1 files changed, 146 insertions, 62 deletions

diff --git a/fs/ocfs2/buffer_head_io.c b/fs/ocfs2/buffer_head_io.c
index f136639f5b4..1edcb141f63 100644
--- a/fs/ocfs2/buffer_head_io.c
+++ b/fs/ocfs2/buffer_head_io.c
@@ -25,7 +25,6 @@
 
 #include <linux/fs.h>
 #include <linux/types.h>
-#include <linux/slab.h>
 #include <linux/highmem.h>
 
 #include <cluster/masklog.h>
@@ -36,16 +35,27 @@
 #include "inode.h"
 #include "journal.h"
 #include "uptodate.h"
-
 #include "buffer_head_io.h"
+#include "ocfs2_trace.h"
+
+/*
+ * Bits on bh->b_state used by ocfs2.
+ *
+ * These MUST be after the JBD2 bits.  Hence, we use BH_JBDPrivateStart.
+ */
+enum ocfs2_state_bits {
+	BH_NeedsValidate = BH_JBDPrivateStart,
+};
+
+/* Expand the magic b_state functions */
+BUFFER_FNS(NeedsValidate, needs_validate);
 
 int ocfs2_write_block(struct ocfs2_super *osb, struct buffer_head *bh,
-		      struct inode *inode)
+		      struct ocfs2_caching_info *ci)
 {
 	int ret = 0;
 
-	mlog_entry("(bh->b_blocknr = %llu, inode=%p)\n",
-		   (unsigned long long)bh->b_blocknr, inode);
+	trace_ocfs2_write_block((unsigned long long)bh->b_blocknr, ci);
 
 	BUG_ON(bh->b_blocknr < OCFS2_SUPER_BLOCK_BLKNO);
 	BUG_ON(buffer_jbd(bh));
@@ -55,10 +65,11 @@ int ocfs2_write_block(struct ocfs2_super *osb, struct buffer_head *bh,
 	 * can get modified during recovery even if read-only. */
 	if (ocfs2_is_hard_readonly(osb)) {
 		ret = -EROFS;
+		mlog_errno(ret);
 		goto out;
 	}
 
-	mutex_lock(&OCFS2_I(inode)->ip_io_mutex);
+	ocfs2_metadata_cache_io_lock(ci);
 
 	lock_buffer(bh);
 	set_buffer_uptodate(bh);
@@ -66,44 +77,119 @@ int ocfs2_write_block(struct ocfs2_super *osb, struct buffer_head *bh,
 	/* remove from dirty list before I/O. */
 	clear_buffer_dirty(bh);
 
-	get_bh(bh); /* for end_buffer_write_sync() */                   
+	get_bh(bh); /* for end_buffer_write_sync() */
 	bh->b_end_io = end_buffer_write_sync;
 	submit_bh(WRITE, bh);
 
 	wait_on_buffer(bh);
 
 	if (buffer_uptodate(bh)) {
-		ocfs2_set_buffer_uptodate(inode, bh);
+		ocfs2_set_buffer_uptodate(ci, bh);
 	} else {
 		/* We don't need to remove the clustered uptodate
 		 * information for this bh as it's not marked locally
 		 * uptodate. */
 		ret = -EIO;
-		put_bh(bh);
+		mlog_errno(ret);
 	}
 
-	mutex_unlock(&OCFS2_I(inode)->ip_io_mutex);
+	ocfs2_metadata_cache_io_unlock(ci);
 out:
-	mlog_exit(ret);
 	return ret;
 }
 
