Hi David,

David Baron wrote:

> [Subject: Bug#659519: Other card works the same way!]

Please keep in mind that when these appear as emails in a crowded
inbox, the subject field can be a good way to provide context.

> I disabled the on-board audio and resinstalled my rusty-truesty ens1371.
>
> Guess what? Boot up the 3.2 kernel and one still cannot play from a browser.
>
> The hard-wired line, however, does work which might indicate a hw problem on
> the mobo's audio or jacks.

Thanks. Could you attach "dmesg" output from booting and trying this?
What's the last version of the kernel you've tried that worked with
the ens1371?

[...]
> ** Tainted: PO (4097)
> * Proprietary module has been loaded.
> * Out-of-tree module has been loaded.

Can you reproduce this without the nvidia and virtualbox drivers?
(That would help us get help from upstream.)

Hope that helps,
Jonathan



--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 20120513145636.GK9819@burratino">http://lists.debian.org/20120513145636.GK9819@burratino

On Sunday 13 May 2012 17:56:36 Jonathan Nieder wrote:
> Hi David,
>
> David Baron wrote:
> > [Subject: Bug#659519: Other card works the same way!]
>
> Please keep in mind that when these appear as emails in a crowded
> inbox, the subject field can be a good way to provide context.
>
> > I disabled the on-board audio and resinstalled my rusty-truesty ens1371.
> >
> > Guess what? Boot up the 3.2 kernel and one still cannot play from a
> > browser.
> >
> > The hard-wired line, however, does work which might indicate a hw problem
> > on the mobo's audio or jacks.
>
> Thanks. Could you attach "dmesg" output from booting and trying this?
> What's the last version of the kernel you've tried that worked with
> the ens1371?
>
Here is the output of dmesg | grep -i ens1371 for the 3.2 kernel boot:

[ 7.208063] snd_ens1371 0000:04:0b.0: BAR 0: set to [io 0xdf00-0xdf3f]
(PCI address [0xdf00-0xdf3f])
[ 7.208826] snd_ens1371 0000:04:0b.0: PCI INT A -> GSI 23 (level, low) ->
IRQ 23

Nothing very illuminating here. The intel had something about a "generic
initialization."

> [...]
>
> > ** Tainted: PO (4097)
> >
> > * Proprietary module has been loaded.
> > * Out-of-tree module has been loaded.
I see nothing like this. The nvidia module taints but this boot is not using
it.

> Can you reproduce this without the nvidia and virtualbox drivers?
> (That would help us get help from upstream.)

This was with nouveau.

Could I try without the virtualbox by simply blacklisting in /etc/modprob.d ?



--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 201205141703.20610.d_baron@012.net.il">http://lists.debian.org/201205141703.20610.d_baron@012.net.il

David Baron wrote:
> On Sunday 13 May 2012 17:56:36 Jonathan Nieder wrote:

>> Thanks. Could you attach "dmesg" output from booting and trying this?
>> What's the last version of the kernel you've tried that worked with
>> the ens1371?
>
> Here is the output of dmesg | grep -i ens1371 for the 3.2 kernel boot:
>
> [ 7.208063] snd_ens1371 0000:04:0b.0: BAR 0: set to [io 0xdf00-0xdf3f] (PCI address [0xdf00-0xdf3f])
> [ 7.208826] snd_ens1371 0000:04:0b.0: PCI INT A -> GSI 23 (level, low) -> IRQ 23
>
> Nothing very illuminating here.

Sorry for the lack of clarity. I meant a full kernel log (as an
attachment).

[...]
> Could I try without the virtualbox by simply blacklisting in /etc/modprob.d ?

Yes, that works.

Thanks,
Jonathan



--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 20120514143040.GC12418@burratino">http://lists.debian.org/20120514143040.GC12418@burratino

David Baron wrote:

> This is kind of large but here it is

Thanks!

