From: Kemeng Shi <shikemeng@huaweicloud.com>
To: Theodore Ts'o <tytso@mit.edu>
Cc: adilger.kernel@dilger.ca, ojaswin@linux.ibm.com,
linux-ext4@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH v4 13/19] ext4: call ext4_mb_mark_group_bb in ext4_free_blocks_simple
Date: Mon, 12 Jun 2023 10:24:55 +0800 [thread overview]
Message-ID: <cd236fdb-c48c-69b0-10a3-3df8a34f0a6e@huaweicloud.com> (raw)
In-Reply-To: <20230611050532.GE1436857@mit.edu>
on 6/11/2023 1:05 PM, Theodore Ts'o wrote:
> On Sat, Jun 03, 2023 at 11:03:21PM +0800, Kemeng Shi wrote:
>> call ext4_mb_mark_group_bb in ext4_free_blocks_simple to:
>> 1. remove repeat code
>> 2. pair update of free_clusters in ext4_mb_new_blocks_simple.
>> 3. add missing ext4_lock_group/ext4_unlock_group protection.
>>
>> Signed-off-by: Kemeng Shi <shikemeng@huaweicloud.com>
>> Reviewed-by: Ojaswin Mujoo <ojaswin@linux.ibm.com>
>
> Note: after bisecting, I've found that this commit is causing a OOPS
> when running "kvm-xfstests -c ext4/adv generic/468". It appears to be
> an issue with the fast commit feature not playing nice with this
> patch. The stack trace looks like this:
>
> [ 7.409663] ------------[ cut here ]------------
> [ 7.409969] WARNING: CPU: 0 PID: 3069 at fs/ext4/mballoc.c:3801 ext4_mb_mark_group_bb+0x48e/0x4a0
> [ 7.410480] CPU: 0 PID: 3069 Comm: mount Not tainted 6.4.0-rc5-xfstests-lockdep-00021-g60ba685c5998 #146
> [ 7.411067] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.16.2-debian-1.16.2-1 04/01/2014
> [ 7.411639] RIP: 0010:ext4_mb_mark_group_bb+0x48e/0x4a0
> [ 7.411968] Code: 48 c7 c7 35 b0 88 82 c6 05 16 f4 9b 01 01 e8 f9 16 c9 ff e9 7f fe ff ff 8b 45 08 c7 44 24 10 00 00 00 00 31 c9 e9 ef fc ff ff <0f> 0b e9 76 fc ff ff e8 96 64 b6 00 66 0f 1f 44 00 00 90 90 90 90
> [ 7.413128] RSP: 0018:ffffc90003b0f9f8 EFLAGS: 00010246
> [ 7.413458] RAX: 0000000000000003 RBX: 0000000000006002 RCX: 0000000000000001
> [ 7.413902] RDX: ffff88800965b000 RSI: 0000000000000000 RDI: ffff88800d690100
> [ 7.414346] RBP: ffffc90003b0fa68 R08: 000000000aebbd6e R09: 0000000000000246
> [ 7.414791] R10: 00000000d148c994 R11: 00000000941da2bb R12: ffff88800d7fd000
> [ 7.415234] R13: 0000000000000000 R14: ffff88800f3e4080 R15: ffff88800b5ca160
> [ 7.415724] FS: 00007f3d04516840(0000) GS:ffff88807da00000(0000) knlGS:0000000000000000
> [ 7.416227] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> [ 7.416588] CR2: 00007ffcb3979ac8 CR3: 000000000f290003 CR4: 0000000000770ef0
> [ 7.417032] PKRU: 55555554
> [ 7.417205] Call Trace:
> [ 7.417363] <TASK>
> [ 7.417502] ? ext4_mb_mark_group_bb+0x48e/0x4a0
> [ 7.417807] ? __warn+0x80/0x170
> [ 7.418051] ? ext4_mb_mark_group_bb+0x48e/0x4a0
> [ 7.418337] ? report_bug+0x173/0x1d0
> [ 7.418567] ? handle_bug+0x3c/0x70
> [ 7.418797] ? exc_invalid_op+0x17/0x70
> [ 7.419037] ? asm_exc_invalid_op+0x1a/0x20
> [ 7.419226] ? ext4_mb_mark_group_bb+0x48e/0x4a0
> [ 7.419437] ? ext4_mb_mark_group_bb+0xae/0x4a0
> [ 7.419708] ext4_mb_mark_bb+0xc0/0x120
> [ 7.419946] ext4_ext_clear_bb+0x210/0x280
> [ 7.420198] ext4_fc_replay_inode+0xa1/0x380
> [ 7.420466] ext4_fc_replay+0x435/0x880
> [ 7.420703] ? __getblk_gfp+0x37/0x110
> [ 7.420938] ? jread+0x7a/0x180
> [ 7.421138] do_one_pass+0x7df/0x1040
> [ 7.421365] jbd2_journal_recover+0x150/0x250
> [ 7.421637] jbd2_journal_load+0xbe/0x190
> [ 7.421886] ext4_load_journal+0x214/0x610
> [ 7.422152] ext4_load_and_init_journal+0x29/0x380
> [ 7.422490] __ext4_fill_super+0x15ca/0x15e0
> [ 7.422756] ? __pfx_ext4_fill_super+0x10/0x10
> [ 7.423032] ext4_fill_super+0xcf/0x280
> [ 7.423270] get_tree_bdev+0x188/0x290
> [ 7.423505] vfs_get_tree+0x29/0xe0
> [ 7.423723] ? capable+0x37/0x70
> [ 7.423927] do_new_mount+0x174/0x300
> [ 7.424157] __x64_sys_mount+0x11a/0x150
> [ 7.424401] do_syscall_64+0x3b/0x90
> [ 7.424624] entry_SYSCALL_64_after_hwframe+0x72/0xdc
> [ 7.424935] RIP: 0033:0x7f3d0475562a
> [ 7.425160] Code: 48 8b 0d 69 18 0d 00 f7 d8 64 89 01 48 83 c8 ff c3 66 2e 0f 1f 84 00 00 00 00 00 0f 1f 44 00 00 49 89 ca b8 a5 00 00 00 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 8b 0d 36 18 0d 00 f7 d8 64 89 01 48
> [ 7.426298] RSP: 002b:00007ffcb397aaf8 EFLAGS: 00000246 ORIG_RAX: 00000000000000a5
> [ 7.426761] RAX: ffffffffffffffda RBX: 00007f3d04889264 RCX: 00007f3d0475562a
> [ 7.427197] RDX: 0000558ea381db90 RSI: 0000558ea381dbb0 RDI: 0000558ea381dbd0
> [ 7.427631] RBP: 0000558ea381d960 R08: 0000558ea381dbf0 R09: 00007f3d04827be0
> [ 7.428063] R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000000000
> [ 7.428499] R13: 0000558ea381dbd0 R14: 0000558ea381db90 R15: 0000558ea381d960
> [ 7.428941] </TASK>
> [ 7.429083] irq event stamp: 10951
> [ 7.429296] hardirqs last enabled at (10959): [<ffffffff811643c2>] __up_console_sem+0x52/0x60
> [ 7.429824] hardirqs last disabled at (10966): [<ffffffff811643a7>] __up_console_sem+0x37/0x60
> [ 7.430325] softirqs last enabled at (10574): [<ffffffff8204a529>] __do_softirq+0x2d9/0x39e
> [ 7.430839] softirqs last disabled at (10407): [<ffffffff810dcc57>] __irq_exit_rcu+0x87/0xb0
> [ 7.431354] ---[ end trace 0000000000000000 ]---
>
>
Hi ted, sorry for this issue. This patch added a WARN_ON for case that we free block
to uninitialized block group which should be invalid.
We can simply remove the WARN_ON to allow free on uninitialized block group as old
way for emergency fix and I will find out why we free blocks to uninitialized block
group in fast commit code path and is it a valid behavior.
