Btrfs: Add zlib compression support

This is a large change for adding compression on reading and writing,
both for inline and regular extents.  It does some fairly large
surgery to the writeback paths.

Compression is off by default and enabled by mount -o compress.  Even
when the -o compress mount option is not used, it is possible to read
compressed extents off the disk.

If compression for a given set of pages fails to make them smaller, the
file is flagged to avoid future compression attempts later.

* While finding delalloc extents, the pages are locked before being sent down
to the delalloc handler.  This allows the delalloc handler to do complex things
such as cleaning the pages, marking them writeback and starting IO on their
behalf.

* Inline extents are inserted at delalloc time now.  This allows us to compress
the data before inserting the inline extent, and it allows us to insert
an inline extent that spans multiple pages.

* All of the in-memory extent representations (extent_map.c, ordered-data.c etc)
are changed to record both an in-memory size and an on disk size, as well
as a flag for compression.

From a disk format point of view, the extent pointers in the file are changed
to record the on disk size of a given extent and some encoding flags.
Space in the disk format is allocated for compression encoding, as well
as encryption and a generic 'other' field.  Neither the encryption or the
'other' field are currently used.

In order to limit the amount of data read for a single random read in the
file, the size of a compressed extent is limited to 128k.  This is a
software only limit, the disk format supports u64 sized compressed extents.

In order to limit the ram consumed while processing extents, the uncompressed
size of a compressed extent is limited to 256k.  This is a software only limit
and will be subject to tuning later.

Checksumming is still done on compressed extents, and it is done on the
uncompressed version of the data.  This way additional encodings can be
layered on without having to figure out which encoding to checksum.

Compression happens at delalloc time, which is basically singled threaded because
it is usually done by a single pdflush thread.  This makes it tricky to
spread the compression load across all the cpus on the box.  We'll have to
look at parallel pdflush walks of dirty inodes at a later time.

Decompression is hooked into readpages and it does spread across CPUs nicely.

