Error "ee1004 3-0050: probe with driver ee1004 failed with error -5" since 6.11

Error "ee1004 3-0050: probe with driver ee1004 failed with error -5" since 6.11

[Diederik de Haas]

[-- Attachment #1: Type: text/plain, Size: 458 bytes --]

Hi,

I updated an amd64 machine to (Debian's) 6.11 kernel and got this error:
``ee1004 3-0050: probe with driver ee1004 failed with error -5``

which I didn't get with 6.10.9.
If needed I could do a git bisect, but maybe you can already directly
tell with commit would've caused this?

If you need more info, please tell me where to get it as this is a
(relatively) new system and I actually don't know where the eeprom is.

Cheers,
  Diederik

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 228 bytes --]

Re: Error "ee1004 3-0050: probe with driver ee1004 failed with error -5" since 6.11

[Greg Kroah-Hartman]

On Thu, Sep 26, 2024 at 02:35:05PM +0200, Diederik de Haas wrote:
> Hi,
> 
> I updated an amd64 machine to (Debian's) 6.11 kernel and got this error:
> ``ee1004 3-0050: probe with driver ee1004 failed with error -5``
> 
> which I didn't get with 6.10.9.
> If needed I could do a git bisect, but maybe you can already directly
> tell with commit would've caused this?
> 
> If you need more info, please tell me where to get it as this is a
> (relatively) new system and I actually don't know where the eeprom is.

bisect would be great, thanks!

greg k-h

Re: Error "ee1004 3-0050: probe with driver ee1004 failed with error -5" since 6.11

[Diederik de Haas]

[-- Attachment #1: Type: text/plain, Size: 6003 bytes --]

On Thu Sep 26, 2024 at 2:44 PM CEST, Greg Kroah-Hartman wrote:
> On Thu, Sep 26, 2024 at 02:35:05PM +0200, Diederik de Haas wrote:
> > I updated an amd64 machine to (Debian's) 6.11 kernel and got this error:
> > ``ee1004 3-0050: probe with driver ee1004 failed with error -5``
> > 
> > which I didn't get with 6.10.9.
>
> bisect would be great, thanks!

I should've known I'd get this response ;-P

$ git bisect log
git bisect start
# status: waiting for both good and bad commits
# good: [6150e5e1ae2d8ad72f52217f8f41fe446cae9e27] eeprom: ee1004: Instantiate jc42 devices for DIMMS implementing Rev.1 SPD
git bisect good 6150e5e1ae2d8ad72f52217f8f41fe446cae9e27
# status: waiting for bad commit, 1 good commit known
# bad: [8400291e289ee6b2bf9779ff1c83a291501f017b] Linux 6.11-rc1
git bisect bad 8400291e289ee6b2bf9779ff1c83a291501f017b
# good: [280e36f0d5b997173d014c07484c03a7f7750668] nsfs: use cleanup guard
git bisect good 280e36f0d5b997173d014c07484c03a7f7750668
# good: [f66b07c56119833b88bffa4ecaf9f983834675de] Merge tag 'vfio-v6.11-rc1' of https://github.com/awilliam/linux-vfio
git bisect good f66b07c56119833b88bffa4ecaf9f983834675de
# good: [2c9b3512402ed192d1f43f4531fb5da947e72bd0] Merge tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm
git bisect good 2c9b3512402ed192d1f43f4531fb5da947e72bd0
# good: [6dc2e98d5f1de162d1777aee97e59d75d70d07c5] s390: Remove protvirt and kvm config guards for uv code
git bisect good 6dc2e98d5f1de162d1777aee97e59d75d70d07c5
# bad: [c2a96b7f187fb6a455836d4a6e113947ff11de97] Merge tag 'driver-core-6.11-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/driver-core
git bisect bad c2a96b7f187fb6a455836d4a6e113947ff11de97
# good: [786c8248dbd33a5a7a07f7c6e55a7bfc68d2ca48] Merge tag 'perf-tools-fixes-for-v6.11-2024-07-23' of git://git.kernel.org/pub/scm/linux/kernel/git/perf/perf-tools
git bisect good 786c8248dbd33a5a7a07f7c6e55a7bfc68d2ca48
# bad: [ad7b0b7b87bb19ea7ee71e165e597f5fbfac76ab] Merge tag 'soundwire-6.11-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/vkoul/soundwire
git bisect bad ad7b0b7b87bb19ea7ee71e165e597f5fbfac76ab
# good: [b8ec9dba02a74797421c52b1226b23a4302362a6] dt-bindings: fsl-qdma: fix interrupts 'if' check logic
git bisect good b8ec9dba02a74797421c52b1226b23a4302362a6
# good: [7a3fad30fd8b4b5e370906b3c554f64026f56c2f] Merge tag 'random-6.11-rc1-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/crng/random
git bisect good 7a3fad30fd8b4b5e370906b3c554f64026f56c2f
# good: [89cc1354d388ba8c8f8b41095736202a83591497] soundwire: amd: simplify return path in hw_params
git bisect good 89cc1354d388ba8c8f8b41095736202a83591497
# good: [13814ed162687be08e34762040cfc2e58831219d] soundwire: intel_ace2x: simplify return path in hw_params
git bisect good 13814ed162687be08e34762040cfc2e58831219d
# good: [b72d4af98cae2f74dc8061befcc3c0c2a174894f] soundwire: cadence: simplify with cleanup.h
git bisect good b72d4af98cae2f74dc8061befcc3c0c2a174894f
# good: [fdd3d14ca3c8c5269174f10d33d6181173cbd0b4] soundwire: debugfs: simplify with cleanup.h
git bisect good fdd3d14ca3c8c5269174f10d33d6181173cbd0b4
# bad: [7a46b17d4c00c2547b5bd82eec9489b19128fd65] Merge tag 'dmaengine-6.11-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/vkoul/dmaengine
git bisect bad 7a46b17d4c00c2547b5bd82eec9489b19128fd65
# first bad commit: [7a46b17d4c00c2547b5bd82eec9489b19128fd65] Merge tag 'dmaengine-6.11-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/vkoul/dmaengine

