linux-stable/fs/xfs/scrub/bmap.c

859 lines
23 KiB
C
Raw Normal View History

// SPDX-License-Identifier: GPL-2.0-or-later
/*
* Copyright (C) 2017-2023 Oracle. All Rights Reserved.
* Author: Darrick J. Wong <djwong@kernel.org>
*/
#include "xfs.h"
#include "xfs_fs.h"
#include "xfs_shared.h"
#include "xfs_format.h"
#include "xfs_trans_resv.h"
#include "xfs_mount.h"
#include "xfs_btree.h"
#include "xfs_bit.h"
#include "xfs_log_format.h"
#include "xfs_trans.h"
#include "xfs_inode.h"
#include "xfs_alloc.h"
#include "xfs_bmap.h"
#include "xfs_bmap_btree.h"
#include "xfs_rmap.h"
#include "xfs_rmap_btree.h"
#include "scrub/scrub.h"
#include "scrub/common.h"
#include "scrub/btree.h"
#include "xfs_ag.h"
/* Set us up with an inode's bmap. */
int
xchk_setup_inode_bmap(
struct xfs_scrub *sc)
{
int error;
if (xchk_need_intent_drain(sc))
xchk_fsgates_enable(sc, XCHK_FSGATES_DRAIN);
error = xchk_iget_for_scrubbing(sc);
if (error)
goto out;
sc->ilock_flags = XFS_IOLOCK_EXCL | XFS_MMAPLOCK_EXCL;
xfs_ilock(sc->ip, sc->ilock_flags);
/*
* We don't want any ephemeral data fork updates sitting around
* while we inspect block mappings, so wait for directio to finish
* and flush dirty data if we have delalloc reservations.
*/
if (S_ISREG(VFS_I(sc->ip)->i_mode) &&
sc->sm->sm_type == XFS_SCRUB_TYPE_BMBTD) {
struct address_space *mapping = VFS_I(sc->ip)->i_mapping;
inode_dio_wait(VFS_I(sc->ip));
/*
* Try to flush all incore state to disk before we examine the
* space mappings for the data fork. Leave accumulated errors
* in the mapping for the writer threads to consume.
*
* On ENOSPC or EIO writeback errors, we continue into the
* extent mapping checks because write failures do not
* necessarily imply anything about the correctness of the file
* metadata. The metadata and the file data could be on
* completely separate devices; a media failure might only
* affect a subset of the disk, etc. We can handle delalloc
* extents in the scrubber, so leaving them in memory is fine.
*/
error = filemap_fdatawrite(mapping);
if (!error)
error = filemap_fdatawait_keep_errors(mapping);
if (error && (error != -ENOSPC && error != -EIO))
goto out;
}
/* Got the inode, lock it and we're ready to go. */
error = xchk_trans_alloc(sc, 0);
if (error)
goto out;
sc->ilock_flags |= XFS_ILOCK_EXCL;
xfs_ilock(sc->ip, XFS_ILOCK_EXCL);
out:
/* scrub teardown will unlock and release the inode */
return error;
}
/*
* Inode fork block mapping (BMBT) scrubber.
* More complex than the others because we have to scrub
* all the extents regardless of whether or not the fork
* is in btree format.
*/
struct xchk_bmap_info {
struct xfs_scrub *sc;
struct xfs_iext_cursor icur;
xfs_fileoff_t lastoff;
bool is_rt;
bool is_shared;
bool was_loaded;
int whichfork;
};
/* Look for a corresponding rmap for this irec. */
static inline bool
xchk_bmap_get_rmap(
struct xchk_bmap_info *info,
struct xfs_bmbt_irec *irec,
xfs_agblock_t agbno,
uint64_t owner,
struct xfs_rmap_irec *rmap)
{
xfs_fileoff_t offset;
unsigned int rflags = 0;
int has_rmap;
int error;
if (info->whichfork == XFS_ATTR_FORK)
rflags |= XFS_RMAP_ATTR_FORK;
if (irec->br_state == XFS_EXT_UNWRITTEN)
rflags |= XFS_RMAP_UNWRITTEN;
/*
* CoW staging extents are owned (on disk) by the refcountbt, so
* their rmaps do not have offsets.
*/
if (info->whichfork == XFS_COW_FORK)
offset = 0;
else
offset = irec->br_startoff;
/*
* If the caller thinks this could be a shared bmbt extent (IOWs,
* any data fork extent of a reflink inode) then we have to use the
* range rmap lookup to make sure we get the correct owner/offset.
