| Message ID | 20170411111011.9437-6-hch@lst.de (mailing list archive) |
|---|---|
| State | Superseded |
| Headers | show |
On Tue, Apr 11, 2017 at 01:10:10PM +0200, Christoph Hellwig wrote: > The main thing that xfs_bmap_remap_alloc does is fixing the AGFL, similar > to what we do in the space allocator. But the reflink code doesn't touch > the allocation btree unlike the normal space allocator, so we couldn't > care less about the state of the AGFL. > > So remove xfs_bmap_remap_alloc and just handle the di_nblocks update in > the caller. Looks ok, will go test. By the way, what release were you targeting with this patchset? AFAICT the only behavioral change is that we no longer ensure the AGFL in the remap step prior to (if necessary) ensuring the AGFL again in the subsequent rmap step. Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> --D > > Signed-off-by: Christoph Hellwig <hch@lst.de> > --- > fs/xfs/libxfs/xfs_bmap.c | 60 ++---------------------------------------------- > fs/xfs/xfs_trace.h | 25 -------------------- > 2 files changed, 2 insertions(+), 83 deletions(-) > > diff --git a/fs/xfs/libxfs/xfs_bmap.c b/fs/xfs/libxfs/xfs_bmap.c > index faf7cb0a28f9..414cfa91af01 100644 > --- a/fs/xfs/libxfs/xfs_bmap.c > +++ b/fs/xfs/libxfs/xfs_bmap.c > @@ -3856,61 +3856,6 @@ xfs_bmap_btalloc( > } > > /* > - * For a remap operation, just "allocate" an extent at the address that the > - * caller passed in, and ensure that the AGFL is the right size. The caller > - * will then map the "allocated" extent into the file somewhere. > - */ > -STATIC int > -xfs_bmap_remap_alloc( > - struct xfs_trans *tp, > - struct xfs_inode *ip, > - xfs_fsblock_t startblock, > - xfs_extlen_t length) > -{ > - struct xfs_mount *mp = tp->t_mountp; > - struct xfs_alloc_arg args; > - int error; > - > - /* > - * validate that the block number is legal - the enables us to detect > - * and handle a silent filesystem corruption rather than crashing. > - */ > - memset(&args, 0, sizeof(struct xfs_alloc_arg)); > - args.tp = tp; > - args.mp = mp; > - args.agno = XFS_FSB_TO_AGNO(mp, startblock); > - args.agbno = XFS_FSB_TO_AGBNO(mp, startblock); > - > - if (args.agno >= mp->m_sb.sb_agcount || > - args.agbno >= mp->m_sb.sb_agblocks) > - return -EFSCORRUPTED; > - > - /* "Allocate" the extent from the range we passed in. */ > - trace_xfs_bmap_remap_alloc(ip, startblock, length); > - > - ip->i_d.di_nblocks += length; > - xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE); > - > - /* Fix the freelist, like a real allocator does. */ > - args.datatype = XFS_ALLOC_USERDATA | XFS_ALLOC_NOBUSY; > - args.pag = xfs_perag_get(mp, args.agno); > - ASSERT(args.pag); > - > - /* > - * The freelist fixing code will decline the allocation if > - * the size and shape of the free space doesn't allow for > - * allocating the extent and updating all the metadata that > - * happens during an allocation. We're remapping, not > - * allocating, so skip that check by pretending to be freeing. > - */ > - error = xfs_alloc_fix_freelist(&args, XFS_ALLOC_FLAG_FREEING); > - xfs_perag_put(args.pag); > - if (error) > - trace_xfs_bmap_remap_alloc_error(ip, error, _RET_IP_); > - return error; > -} > - > -/* > * xfs_bmap_alloc is called by xfs_bmapi to allocate an extent for a file. > * It figures out where to ask the underlying allocator to put the new extent. > */ > @@ -4806,9 +4751,8 @@ xfs_bmapi_remap( > ASSERT(got.br_startoff - bno >= len); > } > > - error = xfs_bmap_remap_alloc(tp, ip, startblock, len); > - if (error) > - goto error0; > + ip->i_d.di_nblocks += len; > + xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE); > > if (ifp->if_flags & XFS_IFBROOT) { > cur = xfs_bmbt_init_cursor(mp, tp, ip, XFS_DATA_FORK); > diff --git a/fs/xfs/xfs_trace.h b/fs/xfs/xfs_trace.h > index 4f96dc953fbe..cba10daf8391 100644 > --- a/fs/xfs/xfs_trace.h > +++ b/fs/xfs/xfs_trace.h > @@ -3003,31 +3003,6 @@ DEFINE_EVENT(xfs_inode_error_class, name, \ > unsigned long caller_ip), \ > TP_ARGS(ip, error, caller_ip)) > > -/* reflink allocator */ > -TRACE_EVENT(xfs_bmap_remap_alloc, > - TP_PROTO(struct xfs_inode *ip, xfs_fsblock_t fsbno, > - xfs_extlen_t len), > - TP_ARGS(ip, fsbno, len), > - TP_STRUCT__entry( > - __field(dev_t, dev) > - __field(xfs_ino_t, ino) > - __field(xfs_fsblock_t, fsbno) > - __field(xfs_extlen_t, len) > - ), > - TP_fast_assign( > - __entry->dev = VFS_I(ip)->i_sb->s_dev; > - __entry->ino = ip->i_ino; > - __entry->fsbno = fsbno; > - __entry->len = len; > - ), > - TP_printk("dev %d:%d ino 0x%llx fsbno 0x%llx len %x", > - MAJOR(__entry->dev), MINOR(__entry->dev), > - __entry->ino, > - __entry->fsbno, > - __entry->len) > -); > -DEFINE_INODE_ERROR_EVENT(xfs_bmap_remap_alloc_error); > - > /* reflink tracepoint classes */ > > /* two-file io tracepoint class */ > -- > 2.11.0 > > -- > To unsubscribe from this list: send the line "unsubscribe linux-xfs" in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html -- To unsubscribe from this list: send the line "unsubscribe linux-xfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Tue, Apr 11, 2017 at 04:02:46PM -0700, Darrick J. Wong wrote: > On Tue, Apr 11, 2017 at 01:10:10PM +0200, Christoph Hellwig wrote: > > The main thing that xfs_bmap_remap_alloc does is fixing the AGFL, similar > > to what we do in the space allocator. But the reflink code doesn't touch > > the allocation btree unlike the normal space allocator, so we couldn't > > care less about the state of the AGFL. > > > > So remove xfs_bmap_remap_alloc and just handle the di_nblocks update in > > the caller. > > Looks ok, will go test. By the way, what release were you targeting > with this patchset? The patches are against for-next. Given how late we are in 4.11 I didn't dare to send them for 4.11, although I'd like to backport it to 4.11-stable and 4.9-stable eventually. > AFAICT the only behavioral change is that we no > longer ensure the AGFL in the remap step prior to (if necessary) > ensuring the AGFL again in the subsequent rmap step. Yes. -- To unsubscribe from this list: send the line "unsubscribe linux-xfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Wed, Apr 12, 2017 at 07:38:39AM +0200, Christoph Hellwig wrote: > On Tue, Apr 11, 2017 at 04:02:46PM -0700, Darrick J. Wong wrote: > > On Tue, Apr 11, 2017 at 01:10:10PM +0200, Christoph Hellwig wrote: > > > The main thing that xfs_bmap_remap_alloc does is fixing the AGFL, similar > > > to what we do in the space allocator. But the reflink code doesn't touch > > > the allocation btree unlike the normal space allocator, so we couldn't > > > care less about the state of the AGFL. > > > > > > So remove xfs_bmap_remap_alloc and just handle the di_nblocks update in > > > the caller. > > > > Looks ok, will go test. By the way, what release were you targeting > > with this patchset? > > The patches are against for-next. Given how late we are in 4.11 I didn't > dare to send them for 4.11, although I'd like to backport it to 4.11-stable > and 4.9-stable eventually. > > > AFAICT the only behavioral change is that we no > > longer ensure the AGFL in the remap step prior to (if necessary) > > ensuring the AGFL again in the subsequent rmap step. > > Yes. Hmmm... I got a crash in generic/187 with: MKFS_OPTIONS="-f -m reflink=1,rmapbt=1, -i sparse=1, -b size=1024, /dev/sdf" generic/187 - output mismatch (see /tmp/xfstests/results//generic/187.out.bad) --- tests/generic/187.out 2017-02-28 09:23:56.058066197 -0800 +++ /tmp/xfstests/results//generic/187.out.bad 2017-04-11 18:14:40.715071575 -0700 @@ -7,8 +7,11 @@ 8d9ea4925db533da10a45d2919d8d2d9 SCRATCH_MNT/test-187/file3 8d9ea4925db533da10a45d2919d8d2d9 SCRATCH_MNT/test-187/file3.chk CoW with multiple extents? +pwrite: No space left on device +/opt/test-187/file3.chk: Input/output error +./common/rc: line 192: 24965 Segmentation fault $MOUNT_PROG '_mount_ops_filter $*' Compare files ... (Run 'diff -u tests/generic/187.out /tmp/xfstests/results//generic/187.out.bad' to see the entire diff) _check_xfs_filesystem: filesystem on /dev/sdf has dirty log (see /tmp/xfstests/results//generic/187.full for details) [ 4307.852659] run fstests generic/187 at 2017-04-11 18:13:44 [ 4308.507527] XFS (sdf): Unmounting Filesystem [ 4308.627786] XFS (sdf): EXPERIMENTAL reverse mapping btree feature enabled. Use at your own risk! [ 4308.628673] XFS (sdf): EXPERIMENTAL reflink feature enabled. Use at your own risk! [ 4308.629822] XFS (sdf): Mounting V5 Filesystem [ 4308.634470] XFS (sdf): Ending clean mount [ 4308.692509] XFS (sdf): Unmounting Filesystem [ 4308.876024] XFS (sdf): EXPERIMENTAL reverse mapping btree feature enabled. Use at your own risk! [ 4308.877205] XFS (sdf): EXPERIMENTAL reflink feature enabled. Use at your own risk! [ 4308.878274] XFS (sdf): Mounting V5 Filesystem [ 4308.883306] XFS (sdf): Ending clean mount [ 4328.077062] XFS (sdf): Unmounting Filesystem [ 4328.128459] XFS (sdf): EXPERIMENTAL reverse mapping btree feature enabled. Use at your own risk! [ 4328.129473] XFS (sdf): EXPERIMENTAL reflink feature enabled. Use at your own risk! [ 4328.130719] XFS (sdf): Mounting V5 Filesystem [ 4328.166304] XFS (sdf): Ending clean mount [ 4362.064634] XFS (sdf): Unmounting Filesystem [ 4362.157219] XFS (sdf): EXPERIMENTAL reverse mapping btree feature enabled. Use at your own risk! [ 4362.159172] XFS (sdf): EXPERIMENTAL reflink feature enabled. Use at your own risk! [ 4362.161274] XFS (sdf): Mounting V5 Filesystem [ 4362.223441] XFS (sdf): Ending clean mount [ 4363.236809] XFS (sdf): xfs_do_force_shutdown(0x1) called from line 236 of file /raid/home/djwong/cdev/work/linux-dgc/fs/xfs/libxfs/xfs_defer.c. Return address = 0xffffffffa0203400 [ 4363.240663] XFS (sdf): I/O Error Detected. Shutting down filesystem [ 4363.241532] XFS (sdf): Please umount the filesystem and rectify the problem(s) [ 4363.375513] XFS (sdf): Unmounting Filesystem [ 4363.422278] XFS (sdf): EXPERIMENTAL reverse mapping btree feature enabled. Use at your own risk! [ 4363.432790] XFS (sdf): EXPERIMENTAL reflink feature enabled. Use at your own risk! [ 4363.435026] XFS (sdf): Mounting V5 Filesystem [ 4363.538385] XFS (sdf): Starting recovery (logdev: internal) [ 4363.630364] XFS: Assertion failed: ifp->if_flags & XFS_IFEXTENTS, file: /raid/home/djwong/cdev/work/linux-dgc/fs/xfs/libxfs/xfs_bmap.c, line: 4706 [ 4363.632502] ------------[ cut here ]------------ [ 4363.633162] kernel BUG at /raid/home/djwong/cdev/work/linux-dgc/fs/xfs/xfs_message.c:113! [ 4363.634304] invalid opcode: 0000 [#1] PREEMPT SMP [ 4363.634969] Dumping ftrace buffer: [ 4363.635463] (ftrace buffer empty) [ 4363.635974] Modules linked in: deadline_iosched dm_snapshot dm_bufio ext4 jbd2 mbcache dm_flakey xfs libcrc32c dax_pmem nd_pmem dax sch_fq_codel af_packet [last unloaded: scsi_debug] [ 4363.638224] CPU: 3 PID: 24965 Comm: mount Tainted: G W 4.11.0-rc4-dgc #1 [ 4363.639316] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS Ubuntu-1.8.2-1ubuntu1djwong0 04/01/2014 [ 4363.640688] task: ffff880074aaa780 task.stack: ffffc90002e50000 [ 4363.641580] RIP: 0010:assfail+0x20/0x30 [xfs] [ 4363.642205] RSP: 0018:ffffc90002e53ac8 EFLAGS: 00010282 [ 4363.642942] RAX: 00000000ffffffea RBX: ffff880072629b00 RCX: 0000000000000001 [ 4363.643927] RDX: 00000000ffffffc0 RSI: 000000000000000a RDI: ffffffffa02a7e3a [ 4363.644989] RBP: ffffc90002e53ac8 R08: 0000000000000000 R09: 0000000000000000 [ 4363.645939] R10: 000000000000000a R11: f000000000000000 R12: 0000000000006740 [ 4363.646964] R13: ffff88007497d000 R14: 00000000000350ee R15: ffff880072629b48 [ 4363.648019] FS: 00007f16ba56d840(0000) GS:ffff88007f600000(0000) knlGS:0000000000000000 [ 4363.649091] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 4363.649851] CR2: 00007f16b9d95170 CR3: 0000000069983000 CR4: 00000000000006e0 [ 4363.650796] Call Trace: [ 4363.651182] xfs_bmap_finish_one+0x5fa/0x6f0 [xfs] [ 4363.651947] ? rcu_read_lock_sched_held+0x72/0x80 [ 4363.652724] ? kmem_zone_alloc+0x81/0x120 [xfs] [ 4363.653373] xfs_trans_log_finish_bmap_update+0x40/0x60 [xfs] [ 4363.654209] xfs_bui_recover+0x265/0x5d0 [xfs] [ 4363.654897] xlog_recover_process_intents+0x293/0x2b0 [xfs] [ 4363.655723] xlog_recover_finish+0x1d/0xa0 [xfs] [ 4363.656423] xfs_log_mount_finish+0x3a/0x80 [xfs] [ 4363.657125] xfs_mountfs+0x65a/0xad0 [xfs] [ 4363.657745] xfs_fs_fill_super+0x483/0x610 [xfs] [ 4363.658419] mount_bdev+0x180/0x1b0 [ 4363.658954] ? xfs_finish_flags+0x150/0x150 [xfs] [ 4363.659660] xfs_fs_mount+0x15/0x20 [xfs] [ 4363.660256] mount_fs+0x14/0xa0 [ 4363.660724] vfs_kern_mount+0x6b/0x160 [ 4363.661269] do_mount+0x195/0xd30 [ 4363.661754] ? _copy_from_user+0x8c/0xd0 [ 4363.662328] ? memdup_user+0x60/0x90 [ 4363.662846] SyS_mount+0x95/0xe0 [ 4363.663326] entry_SYSCALL_64_fastpath+0x1f/0xc2 [ 4363.663988] RIP: 0033:0x7f16b9e4efaa [ 4363.664490] RSP: 002b:00007ffe9202d4a8 EFLAGS: 00000202 ORIG_RAX: 00000000000000a5 [ 4363.665515] RAX: ffffffffffffffda RBX: 00007f16ba14763a RCX: 00007f16b9e4efaa [ 4363.666484] RDX: 0000000002579240 RSI: 0000000002579280 RDI: 0000000002579260 [ 4363.667452] RBP: 0000000002579120 R08: 0000000000000000 R09: 0000000000000012 [ 4363.668431] R10: 00000000c0ed0000 R11: 0000000000000202 R12: 00007f16ba35783c [ 4363.669370] R13: 0000000000000000 R14: 0000000000000000 R15: 0000000000000003 [ 4363.670319] Code: 00 66 2e 0f 1f 84 00 00 00 00 00 66 66 66 66 90 55 48 89 f1 41 89 d0 48 c7 c6 d0 66 2b a0 48 89 fa 31 ff 48 89 e5 e8 50 fa ff ff <0f> 0b 0f 1f 40 00 66 2e 0f 1f 84 00 00 00 00 00 66 66 66 66 90 [ 4363.673053] RIP: assfail+0x20/0x30 [xfs] RSP: ffffc90002e53ac8 [ 4363.674170] ---[ end trace 6f71a70ec6fdf3f5 ]--- Looks like _bmapi_remap needs to be able to _iread_extents() if the data fork hasn't been loaded during log recovery. --D > -- > To unsubscribe from this list: send the line "unsubscribe linux-xfs" in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html -- To unsubscribe from this list: send the line "unsubscribe linux-xfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
> Looks like _bmapi_remap needs to be able to _iread_extents() if the > data fork hasn't been loaded during log recovery. Yeah, probably. I'll respin once more with that included. -- To unsubscribe from this list: send the line "unsubscribe linux-xfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Wed, Apr 12, 2017 at 08:00:26AM +0200, Christoph Hellwig wrote: > > Looks like _bmapi_remap needs to be able to _iread_extents() if the > > data fork hasn't been loaded during log recovery. > > Yeah, probably. I'll respin once more with that included. Unfortunately, even after adding in the necessary loading code I still get -ENOSPC back from xfs_bmap_add_extent_hole_real which causes log recovery to fail. --D -- To unsubscribe from this list: send the line "unsubscribe linux-xfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Tue, Apr 11, 2017 at 11:44:20PM -0700, Darrick J. Wong wrote: > On Wed, Apr 12, 2017 at 08:00:26AM +0200, Christoph Hellwig wrote: > > > Looks like _bmapi_remap needs to be able to _iread_extents() if the > > > data fork hasn't been loaded during log recovery. > > > > Yeah, probably. I'll respin once more with that included. > > Unfortunately, even after adding in the necessary loading code I still > get -ENOSPC back from xfs_bmap_add_extent_hole_real which causes log > recovery to fail. The test in your configuration already fails with -ENOSPC for me on for-next.. -- To unsubscribe from this list: send the line "unsubscribe linux-xfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Wed, Apr 12, 2017 at 10:00:59AM +0200, Christoph Hellwig wrote: > On Tue, Apr 11, 2017 at 11:44:20PM -0700, Darrick J. Wong wrote: > > On Wed, Apr 12, 2017 at 08:00:26AM +0200, Christoph Hellwig wrote: > > > > Looks like _bmapi_remap needs to be able to _iread_extents() if the > > > > data fork hasn't been loaded during log recovery. > > > > > > Yeah, probably. I'll respin once more with that included. > > > > Unfortunately, even after adding in the necessary loading code I still > > get -ENOSPC back from xfs_bmap_add_extent_hole_real which causes log > > recovery to fail. > > The test in your configuration already fails with -ENOSPC for me on > for-next.. Crap, there are actually /two/ problems here: The first problem is that xfs_reflink_end_cow reserves only enough blocks to handle adding 1 extent. This is problematic if we fragment free space, have to do CoW, and then have to perform multiple bmap btree expansions, which g/187 seems to hit. The second problem is that the BUI recovery routine doesn't reserve /any/ blocks to handle btree splits, so when the first problem takes the fs down, recovery also fails. Evidently with 1k blocks we can hit this fairly often in g/187 if the scratch fs is big enough. Ok, patch soon. <sigh> --D > -- > To unsubscribe from this list: send the line "unsubscribe linux-xfs" in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html -- To unsubscribe from this list: send the line "unsubscribe linux-xfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
diff --git a/fs/xfs/libxfs/xfs_bmap.c b/fs/xfs/libxfs/xfs_bmap.c index faf7cb0a28f9..414cfa91af01 100644 --- a/fs/xfs/libxfs/xfs_bmap.c +++ b/fs/xfs/libxfs/xfs_bmap.c @@ -3856,61 +3856,6 @@ xfs_bmap_btalloc( } /* - * For a remap operation, just "allocate" an extent at the address that the - * caller passed in, and ensure that the AGFL is the right size. The caller - * will then map the "allocated" extent into the file somewhere. - */ -STATIC int -xfs_bmap_remap_alloc( - struct xfs_trans *tp, - struct xfs_inode *ip, - xfs_fsblock_t startblock, - xfs_extlen_t length) -{ - struct xfs_mount *mp = tp->t_mountp; - struct xfs_alloc_arg args; - int error; - - /* - * validate that the block number is legal - the enables us to detect - * and handle a silent filesystem corruption rather than crashing. - */ - memset(&args, 0, sizeof(struct xfs_alloc_arg)); - args.tp = tp; - args.mp = mp; - args.agno = XFS_FSB_TO_AGNO(mp, startblock); - args.agbno = XFS_FSB_TO_AGBNO(mp, startblock); - - if (args.agno >= mp->m_sb.sb_agcount || - args.agbno >= mp->m_sb.sb_agblocks) - return -EFSCORRUPTED; - - /* "Allocate" the extent from the range we passed in. */ - trace_xfs_bmap_remap_alloc(ip, startblock, length); - - ip->i_d.di_nblocks += length; - xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE); - - /* Fix the freelist, like a real allocator does. */ - args.datatype = XFS_ALLOC_USERDATA | XFS_ALLOC_NOBUSY; - args.pag = xfs_perag_get(mp, args.agno); - ASSERT(args.pag); - - /* - * The freelist fixing code will decline the allocation if - * the size and shape of the free space doesn't allow for - * allocating the extent and updating all the metadata that - * happens during an allocation. We're remapping, not - * allocating, so skip that check by pretending to be freeing. - */ - error = xfs_alloc_fix_freelist(&args, XFS_ALLOC_FLAG_FREEING); - xfs_perag_put(args.pag); - if (error) - trace_xfs_bmap_remap_alloc_error(ip, error, _RET_IP_); - return error; -} - -/* * xfs_bmap_alloc is called by xfs_bmapi to allocate an extent for a file. * It figures out where to ask the underlying allocator to put the new extent. */ @@ -4806,9 +4751,8 @@ xfs_bmapi_remap( ASSERT(got.br_startoff - bno >= len); } - error = xfs_bmap_remap_alloc(tp, ip, startblock, len); - if (error) - goto error0; + ip->i_d.di_nblocks += len; + xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE); if (ifp->if_flags & XFS_IFBROOT) { cur = xfs_bmbt_init_cursor(mp, tp, ip, XFS_DATA_FORK); diff --git a/fs/xfs/xfs_trace.h b/fs/xfs/xfs_trace.h index 4f96dc953fbe..cba10daf8391 100644 --- a/fs/xfs/xfs_trace.h +++ b/fs/xfs/xfs_trace.h @@ -3003,31 +3003,6 @@ DEFINE_EVENT(xfs_inode_error_class, name, \ unsigned long caller_ip), \ TP_ARGS(ip, error, caller_ip)) -/* reflink allocator */ -TRACE_EVENT(xfs_bmap_remap_alloc, - TP_PROTO(struct xfs_inode *ip, xfs_fsblock_t fsbno, - xfs_extlen_t len), - TP_ARGS(ip, fsbno, len), - TP_STRUCT__entry( - __field(dev_t, dev) - __field(xfs_ino_t, ino) - __field(xfs_fsblock_t, fsbno) - __field(xfs_extlen_t, len) - ), - TP_fast_assign( - __entry->dev = VFS_I(ip)->i_sb->s_dev; - __entry->ino = ip->i_ino; - __entry->fsbno = fsbno; - __entry->len = len; - ), - TP_printk("dev %d:%d ino 0x%llx fsbno 0x%llx len %x", - MAJOR(__entry->dev), MINOR(__entry->dev), - __entry->ino, - __entry->fsbno, - __entry->len) -); -DEFINE_INODE_ERROR_EVENT(xfs_bmap_remap_alloc_error); - /* reflink tracepoint classes */ /* two-file io tracepoint class */
The main thing that xfs_bmap_remap_alloc does is fixing the AGFL, similar to what we do in the space allocator. But the reflink code doesn't touch the allocation btree unlike the normal space allocator, so we couldn't care less about the state of the AGFL. So remove xfs_bmap_remap_alloc and just handle the di_nblocks update in the caller. Signed-off-by: Christoph Hellwig <hch@lst.de> --- fs/xfs/libxfs/xfs_bmap.c | 60 ++---------------------------------------------- fs/xfs/xfs_trace.h | 25 -------------------- 2 files changed, 2 insertions(+), 83 deletions(-)