Commit 26b91c72 authored by Christoph Hellwig's avatar Christoph Hellwig Committed by Darrick J. Wong

xfs: make COW fork unwritten extent conversions more robust

If we have racing buffered and direct I/O COW fork extents under
writeback can have been moved to the data fork by the time we call
xfs_reflink_convert_cow from xfs_submit_ioend.  This would be mostly
harmless as the block numbers don't change by this move, except for
the fact that xfs_bmapi_write will crash or trigger asserts when
not finding existing extents, even despite trying to paper over this
with the XFS_BMAPI_CONVERT_ONLY flag.

Instead of special casing non-transaction conversions in the already
way too complicated xfs_bmapi_write just add a new helper for the much
simpler non-transactional COW fork case, which simplify ignores not
found extents.
Signed-off-by: default avatarChristoph Hellwig <hch@lst.de>
Reviewed-by: default avatarDarrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: default avatarDarrick J. Wong <darrick.wong@oracle.com>
parent db46e604
...@@ -2031,7 +2031,7 @@ xfs_bmap_add_extent_delay_real( ...@@ -2031,7 +2031,7 @@ xfs_bmap_add_extent_delay_real(
/* /*
* Convert an unwritten allocation to a real allocation or vice versa. * Convert an unwritten allocation to a real allocation or vice versa.
*/ */
STATIC int /* error */ int /* error */
xfs_bmap_add_extent_unwritten_real( xfs_bmap_add_extent_unwritten_real(
struct xfs_trans *tp, struct xfs_trans *tp,
xfs_inode_t *ip, /* incore inode pointer */ xfs_inode_t *ip, /* incore inode pointer */
...@@ -4276,9 +4276,7 @@ xfs_bmapi_write( ...@@ -4276,9 +4276,7 @@ xfs_bmapi_write(
ASSERT(*nmap >= 1); ASSERT(*nmap >= 1);
ASSERT(*nmap <= XFS_BMAP_MAX_NMAP); ASSERT(*nmap <= XFS_BMAP_MAX_NMAP);
ASSERT(tp != NULL || ASSERT(tp != NULL);
(flags & (XFS_BMAPI_CONVERT | XFS_BMAPI_COWFORK)) ==
(XFS_BMAPI_CONVERT | XFS_BMAPI_COWFORK));
ASSERT(len > 0); ASSERT(len > 0);
ASSERT(XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_LOCAL); ASSERT(XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_LOCAL);
ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL)); ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL));
...@@ -4352,8 +4350,7 @@ xfs_bmapi_write( ...@@ -4352,8 +4350,7 @@ xfs_bmapi_write(
* First, deal with the hole before the allocated space * First, deal with the hole before the allocated space
* that we found, if any. * that we found, if any.
*/ */
if ((need_alloc || wasdelay) && if (need_alloc || wasdelay) {
!(flags & XFS_BMAPI_CONVERT_ONLY)) {
bma.eof = eof; bma.eof = eof;
bma.conv = !!(flags & XFS_BMAPI_CONVERT); bma.conv = !!(flags & XFS_BMAPI_CONVERT);
bma.wasdel = wasdelay; bma.wasdel = wasdelay;
......
...@@ -95,9 +95,6 @@ struct xfs_extent_free_item ...@@ -95,9 +95,6 @@ struct xfs_extent_free_item
/* Map something in the CoW fork. */ /* Map something in the CoW fork. */
#define XFS_BMAPI_COWFORK 0x200 #define XFS_BMAPI_COWFORK 0x200
/* Only convert unwritten extents, don't allocate new blocks */
#define XFS_BMAPI_CONVERT_ONLY 0x800
/* Skip online discard of freed extents */ /* Skip online discard of freed extents */
#define XFS_BMAPI_NODISCARD 0x1000 #define XFS_BMAPI_NODISCARD 0x1000
...@@ -114,7 +111,6 @@ struct xfs_extent_free_item ...@@ -114,7 +111,6 @@ struct xfs_extent_free_item
{ XFS_BMAPI_ZERO, "ZERO" }, \ { XFS_BMAPI_ZERO, "ZERO" }, \
{ XFS_BMAPI_REMAP, "REMAP" }, \ { XFS_BMAPI_REMAP, "REMAP" }, \
{ XFS_BMAPI_COWFORK, "COWFORK" }, \ { XFS_BMAPI_COWFORK, "COWFORK" }, \
{ XFS_BMAPI_CONVERT_ONLY, "CONVERT_ONLY" }, \
{ XFS_BMAPI_NODISCARD, "NODISCARD" }, \ { XFS_BMAPI_NODISCARD, "NODISCARD" }, \
{ XFS_BMAPI_NORMAP, "NORMAP" } { XFS_BMAPI_NORMAP, "NORMAP" }
...@@ -226,6 +222,10 @@ int xfs_bmapi_reserve_delalloc(struct xfs_inode *ip, int whichfork, ...@@ -226,6 +222,10 @@ int xfs_bmapi_reserve_delalloc(struct xfs_inode *ip, int whichfork,
int xfs_bmapi_convert_delalloc(struct xfs_inode *ip, int whichfork, int xfs_bmapi_convert_delalloc(struct xfs_inode *ip, int whichfork,
xfs_fileoff_t offset_fsb, struct xfs_bmbt_irec *imap, xfs_fileoff_t offset_fsb, struct xfs_bmbt_irec *imap,
