linux-stable.git/fs/ext4, branch v5.17.2

ext4: don't BUG if someone dirty pages without asking ext4 first

2022-04-08T11:58:39+00:00

[ Upstream commit cc5095747edfb054ca2068d01af20be3fcc3634f ]

[un]pin_user_pages_remote is dirtying pages without properly warning
the file system in advance.  A related race was noted by Jan Kara in
2018[1]; however, more recently instead of it being a very hard-to-hit
race, it could be reliably triggered by process_vm_writev(2) which was
discovered by Syzbot[2].

This is technically a bug in mm/gup.c, but arguably ext4 is fragile in
that if some other kernel subsystem dirty pages without properly
notifying the file system using page_mkwrite(), ext4 will BUG, while
other file systems will not BUG (although data will still be lost).

So instead of crashing with a BUG, issue a warning (since there may be
potential data loss) and just mark the page as clean to avoid
unprivileged denial of service attacks until the problem can be
properly fixed.  More discussion and background can be found in the
thread starting at [2].

[1] https://lore.kernel.org/linux-mm/20180103100430.GE4911@quack2.suse.cz
[2] https://lore.kernel.org/r/Yg0m6IjcNmfaSokM@google.com

Reported-by: syzbot+d59332e2db681cf18f0318a06e994ebbb529a8db@syzkaller.appspotmail.com
Reported-by: Lee Jones 
Signed-off-by: Theodore Ts'o 
Link: https://lore.kernel.org/r/YiDS9wVfq4mM2jGK@mit.edu
Signed-off-by: Sasha Levin

ext4: fix ext4_mb_mark_bb() with flex_bg with fast_commit

2022-04-08T11:58:39+00:00

[ Upstream commit bfdc502a4a4c058bf4cbb1df0c297761d528f54d ]

In case of flex_bg feature (which is by default enabled), extents for
any given inode might span across blocks from two different block group.
ext4_mb_mark_bb() only reads the buffer_head of block bitmap once for the
starting block group, but it fails to read it again when the extent length
boundary overflows to another block group. Then in this below loop it
accesses memory beyond the block group bitmap buffer_head and results
into a data abort.

	for (i = 0; i < clen; i++)
		if (!mb_test_bit(blkoff + i, bitmap_bh->b_data) == !state)
			already++;

This patch adds this functionality for checking block group boundary in
ext4_mb_mark_bb() and update the buffer_head(bitmap_bh) for every different
block group.

w/o this patch, I was easily able to hit a data access abort using Power platform.

<...>
[   74.327662] EXT4-fs error (device loop3): ext4_mb_generate_buddy:1141: group 11, block bitmap and bg descriptor inconsistent: 21248 vs 23294 free clusters
[   74.533214] EXT4-fs (loop3): shut down requested (2)
[   74.536705] Aborting journal on device loop3-8.
[   74.702705] BUG: Unable to handle kernel data access on read at 0xc00000005e980000
[   74.703727] Faulting instruction address: 0xc0000000007bffb8
cpu 0xd: Vector: 300 (Data Access) at [c000000015db7060]
    pc: c0000000007bffb8: ext4_mb_mark_bb+0x198/0x5a0
    lr: c0000000007bfeec: ext4_mb_mark_bb+0xcc/0x5a0
    sp: c000000015db7300
   msr: 800000000280b033
   dar: c00000005e980000
 dsisr: 40000000
  current = 0xc000000027af6880
  paca    = 0xc00000003ffd5200   irqmask: 0x03   irq_happened: 0x01
    pid   = 5167, comm = mount
<...>
enter ? for help
[c000000015db7380] c000000000782708 ext4_ext_clear_bb+0x378/0x410
[c000000015db7400] c000000000813f14 ext4_fc_replay+0x1794/0x2000
[c000000015db7580] c000000000833f7c do_one_pass+0xe9c/0x12a0
[c000000015db7710] c000000000834504 jbd2_journal_recover+0x184/0x2d0
[c000000015db77c0] c000000000841398 jbd2_journal_load+0x188/0x4a0
[c000000015db7880] c000000000804de8 ext4_fill_super+0x2638/0x3e10
[c000000015db7a40] c0000000005f8404 get_tree_bdev+0x2b4/0x350
[c000000015db7ae0] c0000000007ef058 ext4_get_tree+0x28/0x40
[c000000015db7b00] c0000000005f6344 vfs_get_tree+0x44/0x100
[c000000015db7b70] c00000000063c408 path_mount+0xdd8/0xe70
[c000000015db7c40] c00000000063c8f0 sys_mount+0x450/0x550
[c000000015db7d50] c000000000035770 system_call_exception+0x4a0/0x4e0
[c000000015db7e10] c00000000000c74c system_call_common+0xec/0x250

