aboutsummaryrefslogtreecommitdiff
path: root/fs/btrfs/compression.c
diff options
context:
space:
mode:
Diffstat (limited to 'fs/btrfs/compression.c')
-rw-r--r--fs/btrfs/compression.c241
1 files changed, 96 insertions, 145 deletions
diff --git a/fs/btrfs/compression.c b/fs/btrfs/compression.c
index 7f390849343b..d2ef9ac2a630 100644
--- a/fs/btrfs/compression.c
+++ b/fs/btrfs/compression.c
@@ -32,6 +32,7 @@
#include <linux/writeback.h>
#include <linux/bit_spinlock.h>
#include <linux/slab.h>
+#include <linux/sched/mm.h>
#include "ctree.h"
#include "disk-io.h"
#include "transaction.h"
@@ -42,48 +43,7 @@
#include "extent_io.h"
#include "extent_map.h"
-struct compressed_bio {
- /* number of bios pending for this compressed extent */
- atomic_t pending_bios;
-
- /* the pages with the compressed data on them */
- struct page **compressed_pages;
-
- /* inode that owns this data */
- struct inode *inode;
-
- /* starting offset in the inode for our pages */
- u64 start;
-
- /* number of bytes in the inode we're working on */
- unsigned long len;
-
- /* number of bytes on disk */
- unsigned long compressed_len;
-
- /* the compression algorithm for this bio */
- int compress_type;
-
- /* number of compressed pages in the array */
- unsigned long nr_pages;
-
- /* IO errors */
- int errors;
- int mirror_num;
-
- /* for reads, this is the bio we are copying the data into */
- struct bio *orig_bio;
-
- /*
- * the start of a variable length array of checksums only
- * used by reads
- */
- u32 sums;
-};
-
-static int btrfs_decompress_bio(int type, struct page **pages_in,
- u64 disk_start, struct bio *orig_bio,
- size_t srclen);
+static int btrfs_decompress_bio(struct compressed_bio *cb);
static inline int compressed_bio_size(struct btrfs_fs_info *fs_info,
unsigned long disk_size)
@@ -94,13 +54,7 @@ static inline int compressed_bio_size(struct btrfs_fs_info *fs_info,
(DIV_ROUND_UP(disk_size, fs_info->sectorsize)) * csum_size;
}
-static struct bio *compressed_bio_alloc(struct block_device *bdev,
- u64 first_byte, gfp_t gfp_flags)
-{
- return btrfs_bio_alloc(bdev, first_byte >> 9, BIO_MAX_PAGES, gfp_flags);
-}
-
-static int check_compressed_csum(struct inode *inode,
+static int check_compressed_csum(struct btrfs_inode *inode,
struct compressed_bio *cb,
u64 disk_start)
{
@@ -111,7 +65,7 @@ static int check_compressed_csum(struct inode *inode,
u32 csum;
u32 *cb_sum = &cb->sums;
- if (BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM)
+ if (inode->flags & BTRFS_INODE_NODATASUM)
return 0;
for (i = 0; i < cb->nr_pages; i++) {
@@ -124,10 +78,8 @@ static int check_compressed_csum(struct inode *inode,
kunmap_atomic(kaddr);
if (csum != *cb_sum) {
- btrfs_info(BTRFS_I(inode)->root->fs_info,
- "csum failed ino %llu extent %llu csum %u wanted %u mirror %d",
- btrfs_ino(inode), disk_start, csum, *cb_sum,
- cb->mirror_num);
+ btrfs_print_data_csum_error(inode, disk_start, csum,
+ *cb_sum, cb->mirror_num);
ret = -EIO;
goto fail;
}
@@ -157,17 +109,17 @@ static void end_compressed_bio_read(struct bio *bio)
unsigned long index;
int ret;
- if (bio->bi_error)
+ if (bio->bi_status)
cb->errors = 1;
/* if there are more bios still pending for this compressed
* extent, just exit
*/
- if (!atomic_dec_and_test(&cb->pending_bios))
+ if (!refcount_dec_and_test(&cb->pending_bios))
goto out;
inode = cb->inode;
- ret = check_compressed_csum(inode, cb,
+ ret = check_compressed_csum(BTRFS_I(inode), cb,
(u64)bio->bi_iter.bi_sector << 9);
if (ret)
goto csum_failed;
@@ -175,11 +127,8 @@ static void end_compressed_bio_read(struct bio *bio)
/* ok, we're the last bio for this extent, lets start
* the decompression.
*/
- ret = btrfs_decompress_bio(cb->compress_type,
- cb->compressed_pages,
- cb->start,
- cb->orig_bio,
- cb->compressed_len);
+ ret = btrfs_decompress_bio(cb);
+
csum_failed:
if (ret)
cb->errors = 1;
@@ -203,6 +152,7 @@ csum_failed:
* we have verified the checksum already, set page
* checked so the end_io handlers know about it
*/
+ ASSERT(!bio_flagged(bio, BIO_CLONED));
bio_for_each_segment_all(bvec, cb->orig_bio, i)
SetPageChecked(bvec->bv_page);
@@ -270,13 +220,13 @@ static void end_compressed_bio_write(struct bio *bio)
struct page *page;
unsigned long index;
- if (bio->bi_error)
+ if (bio->bi_status)
cb->errors = 1;
/* if there are more bios still pending for this compressed
* extent, just exit
*/
- if (!atomic_dec_and_test(&cb->pending_bios))
+ if (!refcount_dec_and_test(&cb->pending_bios))
goto out;
/* ok, we're the last bio for this extent, step one is to
@@ -289,7 +239,7 @@ static void end_compressed_bio_write(struct bio *bio)
cb->start,
cb->start + cb->len - 1,
NULL,
- bio->bi_error ? 0 : 1);
+ bio->bi_status ? 0 : 1);
cb->compressed_pages[0]->mapping = NULL;
end_compressed_writeback(inode, cb);
@@ -322,7 +272,7 @@ out:
* This also checksums the file bytes and gets things ready for
* the end io hooks.
*/
-int btrfs_submit_compressed_write(struct inode *inode, u64 start,
+blk_status_t btrfs_submit_compressed_write(struct inode *inode, u64 start,
unsigned long len, u64 disk_start,
unsigned long compressed_len,
struct page **compressed_pages,
@@ -337,14 +287,14 @@ int btrfs_submit_compressed_write(struct inode *inode, u64 start,
struct page *page;
u64 first_byte = disk_start;
struct block_device *bdev;
- int ret;
+ blk_status_t ret;
int skip_sum = BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM;
WARN_ON(start & ((u64)PAGE_SIZE - 1));
cb = kmalloc(compressed_bio_size(fs_info, compressed_len), GFP_NOFS);
if (!cb)
- return -ENOMEM;
- atomic_set(&cb->pending_bios, 0);
+ return BLK_STS_RESOURCE;
+ refcount_set(&cb->pending_bios, 0);
cb->errors = 0;
cb->inode = inode;
cb->start = start;
@@ -357,30 +307,26 @@ int btrfs_submit_compressed_write(struct inode *inode, u64 start,
bdev = fs_info->fs_devices->latest_bdev;
- bio = compressed_bio_alloc(bdev, first_byte, GFP_NOFS);
- if (!bio) {
- kfree(cb);
- return -ENOMEM;
- }
+ bio = btrfs_bio_alloc(bdev, first_byte);
bio_set_op_attrs(bio, REQ_OP_WRITE, 0);
bio->bi_private = cb;
bio->bi_end_io = end_compressed_bio_write;
- atomic_inc(&cb->pending_bios);
+ refcount_set(&cb->pending_bios, 1);
/* create and submit bios for the compressed pages */
bytes_left = compressed_len;
for (pg_index = 0; pg_index < cb->nr_pages; pg_index++) {
+ int submit = 0;
+
page = compressed_pages[pg_index];
page->mapping = inode->i_mapping;
if (bio->bi_iter.bi_size)
- ret = io_tree->ops->merge_bio_hook(page, 0,
