[alsa-devel] [Fwd: HP DV4 1155 SE - No Sound through speakers...]

[Jean-Pierre André]

Hi Herton,

Thank you for trying to sort his out,

Herton Ronaldo Krzesinski wrote:
> Em Sábado 31 Janeiro 2009, às 05:52:19, Jean-Pierre André escreveu:
>   
>> Hi,
>>
>>     
>>> Hi all, I have an HP DV4 1155-SE, I've been unable to get sound working
>>> through the speakers, although it does work through the headhpones. I ran
>>> alsa-.upgrade which downloaded compiled, and installed 1.0.19 - still no
>>> luck.
>>>   
>>>       
>> I have the same IDT codec 92HD71B7X on another
>> HP DV4 model, and I have tracked the problem to be
>> the kernel configuring IRQ 22 through the APIC, but
>> interrupts triggerring IRQ 20 (on kernel 2.6.27 and
>> alsa 1.0.19). It would be interesting if you could
>> confirm this by analyzing /proc/interrupts.
>>
>> The workarounds I know of are :
>> - adding irqpoll to the boot options,
>> - or adding noapic to the boot options,
>> - using mplayer to play music.
>>
>> There are with option "snd-hda-intel model=hp-m4"
>> (this option is for HP DV5/DV7, but HP DV4
>> is probably different)
>>
>> I have not enough knowledge to tell where the problem
>> lies : kernel, apic, alsa.
>>     
>
> It looks like some of these HP laptops have some sort of IRQ issue indeed, to 
> me it was reported that enabling pci MSI (both system wide and with 
> enable_msi=1 given to snd-hda-intel) cured issues in some cases, along with 
> some changes to sigmatel codec handling code in snd-hda-intel
>
> I have a ticked opened where people started to complain about sound issues 
> with hda-intel+sigmatel codec starting with alsa 1.0.18rc3 (sorry, it's a 
> great mess, type of the ticket everyone says sound doesn't work and report in, 
> but where I tried to filter some things out): 
> https://qa.mandriva.com/show_bug.cgi?id=44855
> Also some reports are hard to understand sometimes.
> With alsa 1.0.18 final in 2.6.27 (or 2.6.28) many issues related on the ticket 
> were fixed, but issues with HP laptops remained.
>
> I don't have any of affected HP laptops for test, but based on sigmatel 
> datasheets, reports of users and pin cfgs used in .ini files of sigmatel 
> windows drivers I made some changes to sigmatel code, plus added patch to set 
> enable_msi. The only issue I have with the changes, and that's why I didn't 
> submit yet, is that I should have split them, don't know if everything is 
> really required, and ask here to clarify some things below; also the changes 
> are outdated compared to current sound git tree.
>   

Thank you for your suggested patches. I had some difficulties in
applying them as your reference version is not from the lasted build.
(for instance from alsa-driver-1.0.19.0.g5aa59.6.gb162b.tar.bz2)
Actually, to get patch_sigmatel to compile, I had to reintegrate
old deprecated code. Attached are the patches I really applied.

And the result is worse than before : no sound at all, using option
"model hp-m4-2", as I see in the patch that this was designed for
my sound device (subsystem id 0x30f7). I am also using option
enable_msi=1.

And still no IRQ 22 seen, IRQ 20 instead, could be another problem,
but what can I do to clarify this ?

With the original code, this is what I get in the syslog,

Feb  3 12:27:18 pavilion2 kernel: HDA Intel 0000:00:1b.0: power state 
changed by ACPI to D0
Feb  3 12:27:18 pavilion2 kernel: HDA Intel 0000:00:1b.0: PCI INT B -> 
GSI 22 (level, low) -> IRQ 22
Feb  3 12:27:19 pavilion2 kernel: ALSA 
/var/tmp/bach-build/BUILD/alsa-driver-1.0.19/pci/hda/hda_intel.c:634: 
hda_intel: azx_get_response timeout, switching to polling mode: last 
cmd=0x000f0000
Feb  3 12:27:19 pavilion2 kernel: input: HDA Digital PCBeep as 
/devices/pci0000:00/0000:00:1b.0/input/input38
Feb  3 12:27:19 pavilion2 kernel: input: HDA Intel at 0xdf300000 irq 22 
Line In at Ext Rear Jack as /devices/pci0000:00/0000:00:1b.0/input/input39
Feb  3 12:27:19 pavilion2 kernel: input: HDA Intel at 0xdf300000 irq 22 
Mic at Ext Front Jack as /devices/pci0000:00/0000:00:1b.0/input/input40
Feb  3 12:27:19 pavilion2 kernel: input: HDA Intel at 0xdf300000 irq 22 
HP Out at Ext Front Jack as /devices/pci0000:00/0000:00:1b.0/input/input41

With the patched one, the last line does not appear, this should
lead to the point where the patch has broken something :

Feb  3 11:23:47 pavilion2 kernel: HDA Intel 0000:00:1b.0: power state 
changed by ACPI to D0
Feb  3 11:23:47 pavilion2 kernel: HDA Intel 0000:00:1b.0: PCI INT B -> 
GSI 22 (level, low) -> IRQ 22
Feb  3 11:23:48 pavilion2 kernel: ALSA 
/home/linux/rpmbuild/BUILD/alsa-driver-1.0.19/pci/hda/hda_intel.c:634: 
hda_intel: azx_get_response timeout, switching to polling mode: last 
cmd=0x000f0000
Feb  3 11:23:48 pavilion2 kernel: input: HDA Digital PCBeep as 
/devices/pci0000:00/0000:00:1b.0/input/input19
Feb  3 11:23:48 pavilion2 kernel: input: HDA Intel at 0xdf300000 irq 22 
Mic at Ext Front Jack as /devices/pci0000:00/0000:00:1b.0/input/input20
Feb  3 11:23:48 pavilion2 kernel: input: HDA Intel at 0xdf300000 irq 22 
HP Out at Ext Front Jack as /devices/pci0000:00/0000:00:1b.0/input/input21

I am open to any further test which you might suggest

Regards

Jean-Pierre

-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: herton-sigmatel.patch
Url: http://mailman.alsa-project.org/pipermail/alsa-devel/attachments/20090203/da4bb996/attachment-0001.bat