| Message ID | 20230419161813.2044576-1-amir73il@gmail.com (mailing list archive) |
|---|---|
| State | Deferred, archived |
| Headers | show |
| Series | [5.10] xfs: drop submit side trans alloc for append ioends | expand |
On Wed, Apr 19, 2023 at 07:18:13PM +0300, Amir Goldstein wrote: > From: Brian Foster <bfoster@redhat.com> > > commit 7cd3099f4925d7c15887d1940ebd65acd66100f5 upstream. > > Per-inode ioend completion batching has a log reservation deadlock > vector between preallocated append transactions and transactions > that are acquired at completion time for other purposes (i.e., > unwritten extent conversion or COW fork remaps). For example, if the > ioend completion workqueue task executes on a batch of ioends that > are sorted such that an append ioend sits at the tail, it's possible > for the outstanding append transaction reservation to block > allocation of transactions required to process preceding ioends in > the list. > > Append ioend completion is historically the common path for on-disk > inode size updates. While file extending writes may have completed > sometime earlier, the on-disk inode size is only updated after > successful writeback completion. These transactions are preallocated > serially from writeback context to mitigate concurrency and > associated log reservation pressure across completions processed by > multi-threaded workqueue tasks. > > However, now that delalloc blocks unconditionally map to unwritten > extents at physical block allocation time, size updates via append > ioends are relatively rare. This means that inode size updates most > commonly occur as part of the preexisting completion time > transaction to convert unwritten extents. As a result, there is no > longer a strong need to preallocate size update transactions. > > Remove the preallocation of inode size update transactions to avoid > the ioend completion processing log reservation deadlock. Instead, > continue to send all potential size extending ioends to workqueue > context for completion and allocate the transaction from that > context. This ensures that no outstanding log reservation is owned > by the ioend completion worker task when it begins to process > ioends. > > Signed-off-by: Brian Foster <bfoster@redhat.com> > Reviewed-by: Christoph Hellwig <hch@lst.de> > Reviewed-by: Darrick J. Wong <djwong@kernel.org> > Signed-off-by: Darrick J. Wong <djwong@kernel.org> > Reported-by: Christian Theune <ct@flyingcircus.io> > Link: https://lore.kernel.org/linux-xfs/CAOQ4uxjj2UqA0h4Y31NbmpHksMhVrXfXjLG4Tnz3zq_UR-3gSA@mail.gmail.com/ > Signed-off-by: Amir Goldstein <amir73il@gmail.com> > Acked-by: Darrick J. Wong <djwong@kernel.org> > --- > > Greg, > > One more fix from v5.13 that I missed from my backports. Now queued up, thanks. greg k-h
diff --git a/fs/xfs/xfs_aops.c b/fs/xfs/xfs_aops.c index 953de843d9c3..e341d6531e68 100644 --- a/fs/xfs/xfs_aops.c +++ b/fs/xfs/xfs_aops.c @@ -39,33 +39,6 @@ static inline bool xfs_ioend_is_append(struct iomap_ioend *ioend) XFS_I(ioend->io_inode)->i_d.di_size; } -STATIC int -xfs_setfilesize_trans_alloc( - struct iomap_ioend *ioend) -{ - struct xfs_mount *mp = XFS_I(ioend->io_inode)->i_mount; - struct xfs_trans *tp; - int error; - - error = xfs_trans_alloc(mp, &M_RES(mp)->tr_fsyncts, 0, 0, 0, &tp); - if (error) - return error; - - ioend->io_private = tp; - - /* - * We may pass freeze protection with a transaction. So tell lockdep - * we released it. - */ - __sb_writers_release(ioend->io_inode->i_sb, SB_FREEZE_FS); - /* - * We hand off the transaction to the completion thread now, so - * clear the flag here. - */ - xfs_trans_clear_context(tp); - return 0; -} - /* * Update on-disk file size now that data has been written to disk. */ @@ -191,12 +164,10 @@ xfs_end_ioend( error = xfs_reflink_end_cow(ip, offset, size); else if (ioend->io_type == IOMAP_UNWRITTEN) error = xfs_iomap_write_unwritten(ip, offset, size, false); - else - ASSERT(!xfs_ioend_is_append(ioend) || ioend->io_private); + if (!error && xfs_ioend_is_append(ioend)) + error = xfs_setfilesize(ip, ioend->io_offset, ioend->io_size); done: - if (ioend->io_private) - error = xfs_setfilesize_ioend(ioend, error); iomap_finish_ioends(ioend, error); memalloc_nofs_restore(nofs_flag); } @@ -246,7 +217,7 @@ xfs_end_io( static inline bool xfs_ioend_needs_workqueue(struct iomap_ioend *ioend) { - return ioend->io_private || + return xfs_ioend_is_append(ioend) || ioend->io_type == IOMAP_UNWRITTEN || (ioend->io_flags & IOMAP_F_SHARED); } @@ -259,8 +230,6 @@ xfs_end_bio( struct xfs_inode *ip = XFS_I(ioend->io_inode); unsigned long flags; - ASSERT(xfs_ioend_needs_workqueue(ioend)); - spin_lock_irqsave(&ip->i_ioend_lock, flags); if (list_empty(&ip->i_ioend_list)) WARN_ON_ONCE(!queue_work(ip->i_mount->m_unwritten_workqueue, @@ -510,14 +479,6 @@ xfs_prepare_ioend( ioend->io_offset, ioend->io_size); } - /* Reserve log space if we might write beyond the on-disk inode size. */ - if (!status && - ((ioend->io_flags & IOMAP_F_SHARED) || - ioend->io_type != IOMAP_UNWRITTEN) && - xfs_ioend_is_append(ioend) && - !ioend->io_private) - status = xfs_setfilesize_trans_alloc(ioend); - memalloc_nofs_restore(nofs_flag); if (xfs_ioend_needs_workqueue(ioend))