*/
if (info->is_shared) {
error = xfs_rmap_lookup_le_range(info->sc->sa.rmap_cur, agbno,
owner, offset, rflags, rmap, &has_rmap);
} else {
error = xfs_rmap_lookup_le(info->sc->sa.rmap_cur, agbno,
owner, offset, rflags, rmap, &has_rmap);
}
if (!xchk_should_check_xref(info->sc, &error, &info->sc->sa.rmap_cur))
return false;
if (!has_rmap)
xchk_fblock_xref_set_corrupt(info->sc, info->whichfork,
irec->br_startoff);
return has_rmap;
}
static inline bool
xchk_bmap_has_prev(
struct xchk_bmap_info *info,
struct xfs_bmbt_irec *irec)
{
struct xfs_bmbt_irec got;
struct xfs_ifork *ifp;
ifp = xfs_ifork_ptr(info->sc->ip, info->whichfork);
if (!xfs_iext_peek_prev_extent(ifp, &info->icur, &got))
return false;
if (got.br_startoff + got.br_blockcount != irec->br_startoff)
return false;
if (got.br_startblock + got.br_blockcount != irec->br_startblock)
return false;
if (got.br_state != irec->br_state)
return false;
return true;
}
static inline bool
xchk_bmap_has_next(
struct xchk_bmap_info *info,
struct xfs_bmbt_irec *irec)
{
struct xfs_bmbt_irec got;
struct xfs_ifork *ifp;
ifp = xfs_ifork_ptr(info->sc->ip, info->whichfork);
if (!xfs_iext_peek_next_extent(ifp, &info->icur, &got))
return false;
if (irec->br_startoff + irec->br_blockcount != got.br_startoff)
return false;
if (irec->br_startblock + irec->br_blockcount != got.br_startblock)
return false;
if (got.br_state != irec->br_state)
return false;
return true;
}
/* Make sure that we have rmapbt records for this extent. */
STATIC void
xchk_bmap_xref_rmap(
struct xchk_bmap_info *info,
struct xfs_bmbt_irec *irec,
xfs_agblock_t agbno)
{
struct xfs_rmap_irec rmap;
unsigned long long rmap_end;
uint64_t owner;
if (!info->sc->sa.rmap_cur || xchk_skip_xref(info->sc->sm))
return;
if (info->whichfork == XFS_COW_FORK)
owner = XFS_RMAP_OWN_COW;
else
owner = info->sc->ip->i_ino;
/* Find the rmap record for this irec. */
if (!xchk_bmap_get_rmap(info, irec, agbno, owner, &rmap))
return;
/* Check the rmap. */
rmap_end = (unsigned long long)rmap.rm_startblock + rmap.rm_blockcount;
if (rmap.rm_startblock > agbno ||
agbno + irec->br_blockcount > rmap_end)
xchk_fblock_xref_set_corrupt(info->sc, info->whichfork,
irec->br_startoff);
/*
* Check the logical offsets if applicable. CoW staging extents
* don't track logical offsets since the mappings only exist in
* memory.
*/
if (info->whichfork != XFS_COW_FORK) {
rmap_end = (unsigned long long)rmap.rm_offset +
rmap.rm_blockcount;
if (rmap.rm_offset > irec->br_startoff ||
irec->br_startoff + irec->br_blockcount > rmap_end)
xchk_fblock_xref_set_corrupt(info->sc,
info->whichfork, irec->br_startoff);
}
if (rmap.rm_owner != owner)
xchk_fblock_xref_set_corrupt(info->sc, info->whichfork,
irec->br_startoff);
/*
* Check for discrepancies between the unwritten flag in the irec and
* the rmap. Note that the (in-memory) CoW fork distinguishes between
* unwritten and written extents, but we don't track that in the rmap
* records because the blocks are owned (on-disk) by the refcountbt,
* which doesn't track unwritten state.
*/
if (owner != XFS_RMAP_OWN_COW &&
!!(irec->br_state == XFS_EXT_UNWRITTEN) !=
!!(rmap.rm_flags & XFS_RMAP_UNWRITTEN))
xchk_fblock_xref_set_corrupt(info->sc, info->whichfork,
irec->br_startoff);
if (!!(info->whichfork == XFS_ATTR_FORK) !=
!!(rmap.rm_flags & XFS_RMAP_ATTR_FORK))
xchk_fblock_xref_set_corrupt(info->sc, info->whichfork,
irec->br_startoff);
if (rmap.rm_flags & XFS_RMAP_BMBT_BLOCK)
xchk_fblock_xref_set_corrupt(info->sc, info->whichfork,
irec->br_startoff);
/*
* If the rmap starts before this bmbt record, make sure there's a bmbt
* record for the previous offset that is contiguous with this mapping.
* Skip this for CoW fork extents because the refcount btree (and not
* the inode) is the ondisk owner for those extents.
*/
if (info->whichfork != XFS_COW_FORK && rmap.rm_startblock < agbno &&
!xchk_bmap_has_prev(info, irec)) {
xchk_fblock_xref_set_corrupt(info->sc, info->whichfork,
irec->br_startoff);
return;
}
/*
* If the rmap ends after this bmbt record, make sure there's a bmbt
* record for the next offset that is contiguous with this mapping.
* Skip this for CoW fork extents because the refcount btree (and not
* the inode) is the ondisk owner for those extents.