-int ocfs2_read_blocks(struct ocfs2_super *osb, u64 block, int nr,
+int ocfs2_read_blocks_sync(struct ocfs2_super *osb, u64 block,
+			   unsigned int nr, struct buffer_head *bhs[])
+{
+	int status = 0;
+	unsigned int i;
+	struct buffer_head *bh;
+
+	trace_ocfs2_read_blocks_sync((unsigned long long)block, nr);
+
+	if (!nr)
+		goto bail;
+
+	for (i = 0 ; i < nr ; i++) {
+		if (bhs[i] == NULL) {
+			bhs[i] = sb_getblk(osb->sb, block++);
+			if (bhs[i] == NULL) {
+				status = -ENOMEM;
+				mlog_errno(status);
+				goto bail;
+			}
+		}
+		bh = bhs[i];
+
+		if (buffer_jbd(bh)) {
+			trace_ocfs2_read_blocks_sync_jbd(
+					(unsigned long long)bh->b_blocknr);
+			continue;
+		}
+
+		if (buffer_dirty(bh)) {
+			/* This should probably be a BUG, or
+			 * at least return an error. */
+			mlog(ML_ERROR,
+			     "trying to sync read a dirty "
+			     "buffer! (blocknr = %llu), skipping\n",
+			     (unsigned long long)bh->b_blocknr);
+			continue;
+		}
+
+		lock_buffer(bh);
+		if (buffer_jbd(bh)) {
+			mlog(ML_ERROR,
+			     "block %llu had the JBD bit set "
+			     "while I was in lock_buffer!",
+			     (unsigned long long)bh->b_blocknr);
+			BUG();
+		}
+
+		clear_buffer_uptodate(bh);
+		get_bh(bh); /* for end_buffer_read_sync() */
+		bh->b_end_io = end_buffer_read_sync;
+		submit_bh(READ, bh);
+	}
+
+	for (i = nr; i > 0; i--) {
+		bh = bhs[i - 1];
+
+		/* No need to wait on the buffer if it's managed by JBD. */
+		if (!buffer_jbd(bh))
+			wait_on_buffer(bh);
+
+		if (!buffer_uptodate(bh)) {
+			/* Status won't be cleared from here on out,
+			 * so we can safely record this and loop back
+			 * to cleanup the other buffers. */
+			status = -EIO;
+			put_bh(bh);
+			bhs[i - 1] = NULL;
+		}
+	}
+
+bail:
+	return status;
+}
+
+int ocfs2_read_blocks(struct ocfs2_caching_info *ci, u64 block, int nr,
 		      struct buffer_head *bhs[], int flags,
-		      struct inode *inode)
+		      int (*validate)(struct super_block *sb,
+				      struct buffer_head *bh))
 {
 	int status = 0;
-	struct super_block *sb;
 	int i, ignore_cache = 0;
 	struct buffer_head *bh;
+	struct super_block *sb = ocfs2_metadata_cache_get_super(ci);
 
-	mlog_entry("(block=(%llu), nr=(%d), flags=%d, inode=%p)\n",
-		   (unsigned long long)block, nr, flags, inode);
+	trace_ocfs2_read_blocks_begin(ci, (unsigned long long)block, nr, flags);
 
+	BUG_ON(!ci);
 	BUG_ON((flags & OCFS2_BH_READAHEAD) &&
-	       (!inode || !(flags & OCFS2_BH_CACHED)));
+	       (flags & OCFS2_BH_IGNORE_CACHE));
 
-	if (osb == NULL || osb->sb == NULL || bhs == NULL) {
+	if (bhs == NULL) {
 		status = -EINVAL;
 		mlog_errno(status);
 		goto bail;
@@ -117,31 +203,23 @@ int ocfs2_read_blocks(struct ocfs2_super *osb, u64 block, int nr,
 	}
 
 	if (nr == 0) {
-		mlog(ML_BH_IO, "No buffers will be read!\n");
 		status = 0;
 		goto bail;
 	}
 