Signed-off-by: Ritesh Harjani 
Reviewed-by: Jan Kara 
Link: https://lore.kernel.org/r/2609bc8f66fc15870616ee416a18a3d392a209c4.1644992609.git.riteshh@linux.ibm.com
Signed-off-by: Theodore Ts'o 
Signed-off-by: Sasha Levin

ext4: correct cluster len and clusters changed accounting in ext4_mb_mark_bb

2022-04-08T11:58:39+00:00

[ Upstream commit a5c0e2fdf7cea535ba03259894dc184e5a4c2800 ]

ext4_mb_mark_bb() currently wrongly calculates cluster len (clen) and
flex_group->free_clusters. This patch fixes that.

Identified based on code review of ext4_mb_mark_bb() function.

Signed-off-by: Ritesh Harjani 
Reviewed-by: Jan Kara 
Link: https://lore.kernel.org/r/a0b035d536bafa88110b74456853774b64c8ac40.1644992609.git.riteshh@linux.ibm.com
Signed-off-by: Theodore Ts'o 
Signed-off-by: Sasha Levin

ext4: fix remount with 'abort' option

2022-04-08T11:57:40+00:00

[ Upstream commit e3952fcce1aad934f1322843b564ff86256444b2 ]

After commit 6e47a3cc68fc ("ext4: get rid of super block and sbi from
handle_mount_ops()") the 'abort' options stopped working. This is
because we're using ctx_set_mount_flags() helper that's expecting an
argument with the appropriate bit set, but instead got
EXT4_MF_FS_ABORTED which is a bit position. ext4_set_mount_flag() is
using set_bit() while ctx_set_mount_flags() was using bitwise OR.

Create a separate helper ctx_set_mount_flag() to handle setting the
mount_flags correctly.

While we're at it clean up the EXT4_SET_CTX macros so that we're only
creating helpers that we actually use to avoid warnings.

Fixes: 6e47a3cc68fc ("ext4: get rid of super block and sbi from handle_mount_ops()")
Signed-off-by: Lukas Czerner 
Cc: Ye Bin 
Reviewed-by: Eric Sandeen 
Tested-by: Gabriel Krisman Bertazi 
Link: https://lore.kernel.org/r/20220201131345.77591-1-lczerner@redhat.com
Signed-off-by: Theodore Ts'o 
Signed-off-by: Sasha Levin

ext4: make mb_optimize_scan performance mount option work with extents

2022-04-08T11:57:25+00:00

commit 077d0c2c78df6f7260cdd015a991327efa44d8ad upstream.

Currently mb_optimize_scan scan feature which improves filesystem
performance heavily (when FS is fragmented), seems to be not working
with files with extents (ext4 by default has files with extents).

This patch fixes that and makes mb_optimize_scan feature work
for files with extents.

Below are some performance numbers obtained when allocating a 10M and 100M
file with and w/o this patch on a filesytem with no 1M contiguous block.