*/
rmap_end = (unsigned long long)rmap.rm_startblock + rmap.rm_blockcount;
if (info->whichfork != XFS_COW_FORK &&
rmap_end > agbno + irec->br_blockcount &&
!xchk_bmap_has_next(info, irec)) {
xchk_fblock_xref_set_corrupt(info->sc, info->whichfork,
irec->br_startoff);
return;
}
}
/* Cross-reference a single rtdev extent record. */
STATIC void
xchk_bmap_rt_iextent_xref(
struct xfs_inode *ip,
struct xchk_bmap_info *info,
struct xfs_bmbt_irec *irec)
{
xchk_xref_is_used_rt_space(info->sc, irec->br_startblock,
irec->br_blockcount);
}
/* Cross-reference a single datadev extent record. */
STATIC void
xchk_bmap_iextent_xref(
struct xfs_inode *ip,
struct xchk_bmap_info *info,
struct xfs_bmbt_irec *irec)
{
struct xfs_owner_info oinfo;
struct xfs_mount *mp = info->sc->mp;
xfs_agnumber_t agno;
xfs_agblock_t agbno;
xfs_extlen_t len;
int error;
agno = XFS_FSB_TO_AGNO(mp, irec->br_startblock);
agbno = XFS_FSB_TO_AGBNO(mp, irec->br_startblock);
len = irec->br_blockcount;
error = xchk_ag_init_existing(info->sc, agno, &info->sc->sa);
if (!xchk_fblock_process_error(info->sc, info->whichfork,
irec->br_startoff, &error))
xfs: fix perag structure refcounting error when scrub fails The kernel test robot found the following bug when running xfs/355 to scrub a bmap btree: XFS: Assertion failed: !sa->pag, file: fs/xfs/scrub/common.c, line: 412 ------------[ cut here ]------------ kernel BUG at fs/xfs/xfs_message.c:110! invalid opcode: 0000 [#1] SMP PTI CPU: 2 PID: 1415 Comm: xfs_scrub Not tainted 5.14.0-rc4-00021-g48c6615cc557 #1 Hardware name: Hewlett-Packard p6-1451cx/2ADA, BIOS 8.15 02/05/2013 RIP: 0010:assfail+0x23/0x28 [xfs] RSP: 0018:ffffc9000aacb890 EFLAGS: 00010202 RAX: 0000000000000000 RBX: ffffc9000aacbcc8 RCX: 0000000000000000 RDX: 00000000ffffffc0 RSI: 000000000000000a RDI: ffffffffc09e7dcd RBP: ffffc9000aacbc80 R08: ffff8881fdf17d50 R09: 0000000000000000 R10: 000000000000000a R11: f000000000000000 R12: 0000000000000000 R13: ffff88820c7ed000 R14: 0000000000000001 R15: ffffc9000aacb980 FS: 00007f185b955700(0000) GS:ffff8881fdf00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 00007f7f6ef43000 CR3: 000000020de38002 CR4: 00000000001706e0 Call Trace: xchk_ag_read_headers+0xda/0x100 [xfs] xchk_ag_init+0x15/0x40 [xfs] xchk_btree_check_block_owner+0x76/0x180 [xfs] xchk_btree_get_block+0xd0/0x140 [xfs] xchk_btree+0x32e/0x440 [xfs] xchk_bmap_btree+0xd4/0x140 [xfs] xchk_bmap+0x1eb/0x3c0 [xfs] xfs_scrub_metadata+0x227/0x4c0 [xfs] xfs_ioc_scrub_metadata+0x50/0xc0 [xfs] xfs_file_ioctl+0x90c/0xc40 [xfs] __x64_sys_ioctl+0x83/0xc0 do_syscall_64+0x3b/0xc0 The unusual handling of errors while initializing struct xchk_ag is the root cause here. Since the beginning of xfs_scrub, the goal of xchk_ag_read_headers has been to read all three AG header buffers and attach them both to the xchk_ag structure and the scrub transaction. Corruption errors on any of the three headers doesn't necessarily trigger an immediate return to userspace, because xfs_scrub can also tell us to /fix/ the problem. In other words, it's possible for the xchk_ag init functions to return an error code and a partially filled out structure so that scrub can use however much information it managed to pull. Before 5.15, it was sufficient to cancel (or commit) the scrub transaction on the way out of the scrub code to release the buffers. Ccommit 48c6615cc557 added a reference to the perag structure to struct xchk_ag. Since perag structures are not attached to transactions like buffers are, this adds the requirement that the perag ref be released explicitly. The scrub teardown function xchk_teardown was amended to do this for the xchk_ag embedded in struct xfs_scrub. Unfortunately, I forgot that certain parts of the scrub code probe multiple AGs and therefore handle the initialization and cleanup on their own. Specifically, the bmbt scrubber will initialize it long enough to cross-reference AG metadata for btree blocks and for the extent mappings in the bmbt. If one of the AG headers is corrupt, the init function returns with a live perag structure reference and some of the AG header buffers. If an error occurs, the cross referencing will be noted as XCORRUPTion and skipped, but the main scrub process will move on to the next record. It is now necessary to release the perag reference before we try to analyze something from a different AG, or else we'll trip over the assertion noted above. Fixes: 48c6615cc557 ("xfs: grab active perag ref when reading AG headers") Reported-by: kernel test robot <oliver.sang@intel.com> Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Chandan Babu R <chandanrlinux@gmail.com>
2021-08-19 19:07:49 +00:00
goto out_free;
xchk_xref_is_used_space(info->sc, agbno, len);