-	sb = osb->sb;
-
-	if (flags & OCFS2_BH_CACHED && !inode)
-		flags &= ~OCFS2_BH_CACHED;
-
-	if (inode)
-		mutex_lock(&OCFS2_I(inode)->ip_io_mutex);
+	ocfs2_metadata_cache_io_lock(ci);
 	for (i = 0 ; i < nr ; i++) {
 		if (bhs[i] == NULL) {
 			bhs[i] = sb_getblk(sb, block++);
 			if (bhs[i] == NULL) {
-				if (inode)
-					mutex_unlock(&OCFS2_I(inode)->ip_io_mutex);
-				status = -EIO;
+				ocfs2_metadata_cache_io_unlock(ci);
+				status = -ENOMEM;
 				mlog_errno(status);
 				goto bail;
 			}
 		}
 		bh = bhs[i];
-		ignore_cache = 0;
+		ignore_cache = (flags & OCFS2_BH_IGNORE_CACHE);
 
 		/* There are three read-ahead cases here which we need to
 		 * be concerned with. All three assume a buffer has
@@ -167,32 +245,26 @@ int ocfs2_read_blocks(struct ocfs2_super *osb, u64 block, int nr,
 		 *    before our is-it-in-flight check.
 		 */
 
-		if (flags & OCFS2_BH_CACHED &&
-		    !ocfs2_buffer_uptodate(inode, bh)) {
-			mlog(ML_UPTODATE,
-			     "bh (%llu), inode %llu not uptodate\n",
+		if (!ignore_cache && !ocfs2_buffer_uptodate(ci, bh)) {
+			trace_ocfs2_read_blocks_from_disk(
 			     (unsigned long long)bh->b_blocknr,
-			     (unsigned long long)OCFS2_I(inode)->ip_blkno);
+			     (unsigned long long)ocfs2_metadata_cache_owner(ci));
+			/* We're using ignore_cache here to say
+			 * "go to disk" */
 			ignore_cache = 1;
 		}
 
-		/* XXX: Can we ever get this and *not* have the cached
-		 * flag set? */
+		trace_ocfs2_read_blocks_bh((unsigned long long)bh->b_blocknr,
+			ignore_cache, buffer_jbd(bh), buffer_dirty(bh));
+
 		if (buffer_jbd(bh)) {
-			if (!(flags & OCFS2_BH_CACHED) || ignore_cache)
-				mlog(ML_BH_IO, "trying to sync read a jbd "
-					       "managed bh (blocknr = %llu)\n",
-				     (unsigned long long)bh->b_blocknr);
 			continue;
 		}
 
-		if (!(flags & OCFS2_BH_CACHED) || ignore_cache) {
+		if (ignore_cache) {
 			if (buffer_dirty(bh)) {
 				/* This should probably be a BUG, or
 				 * at least return an error. */
-				mlog(ML_BH_IO, "asking me to sync read a dirty "
-					       "buffer! (blocknr = %llu)\n",
-				     (unsigned long long)bh->b_blocknr);
 				continue;
 			}
 
@@ -201,7 +273,7 @@ int ocfs2_read_blocks(struct ocfs2_super *osb, u64 block, int nr,
 			 * previously submitted request than we are
 			 * done here. */
 			if ((flags & OCFS2_BH_READAHEAD)
-			    && ocfs2_buffer_read_ahead(inode, bh))
+			    && ocfs2_buffer_read_ahead(ci, bh))
 				continue;
 
 			lock_buffer(bh);
@@ -221,15 +293,17 @@ int ocfs2_read_blocks(struct ocfs2_super *osb, u64 block, int nr,
 			 * previously read-ahead buffer may have
 			 * completed I/O while we were waiting for the
 			 * buffer lock. */
-			if ((flags & OCFS2_BH_CACHED)
+			if (!(flags & OCFS2_BH_IGNORE_CACHE)
 			    && !(flags & OCFS2_BH_READAHEAD)
-			    && ocfs2_buffer_uptodate(inode, bh)) {
+			    && ocfs2_buffer_uptodate(ci, bh)) {
 				unlock_buffer(bh);
 				continue;
 			}
 
