reservation to remove + * + */ +static void __rs_deltree(struct gfs2_blkreserv *rs) { - struct list_head *head; struct gfs2_rgrpd *rgd; - struct gfs2_glock *gl; - spin_lock(&sdp->sd_rindex_spin); - sdp->sd_rindex_forward = NULL; - spin_unlock(&sdp->sd_rindex_spin); + if (!gfs2_rs_active(rs)) + return; + + rgd = rs->rs_rbm.rgd; + trace_gfs2_rs(rs, TRACE_RS_TREEDEL); + rb_erase(&rs->rs_node, &rgd->rd_rstree); + RB_CLEAR_NODE(&rs->rs_node); + + if (rs->rs_free) { + struct gfs2_bitmap *bi = rbm_bi(&rs->rs_rbm); + + /* return reserved blocks to the rgrp */ + BUG_ON(rs->rs_rbm.rgd->rd_reserved < rs->rs_free); + rs->rs_rbm.rgd->rd_reserved -= rs->rs_free; + /* The rgrp extent failure point is likely not to increase; + it will only do so if the freed blocks are somehow + contiguous with a span of free blocks that follows. Still, + it will force the number to be recalculated later. */ + rgd->rd_extfail_pt += rs->rs_free; + rs->rs_free = 0; + clear_bit(GBF_FULL, &bi->bi_flags); + } +} + +/** + * gfs2_rs_deltree - remove a multi-block reservation from the rgd tree + * @rs: The reservation to remove + * + */ +void gfs2_rs_deltree(struct gfs2_blkreserv *rs) +{ + struct gfs2_rgrpd *rgd; - head = &sdp->sd_rindex_list; - while (!list_empty(head)) { - rgd = list_entry(head->next, struct gfs2_rgrpd, rd_list); + rgd = rs->rs_rbm.rgd; + if (rgd) { + spin_lock(&rgd->rd_rsspin); + __rs_deltree(rs); + spin_unlock(&rgd->rd_rsspin); + } +} + +/** + * gfs2_rs_delete - delete a multi-block reservation + * @ip: The inode for this reservation + * @wcount: The inode's write count, or NULL + * + */ +void gfs2_rs_delete(struct gfs2_inode *ip, atomic_t *wcount) +{ + down_write(&ip->i_rw_mutex); + if (ip->i_res && ((wcount == NULL) || (atomic_read(wcount) <= 1))) { + gfs2_rs_deltree(ip->i_res); + BUG_ON(ip->i_res->rs_free); + kmem_cache_free(gfs2_rsrv_cachep, ip->i_res); + ip->i_res = NULL; + } + up_write(&ip->i_rw_mutex); +} + +/** + * return_all_reservations - return all reserved blocks back to the rgrp. + * @rgd: the rgrp that needs its space back + * + * We previously reserved a bunch of blocks for allocation. Now we need to + * give them back. This leave the reservation structures in tact, but removes + * all of their corresponding "no-fly zones". + */ +static void return_all_reservations(struct gfs2_rgrpd *rgd) +{ + struct rb_node *n; + struct gfs2_blkreserv *rs; + + spin_lock(&rgd->rd_rsspin); + while ((n = rb_first(&rgd->rd_rstree))) { + rs = rb_entry(n, struct gfs2_blkreserv, rs_node); + __rs_deltree(rs); + } + spin_unlock(&rgd->rd_rsspin); +} + +void gfs2_clear_rgrpd(struct gfs2_sbd *sdp) +{ + struct rb_node *n; + struct gfs2_rgrpd *rgd; + struct gfs2_glock *gl; + + while ((n = rb_first(&sdp->sd_rindex_tree))) { + rgd = rb_entry(n, struct gfs2_rgrpd, rd_node); gl = rgd->rd_gl; - list_del(&rgd->rd_list); - list_del(&rgd->rd_list_mru); + rb_erase(n, &sdp->sd_rindex_tree); if (gl) { + spin_lock(&gl->gl_spin); gl->gl_object = NULL; + spin_unlock(&gl->gl_spin); + gfs2_glock_add_to_lru(gl); gfs2_glock_put(gl); } + gfs2_free_clones(rgd); kfree(rgd->rd_bits); + return_all_reservations(rgd); kmem_cache_free(gfs2_rgrpd_cachep, rgd); } } -void gfs2_clear_rgrpd(struct gfs2_sbd *sdp) -{ - mutex_lock(&sdp->sd_rindex_mutex); - clear_rgrpdi(sdp); - mutex_unlock(&sdp->sd_rindex_mutex); -} - static void gfs2_rindex_print(const struct gfs2_rgrpd *rgd) { - printk(KERN_INFO " ri_addr = %llu\n", (unsigned long long)rgd->rd_addr); - printk(KERN_INFO " ri_length = %u\n", rgd->rd_length); - printk(KERN_INFO " ri_data0 = %llu\n", (unsigned long long)rgd->rd_data0); - printk(KERN_INFO " ri_data = %u\n", rgd->rd_data); - printk(KERN_INFO " ri_bitbytes = %u\n", rgd->rd_bitbytes); + pr_info("ri_addr = %llu\n", (unsigned long long)rgd->rd_addr); + pr_info("ri_length = %u\n", rgd->rd_length); + pr_info("ri_data0 = %llu\n", (unsigned long long)rgd->rd_data0); + pr_info("ri_data = %u\n", rgd->rd_data); + pr_info("ri_bitbytes = %u\n", rgd->rd_bitbytes); } /** @@ -441,18 +781,21 @@ static int compute_bitstructs(struct gfs2_rgrpd *rgd) bi->bi_offset = sizeof(struct gfs2_rgrp); bi->bi_start = 0; bi->bi_len = bytes; + bi->bi_blocks = bytes * GFS2_NBBY; /* header block */ } else if (x == 0) { bytes = sdp->sd_sb.sb_bsize - sizeof(struct gfs2_rgrp); bi->bi_offset = sizeof(struct gfs2_rgrp); bi->bi_start = 0; bi->bi_len = bytes; + bi->bi_blocks = bytes * GFS2_NBBY; /* last block */ } else if (x + 1 == length) { bytes = bytes_left; bi->bi_offset = sizeof(struct gfs2_meta_header); bi->bi_start = rgd->rd_bitbytes - bytes_left; bi->bi_len = bytes; + bi->bi_blocks = bytes * GFS2_NBBY; /* other blocks */ } else { bytes = sdp->sd_sb.sb_bsize - @@ -460,6 +803,7 @@ static int compute_bitstructs(struct gfs2_rgrpd *rgd) bi->bi_offset = sizeof(struct gfs2_meta_header); bi->bi_start = rgd->rd_bitbytes - bytes_left; bi->bi_len = bytes; + bi->bi_blocks = bytes * GFS2_NBBY; } bytes_left -= bytes; @@ -484,6 +828,7 @@ static int compute_bitstructs(struct gfs2_rgrpd *rgd) /** * gfs2_ri_total - Total up the file system space, according to the rindex. + * @sdp: the filesystem * */ u64 gfs2_ri_total(struct gfs2_sbd *sdp) @@ -492,87 +837,113 @@ u64 gfs2_ri_total(struct gfs2_sbd *sdp) struct inode *inode = sdp->sd_rindex; struct gfs2_inode *ip = GFS2_I(inode); char buf[sizeof(struct gfs2_rindex)]; - struct file_ra_state ra_state; int error, rgrps; - mutex_lock(&sdp->sd_rindex_mutex); - file_ra_state_init(&ra_state, inode->i_mapping); for (rgrps = 0;; rgrps++) { loff_t pos = rgrps * sizeof(struct gfs2_rindex); if (pos + sizeof(struct gfs2_rindex) > i_size_read(inode)) break; - error = gfs2_internal_read(ip, &ra_state, buf, &pos, + error = gfs2_internal_read(ip, buf, &pos, sizeof(struct gfs2_rindex)); if (error != sizeof(struct gfs2_rindex)) break; total_data += be32_to_cpu(((struct gfs2_rindex *)buf)->ri_data); } - mutex_unlock(&sdp->sd_rindex_mutex); return total_data; } -static void gfs2_rindex_in(struct gfs2_rgrpd *rgd, const void *buf) +static int rgd_insert(struct gfs2_rgrpd *rgd) { - const struct gfs2_rindex *str = buf; + struct gfs2_sbd *sdp = rgd->rd_sbd; + struct rb_node **newn = &sdp->sd_rindex_tree.rb_node, *parent = NULL; + + /* Figure out where to put new node */ + while (*newn) { + struct gfs2_rgrpd *cur = rb_entry(*newn, struct gfs2_rgrpd, + rd_node); + + parent = *newn; + if (rgd->rd_addr < cur->rd_addr) + newn = &((*newn)->rb_left); + else if (rgd->rd_addr > cur->rd_addr) + newn = &((*newn)->rb_right); + else + return -EEXIST; + } - rgd->rd_addr = be64_to_cpu(str->ri_addr); - rgd->rd_length = be32_to_cpu(str->ri_length); - rgd->rd_data0 = be64_to_cpu(str->ri_data0); - rgd->rd_data = be32_to_cpu(str->ri_data); - rgd->rd_bitbytes = be32_to_cpu(str->ri_bitbytes); + rb_link_node(&rgd->rd_node, parent, newn); + rb_insert_color(&rgd->rd_node, &sdp->sd_rindex_tree); + sdp->sd_rgrps++; + return 0; } /** * read_rindex_entry - Pull in a new resource index entry from the disk - * @gl: The glock covering the rindex inode + * @ip: Pointer to the rindex inode * - * Returns: 0 on success, error code otherwise + * Returns: 0 on success, > 0 on EOF, error code otherwise */ -static int read_rindex_entry(struct gfs2_inode *ip, - struct file_ra_state *ra_state) +static int read_rindex_entry(struct gfs2_inode *ip) { struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); + const unsigned bsize = sdp->sd_sb.sb_bsize; loff_t pos = sdp->sd_rgrps * sizeof(struct gfs2_rindex); - char buf[sizeof(struct gfs2_rindex)]; + struct gfs2_rindex buf; int error; struct gfs2_rgrpd *rgd; - error = gfs2_internal_read(ip, ra_state, buf, &pos, + if (pos >= i_size_read(&ip->i_inode)) + return 1; + + error = gfs2_internal_read(ip, (char *)&buf, &pos, sizeof(struct gfs2_rindex)); - if (!error) - return 0; - if (error != sizeof(struct gfs2_rindex)) { - if (error > 0) - error = -EIO; - return error; - } + + if (error != sizeof(struct gfs2_rindex)) + return (error == 0) ? 1 : error; rgd = kmem_cache_zalloc(gfs2_rgrpd_cachep, GFP_NOFS); error = -ENOMEM; if (!rgd) return error; - mutex_init(&rgd->rd_mutex); - lops_init_le(&rgd->rd_le, &gfs2_rg_lops); rgd->rd_sbd = sdp; + rgd->rd_addr = be64_to_cpu(buf.ri_addr); + rgd->rd_length = be32_to_cpu(buf.ri_length); + rgd->rd_data0 = be64_to_cpu(buf.ri_data0); + rgd->rd_data = be32_to_cpu(buf.ri_data); + rgd->rd_bitbytes = be32_to_cpu(buf.ri_bitbytes); + spin_lock_init(&rgd->rd_rsspin); - list_add_tail(&rgd->rd_list, &sdp->sd_rindex_list); - list_add_tail(&rgd->rd_list_mru, &sdp->sd_rindex_mru_list); - - gfs2_rindex_in(rgd, buf); error = compute_bitstructs(rgd); if (error) - return error; + goto fail; error = gfs2_glock_get(sdp, rgd->rd_addr, &gfs2_rgrp_glops, CREATE, &rgd->rd_gl); if (error) - return error; + goto fail; rgd->rd_gl->gl_object = rgd; + rgd->rd_gl->gl_vm.start = rgd->rd_addr * bsize; + rgd->rd_gl->gl_vm.end = rgd->rd_gl->gl_vm.start + (rgd->rd_length * bsize) - 1; + rgd->rd_rgl = (struct gfs2_rgrp_lvb *)rgd->rd_gl->gl_lksb.sb_lvbptr; rgd->rd_flags &= ~GFS2_RDF_UPTODATE; + if (rgd->rd_data > sdp->sd_max_rg_data) + sdp->sd_max_rg_data = rgd->rd_data; + spin_lock(&sdp->sd_rindex_spin); + error = rgd_insert(rgd); + spin_unlock(&sdp->sd_rindex_spin); + if (!error) + return 0; + + error = 0; /* someone else read in the rgrp; free it and ignore it */ + gfs2_glock_put(rgd->rd_gl); + +fail: + kfree(rgd->rd_bits); + kmem_cache_free(gfs2_rgrpd_cachep, rgd); return error; } @@ -583,40 +954,25 @@ static int read_rindex_entry(struct gfs2_inode *ip, * Returns: 0 on successful update, error code otherwise */ -int gfs2_ri_update(struct gfs2_inode *ip) +static int gfs2_ri_update(struct gfs2_inode *ip) { struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); - struct inode *inode = &ip->i_inode; - struct file_ra_state ra_state; - u64 rgrp_count = i_size_read(inode); - struct gfs2_rgrpd *rgd; - unsigned int max_data = 0; int error; - do_div(rgrp_count, sizeof(struct gfs2_rindex)); - clear_rgrpdi(sdp); + do { + error = read_rindex_entry(ip); + } while (error == 0); - file_ra_state_init(&ra_state, inode->i_mapping); - for (sdp->sd_rgrps = 0; sdp->sd_rgrps < rgrp_count; sdp->sd_rgrps++) { - error = read_rindex_entry(ip, &ra_state); - if (error) { - clear_rgrpdi(sdp); - return error; - } - } + if (error < 0) + return error; - list_for_each_entry(rgd, &sdp->sd_rindex_list, rd_list) - if (rgd->rd_data > max_data) - max_data = rgd->rd_data; - sdp->sd_max_rg_data = max_data; sdp->sd_rindex_uptodate = 1; return 0; } /** - * gfs2_rindex_hold - Grab a lock on the rindex + * gfs2_rindex_update - Update the rindex if required * @sdp: The GFS2 superblock - * @ri_gh: the glock holder * * We grab a lock on the rindex inode to make sure that it doesn't * change whilst we are performing an operation. We keep this lock @@ -628,28 +984,29 @@ int gfs2_ri_update(struct gfs2_inode *ip) * special file, which might have been updated if someone expanded the * filesystem (via gfs2_grow utility), which adds new resource groups. * - * Returns: 0 on success, error code otherwise + * Returns: 0 on succeess, error code otherwise */ -int gfs2_rindex_hold(struct gfs2_sbd *sdp, struct gfs2_holder *ri_gh) +int gfs2_rindex_update(struct gfs2_sbd *sdp) { struct gfs2_inode *ip = GFS2_I(sdp->sd_rindex); struct gfs2_glock *gl = ip->i_gl; - int error; - - error = gfs2_glock_nq_init(gl, LM_ST_SHARED, 0, ri_gh); - if (error) - return error; + struct gfs2_holder ri_gh; + int error = 0; + int unlock_required = 0; /* Read new copy from disk if we don't have the latest */ if (!sdp->sd_rindex_uptodate) { - mutex_lock(&sdp->sd_rindex_mutex); - if (!sdp->sd_rindex_uptodate) { - error = gfs2_ri_update(ip); + if (!gfs2_glock_is_locked_by_me(gl)) { + error = gfs2_glock_nq_init(gl, LM_ST_SHARED, 0, &ri_gh); if (error) - gfs2_glock_dq_uninit(ri_gh); + return error; + unlock_required = 1; } - mutex_unlock(&sdp->sd_rindex_mutex); + if (!sdp->sd_rindex_uptodate) + error = gfs2_ri_update(ip); + if (unlock_required) + gfs2_glock_dq_uninit(&ri_gh); } return error; @@ -681,6 +1038,62 @@ static void gfs2_rgrp_out(struct gfs2_rgrpd *rgd, void *buf) memset(&str->rg_reserved, 0, sizeof(str->rg_reserved)); } +static int gfs2_rgrp_lvb_valid(struct gfs2_rgrpd *rgd) +{ + struct gfs2_rgrp_lvb *rgl = rgd->rd_rgl; + struct gfs2_rgrp *str = (struct gfs2_rgrp *)rgd->rd_bits[0].bi_bh->b_data; + + if (rgl->rl_flags != str->rg_flags || rgl->rl_free != str->rg_free || + rgl->rl_dinodes != str->rg_dinodes || + rgl->rl_igeneration != str->rg_igeneration) + return 0; + return 1; +} + +static void gfs2_rgrp_ondisk2lvb(struct gfs2_rgrp_lvb *rgl, const void *buf) +{ + const struct gfs2_rgrp *str = buf; + + rgl->rl_magic = cpu_to_be32(GFS2_MAGIC); + rgl->rl_flags = str->rg_flags; + rgl->rl_free = str->rg_free; + rgl->rl_dinodes = str->rg_dinodes; + rgl->rl_igeneration = str->rg_igeneration; + rgl->__pad = 0UL; +} + +static void update_rgrp_lvb_unlinked(struct gfs2_rgrpd *rgd, u32 change) +{ + struct gfs2_rgrp_lvb *rgl = rgd->rd_rgl; + u32 unlinked = be32_to_cpu(rgl->rl_unlinked) + change; + rgl->rl_unlinked = cpu_to_be32(unlinked); +} + +static u32 count_unlinked(struct gfs2_rgrpd *rgd) +{ + struct gfs2_bitmap *bi; + const u32 length = rgd->rd_length; + const u8 *buffer = NULL; + u32 i, goal, count = 0; + + for (i = 0, bi = rgd->rd_bits; i < length; i++, bi++) { + goal = 0; + buffer = bi->bi_bh->b_data + bi->bi_offset; + WARN_ON(!buffer_uptodate(bi->bi_bh)); + while (goal < bi->bi_len * GFS2_NBBY) { + goal = gfs2_bitfit(buffer, bi->bi_len, goal, + GFS2_BLKST_UNLINKED); + if (goal == BFITNOENT) + break; + count++; + goal++; + } + } + + return count; +} + + /** * gfs2_rgrp_bh_get - Read in a RG's header and bitmaps * @rgd: the struct gfs2_rgrpd describing the RG to read in @@ -691,7 +1104,7 @@ static void gfs2_rgrp_out(struct gfs2_rgrpd *rgd, void *buf) * Returns: errno */ -int gfs2_rgrp_bh_get(struct gfs2_rgrpd *rgd) +static int gfs2_rgrp_bh_get(struct gfs2_rgrpd *rgd) { struct gfs2_sbd *sdp = rgd->rd_sbd; struct gfs2_glock *gl = rgd->rd_gl; @@ -700,16 +1113,8 @@ int gfs2_rgrp_bh_get(struct gfs2_rgrpd *rgd) unsigned int x, y; int error; - mutex_lock(&rgd->rd_mutex); - - spin_lock(&sdp->sd_rindex_spin); - if (rgd->rd_bh_count) { - rgd->rd_bh_count++; - spin_unlock(&sdp->sd_rindex_spin); - mutex_unlock(&rgd->rd_mutex); + if (rgd->rd_bits[0].bi_bh != NULL) return 0; - } - spin_unlock(&sdp->sd_rindex_spin); for (x = 0; x < length; x++) { bi = rgd->rd_bits + x; @@ -735,15 +1140,24 @@ int gfs2_rgrp_bh_get(struct gfs2_rgrpd *rgd) clear_bit(GBF_FULL, &rgd->rd_bits[x].bi_flags); gfs2_rgrp_in(rgd, (rgd->rd_bits[0].bi_bh)->b_data); rgd->rd_flags |= (GFS2_RDF_UPTODATE | GFS2_RDF_CHECK); + rgd->rd_free_clone = rgd->rd_free; + /* max out the rgrp allocation failure point */ + rgd->rd_extfail_pt = rgd->rd_free; + } + if (cpu_to_be32(GFS2_MAGIC) != rgd->rd_rgl->rl_magic) { + rgd->rd_rgl->rl_unlinked = cpu_to_be32(count_unlinked(rgd)); + gfs2_rgrp_ondisk2lvb(rgd->rd_rgl, + rgd->rd_bits[0].bi_bh->b_data); + } + else if (sdp->sd_args.ar_rgrplvb) { + if (!gfs2_rgrp_lvb_valid(rgd)){ + gfs2_consist_rgrpd(rgd); + error = -EIO; + goto fail; + } + if (rgd->rd_rgl->rl_unlinked == 0) + rgd->rd_flags &= ~GFS2_RDF_CHECK; } - - spin_lock(&sdp->sd_rindex_spin); - rgd->rd_free_clone = rgd->rd_free; - rgd->rd_bh_count++; - spin_unlock(&sdp->sd_rindex_spin); - - mutex_unlock(&rgd->rd_mutex); - return 0; fail: @@ -753,171 +1167,556 @@ fail: bi->bi_bh = NULL; gfs2_assert_warn(sdp, !bi->bi_clone); } - mutex_unlock(&rgd->rd_mutex); return error; } -void gfs2_rgrp_bh_hold(struct gfs2_rgrpd *rgd) +static int update_rgrp_lvb(struct gfs2_rgrpd *rgd) +{ + u32 rl_flags; + + if (rgd->rd_flags & GFS2_RDF_UPTODATE) + return 0; + + if (cpu_to_be32(GFS2_MAGIC) != rgd->rd_rgl->rl_magic) + return gfs2_rgrp_bh_get(rgd); + + rl_flags = be32_to_cpu(rgd->rd_rgl->rl_flags); + rl_flags &= ~GFS2_RDF_MASK; + rgd->rd_flags &= GFS2_RDF_MASK; + rgd->rd_flags |= (rl_flags | GFS2_RDF_UPTODATE | GFS2_RDF_CHECK); + if (rgd->rd_rgl->rl_unlinked == 0) + rgd->rd_flags &= ~GFS2_RDF_CHECK; + rgd->rd_free = be32_to_cpu(rgd->rd_rgl->rl_free); + rgd->rd_free_clone = rgd->rd_free; + rgd->rd_dinodes = be32_to_cpu(rgd->rd_rgl->rl_dinodes); + rgd->rd_igeneration = be64_to_cpu(rgd->rd_rgl->rl_igeneration); + return 0; +} + +int gfs2_rgrp_go_lock(struct gfs2_holder *gh) { + struct gfs2_rgrpd *rgd = gh->gh_gl->gl_object; struct gfs2_sbd *sdp = rgd->rd_sbd; - spin_lock(&sdp->sd_rindex_spin); - gfs2_assert_warn(rgd->rd_sbd, rgd->rd_bh_count); - rgd->rd_bh_count++; - spin_unlock(&sdp->sd_rindex_spin); + if (gh->gh_flags & GL_SKIP && sdp->sd_args.ar_rgrplvb) + return 0; + return gfs2_rgrp_bh_get(rgd); } /** - * gfs2_rgrp_bh_put - Release RG bitmaps read in with gfs2_rgrp_bh_get() - * @rgd: the struct gfs2_rgrpd describing the RG to read in + * gfs2_rgrp_go_unlock - Release RG bitmaps read in with gfs2_rgrp_bh_get() + * @gh: The glock holder for the resource group * */ -void gfs2_rgrp_bh_put(struct gfs2_rgrpd *rgd) +void gfs2_rgrp_go_unlock(struct gfs2_holder *gh) { - struct gfs2_sbd *sdp = rgd->rd_sbd; + struct gfs2_rgrpd *rgd = gh->gh_gl->gl_object; int x, length = rgd->rd_length; - spin_lock(&sdp->sd_rindex_spin); - gfs2_assert_warn(rgd->rd_sbd, rgd->rd_bh_count); - if (--rgd->rd_bh_count) { - spin_unlock(&sdp->sd_rindex_spin); - return; - } - for (x = 0; x < length; x++) { struct gfs2_bitmap *bi = rgd->rd_bits + x; - kfree(bi->bi_clone); - bi->bi_clone = NULL; - brelse(bi->bi_bh); - bi->bi_bh = NULL; + if (bi->bi_bh) { + brelse(bi->bi_bh); + bi->bi_bh = NULL; + } } - spin_unlock(&sdp->sd_rindex_spin); } -static void gfs2_rgrp_send_discards(struct gfs2_sbd *sdp, u64 offset, - const struct gfs2_bitmap *bi) +int gfs2_rgrp_send_discards(struct gfs2_sbd *sdp, u64 offset, + struct buffer_head *bh, + const struct gfs2_bitmap *bi, unsigned minlen, u64 *ptrimmed) { struct super_block *sb = sdp->sd_vfs; - struct block_device *bdev = sb->s_bdev; - const unsigned int sects_per_blk = sdp->sd_sb.sb_bsize / - bdev_logical_block_size(sb->s_bdev); u64 blk; sector_t start = 0; - sector_t nr_sects = 0; + sector_t nr_blks = 0; int rv; unsigned int x; + u32 trimmed = 0; + u8 diff; for (x = 0; x < bi->bi_len; x++) { - const u8 *orig = bi->bi_bh->b_data + bi->bi_offset + x; - const u8 *clone = bi->bi_clone + bi->bi_offset + x; - u8 diff = ~(*orig | (*orig >> 1)) & (*clone | (*clone >> 1)); + const u8 *clone = bi->bi_clone ? bi->bi_clone : bi->bi_bh->b_data; + clone += bi->bi_offset; + clone += x; + if (bh) { + const u8 *orig = bh->b_data + bi->bi_offset + x; + diff = ~(*orig | (*orig >> 1)) & (*clone | (*clone >> 1)); + } else { + diff = ~(*clone | (*clone >> 1)); + } diff &= 0x55; if (diff == 0) continue; blk = offset + ((bi->bi_start + x) * GFS2_NBBY); - blk *= sects_per_blk; /* convert to sectors */ while(diff) { if (diff & 1) { - if (nr_sects == 0) + if (nr_blks == 0) goto start_new_extent; - if ((start + nr_sects) != blk) { - rv = blkdev_issue_discard(bdev, start, - nr_sects, GFP_NOFS, - 0); - if (rv) - goto fail; - nr_sects = 0; + if ((start + nr_blks) != blk) { + if (nr_blks >= minlen) { + rv = sb_issue_discard(sb, + start, nr_blks, + GFP_NOFS, 0); + if (rv) + goto fail; + trimmed += nr_blks; + } + nr_blks = 0; start_new_extent: start = blk; } - nr_sects += sects_per_blk; + nr_blks++; } diff >>= 2; - blk += sects_per_blk; + blk++; } } - if (nr_sects) { - rv = blkdev_issue_discard(bdev, start, nr_sects, GFP_NOFS, 0); + if (nr_blks >= minlen) { + rv = sb_issue_discard(sb, start, nr_blks, GFP_NOFS, 0); if (rv) goto fail; + trimmed += nr_blks; } - return; + if (ptrimmed) + *ptrimmed = trimmed; + return 0; + fail: - fs_warn(sdp, "error %d on discard request, turning discards off for this filesystem", rv); + if (sdp->sd_args.ar_discard) + fs_warn(sdp, "error %d on discard request, turning discards off for this filesystem", rv); sdp->sd_args.ar_discard = 0; + return -EIO; } -void gfs2_rgrp_repolish_clones(struct gfs2_rgrpd *rgd) +/** + * gfs2_fitrim - Generate discard requests for unused bits of the filesystem + * @filp: Any file on the filesystem + * @argp: Pointer to the arguments (also used to pass result) + * + * Returns: 0 on success, otherwise error code + */ + +int gfs2_fitrim(struct file *filp, void __user *argp) { - struct gfs2_sbd *sdp = rgd->rd_sbd; - unsigned int length = rgd->rd_length; + struct inode *inode = file_inode(filp); + struct gfs2_sbd *sdp = GFS2_SB(inode); + struct request_queue *q = bdev_get_queue(sdp->sd_vfs->s_bdev); + struct buffer_head *bh; + struct gfs2_rgrpd *rgd; + struct gfs2_rgrpd *rgd_end; + struct gfs2_holder gh; + struct fstrim_range r; + int ret = 0; + u64 amt; + u64 trimmed = 0; + u64 start, end, minlen; unsigned int x; + unsigned bs_shift = sdp->sd_sb.sb_bsize_shift; - for (x = 0; x < length; x++) { - struct gfs2_bitmap *bi = rgd->rd_bits + x; - if (!bi->bi_clone) - continue; - if (sdp->sd_args.ar_discard) - gfs2_rgrp_send_discards(sdp, rgd->rd_data0, bi); - clear_bit(GBF_FULL, &bi->bi_flags); - memcpy(bi->bi_clone + bi->bi_offset, - bi->bi_bh->b_data + bi->bi_offset, bi->bi_len); + if (!capable(CAP_SYS_ADMIN)) + return -EPERM; + + if (!blk_queue_discard(q)) + return -EOPNOTSUPP; + + if (copy_from_user(&r, argp, sizeof(r))) + return -EFAULT; + + ret = gfs2_rindex_update(sdp); + if (ret) + return ret; + + start = r.start >> bs_shift; + end = start + (r.len >> bs_shift); + minlen = max_t(u64, r.minlen, + q->limits.discard_granularity) >> bs_shift; + + if (end <= start || minlen > sdp->sd_max_rg_data) + return -EINVAL; + + rgd = gfs2_blk2rgrpd(sdp, start, 0); + rgd_end = gfs2_blk2rgrpd(sdp, end, 0); + + if ((gfs2_rgrpd_get_first(sdp) == gfs2_rgrpd_get_next(rgd_end)) + && (start > rgd_end->rd_data0 + rgd_end->rd_data)) + return -EINVAL; /* start is beyond the end of the fs */ + + while (1) { + + ret = gfs2_glock_nq_init(rgd->rd_gl, LM_ST_EXCLUSIVE, 0, &gh); + if (ret) + goto out; + + if (!(rgd->rd_flags & GFS2_RGF_TRIMMED)) { + /* Trim each bitmap in the rgrp */ + for (x = 0; x < rgd->rd_length; x++) { + struct gfs2_bitmap *bi = rgd->rd_bits + x; + ret = gfs2_rgrp_send_discards(sdp, + rgd->rd_data0, NULL, bi, minlen, + &amt); + if (ret) { + gfs2_glock_dq_uninit(&gh); + goto out; + } + trimmed += amt; + } + + /* Mark rgrp as having been trimmed */ + ret = gfs2_trans_begin(sdp, RES_RG_HDR, 0); + if (ret == 0) { + bh = rgd->rd_bits[0].bi_bh; + rgd->rd_flags |= GFS2_RGF_TRIMMED; + gfs2_trans_add_meta(rgd->rd_gl, bh); + gfs2_rgrp_out(rgd, bh->b_data); + gfs2_rgrp_ondisk2lvb(rgd->rd_rgl, bh->b_data); + gfs2_trans_end(sdp); + } + } + gfs2_glock_dq_uninit(&gh); + + if (rgd == rgd_end) + break; + + rgd = gfs2_rgrpd_get_next(rgd); } - spin_lock(&sdp->sd_rindex_spin); - rgd->rd_free_clone = rgd->rd_free; - spin_unlock(&sdp->sd_rindex_spin); +out: + r.len = trimmed << bs_shift; + if (copy_to_user(argp, &r, sizeof(r))) + return -EFAULT; + + return ret; } /** - * gfs2_alloc_get - get the struct gfs2_alloc structure for an inode - * @ip: the incore GFS2 inode structure + * rs_insert - insert a new multi-block reservation into the rgrp's rb_tree + * @ip: the inode structure * - * Returns: the struct gfs2_alloc + */ +static void rs_insert(struct gfs2_inode *ip) +{ + struct rb_node **newn, *parent = NULL; + int rc; + struct gfs2_blkreserv *rs = ip->i_res; + struct gfs2_rgrpd *rgd = rs->rs_rbm.rgd; + u64 fsblock = gfs2_rbm_to_block(&rs->rs_rbm); + + BUG_ON(gfs2_rs_active(rs)); + + spin_lock(&rgd->rd_rsspin); + newn = &rgd->rd_rstree.rb_node; + while (*newn) { + struct gfs2_blkreserv *cur = + rb_entry(*newn, struct gfs2_blkreserv, rs_node); + + parent = *newn; + rc = rs_cmp(fsblock, rs->rs_free, cur); + if (rc > 0) + newn = &((*newn)->rb_right); + else if (rc < 0) + newn = &((*newn)->rb_left); + else { + spin_unlock(&rgd->rd_rsspin); + WARN_ON(1); + return; + } + } + + rb_link_node(&rs->rs_node, parent, newn); + rb_insert_color(&rs->rs_node, &rgd->rd_rstree); + + /* Do our rgrp accounting for the reservation */ + rgd->rd_reserved += rs->rs_free; /* blocks reserved */ + spin_unlock(&rgd->rd_rsspin); + trace_gfs2_rs(rs, TRACE_RS_INSERT); +} + +/** + * rg_mblk_search - find a group of multiple free blocks to form a reservation + * @rgd: the resource group descriptor + * @ip: pointer to the inode for which we're reserving blocks + * @ap: the allocation parameters + * + */ + +static void rg_mblk_search(struct gfs2_rgrpd *rgd, struct gfs2_inode *ip, + const struct gfs2_alloc_parms *ap) +{ + struct gfs2_rbm rbm = { .rgd = rgd, }; + u64 goal; + struct gfs2_blkreserv *rs = ip->i_res; + u32 extlen; + u32 free_blocks = rgd->rd_free_clone - rgd->rd_reserved; + int ret; + struct inode *inode = &ip->i_inode; + + if (S_ISDIR(inode->i_mode)) + extlen = 1; + else { + extlen = max_t(u32, atomic_read(&rs->rs_sizehint), ap->target); + extlen = clamp(extlen, RGRP_RSRV_MINBLKS, free_blocks); + } + if ((rgd->rd_free_clone < rgd->rd_reserved) || (free_blocks < extlen)) + return; + + /* Find bitmap block that contains bits for goal block */ + if (rgrp_contains_block(rgd, ip->i_goal)) + goal = ip->i_goal; + else + goal = rgd->rd_last_alloc + rgd->rd_data0; + + if (WARN_ON(gfs2_rbm_from_block(&rbm, goal))) + return; + + ret = gfs2_rbm_find(&rbm, GFS2_BLKST_FREE, &extlen, ip, true, ap); + if (ret == 0) { + rs->rs_rbm = rbm; + rs->rs_free = extlen; + rs->rs_inum = ip->i_no_addr; + rs_insert(ip); + } else { + if (goal == rgd->rd_last_alloc + rgd->rd_data0) + rgd->rd_last_alloc = 0; + } +} + +/** + * gfs2_next_unreserved_block - Return next block that is not reserved + * @rgd: The resource group + * @block: The starting block + * @length: The required length + * @ip: Ignore any reservations for this inode + * + * If the block does not appear in any reservation, then return the + * block number unchanged. If it does appear in the reservation, then + * keep looking through the tree of reservations in order to find the + * first block number which is not reserved. */ -struct gfs2_alloc *gfs2_alloc_get(struct gfs2_inode *ip) +static u64 gfs2_next_unreserved_block(struct gfs2_rgrpd *rgd, u64 block, + u32 length, + const struct gfs2_inode *ip) { - BUG_ON(ip->i_alloc != NULL); - ip->i_alloc = kzalloc(sizeof(struct gfs2_alloc), GFP_NOFS); - return ip->i_alloc; + struct gfs2_blkreserv *rs; + struct rb_node *n; + int rc; + + spin_lock(&rgd->rd_rsspin); + n = rgd->rd_rstree.rb_node; + while (n) { + rs = rb_entry(n, struct gfs2_blkreserv, rs_node); + rc = rs_cmp(block, length, rs); + if (rc < 0) + n = n->rb_left; + else if (rc > 0) + n = n->rb_right; + else + break; + } + + if (n) { + while ((rs_cmp(block, length, rs) == 0) && (ip->i_res != rs)) { + block = gfs2_rbm_to_block(&rs->rs_rbm) + rs->rs_free; + n = n->rb_right; + if (n == NULL) + break; + rs = rb_entry(n, struct gfs2_blkreserv, rs_node); + } + } + + spin_unlock(&rgd->rd_rsspin); + return block; } /** - * try_rgrp_fit - See if a given reservation will fit in a given RG - * @rgd: the RG data - * @al: the struct gfs2_alloc structure describing the reservation + * gfs2_reservation_check_and_update - Check for reservations during block alloc + * @rbm: The current position in the resource group + * @ip: The inode for which we are searching for blocks + * @minext: The minimum extent length + * @maxext: A pointer to the maximum extent structure * - * If there's room for the requested blocks to be allocated from the RG: - * Sets the $al_rgd field in @al. + * This checks the current position in the rgrp to see whether there is + * a reservation covering this block. If not then this function is a + * no-op. If there is, then the position is moved to the end of the + * contiguous reservation(s) so that we are pointing at the first + * non-reserved block. * - * Returns: 1 on success (it fits), 0 on failure (it doesn't fit) + * Returns: 0 if no reservation, 1 if @rbm has changed, otherwise an error */ -static int try_rgrp_fit(struct gfs2_rgrpd *rgd, struct gfs2_alloc *al) +static int gfs2_reservation_check_and_update(struct gfs2_rbm *rbm, + const struct gfs2_inode *ip, + u32 minext, + struct gfs2_extent *maxext) { - struct gfs2_sbd *sdp = rgd->rd_sbd; - int ret = 0; + u64 block = gfs2_rbm_to_block(rbm); + u32 extlen = 1; + u64 nblock; + int ret; + + /* + * If we have a minimum extent length, then skip over any extent + * which is less than the min extent length in size. + */ + if (minext) { + extlen = gfs2_free_extlen(rbm, minext); + if (extlen <= maxext->len) + goto fail; + } - if (rgd->rd_flags & (GFS2_RGF_NOALLOC | GFS2_RDF_ERROR)) - return 0; + /* + * Check the extent which has been found against the reservations + * and skip if parts of it are already reserved + */ + nblock = gfs2_next_unreserved_block(rbm->rgd, block, extlen, ip); + if (nblock == block) { + if (!minext || extlen >= minext) + return 0; + + if (extlen > maxext->len) { + maxext->len = extlen; + maxext->rbm = *rbm; + } +fail: + nblock = block + extlen; + } + ret = gfs2_rbm_from_block(rbm, nblock); + if (ret < 0) + return ret; + return 1; +} - spin_lock(&sdp->sd_rindex_spin); - if (rgd->rd_free_clone >= al->al_requested) { - al->al_rgd = rgd; - ret = 1; +/** + * gfs2_rbm_find - Look for blocks of a particular state + * @rbm: Value/result starting position and final position + * @state: The state which we want to find + * @minext: Pointer to the requested extent length (NULL for a single block) + * This is updated to be the actual reservation size. + * @ip: If set, check for reservations + * @nowrap: Stop looking at the end of the rgrp, rather than wrapping + * around until we've reached the starting point. + * @ap: the allocation parameters + * + * Side effects: + * - If looking for free blocks, we set GBF_FULL on each bitmap which + * has no free blocks in it. + * - If looking for free blocks, we set rd_extfail_pt on each rgrp which + * has come up short on a free block search. + * + * Returns: 0 on success, -ENOSPC if there is no block of the requested state + */ + +static int gfs2_rbm_find(struct gfs2_rbm *rbm, u8 state, u32 *minext, + const struct gfs2_inode *ip, bool nowrap, + const struct gfs2_alloc_parms *ap) +{ + struct buffer_head *bh; + int initial_bii; + u32 initial_offset; + int first_bii = rbm->bii; + u32 first_offset = rbm->offset; + u32 offset; + u8 *buffer; + int n = 0; + int iters = rbm->rgd->rd_length; + int ret; + struct gfs2_bitmap *bi; + struct gfs2_extent maxext = { .rbm.rgd = rbm->rgd, }; + + /* If we are not starting at the beginning of a bitmap, then we + * need to add one to the bitmap count to ensure that we search + * the starting bitmap twice. + */ + if (rbm->offset != 0) + iters++; + + while(1) { + bi = rbm_bi(rbm); + if (test_bit(GBF_FULL, &bi->bi_flags) && + (state == GFS2_BLKST_FREE)) + goto next_bitmap; + + bh = bi->bi_bh; + buffer = bh->b_data + bi->bi_offset; + WARN_ON(!buffer_uptodate(bh)); + if (state != GFS2_BLKST_UNLINKED && bi->bi_clone) + buffer = bi->bi_clone + bi->bi_offset; + initial_offset = rbm->offset; + offset = gfs2_bitfit(buffer, bi->bi_len, rbm->offset, state); + if (offset == BFITNOENT) + goto bitmap_full; + rbm->offset = offset; + if (ip == NULL) + return 0; + + initial_bii = rbm->bii; + ret = gfs2_reservation_check_and_update(rbm, ip, + minext ? *minext : 0, + &maxext); + if (ret == 0) + return 0; + if (ret > 0) { + n += (rbm->bii - initial_bii); + goto next_iter; + } + if (ret == -E2BIG) { + rbm->bii = 0; + rbm->offset = 0; + n += (rbm->bii - initial_bii); + goto res_covered_end_of_rgrp; + } + return ret; + +bitmap_full: /* Mark bitmap as full and fall through */ + if ((state == GFS2_BLKST_FREE) && initial_offset == 0) { + struct gfs2_bitmap *bi = rbm_bi(rbm); + set_bit(GBF_FULL, &bi->bi_flags); + } + +next_bitmap: /* Find next bitmap in the rgrp */ + rbm->offset = 0; + rbm->bii++; + if (rbm->bii == rbm->rgd->rd_length) + rbm->bii = 0; +res_covered_end_of_rgrp: + if ((rbm->bii == 0) && nowrap) + break; + n++; +next_iter: + if (n >= iters) + break; } - spin_unlock(&sdp->sd_rindex_spin); - return ret; + if (minext == NULL || state != GFS2_BLKST_FREE) + return -ENOSPC; + + /* If the extent was too small, and it's smaller than the smallest + to have failed before, remember for future reference that it's + useless to search this rgrp again for this amount or more. */ + if ((first_offset == 0) && (first_bii == 0) && + (*minext < rbm->rgd->rd_extfail_pt)) + rbm->rgd->rd_extfail_pt = *minext; + + /* If the maximum extent we found is big enough to fulfill the + minimum requirements, use it anyway. */ + if (maxext.len) { + *rbm = maxext.rbm; + *minext = maxext.len; + return 0; + } + + return -ENOSPC; } /** * try_rgrp_unlink - Look for any unlinked, allocated, but unused inodes * @rgd: The rgrp + * @last_unlinked: block address of the last dinode we unlinked + * @skip: block address we should explicitly not unlink * * Returns: 0 if no error * The inode, if one has been found, in inode. @@ -925,34 +1724,34 @@ static int try_rgrp_fit(struct gfs2_rgrpd *rgd, struct gfs2_alloc *al) static void try_rgrp_unlink(struct gfs2_rgrpd *rgd, u64 *last_unlinked, u64 skip) { - u32 goal = 0, block; - u64 no_addr; + u64 block; struct gfs2_sbd *sdp = rgd->rd_sbd; - unsigned int n; struct gfs2_glock *gl; struct gfs2_inode *ip; int error; int found = 0; + struct gfs2_rbm rbm = { .rgd = rgd, .bii = 0, .offset = 0 }; - while (goal < rgd->rd_data) { + while (1) { down_write(&sdp->sd_log_flush_lock); - n = 1; - block = rgblk_search(rgd, goal, GFS2_BLKST_UNLINKED, - GFS2_BLKST_UNLINKED, &n); + error = gfs2_rbm_find(&rbm, GFS2_BLKST_UNLINKED, NULL, NULL, + true, NULL); up_write(&sdp->sd_log_flush_lock); - if (block == BFITNOENT) + if (error == -ENOSPC) break; - /* rgblk_search can return a block < goal, so we need to - keep it marching forward. */ - no_addr = block + rgd->rd_data0; - goal++; - if (*last_unlinked != NO_BLOCK && no_addr <= *last_unlinked) + if (WARN_ON_ONCE(error)) + break; + + block = gfs2_rbm_to_block(&rbm); + if (gfs2_rbm_from_block(&rbm, block + 1)) + break; + if (*last_unlinked != NO_BLOCK && block <= *last_unlinked) continue; - if (no_addr == skip) + if (block == skip) continue; - *last_unlinked = no_addr; + *last_unlinked = block; - error = gfs2_glock_get(sdp, no_addr, &gfs2_inode_glops, CREATE, &gl); + error = gfs2_glock_get(sdp, block, &gfs2_inode_glops, CREATE, &gl); if (error) continue; @@ -971,7 +1770,7 @@ static void try_rgrp_unlink(struct gfs2_rgrpd *rgd, u64 *last_unlinked, u64 skip found++; /* Limit reclaim to sensible number of tasks */ - if (found > 2*NR_CPUS) + if (found > NR_CPUS) return; } @@ -980,244 +1779,232 @@ static void try_rgrp_unlink(struct gfs2_rgrpd *rgd, u64 *last_unlinked, u64 skip } /** - * recent_rgrp_next - get next RG from "recent" list - * @cur_rgd: current rgrp + * gfs2_rgrp_congested - Use stats to figure out whether an rgrp is congested + * @rgd: The rgrp in question + * @loops: An indication of how picky we can be (0=very, 1=less so) + * + * This function uses the recently added glock statistics in order to + * figure out whether a parciular resource group is suffering from + * contention from multiple nodes. This is done purely on the basis + * of timings, since this is the only data we have to work with and + * our aim here is to reject a resource group which is highly contended + * but (very important) not to do this too often in order to ensure that + * we do not land up introducing fragmentation by changing resource + * groups when not actually required. + * + * The calculation is fairly simple, we want to know whether the SRTTB + * (i.e. smoothed round trip time for blocking operations) to acquire + * the lock for this rgrp's glock is significantly greater than the + * time taken for resource groups on average. We introduce a margin in + * the form of the variable @var which is computed as the sum of the two + * respective variences, and multiplied by a factor depending on @loops + * and whether we have a lot of data to base the decision on. This is + * then tested against the square difference of the means in order to + * decide whether the result is statistically significant or not. * - * Returns: The next rgrp in the recent list + * Returns: A boolean verdict on the congestion status */ -static struct gfs2_rgrpd *recent_rgrp_next(struct gfs2_rgrpd *cur_rgd) +static bool gfs2_rgrp_congested(const struct gfs2_rgrpd *rgd, int loops) { - struct gfs2_sbd *sdp = cur_rgd->rd_sbd; - struct list_head *head; - struct gfs2_rgrpd *rgd; - - spin_lock(&sdp->sd_rindex_spin); - head = &sdp->sd_rindex_mru_list; - if (unlikely(cur_rgd->rd_list_mru.next == head)) { - spin_unlock(&sdp->sd_rindex_spin); - return NULL; - } - rgd = list_entry(cur_rgd->rd_list_mru.next, struct gfs2_rgrpd, rd_list_mru); - spin_unlock(&sdp->sd_rindex_spin); - return rgd; + const struct gfs2_glock *gl = rgd->rd_gl; + const struct gfs2_sbd *sdp = gl->gl_sbd; + struct gfs2_lkstats *st; + s64 r_dcount, l_dcount; + s64 r_srttb, l_srttb; + s64 srttb_diff; + s64 sqr_diff; + s64 var; + + preempt_disable(); + st = &this_cpu_ptr(sdp->sd_lkstats)->lkstats[LM_TYPE_RGRP]; + r_srttb = st->stats[GFS2_LKS_SRTTB]; + r_dcount = st->stats[GFS2_LKS_DCOUNT]; + var = st->stats[GFS2_LKS_SRTTVARB] + + gl->gl_stats.stats[GFS2_LKS_SRTTVARB]; + preempt_enable(); + + l_srttb = gl->gl_stats.stats[GFS2_LKS_SRTTB]; + l_dcount = gl->gl_stats.stats[GFS2_LKS_DCOUNT]; + + if ((l_dcount < 1) || (r_dcount < 1) || (r_srttb == 0)) + return false; + + srttb_diff = r_srttb - l_srttb; + sqr_diff = srttb_diff * srttb_diff; + + var *= 2; + if (l_dcount < 8 || r_dcount < 8) + var *= 2; + if (loops == 1) + var *= 2; + + return ((srttb_diff < 0) && (sqr_diff > var)); } /** - * forward_rgrp_get - get an rgrp to try next from full list - * @sdp: The GFS2 superblock + * gfs2_rgrp_used_recently + * @rs: The block reservation with the rgrp to test + * @msecs: The time limit in milliseconds * - * Returns: The rgrp to try next + * Returns: True if the rgrp glock has been used within the time limit */ - -static struct gfs2_rgrpd *forward_rgrp_get(struct gfs2_sbd *sdp) +static bool gfs2_rgrp_used_recently(const struct gfs2_blkreserv *rs, + u64 msecs) { - struct gfs2_rgrpd *rgd; - unsigned int journals = gfs2_jindex_size(sdp); - unsigned int rg = 0, x; - - spin_lock(&sdp->sd_rindex_spin); - - rgd = sdp->sd_rindex_forward; - if (!rgd) { - if (sdp->sd_rgrps >= journals) - rg = sdp->sd_rgrps * sdp->sd_jdesc->jd_jid / journals; + u64 tdiff; - for (x = 0, rgd = gfs2_rgrpd_get_first(sdp); x < rg; - x++, rgd = gfs2_rgrpd_get_next(rgd)) - /* Do Nothing */; + tdiff = ktime_to_ns(ktime_sub(ktime_get_real(), + rs->rs_rbm.rgd->rd_gl->gl_dstamp)); - sdp->sd_rindex_forward = rgd; - } + return tdiff > (msecs * 1000 * 1000); +} - spin_unlock(&sdp->sd_rindex_spin); +static u32 gfs2_orlov_skip(const struct gfs2_inode *ip) +{ + const struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); + u32 skip; - return rgd; + get_random_bytes(&skip, sizeof(skip)); + return skip % sdp->sd_rgrps; } -/** - * forward_rgrp_set - set the forward rgrp pointer - * @sdp: the filesystem - * @rgd: The new forward rgrp - * - */ - -static