[...]
> [ 6.700526] [drm] Initialized nouveau 0.0.16 20090420 for 0000:01:00.0 on minor 0
> [ 7.208063] snd_ens1371 0000:04:0b.0: BAR 0: set to [io 0xdf00-0xdf3f] (PCI address [0xdf00-0xdf3f])
> [ 7.208826] snd_ens1371 0000:04:0b.0: PCI INT A -> GSI 23 (level, low) -> IRQ 23
> [ 7.366048] nvidia: module license 'NVIDIA' taints kernel.
> [ 7.366800] Disabling lock debugging due to kernel taint
> [ 8.025924] NVRM: The NVIDIA probe routine was not called for 1 device(s).
> [ 8.026809] NVRM: This can occur when a driver such as nouveau, rivafb,
> [ 8.026811] NVRM: nvidiafb, or rivatv was loaded and obtained ownership of
> [ 8.026813] NVRM: the NVIDIA device(s).
> [ 8.029549] NVRM: Try unloading the conflicting kernel module (and/or

Probably blacklisting the nvidia driver through /etc/modprobe.d would
prevent these messages about the same. But I don't think it's the
cause of the problem.

[...]
> [ 2656.878112] ABORTED IN=eth2 OUT= MAC=00:e0:4c:68:00:c5:00:90:8f:2c:50:c9:08:00 SRC=208.83.137.114 DST=10.100.101.101 LEN=40 TOS=0x00 PREC=0x00 TTL=47 ID=61218 DF PROTO=TCP SPT=2703 DPT=9081 SEQ=1749643406 ACK=716693663 WINDOW=46 RES=0x00 ACK RST URGP=0
> [ 2956.105668] CPU0: Core temperature above threshold, cpu clock throttled (total events = 1)
> [ 2956.105692] CPU1: Core temperature above threshold, cpu clock throttled (total events = 1)
> [ 2956.106875] CPU0: Core temperature/speed normal
> [ 2956.106884] CPU1: Core temperature/speed normal
> [ 2999.988022] [Hardware Error]: Machine check events logged
(once)

Was the machine especially active, or is there a cooling problem?

[...]
> [22172.918562] [drm] nouveau 0000:01:00.0: Setting dpms mode 0 on vga encoder (output 0)
> [22296.194815] CPU0: Core temperature above threshold, cpu clock throttled (total events = 1694)
> [22296.194841] CPU1: Core temperature above threshold, cpu clock throttled (total events = 1694)
> [22296.195873] CPU1: Core temperature/speed normal
> [22296.195878] CPU0: Core temperature/speed normal
> [22349.988028] [Hardware Error]: Machine check events logged
> [22636.814452] CPU1: Core temperature above threshold, cpu clock throttled (total events = 5487)
> [22636.814478] CPU0: Core temperature above threshold, cpu clock throttled (total events = 5486)
> [22636.815515] CPU1: Core temperature/speed normal
> [22636.815521] CPU0: Core temperature/speed normal
> [22799.988028] [Hardware Error]: Machine check events logged
> [23002.995857] CPU1: Core temperature above threshold, cpu clock throttled (total events = 8649)
> [23002.995886] CPU0: Core temperature above threshold, cpu clock throttled (total events = 8648)
> [23002.997076] CPU1: Core temperature/speed normal
> [23002.997083] CPU0: Core temperature/speed normal
> [23100.000210] [Hardware Error]: Machine check events logged
> [23504.371700] CPU1: Core temperature above threshold, cpu clock throttled (total events = 9153)
> [23504.371723] CPU0: Core temperature above threshold, cpu clock throttled (total events = 9152)
> [23504.372762] CPU1: Core temperature/speed normal
> [23504.372768] CPU0: Core temperature/speed normal
> [23549.988022] [Hardware Error]: Machine check events logged
[...]