xchk_xref_is_not_inode_chunk(info->sc, agbno, len);
xchk_bmap_xref_rmap(info, irec, agbno);
switch (info->whichfork) {
case XFS_DATA_FORK:
if (!xfs_is_reflink_inode(info->sc->ip)) {
xfs_rmap_ino_owner(&oinfo, info->sc->ip->i_ino,
info->whichfork, irec->br_startoff);
xchk_xref_is_only_owned_by(info->sc, agbno,
irec->br_blockcount, &oinfo);
xchk_xref_is_not_shared(info->sc, agbno,
irec->br_blockcount);
}
xchk_xref_is_not_cow_staging(info->sc, agbno,
irec->br_blockcount);
break;
case XFS_ATTR_FORK:
xfs_rmap_ino_owner(&oinfo, info->sc->ip->i_ino,
info->whichfork, irec->br_startoff);
xchk_xref_is_only_owned_by(info->sc, agbno, irec->br_blockcount,
&oinfo);
xchk_xref_is_not_shared(info->sc, agbno,
irec->br_blockcount);
xchk_xref_is_not_cow_staging(info->sc, agbno,
irec->br_blockcount);
break;
case XFS_COW_FORK:
xchk_xref_is_only_owned_by(info->sc, agbno, irec->br_blockcount,
&XFS_RMAP_OINFO_COW);
xchk_xref_is_cow_staging(info->sc, agbno,
irec->br_blockcount);
xchk_xref_is_not_shared(info->sc, agbno,
irec->br_blockcount);
break;
}
xfs: fix perag structure refcounting error when scrub fails The kernel test robot found the following bug when running xfs/355 to scrub a bmap btree: XFS: Assertion failed: !sa->pag, file: fs/xfs/scrub/common.c, line: 412 ------------[ cut here ]------------ kernel BUG at fs/xfs/xfs_message.c:110! invalid opcode: 0000 [#1] SMP PTI CPU: 2 PID: 1415 Comm: xfs_scrub Not tainted 5.14.0-rc4-00021-g48c6615cc557 #1 Hardware name: Hewlett-Packard p6-1451cx/2ADA, BIOS 8.15 02/05/2013 RIP: 0010:assfail+0x23/0x28 [xfs] RSP: 0018:ffffc9000aacb890 EFLAGS: 00010202 RAX: 0000000000000000 RBX: ffffc9000aacbcc8 RCX: 0000000000000000 RDX: 00000000ffffffc0 RSI: 000000000000000a RDI: ffffffffc09e7dcd RBP: ffffc9000aacbc80 R08: ffff8881fdf17d50 R09: 0000000000000000 R10: 000000000000000a R11: f000000000000000 R12: 0000000000000000 R13: ffff88820c7ed000 R14: 0000000000000001 R15: ffffc9000aacb980 FS: 00007f185b955700(0000) GS:ffff8881fdf00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 00007f7f6ef43000 CR3: 000000020de38002 CR4: 00000000001706e0 Call Trace: xchk_ag_read_headers+0xda/0x100 [xfs] xchk_ag_init+0x15/0x40 [xfs] xchk_btree_check_block_owner+0x76/0x180 [xfs] xchk_btree_get_block+0xd0/0x140 [xfs] xchk_btree+0x32e/0x440 [xfs] xchk_bmap_btree+0xd4/0x140 [xfs] xchk_bmap+0x1eb/0x3c0 [xfs] xfs_scrub_metadata+0x227/0x4c0 [xfs] xfs_ioc_scrub_metadata+0x50/0xc0 [xfs] xfs_file_ioctl+0x90c/0xc40 [xfs] __x64_sys_ioctl+0x83/0xc0 do_syscall_64+0x3b/0xc0 The unusual handling of errors while initializing struct xchk_ag is the root cause here. Since the beginning of xfs_scrub, the goal of xchk_ag_read_headers has been to read all three AG header buffers and attach them both to the xchk_ag structure and the scrub transaction. Corruption errors on any of the three headers doesn't necessarily trigger an immediate return to userspace, because xfs_scrub can also tell us to /fix/ the problem. In other words, it's possible for the xchk_ag init functions to return an error code and a partially filled out structure so that scrub can use however much information it managed to pull. Before 5.15, it was sufficient to cancel (or commit) the scrub transaction on the way out of the scrub code to release the buffers. Ccommit 48c6615cc557 added a reference to the perag structure to struct xchk_ag. Since perag structures are not attached to transactions like buffers are, this adds the requirement that the perag ref be released explicitly. The scrub teardown function xchk_teardown was amended to do this for the xchk_ag embedded in struct xfs_scrub. Unfortunately, I forgot that certain parts of the scrub code probe multiple AGs and therefore handle the initialization and cleanup on their own. Specifically, the bmbt scrubber will initialize it long enough to cross-reference AG metadata for btree blocks and for the extent mappings in the bmbt. If one of the AG headers is corrupt, the init function returns with a live perag structure reference and some of the AG header buffers. If an error occurs, the cross referencing will be noted as XCORRUPTion and skipped, but the main scrub process will move on to the next record. It is now necessary to release the perag reference before we try to analyze something from a different AG, or else we'll trip over the assertion noted above. Fixes: 48c6615cc557 ("xfs: grab active perag ref when reading AG headers") Reported-by: kernel test robot <oliver.sang@intel.com> Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Chandan Babu R <chandanrlinux@gmail.com>
2021-08-19 19:07:49 +00:00
out_free:
xchk_ag_free(info->sc, &info->sc->sa);
}
/*
* Directories and attr forks should never have blocks that can't be addressed
* by a xfs_dablk_t.
