From dca94251f617942f05c7c6ff30a299f6b7dff770 Mon Sep 17 00:00:00 2001 From: "Darrick J. Wong" Date: Sun, 3 Nov 2024 20:19:16 -0800 Subject: [PATCH 1/2] xfs: fix rt device offset calculations for FITRIM FITRIM on xfs has this bizarro uapi where we flatten all the physically addressable storage across two block devices into a linear address space. In this address space, the realtime device comes immediately after the data device. Therefore, the xfs_trim_rtdev_extents has to convert its input parameters from the linear address space to actual rtdev block addresses on the realtime volume. Right now the address space conversion is done in units of rtblocks. However, a future patchset will convert xfs_rtblock_t to be a segmented address space (group:blkno) like the data device. Change the conversion code to be done in units of daddrs since those will never be segmented. Signed-off-by: Darrick J. Wong Reviewed-by: Christoph Hellwig --- fs/xfs/xfs_discard.c | 19 +++++++++++-------- 1 file changed, 11 insertions(+), 8 deletions(-) diff --git a/fs/xfs/xfs_discard.c b/fs/xfs/xfs_discard.c index 4f3e4736f13e..42b8b5e0e931 100644 --- a/fs/xfs/xfs_discard.c +++ b/fs/xfs/xfs_discard.c @@ -480,7 +480,7 @@ xfs_discard_rtdev_extents( trace_xfs_discard_rtextent(mp, busyp->bno, busyp->length); error = __blkdev_issue_discard(bdev, - XFS_FSB_TO_BB(mp, busyp->bno), + xfs_rtb_to_daddr(mp, busyp->bno), XFS_FSB_TO_BB(mp, busyp->length), GFP_NOFS, &bio); if (error) @@ -612,22 +612,25 @@ xfs_trim_rtdev_extents( xfs_rtblock_t start_rtbno, end_rtbno; xfs_rtxnum_t start_rtx, end_rtx; xfs_rgnumber_t start_rgno, end_rgno; + xfs_daddr_t daddr_offset; int last_error = 0, error; struct xfs_rtgroup *rtg = NULL; /* Shift the start and end downwards to match the rt device. */ - start_rtbno = xfs_daddr_to_rtb(mp, start); - if (start_rtbno > mp->m_sb.sb_dblocks) - start_rtbno -= mp->m_sb.sb_dblocks; + daddr_offset = XFS_FSB_TO_BB(mp, mp->m_sb.sb_dblocks); + if (start > daddr_offset) + start -= daddr_offset; else - start_rtbno = 0; + start = 0; + start_rtbno = xfs_daddr_to_rtb(mp, start); start_rtx = xfs_rtb_to_rtx(mp, start_rtbno); start_rgno = xfs_rtb_to_rgno(mp, start_rtbno); - end_rtbno = xfs_daddr_to_rtb(mp, end); - if (end_rtbno <= mp->m_sb.sb_dblocks) + if (end <= daddr_offset) return 0; - end_rtbno -= mp->m_sb.sb_dblocks; + else + end -= daddr_offset; + end_rtbno = xfs_daddr_to_rtb(mp, end); end_rtx = xfs_rtb_to_rtx(mp, end_rtbno + mp->m_sb.sb_rextsize - 1); end_rgno = xfs_rtb_to_rgno(mp, end_rtbno); From 64c58d7c99343a910edf995e15d8037e19ec5777 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Sun, 3 Nov 2024 20:19:16 -0800 Subject: [PATCH 2/2] iomap: add a merge boundary flag File systems might have boundaries over which merges aren't possible. In fact these are very common, although most of the time some kind of header at the beginning of this region (e.g. XFS alloation groups, ext4 block groups) automatically create a merge barrier. But if that is not present, say for a device purely used for data we need to manually communicate that to iomap. Add a IOMAP_F_BOUNDARY flag to never merge I/O into a previous mapping. Signed-off-by: Christoph Hellwig Reviewed-by: Darrick J. Wong Signed-off-by: Darrick J. Wong --- fs/iomap/buffered-io.c | 6 ++++++ include/linux/iomap.h | 4 ++++ 2 files changed, 10 insertions(+) diff --git a/fs/iomap/buffered-io.c b/fs/iomap/buffered-io.c index ef0b68bccbb6..fcadd31017d1 100644 --- a/fs/iomap/buffered-io.c +++ b/fs/iomap/buffered-io.c @@ -1601,6 +1601,8 @@ iomap_ioend_can_merge(struct iomap_ioend *ioend, struct iomap_ioend *next) { if (ioend->io_bio.bi_status != next->io_bio.bi_status) return false; + if (next->io_flags & IOMAP_F_BOUNDARY) + return false; if ((ioend->io_flags & IOMAP_F_SHARED) ^ (next->io_flags & IOMAP_F_SHARED)) return false; @@ -1720,6 +1722,8 @@ static struct iomap_ioend *iomap_alloc_ioend(struct iomap_writepage_ctx *wpc, INIT_LIST_HEAD(&ioend->io_list); ioend->io_type = wpc->iomap.type; ioend->io_flags = wpc->iomap.flags; + if (pos > wpc->iomap.offset) + wpc->iomap.flags &= ~IOMAP_F_BOUNDARY; ioend->io_inode = inode; ioend->io_size = 0; ioend->io_offset = pos; @@ -1731,6 +1735,8 @@ static struct iomap_ioend *iomap_alloc_ioend(struct iomap_writepage_ctx *wpc, static bool iomap_can_add_to_ioend(struct iomap_writepage_ctx *wpc, loff_t pos) { + if (wpc->iomap.offset == pos && (wpc->iomap.flags & IOMAP_F_BOUNDARY)) + return false; if ((wpc->iomap.flags & IOMAP_F_SHARED) != (wpc->ioend->io_flags & IOMAP_F_SHARED)) return false; diff --git a/include/linux/iomap.h b/include/linux/iomap.h index f61407e3b121..9ecb8ea7714c 100644 --- a/include/linux/iomap.h +++ b/include/linux/iomap.h @@ -53,6 +53,9 @@ struct vm_fault; * * IOMAP_F_XATTR indicates that the iomap is for an extended attribute extent * rather than a file data extent. + * + * IOMAP_F_BOUNDARY indicates that I/O and I/O completions for this iomap must + * never be merged with the mapping before it. */ #define IOMAP_F_NEW (1U << 0) #define IOMAP_F_DIRTY (1U << 1) @@ -64,6 +67,7 @@ struct vm_fault; #define IOMAP_F_BUFFER_HEAD 0 #endif /* CONFIG_BUFFER_HEAD */ #define IOMAP_F_XATTR (1U << 5) +#define IOMAP_F_BOUNDARY (1U << 6) /* * Flags set by the core iomap code during operations: