similar to: [PATCH] Btrfs: use bytes_may_use for all ENOSPC reservations

We have been using bytes_reserved for metadata reservations, which is wrong since we use that to keep track of outstanding reservations from the allocator. This resulted in us doing a lot of silly things to make sure we don''t allocate a bunch of metadata chunks since we never had a real view of how much space was actually in use by metadata. There are a lot of fixes in here to make this

Reserve metadata space for extent tree, checksum tree and root tree Signed-off-by: Yan Zheng <zheng.yan@oracle.com> --- diff -urp 8/fs/btrfs/ctree.h 9/fs/btrfs/ctree.h --- 8/fs/btrfs/ctree.h 2010-04-18 10:26:38.327697818 +0800 +++ 9/fs/btrfs/ctree.h 2010-04-18 10:30:01.883697869 +0800 @@ -682,21 +682,15 @@ struct btrfs_space_info { u64 bytes_reserved; /* total bytes the allocator has

This patch finally gives us full proper -ENOSPC handling for btrfs. Now whenever you do a btrfs_start_transaction, you must specify the number of items you are planning to add/delete/modify. The worst case number of blocks that could be modified by changing that number of items will be calculated and checked against the amount of free space in the space_info where the root you are modifying

I''ve been working on running down intermittent ENOSPC issues. I can only seem to replicate ENOSPC errors when running zlib compression. However, I have been seeing similar ENOSPC errors to a lesser extent when playing with the LZ4HC patches. I apologize for not following up on this sooner, but I had drifted away from using zlib, and didn''t notice there was still an issue. My

Hello, So there is an odd case where we can possibly return -ENOSPC when there is in fact space to be had. I think I finally have a hold on what the problem is, it only happens with Metadata writes, and happens _very_ infrequently. What has to happen is we have to allocate have allocated out of the first logical byte on the disk, which would set last_alloc to first_logical_byte(root, 0), so

I suspect this might be related to previous btrfs errors I''ve had on the same filesystem. See: http://python.ca/nas/linux/btrfs_bug.txt The most recent kernel message is: WARNING: at fs/btrfs/extent-tree.c:8239 btrfs_free_block_groups+0x218/0x275() Hardware name: MS-7388 Modules linked in: udf crc_itu_t isofs loop nls_iso8859_1 vboxnetflt vboxdrv nls_utf8 nls_cp437

Document those places in the btrfs code which are BUGing on non-fatal error conditions that should be handled by proper error paths. This makes it easier to distinguish between what needs fixing versus which BUG_ON''s we might want to keep (to trap code bugs, unexpected inconsistencies, etc). Do this with a trivial macro, ''btrfs_fixable_bug_on'' which just defines to

The whole loop=0,1,2 thing was kind of odd and not very self explanatory. I''ve replaced it with a list_for_each_entry on space_info->block_groups. If we have a hint we just jump into the loop with the block group and start looking for space. If we don''t find anything we start at the beginning and start looking. We never come out of the loop with a ref on the block_group

writeback_inodes_sb(_nr)_if_idle() is re-implemented by replacing down_read() with down_read_trylock() because - If ->s_umount is write locked, then the sb is not idle. That is writeback_inodes_sb(_nr)_if_idle() needn''t wait for the lock. - writeback_inodes_sb(_nr)_if_idle() grabs s_umount lock when it want to start writeback, it may bring us deadlock problem when doing umount.

I fixed a problem where we weren''t reserving space for an orphan item when we had to fallback to using the global reserve for an unlink, but I introduced another problem. I was migrating the bytes from the transaction reserve to the global reserve and then releasing from the global reserve in btrfs_end_transaction(). The problem with this is that a migrate will jack up the size for the

These are all the cases where a variable is set, but not read which are really bugs. - Couple of incorrect error handling fixed. - One incorrect use of a allocation policy - Some other things Still needs more review. Found by gcc 4.6''s new warnings Cc: chris.mason@oracle.com cc: linux-btrfs@vger.kernel.org Signed-off-by: Andi Kleen <ak@linux.intel.com> ---

This reverts commit 5500cdbe14d7435e04f66ff3cfb8ecd8b8e44ebf. We had numerous reports of premature ENOSPC that were bisected to this patch. Reverting will not break things but a warning in ''use_block_rsv'' may show up in the syslog. There''s no alternative fix in sight and the ENOSPC problem affects all 3.3 btrfs users during normal filesystem use. CC:

An user reported that he has hit an annoying deadlock while playing with ceph based on btrfs. Current updating device tree requires space from METADATA chunk, so we -may- need to do a recursive chunk allocation when adding/updating dev extent, that is where the deadlock comes from. If we use SYSTEM metadata to update device tree, we can avoid the recursive stuff. Reported-by: Jim Schutt

In trying to track down where we were leaking reserved space I noticed our reserve extent tracepoints are a little off. First we were saying that the reserved space had been alloced in btrfs_reserve_extent, which isn''t the case, this needs to be triggered when we actually allocate the space when we run the delayed ref. We were also missing a few places where we should have been tracing

Josef''s fs_mark test fs_mark -d /mnt/btrfs-test -D 512 -t 16 -n 4096 -F -S0 on a 2GB single metadata fs leaves about 400Mb of metadata almost unused. This patch reduces metadata chunk allocations by considering the proper metadata chunk size of 200MB in should_alloc_chunk(), not the default 256MB which is set in __btrfs_alloc_chunk(). Signed-off-by: Itaru Kitayama

The biggest one is a fix for fstrim, and there''s a fix for on-disk free space cache. Others are small fixes and cleanups. The last three have been sent weeks ago. The patchset is also available in this repo: git://repo.or.cz/linux-btrfs-devel.git for-chris Note there''s a small confict with Al Viro''s vfs changes. Li Zefan (11): Btrfs: add pinned extents to

Alex pointed out the consequences after a transaction is not committed when a subvolume is deleted, so in case of a crash before an actual commit happens will let the subvolume reappear. Original post: http://www.spinics.net/lists/linux-btrfs/msg22088.html Josef''s objections: http://www.spinics.net/lists/linux-btrfs/msg22256.html While there''s no need to do a full commit for

If we flush inodes with pending delalloc in a transaction, we may join the same transaction handler more than 2 times. The reason is: Task use_count of trans handle commit_transaction 1 |-> btrfs_start_delalloc_inodes 1 |-> run_delalloc_nocow 1 |-> join_transaction 2 |-> cow_file_range 2 |-> join_transaction 3 In fact, cow_file_range needn''t

the tree log recover code expects no free space cached before it executes. Signed-off-by: Yan Zheng <zheng.yan@oracle.com> --- diff -urp 4/fs/btrfs/extent-tree.c 8/fs/btrfs/extent-tree.c --- 4/fs/btrfs/extent-tree.c 2010-05-11 14:15:29.174108554 +0800 +++ 8/fs/btrfs/extent-tree.c 2010-05-11 13:26:38.036107000 +0800 @@ -316,11 +329,6 @@ static int caching_kthread(void *data) if (!path)

The reason the deadlock is that: Task Btrfs-cleaner umount() down_write(&s->s_umount) sync_filesystem() do auto-defragment and produce lots of dirty pages close_ctree() wait for the end of btrfs-cleaner start_transaction reserve space shrink_delalloc() writeback_inodes_sb_nr_if_idle()