reiserfs-devel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Edward Shishkin <edward.shishkin@gmail.com>
To: Mike Fleetwood <mike.fleetwood@googlemail.com>
Cc: reiserfs-devel@vger.kernel.org
Subject: Re: Occasionally mkfs.reiser4 creates a file system without a UUID
Date: Tue, 16 Mar 2021 00:23:05 +0100	[thread overview]
Message-ID: <dd354e65-84ed-02e3-9800-04ff65937de5@gmail.com> (raw)
In-Reply-To: <CAMU1PDi9xiePgzDpd1cznGaBEGoSJjKafvBxQStaYM=8HLKtSQ@mail.gmail.com>



On 03/15/2021 11:20 PM, Mike Fleetwood wrote:
> On Sat, 13 Mar 2021 at 14:28, Edward Shishkin <edward.shishkin@gmail.com> wrote:
>>
>> On 03/13/2021 12:37 PM, Edward Shishkin wrote:
>>> On 03/12/2021 08:36 PM, Mike Fleetwood wrote:
>>>> Hi,
>>>>
>>>> Occasionally mkfs.reiser4 creates a file system without a UUID,
>>>> debugfs.reiser4 reports UUID as '<none>'.  This is being detected in
>>>> GParted's file system interface CI tests [1] which exercises the resier4
>>>> commands.
>>>>
>>>> In a Ubuntu 20.04 LTS VM with both resier4progs-1.2.1-1 package and
>>>> current code from GIT, the fault can easily be replicated like this:
>>>>
>>>> truncate -s 256M test.img
>>>> i=1
>>>> while :
>>>> do
>>>>     mkfs.reiser4 --force --yes --label '' test.img
>>>>     line=`debugfs.reiser4 test.img 2> /dev/null | egrep '^uuid:'`
>>>>     echo "[$i] $line"
>>>>     ((i++))
>>>>     echo "$line" | grep -q '<none>' && break
>>>> done
>>>>
>>>> Output fragment:
>>>> [1] uuid:        17073919-e41d-4892-9b22-4294d1544c4a
>>>> [2] uuid:        af2821de-ea85-4f20-9621-4fbd128b3fb8
>>>> [3] uuid:        c0fb805b-e224-4695-a504-d87460d158ae
>>>> ...
>>>> [35] uuid:        d604794d-097f-4810-bbb3-01a1518f3ef1
>>>> [36] uuid:        9634100c-1f98-42b3-a684-c9df77ab54e2
>>>> [37] uuid:        <none>
>>>>
>>>> Performing the same test with resierfs and ext4 runs for many hours
>>>> doing 100,000s of iterations without issue.
>>>>
>>>
>>>
>>> It seems that debugfs.reiser4 prints uuid incorrectly: it assumes that
>>> binary uuid can't contain '\0' symbols, which apparently is wrong.
>>> Thus, if the first uuid's symbol is '\0', debugfs thinks that uuid is
>>> not set and prints "none" ;)
>>>
>>> I'll try to prepare a fixup a bit later.
>>> Special thanks for the nice testing script!
>>>
>>
>>
>> Here is the fixup:
>> https://github.com/edward6/reiser4progs/commit/4802cdb18ae03031d0e51a58b6655f3b99021ec2
> 
> Your fix to debugfs.reiser4 allows for printing of UUIDs correctly.
> Here's my pull request to fix mkfs.reiser4 for not writing UUIDs
> correctly in the first place.
> https://github.com/edward6/reiser4progs/pull/1


Mike, thanks a lot!

For me personally, this is the worst kind of bugs (hardly fishable by
code reviews plus not always reproducible).

I cherry pick-ed your commit also for the parallel branch "format41":
https://github.com/edward6/reiser4progs/commit/01a5e0e41ec8d5147091d44e902d60f61c336f30

Edward.

      reply	other threads:[~2021-03-15 23:23 UTC|newest]

Thread overview: 6+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-03-12 19:36 Occasionally mkfs.reiser4 creates a file system without a UUID Mike Fleetwood
2021-03-12 19:48 ` Edward Shishkin
2021-03-13 11:37 ` Edward Shishkin
2021-03-13 14:28   ` Edward Shishkin
2021-03-15 22:20     ` Mike Fleetwood
2021-03-15 23:23       ` Edward Shishkin [this message]

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=dd354e65-84ed-02e3-9800-04ff65937de5@gmail.com \
    --to=edward.shishkin@gmail.com \
    --cc=mike.fleetwood@googlemail.com \
    --cc=reiserfs-devel@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;
as well as URLs for NNTP newsgroup(s).