*/
STATIC void
xchk_bmap_dirattr_extent(
struct xfs_inode *ip,
struct xchk_bmap_info *info,
struct xfs_bmbt_irec *irec)
{
struct xfs_mount *mp = ip->i_mount;
xfs_fileoff_t off;
if (!S_ISDIR(VFS_I(ip)->i_mode) && info->whichfork != XFS_ATTR_FORK)
return;
if (!xfs_verify_dablk(mp, irec->br_startoff))
xchk_fblock_set_corrupt(info->sc, info->whichfork,
irec->br_startoff);
off = irec->br_startoff + irec->br_blockcount - 1;
if (!xfs_verify_dablk(mp, off))
xchk_fblock_set_corrupt(info->sc, info->whichfork, off);
}
/* Scrub a single extent record. */
STATIC void
xchk_bmap_iextent(
struct xfs_inode *ip,
struct xchk_bmap_info *info,
struct xfs_bmbt_irec *irec)
{
struct xfs_mount *mp = info->sc->mp;
/*
* Check for out-of-order extents. This record could have come
* from the incore list, for which there is no ordering check.
*/
if (irec->br_startoff < info->lastoff)
xchk_fblock_set_corrupt(info->sc, info->whichfork,
irec->br_startoff);
if (!xfs_verify_fileext(mp, irec->br_startoff, irec->br_blockcount))
xchk_fblock_set_corrupt(info->sc, info->whichfork,
irec->br_startoff);
xchk_bmap_dirattr_extent(ip, info, irec);
/* There should never be a "hole" extent in either extent list. */
if (irec->br_startblock == HOLESTARTBLOCK)
xchk_fblock_set_corrupt(info->sc, info->whichfork,
irec->br_startoff);
/* Make sure the extent points to a valid place. */
if (irec->br_blockcount > XFS_MAX_BMBT_EXTLEN)
xchk_fblock_set_corrupt(info->sc, info->whichfork,
irec->br_startoff);
if (info->is_rt &&
!xfs_verify_rtext(mp, irec->br_startblock, irec->br_blockcount))
xchk_fblock_set_corrupt(info->sc, info->whichfork,
irec->br_startoff);
if (!info->is_rt &&
!xfs_verify_fsbext(mp, irec->br_startblock, irec->br_blockcount))
xchk_fblock_set_corrupt(info->sc, info->whichfork,
irec->br_startoff);
/* We don't allow unwritten extents on attr forks. */
if (irec->br_state == XFS_EXT_UNWRITTEN &&
info->whichfork == XFS_ATTR_FORK)
xchk_fblock_set_corrupt(info->sc, info->whichfork,
irec->br_startoff);
if (info->sc->sm->sm_flags & XFS_SCRUB_OFLAG_CORRUPT)
return;
if (info->is_rt)
xchk_bmap_rt_iextent_xref(ip, info, irec);
else
xchk_bmap_iextent_xref(ip, info, irec);
}
/* Scrub a bmbt record. */
STATIC int
xchk_bmapbt_rec(
struct xchk_btree *bs,
const union xfs_btree_rec *rec)
{
struct xfs_bmbt_irec irec;
struct xfs_bmbt_irec iext_irec;
struct xfs_iext_cursor icur;
struct xchk_bmap_info *info = bs->private;
struct xfs_inode *ip = bs->cur->bc_ino.ip;
struct xfs_buf *bp = NULL;
struct xfs_btree_block *block;
struct xfs_ifork *ifp = xfs_ifork_ptr(ip, info->whichfork);
uint64_t owner;
int i;
/*
* Check the owners of the btree blocks up to the level below
* the root since the verifiers don't do that.
*/
if (xfs_has_crc(bs->cur->bc_mp) &&
bs->cur->bc_levels[0].ptr == 1) {
for (i = 0; i < bs->cur->bc_nlevels - 1; i++) {
block = xfs_btree_get_block(bs->cur, i, &bp);
owner = be64_to_cpu(block->bb_u.l.bb_owner);
if (owner != ip->i_ino)
xchk_fblock_set_corrupt(bs->sc,
info->whichfork, 0);
}
}
/*
* Check that the incore extent tree contains an extent that matches
* this one exactly. We validate those cached bmaps later, so we don't
* need to check them here. If the incore extent tree was just loaded
* from disk by the scrubber, we assume that its contents match what's
* on disk (we still hold the ILOCK) and skip the equivalence check.