 			clear_buffer_uptodate(bh);
 			get_bh(bh); /* for end_buffer_read_sync() */
+			if (validate)
+				set_buffer_needs_validate(bh);
 			bh->b_end_io = end_buffer_read_sync;
 			submit_bh(READ, bh);
 			continue;
@@ -243,7 +317,7 @@ int ocfs2_read_blocks(struct ocfs2_super *osb, u64 block, int nr,
 
 		if (!(flags & OCFS2_BH_READAHEAD)) {
 			/* We know this can't have changed as we hold the
-			 * inode sem. Avoid doing any work on the bh if the
+			 * owner sem. Avoid doing any work on the bh if the
 			 * journal has it. */
 			if (!buffer_jbd(bh))
 				wait_on_buffer(bh);
@@ -260,24 +334,34 @@ int ocfs2_read_blocks(struct ocfs2_super *osb, u64 block, int nr,
 				bhs[i] = NULL;
 				continue;
 			}
+
+			if (buffer_needs_validate(bh)) {
+				/* We never set NeedsValidate if the
+				 * buffer was held by the journal, so
+				 * that better not have changed */
+				BUG_ON(buffer_jbd(bh));
+				clear_buffer_needs_validate(bh);
+				status = validate(sb, bh);
+				if (status) {
+					put_bh(bh);
+					bhs[i] = NULL;
+					continue;
+				}
+			}
 		}
 
 		/* Always set the buffer in the cache, even if it was
 		 * a forced read, or read-ahead which hasn't yet
 		 * completed. */
-		if (inode)
-			ocfs2_set_buffer_uptodate(inode, bh);
+		ocfs2_set_buffer_uptodate(ci, bh);
 	}
-	if (inode)
-		mutex_unlock(&OCFS2_I(inode)->ip_io_mutex);
+	ocfs2_metadata_cache_io_unlock(ci);
 
-	mlog(ML_BH_IO, "block=(%llu), nr=(%d), cached=%s, flags=0x%x\n", 
-	     (unsigned long long)block, nr,
-	     (!(flags & OCFS2_BH_CACHED) || ignore_cache) ? "no" : "yes", flags);
+	trace_ocfs2_read_blocks_end((unsigned long long)block, nr,
+				    flags, ignore_cache);
 
 bail:
 
-	mlog_exit(status);
 	return status;
 }
 
@@ -302,21 +386,21 @@ static void ocfs2_check_super_or_backup(struct super_block *sb,
 
 /*
  * Write super block and backups doesn't need to collaborate with journal,
- * so we don't need to lock ip_io_mutex and inode doesn't need to bea passed
+ * so we don't need to lock ip_io_mutex and ci doesn't need to bea passed
  * into this function.
  */
 int ocfs2_write_super_or_backup(struct ocfs2_super *osb,
 				struct buffer_head *bh)
 {
 	int ret = 0;
-
-	mlog_entry_void();
+	struct ocfs2_dinode *di = (struct ocfs2_dinode *)bh->b_data;
 
 	BUG_ON(buffer_jbd(bh));
 	ocfs2_check_super_or_backup(osb->sb, bh->b_blocknr);
 
 	if (ocfs2_is_hard_readonly(osb) || ocfs2_is_soft_readonly(osb)) {
 		ret = -EROFS;
+		mlog_errno(ret);
 		goto out;
 	}
 
@@ -328,16 +412,16 @@ int ocfs2_write_super_or_backup(struct ocfs2_super *osb,
 
 	get_bh(bh); /* for end_buffer_write_sync() */
 	bh->b_end_io = end_buffer_write_sync;
+	ocfs2_compute_meta_ecc(osb->sb, bh->b_data, &di->i_check);
 	submit_bh(WRITE, bh);
 
 	wait_on_buffer(bh);
 
 	if (!buffer_uptodate(bh)) {
 		ret = -EIO;
-		put_bh(bh);
+		mlog_errno(ret);
 	}
 
 out:
-	mlog_exit(ret);
 	return ret;
 }