void forward_rgrp_set(struct gfs2_sbd *sdp, struct gfs2_rgrpd *rgd) +static bool gfs2_select_rgrp(struct gfs2_rgrpd **pos, const struct gfs2_rgrpd *begin) { - spin_lock(&sdp->sd_rindex_spin); - sdp->sd_rindex_forward = rgd; - spin_unlock(&sdp->sd_rindex_spin); + struct gfs2_rgrpd *rgd = *pos; + struct gfs2_sbd *sdp = rgd->rd_sbd; + + rgd = gfs2_rgrpd_get_next(rgd); + if (rgd == NULL) + rgd = gfs2_rgrpd_get_first(sdp); + *pos = rgd; + if (rgd != begin) /* If we didn't wrap */ + return true; + return false; } /** - * get_local_rgrp - Choose and lock a rgrp for allocation + * gfs2_inplace_reserve - Reserve space in the filesystem * @ip: the inode to reserve space for - * @rgp: the chosen and locked rgrp - * - * Try to acquire rgrp in way which avoids contending with others. + * @ap: the allocation parameters * * Returns: errno */ -static int get_local_rgrp(struct gfs2_inode *ip, u64 *last_unlinked) +int gfs2_inplace_reserve(struct gfs2_inode *ip, const struct gfs2_alloc_parms *ap) { struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); - struct gfs2_rgrpd *rgd, *begin = NULL; - struct gfs2_alloc *al = ip->i_alloc; - int flags = LM_FLAG_TRY; - int skipped = 0; + struct gfs2_rgrpd *begin = NULL; + struct gfs2_blkreserv *rs = ip->i_res; + int error = 0, rg_locked, flags = 0; + u64 last_unlinked = NO_BLOCK; int loops = 0; - int error, rg_locked; - - rgd = gfs2_blk2rgrpd(sdp, ip->i_goal); - - while (rgd) { - rg_locked = 0; - - if (gfs2_glock_is_locked_by_me(rgd->rd_gl)) { - rg_locked = 1; - error = 0; - } else { - error = gfs2_glock_nq_init(rgd->rd_gl, LM_ST_EXCLUSIVE, - LM_FLAG_TRY, &al->al_rgd_gh); - } - switch (error) { - case 0: - if (try_rgrp_fit(rgd, al)) - goto out; - if (rgd->rd_flags & GFS2_RDF_CHECK) - try_rgrp_unlink(rgd, last_unlinked, ip->i_no_addr); - if (!rg_locked) - gfs2_glock_dq_uninit(&al->al_rgd_gh); - /* fall through */ - case GLR_TRYFAILED: - rgd = recent_rgrp_next(rgd); - break; + u32 skip = 0; - default: - return error; - } + if (sdp->sd_args.ar_rgrplvb) + flags |= GL_SKIP; + if (gfs2_assert_warn(sdp, ap->target)) + return -EINVAL; + if (gfs2_rs_active(rs)) { + begin = rs->rs_rbm.rgd; + } else if (ip->i_rgd && rgrp_contains_block(ip->i_rgd, ip->i_goal)) { + rs->rs_rbm.rgd = begin = ip->i_rgd; + } else { + rs->rs_rbm.rgd = begin = gfs2_blk2rgrpd(sdp, ip->i_goal, 1); } - - /* Go through full list of rgrps */ - - begin = rgd = forward_rgrp_get(sdp); - - for (;;) { - rg_locked = 0; - - if (gfs2_glock_is_locked_by_me(rgd->rd_gl)) { - rg_locked = 1; - error = 0; - } else { - error = gfs2_glock_nq_init(rgd->rd_gl, LM_ST_EXCLUSIVE, flags, - &al->al_rgd_gh); + if (S_ISDIR(ip->i_inode.i_mode) && (ap->aflags & GFS2_AF_ORLOV)) + skip = gfs2_orlov_skip(ip); + if (rs->rs_rbm.rgd == NULL) + return -EBADSLT; + + while (loops < 3) { + rg_locked = 1; + + if (!gfs2_glock_is_locked_by_me(rs->rs_rbm.rgd->rd_gl)) { + rg_locked = 0; + if (skip && skip--) + goto next_rgrp; + if (!gfs2_rs_active(rs) && (loops < 2) && + gfs2_rgrp_used_recently(rs, 1000) && + gfs2_rgrp_congested(rs->rs_rbm.rgd, loops)) + goto next_rgrp; + error = gfs2_glock_nq_init(rs->rs_rbm.rgd->rd_gl, + LM_ST_EXCLUSIVE, flags, + &rs->rs_rgd_gh); + if (unlikely(error)) + return error; + if (!gfs2_rs_active(rs) && (loops < 2) && + gfs2_rgrp_congested(rs->rs_rbm.rgd, loops)) + goto skip_rgrp; + if (sdp->sd_args.ar_rgrplvb) { + error = update_rgrp_lvb(rs->rs_rbm.rgd); + if (unlikely(error)) { + gfs2_glock_dq_uninit(&rs->rs_rgd_gh); + return error; + } + } } - switch (error) { - case 0: - if (try_rgrp_fit(rgd, al)) - goto out; - if (rgd->rd_flags & GFS2_RDF_CHECK) - try_rgrp_unlink(rgd, last_unlinked, ip->i_no_addr); - if (!rg_locked) - gfs2_glock_dq_uninit(&al->al_rgd_gh); - break; - case GLR_TRYFAILED: - skipped++; - break; + /* Skip unuseable resource groups */ + if ((rs->rs_rbm.rgd->rd_flags & (GFS2_RGF_NOALLOC | + GFS2_RDF_ERROR)) || + (ap->target > rs->rs_rbm.rgd->rd_extfail_pt)) + goto skip_rgrp; - default: - return error; - } + if (sdp->sd_args.ar_rgrplvb) + gfs2_rgrp_bh_get(rs->rs_rbm.rgd); - rgd = gfs2_rgrpd_get_next(rgd); - if (!rgd) - rgd = gfs2_rgrpd_get_first(sdp); - - if (rgd == begin) { - if (++loops >= 3) - return -ENOSPC; - if (!skipped) - loops++; - flags = 0; - if (loops == 2) - gfs2_log_flush(sdp, NULL); - } - } + /* Get a reservation if we don't already have one */ + if (!gfs2_rs_active(rs)) + rg_mblk_search(rs->rs_rbm.rgd, ip, ap); -out: - if (begin) { - spin_lock(&sdp->sd_rindex_spin); - list_move(&rgd->rd_list_mru, &sdp->sd_rindex_mru_list); - spin_unlock(&sdp->sd_rindex_spin); - rgd = gfs2_rgrpd_get_next(rgd); - if (!rgd) - rgd = gfs2_rgrpd_get_first(sdp); - forward_rgrp_set(sdp, rgd); - } - - return 0; -} + /* Skip rgrps when we can't get a reservation on first pass */ + if (!gfs2_rs_active(rs) && (loops < 1)) + goto check_rgrp; -/** - * gfs2_inplace_reserve_i - Reserve space in the filesystem - * @ip: the inode to reserve space for - * - * Returns: errno - */ - -int gfs2_inplace_reserve_i(struct gfs2_inode *ip, int hold_rindex, - char *file, unsigned int line) -{ - struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); - struct gfs2_alloc *al = ip->i_alloc; - int error = 0; - u64 last_unlinked = NO_BLOCK; - int tries = 0; + /* If rgrp has enough free space, use it */ + if (rs->rs_rbm.rgd->rd_free_clone >= ap->target) { + ip->i_rgd = rs->rs_rbm.rgd; + return 0; + } - if (gfs2_assert_warn(sdp, al->al_requested)) - return -EINVAL; +check_rgrp: + /* Check for unlinked inodes which can be reclaimed */ + if (rs->rs_rbm.rgd->rd_flags & GFS2_RDF_CHECK) + try_rgrp_unlink(rs->rs_rbm.rgd, &last_unlinked, + ip->i_no_addr); +skip_rgrp: + /* Drop reservation, if we couldn't use reserved rgrp */ + if (gfs2_rs_active(rs)) + gfs2_rs_deltree(rs); + + /* Unlock rgrp if required */ + if (!rg_locked) + gfs2_glock_dq_uninit(&rs->rs_rgd_gh); +next_rgrp: + /* Find the next rgrp, and continue looking */ + if (gfs2_select_rgrp(&rs->rs_rbm.rgd, begin)) + continue; + if (skip) + continue; - if (hold_rindex) { - /* We need to hold the rindex unless the inode we're using is - the rindex itself, in which case it's already held. */ - if (ip != GFS2_I(sdp->sd_rindex)) - error = gfs2_rindex_hold(sdp, &al->al_ri_gh); - else if (!sdp->sd_rgrps) /* We may not have the rindex read - in, so: */ + /* If we've scanned all the rgrps, but found no free blocks + * then this checks for some less likely conditions before + * trying again. + */ + loops++; + /* Check that fs hasn't grown if writing to rindex */ + if (ip == GFS2_I(sdp->sd_rindex) && !sdp->sd_rindex_uptodate) { error = gfs2_ri_update(ip); - if (error) - return error; - } - -try_again: - do { - error = get_local_rgrp(ip, &last_unlinked); - /* If there is no space, flushing the log may release some */ - if (error) { - if (ip == GFS2_I(sdp->sd_rindex) && - !sdp->sd_rindex_uptodate) { - error = gfs2_ri_update(ip); - if (error) - return error; - goto try_again; - } - gfs2_log_flush(sdp, NULL); + if (error) + return error; } - } while (error && tries++ < 3); - - if (error) { - if (hold_rindex && ip != GFS2_I(sdp->sd_rindex)) - gfs2_glock_dq_uninit(&al->al_ri_gh); - return error; + /* Flushing the log may release space */ + if (loops == 2) + gfs2_log_flush(sdp, NULL, NORMAL_FLUSH); } - /* no error, so we have the rgrp set in the inode's allocation. */ - al->al_file = file; - al->al_line = line; - - return 0; + return -ENOSPC; } /** @@ -1229,20 +2016,10 @@ try_again: void gfs2_inplace_release(struct gfs2_inode *ip) { - struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); - struct gfs2_alloc *al = ip->i_alloc; - - if (gfs2_assert_warn(sdp, al->al_alloced <= al->al_requested) == -1) - fs_warn(sdp, "al_alloced = %u, al_requested = %u " - "al_file = %s, al_line = %u\n", - al->al_alloced, al->al_requested, al->al_file, - al->al_line); + struct gfs2_blkreserv *rs = ip->i_res; - al->al_rgd = NULL; - if (al->al_rgd_gh.gh_gl) - gfs2_glock_dq_uninit(&al->al_rgd_gh); - if (ip != GFS2_I(sdp->sd_rindex) && al->al_ri_gh.gh_gl) - gfs2_glock_dq_uninit(&al->al_ri_gh); + if (rs->rs_rgd_gh.gh_gl) + gfs2_glock_dq_uninit(&rs->rs_rgd_gh); } /** @@ -1255,131 +2032,47 @@ void gfs2_inplace_release(struct gfs2_inode *ip) static unsigned char gfs2_get_block_type(struct gfs2_rgrpd *rgd, u64 block) { - struct gfs2_bitmap *bi = NULL; - u32 length, rgrp_block, buf_block; - unsigned int buf; - unsigned char type; - - length = rgd->rd_length; - rgrp_block = block - rgd->rd_data0; - - for (buf = 0; buf < length; buf++) { - bi = rgd->rd_bits + buf; - if (rgrp_block < (bi->bi_start + bi->bi_len) * GFS2_NBBY) - break; - } - - gfs2_assert(rgd->rd_sbd, buf < length); - buf_block = rgrp_block - bi->bi_start * GFS2_NBBY; + struct gfs2_rbm rbm = { .rgd = rgd, }; + int ret; - type = gfs2_testbit(rgd, bi->bi_bh->b_data + bi->bi_offset, - bi->bi_len, buf_block); + ret = gfs2_rbm_from_block(&rbm, block); + WARN_ON_ONCE(ret != 0); - return type; + return gfs2_testbit(&rbm); } + /** - * rgblk_search - find a block in @old_state, change allocation - * state to @new_state - * @rgd: the resource group descriptor - * @goal: the goal block within the RG (start here to search for avail block) - * @old_state: GFS2_BLKST_XXX the before-allocation state to find - * @new_state: GFS2_BLKST_XXX the after-allocation block state - * @n: The extent length + * gfs2_alloc_extent - allocate an extent from a given bitmap + * @rbm: the resource group information + * @dinode: TRUE if the first block we allocate is for a dinode + * @n: The extent length (value/result) * - * Walk rgrp's bitmap to find bits that represent a block in @old_state. - * Add the found bitmap buffer to the transaction. + * Add the bitmap buffer to the transaction. * Set the found bits to @new_state to change block's allocation state. - * - * This function never fails, because we wouldn't call it unless we - * know (from reservation results, etc.) that a block is available. - * - * Scope of @goal and returned block is just within rgrp, not the whole - * filesystem. - * - * Returns: the block number allocated */ - -static u32 rgblk_search(struct gfs2_rgrpd *rgd, u32 goal, - unsigned char old_state, unsigned char new_state, - unsigned int *n) +static void gfs2_alloc_extent(const struct gfs2_rbm *rbm, bool dinode, + unsigned int *n) { - struct gfs2_bitmap *bi = NULL; - const u32 length = rgd->rd_length; - u32 blk = BFITNOENT; - unsigned int buf, x; + struct gfs2_rbm pos = { .rgd = rbm->rgd, }; const unsigned int elen = *n; - const u8 *buffer = NULL; - - *n = 0; - /* Find bitmap block that contains bits for goal block */ - for (buf = 0; buf < length; buf++) { - bi = rgd->rd_bits + buf; - /* Convert scope of "goal" from rgrp-wide to within found bit block */ - if (goal < (bi->bi_start + bi->bi_len) * GFS2_NBBY) { - goal -= bi->bi_start * GFS2_NBBY; - goto do_search; - } - } - buf = 0; - goal = 0; - -do_search: - /* Search (up to entire) bitmap in this rgrp for allocatable block. - "x <= length", instead of "x < length", because we typically start - the search in the middle of a bit block, but if we can't find an - allocatable block anywhere else, we want to be able wrap around and - search in the first part of our first-searched bit block. */ - for (x = 0; x <= length; x++) { - bi = rgd->rd_bits + buf; - - if (test_bit(GBF_FULL, &bi->bi_flags) && - (old_state == GFS2_BLKST_FREE)) - goto skip; - - /* The GFS2_BLKST_UNLINKED state doesn't apply to the clone - bitmaps, so we must search the originals for that. */ - buffer = bi->bi_bh->b_data + bi->bi_offset; - if (old_state != GFS2_BLKST_UNLINKED && bi->bi_clone) - buffer = bi->bi_clone + bi->bi_offset; - - blk = gfs2_bitfit(buffer, bi->bi_len, goal, old_state); - if (blk != BFITNOENT) - break; - - if ((goal == 0) && (old_state == GFS2_BLKST_FREE)) - set_bit(GBF_FULL, &bi->bi_flags); - - /* Try next bitmap block (wrap back to rgrp header if at end) */ -skip: - buf++; - buf %= length; - goal = 0; - } + u64 block; + int ret; - if (blk == BFITNOENT) - return blk; *n = 1; - if (old_state == new_state) - goto out; - - gfs2_trans_add_bh(rgd->rd_gl, bi->bi_bh, 1); - gfs2_setbit(rgd, bi->bi_bh->b_data, bi->bi_clone, bi->bi_offset, - bi->bi_len, blk, new_state); - goal = blk; + block = gfs2_rbm_to_block(rbm); + gfs2_trans_add_meta(rbm->rgd->rd_gl, rbm_bi(rbm)->bi_bh); + gfs2_setbit(rbm, true, dinode ? GFS2_BLKST_DINODE : GFS2_BLKST_USED); + block++; while (*n < elen) { - goal++; - if (goal >= (bi->bi_len * GFS2_NBBY)) - break; - if (gfs2_testbit(rgd, buffer, bi->bi_len, goal) != - GFS2_BLKST_FREE) + ret = gfs2_rbm_from_block(&pos, block); + if (ret || gfs2_testbit(&pos) != GFS2_BLKST_FREE) break; - gfs2_setbit(rgd, bi->bi_bh->b_data, bi->bi_clone, bi->bi_offset, - bi->bi_len, goal, new_state); + gfs2_trans_add_meta(pos.rgd->rd_gl, rbm_bi(&pos)->bi_bh); + gfs2_setbit(&pos, true, GFS2_BLKST_USED); (*n)++; + block++; } -out: - return (bi->bi_start * GFS2_NBBY) + blk; } /** @@ -1395,47 +2088,31 @@ out: static struct gfs2_rgrpd *rgblk_free(struct gfs2_sbd *sdp, u64 bstart, u32 blen, unsigned char new_state) { - struct gfs2_rgrpd *rgd; - struct gfs2_bitmap *bi = NULL; - u32 length, rgrp_blk, buf_blk; - unsigned int buf; + struct gfs2_rbm rbm; + struct gfs2_bitmap *bi; - rgd = gfs2_blk2rgrpd(sdp, bstart); - if (!rgd) { + rbm.rgd = gfs2_blk2rgrpd(sdp, bstart, 1); + if (!rbm.rgd) { if (gfs2_consist(sdp)) fs_err(sdp, "block = %llu\n", (unsigned long long)bstart); return NULL; } - length = rgd->rd_length; - - rgrp_blk = bstart - rgd->rd_data0; - while (blen--) { - for (buf = 0; buf < length; buf++) { - bi = rgd->rd_bits + buf; - if (rgrp_blk < (bi->bi_start + bi->bi_len) * GFS2_NBBY) - break; - } - - gfs2_assert(rgd->rd_sbd, buf < length); - - buf_blk = rgrp_blk - bi->bi_start * GFS2_NBBY; - rgrp_blk++; - + gfs2_rbm_from_block(&rbm, bstart); + bi = rbm_bi(&rbm); + bstart++; if (!bi->bi_clone) { bi->bi_clone = kmalloc(bi->bi_bh->b_size, GFP_NOFS | __GFP_NOFAIL); memcpy(bi->bi_clone + bi->bi_offset, - bi->bi_bh->b_data + bi->bi_offset, - bi->bi_len); + bi->bi_bh->b_data + bi->bi_offset, bi->bi_len); } - gfs2_trans_add_bh(rgd->rd_gl, bi->bi_bh, 1); - gfs2_setbit(rgd, bi->bi_bh->b_data, NULL, bi->bi_offset, - bi->bi_len, buf_blk, new_state); + gfs2_trans_add_meta(rbm.rgd->rd_gl, bi->bi_bh); + gfs2_setbit(&rbm, false, new_state); } - return rgd; + return rbm.rgd; } /** @@ -1445,15 +2122,24 @@ static struct gfs2_rgrpd *rgblk_free(struct gfs2_sbd *sdp, u64 bstart, * */ -int gfs2_rgrp_dump(struct seq_file *seq, const struct gfs2_glock *gl) +void gfs2_rgrp_dump(struct seq_file *seq, const struct gfs2_glock *gl) { - const struct gfs2_rgrpd *rgd = gl->gl_object; + struct gfs2_rgrpd *rgd = gl->gl_object; + struct gfs2_blkreserv *trs; + const struct rb_node *n; + if (rgd == NULL) - return 0; - gfs2_print_dbg(seq, " R: n:%llu f:%02x b:%u/%u i:%u\n", + return; + gfs2_print_dbg(seq, " R: n:%llu f:%02x b:%u/%u i:%u r:%u e:%u\n", (unsigned long long)rgd->rd_addr, rgd->rd_flags, - rgd->rd_free, rgd->rd_free_clone, rgd->rd_dinodes); - return 0; + rgd->rd_free, rgd->rd_free_clone, rgd->rd_dinodes, + rgd->rd_reserved, rgd->rd_extfail_pt); + spin_lock(&rgd->rd_rsspin); + for (n = rb_first(&rgd->rd_rstree); n; n = rb_next(&trs->rs_node)) { + trs = rb_entry(n, struct gfs2_blkreserv, rs_node); + dump_rs(seq, trs); + } + spin_unlock(&rgd->rd_rsspin); } static void gfs2_rgrp_error(struct gfs2_rgrpd *rgd) @@ -1467,142 +2153,176 @@ static void gfs2_rgrp_error(struct gfs2_rgrpd *rgd) } /** - * gfs2_alloc_block - Allocate one or more blocks - * @ip: the inode to allocate the block for - * @bn: Used to return the starting block number - * @n: requested number of blocks/extent length (value/result) + * gfs2_adjust_reservation - Adjust (or remove) a reservation after allocation + * @ip: The inode we have just allocated blocks for + * @rbm: The start of the allocated blocks + * @len: The extent length * - * Returns: 0 or error + * Adjusts a reservation after an allocation has taken place. If the + * reservation does not match the allocation, or if it is now empty + * then it is removed. */ -int gfs2_alloc_block(struct gfs2_inode *ip, u64 *bn, unsigned int *n) +static void gfs2_adjust_reservation(struct gfs2_inode *ip, + const struct gfs2_rbm *rbm, unsigned len) { - struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); - struct buffer_head *dibh; - struct gfs2_alloc *al = ip->i_alloc; - struct gfs2_rgrpd *rgd; - u32 goal, blk; + struct gfs2_blkreserv *rs = ip->i_res; + struct gfs2_rgrpd *rgd = rbm->rgd; + unsigned rlen; u64 block; - int error; - - /* Only happens if there is a bug in gfs2, return something distinctive - * to ensure that it is noticed. - */ - if (al == NULL) - return -ECANCELED; - - rgd = al->al_rgd; - - if (rgrp_contains_block(rgd, ip->i_goal)) - goal = ip->i_goal - rgd->rd_data0; - else - goal = rgd->rd_last_alloc; - - blk = rgblk_search(rgd, goal, GFS2_BLKST_FREE, GFS2_BLKST_USED, n); - - /* Since all blocks are reserved in advance, this shouldn't happen */ - if (blk == BFITNOENT) - goto rgrp_error; - - rgd->rd_last_alloc = blk; - block = rgd->rd_data0 + blk; - ip->i_goal = block; - error = gfs2_meta_inode_buffer(ip, &dibh); - if (error == 0) { - struct gfs2_dinode *di = (struct gfs2_dinode *)dibh->b_data; - gfs2_trans_add_bh(ip->i_gl, dibh, 1); - di->di_goal_meta = di->di_goal_data = cpu_to_be64(ip->i_goal); - brelse(dibh); + int ret; + + spin_lock(&rgd->rd_rsspin); + if (gfs2_rs_active(rs)) { + if (gfs2_rbm_eq(&rs->rs_rbm, rbm)) { + block = gfs2_rbm_to_block(rbm); + ret = gfs2_rbm_from_block(&rs->rs_rbm, block + len); + rlen = min(rs->rs_free, len); + rs->rs_free -= rlen; + rgd->rd_reserved -= rlen; + trace_gfs2_rs(rs, TRACE_RS_CLAIM); + if (rs->rs_free && !ret) + goto out; + } + __rs_deltree(rs); } - if (rgd->rd_free < *n) - goto rgrp_error; - - rgd->rd_free -= *n; +out: + spin_unlock(&rgd->rd_rsspin); +} - gfs2_trans_add_bh(rgd->rd_gl, rgd->rd_bits[0].bi_bh, 1); - gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data); +/** + * gfs2_set_alloc_start - Set starting point for block allocation + * @rbm: The rbm which will be set to the required location + * @ip: The gfs2 inode + * @dinode: Flag to say if allocation includes a new inode + * + * This sets the starting point from the reservation if one is active + * otherwise it falls back to guessing a start point based on the + * inode's goal block or the last allocation point in the rgrp. + */ - al->al_alloced += *n; +static void gfs2_set_alloc_start(struct gfs2_rbm *rbm, + const struct gfs2_inode *ip, bool dinode) +{ + u64 goal; - gfs2_statfs_change(sdp, 0, -(s64)*n, 0); - gfs2_quota_change(ip, *n, ip->i_inode.i_uid, ip->i_inode.i_gid); + if (gfs2_rs_active(ip->i_res)) { + *rbm = ip->i_res->rs_rbm; + return; + } - spin_lock(&sdp->sd_rindex_spin); - rgd->rd_free_clone -= *n; - spin_unlock(&sdp->sd_rindex_spin); - trace_gfs2_block_alloc(ip, block, *n, GFS2_BLKST_USED); - *bn = block; - return 0; + if (!dinode && rgrp_contains_block(rbm->rgd, ip->i_goal)) + goal = ip->i_goal; + else + goal = rbm->rgd->rd_last_alloc + rbm->rgd->rd_data0; -rgrp_error: - gfs2_rgrp_error(rgd); - return -EIO; + gfs2_rbm_from_block(rbm, goal); } /** - * gfs2_alloc_di - Allocate a dinode - * @dip: the directory that the inode is going in - * @bn: the block number which is allocated + * gfs2_alloc_blocks - Allocate one or more blocks of data and/or a dinode + * @ip: the inode to allocate the block for + * @bn: Used to return the starting block number + * @nblocks: requested number of blocks/extent length (value/result) + * @dinode: 1 if we're allocating a dinode block, else 0 * @generation: the generation number of the inode * - * Returns: 0 on success or error + * Returns: 0 or error */ -int gfs2_alloc_di(struct gfs2_inode *dip, u64 *bn, u64 *generation) +int gfs2_alloc_blocks(struct gfs2_inode *ip, u64 *bn, unsigned int *nblocks, + bool dinode, u64 *generation) { - struct gfs2_sbd *sdp = GFS2_SB(&dip->i_inode); - struct gfs2_alloc *al = dip->i_alloc; - struct gfs2_rgrpd *rgd = al->al_rgd; - u32 blk; - u64 block; - unsigned int n = 1; + struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); + struct buffer_head *dibh; + struct gfs2_rbm rbm = { .rgd = ip->i_rgd, }; + unsigned int ndata; + u64 block; /* block, within the file system scope */ + int error; - blk = rgblk_search(rgd, rgd->rd_last_alloc, - GFS2_BLKST_FREE, GFS2_BLKST_DINODE, &n); + gfs2_set_alloc_start(&rbm, ip, dinode); + error = gfs2_rbm_find(&rbm, GFS2_BLKST_FREE, NULL, ip, false, NULL); + + if (error == -ENOSPC) { + gfs2_set_alloc_start(&rbm, ip, dinode); + error = gfs2_rbm_find(&rbm, GFS2_BLKST_FREE, NULL, NULL, false, + NULL); + } /* Since all blocks are reserved in advance, this shouldn't happen */ - if (blk == BFITNOENT) + if (error) { + fs_warn(sdp, "inum=%llu error=%d, nblocks=%u, full=%d fail_pt=%d\n", + (unsigned long long)ip->i_no_addr, error, *nblocks, + test_bit(GBF_FULL, &rbm.rgd->rd_bits->bi_flags), + rbm.rgd->rd_extfail_pt); goto rgrp_error; + } - rgd->rd_last_alloc = blk; - block = rgd->rd_data0 + blk; - if (rgd->rd_free == 0) + gfs2_alloc_extent(&rbm, dinode, nblocks); + block = gfs2_rbm_to_block(&rbm); + rbm.rgd->rd_last_alloc = block - rbm.rgd->rd_data0; + if (gfs2_rs_active(ip->i_res)) + gfs2_adjust_reservation(ip, &rbm, *nblocks); + ndata = *nblocks; + if (dinode) + ndata--; + + if (!dinode) { + ip->i_goal = block + ndata - 1; + error = gfs2_meta_inode_buffer(ip, &dibh); + if (error == 0) { + struct gfs2_dinode *di = + (struct gfs2_dinode *)dibh->b_data; + gfs2_trans_add_meta(ip->i_gl, dibh); + di->di_goal_meta = di->di_goal_data = + cpu_to_be64(ip->i_goal); + brelse(dibh); + } + } + if (rbm.rgd->rd_free < *nblocks) { + pr_warn("nblocks=%u\n", *nblocks); goto rgrp_error; + } - rgd->rd_free--; - rgd->rd_dinodes++; - *generation = rgd->rd_igeneration++; - if (*generation == 0) - *generation = rgd->rd_igeneration++; - gfs2_trans_add_bh(rgd->rd_gl, rgd->rd_bits[0].bi_bh, 1); - gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data); + rbm.rgd->rd_free -= *nblocks; + if (dinode) { + rbm.rgd->rd_dinodes++; + *generation = rbm.rgd->rd_igeneration++; + if (*generation == 0) + *generation = rbm.rgd->rd_igeneration++; + } - al->al_alloced++; + gfs2_trans_add_meta(rbm.rgd->rd_gl, rbm.rgd->rd_bits[0].bi_bh); + gfs2_rgrp_out(rbm.rgd, rbm.rgd->rd_bits[0].bi_bh->b_data); + gfs2_rgrp_ondisk2lvb(rbm.rgd->rd_rgl, rbm.rgd->rd_bits[0].bi_bh->b_data); - gfs2_statfs_change(sdp, 0, -1, +1); - gfs2_trans_add_unrevoke(sdp, block, 1); + gfs2_statfs_change(sdp, 0, -(s64)*nblocks, dinode ? 1 : 0); + if (dinode) + gfs2_trans_add_unrevoke(sdp, block, *nblocks); - spin_lock(&sdp->sd_rindex_spin); - rgd->rd_free_clone--; - spin_unlock(&sdp->sd_rindex_spin); - trace_gfs2_block_alloc(dip, block, 1, GFS2_BLKST_DINODE); + gfs2_quota_change(ip, *nblocks, ip->i_inode.i_uid, ip->i_inode.i_gid); + + rbm.rgd->rd_free_clone -= *nblocks; + trace_gfs2_block_alloc(ip, rbm.rgd, block, *nblocks, + dinode ? GFS2_BLKST_DINODE : GFS2_BLKST_USED); *bn = block; return 0; rgrp_error: - gfs2_rgrp_error(rgd); + gfs2_rgrp_error(rbm.rgd); return -EIO; } /** - * gfs2_free_data - free a contiguous run of data block(s) + * __gfs2_free_blocks - free a contiguous run of block(s) * @ip: the inode these blocks are being freed from * @bstart: first block of a run of contiguous blocks * @blen: the length of the block run + * @meta: 1 if the blocks represent metadata * */ -void gfs2_free_data(struct gfs2_inode *ip, u64 bstart, u32 blen) +void __gfs2_free_blocks(struct gfs2_inode *ip, u64 bstart, u32 blen, int meta) { struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); struct gfs2_rgrpd *rgd; @@ -1610,16 +2330,16 @@ void gfs2_free_data(struct gfs2_inode *ip, u64 bstart, u32 blen) rgd = rgblk_free(sdp, bstart, blen, GFS2_BLKST_FREE); if (!rgd) return; - trace_gfs2_block_alloc(ip, bstart, blen, GFS2_BLKST_FREE); + trace_gfs2_block_alloc(ip, rgd, bstart, blen, GFS2_BLKST_FREE); rgd->rd_free += blen; - - gfs2_trans_add_bh(rgd->rd_gl, rgd->rd_bits[0].bi_bh, 1); + rgd->rd_flags &= ~GFS2_RGF_TRIMMED; + gfs2_trans_add_meta(rgd->rd_gl, rgd->rd_bits[0].bi_bh); gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data); + gfs2_rgrp_ondisk2lvb(rgd->rd_rgl, rgd->rd_bits[0].bi_bh->b_data); - gfs2_trans_add_rg(rgd); - - gfs2_statfs_change(sdp, 0, +blen, 0); - gfs2_quota_change(ip, -(s64)blen, ip->i_inode.i_uid, ip->i_inode.i_gid); + /* Directories keep their data in the metadata address space */ + if (meta || ip->i_depth) + gfs2_meta_wipe(ip, bstart, blen); } /** @@ -1633,22 +2353,10 @@ void gfs2_free_data(struct gfs2_inode *ip, u64 bstart, u32 blen) void gfs2_free_meta(struct gfs2_inode *ip, u64 bstart, u32 blen) { struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); - struct gfs2_rgrpd *rgd; - - rgd = rgblk_free(sdp, bstart, blen, GFS2_BLKST_FREE); - if (!rgd) - return; - trace_gfs2_block_alloc(ip, bstart, blen, GFS2_BLKST_FREE); - rgd->rd_free += blen; - - gfs2_trans_add_bh(rgd->rd_gl, rgd->rd_bits[0].bi_bh, 1); - gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data); - - gfs2_trans_add_rg(rgd); + __gfs2_free_blocks(ip, bstart, blen, 1); gfs2_statfs_change(sdp, 0, +blen, 0); gfs2_quota_change(ip, -(s64)blen, ip->i_inode.i_uid, ip->i_inode.i_gid); - gfs2_meta_wipe(ip, bstart, blen); } void gfs2_unlink_di(struct inode *inode) @@ -1661,10 +2369,11 @@ void gfs2_unlink_di(struct inode *inode) rgd = rgblk_free(sdp, blkno, 1, GFS2_BLKST_UNLINKED); if (!rgd) return; - trace_gfs2_block_alloc(ip, blkno, 1, GFS2_BLKST_UNLINKED); - gfs2_trans_add_bh(rgd->rd_gl, rgd->rd_bits[0].bi_bh, 1); + trace_gfs2_block_alloc(ip, rgd, blkno, 1, GFS2_BLKST_UNLINKED); + gfs2_trans_add_meta(rgd->rd_gl, rgd->rd_bits[0].bi_bh); gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data); - gfs2_trans_add_rg(rgd); + gfs2_rgrp_ondisk2lvb(rgd->rd_rgl, rgd->rd_bits[0].bi_bh->b_data); + update_rgrp_lvb_unlinked(rgd, 1); } static void gfs2_free_uninit_di(struct gfs2_rgrpd *rgd, u64 blkno) @@ -1682,18 +2391,19 @@ static void gfs2_free_uninit_di(struct gfs2_rgrpd *rgd, u64 blkno) rgd->rd_dinodes--; rgd->rd_free++; - gfs2_trans_add_bh(rgd->rd_gl, rgd->rd_bits[0].bi_bh, 1); + gfs2_trans_add_meta(rgd->rd_gl, rgd->rd_bits[0].bi_bh); gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data); + gfs2_rgrp_ondisk2lvb(rgd->rd_rgl, rgd->rd_bits[0].bi_bh->b_data); + update_rgrp_lvb_unlinked(rgd, -1); gfs2_statfs_change(sdp, 0, +1, -1); - gfs2_trans_add_rg(rgd); } void gfs2_free_di(struct gfs2_rgrpd *rgd, struct gfs2_inode *ip) { gfs2_free_uninit_di(rgd, ip->i_no_addr); - trace_gfs2_block_alloc(ip, ip->i_no_addr, 1, GFS2_BLKST_FREE); + trace_gfs2_block_alloc(ip, rgd, ip->i_no_addr, 1, GFS2_BLKST_FREE); gfs2_quota_change(ip, -1, ip->i_inode.i_uid, ip->i_inode.i_gid); gfs2_meta_wipe(ip, ip->i_no_addr, 1); } @@ -1712,41 +2422,28 @@ void gfs2_free_di(struct gfs2_rgrpd *rgd, struct gfs2_inode *ip) int gfs2_check_blk_type(struct gfs2_sbd *sdp, u64 no_addr, unsigned int type) { struct gfs2_rgrpd *rgd; - struct gfs2_holder ri_gh, rgd_gh; - struct gfs2_inode *ip = GFS2_I(sdp->sd_rindex); - int ri_locked = 0; - int error; + struct gfs2_holder rgd_gh; + int error = -EINVAL; - if (!gfs2_glock_is_locked_by_me(ip->i_gl)) { - error = gfs2_rindex_hold(sdp, &ri_gh); - if (error) - goto fail; - ri_locked = 1; - } - - error = -EINVAL; - rgd = gfs2_blk2rgrpd(sdp, no_addr); + rgd = gfs2_blk2rgrpd(sdp, no_addr, 1); if (!rgd) - goto fail_rindex; + goto fail; error = gfs2_glock_nq_init(rgd->rd_gl, LM_ST_SHARED, 0, &rgd_gh); if (error) - goto fail_rindex; + goto fail; if (gfs2_get_block_type(rgd, no_addr) != type) error = -ESTALE; gfs2_glock_dq_uninit(&rgd_gh); -fail_rindex: - if (ri_locked) - gfs2_glock_dq_uninit(&ri_gh); fail: return error; } /** * gfs2_rlist_add - add a RG to a list of RGs - * @sdp: the filesystem + * @ip: the inode * @rlist: the list of resource groups * @block: the block * @@ -1756,9 +2453,10 @@ fail: * */ -void gfs2_rlist_add(struct gfs2_sbd *sdp, struct gfs2_rgrp_list *rlist, +void gfs2_rlist_add(struct gfs2_inode *ip, struct gfs2_rgrp_list *rlist, u64 block) { + struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); struct gfs2_rgrpd *rgd; struct gfs2_rgrpd **tmp; unsigned int new_space; @@ -1767,12 +2465,15 @@ void gfs2_rlist_add(struct gfs2_sbd *sdp, struct gfs2_rgrp_list *rlist, if (gfs2_assert_warn(sdp, !rlist->rl_ghs)) return; - rgd = gfs2_blk2rgrpd(sdp, block); + if (ip->i_rgd && rgrp_contains_block(ip->i_rgd, block)) + rgd = ip->i_rgd; + else + rgd = gfs2_blk2rgrpd(sdp, block, 1); if (!rgd) { - if (gfs2_consist(sdp)) - fs_err(sdp, "block = %llu\n", (unsigned long long)block); + fs_err(sdp, "rlist_add: no rgrp for block %llu\n", (unsigned long long)block); return; } + ip->i_rgd = rgd; for (x = 0; x < rlist->rl_rgrps; x++) if (rlist->rl_rgd[x] == rgd) @@ -1802,7 +2503,6 @@ void gfs2_rlist_add(struct gfs2_sbd *sdp, struct gfs2_rgrp_list *rlist, * and initialize an array of glock holders for them * @rlist: the list of resource groups * @state: the lock state to acquire the RG lock in - * @flags: the modifier flags for the holder structures * * FIXME: Don't use NOFAIL * @@ -1822,7 +2522,7 @@ void gfs2_rlist_alloc(struct gfs2_rgrp_list *rlist, unsigned int state) /** * gfs2_rlist_free - free a resource group list - * @list: the list of resource groups + * @rlist: the list of resource groups * */ @@ -1836,6 +2536,7 @@ void gfs2_rlist_free(struct gfs2_rgrp_list *rlist) for (x = 0; x < rlist->rl_rgrps; x++) gfs2_holder_uninit(&rlist->rl_ghs[x]); kfree(rlist->rl_ghs); + rlist->rl_ghs = NULL; } } diff --git a/fs/gfs2/rgrp.h b/fs/gfs2/rgrp.h index 50c2bb04369..463ab2e95d1 100644 --- a/fs/gfs2/rgrp.h +++ b/fs/gfs2/rgrp.h @@ -11,6 +11,15 @@ #define __RGRP_DOT_H__ #include <linux/slab.h> +#include <linux/uaccess.h> + +/* Since each block in the file system is represented by two bits in the + * bitmap, one 64-bit word in the bitmap will represent 32 blocks. + * By reserving 32 blocks at a time, we can optimize / shortcut how we search + * through the bitmaps by looking a word at a time. + */ +#define RGRP_RSRV_MINBYTES 8 +#define RGRP_RSRV_MINBLKS ((u32)(RGRP_RSRV_MINBYTES * GFS2_NBBY)) struct gfs2_rgrpd; struct gfs2_sbd; @@ -18,41 +27,29 @@ struct gfs2_holder; extern void gfs2_rgrp_verify(struct gfs2_rgrpd *rgd); -struct gfs2_rgrpd *gfs2_blk2rgrpd(struct gfs2_sbd *sdp, u64 blk); -struct gfs2_rgrpd *gfs2_rgrpd_get_first(struct gfs2_sbd *sdp); -struct gfs2_rgrpd *gfs2_rgrpd_get_next(struct gfs2_rgrpd *rgd); +extern struct gfs2_rgrpd *gfs2_blk2rgrpd(struct gfs2_sbd *sdp, u64 blk, bool exact); +extern struct gfs2_rgrpd *gfs2_rgrpd_get_first(struct gfs2_sbd *sdp); +extern struct gfs2_rgrpd *gfs2_rgrpd_get_next(struct gfs2_rgrpd *rgd); extern void gfs2_clear_rgrpd(struct gfs2_sbd *sdp); -extern int gfs2_rindex_hold(struct gfs2_sbd *sdp, struct gfs2_holder *ri_gh); - -extern int gfs2_rgrp_bh_get(struct gfs2_rgrpd *rgd); -extern void gfs2_rgrp_bh_hold(struct gfs2_rgrpd *rgd); -extern void gfs2_rgrp_bh_put(struct gfs2_rgrpd *rgd); - -extern void gfs2_rgrp_repolish_clones(struct gfs2_rgrpd *rgd); +extern int gfs2_rindex_update(struct gfs2_sbd *sdp); +extern void gfs2_free_clones(struct gfs2_rgrpd *rgd); +extern int gfs2_rgrp_go_lock(struct gfs2_holder *gh); +extern void gfs2_rgrp_go_unlock(struct gfs2_holder *gh); extern struct gfs2_alloc *gfs2_alloc_get(struct gfs2_inode *ip); -static inline void gfs2_alloc_put(struct gfs2_inode *ip) -{ - BUG_ON(ip->i_alloc == NULL); - kfree(ip->i_alloc); - ip->i_alloc = NULL; -} - -extern int gfs2_inplace_reserve_i(struct gfs2_inode *ip, int hold_rindex, - char *file, unsigned int line); -#define gfs2_inplace_reserve(ip) \ - gfs2_inplace_reserve_i((ip), 1, __FILE__, __LINE__) -#define gfs2_inplace_reserve_ri(ip) \ - gfs2_inplace_reserve_i((ip), 0, __FILE__, __LINE__) +#define GFS2_AF_ORLOV 1 +extern int gfs2_inplace_reserve(struct gfs2_inode *ip, const struct gfs2_alloc_parms *ap); extern void gfs2_inplace_release(struct gfs2_inode *ip); -extern int gfs2_ri_update(struct gfs2_inode *ip); -extern int gfs2_alloc_block(struct gfs2_inode *ip, u64 *bn, unsigned int *n); -extern int gfs2_alloc_di(struct gfs2_inode *ip, u64 *bn, u64 *generation); +extern int gfs2_alloc_blocks(struct gfs2_inode *ip, u64 *bn, unsigned int *n, + bool dinode, u64 *generation); -extern void gfs2_free_data(struct gfs2_inode *ip, u64 bstart, u32 blen); +extern int gfs2_rs_alloc(struct gfs2_inode *ip); +extern void gfs2_rs_deltree(struct gfs2_blkreserv *rs); +extern void gfs2_rs_delete(struct gfs2_inode *ip, atomic_t *wcount); +extern void __gfs2_free_blocks(struct gfs2_inode *ip, u64 bstart, u32 blen, int meta); extern void gfs2_free_meta(struct gfs2_inode *ip, u64 bstart, u32 blen); extern void gfs2_free_di(struct gfs2_rgrpd *rgd, struct gfs2_inode *ip); extern void gfs2_unlink_di(struct inode *inode); @@ -66,11 +63,21 @@ struct gfs2_rgrp_list { struct gfs2_holder *rl_ghs; }; -extern void gfs2_rlist_add(struct gfs2_sbd *sdp, struct gfs2_rgrp_list *rlist, +extern void gfs2_rlist_add(struct gfs2_inode *ip, struct gfs2_rgrp_list *rlist, u64 block); extern void gfs2_rlist_alloc(struct gfs2_rgrp_list *rlist, unsigned int state); extern void gfs2_rlist_free(struct gfs2_rgrp_list *rlist); extern u64 gfs2_ri_total(struct gfs2_sbd *sdp); -extern int gfs2_rgrp_dump(struct seq_file *seq, const struct gfs2_glock *gl); +extern void gfs2_rgrp_dump(struct seq_file *seq, const struct gfs2_glock *gl); +extern int gfs2_rgrp_send_discards(struct gfs2_sbd *sdp, u64 offset, + struct buffer_head *bh, + const struct gfs2_bitmap *bi, unsigned minlen, u64 *ptrimmed); +extern int gfs2_fitrim(struct file *filp, void __user *argp); + +/* This is how to tell if a reservation is in the rgrp tree: */ +static inline bool gfs2_rs_active(struct gfs2_blkreserv *rs) +{ + return rs && !RB_EMPTY_NODE(&rs->rs_node); +} #endif /* __RGRP_DOT_H__ */ diff --git a/fs/gfs2/super.c b/fs/gfs2/super.c index ec73ed70bae..1319b5c4ec6 100644 --- a/fs/gfs2/super.c +++ b/fs/gfs2/super.c @@ -7,6 +7,8 @@ * of the GNU General Public License version 2. */ +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + #include <linux/bio.h> #include <linux/sched.h> #include <linux/slab.h> @@ -23,6 +25,7 @@ #include <linux/time.h> #include <linux/wait.h> #include <linux/writeback.h> +#include <linux/backing-dev.h> #include "gfs2.h" #include "incore.h" @@ -77,6 +80,8 @@ enum { Opt_quota_quantum, Opt_barrier, Opt_nobarrier, + Opt_rgrplvb, + Opt_norgrplvb, Opt_error, }; @@ -114,6 +119,8 @@ static const match_table_t tokens = { {Opt_quota_quantum, "quota_quantum=%d"}, {Opt_barrier, "barrier"}, {Opt_nobarrier, "nobarrier"}, + {Opt_rgrplvb, "rgrplvb"}, + {Opt_norgrplvb, "norgrplvb"}, {Opt_error, NULL} }; @@ -170,8 +177,7 @@ int gfs2_mount_args(struct gfs2_args *args, char *options) break; case Opt_debug: if (args->ar_errors == GFS2_ERRORS_PANIC) { - printk(KERN_WARNING "GFS2: -o debug and -o errors=panic " - "are mutually exclusive.\n"); + pr_warn("-o debug and -o errors=panic are mutually exclusive\n"); return -EINVAL; } args->ar_debug = 1; @@ -223,21 +229,21 @@ int gfs2_mount_args(struct gfs2_args *args, char *options) case Opt_commit: rv = match_int(&tmp[0], &args->ar_commit); if (rv || args->ar_commit <= 0) { - printk(KERN_WARNING "GFS2: commit mount option requires a positive numeric argument\n"); + pr_warn("commit mount option requires a positive numeric argument\n"); return rv ? rv : -EINVAL; } break; case Opt_statfs_quantum: rv = match_int(&tmp[0], &args->ar_statfs_quantum); if (rv || args->ar_statfs_quantum < 0) { - printk(KERN_WARNING "GFS2: statfs_quantum mount option requires a non-negative numeric argument\n"); + pr_warn("statfs_quantum mount option requires a non-negative numeric argument\n"); return rv ? rv : -EINVAL; } break; case Opt_quota_quantum: rv = match_int(&tmp[0], &args->ar_quota_quantum); if (rv || args->ar_quota_quantum <= 0) { - printk(KERN_WARNING "GFS2: quota_quantum mount option requires a positive numeric argument\n"); + pr_warn("quota_quantum mount option requires a positive numeric argument\n"); return rv ? rv : -EINVAL; } break; @@ -245,7 +251,7 @@ int gfs2_mount_args(struct gfs2_args *args, char *options) rv = match_int(&tmp[0], &args->ar_statfs_percent); if (rv || args->ar_statfs_percent < 0 || args->ar_statfs_percent > 100) { - printk(KERN_WARNING "statfs_percent mount option requires a numeric argument between 0 and 100\n"); + pr_warn("statfs_percent mount option requires a numeric argument between 0 and 100\n"); return rv ? rv : -EINVAL; } break; @@ -254,8 +260,7 @@ int gfs2_mount_args(struct gfs2_args *args, char *options) break; case Opt_err_panic: if (args->ar_debug) { - printk(KERN_WARNING "GFS2: -o debug and -o errors=panic " - "are mutually exclusive.