*/
if (!info->was_loaded)
return 0;
xfs: use a b+tree for the in-core extent list Replace the current linear list and the indirection array for the in-core extent list with a b+tree to avoid the need for larger memory allocations for the indirection array when lots of extents are present. The current extent list implementations leads to heavy pressure on the memory allocator when modifying files with a high extent count, and can lead to high latencies because of that. The replacement is a b+tree with a few quirks. The leaf nodes directly store the extent record in two u64 values. The encoding is a little bit different from the existing in-core extent records so that the start offset and length which are required for lookups can be retreived with simple mask operations. The inner nodes store a 64-bit key containing the start offset in the first half of the node, and the pointers to the next lower level in the second half. In either case we walk the node from the beginninig to the end and do a linear search, as that is more efficient for the low number of cache lines touched during a search (2 for the inner nodes, 4 for the leaf nodes) than a binary search. We store termination markers (zero length for the leaf nodes, an otherwise impossible high bit for the inner nodes) to terminate the key list / records instead of storing a count to use the available cache lines as efficiently as possible. One quirk of the algorithm is that while we normally split a node half and half like usual btree implementations we just spill over entries added at the very end of the list to a new node on its own. This means we get a 100% fill grade for the common cases of bulk insertion when reading an inode into memory, and when only sequentially appending to a file. The downside is a slightly higher chance of splits on the first random insertions. Both insert and removal manually recurse into the lower levels, but the bulk deletion of the whole tree is still implemented as a recursive function call, although one limited by the overall depth and with very little stack usage in every iteration. For the first few extents we dynamically grow the list from a single extent to the next powers of two until we have a first full leaf block and that building the actual tree. The code started out based on the generic lib/btree.c code from Joern Engel based on earlier work from Peter Zijlstra, but has since been rewritten beyond recognition. Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2017-11-03 17:34:46 +00:00
xfs_bmbt_disk_get_all(&rec->bmbt, &irec);
if (xfs_bmap_validate_extent(ip, info->whichfork, &irec) != NULL) {
xchk_fblock_set_corrupt(bs->sc, info->whichfork,
irec.br_startoff);
return 0;
}
if (!xfs_iext_lookup_extent(ip, ifp, irec.br_startoff, &icur,
&iext_irec) ||
irec.br_startoff != iext_irec.br_startoff ||
irec.br_startblock != iext_irec.br_startblock ||
irec.br_blockcount != iext_irec.br_blockcount ||
irec.br_state != iext_irec.br_state)
xchk_fblock_set_corrupt(bs->sc, info->whichfork,
irec.br_startoff);
return 0;
}
/* Scan the btree records. */
STATIC int
xchk_bmap_btree(
struct xfs_scrub *sc,
int whichfork,
struct xchk_bmap_info *info)
{
struct xfs_owner_info oinfo;
struct xfs_ifork *ifp = xfs_ifork_ptr(sc->ip, whichfork);
struct xfs_mount *mp = sc->mp;
struct xfs_inode *ip = sc->ip;
struct xfs_btree_cur *cur;
int error;
/* Load the incore bmap cache if it's not loaded. */
info->was_loaded = !xfs_need_iread_extents(ifp);
error = xfs_iread_extents(sc->tp, ip, whichfork);
if (!xchk_fblock_process_error(sc, whichfork, 0, &error))
goto out;
/* Check the btree structure. */
cur = xfs_bmbt_init_cursor(mp, sc->tp, ip, whichfork);
xfs_rmap_ino_bmbt_owner(&oinfo, ip->i_ino, whichfork);
error = xchk_btree(sc, cur, xchk_bmapbt_rec, &oinfo, info);
xfs_btree_del_cursor(cur, error);
out:
return error;
}
struct xchk_bmap_check_rmap_info {
struct xfs_scrub *sc;
int whichfork;
struct xfs_iext_cursor icur;
};
/* Can we find bmaps that fit this rmap? */
STATIC int
xchk_bmap_check_rmap(
struct xfs_btree_cur *cur,
const struct xfs_rmap_irec *rec,
void *priv)
{
struct xfs_bmbt_irec irec;
struct xfs_rmap_irec check_rec;
struct xchk_bmap_check_rmap_info *sbcri = priv;
struct xfs_ifork *ifp;
struct xfs_scrub *sc = sbcri->sc;
bool have_map;
/* Is this even the right fork? */
if (rec->rm_owner != sc->ip->i_ino)
return 0;
if ((sbcri->whichfork == XFS_ATTR_FORK) ^
!!(rec->rm_flags & XFS_RMAP_ATTR_FORK))
return 0;
if (rec->rm_flags & XFS_RMAP_BMBT_BLOCK)
return 0;
/* Now look up the bmbt record. */
ifp = xfs_ifork_ptr(sc->ip, sbcri->whichfork);
if (!ifp) {
xchk_fblock_set_corrupt(sc, sbcri->whichfork,
rec->rm_offset);
goto out;
}
have_map = xfs_iext_lookup_extent(sc->ip, ifp, rec->rm_offset,
&sbcri->icur, &irec);
if (!have_map)
xchk_fblock_set_corrupt(sc, sbcri->whichfork,
rec->rm_offset);
/*
* bmap extent record lengths are constrained to 2^21 blocks in length
* because of space constraints in the on-disk metadata structure.
* However, rmap extent record lengths are constrained only by AG
* length, so we have to loop through the bmbt to make sure that the
* entire rmap is covered by bmbt records.