During the bisect operation I noticed the following messages:
[    4.920273] ee1004 1-0050: 512 byte EE1004-compliant SPD EEPROM, read-only
[    4.922699] ee1004 1-0051: 512 byte EE1004-compliant SPD EEPROM, read-only
[    4.924960] ee1004 1-0052: 512 byte EE1004-compliant SPD EEPROM, read-only
[    4.927236] ee1004 1-0053: 512 byte EE1004-compliant SPD EEPROM, read-only

but only on some. I didn't get those before, but they're informational
so not a problem. (But I don't care for them either)

Good commits, no info message:
89cc1354d388 soundwire: amd: simplify return path in hw_params
13814ed16268 soundwire: intel_ace2x: simplify return path in hw_params
b72d4af98cae soundwire: cadence: simplify with cleanup.h
fdd3d14ca3c8 soundwire: debugfs: simplify with cleanup.h
b8ec9dba02a7 dt-bindings: fsl-qdma: fix interrupts 'if' check logic
6150e5e1ae2d eeprom: ee1004: Instantiate jc42 devices for DIMMS implementing Rev.1 SPD
280e36f0d5b9 nsfs: use cleanup guard
f66b07c56119 Merge tag 'vfio-v6.11-rc1' of https://github.com/awilliam/linux-vfio
2c9b3512402e Merge tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm
6dc2e98d5f1d s390: Remove protvirt and kvm config guards for uv code

Good commits, with info message:
786c8248dbd3 Merge tag 'perf-tools-fixes-for-v6.11-2024-07-23' of git://git.kernel.org/pub/scm/linux/kernel/git/perf/perf-tools
7a3fad30fd8b Merge tag 'random-6.11-rc1-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/crng/random

Bad commits, all with info message:
7a46b17d4c00 Merge tag 'dmaengine-6.11-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/vkoul/dmaengine
ad7b0b7b87bb Merge tag 'soundwire-6.11-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/vkoul/soundwire
c2a96b7f187f Merge tag 'driver-core-6.11-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/driver-core
8400291e289e Linux 6.11-rc1

``git blame -L '/static int ee1004_probe/' drivers/misc/eeprom/ee1004.c``
showed the ``dev_info`` line has been present since 2018-10-07, but I
guess it never used the ``ee1004_probe`` function before?

Anticipating the likely next question "Can you still reproduce it in
current master/HEAD?" I build a new kernel from this:
075dbe9f6e3c ("Merge tag 'soc-ep93xx-dt-6.12' of git://git.kernel.org/pub/scm/linux/kernel/git/soc/soc")

And that showed the info messages, but not the error.
So that looks like the bug has been fixed for 6.12.