+ submit = io_tree->ops->merge_bio_hook(page, 0,
PAGE_SIZE,
bio, 0);
- else
- ret = 0;
page->mapping = NULL;
- if (ret || bio_add_page(bio, page, PAGE_SIZE, 0) <
+ if (submit || bio_add_page(bio, page, PAGE_SIZE, 0) <
PAGE_SIZE) {
bio_get(bio);
@@ -390,7 +336,7 @@ int btrfs_submit_compressed_write(struct inode *inode, u64 start,
* we inc the count. Otherwise, the cb might get
* freed before we're done setting it up
*/
- atomic_inc(&cb->pending_bios);
+ refcount_inc(&cb->pending_bios);
ret = btrfs_bio_wq_end_io(fs_info, bio,
BTRFS_WQ_ENDIO_DATA);
BUG_ON(ret); /* -ENOMEM */
@@ -402,14 +348,13 @@ int btrfs_submit_compressed_write(struct inode *inode, u64 start,
ret = btrfs_map_bio(fs_info, bio, 0, 1);
if (ret) {
- bio->bi_error = ret;
+ bio->bi_status = ret;
bio_endio(bio);
}
bio_put(bio);
- bio = compressed_bio_alloc(bdev, first_byte, GFP_NOFS);
- BUG_ON(!bio);
+ bio = btrfs_bio_alloc(bdev, first_byte);
bio_set_op_attrs(bio, REQ_OP_WRITE, 0);
bio->bi_private = cb;
bio->bi_end_io = end_compressed_bio_write;
@@ -436,7 +381,7 @@ int btrfs_submit_compressed_write(struct inode *inode, u64 start,
ret = btrfs_map_bio(fs_info, bio, 0, 1);
if (ret) {
- bio->bi_error = ret;
+ bio->bi_status = ret;
bio_endio(bio);
}
@@ -571,7 +516,7 @@ next:
* After the compressed pages are read, we copy the bytes into the
* bio we were passed and then call the bio end_io calls
*/
-int btrfs_submit_compressed_read(struct inode *inode, struct bio *bio,
+blk_status_t btrfs_submit_compressed_read(struct inode *inode, struct bio *bio,
int mirror_num, unsigned long bio_flags)
{
struct btrfs_fs_info *fs_info = btrfs_sb(inode->i_sb);
@@ -588,7 +533,7 @@ int btrfs_submit_compressed_read(struct inode *inode, struct bio *bio,
u64 em_len;
u64 em_start;
struct extent_map *em;
- int ret = -ENOMEM;
+ blk_status_t ret = BLK_STS_RESOURCE;
int faili = 0;
u32 *sums;
@@ -602,14 +547,14 @@ int btrfs_submit_compressed_read(struct inode *inode, struct bio *bio,
PAGE_SIZE);
read_unlock(&em_tree->lock);
if (!em)
- return -EIO;
+ return BLK_STS_IOERR;
compressed_len = em->block_len;
cb = kmalloc(compressed_bio_size(fs_info, compressed_len), GFP_NOFS);
if (!cb)
goto out;
- atomic_set(&cb->pending_bios, 0);
+ refcount_set(&cb->pending_bios, 0);
cb->errors = 0;
cb->inode = inode;
cb->mirror_num = mirror_num;
@@ -640,7 +585,7 @@ int btrfs_submit_compressed_read(struct inode *inode, struct bio *bio,
__GFP_HIGHMEM);
if (!cb->compressed_pages[pg_index]) {
faili = pg_index - 1;
- ret = -ENOMEM;
+ ret = BLK_STS_RESOURCE;
goto fail2;
}
}
@@ -652,28 +597,26 @@ int btrfs_submit_compressed_read(struct inode *inode, struct bio *bio,
/* include any pages we added in add_ra-bio_pages */
cb->len = bio->bi_iter.bi_size;
- comp_bio = compressed_bio_alloc(bdev, cur_disk_byte, GFP_NOFS);
- if (!comp_bio)
- goto fail2;
+ comp_bio = btrfs_bio_alloc(bdev, cur_disk_byte);
bio_set_op_attrs (comp_bio, REQ_OP_READ, 0);
comp_bio->bi_private = cb;
comp_bio->bi_end_io = end_compressed_bio_read;
- atomic_inc(&cb->pending_bios);
+ refcount_set(&cb->pending_bios, 1);
for (pg_index = 0; pg_index < nr_pages; pg_index++) {
+ int submit = 0;
+
page = cb->compressed_pages[pg_index];
page->mapping = inode->i_mapping;
page->index = em_start >> PAGE_SHIFT;
if (comp_bio->bi_iter.bi_size)
- ret = tree->ops->merge_bio_hook(page, 0,
+ submit = tree->ops->merge_bio_hook(page, 0,
PAGE_SIZE,
comp_bio, 0);
- else
- ret = 0;
page->mapping = NULL;
- if (ret || bio_add_page(comp_bio, page, PAGE_SIZE, 0) <
+ if (submit || bio_add_page(comp_bio, page, PAGE_SIZE, 0) <
PAGE_SIZE) {
bio_get(comp_bio);
@@ -687,7 +630,7 @@ int btrfs_submit_compressed_read(struct inode *inode, struct bio *bio,
* we inc the count. Otherwise, the cb might get
* freed before we're done setting it up
*/
- atomic_inc(&cb->pending_bios);
+ refcount_inc(&cb->pending_bios);
if (!(BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM)) {
ret = btrfs_lookup_bio_sums(inode, comp_bio,
@@ -699,15 +642,13 @@ int btrfs_submit_compressed_read(struct inode *inode, struct bio *bio,
ret = btrfs_map_bio(fs_info, comp_bio, mirror_num, 0);
if (ret) {
- comp_bio->bi_error = ret;
+ comp_bio->bi_status = ret;
bio_endio(comp_bio);
}
bio_put(comp_bio);
- comp_bio = compressed_bio_alloc(bdev, cur_disk_byte,
- GFP_NOFS);
- BUG_ON(!comp_bio);
+ comp_bio = btrfs_bio_alloc(bdev, cur_disk_byte);
bio_set_op_attrs(comp_bio, REQ_OP_READ, 0);
comp_bio->bi_private = cb;
comp_bio->bi_end_io = end_compressed_bio_read;
@@ -728,7 +669,7 @@ int btrfs_submit_compressed_read(struct inode *inode, struct bio *bio,
ret = btrfs_map_bio(fs_info, comp_bio, mirror_num, 0);
if (ret) {
- comp_bio->bi_error = ret;
+ comp_bio->bi_status = ret;
bio_endio(comp_bio);
}
@@ -803,6 +744,7 @@ static struct list_head *find_workspace(int type)
struct list_head *workspace;
int cpus = num_online_cpus();
int idx = type - 1;
+ unsigned nofs_flag;
struct list_head *idle_ws = &btrfs_comp_ws[idx].idle_ws;
spinlock_t *ws_lock = &btrfs_comp_ws[idx].ws_lock;
@@ -832,7 +774,15 @@ again:
atomic_inc(total_ws);
spin_unlock(ws_lock);
+ /*
+ * Allocation helpers call vmalloc that can't use GFP_NOFS, so we have
+ * to turn it off here because we might get called from the restricted
+ * context of btrfs_compress_bio/btrfs_compress_pages
+ */
+ nofs_flag = memalloc_nofs_save();
workspace = btrfs_compress_op[idx]->alloc_workspace();
+ memalloc_nofs_restore(nofs_flag);
+
if (IS_ERR(workspace)) {
atomic_dec(total_ws);
wake_up(ws_wait);
@@ -913,32 +863,28 @@ static void free_workspaces(void)
}
/*
- * given an address space and start/len, compress the bytes.
+ * Given an address space and start and length, compress the bytes into @pages
+ * that are allocated on demand.
*
- * pages are allocated to hold the compressed result and stored
- * in 'pages'
+ * @out_pages is an in/out parameter, holds maximum number of pages to allocate
+ * and returns number of actually allocated 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
+ * @total_in is used to return the number of bytes actually read. It
+ * may be smaller than the input length 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
+ * @total_out is an in/out parameter, must be set to the input length and will
+ * be also used to return the total number of compressed bytes
*
- * max_out tells us the max number of bytes that we're allowed to
+ * @max_out tells us the max number of bytes that we're allowed to
* stuff into pages
*/
int btrfs_compress_pages(int type, struct address_space *mapping,
- u64 start, unsigned long len,
- struct page **pages,
- unsigned long nr_dest_pages,
+ u64 start, struct page **pages,
unsigned long *out_pages,
unsigned long *total_in,
- unsigned long *total_out,
- unsigned long max_out)
+ unsigned long *total_out)
{
struct list_head *workspace;
int ret;
@@ -946,10 +892,9 @@ int btrfs_compress_pages(int type, struct address_space *mapping,
workspace = find_workspace(type);
ret = btrfs_compress_op[type-1]->compress_pages(workspace, mapping,
- start, len, pages,
- nr_dest_pages, out_pages,
- total_in, total_out,
- max_out);
+ start, pages,
+ out_pages,
+ total_in, total_out);
free_workspace(type, workspace);
return ret;
}
@@ -968,19 +913,16 @@ int btrfs_compress_pages(int type, struct address_space *mapping,
* be contiguous. They all correspond to the range of bytes covered by
* the compressed extent.
*/
-static int btrfs_decompress_bio(int type, struct page **pages_in,
- u64 disk_start, struct bio *orig_bio,
- size_t srclen)
+static int btrfs_decompress_bio(struct compressed_bio *cb)
{
struct list_head *workspace;
int ret;
+ int type = cb->compress_type;
workspace = find_workspace(type);
-
- ret = btrfs_compress_op[type-1]->decompress_bio(workspace, pages_in,
- disk_start, orig_bio,
- srclen);
+ ret = btrfs_compress_op[type - 1]->decompress_bio(workspace, cb);
free_workspace(type, workspace);
+
return ret;
}
@@ -1017,13 +959,14 @@ void btrfs_exit_compress(void)
*
* total_out is the last byte of the buffer
*/
-int btrfs_decompress_buf2page(char *buf, unsigned long buf_start,
+int btrfs_decompress_buf2page(const char *buf, unsigned long buf_start,
unsigned long total_out, u64 disk_start,
struct bio *bio)
{
unsigned long buf_offset;
unsigned long current_buf_start;
unsigned long start_byte;
+ unsigned long prev_start_byte;
unsigned long working_bytes = total_out - buf_start;
unsigned long bytes;
char *kaddr;
@@ -1071,26 +1014,34 @@ int btrfs_decompress_buf2page(char *buf, unsigned long buf_start,
if (!bio->bi_iter.bi_size)
return 0;
bvec = bio_iter_iovec(bio, bio->bi_iter);
-
+ prev_start_byte = start_byte;
start_byte = page_offset(bvec.bv_page) - disk_start;
/*
- * make sure our new page is covered by this
- * working buffer
+ * We need to make sure we're only adjusting
+ * our offset into compression working buffer when
+ * we're switching pages. Otherwise we can incorrectly
+ * keep copying when we were actually done.
*/
- if (total_out <= start_byte)
- return 1;
+ if (start_byte != prev_start_byte) {
+ /*
+ * make sure our new page is covered by this
+ * working buffer
+ */
+ if (total_out <= start_byte)
+ return 1;
- /*
- * 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;
+ /*
+ * 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;
+ }
}
}
Powered by cgit, Gruvboxed by Blaster4385.