===============
Workload: dd if=/dev/urandom of=test conv=fsync bs=1M count=10/100

Time taken
=====================================================
no.     Size   without-patch     with-patch    Diff(%)
1       10M      0m8.401s         0m5.623s     33.06%
2       100M     1m40.465s        1m14.737s    25.6%


=============
w/o patch:
  mballoc:
    reqs: 17056
    success: 11407
    groups_scanned: 13643
    cr0_stats:
            hits: 37
            groups_considered: 9472
            useless_loops: 36
            bad_suggestions: 0
    cr1_stats:
            hits: 11418
            groups_considered: 908560
            useless_loops: 1894
            bad_suggestions: 0
    cr2_stats:
            hits: 1873
            groups_considered: 6913
            useless_loops: 21
    cr3_stats:
            hits: 21
            groups_considered: 5040
            useless_loops: 21
    extents_scanned: 417364
            goal_hits: 3707
            2^n_hits: 37
            breaks: 1873
            lost: 0
    buddies_generated: 239/240
    buddies_time_used: 651080
    preallocated: 705
    discarded: 478

with patch:
  mballoc:
    reqs: 12768
    success: 11305
    groups_scanned: 12768
    cr0_stats:
            hits: 1
            groups_considered: 18
            useless_loops: 0
            bad_suggestions: 0
    cr1_stats:
            hits: 5829
            groups_considered: 50626
            useless_loops: 0
            bad_suggestions: 0
    cr2_stats:
            hits: 6938
            groups_considered: 580363
            useless_loops: 0
    cr3_stats:
            hits: 0
            groups_considered: 0
            useless_loops: 0
    extents_scanned: 309059
            goal_hits: 0
            2^n_hits: 1
            breaks: 1463
            lost: 0
    buddies_generated: 239/240
    buddies_time_used: 791392
    preallocated: 673
    discarded: 446

Fixes: 196e402 (ext4: improve cr 0 / cr 1 group scanning)
Cc: stable@kernel.org
Reported-by: Geetika Moolchandani 
Reported-by: Nageswara R Sastry 
Suggested-by: Ritesh Harjani 
Signed-off-by: Ojaswin Mujoo 
Link: https://lore.kernel.org/r/fc9a48f7f8dcfc83891a8b21f6dd8cdf056ed810.1646732698.git.ojaswin@linux.ibm.com
Signed-off-by: Theodore Ts'o 
Signed-off-by: Greg Kroah-Hartman

ext4: make mb_optimize_scan option work with set/unset mount cmd

2022-04-08T11:57:25+00:00

commit 27b38686a3bb601db48901dbc4e2fc5d77ffa2c1 upstream.

After moving to the new mount API, mb_optimize_scan mount option
handling was not working as expected due to the parsed value always
being overwritten by default. Refactor and fix this to the expected
behavior described below:

*  mb_optimize_scan=1 - On
*  mb_optimize_scan=0 - Off
*  mb_optimize_scan not passed - On if no. of BGs > threshold else off
*  Remounts retain previous value unless we explicitly pass the option
   with a new value

Fixes: cebe85d570cf ("ext4: switch to the new mount api")
Cc: stable@kernel.org
Reported-by: Ritesh Harjani 
Signed-off-by: Ojaswin Mujoo 
Reviewed-by: Ritesh Harjani 
Link: https://lore.kernel.org/r/c98970fe99f26718586d02e942f293300fb48ef3.1646732698.git.ojaswin@linux.ibm.com
Signed-off-by: Theodore Ts'o 
Signed-off-by: Greg Kroah-Hartman

ext4: fix fs corruption when tring to remove a non-empty directory with IO error

2022-04-08T11:57:25+00:00

commit 7aab5c84a0f6ec2290e2ba4a6b245178b1bf949a upstream.