Signed-off-by: Chris Mason <chris.mason@oracle.com>

1 files changed, 637 insertions, 0 deletions

diff --git a/fs/btrfs/zlib.c b/fs/btrfs/zlib.c
new file mode 100644
index 00000000000..e99309180a1
--- /dev/null
+++ b/fs/btrfs/zlib.c
@@ -0,0 +1,637 @@
+/*
+ * Copyright (C) 2008 Oracle.  All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public
+ * License v2 as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public
+ * License along with this program; if not, write to the
+ * Free Software Foundation, Inc., 59 Temple Place - Suite 330,
+ * Boston, MA 021110-1307, USA.
+ *
+ * Based on jffs2 zlib code:
+ * Copyright © 2001-2007 Red Hat, Inc.
+ * Created by David Woodhouse <dwmw2@infradead.org>
+ */
+
+#include <linux/kernel.h>
+#include <linux/slab.h>
+#include <linux/zlib.h>
+#include <linux/zutil.h>
+#include <linux/vmalloc.h>
+#include <linux/init.h>
+#include <linux/err.h>
+#include <linux/sched.h>
+#include <linux/pagemap.h>
+#include <linux/bio.h>
+
+/* Plan: call deflate() with avail_in == *sourcelen,
+	avail_out = *dstlen - 12 and flush == Z_FINISH.
+	If it doesn't manage to finish,	call it again with
+	avail_in == 0 and avail_out set to the remaining 12
+	bytes for it to clean up.
+   Q: Is 12 bytes sufficient?
+*/
+#define STREAM_END_SPACE 12
+
+struct workspace {
+	z_stream inf_strm;
+	z_stream def_strm;
+	char *buf;
+	struct list_head list;
+};
+
+static LIST_HEAD(idle_workspace);
+static DEFINE_SPINLOCK(workspace_lock);
+static unsigned long num_workspace;
+static atomic_t alloc_workspace = ATOMIC_INIT(0);
+static DECLARE_WAIT_QUEUE_HEAD(workspace_wait);
+
+/*
+ * this finds an available zlib workspace or allocates a new one
+ * NULL or an ERR_PTR is returned if things go bad.
+ */
+static struct workspace *find_zlib_workspace(void)
+{
+	struct workspace *workspace;
+	int ret;
+	int cpus = num_online_cpus();
+
+again:
+	spin_lock(&workspace_lock);
+	if (!list_empty(&idle_workspace)) {
+		workspace = list_entry(idle_workspace.next, struct workspace,
+				       list);
+		list_del(&workspace->list);
+		num_workspace--;
+		spin_unlock(&workspace_lock);
+		return workspace;
+
+	}
+	spin_unlock(&workspace_lock);
+	if (atomic_read(&alloc_workspace) > cpus) {
+		DEFINE_WAIT(wait);
+		prepare_to_wait(&workspace_wait, &wait, TASK_UNINTERRUPTIBLE);
+		if (atomic_read(&alloc_workspace) > cpus)
+			schedule();
+		finish_wait(&workspace_wait, &wait);
+		goto again;
+	}
+	atomic_inc(&alloc_workspace);
+	workspace = kzalloc(sizeof(*workspace), GFP_NOFS);
+	if (!workspace) {
+		ret = -ENOMEM;
+		goto fail;
+	}
+
+	workspace->def_strm.workspace = vmalloc(zlib_deflate_workspacesize());
+	if (!workspace->def_strm.workspace) {
+		ret = -ENOMEM;
+		goto fail;
+	}
+	workspace->inf_strm.workspace = vmalloc(zlib_inflate_workspacesize());
+	if (!workspace->inf_strm.workspace) {
+		ret = -ENOMEM;
+		goto fail_inflate;
+	}
+	workspace->buf = kmalloc(PAGE_CACHE_SIZE, GFP_NOFS);
+	if (!workspace->buf) {
+		ret = -ENOMEM;
+		goto fail_kmalloc;
+	}
+	return workspace;
+
+fail_kmalloc:
+	vfree(workspace->inf_strm.workspace);
+fail_inflate:
+	vfree(workspace->def_strm.workspace);
+fail:
+	kfree(workspace);
+	atomic_dec(&alloc_workspace);
+	wake_up(&workspace_wait);
+	return ERR_PTR(ret);
+}
+
+/*
+ * put a workspace struct back on the list or free it if we have enough
+ * idle ones sitting around
+ */
+static int free_workspace(struct workspace *workspace)
+{