*/
check_rec = *rec;
while (have_map) {
if (irec.br_startoff != check_rec.rm_offset)
xchk_fblock_set_corrupt(sc, sbcri->whichfork,
check_rec.rm_offset);
if (irec.br_startblock != XFS_AGB_TO_FSB(sc->mp,
cur->bc_ag.pag->pag_agno,
check_rec.rm_startblock))
xchk_fblock_set_corrupt(sc, sbcri->whichfork,
check_rec.rm_offset);
if (irec.br_blockcount > check_rec.rm_blockcount)
xchk_fblock_set_corrupt(sc, sbcri->whichfork,
check_rec.rm_offset);
if (sc->sm->sm_flags & XFS_SCRUB_OFLAG_CORRUPT)
break;
check_rec.rm_startblock += irec.br_blockcount;
check_rec.rm_offset += irec.br_blockcount;
check_rec.rm_blockcount -= irec.br_blockcount;
if (check_rec.rm_blockcount == 0)
break;
have_map = xfs_iext_next_extent(ifp, &sbcri->icur, &irec);
if (!have_map)
xchk_fblock_set_corrupt(sc, sbcri->whichfork,
check_rec.rm_offset);
}
out:
if (sc->sm->sm_flags & XFS_SCRUB_OFLAG_CORRUPT)
return -ECANCELED;
return 0;
}
/* Make sure each rmap has a corresponding bmbt entry. */
STATIC int
xchk_bmap_check_ag_rmaps(
struct xfs_scrub *sc,
int whichfork,
struct xfs_perag *pag)
{
struct xchk_bmap_check_rmap_info sbcri;
struct xfs_btree_cur *cur;
struct xfs_buf *agf;
int error;
error = xfs_alloc_read_agf(pag, sc->tp, 0, &agf);
if (error)
return error;
cur = xfs_rmapbt_init_cursor(sc->mp, sc->tp, agf, pag);
sbcri.sc = sc;
sbcri.whichfork = whichfork;
error = xfs_rmap_query_all(cur, xchk_bmap_check_rmap, &sbcri);
if (error == -ECANCELED)
error = 0;
xfs_btree_del_cursor(cur, error);
xfs_trans_brelse(sc->tp, agf);
return error;
}
/* Make sure each rmap has a corresponding bmbt entry. */
STATIC int
xchk_bmap_check_rmaps(
struct xfs_scrub *sc,
int whichfork)
{
struct xfs_ifork *ifp = xfs_ifork_ptr(sc->ip, whichfork);
struct xfs_perag *pag;
xfs_agnumber_t agno;
bool zero_size;
int error;
if (!xfs_has_rmapbt(sc->mp) ||
whichfork == XFS_COW_FORK ||
(sc->sm->sm_flags & XFS_SCRUB_OFLAG_CORRUPT))
return 0;
/* Don't support realtime rmap checks yet. */
if (XFS_IS_REALTIME_INODE(sc->ip) && whichfork == XFS_DATA_FORK)
return 0;
ASSERT(xfs_ifork_ptr(sc->ip, whichfork) != NULL);
/*
* Only do this for complex maps that are in btree format, or for
* situations where we would seem to have a size but zero extents.
* The inode repair code can zap broken iforks, which means we have
* to flag this bmap as corrupt if there are rmaps that need to be
* reattached.
*/
if (whichfork == XFS_DATA_FORK)
zero_size = i_size_read(VFS_I(sc->ip)) == 0;
else
zero_size = false;
if (ifp->if_format != XFS_DINODE_FMT_BTREE &&
(zero_size || ifp->if_nextents > 0))
return 0;
for_each_perag(sc->mp, agno, pag) {
error = xchk_bmap_check_ag_rmaps(sc, whichfork, pag);
if (error ||
(sc->sm->sm_flags & XFS_SCRUB_OFLAG_CORRUPT)) {
xfs: active perag reference counting We need to be able to dynamically remove instantiated AGs from memory safely, either for shrinking the filesystem or paging AG state in and out of memory (e.g. supporting millions of AGs). This means we need to be able to safely exclude operations from accessing perags while dynamic removal is in progress. To do this, introduce the concept of active and passive references. Active references are required for high level operations that make use of an AG for a given operation (e.g. allocation) and pin the perag in memory for the duration of the operation that is operating on the perag (e.g. transaction scope). This means we can fail to get an active reference to an AG, hence callers of the new active reference API must be able to handle lookup failure gracefully. Passive references are used in low level code, where we might need to access the perag structure for the purposes of completing high level operations. For example, buffers need to use passive references because: - we need to be able to do metadata IO during operations like grow and shrink transactions where high level active references to the AG have already been blocked - buffers need to pin the perag until they are reclaimed from memory, something that high level code has no direct control over. - unused cached buffers should not prevent a shrink from being started. Hence we have active references that will form exclusion barriers for operations to be performed on an AG, and passive references that will prevent reclaim of the perag until all objects with passive references have been reclaimed themselves. This patch introduce xfs_perag_grab()/xfs_perag_rele() as the API for active AG reference functionality. We also need to convert the for_each_perag*() iterators to use active references, which will start the process of converting high level code over to using active references. Conversion of non-iterator based code to active references will be done in followup patches. Note that the implementation using reference counting is really just a development vehicle for the API to ensure we don't have any leaks in the callers. Once we need to remove perag structures from memory dyanmically, we will need a much more robust per-ag state transition mechanism for preventing new references from being taken while we wait for existing references to drain before removal from memory can occur.... Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Allison Henderson <allison.henderson@oracle.com> Reviewed-by: Darrick J. Wong <djwong@kernel.org>
2023-02-12 22:14:42 +00:00
xfs_perag_rele(pag);
return error;
}
}
return 0;
}
/* Scrub a delalloc reservation from the incore extent map tree. */
STATIC void
xchk_bmap_iextent_delalloc(
struct xfs_inode *ip,
struct xchk_bmap_info *info,
struct xfs_bmbt_irec *irec)
{
struct xfs_mount *mp = info->sc->mp;
/*
* Check for out-of-order extents. This record could have come
* from the incore list, for which there is no ordering check.