\n"); + pr_warn("-o debug and -o errors=panic are mutually exclusive\n"); return -EINVAL; } args->ar_errors = GFS2_ERRORS_PANIC; @@ -266,9 +271,15 @@ int gfs2_mount_args(struct gfs2_args *args, char *options) case Opt_nobarrier: args->ar_nobarrier = 1; break; + case Opt_rgrplvb: + args->ar_rgrplvb = 1; + break; + case Opt_norgrplvb: + args->ar_rgrplvb = 0; + break; case Opt_error: default: - printk(KERN_WARNING "GFS2: invalid mount option: %s\n", o); + pr_warn("invalid mount option: %s\n", o); return -EINVAL; } } @@ -284,9 +295,8 @@ int gfs2_mount_args(struct gfs2_args *args, char *options) void gfs2_jindex_free(struct gfs2_sbd *sdp) { - struct list_head list, *head; + struct list_head list; struct gfs2_jdesc *jd; - struct gfs2_journal_extent *jext; spin_lock(&sdp->sd_jindex_spin); list_add(&list, &sdp->sd_jindex_list); @@ -296,14 +306,7 @@ void gfs2_jindex_free(struct gfs2_sbd *sdp) while (!list_empty(&list)) { jd = list_entry(list.next, struct gfs2_jdesc, jd_list); - head = &jd->extent_list; - while (!list_empty(head)) { - jext = list_entry(head->next, - struct gfs2_journal_extent, - extent_list); - list_del(&jext->extent_list); - kfree(jext); - } + gfs2_free_journal_extents(jd); list_del(&jd->jd_list); iput(jd->jd_inode); kfree(jd); @@ -358,6 +361,33 @@ int gfs2_jdesc_check(struct gfs2_jdesc *jd) return 0; } +static int init_threads(struct gfs2_sbd *sdp) +{ + struct task_struct *p; + int error = 0; + + p = kthread_run(gfs2_logd, sdp, "gfs2_logd"); + if (IS_ERR(p)) { + error = PTR_ERR(p); + fs_err(sdp, "can't start logd thread: %d\n", error); + return error; + } + sdp->sd_logd_process = p; + + p = kthread_run(gfs2_quotad, sdp, "gfs2_quotad"); + if (IS_ERR(p)) { + error = PTR_ERR(p); + fs_err(sdp, "can't start quotad thread: %d\n", error); + goto fail; + } + sdp->sd_quotad_process = p; + return 0; + +fail: + kthread_stop(sdp->sd_logd_process); + return error; +} + /** * gfs2_make_fs_rw - Turn a Read-Only FS into a Read-Write one * @sdp: the filesystem @@ -369,14 +399,19 @@ int gfs2_make_fs_rw(struct gfs2_sbd *sdp) { struct gfs2_inode *ip = GFS2_I(sdp->sd_jdesc->jd_inode); struct gfs2_glock *j_gl = ip->i_gl; - struct gfs2_holder t_gh; + struct gfs2_holder thaw_gh; struct gfs2_log_header_host head; int error; - error = gfs2_glock_nq_init(sdp->sd_trans_gl, LM_ST_SHARED, 0, &t_gh); + error = init_threads(sdp); if (error) return error; + error = gfs2_glock_nq_init(sdp->sd_freeze_gl, LM_ST_SHARED, 0, + &thaw_gh); + if (error) + goto fail_threads; + j_gl->gl_ops->go_inval(j_gl, DIO_METADATA); error = gfs2_find_jhead(sdp->sd_jdesc, &head); @@ -399,14 +434,16 @@ int gfs2_make_fs_rw(struct gfs2_sbd *sdp) set_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags); - gfs2_glock_dq_uninit(&t_gh); + gfs2_glock_dq_uninit(&thaw_gh); return 0; fail: - t_gh.gh_flags |= GL_NOCACHE; - gfs2_glock_dq_uninit(&t_gh); - + thaw_gh.gh_flags |= GL_NOCACHE; + gfs2_glock_dq_uninit(&thaw_gh); +fail_threads: + kthread_stop(sdp->sd_quotad_process); + kthread_stop(sdp->sd_logd_process); return error; } @@ -489,7 +526,7 @@ void gfs2_statfs_change(struct gfs2_sbd *sdp, s64 total, s64 free, if (error) return; - gfs2_trans_add_bh(l_ip->i_gl, l_bh, 1); + gfs2_trans_add_meta(l_ip->i_gl, l_bh); spin_lock(&sdp->sd_statfs_spin); l_sc->sc_total += total; @@ -517,7 +554,7 @@ void update_statfs(struct gfs2_sbd *sdp, struct buffer_head *m_bh, struct gfs2_statfs_change_host *m_sc = &sdp->sd_statfs_master; struct gfs2_statfs_change_host *l_sc = &sdp->sd_statfs_local; - gfs2_trans_add_bh(l_ip->i_gl, l_bh, 1); + gfs2_trans_add_meta(l_ip->i_gl, l_bh); spin_lock(&sdp->sd_statfs_spin); m_sc->sc_total += l_sc->sc_total; @@ -528,7 +565,7 @@ void update_statfs(struct gfs2_sbd *sdp, struct buffer_head *m_bh, 0, sizeof(struct gfs2_statfs_change)); spin_unlock(&sdp->sd_statfs_spin); - gfs2_trans_add_bh(m_ip->i_gl, m_bh, 1); + gfs2_trans_add_meta(m_ip->i_gl, m_bh); gfs2_statfs_change_out(m_sc, m_bh->b_data + sizeof(struct gfs2_dinode)); } @@ -599,15 +636,21 @@ struct lfcc { */ static int gfs2_lock_fs_check_clean(struct gfs2_sbd *sdp, - struct gfs2_holder *t_gh) + struct gfs2_holder *freeze_gh) { struct gfs2_inode *ip; struct gfs2_jdesc *jd; struct lfcc *lfcc; LIST_HEAD(list); struct gfs2_log_header_host lh; + struct gfs2_inode *dip = GFS2_I(sdp->sd_root_dir->d_inode); int error; + error = gfs2_glock_nq_init(dip->i_gl, LM_ST_SHARED, 0, + &sdp->sd_freeze_root_gh); + if (error) + return error; + atomic_set(&sdp->sd_frozen_root, 1); list_for_each_entry(jd, &sdp->sd_jindex_list, jd_list) { lfcc = kmalloc(sizeof(struct lfcc), GFP_KERNEL); if (!lfcc) { @@ -623,8 +666,8 @@ static int gfs2_lock_fs_check_clean(struct gfs2_sbd *sdp, list_add(&lfcc->list, &list); } - error = gfs2_glock_nq_init(sdp->sd_trans_gl, LM_ST_DEFERRED, - GL_NOCACHE, t_gh); + error = gfs2_glock_nq_init(sdp->sd_freeze_gl, LM_ST_EXCLUSIVE, + GL_NOCACHE, freeze_gh); list_for_each_entry(jd, &sdp->sd_jindex_list, jd_list) { error = gfs2_jdesc_check(jd); @@ -640,7 +683,7 @@ static int gfs2_lock_fs_check_clean(struct gfs2_sbd *sdp, } if (error) - gfs2_glock_dq_uninit(t_gh); + gfs2_glock_dq_uninit(freeze_gh); out: while (!list_empty(&list)) { @@ -649,103 +692,137 @@ out: gfs2_glock_dq_uninit(&lfcc->gh); kfree(lfcc); } + if (error) { + atomic_dec(&sdp->sd_frozen_root); + wait_event(sdp->sd_frozen_root_wait, atomic_read(&sdp->sd_frozen_root) == 0); + gfs2_glock_dq_uninit(&sdp->sd_freeze_root_gh); + } return error; } -/** - * gfs2_freeze_fs - freezes the file system - * @sdp: the file system - * - * This function flushes data and meta data for all machines by - * aquiring the transaction log exclusively. All journals are - * ensured to be in a clean state as well. - * - * Returns: errno - */ - -int gfs2_freeze_fs(struct gfs2_sbd *sdp) +void gfs2_dinode_out(const struct gfs2_inode *ip, void *buf) { - int error = 0; - - mutex_lock(&sdp->sd_freeze_lock); - - if (!sdp->sd_freeze_count++) { - error = gfs2_lock_fs_check_clean(sdp, &sdp->sd_freeze_gh); - if (error) - sdp->sd_freeze_count--; - } - - mutex_unlock(&sdp->sd_freeze_lock); - - return error; + struct gfs2_dinode *str = buf; + + str->di_header.mh_magic = cpu_to_be32(GFS2_MAGIC); + str->di_header.mh_type = cpu_to_be32(GFS2_METATYPE_DI); + str->di_header.mh_format = cpu_to_be32(GFS2_FORMAT_DI); + str->di_num.no_addr = cpu_to_be64(ip->i_no_addr); + str->di_num.no_formal_ino = cpu_to_be64(ip->i_no_formal_ino); + str->di_mode = cpu_to_be32(ip->i_inode.i_mode); + str->di_uid = cpu_to_be32(i_uid_read(&ip->i_inode)); + str->di_gid = cpu_to_be32(i_gid_read(&ip->i_inode)); + str->di_nlink = cpu_to_be32(ip->i_inode.i_nlink); + str->di_size = cpu_to_be64(i_size_read(&ip->i_inode)); + str->di_blocks = cpu_to_be64(gfs2_get_inode_blocks(&ip->i_inode)); + str->di_atime = cpu_to_be64(ip->i_inode.i_atime.tv_sec); + str->di_mtime = cpu_to_be64(ip->i_inode.i_mtime.tv_sec); + str->di_ctime = cpu_to_be64(ip->i_inode.i_ctime.tv_sec); + + str->di_goal_meta = cpu_to_be64(ip->i_goal); + str->di_goal_data = cpu_to_be64(ip->i_goal); + str->di_generation = cpu_to_be64(ip->i_generation); + + str->di_flags = cpu_to_be32(ip->i_diskflags); + str->di_height = cpu_to_be16(ip->i_height); + str->di_payload_format = cpu_to_be32(S_ISDIR(ip->i_inode.i_mode) && + !(ip->i_diskflags & GFS2_DIF_EXHASH) ? + GFS2_FORMAT_DE : 0); + str->di_depth = cpu_to_be16(ip->i_depth); + str->di_entries = cpu_to_be32(ip->i_entries); + + str->di_eattr = cpu_to_be64(ip->i_eattr); + str->di_atime_nsec = cpu_to_be32(ip->i_inode.i_atime.tv_nsec); + str->di_mtime_nsec = cpu_to_be32(ip->i_inode.i_mtime.tv_nsec); + str->di_ctime_nsec = cpu_to_be32(ip->i_inode.i_ctime.tv_nsec); } /** - * gfs2_unfreeze_fs - unfreezes the file system - * @sdp: the file system - * - * This function allows the file system to proceed by unlocking - * the exclusively held transaction lock. Other GFS2 nodes are - * now free to acquire the lock shared and go on with their lives. + * gfs2_write_inode - Make sure the inode is stable on the disk + * @inode: The inode + * @wbc: The writeback control structure * + * Returns: errno */ -void gfs2_unfreeze_fs(struct gfs2_sbd *sdp) +static int gfs2_write_inode(struct inode *inode, struct writeback_control *wbc) { - mutex_lock(&sdp->sd_freeze_lock); - - if (sdp->sd_freeze_count && !--sdp->sd_freeze_count) - gfs2_glock_dq_uninit(&sdp->sd_freeze_gh); + struct gfs2_inode *ip = GFS2_I(inode); + struct gfs2_sbd *sdp = GFS2_SB(inode); + struct address_space *metamapping = gfs2_glock2aspace(ip->i_gl); + struct backing_dev_info *bdi = metamapping->backing_dev_info; + int ret = 0; - mutex_unlock(&sdp->sd_freeze_lock); + if (wbc->sync_mode == WB_SYNC_ALL) + gfs2_log_flush(GFS2_SB(inode), ip->i_gl, NORMAL_FLUSH); + if (bdi->dirty_exceeded) + gfs2_ail1_flush(sdp, wbc); + else + filemap_fdatawrite(metamapping); + if (wbc->sync_mode == WB_SYNC_ALL) + ret = filemap_fdatawait(metamapping); + if (ret) + mark_inode_dirty_sync(inode); + return ret; } - /** - * gfs2_write_inode - Make sure the inode is stable on the disk - * @inode: The inode - * @sync: synchronous write flag + * gfs2_dirty_inode - check for atime updates + * @inode: The inode in question + * @flags: The type of dirty * - * Returns: errno + * Unfortunately it can be called under any combination of inode + * glock and transaction lock, so we have to check carefully. + * + * At the moment this deals only with atime - it should be possible + * to expand that role in future, once a review of the locking has + * been carried out. */ -static int gfs2_write_inode(struct inode *inode, struct writeback_control *wbc) +static void gfs2_dirty_inode(struct inode *inode, int flags) { struct gfs2_inode *ip = GFS2_I(inode); struct gfs2_sbd *sdp = GFS2_SB(inode); - struct gfs2_holder gh; struct buffer_head *bh; - struct timespec atime; - struct gfs2_dinode *di; - int ret = 0; + struct gfs2_holder gh; + int need_unlock = 0; + int need_endtrans = 0; + int ret; + + if (!(flags & (I_DIRTY_DATASYNC|I_DIRTY_SYNC))) + return; + + if (!gfs2_glock_is_locked_by_me(ip->i_gl)) { + ret = gfs2_glock_nq_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, &gh); + if (ret) { + fs_err(sdp, "dirty_inode: glock %d\n", ret); + return; + } + need_unlock = 1; + } else if (WARN_ON_ONCE(ip->i_gl->gl_state != LM_ST_EXCLUSIVE)) + return; + + if (current->journal_info == NULL) { + ret = gfs2_trans_begin(sdp, RES_DINODE, 0); + if (ret) { + fs_err(sdp, "dirty_inode: gfs2_trans_begin %d\n", ret); + goto out; + } + need_endtrans = 1; + } - /* Check this is a "normal" inode, etc */ - if (current->flags & PF_MEMALLOC) - return 0; - ret = gfs2_glock_nq_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, &gh); - if (ret) - goto do_flush; - ret = gfs2_trans_begin(sdp, RES_DINODE, 0); - if (ret) - goto do_unlock; ret = gfs2_meta_inode_buffer(ip, &bh); if (ret == 0) { - di = (struct gfs2_dinode *)bh->b_data; - atime.tv_sec = be64_to_cpu(di->di_atime); - atime.tv_nsec = be32_to_cpu(di->di_atime_nsec); - if (timespec_compare(&inode->i_atime, &atime) > 0) { - gfs2_trans_add_bh(ip->i_gl, bh, 1); - gfs2_dinode_out(ip, bh->b_data); - } + gfs2_trans_add_meta(ip->i_gl, bh); + gfs2_dinode_out(ip, bh->b_data); brelse(bh); } - gfs2_trans_end(sdp); -do_unlock: - gfs2_glock_dq_uninit(&gh); -do_flush: - if (wbc->sync_mode == WB_SYNC_ALL) - gfs2_log_flush(GFS2_SB(inode), ip->i_gl); - return ret; + + if (need_endtrans) + gfs2_trans_end(sdp); +out: + if (need_unlock) + gfs2_glock_dq_uninit(&gh); } /** @@ -757,25 +834,30 @@ do_flush: static int gfs2_make_fs_ro(struct gfs2_sbd *sdp) { - struct gfs2_holder t_gh; + struct gfs2_holder thaw_gh; int error; - flush_workqueue(gfs2_delete_workqueue); - gfs2_quota_sync(sdp->sd_vfs, 0, 1); - gfs2_statfs_sync(sdp->sd_vfs, 0); - - error = gfs2_glock_nq_init(sdp->sd_trans_gl, LM_ST_SHARED, GL_NOCACHE, - &t_gh); + error = gfs2_glock_nq_init(sdp->sd_freeze_gl, LM_ST_SHARED, GL_NOCACHE, + &thaw_gh); if (error && !test_bit(SDF_SHUTDOWN, &sdp->sd_flags)) return error; - gfs2_meta_syncfs(sdp); - gfs2_log_shutdown(sdp); - + down_write(&sdp->sd_log_flush_lock); clear_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags); + up_write(&sdp->sd_log_flush_lock); + + kthread_stop(sdp->sd_quotad_process); + kthread_stop(sdp->sd_logd_process); + + flush_workqueue(gfs2_delete_workqueue); + gfs2_quota_sync(sdp->sd_vfs, 0); + gfs2_statfs_sync(sdp->sd_vfs, 0); + + gfs2_log_flush(sdp, NULL, SHUTDOWN_FLUSH); + gfs2_assert_warn(sdp, atomic_read(&sdp->sd_log_blks_free) == sdp->sd_jdesc->jd_blocks); - if (t_gh.gh_gl) - gfs2_glock_dq_uninit(&t_gh); + if (thaw_gh.gh_gl) + gfs2_glock_dq_uninit(&thaw_gh); gfs2_quota_cleanup(sdp); @@ -800,13 +882,6 @@ static void gfs2_put_super(struct super_block *sb) int error; struct gfs2_jdesc *jd; - /* Unfreeze the filesystem, if we need to */ - - mutex_lock(&sdp->sd_freeze_lock); - if (sdp->sd_freeze_count) - gfs2_glock_dq_uninit(&sdp->sd_freeze_gh); - mutex_unlock(&sdp->sd_freeze_lock); - /* No more recovery requests */ set_bit(SDF_NORECOVERY, &sdp->sd_flags); smp_mb(); @@ -824,9 +899,6 @@ restart: } spin_unlock(&sdp->sd_jindex_spin); - kthread_stop(sdp->sd_quotad_process); - kthread_stop(sdp->sd_logd_process); - if (!(sb->s_flags & MS_RDONLY)) { error = gfs2_make_fs_ro(sdp); if (error) @@ -842,7 +914,7 @@ restart: iput(sdp->sd_quota_inode); gfs2_glock_put(sdp->sd_rename_gl); - gfs2_glock_put(sdp->sd_trans_gl); + gfs2_glock_put(sdp->sd_freeze_gl); if (!sdp->sd_args.ar_spectator) { gfs2_glock_dq_uninit(&sdp->sd_journal_gh); @@ -874,8 +946,11 @@ restart: static int gfs2_sync_fs(struct super_block *sb, int wait) { - if (wait && sb->s_fs_info) - gfs2_log_flush(sb->s_fs_info, NULL); + struct gfs2_sbd *sdp = sb->s_fs_info; + + gfs2_quota_sync(sb, -1); + if (wait && sdp && !atomic_read(&sdp->sd_log_freeze)) + gfs2_log_flush(sdp, NULL, NORMAL_FLUSH); return 0; } @@ -894,7 +969,7 @@ static int gfs2_freeze(struct super_block *sb) return -EINVAL; for (;;) { - error = gfs2_freeze_fs(sdp); + error = gfs2_lock_fs_check_clean(sdp, &sdp->sd_freeze_gh); if (!error) break; @@ -922,7 +997,12 @@ static int gfs2_freeze(struct super_block *sb) static int gfs2_unfreeze(struct super_block *sb) { - gfs2_unfreeze_fs(sb->s_fs_info); + struct gfs2_sbd *sdp = sb->s_fs_info; + + gfs2_glock_dq_uninit(&sdp->sd_freeze_gh); + atomic_dec(&sdp->sd_frozen_root); + wait_event(sdp->sd_frozen_root_wait, atomic_read(&sdp->sd_frozen_root) == 0); + gfs2_glock_dq_uninit(&sdp->sd_freeze_root_gh); return 0; } @@ -959,7 +1039,6 @@ static int statfs_slow_fill(struct gfs2_rgrpd *rgd, static int gfs2_statfs_slow(struct gfs2_sbd *sdp, struct gfs2_statfs_change_host *sc) { - struct gfs2_holder ri_gh; struct gfs2_rgrpd *rgd_next; struct gfs2_holder *gha, *gh; unsigned int slots = 64; @@ -972,10 +1051,6 @@ static int gfs2_statfs_slow(struct gfs2_sbd *sdp, struct gfs2_statfs_change_host if (!gha) return -ENOMEM; - error = gfs2_rindex_hold(sdp, &ri_gh); - if (error) - goto out; - rgd_next = gfs2_rgrpd_get_first(sdp); for (;;) { @@ -1018,9 +1093,6 @@ static int gfs2_statfs_slow(struct gfs2_sbd *sdp, struct gfs2_statfs_change_host yield(); } - gfs2_glock_dq_uninit(&ri_gh); - -out: kfree(gha); return error; } @@ -1072,6 +1144,10 @@ static int gfs2_statfs(struct dentry *dentry, struct kstatfs *buf) struct gfs2_statfs_change_host sc; int error; + error = gfs2_rindex_update(sdp); + if (error) + return error; + if (gfs2_tune_get(sdp, gt_statfs_slow)) error = gfs2_statfs_slow(sdp, &sc); else @@ -1108,6 +1184,8 @@ static int gfs2_remount_fs(struct super_block *sb, int *flags, char *data) struct gfs2_tune *gt = &sdp->sd_tune; int error; + sync_filesystem(sb); + spin_lock(>->gt_spin); args.ar_commit = gt->gt_logd_secs; args.ar_quota_quantum = gt->gt_quota_quantum; @@ -1189,7 +1267,7 @@ static int gfs2_drop_inode(struct inode *inode) { struct gfs2_inode *ip = GFS2_I(inode); - if (inode->i_nlink) { + if (!test_bit(GIF_FREE_VFS_INODE, &ip->i_flags) && inode->i_nlink) { struct gfs2_glock *gl = ip->i_iopen_gh.gh_gl; if (gl && test_bit(GLF_DEMOTE, &gl->gl_flags)) clear_nlink(inode); @@ -1210,18 +1288,18 @@ static int is_ancestor(const struct dentry *d1, const struct dentry *d2) /** * gfs2_show_options - Show mount options for /proc/mounts * @s: seq_file structure - * @mnt: vfsmount + * @root: root of this (sub)tree * * Returns: 0 on success or error code */ -static int gfs2_show_options(struct seq_file *s, struct vfsmount *mnt) +static int gfs2_show_options(struct seq_file *s, struct dentry *root) { - struct gfs2_sbd *sdp = mnt->mnt_sb->s_fs_info; + struct gfs2_sbd *sdp = root->d_sb->s_fs_info; struct gfs2_args *args = &sdp->sd_args; int val; - if (is_ancestor(mnt->mnt_root, sdp->sd_master_dir)) + if (is_ancestor(root, sdp->sd_master_dir)) seq_printf(s, ",meta"); if (args->ar_lockproto[0]) seq_printf(s, ",lockproto=%s", args->ar_lockproto); @@ -1280,6 +1358,8 @@ static int gfs2_show_options(struct seq_file *s, struct vfsmount *mnt) val = sdp->sd_tune.gt_statfs_quantum; if (val != 30) seq_printf(s, ",statfs_quantum=%d", val); + else if (sdp->sd_tune.gt_statfs_slow) + seq_puts(s, ",statfs_quantum=0"); val = sdp->sd_tune.gt_quota_quantum; if (val != 60) seq_printf(s, ",quota_quantum=%d", val); @@ -1305,10 +1385,88 @@ static int gfs2_show_options(struct seq_file *s, struct vfsmount *mnt) seq_printf(s, ",nobarrier"); if (test_bit(SDF_DEMOTE, &sdp->sd_flags)) seq_printf(s, ",demote_interface_used"); + if (args->ar_rgrplvb) + seq_printf(s, ",rgrplvb"); return 0; } -/* +static void gfs2_final_release_pages(struct gfs2_inode *ip) +{ + struct inode *inode = &ip->i_inode; + struct gfs2_glock *gl = ip->i_gl; + + truncate_inode_pages(gfs2_glock2aspace(ip->i_gl), 0); + truncate_inode_pages(&inode->i_data, 0); + + if (atomic_read(&gl->gl_revokes) == 0) { + clear_bit(GLF_LFLUSH, &gl->gl_flags); + clear_bit(GLF_DIRTY, &gl->gl_flags); + } +} + +static int gfs2_dinode_dealloc(struct gfs2_inode *ip) +{ + struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); + struct gfs2_rgrpd *rgd; + struct gfs2_holder gh; + int error; + + if (gfs2_get_inode_blocks(&ip->i_inode) != 1) { + gfs2_consist_inode(ip); + return -EIO; + } + + error = gfs2_rindex_update(sdp); + if (error) + return error; + + error = gfs2_quota_hold(ip, NO_UID_QUOTA_CHANGE, NO_GID_QUOTA_CHANGE); + if (error) + return error; + + rgd = gfs2_blk2rgrpd(sdp, ip->i_no_addr, 1); + if (!rgd) { + gfs2_consist_inode(ip); + error = -EIO; + goto out_qs; + } + + error = gfs2_glock_nq_init(rgd->rd_gl, LM_ST_EXCLUSIVE, 0, &gh); + if (error) + goto out_qs; + + error = gfs2_trans_begin(sdp, RES_RG_BIT + RES_STATFS + RES_QUOTA, + sdp->sd_jdesc->jd_blocks); + if (error) + goto out_rg_gunlock; + + gfs2_free_di(rgd, ip); + + gfs2_final_release_pages(ip); + + gfs2_trans_end(sdp); + +out_rg_gunlock: + gfs2_glock_dq_uninit(&gh); +out_qs: + gfs2_quota_unhold(ip); + return error; +} + +/** + * gfs2_evict_inode - Remove an inode from cache + * @inode: The inode to evict + * + * There are three cases to consider: + * 1. i_nlink == 0, we are final opener (and must deallocate) + * 2. i_nlink == 0, we are not the final opener (and cannot deallocate) + * 3. i_nlink > 0 + * + * If the fs is read only, then we have to treat all cases as per #3 + * since we are unable to do any deallocation. The inode will be + * deallocated by the next read/write node to attempt an allocation + * in the same resource group + * * We have to (at the moment) hold the inodes main lock to cover * the gap between unlocking the shared lock on the iopen lock and * taking the exclusive lock. I'd rather do a shared -> exclusive @@ -1318,23 +1476,39 @@ static int gfs2_show_options(struct seq_file *s, struct vfsmount *mnt) static void gfs2_evict_inode(struct inode *inode) { - struct gfs2_sbd *sdp = inode->i_sb->s_fs_info; + struct super_block *sb = inode->i_sb; + struct gfs2_sbd *sdp = sb->s_fs_info; struct gfs2_inode *ip = GFS2_I(inode); struct gfs2_holder gh; int error; - if (inode->i_nlink) + if (test_bit(GIF_FREE_VFS_INODE, &ip->i_flags)) { + clear_inode(inode); + return; + } + + if (inode->i_nlink || (sb->s_flags & MS_RDONLY)) goto out; - error = gfs2_glock_nq_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, &gh); + /* Must not read inode block until block type has been verified */ + error = gfs2_glock_nq_init(ip->i_gl, LM_ST_EXCLUSIVE, GL_SKIP, &gh); if (unlikely(error)) { + ip->i_iopen_gh.gh_flags |= GL_NOCACHE; gfs2_glock_dq_uninit(&ip->i_iopen_gh); goto out; } - error = gfs2_check_blk_type(sdp, ip->i_no_addr, GFS2_BLKST_UNLINKED); - if (error) - goto out_truncate; + if (!test_bit(GIF_ALLOC_FAILED, &ip->i_flags)) { + error = gfs2_check_blk_type(sdp, ip->i_no_addr, GFS2_BLKST_UNLINKED); + if (error) + goto out_truncate; + } + + if (test_bit(GIF_INVALID, &ip->i_flags)) { + error = gfs2_inode_refresh(ip); + if (error) + goto out_truncate; + } ip->i_iopen_gh.gh_flags |= GL_NOCACHE; gfs2_glock_dq_wait(&ip->i_iopen_gh); @@ -1343,6 +1517,8 @@ static void gfs2_evict_inode(struct inode *inode) if (error) goto out_truncate; + /* Case 1 starts here */ + if (S_ISDIR(inode->i_mode) && (ip->i_diskflags & GFS2_DIF_EXHASH)) { error = gfs2_dir_exhash_dealloc(ip); @@ -1363,10 +1539,19 @@ static void gfs2_evict_inode(struct inode *inode) } error = gfs2_dinode_dealloc(ip); - if (error) - goto out_unlock; + goto out_unlock; out_truncate: + gfs2_log_flush(sdp, ip->i_gl, NORMAL_FLUSH); + if (test_bit(GLF_DIRTY, &ip->i_gl->gl_flags)) { + struct address_space *metamapping = gfs2_glock2aspace(ip->i_gl); + filemap_fdatawrite(metamapping); + filemap_fdatawait(metamapping); + } + write_inode_now(inode, 1); + gfs2_ail_flush(ip->i_gl, 0); + + /* Case 2 starts here */ error = gfs2_trans_begin(sdp, 0, sdp->sd_jdesc->jd_blocks); if (error) goto out_unlock; @@ -1375,21 +1560,33 @@ out_truncate: gfs2_trans_end(sdp); out_unlock: - if (test_bit(HIF_HOLDER, &ip->i_iopen_gh.gh_iflags)) + /* Error path for case 1 */ + if (gfs2_rs_active(ip->i_res)) + gfs2_rs_deltree(ip->i_res); + + if (test_bit(HIF_HOLDER, &ip->i_iopen_gh.gh_iflags)) { + ip->i_iopen_gh.gh_flags |= GL_NOCACHE; gfs2_glock_dq(&ip->i_iopen_gh); + } gfs2_holder_uninit(&ip->i_iopen_gh); gfs2_glock_dq_uninit(&gh); if (error && error != GLR_TRYFAILED && error != -EROFS) fs_warn(sdp, "gfs2_evict_inode: %d\n", error); out: - truncate_inode_pages(&inode->i_data, 0); - end_writeback(inode); - + /* Case 3 starts here */ + truncate_inode_pages_final(&inode->i_data); + gfs2_rs_delete(ip, NULL); + gfs2_ordered_del_inode(ip); + clear_inode(inode); + gfs2_dir_hash_inval(ip); ip->i_gl->gl_object = NULL; + flush_delayed_work(&ip->i_gl->gl_work); + gfs2_glock_add_to_lru(ip->i_gl); gfs2_glock_put(ip->i_gl); ip->i_gl = NULL; if (ip->i_iopen_gh.gh_gl) { ip->i_iopen_gh.gh_gl->gl_object = NULL; + ip->i_iopen_gh.gh_flags |= GL_NOCACHE; gfs2_glock_dq_uninit(&ip->i_iopen_gh); } } @@ -1402,6 +1599,8 @@ static struct inode *gfs2_alloc_inode(struct super_block *sb) if (ip) { ip->i_flags = 0; ip->i_gl = NULL; + ip->i_rgd = NULL; + ip->i_res = NULL; } return &ip->i_inode; } @@ -1409,7 +1608,6 @@ static struct inode *gfs2_alloc_inode(struct super_block *sb) static void gfs2_i_callback(struct rcu_head *head) { struct inode *inode = container_of(head, struct inode, i_rcu); - INIT_LIST_HEAD(&inode->i_dentry); kmem_cache_free(gfs2_inode_cachep, inode); } @@ -1422,6 +1620,7 @@ const struct super_operations gfs2_super_ops = { .alloc_inode = gfs2_alloc_inode, .destroy_inode = gfs2_destroy_inode, .write_inode = gfs2_write_inode, + .dirty_inode = gfs2_dirty_inode, .evict_inode = gfs2_evict_inode, .put_super = gfs2_put_super, .sync_fs = gfs2_sync_fs, diff --git a/fs/gfs2/super.h b/fs/gfs2/super.h index a0464680af0..90e3322ffa1 100644 --- a/fs/gfs2/super.h +++ b/fs/gfs2/super.h @@ -46,9 +46,6 @@ extern void update_statfs(struct gfs2_sbd *sdp, struct buffer_head *m_bh, struct buffer_head *l_bh); extern int gfs2_statfs_sync(struct super_block *sb, int type); -extern int gfs2_freeze_fs(struct gfs2_sbd *sdp); -extern void gfs2_unfreeze_fs(struct gfs2_sbd *sdp); - extern struct file_system_type gfs2_fs_type; extern struct file_system_type gfs2meta_fs_type; extern const struct export_operations gfs2_export_ops; diff --git a/fs/gfs2/sys.c b/fs/gfs2/sys.c index 748ccb557c1..3ab566ba569 100644 --- a/fs/gfs2/sys.c +++ b/fs/gfs2/sys.c @@ -7,6 +7,8 @@ * of the GNU General Public License version 2. */ +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + #include <linux/sched.h> #include <linux/spinlock.h> #include <linux/completion.h> @@ -81,7 +83,8 @@ static int gfs2_uuid_valid(const u8 *uuid) static ssize_t uuid_show(struct gfs2_sbd *sdp, char *buf) { - const u8 *uuid = sdp->sd_sb.sb_uuid; + struct super_block *s = sdp->sd_vfs; + const u8 *uuid = s->s_uuid; buf[0] = '\0'; if (!gfs2_uuid_valid(uuid)) return 0; @@ -90,39 +93,37 @@ static ssize_t uuid_show(struct gfs2_sbd *sdp, char *buf) static ssize_t freeze_show(struct gfs2_sbd *sdp, char *buf) { - unsigned int count; - - mutex_lock(&sdp->sd_freeze_lock); - count = sdp->sd_freeze_count; - mutex_unlock(&sdp->sd_freeze_lock); + struct super_block *sb = sdp->sd_vfs; + int frozen = (sb->s_writers.frozen == SB_UNFROZEN) ? 0 : 1; - return snprintf(buf, PAGE_SIZE, "%u\n", count); + return snprintf(buf, PAGE_SIZE, "%u\n", frozen); } static ssize_t freeze_store(struct gfs2_sbd *sdp, const char *buf, size_t len) { - ssize_t ret = len; - int error = 0; + int error; int n = simple_strtol(buf, NULL, 0); if (!capable(CAP_SYS_ADMIN)) - return -EACCES; + return -EPERM; switch (n) { case 0: - gfs2_unfreeze_fs(sdp); + error = thaw_super(sdp->sd_vfs); break; case 1: - error = gfs2_freeze_fs(sdp); + error = freeze_super(sdp->sd_vfs); break; default: - ret = -EINVAL; + return -EINVAL; } - if (error) + if (error) { fs_warn(sdp, "freeze %d error %d", n, error); + return error; + } - return ret; + return len; } static ssize_t withdraw_show(struct gfs2_sbd *sdp, char *buf) @@ -134,14 +135,13 @@ static ssize_t withdraw_show(struct gfs2_sbd *sdp, char *buf) static ssize_t withdraw_store(struct gfs2_sbd *sdp, const char *buf, size_t len) { if (!capable(CAP_SYS_ADMIN)) - return -EACCES; + return -EPERM; if (simple_strtol(buf, NULL, 0) != 1) return -EINVAL; - gfs2_lm_withdraw(sdp, - "GFS2: fsid=%s: withdrawing from cluster at user's request\n", - sdp->sd_fsname); + gfs2_lm_withdraw(sdp, "withdrawing from cluster at user's request\n"); + return len; } @@ -149,7 +149,7 @@ static ssize_t statfs_sync_store(struct gfs2_sbd *sdp, const char *buf, size_t len) { if (!capable(CAP_SYS_ADMIN)) - return -EACCES; + return -EPERM; if (simple_strtol(buf, NULL, 0) != 1) return -EINVAL; @@ -162,42 +162,52 @@ static ssize_t quota_sync_store(struct gfs2_sbd *sdp, const char *buf, size_t len) { if (!capable(CAP_SYS_ADMIN)) - return -EACCES; + return -EPERM; if (simple_strtol(buf, NULL, 0) != 1) return -EINVAL; - gfs2_quota_sync(sdp->sd_vfs, 0, 1); + gfs2_quota_sync(sdp->sd_vfs, 0); return len; } static ssize_t quota_refresh_user_store(struct gfs2_sbd *sdp, const char *buf, size_t len) { + struct kqid qid; int error; u32 id; if (!capable(CAP_SYS_ADMIN)) - return -EACCES; + return -EPERM; id = simple_strtoul(buf, NULL, 0); - error = gfs2_quota_refresh(sdp, 1, id); + qid = make_kqid(current_user_ns(), USRQUOTA, id); + if (!qid_valid(qid)) + return -EINVAL; + + error = gfs2_quota_refresh(sdp, qid); return error ? error : len; } static ssize_t quota_refresh_group_store(struct gfs2_sbd *sdp, const char *buf, size_t len) { + struct kqid qid; int error; u32 id; if (!capable(CAP_SYS_ADMIN)) - return -EACCES; + return -EPERM; id = simple_strtoul(buf, NULL, 0); - error = gfs2_quota_refresh(sdp, 0, id); + qid = make_kqid(current_user_ns(), GRPQUOTA, id); + if (!qid_valid(qid)) + return -EINVAL; + + error = gfs2_quota_refresh(sdp, qid); return error ? error : len; } @@ -212,7 +222,7 @@ static ssize_t demote_rq_store(struct gfs2_sbd *sdp, const char *buf, size_t len int rv; if (!capable(CAP_SYS_ADMIN)) - return -EACCES; + return -EPERM; rv = sscanf(buf, "%u:%llu %15s", &gltype, &glnum, mode); @@ -230,8 +240,8 @@ static ssize_t demote_rq_store(struct gfs2_sbd *sdp, const char *buf, size_t len if (gltype > LM_TYPE_JOURNAL) return -EINVAL; - if (gltype == LM_TYPE_NONDISK && glnum == GFS2_TRANS_LOCK) - glops = &gfs2_trans_glops; + if (gltype == LM_TYPE_NONDISK && glnum == GFS2_FREEZE_LOCK) + glops = &gfs2_freeze_glops; else glops = gfs2_glops_list[gltype]; if (glops == NULL) @@ -275,7 +285,15 @@ static struct attribute *gfs2_attrs[] = { NULL, }; +static void gfs2_sbd_release(struct kobject *kobj) +{ + struct gfs2_sbd *sdp = container_of(kobj, struct gfs2_sbd, sd_kobj); + + kfree(sdp); +} + static struct kobj_type gfs2_ktype = { + .release = gfs2_sbd_release, .default_attrs = gfs2_attrs, .sysfs_ops = &gfs2_attr_ops, }; @@ -297,7 +315,7 @@ static ssize_t block_show(struct gfs2_sbd *sdp, char *buf) ssize_t ret; int val = 0; - if (test_bit(DFL_BLOCK_LOCKS, &ls->ls_flags)) + if (test_bit(DFL_BLOCK_LOCKS, &ls->ls_recover_flags)) val = 1; ret = sprintf(buf, "%d\n", val); return ret; @@ -312,10 +330,10 @@ static ssize_t block_store(struct gfs2_sbd *sdp, const char *buf, size_t len) val = simple_strtol(buf, NULL, 0); if (val == 1) - set_bit(DFL_BLOCK_LOCKS, &ls->ls_flags); + set_bit(DFL_BLOCK_LOCKS, &ls->ls_recover_flags); else if (val == 0) { - clear_bit(DFL_BLOCK_LOCKS, &ls->ls_flags); - smp_mb__after_clear_bit(); + clear_bit(DFL_BLOCK_LOCKS, &ls->ls_recover_flags); + smp_mb__after_atomic(); gfs2_glock_thaw(sdp); } else { ret = -EINVAL; @@ -323,6 +341,28 @@ static ssize_t block_store(struct gfs2_sbd *sdp, const char *buf, size_t len) return ret; } +static ssize_t wdack_show(struct gfs2_sbd *sdp, char *buf) +{ + int val = completion_done(&sdp->sd_wdack) ? 1 : 0; + + return sprintf(buf, "%d\n", val); +} + +static ssize_t wdack_store(struct gfs2_sbd *sdp, const char *buf, size_t len) +{ + ssize_t ret = len; + int val; + + val = simple_strtol(buf, NULL, 0); + + if ((val == 1) && + !strcmp(sdp->sd_lockstruct.ls_ops->lm_proto_name, "lock_dlm")) + complete(&sdp->sd_wdack); + else + ret = -EINVAL; + return ret; +} + static ssize_t lkfirst_show(struct gfs2_sbd *sdp, char *buf) { struct lm_lockstruct *ls = &sdp->sd_lockstruct; @@ -337,6 +377,9 @@ static ssize_t lkfirst_store(struct gfs2_sbd *sdp, const char *buf, size_t len) rv = sscanf(buf, "%u", &first); if (rv != 1 || first > 1) return -EINVAL; + rv = wait_for_completion_killable(&sdp->sd_locking_init); + if (rv) + return rv; spin_lock(&sdp->sd_jindex_spin); rv = -EBUSY; if (test_bit(SDF_NOJOURNALID, &sdp->sd_flags) == 0) @@ -346,8 +389,8 @@ static ssize_t lkfirst_store(struct gfs2_sbd *sdp, const char *buf, size_t len) goto out; if (sdp->sd_lockstruct.ls_ops->lm_mount == NULL) goto out; - sdp->sd_lockstruct.ls_first = first; - rv = 0; + sdp->sd_lockstruct.ls_first = first; + rv = 0; out: spin_unlock(&sdp->sd_jindex_spin); return rv ? rv : len; @@ -356,23 +399,18 @@ out: static ssize_t first_done_show(struct gfs2_sbd *sdp, char *buf) { struct lm_lockstruct *ls = &sdp->sd_lockstruct; - return sprintf(buf, "%d\n", ls->ls_first_done); + return sprintf(buf, "%d\n", !!test_bit(DFL_FIRST_MOUNT_DONE, &ls->ls_recover_flags)); } -static ssize_t recover_store(struct gfs2_sbd *sdp, const char *buf, size_t len) +int gfs2_recover_set(struct gfs2_sbd *sdp, unsigned jid) { - unsigned jid; struct gfs2_jdesc *jd; int rv; - rv = sscanf(buf, "%u", &jid); - if (rv != 1) - return -EINVAL; + /* Wait for our primary journal to be initialized */ + wait_for_completion(&sdp->sd_journal_ready); - rv = -ESHUTDOWN; spin_lock(&sdp->sd_jindex_spin); - if (test_bit(SDF_NORECOVERY, &sdp->sd_flags)) - goto out; rv = -EBUSY; if (sdp->sd_jdesc->jd_jid == jid) goto out; @@ -385,6 +423,25 @@ static ssize_t recover_store(struct gfs2_sbd *sdp, const char *buf, size_t len) } out: spin_unlock(&sdp->sd_jindex_spin); + return rv; +} + +static ssize_t recover_store(struct gfs2_sbd *sdp, const char *buf, size_t len) +{ + unsigned jid; + int rv; + + rv = sscanf(buf, "%u", &jid); + if (rv != 1) + return -EINVAL; + + if (test_bit(SDF_NORECOVERY, &sdp->sd_flags)) { + rv = -ESHUTDOWN; + goto out; + } + + rv = gfs2_recover_set(sdp, jid); +out: return rv ? rv : len; } @@ -413,7 +470,9 @@ static ssize_t jid_store(struct gfs2_sbd *sdp, const char *buf, size_t len) rv = sscanf(buf, "%d", &jid); if (rv != 1) return -EINVAL; - + rv = wait_for_completion_killable(&sdp->sd_locking_init); + if (rv) + return rv; spin_lock(&sdp->sd_jindex_spin); rv = -EINVAL; if (sdp->sd_lockstruct.ls_ops->lm_mount == NULL) @@ -426,7 +485,7 @@ static ssize_t jid_store(struct gfs2_sbd *sdp, const char *buf, size_t len) rv = jid = -EINVAL; sdp->sd_lockstruct.ls_jid = jid; clear_bit(SDF_NOJOURNALID, &sdp->sd_flags); - smp_mb__after_clear_bit(); + smp_mb__after_atomic(); wake_up_bit(&sdp->sd_flags, SDF_NOJOURNALID); out: spin_unlock(&sdp->sd_jindex_spin); @@ -438,7 +497,7 @@ static struct gfs2_attr gdlm_attr_##_name = __ATTR(_name,_mode,_show,_store) GDLM_ATTR(proto_name, 0444, proto_name_show, NULL); GDLM_ATTR(block, 0644, block_show, block_store); -GDLM_ATTR(withdraw, 0644, withdraw_show, withdraw_store); +GDLM_ATTR(withdraw, 0644, wdack_show, wdack_store); GDLM_ATTR(jid, 0644, jid_show, jid_store); GDLM_ATTR(first, 0644, lkfirst_show, lkfirst_store); GDLM_ATTR(first_done, 0444, first_done_show, NULL); @@ -477,7 +536,7 @@ static ssize_t quota_scale_store(struct gfs2_sbd *sdp, const char *buf, unsigned int x, y; if (!capable(CAP_SYS_ADMIN)) - return -EACCES; + return -EPERM; if (sscanf(buf, "%u %u", &x, &y) != 2 || !y) return -EINVAL; @@ -496,7 +555,7 @@ static ssize_t tune_set(struct gfs2_sbd *sdp, unsigned int *field, unsigned int x; if (!capable(CAP_SYS_ADMIN)) - return -EACCES; + return -EPERM; x = simple_strtoul(buf, NULL, 0); @@ -532,7 +591,6 @@ TUNE_ATTR(max_readahead, 0); TUNE_ATTR(complain_secs, 0); TUNE_ATTR(statfs_slow, 0); TUNE_ATTR(new_files_jdata, 0); -TUNE_ATTR(quota_simul_sync, 1); TUNE_ATTR(statfs_quantum, 1); TUNE_ATTR_3(quota_scale, quota_scale_show, quota_scale_store); @@ -542,7 +600,6 @@ static struct attribute *tune_attrs[] = { &tune_attr_max_readahead.attr, &tune_attr_complain_secs.attr, &tune_attr_statfs_slow.attr, - &tune_attr_quota_simul_sync.attr, &tune_attr_statfs_quantum.attr, &tune_attr_quota_scale.attr, &tune_attr_new_files_jdata.attr, @@ -566,6 +623,7 @@ int gfs2_sys_fs_add(struct gfs2_sbd *sdp) char ro[20]; char spectator[20]; char *envp[] = { ro, spectator, NULL }; + int sysfs_frees_sdp = 0; sprintf(ro, "RDONLY=%d", (sb->s_flags & MS_RDONLY) ? 1 : 0); sprintf(spectator, "SPECTATOR=%d", sdp->sd_args.ar_spectator ? 1 : 0); @@ -574,8 +632,10 @@ int gfs2_sys_fs_add(struct gfs2_sbd *sdp) error = kobject_init_and_add(&sdp->sd_kobj, &gfs2_ktype, NULL, "%s", sdp->sd_table_name); if (error) - goto fail; + goto fail_reg; + sysfs_frees_sdp = 1; /* Freeing sdp is now done by sysfs calling + function gfs2_sbd_release. */ error = sysfs_create_group(&sdp->sd_kobj, &tune_group); if (error) goto fail_reg; @@ -598,9 +658,13 @@ fail_lock_module: fail_tune: sysfs_remove_group(&sdp->sd_kobj, &tune_group); fail_reg: - kobject_put(&sdp->sd_kobj); -fail: + free_percpu(sdp->sd_lkstats); fs_err(sdp, "error %d adding sysfs files", error); + if (sysfs_frees_sdp) + kobject_put(&sdp->sd_kobj); + else + kfree(sdp); + sb->s_fs_info = NULL; return error; } @@ -616,7 +680,8 @@ static int gfs2_uevent(struct kset *kset, struct kobject *kobj, struct kobj_uevent_env *env) { struct gfs2_sbd *sdp = container_of(kobj, struct gfs2_sbd, sd_kobj); - const u8 *uuid = sdp->sd_sb.sb_uuid; + struct super_block *s = sdp->sd_vfs; + const u8 *uuid = s->s_uuid; add_uevent_var(env, "LOCKTABLE=%s", sdp->sd_table_name); add_uevent_var(env, "LOCKPROTO=%s", sdp->sd_proto_name); diff --git a/fs/gfs2/sys.h b/fs/gfs2/sys.h index e94560e836d..79182d6ad6a 100644 --- a/fs/gfs2/sys.h +++ b/fs/gfs2/sys.h @@ -19,5 +19,7 @@ void gfs2_sys_fs_del(struct gfs2_sbd *sdp); int gfs2_sys_init(void); void gfs2_sys_uninit(void); +int gfs2_recover_set(struct gfs2_sbd *sdp, unsigned jid); + #endif /* __SYS_DOT_H__ */ diff --git a/fs/gfs2/trace_gfs2.h b/fs/gfs2/trace_gfs2.h index cedb0bb96d9..20c007d747a 100644 --- a/fs/gfs2/trace_gfs2.h +++ b/fs/gfs2/trace_gfs2.h @@ -10,8 +10,11 @@ #include <linux/buffer_head.h> #include <linux/dlmconstants.h> #include <linux/gfs2_ondisk.h> +#include <linux/writeback.h> +#include <linux/ktime.h> #include "incore.h" #include "glock.h" +#include "rgrp.h" #define dlm_state_name(nn) { DLM_LOCK_##nn, #nn } #define glock_trace_name(x) __print_symbolic(x, \ @@ -29,6 +32,17 @@ { GFS2_BLKST_DINODE, "dinode" }, \ { GFS2_BLKST_UNLINKED, "unlinked" }) +#define TRACE_RS_DELETE 0 +#define TRACE_RS_TREEDEL 1 +#define TRACE_RS_INSERT 2 +#define TRACE_RS_CLAIM 3 + +#define rs_func_name(x) __print_symbolic(x, \ + { 0, "del " }, \ + { 1, "tdel" }, \ + { 2, "ins " }, \ + { 3, "clm " }) + #define show_glock_flags(flags) __print_flags(flags, "", \ {(1UL << GLF_LOCK), "l" }, \ {(1UL << GLF_DEMOTE), "D" }, \ @@ -40,7 +54,10 @@ {(1UL << GLF_REPLY_PENDING), "r" }, \ {(1UL << GLF_INITIAL), "I" }, \ {(1UL << GLF_FROZEN), "F" }, \ - {(1UL << GLF_QUEUED), "q" }) + {(1UL << GLF_QUEUED), "q" }, \ + {(1UL << GLF_LRU), "L" }, \ + {(1UL << GLF_OBJECT), "o" }, \ + {(1UL << GLF_BLOCKING), "b" }) #ifndef NUMPTY #define NUMPTY @@ -94,7 +111,7 @@ TRACE_EVENT(gfs2_glock_state_change, __entry->new_state = glock_trace_state(new_state); __entry->tgt_state = glock_trace_state(gl->gl_target); __entry->dmt_state = glock_trace_state(gl->gl_demote_state); - __entry->flags = gl->gl_flags; + __entry->flags = gl->gl_flags | (gl->gl_object ? (1UL<<GLF_OBJECT) : 0); ), TP_printk("%u,%u glock %d:%lld state %s to %s tgt:%s dmt:%s flags:%s", @@ -127,7 +144,7 @@ TRACE_EVENT(gfs2_glock_put, __entry->gltype = gl->gl_name.ln_type; __entry->glnum = gl->gl_name.ln_number; __entry->cur_state = glock_trace_state(gl->gl_state); - __entry->flags = gl->gl_flags; + __entry->flags = gl->gl_flags | (gl->gl_object ? (1UL<<GLF_OBJECT) : 0); ), TP_printk("%u,%u glock %d:%lld state %s => %s flags:%s", @@ -142,9 +159,9 @@ TRACE_EVENT(gfs2_glock_put, /* Callback (local or remote) requesting lock demotion */ TRACE_EVENT(gfs2_demote_rq, - TP_PROTO(const struct gfs2_glock *gl), + TP_PROTO(const struct gfs2_glock *gl, bool remote), - TP_ARGS(gl), + TP_ARGS(gl, remote), TP_STRUCT__entry( __field( dev_t, dev ) @@ -153,6 +170,7 @@ TRACE_EVENT(gfs2_demote_rq, __field( u8, cur_state ) __field( u8, dmt_state ) __field( unsigned long, flags ) + __field( bool, remote ) ), TP_fast_assign( @@ -161,15 +179,17 @@ TRACE_EVENT(gfs2_demote_rq, __entry->glnum = gl->gl_name.ln_number; __entry->cur_state = glock_trace_state(gl->gl_state); __entry->dmt_state = glock_trace_state(gl->gl_demote_state); - __entry->flags = gl->gl_flags; + __entry->flags = gl->gl_flags | (gl->gl_object ? (1UL<<GLF_OBJECT) : 0); + __entry->remote = remote; ), - TP_printk("%u,%u glock %d:%lld demote %s to %s flags:%s", + TP_printk("%u,%u glock %d:%lld demote %s to %s flags:%s %s", MAJOR(__entry->dev), MINOR(__entry->dev), __entry->gltype, (unsigned long long)__entry->glnum, glock_trace_name(__entry->cur_state), glock_trace_name(__entry->dmt_state), - show_glock_flags(__entry->flags)) + show_glock_flags(__entry->flags), + __entry->remote ? "remote" : "local") ); @@ -233,6 +253,62 @@ TRACE_EVENT(gfs2_glock_queue, glock_trace_name(__entry->state)) ); +/* DLM sends a reply to GFS2 */ +TRACE_EVENT(gfs2_glock_lock_time, + + TP_PROTO(const struct gfs2_glock *gl, s64 tdiff), + + TP_ARGS(gl, tdiff), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( u64, glnum ) + __field( u32, gltype ) + __field( int, status ) + __field( char, flags ) + __field( s64, tdiff ) + __field( s64, srtt ) + __field( s64, srttvar ) + __field( s64, srttb ) + __field( s64, srttvarb ) + __field( s64, sirt ) + __field( s64, sirtvar ) + __field( s64, dcount ) + __field( s64, qcount ) + ), + + TP_fast_assign( + __entry->dev = gl->gl_sbd->sd_vfs->s_dev; + __entry->glnum = gl->gl_name.ln_number; + __entry->gltype = gl->gl_name.ln_type; + __entry->status = gl->gl_lksb.sb_status; + __entry->flags = gl->gl_lksb.sb_flags; + __entry->tdiff = tdiff; + __entry->srtt = gl->gl_stats.stats[GFS2_LKS_SRTT]; + __entry->srttvar = gl->gl_stats.stats[GFS2_LKS_SRTTVAR]; + __entry->srttb = gl->gl_stats.stats[GFS2_LKS_SRTTB]; + __entry->srttvarb = gl->gl_stats.stats[GFS2_LKS_SRTTVARB]; + __entry->sirt = gl->gl_stats.stats[GFS2_LKS_SIRT]; + __entry->sirtvar = gl->gl_stats.stats[GFS2_LKS_SIRTVAR]; + __entry->dcount = gl->gl_stats.stats[GFS2_LKS_DCOUNT]; + __entry->qcount = gl->gl_stats.stats[GFS2_LKS_QCOUNT]; + ), + + TP_printk("%u,%u glock %d:%lld status:%d flags:%02x tdiff:%lld srtt:%lld/%lld srttb:%lld/%lld sirt:%lld/%lld dcnt:%lld qcnt:%lld", + MAJOR(__entry->dev), MINOR(__entry->dev), __entry->gltype, + (unsigned long long)__entry->glnum, + __entry->status, __entry->flags, + (long long)__entry->tdiff, + (long long)__entry->srtt, + (long long)__entry->srttvar, + (long long)__entry->srttb, + (long long)__entry->srttvarb, + (long long)__entry->sirt, + (long long)__entry->sirtvar, + (long long)__entry->dcount, + (long long)__entry->qcount) +); + /* Section 2 - Log/journal * * Objectives: @@ -318,6 +394,33 @@ TRACE_EVENT(gfs2_log_blocks, MINOR(__entry->dev), __entry->blocks) ); +/* Writing back the AIL */ +TRACE_EVENT(gfs2_ail_flush, + + TP_PROTO(const struct gfs2_sbd *sdp, const struct