mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Chao Yu <chao@kernel.org>
To: Jaegeuk Kim <jaegeuk@kernel.org>,
	linux-kernel@vger.kernel.org,
	linux-f2fs-devel@lists.sourceforge.net
Subject: Re: [f2fs-dev] [PATCH 1/2] f2fs: evict inode cache for frozen fs
Date: Fri, 11 Mar 2022 11:10:50 +0800	[thread overview]
Message-ID: <a475c925-cfe4-abf7-f909-0d883df391a2@kernel.org> (raw)
In-Reply-To: <2a5564bb-463e-d163-0042-34fad42e2e46@kernel.org>

Jaegeuk,

Could you please send v2 patch to mailing list? as I saw the revised
one has been merged in dev branch.

Otherwise, I've no idea where I should reply "Reviewed-by" tag to...

Thanks,

On 2022/3/10 9:53, Chao Yu wrote:
> On 2022/3/10 5:48, Jaegeuk Kim wrote:
>> Let's purge inode cache in order to avoid the below deadlock.
>>
>> [freeze test]                         shrinkder
>> freeze_super
>>   - pwercpu_down_write(SB_FREEZE_FS)
>>                                         - super_cache_scan
>>                                           - down_read(&sb->s_umount)
>>                                             - prune_icache_sb
>>                                              - dispose_list
>>                                               - evict
>>                                                - f2fs_evict_inode
>> thaw_super
>>   - down_write(&sb->s_umount);
>>                                                - __percpu_down_read(SB_FREEZE_FS)
> 
> Ah, finally we catch this. :)
> 
>>
>> Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
>> ---
>>   fs/f2fs/debug.c | 1 +
>>   fs/f2fs/f2fs.h  | 1 +
>>   fs/f2fs/inode.c | 6 ++++--
>>   fs/f2fs/super.c | 4 ++++
>>   4 files changed, 10 insertions(+), 2 deletions(-)
>>
>> diff --git a/fs/f2fs/debug.c b/fs/f2fs/debug.c
>> index 9a13902c7702..cba5eab24595 100644
>> --- a/fs/f2fs/debug.c
>> +++ b/fs/f2fs/debug.c
>> @@ -338,6 +338,7 @@ static char *s_flag[] = {
>>       [SBI_QUOTA_SKIP_FLUSH]    = " quota_skip_flush",
>>       [SBI_QUOTA_NEED_REPAIR]    = " quota_need_repair",
>>       [SBI_IS_RESIZEFS]    = " resizefs",
>> +    [SBI_IS_FREEZING]    = " freezefs",
> 
> Could you please update description of "sb_status" entry in
> Documentation/ABI/testing/sysfs-fs-f2fs as well?
> 
> Thanks,
> 
>>   };
>>   static int stat_show(struct seq_file *s, void *v)
>> diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h
>> index 68d791ec8b27..da729f53daa8 100644
>> --- a/fs/f2fs/f2fs.h
>> +++ b/fs/f2fs/f2fs.h
>> @@ -1293,6 +1293,7 @@ enum {
>>       SBI_QUOTA_SKIP_FLUSH,            /* skip flushing quota in current CP */
>>       SBI_QUOTA_NEED_REPAIR,            /* quota file may be corrupted */
>>       SBI_IS_RESIZEFS,            /* resizefs is in process */
>> +    SBI_IS_FREEZING,            /* freezefs is in process */
>>   };
>>   enum {
>> diff --git a/fs/f2fs/inode.c b/fs/f2fs/inode.c
>> index ab8e0c06c78c..71f232dcf3c2 100644
>> --- a/fs/f2fs/inode.c
>> +++ b/fs/f2fs/inode.c
>> @@ -778,7 +778,8 @@ void f2fs_evict_inode(struct inode *inode)
>>       f2fs_remove_ino_entry(sbi, inode->i_ino, UPDATE_INO);
>>       f2fs_remove_ino_entry(sbi, inode->i_ino, FLUSH_INO);
>> -    sb_start_intwrite(inode->i_sb);
>> +    if (!is_sbi_flag_set(sbi, SBI_IS_FREEZING))
>> +        sb_start_intwrite(inode->i_sb);
>>       set_inode_flag(inode, FI_NO_ALLOC);
>>       i_size_write(inode, 0);
>>   retry:
>> @@ -809,7 +810,8 @@ void f2fs_evict_inode(struct inode *inode)
>>           if (dquot_initialize_needed(inode))
>>               set_sbi_flag(sbi, SBI_QUOTA_NEED_REPAIR);
>>       }
>> -    sb_end_intwrite(inode->i_sb);
>> +    if (!is_sbi_flag_set(sbi, SBI_IS_FREEZING))
>> +        sb_end_intwrite(inode->i_sb);
>>   no_delete:
>>       dquot_drop(inode);
>> diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c
>> index 8e3840973077..4b570b5c2674 100644
>> --- a/fs/f2fs/super.c
>> +++ b/fs/f2fs/super.c
>> @@ -1663,11 +1663,15 @@ static int f2fs_freeze(struct super_block *sb)
>>       /* ensure no checkpoint required */
>>       if (!llist_empty(&F2FS_SB(sb)->cprc_info.issue_list))
>>           return -EINVAL;
>> +
>> +    /* to avoid deadlock on f2fs_evict_inode->SB_FREEZE_FS */
>> +    set_sbi_flag(F2FS_SB(sb), SBI_IS_FREEZING);
>>       return 0;
>>   }
>>   static int f2fs_unfreeze(struct super_block *sb)
>>   {
>> +    clear_sbi_flag(F2FS_SB(sb), SBI_IS_FREEZING);
>>       return 0;
>>   }
> 
> 
> _______________________________________________
> Linux-f2fs-devel mailing list
> Linux-f2fs-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/linux-f2fs-devel

  reply	other threads:[~2022-03-11  3:11 UTC|newest]

Thread overview: 11+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-03-09 21:48 Jaegeuk Kim
2022-03-09 21:48 ` [PATCH 2/2] f2fs: use spin_lock to avoid hang Jaegeuk Kim
2022-03-10  2:23   ` [f2fs-dev] " Chao Yu
2022-03-10  2:51     ` Jaegeuk Kim
2022-03-16  2:33   ` Chao Yu
2022-03-10  1:53 ` [f2fs-dev] [PATCH 1/2] f2fs: evict inode cache for frozen fs Chao Yu
2022-03-11  3:10   ` Chao Yu [this message]
2022-03-11  5:22     ` Jaegeuk Kim
2022-03-10  2:52 ` [PATCH 1/2 v2] f2fs: do not expose unwritten blocks to user by DIO Jaegeuk Kim
2022-03-11  5:22 ` [PATCH 1/2 v2] f2fs: don't get FREEZE lock in f2fs_evict_inode in frozen fs Jaegeuk Kim
2022-03-11  7:25   ` [f2fs-dev] " Chao Yu

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=a475c925-cfe4-abf7-f909-0d883df391a2@kernel.org \
    --to=chao@kernel.org \
    --cc=jaegeuk@kernel.org \
    --cc=linux-f2fs-devel@lists.sourceforge.net \
    --cc=linux-kernel@vger.kernel.org \
    /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

all inboxes | Powered by JetHome®