unsigned int *seq); unsigned int *seq);
int xfs_bmap_add_extent_unwritten_real(struct xfs_trans *tp,
struct xfs_inode *ip, int whichfork,
struct xfs_iext_cursor *icur, struct xfs_btree_cur **curp,
struct xfs_bmbt_irec *new, int *logflagsp);
static inline void static inline void
xfs_bmap_add_free( xfs_bmap_add_free(
......
...@@ -234,26 +234,42 @@ xfs_reflink_trim_around_shared( ...@@ -234,26 +234,42 @@ xfs_reflink_trim_around_shared(
} }
} }
/* Convert part of an unwritten CoW extent to a real one. */ static int
STATIC int xfs_reflink_convert_cow_locked(
xfs_reflink_convert_cow_extent(
struct xfs_inode *ip, struct xfs_inode *ip,
struct xfs_bmbt_irec *imap,
xfs_fileoff_t offset_fsb, xfs_fileoff_t offset_fsb,
xfs_filblks_t count_fsb) xfs_filblks_t count_fsb)
{ {
int nimaps = 1; struct xfs_iext_cursor icur;
struct xfs_bmbt_irec got;
struct xfs_btree_cur *dummy_cur = NULL;
int dummy_logflags;
int error;
if (imap->br_state == XFS_EXT_NORM) if (!xfs_iext_lookup_extent(ip, ip->i_cowfp, offset_fsb, &icur, &got))
return 0; return 0;
xfs_trim_extent(imap, offset_fsb, count_fsb); do {
trace_xfs_reflink_convert_cow(ip, imap); if (got.br_startoff >= offset_fsb + count_fsb)
if (imap->br_blockcount == 0) break;
return 0; if (got.br_state == XFS_EXT_NORM)
return xfs_bmapi_write(NULL, ip, imap->br_startoff, imap->br_blockcount, continue;
XFS_BMAPI_COWFORK | XFS_BMAPI_CONVERT, 0, imap, if (WARN_ON_ONCE(isnullstartblock(got.br_startblock)))
&nimaps); return -EIO;
xfs_trim_extent(&got, offset_fsb, count_fsb);
if (!got.br_blockcount)
continue;
got.br_state = XFS_EXT_NORM;
error = xfs_bmap_add_extent_unwritten_real(NULL, ip,
XFS_COW_FORK, &icur, &dummy_cur, &got,
&dummy_logflags);
if (error)
return error;
} while (xfs_iext_next_extent(ip->i_cowfp, &icur, &got));
return error;
} }
/* Convert all of the unwritten CoW extents in a file's range to real ones. */ /* Convert all of the unwritten CoW extents in a file's range to real ones. */
...@@ -267,15 +283,12 @@ xfs_reflink_convert_cow( ...@@ -267,15 +283,12 @@ xfs_reflink_convert_cow(
xfs_fileoff_t offset_fsb = XFS_B_TO_FSBT(mp, offset); xfs_fileoff_t offset_fsb = XFS_B_TO_FSBT(mp, offset);
xfs_fileoff_t end_fsb = XFS_B_TO_FSB(mp, offset + count); xfs_fileoff_t end_fsb = XFS_B_TO_FSB(mp, offset + count);
xfs_filblks_t count_fsb = end_fsb - offset_fsb; xfs_filblks_t count_fsb = end_fsb - offset_fsb;
struct xfs_bmbt_irec imap; int error;
int nimaps = 1, error = 0;
ASSERT(count != 0); ASSERT(count != 0);
xfs_ilock(ip, XFS_ILOCK_EXCL); xfs_ilock(ip, XFS_ILOCK_EXCL);
error = xfs_bmapi_write(NULL, ip, offset_fsb, count_fsb, error = xfs_reflink_convert_cow_locked(ip, offset_fsb, count_fsb);
XFS_BMAPI_COWFORK | XFS_BMAPI_CONVERT |
XFS_BMAPI_CONVERT_ONLY, 0, &imap, &nimaps);
xfs_iunlock(ip, XFS_ILOCK_EXCL); xfs_iunlock(ip, XFS_ILOCK_EXCL);
return error; return error;
} }
...@@ -405,14 +418,16 @@ xfs_reflink_allocate_cow( ...@@ -405,14 +418,16 @@ xfs_reflink_allocate_cow(
if (nimaps == 0) if (nimaps == 0)
return -ENOSPC; return -ENOSPC;
convert: convert:
xfs_trim_extent(imap, offset_fsb, count_fsb);
/* /*
* COW fork extents are supposed to remain unwritten until we're ready * COW fork extents are supposed to remain unwritten until we're ready
* to initiate a disk write. For direct I/O we are going to write the * to initiate a disk write. For direct I/O we are going to write the
* data and need the conversion, but for buffered writes we're done. * data and need the conversion, but for buffered writes we're done.
*/ */
if (!(iomap_flags & IOMAP_DIRECT)) if (!(iomap_flags & IOMAP_DIRECT) || imap->br_state == XFS_EXT_NORM)
return 0; return 0;
return xfs_reflink_convert_cow_extent(ip, imap, offset_fsb, count_fsb); trace_xfs_reflink_convert_cow(ip, imap);
return xfs_reflink_convert_cow_locked(ip, offset_fsb, count_fsb);
out_unreserve: out_unreserve:
xfs_trans_unreserve_quota_nblks(tp, ip, (long)resblks, 0, xfs_trans_unreserve_quota_nblks(tp, ip, (long)resblks, 0,
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment