btrfs: rework the order of btrfs_ordered_extent::flags
Subsystems (2)
PATCH v2 1/1 btrfs: rework the order of btrfs_ordered_extent::flags
[BUG]
There is a long existing bug in the last parameter of btrfs_add_ordered_extent(), in commit 771ed689d2cd ("Btrfs: Optimize compressed writeback and reads") back to 2008.
In that ancient commit btrfs_add_ordered_extent() expects the @type parameter to be one of the following:
- BTRFS_ORDERED_REGULAR
- BTRFS_ORDERED_NOCOW
- BTRFS_ORDERED_PREALLOC
- BTRFS_ORDERED_COMPRESSED
But we pass 0 in cow_file_range(), which means BTRFS_ORDERED_IO_DONE.
Ironically extra check in __btrfs_add_ordered_extent() won't set the bit if we're seeing (type == IO_DONE || type == IO_COMPLETE), and avoid any obvious bug.
But this still leads to regular COW ordered extent having no bit to indicate its type in various trace events, rendering REGULAR bit useless.
[FIX]
This patch will change the following aspects to avoid such problem:
- Reorder btrfs_ordered_extent::flags
Now the type bits go first (REGULAR/NOCOW/PREALLCO/COMPRESSED), then DIRECT bit, finally extra status bits like IO_DONE/COMPLETE/IOERR.
- Add extra ASSERT() for btrfs_add_ordered_extent_*()
- Remove @type parameter for btrfs_add_ordered_extent_compress() As the only valid @type here is BTRFS_ORDERED_COMPRESSED.
- Remove the unnecessary special check for IO_DONE/COMPLETE in __btrfs_add_ordered_extent()
This is just to make the code work, with extra ASSERT(), there are limited values can be passed in.
---
Changelog:
v2:
- Put ASSERT() on type check to separate line for each condition
- Unify DirectIO/DIO to "direct IO"
- Update comment on ordered extent type bits
--- fs/btrfs/inode.c | 4 ++-- fs/btrfs/ordered-data.c | 21 +++++++++++++++----- fs/btrfs/ordered-data.h | 37 +++++++++++++++++++++++------------- include/trace/events/btrfs.h | 7 ++++--- 4 files changed, 46 insertions(+), 23 deletions(-)
fs/btrfs/inode.c +2 -2
@@ -917,7 +917,6 @@ static noinline void submit_compressed_extents(struct async_chunk *async_chunk) ins.objectid, async_extent->ram_size, ins.offset, -BTRFS_ORDERED_COMPRESSED , async_extent->compress_type);if (ret) { btrfs_drop_extent_cache(inode, async_extent->start,
@@ -1127,7 +1126,8 @@ static noinline int cow_file_range(struct btrfs_inode *inode,free_extent_map (em); ret =btrfs_add_ordered_extent (inode, start, ins.objectid, - ram_size, cur_alloc_size,0 ); + ram_size, cur_alloc_size, +BTRFS_ORDERED_REGULAR );if (ret) goto out_drop_extent_cache;
fs/btrfs/ordered-data.c +16 -5
@@ -199,8 +199,12 @@ static int __btrfs_add_ordered_extent(struct btrfs_inode *inode, u64 file_offset entry->compress_type = compress_type; entry->truncated_len = (u64)-1 ; entry->qgroup_rsv = ret; -if (type !=BTRFS_ORDERED_IO_DONE && type !=BTRFS_ORDERED_COMPLETE ) -set_bit (type, &entry->flags); + +ASSERT (type ==BTRFS_ORDERED_REGULAR || + type ==BTRFS_ORDERED_NOCOW || + type ==BTRFS_ORDERED_PREALLOC || + type ==BTRFS_ORDERED_COMPRESSED ); +set_bit (type, &entry->flags);if (dio) { percpu_counter_add_batch(&fs_info->dio_bytes, num_bytes,
@@ -256,6 +260,9 @@ int btrfs_add_ordered_extent(struct btrfs_inode *inode, u64 file_offset, u64 disk_bytenr, u64 num_bytes, u64 disk_num_bytes, int type) { + ASSERT(type == BTRFS_ORDERED_REGULAR || + type == BTRFS_ORDERED_NOCOW || + type == BTRFS_ORDERED_PREALLOC);return __btrfs_add_ordered_extent (inode, file_offset, disk_bytenr, num_bytes, disk_num_bytes, type,0 ,BTRFS_COMPRESS_NONE );
@@ -265,6 +272,9 @@ int btrfs_add_ordered_extent_dio(struct btrfs_inode *inode, u64 file_offset, u64 disk_bytenr, u64 num_bytes, u64 disk_num_bytes, int type) { + ASSERT(type == BTRFS_ORDERED_REGULAR || + type == BTRFS_ORDERED_NOCOW || + type == BTRFS_ORDERED_PREALLOC);return __btrfs_add_ordered_extent (inode, file_offset, disk_bytenr, num_bytes, disk_num_bytes, type,1 ,BTRFS_COMPRESS_NONE );
@@ -272,11 +282,12 @@ int btrfs_add_ordered_extent_dio(struct btrfs_inode *inode, u64 file_offset,int btrfs_add_ordered_extent_compress (struct btrfs_inode *inode,u64 file_offset,u64 disk_bytenr,u64 num_bytes, -u64 disk_num_bytes,int type, -int compress_type) +u64 disk_num_bytes,int compress_type) { +ASSERT (compress_type !=BTRFS_COMPRESS_NONE );return __btrfs_add_ordered_extent (inode, file_offset, disk_bytenr, - num_bytes, disk_num_bytes, type,0 , + num_bytes, disk_num_bytes, +BTRFS_ORDERED_COMPRESSED ,0 , compress_type); }
fs/btrfs/ordered-data.h +24 -13
@@ -27,7 +27,7 @@ struct btrfs_ordered_sum { };/* -* bits for the flags field: +* Bits for btrfs_ordered_extent::flags. * * BTRFS_ORDERED_IO_DONE is set when all of the blocks are written. * It is used to make sure metadata is inserted into the tree only once
@@ -38,24 +38,36 @@ struct btrfs_ordered_sum {* IO is done and any metadata is inserted into the tree. */ enum { +/* +* Different types for ordered extent, one and only one of the 4 types +* can be set when creating ordered extent. +* +* REGULAR: For regular non-compressed COW write +* NOCOW: For NOCOW write into existing non-hole extent +* PREALLOC: For NOCOW write into preallocated extent +* COMPRESSED: For compressed COW write +*/ +BTRFS_ORDERED_REGULAR , +BTRFS_ORDERED_NOCOW , +BTRFS_ORDERED_PREALLOC , +BTRFS_ORDERED_COMPRESSED , + +/* +* Extra bit for direct IO, can only be set for +* REGULAR/NOCOW/PREALLOC. No direct IO for compressed extent. +*/ +BTRFS_ORDERED_DIRECT , + +/* Extra status bits for ordered extents */ +/* set when all the pages are written */ BTRFS_ORDERED_IO_DONE ,/* set when removed from the tree */ BTRFS_ORDERED_COMPLETE , -/* set when we want to write in place */ -BTRFS_ORDERED_NOCOW , -/* writing a zlib compressed extent */ -BTRFS_ORDERED_COMPRESSED , -/* set when writing to preallocated extent */ -BTRFS_ORDERED_PREALLOC , -/* set when we're doing DIO with this extent */ -BTRFS_ORDERED_DIRECT ,/* We had an io error when writing this out */ BTRFS_ORDERED_IOERR ,/* Set when we have to truncate an extent */ BTRFS_ORDERED_TRUNCATED , -/* Regular IO for COW */ -BTRFS_ORDERED_REGULAR ,/* Used during fsync to track already logged extents */ BTRFS_ORDERED_LOGGED ,/* We have already logged all the csums of the ordered extent */
@@ -167,8 +179,7 @@ int btrfs_add_ordered_extent_dio(struct btrfs_inode *inode, u64 file_offset, u64 disk_num_bytes, int type);int btrfs_add_ordered_extent_compress (struct btrfs_inode *inode,u64 file_offset,u64 disk_bytenr,u64 num_bytes, -u64 disk_num_bytes,int type, -int compress_type); +u64 disk_num_bytes,int compress_type);void btrfs_add_ordered_sum (struct btrfs_ordered_extent *entry,struct btrfs_ordered_sum *sum); struct btrfs_ordered_extent *btrfs_lookup_ordered_extent(struct btrfs_inode *inode,
include/trace/events/btrfs.h +4 -3
@@ -499,12 +499,13 @@ DEFINE_EVENT(#define show_ordered_flags (flags) \ __print_flags(flags, "|", \ - { (1 << BTRFS_ORDERED_IO_DONE), "IO_DONE" }, \ - { (1 << BTRFS_ORDERED_COMPLETE), "COMPLETE" }, \ + { (1 << BTRFS_ORDERED_REGULAR), "REGULAR" }, \ { (1 << BTRFS_ORDERED_NOCOW), "NOCOW" }, \ - { (1 << BTRFS_ORDERED_COMPRESSED), "COMPRESSED" }, \ { (1 << BTRFS_ORDERED_PREALLOC), "PREALLOC" }, \ + { (1 << BTRFS_ORDERED_COMPRESSED), "COMPRESSED" }, \ { (1 << BTRFS_ORDERED_DIRECT), "DIRECT" }, \ + { (1 << BTRFS_ORDERED_IO_DONE), "IO_DONE" }, \ + { (1 << BTRFS_ORDERED_COMPLETE), "COMPLETE" }, \ { (1 << BTRFS_ORDERED_IOERR), "IOERR" }, \ { (1 << BTRFS_ORDERED_TRUNCATED), "TRUNCATED" })
Signature · 1 line
2.30.0
Quoted 12 lines · Qu Wenruo
fs/btrfs/ordered-data.h +1
-1@@ -27,7 +27,7 @@ struct btrfs_ordered_sum { };/* -* bits for the flags field: +* Bits for btrfs_ordered_extent::flags. * * BTRFS_ORDERED_IO_DONE is set when all of the blocks are written. * It is used to make sure metadata is inserted into the tree only once
fs/btrfs/ordered-data.h struct btrfs_ordered_sum { · line 43
Quoted 6 lines
fs/btrfs/ordered-data.h +3
-0@@ -38,24 +38,36 @@ struct btrfs_ordered_sum {* IO is done and any metadata is inserted into the tree. */ enum { +/* +* Different types for ordered extent, one and only one of the 4 types +* can be set when creating ordered extent.
Sorry, I did not merge this followup in time for the 5.12 freeze and now it would be too late to rebase the whole branch just to fix a comment.
The difference is 'types for direct io' and the 'the 4 type' (typo). Please send an incremental patch, the comment would be incorrect and is worth fixing. Thanks.