We inject IO error when rmdir non empty direcory, then got issue as follows:
step1: mkfs.ext4 -F /dev/sda
step2: mount /dev/sda  test
step3: cd test
step4: mkdir -p 1/2
step5: rmdir 1
	[  110.920551] ext4_empty_dir: inject fault
	[  110.921926] EXT4-fs warning (device sda): ext4_rmdir:3113: inode #12:
	comm rmdir: empty directory '1' has too many links (3)
step6: cd ..
step7: umount test
step8: fsck.ext4 -f /dev/sda
	e2fsck 1.42.9 (28-Dec-2013)
	Pass 1: Checking inodes, blocks, and sizes
	Pass 2: Checking directory structure
	Entry '..' in .../??? (13) has deleted/unused inode 12.  Clear? yes
	Pass 3: Checking directory connectivity
	Unconnected directory inode 13 (...)
	Connect to /lost+found? yes
	Pass 4: Checking reference counts
	Inode 13 ref count is 3, should be 2.  Fix? yes
	Pass 5: Checking group summary information

	/dev/sda: ***** FILE SYSTEM WAS MODIFIED *****
	/dev/sda: 12/131072 files (0.0% non-contiguous), 26157/524288 blocks

ext4_rmdir
	if (!ext4_empty_dir(inode))
		goto end_rmdir;
ext4_empty_dir
	bh = ext4_read_dirblock(inode, 0, DIRENT_HTREE);
	if (IS_ERR(bh))
		return true;
Now if read directory block failed, 'ext4_empty_dir' will return true, assume
directory is empty. Obviously, it will lead to above issue.
To solve this issue, if read directory block failed 'ext4_empty_dir' just
return false. To avoid making things worse when file system is already
corrupted, 'ext4_empty_dir' also return false.

Signed-off-by: Ye Bin 
Cc: stable@kernel.org
Link: https://lore.kernel.org/r/20220228024815.3952506-1-yebin10@huawei.com
Signed-off-by: Theodore Ts'o 
Signed-off-by: Greg Kroah-Hartman

Merge tag 'ext4_for_linus_stable' of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4

2022-02-06T18:34:45+00:00

Pull ext4 fixes from Ted Ts'o:
 "Various bug fixes for ext4 fast commit and inline data handling.

  Also fix regression introduced as part of moving to the new mount API"

* tag 'ext4_for_linus_stable' of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4:
  fs/ext4: fix comments mentioning i_mutex
  ext4: fix incorrect type issue during replay_del_range
  jbd2: fix kernel-doc descriptions for jbd2_journal_shrink_{scan,count}()
  ext4: fix potential NULL pointer dereference in ext4_fill_super()
  jbd2: refactor wait logic for transaction updates into a common function
  jbd2: cleanup unused functions declarations from jbd2.h
  ext4: fix error handling in ext4_fc_record_modified_inode()
  ext4: remove redundant max inline_size check in ext4_da_write_inline_data_begin()
  ext4: fix error handling in ext4_restore_inline_data()
  ext4: fast commit may miss file actions
  ext4: fast commit may not fallback for ineligible commit
  ext4: modify the logic of ext4_mb_new_blocks_simple
  ext4: prevent used blocks from being allocated during fast commit replay

fs/ext4: fix comments mentioning i_mutex

2022-02-03T15:57:53+00:00

inode->i_mutex has been replaced with inode->i_rwsem long ago. Fix
comments still mentioning i_mutex.

Signed-off-by: hongnanli 
Link: https://lore.kernel.org/r/20220121070611.21618-1-hongnan.li@linux.alibaba.com
Signed-off-by: Theodore Ts'o

ext4: fix incorrect type issue during replay_del_range

2022-02-03T15:57:53+00:00

should not use fast commit log data directly, add le32_to_cpu().

Reported-by: kernel test robot 
Fixes: 0b5b5a62b945 ("ext4: use ext4_ext_remove_space() for fast commit replay delete range")
Cc: stable@kernel.org
Signed-off-by: Xin Yin 
Reviewed-by: Ritesh Harjani 
Link: https://lore.kernel.org/r/20220126063146.2302-1-yinxin.x@bytedance.com
Signed-off-by: Theodore Ts'o