--
Best wishes
Kemeng Shi
next prev parent reply other threads:[~2023-06-12 2:25 UTC|newest]
Thread overview: 29+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-06-03 15:03 [PATCH v4 00/19] Fixes, cleanups and unit test for mballoc Kemeng Shi
2023-06-03 15:03 ` [PATCH v4 01/19] ext4: fix wrong unit use in ext4_mb_normalize_request Kemeng Shi
2023-06-03 15:03 ` [PATCH v4 02/19] ext4: fix unit mismatch in ext4_mb_new_blocks_simple Kemeng Shi
2023-06-03 15:03 ` [PATCH v4 03/19] ext4: fix wrong unit use in ext4_mb_find_by_goal Kemeng Shi
2023-06-03 15:03 ` [PATCH v4 04/19] ext4: treat stripe in block unit Kemeng Shi
2023-06-03 15:03 ` [PATCH v4 05/19] ext4: add EXT4_MB_HINT_GOAL_ONLY test in ext4_mb_use_preallocated Kemeng Shi
2023-06-03 15:03 ` [PATCH v4 06/19] ext4: remove ext4_block_group and ext4_block_group_offset declaration Kemeng Shi
2023-06-03 15:03 ` [PATCH v4 07/19] ext4: try all groups in ext4_mb_new_blocks_simple Kemeng Shi
2023-06-03 15:03 ` [PATCH v4 08/19] ext4: get block from bh before pass it to ext4_free_blocks_simple in ext4_free_blocks Kemeng Shi
2023-06-03 15:03 ` [PATCH v4 09/19] ext4: remove unsed parameter and unnecessary forward declaration of ext4_mb_new_blocks_simple Kemeng Shi
2023-06-03 15:03 ` [PATCH v4 10/19] ext4: fix wrong unit use in ext4_mb_clear_bb Kemeng Shi
2023-06-03 15:03 ` [PATCH v4 11/19] ext4: fix wrong unit use in ext4_mb_new_blocks Kemeng Shi
2023-06-03 15:03 ` [PATCH v4 12/19] ext4: factor out codes to update block bitmap and group descriptor on disk from ext4_mb_mark_bb Kemeng Shi
2023-06-03 15:03 ` [PATCH v4 13/19] ext4: call ext4_mb_mark_group_bb in ext4_free_blocks_simple Kemeng Shi
2023-06-11 5:05 ` Theodore Ts'o
2023-06-12 2:24 ` Kemeng Shi [this message]
2023-06-12 3:49 ` Theodore Ts'o
2023-06-13 1:22 ` Kemeng Shi
2023-06-20 1:50 ` Kemeng Shi
2023-06-03 15:03 ` [PATCH v4 14/19] ext4: extent ext4_mb_mark_group_bb to support allocation under journal Kemeng Shi
2023-06-03 15:03 ` [PATCH v4 15/19] ext4: call ext4_mb_mark_group_bb in ext4_mb_mark_diskspace_used Kemeng Shi
2023-06-03 15:03 ` [PATCH v4 16/19] ext4: call ext4_mb_mark_group_bb in ext4_mb_clear_bb Kemeng Shi
2023-06-06 9:39 ` Ojaswin Mujoo
2023-06-06 14:16 ` Kemeng Shi
2023-06-08 6:16 ` Ojaswin Mujoo
2023-06-03 15:03 ` [PATCH v4 17/19] ext4: call ext4_mb_mark_group_bb in ext4_group_add_blocks Kemeng Shi
2023-06-03 15:03 ` [PATCH v4 18/19] ext4: add some kunit stub for mballoc kunit test Kemeng Shi
2023-06-03 15:03 ` [PATCH v4 19/19] ext4: add first unit test for ext4_mb_new_blocks_simple in mballoc Kemeng Shi
2023-06-09 3:14 ` [PATCH v4 00/19] Fixes, cleanups and unit test for mballoc Theodore Ts'o
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=cd236fdb-c48c-69b0-10a3-3df8a34f0a6e@huaweicloud.com \
--to=shikemeng@huaweicloud.com \
--cc=adilger.kernel@dilger.ca \
--cc=linux-ext4@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=ojaswin@linux.ibm.com \
--cc=tytso@mit.edu \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).