+	spin_lock(&workspace_lock);
+	if (num_workspace < num_online_cpus()) {
+		list_add_tail(&workspace->list, &idle_workspace);
+		num_workspace++;
+		spin_unlock(&workspace_lock);
+		if (waitqueue_active(&workspace_wait))
+			wake_up(&workspace_wait);
+		return 0;
+	}
+	spin_unlock(&workspace_lock);
+	vfree(workspace->def_strm.workspace);
+	vfree(workspace->inf_strm.workspace);
+	kfree(workspace->buf);
+	kfree(workspace);
+
+	atomic_dec(&alloc_workspace);
+	if (waitqueue_active(&workspace_wait))
+		wake_up(&workspace_wait);
+	return 0;
+}
+
+/*
+ * cleanup function for module exit
+ */
+static void free_workspaces(void)
+{
+	struct workspace *workspace;
+	while(!list_empty(&idle_workspace)) {
+		workspace = list_entry(idle_workspace.next, struct workspace,
+				       list);
+		list_del(&workspace->list);
+		vfree(workspace->def_strm.workspace);
+		vfree(workspace->inf_strm.workspace);
+		kfree(workspace->buf);
+		kfree(workspace);
+		atomic_dec(&alloc_workspace);
+	}
+}
+
+/*
+ * given an address space and start/len, compress the bytes.
+ *
+ * pages are allocated to hold the compressed result and stored
+ * in 'pages'
+ *
+ * out_pages is used to return the number of pages allocated.  There
+ * may be pages allocated even if we return an error
+ *
+ * total_in is used to return the number of bytes actually read.  It
+ * may be smaller then len if we had to exit early because we
+ * ran out of room in the pages array or because we cross the
+ * max_out threshold.
+ *
+ * total_out is used to return the total number of compressed bytes
+ *
+ * max_out tells us the max number of bytes that we're allowed to
+ * stuff into pages
+ */
+int btrfs_zlib_compress_pages(struct address_space *mapping,
+			      u64 start, unsigned long len,
+			      struct page **pages,
+			      unsigned long nr_dest_pages,
+			      unsigned long *out_pages,
+			      unsigned long *total_in,
+			      unsigned long *total_out,
+			      unsigned long max_out)
+{
+	int ret;
+	struct workspace *workspace;
+	char *data_in;
+	char *cpage_out;
+	int nr_pages = 0;
+	struct page *in_page = NULL;
+	struct page *out_page = NULL;
+	int out_written = 0;
+	int in_read = 0;
+	unsigned long bytes_left;
+
+	*out_pages = 0;
+	*total_out = 0;
+	*total_in = 0;
+
+	workspace = find_zlib_workspace();
+	if (!workspace)
+		return -1;
+
+	if (Z_OK != zlib_deflateInit(&workspace->def_strm, 3)) {
+		printk(KERN_WARNING "deflateInit failed\n");
+		ret = -1;
+		goto out;
+	}
+
+	workspace->def_strm.total_in = 0;
+	workspace->def_strm.total_out = 0;
+
+	in_page = find_get_page(mapping, start >> PAGE_CACHE_SHIFT);
+	data_in = kmap(in_page);
+
+	out_page = alloc_page(GFP_NOFS | __GFP_HIGHMEM);
+	cpage_out = kmap(out_page);
+	pages[0] = out_page;
+	nr_pages = 1;
+
+	workspace->def_strm.next_in = data_in;
+	workspace->def_strm.next_out = cpage_out;
+	workspace->def_strm.avail_out = PAGE_CACHE_SIZE;
+	workspace->def_strm.avail_in = min(len, PAGE_CACHE_SIZE);
+
+	out_written = 0;
+	in_read = 0;
+
+	while (workspace->def_strm.total_in < len) {
+		ret = zlib_deflate(&workspace->def_strm, Z_SYNC_FLUSH);
+		if (ret != Z_OK) {
+			printk(KERN_DEBUG "btrfs deflate in loop returned %d\n",
+			       ret);
+			zlib_deflateEnd(&workspace->def_strm);
+			ret = -1;
+			goto out;
+		}
+
+		/* we're making it bigger, give up */
+		if (workspace->def_strm.total_in > 8192 &&
+		    workspace->def_strm.total_in <
+		    workspace->def_strm.total_out) {
+			ret = -1;
+			goto out;
+		}
+		/* we need another page for writing out.  Test this
+		 * before the total_in so we will pull in a new page for
+		 * the stream end if required
+		 */
+		if (workspace->def_strm.avail_out == 0) {
+			kunmap(out_page);
+			if (nr_pages == nr_dest_pages) {
+				out_page = NULL;
+				ret = -1;
+				goto out;
+			}
+			out_page = alloc_page(GFP_NOFS | __GFP_HIGHMEM);
+			cpage_out = kmap(out_page);
+			pages[nr_pages] = out_page;
+			nr_pages++;
+			workspace->def_strm.avail_out = PAGE_CACHE_SIZE;
+			workspace->def_strm.next_out = cpage_out;
+		}
+		/* we're all done */
+		if (workspace->def_strm.total_in >= len)
+			break;
+
+		/* we've read in a full page, get a new one */
+		if (workspace->def_strm.avail_in == 0) {
+			if (workspace->def_strm.total_out > max_out)
+				break;
+
+			bytes_left = len - workspace->def_strm.total_in;
+			kunmap(in_page);
+			page_cache_release(in_page);
+
+			start += PAGE_CACHE_SIZE;
+			in_page = find_get_page(mapping,
+						start >> PAGE_CACHE_SHIFT);
+			data_in = kmap(in_page);
+			workspace->def_strm.avail_in = min(bytes_left,
+							   PAGE_CACHE_SIZE);
+			workspace->def_strm.next_in = data_in;
+		}
+	}
+	workspace->def_strm.avail_in = 0;
+	ret = zlib_deflate(&workspace->def_strm, Z_FINISH);
+	zlib_deflateEnd(&workspace->def_strm);
+
+	if (ret != Z_STREAM_END) {
+		ret = -1;
+		goto out;
+	}
+
+	if (workspace->def_strm.total_out >= workspace->def_strm.total_in) {
+		ret = -1;
+		goto out;
+	}
+
+	ret = 0;
+	*total_out = workspace->def_strm.total_out;
+	*total_in = workspace->def_strm.total_in;
+out:
+	*out_pages = nr_pages;
+	if (out_page)
+		kunmap(out_page);
+
+	if (in_page) {
+		kunmap(in_page);
+		page_cache_release(in_page);
+	}
+	free_workspace(workspace);
+	return ret;
+}
+
+/*
+ * pages_in is an array of pages with compressed data.
+ *
+ * disk_start is the starting logical offset of this array in the file
+ *
+ * bvec is a bio_vec of pages from the file that we want to decompress into
+ *
+ * vcnt is the count of pages in the biovec
+ *
+ * srclen is the number of bytes in pages_in
+ *
+ * The basic idea is that we have a bio that was created by readpages.
+ * The pages in the bio are for the uncompressed data, and they may not
+ * be contiguous.  They all correspond to the range of bytes covered by
+ * the compressed extent.
+ */
+int btrfs_zlib_decompress_biovec(struct page **pages_in,
+			      u64 disk_start,
+			      struct bio_vec *bvec,
+			      int vcnt,
+			      size_t srclen)
+{
+	int ret = 0;
+	int wbits = MAX_WBITS;
+	struct workspace *workspace;
+	char *data_in;
+	size_t total_out = 0;
+	unsigned long page_bytes_left;
+	unsigned long page_in_index = 0;
+	unsigned long page_out_index = 0;
+	struct page *page_out;
+	unsigned long total_pages_in = (srclen + PAGE_CACHE_SIZE - 1) /
+					PAGE_CACHE_SIZE;
+	unsigned long buf_start;
+	unsigned long buf_offset;
+	unsigned long bytes;
+	unsigned long working_bytes;
+	unsigned long pg_offset;
+	unsigned long start_byte;
+	unsigned long current_buf_start;
+	char *kaddr;
+
+	workspace = find_zlib_workspace();
+	if (!workspace)
+		return -ENOMEM;
+
+	data_in = kmap(pages_in[page_in_index]);
+	workspace->inf_strm.next_in = data_in;
+	workspace->inf_strm.avail_in = min(srclen, PAGE_CACHE_SIZE);
+	workspace->inf_strm.total_in = 0;
+
+	workspace->inf_strm.total_out = 0;
+	workspace->inf_strm.next_out = workspace->buf;