This time it keeps happening, until

> [24835.817176] CPU0: Core temperature above threshold, cpu clock throttled (total events = 9596)
> [24835.817200] CPU1: Core temperature above threshold, cpu clock throttled (total events = 9597)
> [24835.818353] CPU1: Core temperature/speed normal
> [24835.818356] CPU0: Core temperature/speed normal
[...]
> [25049.988081] [Hardware Error]: Machine check events logged
[...]
> [25280.229943] usb 1-3: USB disconnect, device number 6
> [25289.864027] usb 1-3: new high-speed USB device number 7 using ehci_hcd
> [25289.998233] usb 1-3: New USB device found, idVendor=1004, idProduct=618e
> [25289.998239] usb 1-3: New USB device strings: Mfr=1, Product=2, SerialNumber=3
> [25289.998243] usb 1-3: Product: LG Android USB Device
> [25289.998246] usb 1-3: Manufacturer: LG Electronics Inc.

Perhaps the machine was in a hot place at the same time as the phone
was plugged in. In any case, it would be useful to rule out cooling
problems, the nvidia driver, and the virtualbox driver as causes.

If 3.1.y (e.g., from snapshot.debian.org) still works fine, the above
look like red herrings, so still no idea what's actually wrong. Would
you be able to bisect to find the change that introduced this
regression (I can list what commands do so)?

Sorry for the trouble,
Jonathan



--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 20120514214408.GC7439@burratino">http://lists.debian.org/20120514214408.GC7439@burratino

On Tuesday 15 May 2012 00:44:08 Jonathan Nieder wrote:
> David Baron wrote:
> > This is kind of large but here it is
>
> Thanks!
>
> [...]
>
> > [ 6.700526] [drm] Initialized nouveau 0.0.16 20090420 for 0000:01:00.0
> > on minor 0 [ 7.208063] snd_ens1371 0000:04:0b.0: BAR 0: set to [io
> > 0xdf00-0xdf3f] (PCI address [0xdf00-0xdf3f]) [ 7.208826] snd_ens1371
> > 0000:04:0b.0: PCI INT A -> GSI 23 (level, low) -> IRQ 23 [ 7.366048]
> > nvidia: module license 'NVIDIA' taints kernel.
> > [ 7.366800] Disabling lock debugging due to kernel taint
> > [ 8.025924] NVRM: The NVIDIA probe routine was not called for 1
> > device(s). [ 8.026809] NVRM: This can occur when a driver such as
> > nouveau, rivafb, [ 8.026811] NVRM: nvidiafb, or rivatv was loaded and
> > obtained ownership of [ 8.026813] NVRM: the NVIDIA device(s).
> > [ 8.029549] NVRM: Try unloading the conflicting kernel module (and/or
>
> Probably blacklisting the nvidia driver through /etc/modprobe.d would
> prevent these messages about the same. But I don't think it's the
> cause of the problem.
>
These messages happen because Nouveau grabs the interfac unless IT is
blacklisted or nomodeset

> [...]
>
> > [ 2656.878112] ABORTED IN=eth2 OUT=
> > MAC=00:e0:4c:68:00:c5:00:90:8f:2c:50:c9:08:00 SRC=208.83.137.114
> > DST=10.100.101.101 LEN=40 TOS=0x00 PREC=0x00 TTL=47 ID=61218 DF
> > PROTO=TCP SPT=2703 DPT=9081 SEQ=1749643406 ACK=716693663 WINDOW=46
> > RES=0x00 ACK RST URGP=0 [ 2956.105668] CPU0: Core temperature above
> > threshold, cpu clock throttled (total events = 1) [ 2956.105692] CPU1:
> > Core temperature above threshold, cpu clock throttled (total events = 1)
> > [ 2956.106875] CPU0: Core temperature/speed normal
> > [ 2956.106884] CPU1: Core temperature/speed normal
> > [ 2999.988022] [Hardware Error]: Machine check events logged
>
> (once)
>
> Was the machine especially active, or is there a cooling problem?

These happen a lot, seem not to be dependent on room temperature. The CPU fan
works, I have de-dusted it, the case is open and still.