*/
if (irec->br_startoff < info->lastoff)
xchk_fblock_set_corrupt(info->sc, info->whichfork,
irec->br_startoff);
if (!xfs_verify_fileext(mp, irec->br_startoff, irec->br_blockcount))
xchk_fblock_set_corrupt(info->sc, info->whichfork,
irec->br_startoff);
/* Make sure the extent points to a valid place. */
if (irec->br_blockcount > XFS_MAX_BMBT_EXTLEN)
xchk_fblock_set_corrupt(info->sc, info->whichfork,
irec->br_startoff);
}
/*
* Scrub an inode fork's block mappings.
*
* First we scan every record in every btree block, if applicable.
* Then we unconditionally scan the incore extent cache.
*/
STATIC int
xchk_bmap(
struct xfs_scrub *sc,
int whichfork)
{
struct xfs_bmbt_irec irec;
struct xchk_bmap_info info = { NULL };
struct xfs_mount *mp = sc->mp;
struct xfs_inode *ip = sc->ip;
struct xfs_ifork *ifp = xfs_ifork_ptr(ip, whichfork);
xfs_fileoff_t endoff;
int error = 0;
/* Non-existent forks can be ignored. */
if (!ifp)
goto out;
info.is_rt = whichfork == XFS_DATA_FORK && XFS_IS_REALTIME_INODE(ip);
info.whichfork = whichfork;
info.is_shared = whichfork == XFS_DATA_FORK && xfs_is_reflink_inode(ip);
info.sc = sc;
switch (whichfork) {
case XFS_COW_FORK:
/* No CoW forks on non-reflink inodes/filesystems. */
if (!xfs_is_reflink_inode(ip)) {
xchk_ino_set_corrupt(sc, sc->ip->i_ino);
goto out;
}
break;
case XFS_ATTR_FORK:
if (!xfs_has_attr(mp) && !xfs_has_attr2(mp))
xchk_ino_set_corrupt(sc, sc->ip->i_ino);
break;
default:
ASSERT(whichfork == XFS_DATA_FORK);
break;
}
/* Check the fork values */
switch (ifp->if_format) {
case XFS_DINODE_FMT_UUID:
case XFS_DINODE_FMT_DEV:
case XFS_DINODE_FMT_LOCAL:
/* No mappings to check. */
if (whichfork == XFS_COW_FORK)
xchk_fblock_set_corrupt(sc, whichfork, 0);
goto out;
case XFS_DINODE_FMT_EXTENTS:
break;
case XFS_DINODE_FMT_BTREE:
if (whichfork == XFS_COW_FORK) {
xchk_fblock_set_corrupt(sc, whichfork, 0);
goto out;
}
error = xchk_bmap_btree(sc, whichfork, &info);
if (error)
goto out;
break;
default:
xchk_fblock_set_corrupt(sc, whichfork, 0);
goto out;
}
if (sc->sm->sm_flags & XFS_SCRUB_OFLAG_CORRUPT)
goto out;
/* Find the offset of the last extent in the mapping. */
error = xfs_bmap_last_offset(ip, &endoff, whichfork);
if (!xchk_fblock_process_error(sc, whichfork, 0, &error))
goto out;
/* Scrub extent records. */
info.lastoff = 0;
ifp = xfs_ifork_ptr(ip, whichfork);
for_each_xfs_iext(ifp, &info.icur, &irec) {
if (xchk_should_terminate(sc, &error) ||
(sc->sm->sm_flags & XFS_SCRUB_OFLAG_CORRUPT))
goto out;
if (irec.br_startoff >= endoff) {
xchk_fblock_set_corrupt(sc, whichfork,
irec.br_startoff);
goto out;
}
if (isnullstartblock(irec.br_startblock))
xchk_bmap_iextent_delalloc(ip, &info, &irec);
else
xchk_bmap_iextent(ip, &info, &irec);
info.lastoff = irec.br_startoff + irec.br_blockcount;
}
error = xchk_bmap_check_rmaps(sc, whichfork);
if (!xchk_fblock_xref_process_error(sc, whichfork, 0, &error))
goto out;
out:
return error;
}
/* Scrub an inode's data fork. */
int
xchk_bmap_data(
struct xfs_scrub *sc)
{
return xchk_bmap(sc, XFS_DATA_FORK);
}
/* Scrub an inode's attr fork. */
int
xchk_bmap_attr(
struct xfs_scrub *sc)
{
return xchk_bmap(sc, XFS_ATTR_FORK);
}
/* Scrub an inode's CoW fork. */
int
xchk_bmap_cow(
struct xfs_scrub *sc)
{
if (!xfs_is_reflink_inode(sc->ip))
return -ENOENT;
return xchk_bmap(sc, XFS_COW_FORK);
}