+	workspace->inf_strm.avail_out = PAGE_CACHE_SIZE;
+	page_out = bvec[page_out_index].bv_page;
+	page_bytes_left = PAGE_CACHE_SIZE;
+	pg_offset = 0;
+
+	/* If it's deflate, and it's got no preset dictionary, then
+	   we can tell zlib to skip the adler32 check. */
+	if (srclen > 2 && !(data_in[1] & PRESET_DICT) &&
+	    ((data_in[0] & 0x0f) == Z_DEFLATED) &&
+	    !(((data_in[0]<<8) + data_in[1]) % 31)) {
+
+		wbits = -((data_in[0] >> 4) + 8);
+		workspace->inf_strm.next_in += 2;
+		workspace->inf_strm.avail_in -= 2;
+	}
+
+	if (Z_OK != zlib_inflateInit2(&workspace->inf_strm, wbits)) {
+		printk(KERN_WARNING "inflateInit failed\n");
+		ret = -1;
+		goto out;
+	}
+	while(workspace->inf_strm.total_in < srclen) {
+		ret = zlib_inflate(&workspace->inf_strm, Z_NO_FLUSH);
+		if (ret != Z_OK && ret != Z_STREAM_END) {
+			break;
+		}
+
+		/*
+		 * buf start is the byte offset we're of the start of
+		 * our workspace buffer
+		 */
+		buf_start = total_out;
+
+		/* total_out is the last byte of the workspace buffer */
+		total_out = workspace->inf_strm.total_out;
+
+		working_bytes = total_out - buf_start;
+
+		/*
+		 * start byte is the first byte of the page we're currently
+		 * copying into relative to the start of the compressed data.
+		 */
+		start_byte = page_offset(page_out) - disk_start;
+
+		if (working_bytes == 0) {
+			/* we didn't make progress in this inflate
+			 * call, we're done
+			 */
+			if (ret != Z_STREAM_END)
+				ret = -1;
+			break;
+		}
+
+		/* we haven't yet hit data corresponding to this page */
+		if (total_out <= start_byte) {
+			goto next;
+		}
+
+		/*
+		 * the start of the data we care about is offset into
+		 * the middle of our working buffer
+		 */
+		if (total_out > start_byte && buf_start < start_byte) {
+			buf_offset = start_byte - buf_start;
+			working_bytes -= buf_offset;
+		} else {
+			buf_offset = 0;
+		}
+		current_buf_start = buf_start;
+
+		/* copy bytes from the working buffer into the pages */
+		while(working_bytes > 0) {
+			bytes = min(PAGE_CACHE_SIZE - pg_offset,
+				    PAGE_CACHE_SIZE - buf_offset);
+			bytes = min(bytes, working_bytes);
+			kaddr = kmap_atomic(page_out, KM_USER0);
+			memcpy(kaddr + pg_offset, workspace->buf + buf_offset,
+			       bytes);
+			kunmap_atomic(kaddr, KM_USER0);
+			flush_dcache_page(page_out);
+
+			pg_offset += bytes;
+			page_bytes_left -= bytes;
+			buf_offset += bytes;
+			working_bytes -= bytes;
+			current_buf_start += bytes;
+
+			/* check if we need to pick another page */
+			if (page_bytes_left == 0) {
+				page_out_index++;
+				if (page_out_index >= vcnt) {
+					ret = 0;
+					goto done;
+				}
+				page_out = bvec[page_out_index].bv_page;
+				pg_offset = 0;
+				page_bytes_left = PAGE_CACHE_SIZE;
+				start_byte = page_offset(page_out) - disk_start;
+
+				/*
+				 * make sure our new page is covered by this
+				 * working buffer
+				 */
+				if (total_out <= start_byte) {
+					goto next;
+				}
+
+				/* the next page in the biovec might not
+				 * be adjacent to the last page, but it
+				 * might still be found inside this working
+				 * buffer.  bump our offset pointer
+				 */
+				if (total_out > start_byte &&
+				    current_buf_start < start_byte) {
+					buf_offset = start_byte - buf_start;
+					working_bytes = total_out - start_byte;
+					current_buf_start = buf_start +
+						buf_offset;
+				}
+			}
+		}
+next:
+		workspace->inf_strm.next_out = workspace->buf;
+		workspace->inf_strm.avail_out = PAGE_CACHE_SIZE;
+
+		if (workspace->inf_strm.avail_in == 0) {
+			unsigned long tmp;
+			kunmap(pages_in[page_in_index]);
+			page_in_index++;
+			if (page_in_index >= total_pages_in) {
+				data_in = NULL;
+				break;
+			}
+			data_in = kmap(pages_in[page_in_index]);
+			workspace->inf_strm.next_in = data_in;
+			tmp = srclen - workspace->inf_strm.total_in;
+			workspace->inf_strm.avail_in = min(tmp,
+							   PAGE_CACHE_SIZE);
+		}
+	}
+	if (ret != Z_STREAM_END) {
+		ret = -1;
+	} else {
+		ret = 0;
+	}
+done:
+	zlib_inflateEnd(&workspace->inf_strm);
+	if (data_in)
+		kunmap(pages_in[page_in_index]);
+out:
+	free_workspace(workspace);
+	return ret;
+}
+
+/*
+ * a less complex decompression routine.  Our compressed data fits in a
+ * single page, and we want to read a single page out of it.
+ * start_byte tells us the offset into the compressed data we're interested in
+ */
+int btrfs_zlib_decompress(unsigned char *data_in,
+			  struct page *dest_page,
+			  unsigned long start_byte,
+			  size_t srclen, size_t destlen)
+{
+	int ret = 0;
+	int wbits = MAX_WBITS;
+	struct workspace *workspace;
+	unsigned long bytes_left = destlen;
+	unsigned long total_out = 0;
+	char *kaddr;
+
+	if (destlen > PAGE_CACHE_SIZE)
+		return -ENOMEM;
+
+	workspace = find_zlib_workspace();
+	if (!workspace)
+		return -ENOMEM;
+
+	workspace->inf_strm.next_in = data_in;
+	workspace->inf_strm.avail_in = srclen;
+	workspace->inf_strm.total_in = 0;
+
+	workspace->inf_strm.next_out = workspace->buf;
+	workspace->inf_strm.avail_out = PAGE_CACHE_SIZE;
+	workspace->inf_strm.total_out = 0;
+	/* If it's deflate, and it's got no preset dictionary, then
+	   we can tell zlib to skip the adler32 check. */
+	if (srclen > 2 && !(data_in[1] & PRESET_DICT) &&
+	    ((data_in[0] & 0x0f) == Z_DEFLATED) &&
+	    !(((data_in[0]<<8) + data_in[1]) % 31)) {
+
+		wbits = -((data_in[0] >> 4) + 8);
+		workspace->inf_strm.next_in += 2;
+		workspace->inf_strm.avail_in -= 2;
+	}
+
+	if (Z_OK != zlib_inflateInit2(&workspace->inf_strm, wbits)) {
+		printk(KERN_WARNING "inflateInit failed\n");
+		ret = -1;
+		goto out;
+	}
+
+	while(bytes_left > 0) {
+		unsigned long buf_start;
+		unsigned long buf_offset;
+		unsigned long bytes;
+		unsigned long pg_offset = 0;
+
+		ret = zlib_inflate(&workspace->inf_strm, Z_NO_FLUSH);
+		if (ret != Z_OK && ret != Z_STREAM_END) {
+			break;
+		}
+
+		buf_start = total_out;
+		total_out = workspace->inf_strm.total_out;
+
+		if (total_out == buf_start) {
+			ret = -1;
+			break;
+		}
+
+		if (total_out <= start_byte) {
+			goto next;
+		}
+
+		if (total_out > start_byte && buf_start < start_byte) {
+			buf_offset = start_byte - buf_start;
+		} else {
+			buf_offset = 0;
+		}
+
+		bytes = min(PAGE_CACHE_SIZE - pg_offset,
+			    PAGE_CACHE_SIZE - buf_offset);
+		bytes = min(bytes, bytes_left);
+
+		kaddr = kmap_atomic(dest_page, KM_USER0);
+		memcpy(kaddr + pg_offset, workspace->buf + buf_offset, bytes);
+		kunmap_atomic(kaddr, KM_USER0);
+
+		pg_offset += bytes;
+		bytes_left -= bytes;
+next:
+		workspace->inf_strm.next_out = workspace->buf;
+		workspace->inf_strm.avail_out = PAGE_CACHE_SIZE;
+	}
+	if (ret != Z_STREAM_END && bytes_left != 0) {
+		ret = -1;
+	} else {
+		ret = 0;
+	}
+	zlib_inflateEnd(&workspace->inf_strm);
+out:
+	free_workspace(workspace);
+	return ret;
+}
+
+void btrfs_zlib_exit(void)
+{
+    free_workspaces();
+}