Cheers,
  Diederik

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 228 bytes --]

Re: Error "ee1004 3-0050: probe with driver ee1004 failed with error -5" since 6.11

[Uwe Kleine-König]

[-- Attachment #1: Type: text/plain, Size: 5329 bytes --]

Hello Diederik,

On Fri, Sep 27, 2024 at 02:57:30PM +0200, Diederik de Haas wrote:
> On Thu Sep 26, 2024 at 2:44 PM CEST, Greg Kroah-Hartman wrote:
> > On Thu, Sep 26, 2024 at 02:35:05PM +0200, Diederik de Haas wrote:
> > > I updated an amd64 machine to (Debian's) 6.11 kernel and got this error:
> > > ``ee1004 3-0050: probe with driver ee1004 failed with error -5``
> > > 
> > > which I didn't get with 6.10.9.
> >
> > bisect would be great, thanks!
> 
> I should've known I'd get this response ;-P
> 
> $ git bisect log
> git bisect start
> # status: waiting for both good and bad commits
> # good: [6150e5e1ae2d8ad72f52217f8f41fe446cae9e27] eeprom: ee1004: Instantiate jc42 devices for DIMMS implementing Rev.1 SPD
> git bisect good 6150e5e1ae2d8ad72f52217f8f41fe446cae9e27
> # status: waiting for bad commit, 1 good commit known
> # bad: [8400291e289ee6b2bf9779ff1c83a291501f017b] Linux 6.11-rc1
> git bisect bad 8400291e289ee6b2bf9779ff1c83a291501f017b
> # good: [280e36f0d5b997173d014c07484c03a7f7750668] nsfs: use cleanup guard
> git bisect good 280e36f0d5b997173d014c07484c03a7f7750668
> # good: [f66b07c56119833b88bffa4ecaf9f983834675de] Merge tag 'vfio-v6.11-rc1' of https://github.com/awilliam/linux-vfio
> git bisect good f66b07c56119833b88bffa4ecaf9f983834675de
> # good: [2c9b3512402ed192d1f43f4531fb5da947e72bd0] Merge tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm
> git bisect good 2c9b3512402ed192d1f43f4531fb5da947e72bd0
> # good: [6dc2e98d5f1de162d1777aee97e59d75d70d07c5] s390: Remove protvirt and kvm config guards for uv code
> git bisect good 6dc2e98d5f1de162d1777aee97e59d75d70d07c5
> # bad: [c2a96b7f187fb6a455836d4a6e113947ff11de97] Merge tag 'driver-core-6.11-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/driver-core
> git bisect bad c2a96b7f187fb6a455836d4a6e113947ff11de97
> # good: [786c8248dbd33a5a7a07f7c6e55a7bfc68d2ca48] Merge tag 'perf-tools-fixes-for-v6.11-2024-07-23' of git://git.kernel.org/pub/scm/linux/kernel/git/perf/perf-tools
> git bisect good 786c8248dbd33a5a7a07f7c6e55a7bfc68d2ca48
> # bad: [ad7b0b7b87bb19ea7ee71e165e597f5fbfac76ab] Merge tag 'soundwire-6.11-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/vkoul/soundwire
> git bisect bad ad7b0b7b87bb19ea7ee71e165e597f5fbfac76ab
> # good: [b8ec9dba02a74797421c52b1226b23a4302362a6] dt-bindings: fsl-qdma: fix interrupts 'if' check logic
> git bisect good b8ec9dba02a74797421c52b1226b23a4302362a6
> # good: [7a3fad30fd8b4b5e370906b3c554f64026f56c2f] Merge tag 'random-6.11-rc1-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/crng/random
> git bisect good 7a3fad30fd8b4b5e370906b3c554f64026f56c2f
> # good: [89cc1354d388ba8c8f8b41095736202a83591497] soundwire: amd: simplify return path in hw_params
> git bisect good 89cc1354d388ba8c8f8b41095736202a83591497
> # good: [13814ed162687be08e34762040cfc2e58831219d] soundwire: intel_ace2x: simplify return path in hw_params
> git bisect good 13814ed162687be08e34762040cfc2e58831219d
> # good: [b72d4af98cae2f74dc8061befcc3c0c2a174894f] soundwire: cadence: simplify with cleanup.h
> git bisect good b72d4af98cae2f74dc8061befcc3c0c2a174894f
> # good: [fdd3d14ca3c8c5269174f10d33d6181173cbd0b4] soundwire: debugfs: simplify with cleanup.h
> git bisect good fdd3d14ca3c8c5269174f10d33d6181173cbd0b4
> # bad: [7a46b17d4c00c2547b5bd82eec9489b19128fd65] Merge tag 'dmaengine-6.11-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/vkoul/dmaengine
> git bisect bad 7a46b17d4c00c2547b5bd82eec9489b19128fd65
> # first bad commit: [7a46b17d4c00c2547b5bd82eec9489b19128fd65] Merge tag 'dmaengine-6.11-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/vkoul/dmaengine

The following is completely generic. Note I didn't try to to understand
the problem, just giving some hint how to further improve the bisection
result.

If the first bad commit is a merge that isn't usually very helpful. It
just means that either the merge is broken (I guess that's unlikely) or
one commit on the left side of the merge has a problem with one on the
right side.

To identify the two conflicting commits you can do the following:

	badmerge=7a46b17d4c00c2547b5bd82eec9489b19128fd65
	base=$(git merge-base ${badmerge}^ ${badmerge}^2)
	git bisect start ${badmerge}^ ${base}

and then to test a commit that git-bisect asks you to judge, do:

	git merge --no-commit ${badmerge}^2
	build_and_test
	git reset --hard
	git bisect $good_or_bad

The resulting bad commit is the problematic on the left side (i.e. in
${base}..${badmerge}^). To then find the problematic on the right side,
start a new bisection:

	git bisect start ${badmerge}^2 ${base}

and to test these do:

	git merge --no-commit ${bad_commit_on_left_side}
	build_and_test
	git reset --hard
	git bisect $good_or_bad

The first bad there is then the conflicting commit on the right side
(i.e. in ${base}..${badmerge}^2).

Also finding the commit that fixes the issue would be possible, for that
I'd suggest

	git bisect start --term-new=fixed --term-old=broken v6.11-rc1 075dbe9f6e3c

and then after testing the suggested commits use either

	git bisect fixed

(if the problem didn't occur); or

	git bisect broken

(if the problem is still unfixed in the commit to test).

Best regards
Uwe

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

Re: Error "ee1004 3-0050: probe with driver ee1004 failed with error -5" since 6.11

[Diederik de Haas]

[-- Attachment #1: Type: text/plain, Size: 2601 bytes --]

Hi Uwe,

On Fri Sep 27, 2024 at 11:59 PM CEST, Uwe Kleine-König wrote:
> On Fri, Sep 27, 2024 at 02:57:30PM +0200, Diederik de Haas wrote:
> > On Thu Sep 26, 2024 at 2:44 PM CEST, Greg Kroah-Hartman wrote:
> > > On Thu, Sep 26, 2024 at 02:35:05PM +0200, Diederik de Haas wrote:
> > > > I updated an amd64 machine to (Debian's) 6.11 kernel and got this error:
> > > > ``ee1004 3-0050: probe with driver ee1004 failed with error -5``
> > > > 
> > > > which I didn't get with 6.10.9.
> > >
> > > bisect would be great, thanks!
> > 
> > $ git bisect log
> > git bisect start
> > # status: waiting for both good and bad commits
> > # good: [6150e5e1ae2d8ad72f52217f8f41fe446cae9e27] eeprom: ee1004: Instantiate jc42 devices for DIMMS implementing Rev.1 SPD
> > git bisect good 6150e5e1ae2d8ad72f52217f8f41fe446cae9e27
> > # status: waiting for bad commit, 1 good commit known
> > # bad: [8400291e289ee6b2bf9779ff1c83a291501f017b] Linux 6.11-rc1
> > ... [omitted for brevity] ...
> > # bad: [7a46b17d4c00c2547b5bd82eec9489b19128fd65] Merge tag 'dmaengine-6.11-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/vkoul/dmaengine
> > git bisect bad 7a46b17d4c00c2547b5bd82eec9489b19128fd65
> > # first bad commit: [7a46b17d4c00c2547b5bd82eec9489b19128fd65] Merge tag 'dmaengine-6.11-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/vkoul/dmaengine
>
> The following is completely generic. Note I didn't try to to understand
> the problem, just giving some hint how to further improve the bisection
> result.
>
> [ extensive description on what to do when git bisect results
>   in a merge commit as 'first bad commit' ]
>

This is awesome, thank you very much :)

I already thought that this git bisect result wasn't very useful as I
failed to see the logic of dmaengine changes being relevant for this
issue.
And I (indeed) didn't know what to do when a git bisect resulted in a
merge commit as 'first bad commit', so your description will surely come
in handy (someday).

I haven't made up my mind yet whether I'll apply it to this specific
issue. It doesn't cause any actual problems (AFAICT) and most
importantly, it seems to be fixed in 6.12(-rc1) already.
And the 6.11 kernel is only a 'temporary' one which I don't care (much)
about, while it is fixed in the kernel I do care about (6.12).

So if there isn't a follow up from me on this issue, it certainly is NOT
related to your message, which I appreciate a LOT :-)
(And I'm pretty sure there'll be other instances where a git bisect
results in a merge commit ...)

Cheers,
  Diederik


[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 228 bytes --]

Re: Error "ee1004 3-0050: probe with driver ee1004 failed with error -5" since 6.11

[Uwe Kleine-König]

[-- Attachment #1: Type: text/plain, Size: 1132 bytes --]

Hello Diederik,

On Sat, Sep 28, 2024 at 08:54:12AM +0200, Diederik de Haas wrote:
> On Fri Sep 27, 2024 at 11:59 PM CEST, Uwe Kleine-König wrote:
> > [ extensive description on what to do when git bisect results
> >   in a merge commit as 'first bad commit' ]
> 
> This is awesome, thank you very much :)

Were you able to understand that, or did you just bookmark it for the
next time? If you still have questions, don't hesitate to ask.
 
> I haven't made up my mind yet whether I'll apply it to this specific
> issue. It doesn't cause any actual problems (AFAICT) and most
> importantly, it seems to be fixed in 6.12(-rc1) already.
> And the 6.11 kernel is only a 'temporary' one which I don't care (much)
> about, while it is fixed in the kernel I do care about (6.12).

It might be beneficial to find the commit between 6.11 and 6.12-rc1
that's fixing the issue. That would allow to backport the fix to 6.11.x.

(No need for a complicated procedure as I described, just bisect between
6.11 and 6.12-rc1 (maybe with "new" and "old" instead of "bad" and
"good" for less confusion)).

Best regards
Uwe

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

Re: Error "ee1004 3-0050: probe with driver ee1004 failed with error -5" since 6.11

[Diederik de Haas]

[-- Attachment #1: Type: text/plain, Size: 1784 bytes --]

Hi Uwe,

On Mon Oct 7, 2024 at 4:38 PM CEST, Uwe Kleine-König wrote:
> On Sat, Sep 28, 2024 at 08:54:12AM +0200, Diederik de Haas wrote:
> > On Fri Sep 27, 2024 at 11:59 PM CEST, Uwe Kleine-König wrote:
> > > [ extensive description on what to do when git bisect results
> > >   in a merge commit as 'first bad commit' ]
> > 
> > This is awesome, thank you very much :)
>
> Were you able to understand that, or did you just bookmark it for the
> next time? If you still have questions, don't hesitate to ask.

Thanks for the offer :)
  
> > I haven't made up my mind yet whether I'll apply it to this specific
> > issue. It doesn't cause any actual problems (AFAICT) and most
> > importantly, it seems to be fixed in 6.12(-rc1) already.
> > And the 6.11 kernel is only a 'temporary' one which I don't care (much)
> > about, while it is fixed in the kernel I do care about (6.12).
>
> It might be beneficial to find the commit between 6.11 and 6.12-rc1
> that's fixing the issue. That would allow to backport the fix to 6.11.x.

Today I finally had time to try it out ... to find I can no longer
reproduce it. At least not reliably. I've seen the error message once
with 6.11-rc1 today, but not since. When I now boot 6.11-rc1, I do not
see it and I've rebooted several times.

I did upgrade the BIOS to a new (beta) version since my previous
attempts, which may play a factor? But that seems like a minor change.
So it could be that my previous bisect operation wasn't reliable either.

The issue wasn't a big deal to begin with but mostly interesting to
enhance my git bisect skills. But now that I can't reliably reproduce
the issue, that makes even that moot.

So I'm hereby stopping my attempts to bisect this issue.

Cheers,
  Diederik

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 228 bytes --]