Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > linux.debian.kernel > #76173 > unrolled thread

Bug#1017915: btrfs: space cache corruption and potential double allocations

Started byChristoph Anton Mitterer <calestyo@scientia.org>
First post2022-08-22 15:10 +0200
Last post2022-09-02 09:20 +0200
Articles 6 — 3 participants

Back to article view | Back to linux.debian.kernel


Contents

  Bug#1017915: btrfs: space cache corruption and potential double allocations Christoph Anton Mitterer <calestyo@scientia.org> - 2022-08-22 15:10 +0200
    Bug#1017915: btrfs: space cache corruption and potential double allocations Christoph Anton Mitterer <calestyo@scientia.org> - 2022-09-01 15:50 +0200
    Processed: RE: btrfs: space cache corruption and potential double  allocations "Debian Bug Tracking System" <owner@bugs.debian.org> - 2022-09-01 15:50 +0200
    Bug#1017915: marked as done (btrfs: space cache corruption and  potential double allocations) "Debian Bug Tracking System" <owner@bugs.debian.org> - 2022-09-01 16:50 +0200
    Bug#1017915: btrfs: space cache corruption and potential double allocations Christoph Anton Mitterer <calestyo@scientia.org> - 2022-09-01 21:00 +0200
      Bug#1017915: btrfs: space cache corruption and potential double allocations Salvatore Bonaccorso <carnil@debian.org> - 2022-09-02 09:20 +0200

#76173 — Bug#1017915: btrfs: space cache corruption and potential double allocations

FromChristoph Anton Mitterer <calestyo@scientia.org>
Date2022-08-22 15:10 +0200
SubjectBug#1017915: btrfs: space cache corruption and potential double allocations
Message-ID<EXflM-2iDp-17@gated-at.bofh.it>
Source: linux
Version: 5.18.16-1
Severity: important
Tags: upstream patch


Hey.

This is merely a heads up, that a potentially pretty nasty data corruption
issue has been found in btrfs.


As far as I understand:
It was introduced in 5.12, has the highest chances to occur with
free space cache v2, but may also happen with v1 or even nospace_cache.
v2 was made default with btrfs-progs 5.15.

It seems that it may in principle affect data and metadata, so could range from
single file corruptions to complete fs breakage?


The patch, which AFAICS hasn't been submitted to linux-stable yet, can be found
here:
https://lore.kernel.org/linux-btrfs/cover.1660690698.git.osandov@fb.com/
A more in-depth description can be found in it's first commit.


Some others and I myself have asked some further questions about the issue
(how one can detect any corruptions, and so on), which can also be found
in the very same mail thread.


Not sure if the above patch is already production read, but at least it would
be quite helpful if it could be cherry-picked for testing/sid as soon as it has
landed in linux-stable.


Thanks,
Chris.

[toc] | [next] | [standalone]


#76231

FromChristoph Anton Mitterer <calestyo@scientia.org>
Date2022-09-01 15:50 +0200
Message-ID<F0SJX-4zwX-1@gated-at.bofh.it>
In reply to#76173
Control: tags -1 + patch

Hey.

The patch for this has been merged in the stable kernels... could you
please cherry pick respectively update the sid kernel ASAP?

https://cdn.kernel.org/pub/linux/kernel/v5.x/ChangeLog-5.19.6

commit: a2e54eb64229f07f917b05d0c323604fda9b89f7
btrfs: fix space cache corruption and potential double allocations


Thanks,
Chris.

[toc] | [prev] | [next] | [standalone]


#76232 — Processed: RE: btrfs: space cache corruption and potential double allocations

From"Debian Bug Tracking System" <owner@bugs.debian.org>
Date2022-09-01 15:50 +0200
SubjectProcessed: RE: btrfs: space cache corruption and potential double allocations
Message-ID<F0SJX-4zwX-5@gated-at.bofh.it>
In reply to#76173
Processing control commands:

> tags -1 + patch
Bug #1017915 [src:linux] btrfs: space cache corruption and potential double allocations
Ignoring request to alter tags of bug #1017915 to the same tags previously set

-- 
1017915: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1017915
Debian Bug Tracking System
Contact owner@bugs.debian.org with problems

[toc] | [prev] | [next] | [standalone]


#76236 — Bug#1017915: marked as done (btrfs: space cache corruption and potential double allocations)

From"Debian Bug Tracking System" <owner@bugs.debian.org>
Date2022-09-01 16:50 +0200
SubjectBug#1017915: marked as done (btrfs: space cache corruption and potential double allocations)
Message-ID<F0TG1-4A6U-1@gated-at.bofh.it>
In reply to#76173

[Multipart message — attachments visible in raw view] — view raw

Your message dated Thu, 1 Sep 2022 16:40:57 +0200
with message-id <YxDEeZApuNlaSY12@eldamar.lan>
and subject line Re: Bug#1017915: btrfs: space cache corruption and potential double allocations
has caused the Debian Bug report #1017915,
regarding btrfs: space cache corruption and potential double allocations
to be marked as done.

This means that you claim that the problem has been dealt with.
If this is not the case it is now your responsibility to reopen the
Bug report if necessary, and/or fix the problem forthwith.

(NB: If you are a system administrator and have no idea what this
message is talking about, this may indicate a serious mail system
misconfiguration somewhere. Please contact owner@bugs.debian.org
immediately.)


-- 
1017915: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1017915
Debian Bug Tracking System
Contact owner@bugs.debian.org with problems

[toc] | [prev] | [next] | [standalone]


#76238

FromChristoph Anton Mitterer <calestyo@scientia.org>
Date2022-09-01 21:00 +0200
Message-ID<F0XzX-4Cwr-1@gated-at.bofh.it>
In reply to#76173
On Thu, 2022-09-01 at 16:40 +0200, Salvatore Bonaccorso wrote:
> 5.19.6-1 has been uploaded earlier today.

Ah,... okay... I had only seen the UNRELEASED commit in salsa.


Cheers,
Chris.

[toc] | [prev] | [next] | [standalone]


#76239

FromSalvatore Bonaccorso <carnil@debian.org>
Date2022-09-02 09:20 +0200
Message-ID<F1985-4K84-3@gated-at.bofh.it>
In reply to#76238
Hi,

On Thu, Sep 01, 2022 at 06:23:21PM +0200, Christoph Anton Mitterer wrote:
> On Thu, 2022-09-01 at 16:40 +0200, Salvatore Bonaccorso wrote:
> > 5.19.6-1 has been uploaded earlier today.
> 
> Ah,... okay... I had only seen the UNRELEASED commit in salsa.

Was probably then a race between finalizing the update and this
message :)

Almost all builds are done, but the signed packages need to be
triggered yet.

Regards,
Salvatore

[toc] | [prev] | [standalone]


Back to top | Article view | linux.debian.kernel


csiph-web