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

Groups > linux.debian.kernel > #76686 > unrolled thread

Bug#1022126: mpt3sas broken with xen dom0

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

This discussion starts older than the indexed window; earlier articles aren't shown. The article labeled Started by below is the oldest one visible, not the original post.

Contents

  Bug#1022126: mpt3sas broken with xen dom0 Marc-Robin Wendt <wendt@math.tu-berlin.de> - 2022-10-25 11:40 +0200
    Bug#1022126: mpt3sas broken with xen dom0 Diederik de Haas <didi.debian@cknow.org> - 2022-12-08 17:30 +0100
      Processed: Re: Bug#1022126: mpt3sas broken with xen dom0 "Debian Bug Tracking System" <owner@bugs.debian.org> - 2022-12-08 17:30 +0100
      Bug#1022126: mpt3sas broken with xen dom0 Adi Kriegisch <adi@kriegisch.at> - 2022-12-15 11:30 +0100
        Bug#1022126: mpt3sas broken with xen dom0 Diederik de Haas <didi.debian@cknow.org> - 2022-12-15 12:00 +0100
          Bug#1022126: mpt3sas broken with xen dom0 Diederik de Haas <didi.debian@cknow.org> - 2022-12-15 12:10 +0100
            Bug#1022126: mpt3sas broken with xen dom0 Taavi Ansper <taavi.ansper@cyber.ee> - 2023-01-02 15:10 +0100
              Bug#1022126: mpt3sas broken with xen dom0 Salvatore Bonaccorso <carnil@debian.org> - 2023-01-04 11:20 +0100

#76686 — Bug#1022126: mpt3sas broken with xen dom0

Hello,

same problem here with linux-image-5.10.0-19-amd64 version 5.10.149-2.

Went back to kernel linux-image-5.10.0-17-amd64 ver. 5.10.136-1, which
still works fine.

Greetings

Marc

[toc] | [next] | [standalone]

#77514

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

Control: tag -1 =upstream

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

#77515 — Processed: Re: Bug#1022126: mpt3sas broken with xen dom0

Processing control commands:

> tag -1 =upstream
Bug #1022126 [src:linux] mpt3sas broken with xen dom0
Bug #1022268 [src:linux] linux-image-5.10.0-19-amd64: Boot failure under xen as dom0
Bug #1023183 [src:linux] mpt3sas driver does not load
Bug #1025747 [src:linux] linux-image-5.10.0-19: linux-image-5.10.0-19 not boot with XEN (mpt3sas error)
Added tag(s) upstream; removed tag(s) moreinfo.
Added tag(s) upstream; removed tag(s) moreinfo.
Added tag(s) upstream; removed tag(s) moreinfo.
Added tag(s) upstream; removed tag(s) moreinfo.

-- 
1022126: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1022126
1022268: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1022268
1023183: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1023183
1025747: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1025747
Debian Bug Tracking System
Contact owner@bugs.debian.org with problems

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

#77615

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

Tags: patch

Dear maintainers,

after the discussion about the root cause of this issue on the linux-scsi
mailing list[1] there was one simple patch submitted to this list[2] that
just changes the logic of the selection of the dma mask: in case of a 64bit
host operating system, the driver chooses a 63/64bit mask.

We backported and tested this patch on our systems and everything works
just fine. Find our backported patch for 5.10 kernels attached. We tested
that patch with 5.10.149-2/5.10.0-19 and 5.10.158-2/5.10.0-20.

Note, however, this patch has not yet been merged upstream and thus there
is no backport of this to the lts kernels. I am pretty unsure about how to
proceed: currently we're using dkms to build binary module packages with
the attached patch (dkms mkbmdeb) and do deploy these to our servers but
we'd really love to see this patch upstream and downstream! :)

all the best,
    Adi

[1] https://lore.kernel.org/linux-scsi/Y1JkuKTjVYrOWbvm@eldamar.lan/T/
[2] https://lore.kernel.org/linux-scsi/20221028091655.17741-2-sreekanth.reddy@broadcom.com/T/

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

#77618

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

On Thursday, 15 December 2022 10:43:22 CET Adi Kriegisch wrote:
> after the discussion about the root cause of this issue on the linux-scsi
> mailing list[1] there was one simple patch submitted to this list[2] that
> just changes the logic of the selection of the dma mask: in case of a 64bit
> host operating system, the driver chooses a 63/64bit mask.

Was Juergen Gross part of that discussion? I didn't see his name.
I do know he's one of the main Xen contributors and he seemed to have some 
good input in [1].

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

#77619

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

On Thursday, 15 December 2022 11:47:11 CET Diederik de Haas wrote:
> On Thursday, 15 December 2022 10:43:22 CET Adi Kriegisch wrote:
> > after the discussion about the root cause of this issue on the linux-scsi
> > mailing list[1] there was one simple patch submitted to this list[2] that
> > just changes the logic of the selection of the dma mask: in case of a
> > 64bit
> > host operating system, the driver chooses a 63/64bit mask.
> 
> Was Juergen Gross part of that discussion? I didn't see his name.
> I do know he's one of the main Xen contributors and he seemed to have some
> good input in [1].

Probably better to (also) send it to xen-devel@lists.xenproject.org as there 
are more smart people with a lot of Xen knowledge there.

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

#77809

Are there any news regarding when the fix would be implemented for bullseye?

Patiently waiting :)

On 15.12.22 12:57, Diederik de Haas wrote:
> On Thursday, 15 December 2022 11:47:11 CET Diederik de Haas wrote:
>> On Thursday, 15 December 2022 10:43:22 CET Adi Kriegisch wrote:
>>> after the discussion about the root cause of this issue on the linux-scsi
>>> mailing list[1] there was one simple patch submitted to this list[2] that
>>> just changes the logic of the selection of the dma mask: in case of a
>>> 64bit
>>> host operating system, the driver chooses a 63/64bit mask.
>>
>> Was Juergen Gross part of that discussion? I didn't see his name.
>> I do know he's one of the main Xen contributors and he seemed to have some
>> good input in [1].
> 
> Probably better to (also) send it to xen-devel@lists.xenproject.org as there
> are more smart people with a lot of Xen knowledge there.

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

#77821

Hi Taavi,

On Mon, Jan 02, 2023 at 03:54:16PM +0200, Taavi Ansper wrote:
> Are there any news regarding when the fix would be implemented for bullseye?

See
https://lore.kernel.org/linux-scsi/181536c494aa39ca78b190396a97072448739411.camel@suse.com/
for an upstream status. Once the patches land in mainline, they can be
picked up for the stable series.

Then it will be included as well for bullseye as we follow the stable
series and try to rebase at least on each point release.

So I would expect we can see the fix in the next update.

Regards,
Salvatore

[toc] | [prev] | [standalone]

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

csiph-web