There is a "temp1 at 55 c listed on PCI adapter. It seems rock steady, too
steady! I can get it to 59-60 on a heavy 3d hw-accelerated graphics game.
Maybe this sensor is on the nvidia pci-express card, not the cpu!

Others, listed on ISA adapter, m/b, cpu, temp3 are all "disabled," possibly
because of bios settings or problems.

The fan itself, speed not sensored either, has its own thermostat, evident on
cold days.

These messages are annoying, the "throttling," if real, is a problem, but seem
spurious.



> [...]
>
> > [22172.918562] [drm] nouveau 0000:01:00.0: Setting dpms mode 0 on vga
> > encoder (output 0) [22296.194815] CPU0: Core temperature above
> > threshold, cpu clock throttled (total events = 1694) [22296.194841]
> > CPU1: Core temperature above threshold, cpu clock throttled (total
> > events = 1694) [22296.195873] CPU1: Core temperature/speed normal
> > [22296.195878] CPU0: Core temperature/speed normal
> > [22349.988028] [Hardware Error]: Machine check events logged
> > [22636.814452] CPU1: Core temperature above threshold, cpu clock
> > throttled (total events = 5487) [22636.814478] CPU0: Core temperature
> > above threshold, cpu clock throttled (total events = 5486)
> > [22636.815515] CPU1: Core temperature/speed normal
> > [22636.815521] CPU0: Core temperature/speed normal
> > [22799.988028] [Hardware Error]: Machine check events logged
> > [23002.995857] CPU1: Core temperature above threshold, cpu clock
> > throttled (total events = 8649) [23002.995886] CPU0: Core temperature
> > above threshold, cpu clock throttled (total events = 8648)
> > [23002.997076] CPU1: Core temperature/speed normal
> > [23002.997083] CPU0: Core temperature/speed normal
> > [23100.000210] [Hardware Error]: Machine check events logged
> > [23504.371700] CPU1: Core temperature above threshold, cpu clock
> > throttled (total events = 9153) [23504.371723] CPU0: Core temperature
> > above threshold, cpu clock throttled (total events = 9152)
> > [23504.372762] CPU1: Core temperature/speed normal
> > [23504.372768] CPU0: Core temperature/speed normal
> > [23549.988022] [Hardware Error]: Machine check events logged
>
> [...]
>
> This time it keeps happening, until
>
> > [24835.817176] CPU0: Core temperature above threshold, cpu clock
> > throttled (total events = 9596) [24835.817200] CPU1: Core temperature
> > above threshold, cpu clock throttled (total events = 9597)
> > [24835.818353] CPU1: Core temperature/speed normal
> > [24835.818356] CPU0: Core temperature/speed normal
>
> [...]
>
> > [25049.988081] [Hardware Error]: Machine check events logged
>
> [...]
>
> > [25280.229943] usb 1-3: USB disconnect, device number 6
> > [25289.864027] usb 1-3: new high-speed USB device number 7 using ehci_hcd
> > [25289.998233] usb 1-3: New USB device found, idVendor=1004,
> > idProduct=618e [25289.998239] usb 1-3: New USB device strings: Mfr=1,
> > Product=2, SerialNumber=3 [25289.998243] usb 1-3: Product: LG Android
> > USB Device
> > [25289.998246] usb 1-3: Manufacturer: LG Electronics Inc.
>
> Perhaps the machine was in a hot place at the same time as the phone
> was plugged in. In any case, it would be useful to rule out cooling
> problems, the nvidia driver, and the virtualbox driver as causes.
>
> If 3.1.y (e.g., from snapshot.debian.org) still works fine, the above
> look like red herrings, so still no idea what's actually wrong. Would
> you be able to bisect to find the change that introduced this
> regression (I can list what commands do so)?

Since problem occurs with or without nvidia, that is not the problem.
If non-sound drivers are causing this problem, I would guess virtual box
because it needs alsa for sound in its VMs. Further testing needed.



--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 201205151652.55486.d_baron@012.net.il">http://lists.debian.org/201205151652.55486.d_baron@012.net.il

David Baron wrote:
> On Tuesday 15 May 2012 00:44:08 Jonathan Nieder wrote:
>> David Baron wrote:

>>> [ 7.366048] nvidia: module license 'NVIDIA' taints kernel.
[...]
>> Probably blacklisting the nvidia driver through /etc/modprobe.d would
>> prevent these messages about the same. But I don't think it's the
>> cause of the problem.
>
> These messages happen because Nouveau grabs the interfac unless IT is
> blacklisted or nomodeset

Sorry for the lack of clarity. I meant messages indicating that the
nvidia binary-only driver has been loaded, like the one I've left
unsnipped above. (There are others, but that's the clearest one.)

[...]
>> Was the machine especially active, or is there a cooling problem?
>
> These happen a lot, seem not to be dependent on room temperature. The CPU fan
> works, I have de-dusted it, the case is open and still.
>
> There is a "temp1 at 55 c listed on PCI adapter. It seems rock steady, too
> steady! I can get it to 59-60 on a heavy 3d hw-accelerated graphics game.
> Maybe this sensor is on the nvidia pci-express card, not the cpu!

Ah, that makes sense. The overheating might be due to the GPU.
[1] has a rough overview of the current state of nouveau power
management --- NV40 looks like it has ok support, and your card
advertises only one performance level, but probably there is still
something that could be done.

[...]
>> If 3.1.y (e.g., from snapshot.debian.org) still works fine, the above
>> look like red herrings, so still no idea what's actually wrong. Would
>> you be able to bisect to find the change that introduced this
>> regression (I can list what commands do so)?
>
> Since problem occurs with or without nvidia, that is not the problem.
> If non-sound drivers are causing this problem, I would guess virtual box
> because it needs alsa for sound in its VMs. Further testing needed.

Right. We should probably take this upstream soon, so if you get a
chance to try with the virtualbox and nvidia drivers blacklisted, that
would be useful.

Thanks again for the quick feedback.

Sincerely,
Jonathan

[1] http://nouveau.freedesktop.org/wiki/PowerManagement



--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 20120515194305.GA7123@burratino">http://lists.debian.org/20120515194305.GA7123@burratino

On Tue, 2012-05-15 at 16:52 +0300, David Baron wrote:
[...]
> > [...]
> >
> > > [ 2656.878112] ABORTED IN=eth2 OUT=
> > > MAC=00:e0:4c:68:00:c5:00:90:8f:2c:50:c9:08:00 SRC=208.83.137.114
> > > DST=10.100.101.101 LEN=40 TOS=0x00 PREC=0x00 TTL=47 ID=61218 DF
> > > PROTO=TCP SPT=2703 DPT=9081 SEQ=1749643406 ACK=716693663 WINDOW=46
> > > RES=0x00 ACK RST URGP=0 [ 2956.105668] CPU0: Core temperature above
> > > threshold, cpu clock throttled (total events = 1) [ 2956.105692] CPU1:
> > > Core temperature above threshold, cpu clock throttled (total events = 1)
> > > [ 2956.106875] CPU0: Core temperature/speed normal
> > > [ 2956.106884] CPU1: Core temperature/speed normal
> > > [ 2999.988022] [Hardware Error]: Machine check events logged
> >
> > (once)
> >
> > Was the machine especially active, or is there a cooling problem?
>
> These happen a lot, seem not to be dependent on room temperature. The CPU fan
> works, I have de-dusted it, the case is open and still.

Opening a case can actually make fans less effective.

> There is a "temp1 at 55 c listed on PCI adapter. It seems rock steady, too
> steady! I can get it to 59-60 on a heavy 3d hw-accelerated graphics game.
> Maybe this sensor is on the nvidia pci-express card, not the cpu!
[...]

Quite likely, yes. The nouveau driver can expose temperature sensors on
graphics cards.

Ben.

--
Ben Hutchings
The two most common things in the universe are hydrogen and stupidity.