writeback_control *wbc, int start), + + TP_ARGS(sdp, wbc, start), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( int, start ) + __field( int, sync_mode ) + __field( long, nr_to_write ) + ), + + TP_fast_assign( + __entry->dev = sdp->sd_vfs->s_dev; + __entry->start = start; + __entry->sync_mode = wbc->sync_mode; + __entry->nr_to_write = wbc->nr_to_write; + ), + + TP_printk("%u,%u ail flush %s %s %ld", MAJOR(__entry->dev), + MINOR(__entry->dev), __entry->start ? "start" : "end", + __entry->sync_mode == WB_SYNC_ALL ? "all" : "none", + __entry->nr_to_write) +); + /* Section 3 - bmap * * Objectives: @@ -369,10 +472,10 @@ TRACE_EVENT(gfs2_bmap, /* Keep track of blocks as they are allocated/freed */ TRACE_EVENT(gfs2_block_alloc, - TP_PROTO(const struct gfs2_inode *ip, u64 block, unsigned len, - u8 block_state), + TP_PROTO(const struct gfs2_inode *ip, struct gfs2_rgrpd *rgd, + u64 block, unsigned len, u8 block_state), - TP_ARGS(ip, block, len, block_state), + TP_ARGS(ip, rgd, block, len, block_state), TP_STRUCT__entry( __field( dev_t, dev ) @@ -380,22 +483,69 @@ TRACE_EVENT(gfs2_block_alloc, __field( u64, inum ) __field( u32, len ) __field( u8, block_state ) + __field( u64, rd_addr ) + __field( u32, rd_free_clone ) + __field( u32, rd_reserved ) ), TP_fast_assign( - __entry->dev = ip->i_gl->gl_sbd->sd_vfs->s_dev; + __entry->dev = rgd->rd_gl->gl_sbd->sd_vfs->s_dev; __entry->start = block; __entry->inum = ip->i_no_addr; __entry->len = len; __entry->block_state = block_state; + __entry->rd_addr = rgd->rd_addr; + __entry->rd_free_clone = rgd->rd_free_clone; + __entry->rd_reserved = rgd->rd_reserved; ), - TP_printk("%u,%u bmap %llu alloc %llu/%lu %s", + TP_printk("%u,%u bmap %llu alloc %llu/%lu %s rg:%llu rf:%u rr:%lu", MAJOR(__entry->dev), MINOR(__entry->dev), (unsigned long long)__entry->inum, (unsigned long long)__entry->start, (unsigned long)__entry->len, - block_state_name(__entry->block_state)) + block_state_name(__entry->block_state), + (unsigned long long)__entry->rd_addr, + __entry->rd_free_clone, (unsigned long)__entry->rd_reserved) +); + +/* Keep track of multi-block reservations as they are allocated/freed */ +TRACE_EVENT(gfs2_rs, + + TP_PROTO(const struct gfs2_blkreserv *rs, u8 func), + + TP_ARGS(rs, func), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( u64, rd_addr ) + __field( u32, rd_free_clone ) + __field( u32, rd_reserved ) + __field( u64, inum ) + __field( u64, start ) + __field( u32, free ) + __field( u8, func ) + ), + + TP_fast_assign( + __entry->dev = rs->rs_rbm.rgd->rd_sbd->sd_vfs->s_dev; + __entry->rd_addr = rs->rs_rbm.rgd->rd_addr; + __entry->rd_free_clone = rs->rs_rbm.rgd->rd_free_clone; + __entry->rd_reserved = rs->rs_rbm.rgd->rd_reserved; + __entry->inum = rs->rs_inum; + __entry->start = gfs2_rbm_to_block(&rs->rs_rbm); + __entry->free = rs->rs_free; + __entry->func = func; + ), + + TP_printk("%u,%u bmap %llu resrv %llu rg:%llu rf:%lu rr:%lu %s f:%lu", + MAJOR(__entry->dev), MINOR(__entry->dev), + (unsigned long long)__entry->inum, + (unsigned long long)__entry->start, + (unsigned long long)__entry->rd_addr, + (unsigned long)__entry->rd_free_clone, + (unsigned long)__entry->rd_reserved, + rs_func_name(__entry->func), (unsigned long)__entry->free) ); #endif /* _TRACE_GFS2_H */ diff --git a/fs/gfs2/trans.c b/fs/gfs2/trans.c index 9ec73a85411..0546ab4e28e 100644 --- a/fs/gfs2/trans.c +++ b/fs/gfs2/trans.c @@ -7,6 +7,8 @@ * of the GNU General Public License version 2. */ +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + #include <linux/sched.h> #include <linux/slab.h> #include <linux/spinlock.h> @@ -18,6 +20,7 @@ #include "gfs2.h" #include "incore.h" #include "glock.h" +#include "inode.h" #include "log.h" #include "lops.h" #include "meta_io.h" @@ -45,123 +48,203 @@ int gfs2_trans_begin(struct gfs2_sbd *sdp, unsigned int blocks, tr->tr_blocks = blocks; tr->tr_revokes = revokes; tr->tr_reserved = 1; + tr->tr_alloced = 1; if (blocks) tr->tr_reserved += 6 + blocks; if (revokes) tr->tr_reserved += gfs2_struct2blk(sdp, revokes, sizeof(u64)); - INIT_LIST_HEAD(&tr->tr_list_buf); - - gfs2_holder_init(sdp->sd_trans_gl, LM_ST_SHARED, 0, &tr->tr_t_gh); + INIT_LIST_HEAD(&tr->tr_databuf); + INIT_LIST_HEAD(&tr->tr_buf); - error = gfs2_glock_nq(&tr->tr_t_gh); - if (error) - goto fail_holder_uninit; + sb_start_intwrite(sdp->sd_vfs); error = gfs2_log_reserve(sdp, tr->tr_reserved); if (error) - goto fail_gunlock; + goto fail; current->journal_info = tr; return 0; -fail_gunlock: - gfs2_glock_dq(&tr->tr_t_gh); - -fail_holder_uninit: - gfs2_holder_uninit(&tr->tr_t_gh); +fail: + sb_end_intwrite(sdp->sd_vfs); kfree(tr); return error; } -/** - * gfs2_log_release - Release a given number of log blocks - * @sdp: The GFS2 superblock - * @blks: The number of blocks - * - */ - -static void gfs2_log_release(struct gfs2_sbd *sdp, unsigned int blks) +static void gfs2_print_trans(const struct gfs2_trans *tr) { - - atomic_add(blks, &sdp->sd_log_blks_free); - trace_gfs2_log_blocks(sdp, blks); - gfs2_assert_withdraw(sdp, atomic_read(&sdp->sd_log_blks_free) <= - sdp->sd_jdesc->jd_blocks); - up_read(&sdp->sd_log_flush_lock); + pr_warn("Transaction created at: %pSR\n", (void *)tr->tr_ip); + pr_warn("blocks=%u revokes=%u reserved=%u touched=%u\n", + tr->tr_blocks, tr->tr_revokes, tr->tr_reserved, tr->tr_touched); + pr_warn("Buf %u/%u Databuf %u/%u Revoke %u/%u\n", + tr->tr_num_buf_new, tr->tr_num_buf_rm, + tr->tr_num_databuf_new, tr->tr_num_databuf_rm, + tr->tr_num_revoke, tr->tr_num_revoke_rm); } void gfs2_trans_end(struct gfs2_sbd *sdp) { struct gfs2_trans *tr = current->journal_info; - + s64 nbuf; BUG_ON(!tr); current->journal_info = NULL; if (!tr->tr_touched) { gfs2_log_release(sdp, tr->tr_reserved); - if (tr->tr_t_gh.gh_gl) { - gfs2_glock_dq(&tr->tr_t_gh); - gfs2_holder_uninit(&tr->tr_t_gh); + if (tr->tr_alloced) kfree(tr); - } + sb_end_intwrite(sdp->sd_vfs); return; } - if (gfs2_assert_withdraw(sdp, tr->tr_num_buf <= tr->tr_blocks)) { - fs_err(sdp, "tr_num_buf = %u, tr_blocks = %u ", - tr->tr_num_buf, tr->tr_blocks); - print_symbol(KERN_WARNING "GFS2: Transaction created at: %s\n", tr->tr_ip); - } - if (gfs2_assert_withdraw(sdp, tr->tr_num_revoke <= tr->tr_revokes)) { - fs_err(sdp, "tr_num_revoke = %u, tr_revokes = %u ", - tr->tr_num_revoke, tr->tr_revokes); - print_symbol(KERN_WARNING "GFS2: Transaction created at: %s\n", tr->tr_ip); - } + nbuf = tr->tr_num_buf_new + tr->tr_num_databuf_new; + nbuf -= tr->tr_num_buf_rm; + nbuf -= tr->tr_num_databuf_rm; + + if (gfs2_assert_withdraw(sdp, (nbuf <= tr->tr_blocks) && + (tr->tr_num_revoke <= tr->tr_revokes))) + gfs2_print_trans(tr); gfs2_log_commit(sdp, tr); - if (tr->tr_t_gh.gh_gl) { - gfs2_glock_dq(&tr->tr_t_gh); - gfs2_holder_uninit(&tr->tr_t_gh); - kfree(tr); - } + if (tr->tr_alloced && !tr->tr_attached) + kfree(tr); + up_read(&sdp->sd_log_flush_lock); if (sdp->sd_vfs->s_flags & MS_SYNCHRONOUS) - gfs2_log_flush(sdp, NULL); + gfs2_log_flush(sdp, NULL, NORMAL_FLUSH); + sb_end_intwrite(sdp->sd_vfs); +} + +static struct gfs2_bufdata *gfs2_alloc_bufdata(struct gfs2_glock *gl, + struct buffer_head *bh, + const struct gfs2_log_operations *lops) +{ + struct gfs2_bufdata *bd; + + bd = kmem_cache_zalloc(gfs2_bufdata_cachep, GFP_NOFS | __GFP_NOFAIL); + bd->bd_bh = bh; + bd->bd_gl = gl; + bd->bd_ops = lops; + INIT_LIST_HEAD(&bd->bd_list); + bh->b_private = bd; + return bd; } /** - * gfs2_trans_add_bh - Add a to-be-modified buffer to the current transaction - * @gl: the glock the buffer belongs to + * gfs2_trans_add_data - Add a databuf to the transaction. + * @gl: The inode glock associated with the buffer * @bh: The buffer to add - * @meta: True in the case of adding metadata * + * This is used in two distinct cases: + * i) In ordered write mode + * We put the data buffer on a list so that we can ensure that its + * synced to disk at the right time + * ii) In journaled data mode + * We need to journal the data block in the same way as metadata in + * the functions above. The difference is that here we have a tag + * which is two __be64's being the block number (as per meta data) + * and a flag which says whether the data block needs escaping or + * not. This means we need a new log entry for each 251 or so data + * blocks, which isn't an enormous overhead but twice as much as + * for normal metadata blocks. */ +void gfs2_trans_add_data(struct gfs2_glock *gl, struct buffer_head *bh) +{ + struct gfs2_trans *tr = current->journal_info; + struct gfs2_sbd *sdp = gl->gl_sbd; + struct address_space *mapping = bh->b_page->mapping; + struct gfs2_inode *ip = GFS2_I(mapping->host); + struct gfs2_bufdata *bd; + + if (!gfs2_is_jdata(ip)) { + gfs2_ordered_add_inode(ip); + return; + } + + lock_buffer(bh); + gfs2_log_lock(sdp); + bd = bh->b_private; + if (bd == NULL) { + gfs2_log_unlock(sdp); + unlock_buffer(bh); + if (bh->b_private == NULL) + bd = gfs2_alloc_bufdata(gl, bh, &gfs2_databuf_lops); + lock_buffer(bh); + gfs2_log_lock(sdp); + } + gfs2_assert(sdp, bd->bd_gl == gl); + tr->tr_touched = 1; + if (list_empty(&bd->bd_list)) { + set_bit(GLF_LFLUSH, &bd->bd_gl->gl_flags); + set_bit(GLF_DIRTY, &bd->bd_gl->gl_flags); + gfs2_pin(sdp, bd->bd_bh); + tr->tr_num_databuf_new++; + list_add_tail(&bd->bd_list, &tr->tr_databuf); + } + gfs2_log_unlock(sdp); + unlock_buffer(bh); +} + +static void meta_lo_add(struct gfs2_sbd *sdp, struct gfs2_bufdata *bd) +{ + struct gfs2_meta_header *mh; + struct gfs2_trans *tr; -void gfs2_trans_add_bh(struct gfs2_glock *gl, struct buffer_head *bh, int meta) + tr = current->journal_info; + tr->tr_touched = 1; + if (!list_empty(&bd->bd_list)) + return; + set_bit(GLF_LFLUSH, &bd->bd_gl->gl_flags); + set_bit(GLF_DIRTY, &bd->bd_gl->gl_flags); + mh = (struct gfs2_meta_header *)bd->bd_bh->b_data; + if (unlikely(mh->mh_magic != cpu_to_be32(GFS2_MAGIC))) { + pr_err("Attempting to add uninitialised block to journal (inplace block=%lld)\n", + (unsigned long long)bd->bd_bh->b_blocknr); + BUG(); + } + gfs2_pin(sdp, bd->bd_bh); + mh->__pad0 = cpu_to_be64(0); + mh->mh_jid = cpu_to_be32(sdp->sd_jdesc->jd_jid); + list_add(&bd->bd_list, &tr->tr_buf); + tr->tr_num_buf_new++; +} + +void gfs2_trans_add_meta(struct gfs2_glock *gl, struct buffer_head *bh) { + struct gfs2_sbd *sdp = gl->gl_sbd; struct gfs2_bufdata *bd; + lock_buffer(bh); + gfs2_log_lock(sdp); bd = bh->b_private; - if (bd) - gfs2_assert(sdp, bd->bd_gl == gl); - else { - gfs2_attach_bufdata(gl, bh, meta); - bd = bh->b_private; + if (bd == NULL) { + gfs2_log_unlock(sdp); + unlock_buffer(bh); + lock_page(bh->b_page); + if (bh->b_private == NULL) + bd = gfs2_alloc_bufdata(gl, bh, &gfs2_buf_lops); + unlock_page(bh->b_page); + lock_buffer(bh); + gfs2_log_lock(sdp); } - lops_add(sdp, &bd->bd_le); + gfs2_assert(sdp, bd->bd_gl == gl); + meta_lo_add(sdp, bd); + gfs2_log_unlock(sdp); + unlock_buffer(bh); } void gfs2_trans_add_revoke(struct gfs2_sbd *sdp, struct gfs2_bufdata *bd) { - BUG_ON(!list_empty(&bd->bd_le.le_list)); - BUG_ON(!list_empty(&bd->bd_ail_st_list)); - BUG_ON(!list_empty(&bd->bd_ail_gl_list)); - lops_init_le(&bd->bd_le, &gfs2_revoke_lops); - lops_add(sdp, &bd->bd_le); + struct gfs2_trans *tr = current->journal_info; + + BUG_ON(!list_empty(&bd->bd_list)); + gfs2_add_revoke(sdp, bd); + tr->tr_touched = 1; + tr->tr_num_revoke++; } void gfs2_trans_add_unrevoke(struct gfs2_sbd *sdp, u64 blkno, unsigned int len) @@ -171,9 +254,9 @@ void gfs2_trans_add_unrevoke(struct gfs2_sbd *sdp, u64 blkno, unsigned int len) unsigned int n = len; gfs2_log_lock(sdp); - list_for_each_entry_safe(bd, tmp, &sdp->sd_log_le_revoke, bd_le.le_list) { + list_for_each_entry_safe(bd, tmp, &sdp->sd_log_le_revoke, bd_list) { if ((bd->bd_blkno >= blkno) && (bd->bd_blkno < (blkno + len))) { - list_del_init(&bd->bd_le.le_list); + list_del_init(&bd->bd_list); gfs2_assert_withdraw(sdp, sdp->sd_log_num_revoke); sdp->sd_log_num_revoke--; kmem_cache_free(gfs2_bufdata_cachep, bd); @@ -185,8 +268,3 @@ void gfs2_trans_add_unrevoke(struct gfs2_sbd *sdp, u64 blkno, unsigned int len) gfs2_log_unlock(sdp); } -void gfs2_trans_add_rg(struct gfs2_rgrpd *rgd) -{ - lops_add(rgd->rd_sbd, &rgd->rd_le); -} - diff --git a/fs/gfs2/trans.h b/fs/gfs2/trans.h index fb56b783e02..1e6e7da25a1 100644 --- a/fs/gfs2/trans.h +++ b/fs/gfs2/trans.h @@ -28,20 +28,20 @@ struct gfs2_glock; /* reserve either the number of blocks to be allocated plus the rg header * block, or all of the blocks in the rg, whichever is smaller */ -static inline unsigned int gfs2_rg_blocks(const struct gfs2_alloc *al) +static inline unsigned int gfs2_rg_blocks(const struct gfs2_inode *ip, unsigned requested) { - return (al->al_requested < al->al_rgd->rd_length)? - al->al_requested + 1 : al->al_rgd->rd_length; + if (requested < ip->i_rgd->rd_length) + return requested + 1; + return ip->i_rgd->rd_length; } -int gfs2_trans_begin(struct gfs2_sbd *sdp, unsigned int blocks, - unsigned int revokes); +extern int gfs2_trans_begin(struct gfs2_sbd *sdp, unsigned int blocks, + unsigned int revokes); -void gfs2_trans_end(struct gfs2_sbd *sdp); - -void gfs2_trans_add_bh(struct gfs2_glock *gl, struct buffer_head *bh, int meta); -void gfs2_trans_add_revoke(struct gfs2_sbd *sdp, struct gfs2_bufdata *bd); -void gfs2_trans_add_unrevoke(struct gfs2_sbd *sdp, u64 blkno, unsigned int len); -void gfs2_trans_add_rg(struct gfs2_rgrpd *rgd); +extern void gfs2_trans_end(struct gfs2_sbd *sdp); +extern void gfs2_trans_add_data(struct gfs2_glock *gl, struct buffer_head *bh); +extern void gfs2_trans_add_meta(struct gfs2_glock *gl, struct buffer_head *bh); +extern void gfs2_trans_add_revoke(struct gfs2_sbd *sdp, struct gfs2_bufdata *bd); +extern void gfs2_trans_add_unrevoke(struct gfs2_sbd *sdp, u64 blkno, unsigned int len); #endif /* __TRANS_DOT_H__ */ diff --git a/fs/gfs2/util.c b/fs/gfs2/util.c index 53511291fe3..86d2035ac66 100644 --- a/fs/gfs2/util.c +++ b/fs/gfs2/util.c @@ -7,6 +7,8 @@ * of the GNU General Public License version 2. */ +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + #include <linux/spinlock.h> #include <linux/completion.h> #include <linux/buffer_head.h> @@ -25,25 +27,32 @@ struct kmem_cache *gfs2_inode_cachep __read_mostly; struct kmem_cache *gfs2_bufdata_cachep __read_mostly; struct kmem_cache *gfs2_rgrpd_cachep __read_mostly; struct kmem_cache *gfs2_quotad_cachep __read_mostly; +struct kmem_cache *gfs2_rsrv_cachep __read_mostly; +mempool_t *gfs2_page_pool __read_mostly; void gfs2_assert_i(struct gfs2_sbd *sdp) { - printk(KERN_EMERG "GFS2: fsid=%s: fatal assertion failed\n", - sdp->sd_fsname); + fs_emerg(sdp, "fatal assertion failed\n"); } -int gfs2_lm_withdraw(struct gfs2_sbd *sdp, char *fmt, ...) +int gfs2_lm_withdraw(struct gfs2_sbd *sdp, const char *fmt, ...) { struct lm_lockstruct *ls = &sdp->sd_lockstruct; const struct lm_lockops *lm = ls->ls_ops; va_list args; + struct va_format vaf; if (sdp->sd_args.ar_errors == GFS2_ERRORS_WITHDRAW && test_and_set_bit(SDF_SHUTDOWN, &sdp->sd_flags)) return 0; va_start(args, fmt); - vprintk(fmt, args); + + vaf.fmt = fmt; + vaf.va = &args; + + fs_err(sdp, "%pV", &vaf); + va_end(args); if (sdp->sd_args.ar_errors == GFS2_ERRORS_WITHDRAW) { @@ -52,6 +61,9 @@ int gfs2_lm_withdraw(struct gfs2_sbd *sdp, char *fmt, ...) kobject_uevent(&sdp->sd_kobj, KOBJ_OFFLINE); + if (!strcmp(sdp->sd_lockstruct.ls_ops->lm_proto_name, "lock_dlm")) + wait_for_completion(&sdp->sd_wdack); + if (lm->lm_unmount) { fs_err(sdp, "telling LM to unmount\n"); lm->lm_unmount(sdp); @@ -61,7 +73,7 @@ int gfs2_lm_withdraw(struct gfs2_sbd *sdp, char *fmt, ...) } if (sdp->sd_args.ar_errors == GFS2_ERRORS_PANIC) - panic("GFS2: fsid=%s: panic requested.\n", sdp->sd_fsname); + panic("GFS2: fsid=%s: panic requested\n", sdp->sd_fsname); return -1; } @@ -77,10 +89,9 @@ int gfs2_assert_withdraw_i(struct gfs2_sbd *sdp, char *assertion, { int me; me = gfs2_lm_withdraw(sdp, - "GFS2: fsid=%s: fatal: assertion \"%s\" failed\n" - "GFS2: fsid=%s: function = %s, file = %s, line = %u\n", - sdp->sd_fsname, assertion, - sdp->sd_fsname, function, file, line); + "fatal: assertion \"%s\" failed\n" + " function = %s, file = %s, line = %u\n", + assertion, function, file, line); dump_stack(); return (me) ? -1 : -2; } @@ -100,11 +111,8 @@ int gfs2_assert_warn_i(struct gfs2_sbd *sdp, char *assertion, return -2; if (sdp->sd_args.ar_errors == GFS2_ERRORS_WITHDRAW) - printk(KERN_WARNING - "GFS2: fsid=%s: warning: assertion \"%s\" failed\n" - "GFS2: fsid=%s: function = %s, file = %s, line = %u\n", - sdp->sd_fsname, assertion, - sdp->sd_fsname, function, file, line); + fs_warn(sdp, "warning: assertion \"%s\" failed at function = %s, file = %s, line = %u\n", + assertion, function, file, line); if (sdp->sd_args.ar_debug) BUG(); @@ -133,10 +141,8 @@ int gfs2_consist_i(struct gfs2_sbd *sdp, int cluster_wide, const char *function, { int rv; rv = gfs2_lm_withdraw(sdp, - "GFS2: fsid=%s: fatal: filesystem consistency error\n" - "GFS2: fsid=%s: function = %s, file = %s, line = %u\n", - sdp->sd_fsname, - sdp->sd_fsname, function, file, line); + "fatal: filesystem consistency error - function = %s, file = %s, line = %u\n", + function, file, line); return rv; } @@ -152,13 +158,12 @@ int gfs2_consist_inode_i(struct gfs2_inode *ip, int cluster_wide, struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); int rv; rv = gfs2_lm_withdraw(sdp, - "GFS2: fsid=%s: fatal: filesystem consistency error\n" - "GFS2: fsid=%s: inode = %llu %llu\n" - "GFS2: fsid=%s: function = %s, file = %s, line = %u\n", - sdp->sd_fsname, - sdp->sd_fsname, (unsigned long long)ip->i_no_formal_ino, - (unsigned long long)ip->i_no_addr, - sdp->sd_fsname, function, file, line); + "fatal: filesystem consistency error\n" + " inode = %llu %llu\n" + " function = %s, file = %s, line = %u\n", + (unsigned long long)ip->i_no_formal_ino, + (unsigned long long)ip->i_no_addr, + function, file, line); return rv; } @@ -174,12 +179,11 @@ int gfs2_consist_rgrpd_i(struct gfs2_rgrpd *rgd, int cluster_wide, struct gfs2_sbd *sdp = rgd->rd_sbd; int rv; rv = gfs2_lm_withdraw(sdp, - "GFS2: fsid=%s: fatal: filesystem consistency error\n" - "GFS2: fsid=%s: RG = %llu\n" - "GFS2: fsid=%s: function = %s, file = %s, line = %u\n", - sdp->sd_fsname, - sdp->sd_fsname, (unsigned long long)rgd->rd_addr, - sdp->sd_fsname, function, file, line); + "fatal: filesystem consistency error\n" + " RG = %llu\n" + " function = %s, file = %s, line = %u\n", + (unsigned long long)rgd->rd_addr, + function, file, line); return rv; } @@ -195,12 +199,11 @@ int gfs2_meta_check_ii(struct gfs2_sbd *sdp, struct buffer_head *bh, { int me; me = gfs2_lm_withdraw(sdp, - "GFS2: fsid=%s: fatal: invalid metadata block\n" - "GFS2: fsid=%s: bh = %llu (%s)\n" - "GFS2: fsid=%s: function = %s, file = %s, line = %u\n", - sdp->sd_fsname, - sdp->sd_fsname, (unsigned long long)bh->b_blocknr, type, - sdp->sd_fsname, function, file, line); + "fatal: invalid metadata block\n" + " bh = %llu (%s)\n" + " function = %s, file = %s, line = %u\n", + (unsigned long long)bh->b_blocknr, type, + function, file, line); return (me) ? -1 : -2; } @@ -216,12 +219,11 @@ int gfs2_metatype_check_ii(struct gfs2_sbd *sdp, struct buffer_head *bh, { int me; me = gfs2_lm_withdraw(sdp, - "GFS2: fsid=%s: fatal: invalid metadata block\n" - "GFS2: fsid=%s: bh = %llu (type: exp=%u, found=%u)\n" - "GFS2: fsid=%s: function = %s, file = %s, line = %u\n", - sdp->sd_fsname, - sdp->sd_fsname, (unsigned long long)bh->b_blocknr, type, t, - sdp->sd_fsname, function, file, line); + "fatal: invalid metadata block\n" + " bh = %llu (type: exp=%u, found=%u)\n" + " function = %s, file = %s, line = %u\n", + (unsigned long long)bh->b_blocknr, type, t, + function, file, line); return (me) ? -1 : -2; } @@ -236,10 +238,9 @@ int gfs2_io_error_i(struct gfs2_sbd *sdp, const char *function, char *file, { int rv; rv = gfs2_lm_withdraw(sdp, - "GFS2: fsid=%s: fatal: I/O error\n" - "GFS2: fsid=%s: function = %s, file = %s, line = %u\n", - sdp->sd_fsname, - sdp->sd_fsname, function, file, line); + "fatal: I/O error\n" + " function = %s, file = %s, line = %u\n", + function, file, line); return rv; } @@ -254,32 +255,11 @@ int gfs2_io_error_bh_i(struct gfs2_sbd *sdp, struct buffer_head *bh, { int rv; rv = gfs2_lm_withdraw(sdp, - "GFS2: fsid=%s: fatal: I/O error\n" - "GFS2: fsid=%s: block = %llu\n" - "GFS2: fsid=%s: function = %s, file = %s, line = %u\n", - sdp->sd_fsname, - sdp->sd_fsname, (unsigned long long)bh->b_blocknr, - sdp->sd_fsname, function, file, line); + "fatal: I/O error\n" + " block = %llu\n" + " function = %s, file = %s, line = %u\n", + (unsigned long long)bh->b_blocknr, + function, file, line); return rv; } -void gfs2_icbit_munge(struct gfs2_sbd *sdp, unsigned char **bitmap, - unsigned int bit, int new_value) -{ - unsigned int c, o, b = bit; - int old_value; - - c = b / (8 * PAGE_SIZE); - b %= 8 * PAGE_SIZE; - o = b / 8; - b %= 8; - - old_value = (bitmap[c][o] & (1 << b)); - gfs2_assert_withdraw(sdp, !old_value != !new_value); - - if (new_value) - bitmap[c][o] |= 1 << b; - else - bitmap[c][o] &= ~(1 << b); -} - diff --git a/fs/gfs2/util.h b/fs/gfs2/util.h index b432e04600d..cbdcbdf3961 100644 --- a/fs/gfs2/util.h +++ b/fs/gfs2/util.h @@ -10,20 +10,23 @@ #ifndef __UTIL_DOT_H__ #define __UTIL_DOT_H__ -#include "incore.h" - -#define fs_printk(level, fs, fmt, arg...) \ - printk(level "GFS2: fsid=%s: " fmt , (fs)->sd_fsname , ## arg) - -#define fs_info(fs, fmt, arg...) \ - fs_printk(KERN_INFO , fs , fmt , ## arg) +#ifdef pr_fmt +#undef pr_fmt +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt +#endif -#define fs_warn(fs, fmt, arg...) \ - fs_printk(KERN_WARNING , fs , fmt , ## arg) +#include <linux/mempool.h> -#define fs_err(fs, fmt, arg...) \ - fs_printk(KERN_ERR, fs , fmt , ## arg) +#include "incore.h" +#define fs_emerg(fs, fmt, ...) \ + pr_emerg("fsid=%s: " fmt, (fs)->sd_fsname, ##__VA_ARGS__) +#define fs_warn(fs, fmt, ...) \ + pr_warn("fsid=%s: " fmt, (fs)->sd_fsname, ##__VA_ARGS__) +#define fs_err(fs, fmt, ...) \ + pr_err("fsid=%s: " fmt, (fs)->sd_fsname, ##__VA_ARGS__) +#define fs_info(fs, fmt, ...) \ + pr_info("fsid=%s: " fmt, (fs)->sd_fsname, ##__VA_ARGS__) void gfs2_assert_i(struct gfs2_sbd *sdp); @@ -77,23 +80,19 @@ int gfs2_meta_check_ii(struct gfs2_sbd *sdp, struct buffer_head *bh, const char *type, const char *function, char *file, unsigned int line); -static inline int gfs2_meta_check_i(struct gfs2_sbd *sdp, - struct buffer_head *bh, - const char *function, - char *file, unsigned int line) +static inline int gfs2_meta_check(struct gfs2_sbd *sdp, + struct buffer_head *bh) { struct gfs2_meta_header *mh = (struct gfs2_meta_header *)bh->b_data; u32 magic = be32_to_cpu(mh->mh_magic); - if (unlikely(magic != GFS2_MAGIC)) - return gfs2_meta_check_ii(sdp, bh, "magic number", function, - file, line); + if (unlikely(magic != GFS2_MAGIC)) { + pr_err("Magic number missing at %llu\n", + (unsigned long long)bh->b_blocknr); + return -EIO; + } return 0; } -#define gfs2_meta_check(sdp, bh) \ -gfs2_meta_check_i((sdp), (bh), __func__, __FILE__, __LINE__) - - int gfs2_metatype_check_ii(struct gfs2_sbd *sdp, struct buffer_head *bh, u16 type, u16 t, const char *function, @@ -150,6 +149,8 @@ extern struct kmem_cache *gfs2_inode_cachep; extern struct kmem_cache *gfs2_bufdata_cachep; extern struct kmem_cache *gfs2_rgrpd_cachep; extern struct kmem_cache *gfs2_quotad_cachep; +extern struct kmem_cache *gfs2_rsrv_cachep; +extern mempool_t *gfs2_page_pool; static inline unsigned int gfs2_tune_get_i(struct gfs2_tune *gt, unsigned int *p) @@ -164,9 +165,7 @@ static inline unsigned int gfs2_tune_get_i(struct gfs2_tune *gt, #define gfs2_tune_get(sdp, field) \ gfs2_tune_get_i(&(sdp)->sd_tune, &(sdp)->sd_tune.field) -void gfs2_icbit_munge(struct gfs2_sbd *sdp, unsigned char **bitmap, - unsigned int bit, int new_value); -int gfs2_lm_withdraw(struct gfs2_sbd *sdp, char *fmt, ...); +__printf(2, 3) +int gfs2_lm_withdraw(struct gfs2_sbd *sdp, const char *fmt, ...); #endif /* __UTIL_DOT_H__ */ - diff --git a/fs/gfs2/xattr.c b/fs/gfs2/xattr.c index 439b61c0326..0b81f783f78 100644 --- a/fs/gfs2/xattr.c +++ b/fs/gfs2/xattr.c @@ -13,6 +13,7 @@ #include <linux/buffer_head.h> #include <linux/xattr.h> #include <linux/gfs2_ondisk.h> +#include <linux/posix_acl_xattr.h> #include <asm/uaccess.h> #include "gfs2.h" @@ -238,6 +239,10 @@ static int ea_dealloc_unstuffed(struct gfs2_inode *ip, struct buffer_head *bh, unsigned int x; int error; + error = gfs2_rindex_update(sdp); + if (error) + return error; + if (GFS2_EA_IS_STUFFED(ea)) return 0; @@ -251,7 +256,7 @@ static int ea_dealloc_unstuffed(struct gfs2_inode *ip, struct buffer_head *bh, if (!blks) return 0; - rgd = gfs2_blk2rgrpd(sdp, bn); + rgd = gfs2_blk2rgrpd(sdp, bn, 1); if (!rgd) { gfs2_consist_inode(ip); return -EIO; @@ -266,7 +271,7 @@ static int ea_dealloc_unstuffed(struct gfs2_inode *ip, struct buffer_head *bh, if (error) goto out_gunlock; - gfs2_trans_add_bh(ip->i_gl, bh, 1); + gfs2_trans_add_meta(ip->i_gl, bh); dataptrs = GFS2_EA2DATAPTRS(ea); for (x = 0; x < ea->ea_num_ptrs; x++, dataptrs++) { @@ -305,7 +310,7 @@ static int ea_dealloc_unstuffed(struct gfs2_inode *ip, struct buffer_head *bh, error = gfs2_meta_inode_buffer(ip, &dibh); if (!error) { ip->i_inode.i_ctime = CURRENT_TIME; - gfs2_trans_add_bh(ip->i_gl, dibh, 1); + gfs2_trans_add_meta(ip->i_gl, dibh); gfs2_dinode_out(ip, dibh->b_data); brelse(dibh); } @@ -321,29 +326,20 @@ static int ea_remove_unstuffed(struct gfs2_inode *ip, struct buffer_head *bh, struct gfs2_ea_header *ea, struct gfs2_ea_header *prev, int leave) { - struct gfs2_alloc *al; int error; - al = gfs2_alloc_get(ip); - if (!al) - return -ENOMEM; - - error = gfs2_quota_hold(ip, NO_QUOTA_CHANGE, NO_QUOTA_CHANGE); + error = gfs2_rindex_update(GFS2_SB(&ip->i_inode)); if (error) - goto out_alloc; + return error; - error = gfs2_rindex_hold(GFS2_SB(&ip->i_inode), &al->al_ri_gh); + error = gfs2_quota_hold(ip, NO_UID_QUOTA_CHANGE, NO_GID_QUOTA_CHANGE); if (error) - goto out_quota; + goto out_alloc; error = ea_dealloc_unstuffed(ip, bh, ea, prev, (leave) ? &error : NULL); - gfs2_glock_dq_uninit(&al->al_ri_gh); - -out_quota: gfs2_quota_unhold(ip); out_alloc: - gfs2_alloc_put(ip); return error; } @@ -453,17 +449,18 @@ ssize_t gfs2_listxattr(struct dentry *dentry, char *buffer, size_t size) } /** - * ea_get_unstuffed - actually copies the unstuffed data into the - * request buffer + * ea_iter_unstuffed - copies the unstuffed xattr data to/from the + * request buffer * @ip: The GFS2 inode * @ea: The extended attribute header structure - * @data: The data to be copied + * @din: The data to be copied in + * @dout: The data to be copied out (one of din,dout will be NULL) * * Returns: errno */ -static int ea_get_unstuffed(struct gfs2_inode *ip, struct gfs2_ea_header *ea, - char *data) +static int gfs2_iter_unstuffed(struct gfs2_inode *ip, struct gfs2_ea_header *ea, + const char *din, char *dout) { struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); struct buffer_head **bh; @@ -472,6 +469,8 @@ static int ea_get_unstuffed(struct gfs2_inode *ip, struct gfs2_ea_header *ea, __be64 *dataptrs = GFS2_EA2DATAPTRS(ea); unsigned int x; int error = 0; + unsigned char *pos; + unsigned cp_size; bh = kcalloc(nptrs, sizeof(struct buffer_head *), GFP_NOFS); if (!bh) @@ -502,12 +501,21 @@ static int ea_get_unstuffed(struct gfs2_inode *ip, struct gfs2_ea_header *ea, goto out; } - memcpy(data, bh[x]->b_data + sizeof(struct gfs2_meta_header), - (sdp->sd_jbsize > amount) ? amount : sdp->sd_jbsize); + pos = bh[x]->b_data + sizeof(struct gfs2_meta_header); + cp_size = (sdp->sd_jbsize > amount) ? amount : sdp->sd_jbsize; - amount -= sdp->sd_jbsize; - data += sdp->sd_jbsize; + if (dout) { + memcpy(dout, pos, cp_size); + dout += sdp->sd_jbsize; + } + + if (din) { + gfs2_trans_add_meta(ip->i_gl, bh[x]); + memcpy(pos, din, cp_size); + din += sdp->sd_jbsize; + } + amount -= sdp->sd_jbsize; brelse(bh[x]); } @@ -528,7 +536,7 @@ static int gfs2_ea_get_copy(struct gfs2_inode *ip, struct gfs2_ea_location *el, memcpy(data, GFS2_EA2DATA(el->el_ea), len); return len; } - ret = ea_get_unstuffed(ip, el->el_ea, data); + ret = gfs2_iter_unstuffed(ip, el->el_ea, NULL, data); if (ret < 0) return ret; return len; @@ -556,9 +564,10 @@ int gfs2_xattr_acl_get(struct gfs2_inode *ip, const char *name, char **ppdata) goto out; error = gfs2_ea_get_copy(ip, &el, data, len); - if (error == 0) - error = len; - *ppdata = data; + if (error < 0) + kfree(data); + else + *ppdata = data; out: brelse(el.el_bh); return error; @@ -616,12 +625,12 @@ static int ea_alloc_blk(struct gfs2_inode *ip, struct buffer_head **bhp) u64 block; int error; - error = gfs2_alloc_block(ip, &block, &n); + error = gfs2_alloc_blocks(ip, &block, &n, 0, NULL); if (error) return error; gfs2_trans_add_unrevoke(sdp, block, 1); *bhp = gfs2_meta_new(ip->i_gl, block); - gfs2_trans_add_bh(ip->i_gl, *bhp, 1); + gfs2_trans_add_meta(ip->i_gl, *bhp); gfs2_metatype_set(*bhp, GFS2_METATYPE_EA, GFS2_FORMAT_EA); gfs2_buffer_clear_tail(*bhp, sizeof(struct gfs2_meta_header)); @@ -678,12 +687,12 @@ static int ea_write(struct gfs2_inode *ip, struct gfs2_ea_header *ea, int mh_size = sizeof(struct gfs2_meta_header); unsigned int n = 1; - error = gfs2_alloc_block(ip, &block, &n); + error = gfs2_alloc_blocks(ip, &block, &n, 0, NULL); if (error) return error; gfs2_trans_add_unrevoke(sdp, block, 1); bh = gfs2_meta_new(ip->i_gl, block); - gfs2_trans_add_bh(ip->i_gl, bh, 1); + gfs2_trans_add_meta(ip->i_gl, bh); gfs2_metatype_set(bh, GFS2_METATYPE_ED, GFS2_FORMAT_ED); gfs2_add_inode_blocks(&ip->i_inode, 1); @@ -715,26 +724,24 @@ static int ea_alloc_skeleton(struct gfs2_inode *ip, struct gfs2_ea_request *er, unsigned int blks, ea_skeleton_call_t skeleton_call, void *private) { - struct gfs2_alloc *al; + struct gfs2_alloc_parms ap = { .target = blks }; struct buffer_head *dibh; int error; - al = gfs2_alloc_get(ip); - if (!al) - return -ENOMEM; + error = gfs2_rindex_update(GFS2_SB(&ip->i_inode)); + if (error) + return error; error = gfs2_quota_lock_check(ip); if (error) - goto out; - - al->al_requested = blks; + return error; - error = gfs2_inplace_reserve(ip); + error = gfs2_inplace_reserve(ip, &ap); if (error) goto out_gunlock_q; error = gfs2_trans_begin(GFS2_SB(&ip->i_inode), - blks + gfs2_rg_blocks(al) + + blks + gfs2_rg_blocks(ip, blks) + RES_DINODE + RES_STATFS + RES_QUOTA, 0); if (error) goto out_ipres; @@ -746,7 +753,7 @@ static int ea_alloc_skeleton(struct gfs2_inode *ip, struct gfs2_ea_request *er, error = gfs2_meta_inode_buffer(ip, &dibh); if (!error) { ip->i_inode.i_ctime = CURRENT_TIME; - gfs2_trans_add_bh(ip->i_gl, dibh, 1); + gfs2_trans_add_meta(ip->i_gl, dibh); gfs2_dinode_out(ip, dibh->b_data); brelse(dibh); } @@ -757,8 +764,6 @@ out_ipres: gfs2_inplace_release(ip); out_gunlock_q: gfs2_quota_unlock(ip); -out: - gfs2_alloc_put(ip); return error; } @@ -831,7 +836,7 @@ static void ea_set_remove_stuffed(struct gfs2_inode *ip, struct gfs2_ea_header *prev = el->el_prev; u32 len; - gfs2_trans_add_bh(ip->i_gl, el->el_bh, 1); + gfs2_trans_add_meta(ip->i_gl, el->el_bh); if (!prev || !GFS2_EA_IS_STUFFED(ea)) { ea->ea_type = GFS2_EATYPE_UNUSED; @@ -869,7 +874,7 @@ static int ea_set_simple_noalloc(struct gfs2_inode *ip, struct buffer_head *bh, if (error) return error; - gfs2_trans_add_bh(ip->i_gl, bh, 1); + gfs2_trans_add_meta(ip->i_gl, bh); if (es->ea_split) ea = ea_split_ea(ea); @@ -883,7 +888,7 @@ static int ea_set_simple_noalloc(struct gfs2_inode *ip, struct buffer_head *bh, if (error) goto out; ip->i_inode.i_ctime = CURRENT_TIME; - gfs2_trans_add_bh(ip->i_gl, dibh, 1); + gfs2_trans_add_meta(ip->i_gl, dibh); gfs2_dinode_out(ip, dibh->b_data); brelse(dibh); out: @@ -898,7 +903,7 @@ static int ea_set_simple_alloc(struct gfs2_inode *ip, struct gfs2_ea_header *ea = es->es_ea; int error; - gfs2_trans_add_bh(ip->i_gl, es->es_bh, 1); + gfs2_trans_add_meta(ip->i_gl, es->es_bh); if (es->ea_split) ea = ea_split_ea(ea); @@ -994,16 +999,16 @@ static int ea_set_block(struct gfs2_inode *ip, struct gfs2_ea_request *er, goto out; } - gfs2_trans_add_bh(ip->i_gl, indbh, 1); + gfs2_trans_add_meta(ip->i_gl, indbh); } else { u64 blk; unsigned int n = 1; - error = gfs2_alloc_block(ip, &blk, &n); + error = gfs2_alloc_blocks(ip, &blk, &n, 0, NULL); if (error) return error; gfs2_trans_add_unrevoke(sdp, blk, 1); indbh = gfs2_meta_new(ip->i_gl, blk); - gfs2_trans_add_bh(ip->i_gl, indbh, 1); + gfs2_trans_add_meta(ip->i_gl, indbh); gfs2_metatype_set(indbh, GFS2_METATYPE_IN, GFS2_FORMAT_IN); gfs2_buffer_clear_tail(indbh, mh_size); @@ -1089,7 +1094,7 @@ static int ea_remove_stuffed(struct gfs2_inode *ip, struct gfs2_ea_location *el) if (error) return error; - gfs2_trans_add_bh(ip->i_gl, el->el_bh, 1); + gfs2_trans_add_meta(ip->i_gl, el->el_bh); if (prev) { u32 len; @@ -1106,7 +1111,7 @@ static int ea_remove_stuffed(struct gfs2_inode *ip, struct gfs2_ea_location *el) error = gfs2_meta_inode_buffer(ip, &dibh); if (!error) { ip->i_inode.i_ctime = CURRENT_TIME; - gfs2_trans_add_bh(ip->i_gl, dibh, 1); + gfs2_trans_add_meta(ip->i_gl, dibh); gfs2_dinode_out(ip, dibh->b_data); brelse(dibh); } @@ -1229,74 +1234,29 @@ static int gfs2_xattr_set(struct dentry *dentry, const char *name, size, flags, type); } + static int ea_acl_chmod_unstuffed(struct gfs2_inode *ip, struct gfs2_ea_header *ea, char *data) { struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); - struct buffer_head **bh; unsigned int amount = GFS2_EA_DATA_LEN(ea); unsigned int nptrs = DIV_ROUND_UP(amount, sdp->sd_jbsize); - __be64 *dataptrs = GFS2_EA2DATAPTRS(ea); - unsigned int x; - int error; - - bh = kcalloc(nptrs, sizeof(struct buffer_head *), GFP_NOFS); - if (!bh) - return -ENOMEM; - - error = gfs2_trans_begin(sdp, nptrs + RES_DINODE, 0); - if (error) - goto out; - - for (x = 0; x < nptrs; x++) { - error = gfs2_meta_read(ip->i_gl, be64_to_cpu(*dataptrs), 0, - bh + x); - if (error) { - while (x--) - brelse(bh[x]); - goto fail; - } - dataptrs++; - } - - for (x = 0; x < nptrs; x++) { - error = gfs2_meta_wait(sdp, bh[x]); - if (error) { - for (; x < nptrs; x++) - brelse(bh[x]); - goto fail; - } - if (gfs2_metatype_check(sdp, bh[x], GFS2_METATYPE_ED)) { - for (; x < nptrs; x++) - brelse(bh[x]); - error = -EIO; - goto fail; - } - - gfs2_trans_add_bh(ip->i_gl, bh[x], 1); - - memcpy(bh[x]->b_data + sizeof(struct gfs2_meta_header), data, - (sdp->sd_jbsize > amount) ? amount : sdp->sd_jbsize); - - amount -= sdp->sd_jbsize; - data += sdp->sd_jbsize; - - brelse(bh[x]); - } + int ret; -out: - kfree(bh); - return error; + ret = gfs2_trans_begin(sdp, nptrs + RES_DINODE, 0); + if (ret) + return ret; -fail: + ret = gfs2_iter_unstuffed(ip, ea, data, NULL); gfs2_trans_end(sdp); - kfree(bh); - return error; + + return ret; } int gfs2_xattr_acl_chmod(struct gfs2_inode *ip, struct iattr *attr, char *data) { - struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); + struct inode *inode = &ip->i_inode; + struct gfs2_sbd *sdp = GFS2_SB(inode); struct gfs2_ea_location el; int error; @@ -1307,7 +1267,7 @@ int gfs2_xattr_acl_chmod(struct gfs2_inode *ip, struct iattr *attr, char *data) if (GFS2_EA_IS_STUFFED(el.el_ea)) { error = gfs2_trans_begin(sdp, RES_DINODE + RES_EATTR, 0); if (error == 0) { - gfs2_trans_add_bh(ip->i_gl, el.el_bh, 1); + gfs2_trans_add_meta(ip->i_gl, el.el_bh); memcpy(GFS2_EA2DATA(el.el_ea), data, GFS2_EA_DATA_LEN(el.el_ea)); } @@ -1319,7 +1279,7 @@ int gfs2_xattr_acl_chmod(struct gfs2_inode *ip, struct iattr *attr, char *data) if (error) return error; - error = gfs2_setattr_simple(ip, attr); + error = gfs2_setattr_simple(inode, attr); gfs2_trans_end(sdp); return error; } @@ -1337,6 +1297,10 @@ static int ea_dealloc_indirect(struct gfs2_inode *ip) unsigned int x; int error; + error = gfs2_rindex_update(sdp); + if (error) + return error; + memset(&rlist, 0, sizeof(struct gfs2_rgrp_list)); error = gfs2_meta_read(ip->i_gl, ip->i_eattr, DIO_WAIT, &indbh); @@ -1362,14 +1326,14 @@ static int ea_dealloc_indirect(struct gfs2_inode *ip) blen++; else { if (bstart) - gfs2_rlist_add(sdp, &rlist, bstart); + gfs2_rlist_add(ip, &rlist, bstart); bstart = bn; blen = 1; } blks++; } if (bstart) - gfs2_rlist_add(sdp, &rlist, bstart); + gfs2_rlist_add(ip, &rlist, bstart); else goto out; @@ -1390,7 +1354,7 @@ static int ea_dealloc_indirect(struct gfs2_inode *ip) if (error) goto out_gunlock; - gfs2_trans_add_bh(ip->i_gl, indbh, 1); + gfs2_trans_add_meta(ip->i_gl, indbh); eablk = (__be64 *)(indbh->b_data + sizeof(struct gfs2_meta_header)); bstart = 0; @@ -1422,7 +1386,7 @@ static int ea_dealloc_indirect(struct gfs2_inode *ip) error = gfs2_meta_inode_buffer(ip, &dibh); if (!error) { - gfs2_trans_add_bh(ip->i_gl, dibh, 1); + gfs2_trans_add_meta(ip->i_gl, dibh); gfs2_dinode_out(ip, dibh->b_data); brelse(dibh); } @@ -1441,19 +1405,22 @@ out: static int ea_dealloc_block(struct gfs2_inode *ip) { struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); - struct gfs2_alloc *al = ip->i_alloc; struct gfs2_rgrpd *rgd; struct buffer_head *dibh; + struct gfs2_holder gh; int error; - rgd = gfs2_blk2rgrpd(sdp, ip->i_eattr); + error = gfs2_rindex_update(sdp); + if (error) + return error; + + rgd = gfs2_blk2rgrpd(sdp, ip->i_eattr, 1); if (!rgd) { gfs2_consist_inode(ip); return -EIO; } - error = gfs2_glock_nq_init(rgd->rd_gl, LM_ST_EXCLUSIVE, 0, - &al->al_rgd_gh); + error = gfs2_glock_nq_init(rgd->rd_gl, LM_ST_EXCLUSIVE, 0, &gh); if (error) return error; @@ -1469,7 +1436,7 @@ static int ea_dealloc_block(struct gfs2_inode *ip) error = gfs2_meta_inode_buffer(ip, &dibh); if (!error) { - gfs2_trans_add_bh(ip->i_gl, dibh, 1); + gfs2_trans_add_meta(ip->i_gl, dibh); gfs2_dinode_out(ip, dibh->b_data); brelse(dibh); } @@ -1477,7 +1444,7 @@ static int ea_dealloc_block(struct gfs2_inode *ip) gfs2_trans_end(sdp); out_gunlock: - gfs2_glock_dq_uninit(&al->al_rgd_gh); + gfs2_glock_dq_uninit(&gh); return error; } @@ -1490,39 +1457,30 @@ out_gunlock: int gfs2_ea_dealloc(struct gfs2_inode *ip) { - struct gfs2_alloc *al; int error; - al = gfs2_alloc_get(ip); - if (!al) - return -ENOMEM; - - error = gfs2_quota_hold(ip, NO_QUOTA_CHANGE, NO_QUOTA_CHANGE); + error = gfs2_rindex_update(GFS2_SB(&ip->i_inode)); if (error) - goto out_alloc; + return error; - error = gfs2_rindex_hold(GFS2_SB(&ip->i_inode), &al->al_ri_gh); + error = gfs2_quota_hold(ip, NO_UID_QUOTA_CHANGE, NO_GID_QUOTA_CHANGE); if (error) - goto out_quota; + return error; error = ea_foreach(ip, ea_dealloc_unstuffed, NULL); if (error) - goto out_rindex; + goto out_quota; if (ip->i_diskflags & GFS2_DIF_EA_INDIRECT) { error = ea_dealloc_indirect(ip); if (error) - goto out_rindex; + goto out_quota; } error = ea_dealloc_block(ip); -out_rindex: - gfs2_glock_dq_uninit(&al->al_ri_gh); out_quota: gfs2_quota_unhold(ip); -out_alloc: - gfs2_alloc_put(ip); return error; } @@ -1543,7 +1501,8 @@ static const struct xattr_handler gfs2_xattr_security_handler = { const struct xattr_handler *gfs2_xattr_handlers[] = { &gfs2_xattr_user_handler, &gfs2_xattr_security_handler, - &gfs2_xattr_system_handler, + &posix_acl_access_xattr_handler, + &posix_acl_default_xattr_handler, NULL, }; |