* Re: [Alsa-user] arecord under 2.6.15.4-rt17 ->overruns...
[not found] ` <1141509605.14714.11.camel@mindpipe>
@ 2006-03-06 17:19 ` Guennadi Liakhovetski
2006-03-07 18:30 ` Guennadi Liakhovetski
0 siblings, 1 reply; 11+ messages in thread
From: Guennadi Liakhovetski @ 2006-03-06 17:19 UTC (permalink / raw)
To: Lee Revell; +Cc: Sergei Steshenko, alsa-user, linux-kernel, gregkh
(Adding lkml and PCI-maintainer to CC: Short description of the problem -
VERY high latency during audio recording, sometimes more than a second
with 2.6.15, 2.6.0, 2.4.32.)
On Sat, 4 Mar 2006, Lee Revell wrote:
> Yes it's not likely to be a RAM problem, but broken interrupt routing or
> an ACPI issue.
BIOS is already the newest version. Another thing:
linux/Documentation/sound/oss/VIA-chipset:
Running sound cards on VIA chipsets
o There are problems with VIA chipsets and sound cards that appear to
lock the hardware solidly. Test programs under DOS have verified the
problem exists on at least some (but apparently not all) VIA boards
o VIA have so far failed to bother to answer support mail on the subject
so if you are a VIA engineer feeling aggrieved as you read this
document go chase your own people. If there is a workaround please
let us know so we can implement it.
... and later:
Linux implements a workaround providing your chipset is PCI and you compiled
with PCI Quirks enabled. If so you will see a message
"Activating ISA DMA bug workarounds"
during booting. If you have a VIA PCI chipset that hangs when you use the
sound and is not generating this message even with PCI quirks enabled
please report the information to the linux-kernel list (see REPORTING-BUGS).
in drivers/pci/quirks.c:
/*
* VIA Apollo KT133 needs PCI latency patch
* Made according to a windows driver based patch by George E. Breese
* see PCI Latency Adjust on http://www.viahardware.com/download/viatweak.shtm
* Also see http://www.au-ja.org/review-kt133a-1-en.phtml for
* the info on which Mr Breese based his work.
*
* Updated based on further information from the site and also on
* information provided by VIA
*/
and later:
/*
* Ok we have the problem. Now set the PCI master grant to
* occur every master grant. The apparent bug is that under high
* PCI load (quite common in Linux of course) you can get data
* loss when the CPU is held off the bus for 3 bus master requests
* This happens to include the IDE controllers....
*
* VIA only apply this fix when an SB Live! is present but under
* both Linux and Windows this isnt enough, and we have seen
* corruption without SB Live! but with things like 3 UDMA IDE
* controllers. So we ignore that bit of the VIA recommendation..
*/
pci_read_config_byte(dev, 0x76, &busarb);
/* Set bit 4 and bi 5 of byte 76 to 0x01
"Master priority rotation on every PCI master grant */
busarb &= ~(1<<5);
busarb |= (1<<4);
pci_write_config_byte(dev, 0x76, busarb);
printk(KERN_INFO "Applying VIA southbridge workaround\n");
which in itself is wrong - the comment says "set bit 5 & 6 to 1", but
the code clears bit 6. Anyway, on my PC:
0000:00:07.0 ISA bridge: VIA Technologies, Inc. VT82C686 [Apollo Super South] (rev 40)
00: 06 11 86 06 87 00 10 02 40 00 01 06 00 00 80 00
10: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
20: 00 00 00 00 00 00 00 00 00 00 00 00 43 10 e7 80
30: 00 00 00 00 c0 00 00 00 00 00 00 00 00 00 00 00
40: 08 01 00 00 00 80 62 e6 01 01 44 00 00 00 f0 f3
50: 0e 76 34 00 00 b0 5a 90 00 04 ff 08 d0 00 00 00
60: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
70: 00 00 00 00 21 88 42 02 04 e4 00 00 00 00 00 00
^^
80: 01 00 00 00 00 09 00 00 00 60 00 00 00 00 00 00
90: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
a0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
b0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
c0: 01 00 02 00 00 00 00 00 00 00 00 00 00 00 00 00
d0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
e0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
f0: 00 00 00 00 00 00 41 00 00 00 00 00 00 00 00 00
So, bit 4 is not set... I added busarb to the printk above and it prints
0xd2. Why is it reset to 0x42 again later??? Found nothing in
drivers/ide/pci/via82cxxx.c... I could read it back to check it did get
written.
Thanks
Guennadi
---
Guennadi Liakhovetski
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: [Alsa-user] arecord under 2.6.15.4-rt17 ->overruns...
2006-03-06 17:19 ` [Alsa-user] arecord under 2.6.15.4-rt17 ->overruns Guennadi Liakhovetski
@ 2006-03-07 18:30 ` Guennadi Liakhovetski
2006-03-07 18:48 ` Lee Revell
0 siblings, 1 reply; 11+ messages in thread
From: Guennadi Liakhovetski @ 2006-03-07 18:30 UTC (permalink / raw)
To: Lee Revell; +Cc: Sergei Steshenko, alsa-user, linux-kernel, gregkh
On Mon, 6 Mar 2006, Guennadi Liakhovetski wrote:
> in drivers/pci/quirks.c:
>
> /*
> * VIA Apollo KT133 needs PCI latency patch
> * Made according to a windows driver based patch by George E. Breese
> * see PCI Latency Adjust on http://www.viahardware.com/download/viatweak.shtm
> * Also see http://www.au-ja.org/review-kt133a-1-en.phtml for
> * the info on which Mr Breese based his work.
> *
> * Updated based on further information from the site and also on
> * information provided by VIA
> */
>
> and later:
>
> /*
> * Ok we have the problem. Now set the PCI master grant to
> * occur every master grant. The apparent bug is that under high
> * PCI load (quite common in Linux of course) you can get data
> * loss when the CPU is held off the bus for 3 bus master requests
> * This happens to include the IDE controllers....
> *
> * VIA only apply this fix when an SB Live! is present but under
> * both Linux and Windows this isnt enough, and we have seen
> * corruption without SB Live! but with things like 3 UDMA IDE
> * controllers. So we ignore that bit of the VIA recommendation..
> */
>
> pci_read_config_byte(dev, 0x76, &busarb);
> /* Set bit 4 and bi 5 of byte 76 to 0x01
> "Master priority rotation on every PCI master grant */
> busarb &= ~(1<<5);
> busarb |= (1<<4);
> pci_write_config_byte(dev, 0x76, busarb);
> printk(KERN_INFO "Applying VIA southbridge workaround\n");
>
> which in itself is wrong - the comment says "set bit 5 & 6 to 1", but
> the code clears bit 6. Anyway, on my PC:
>
> 0000:00:07.0 ISA bridge: VIA Technologies, Inc. VT82C686 [Apollo Super South] (rev 40)
> 00: 06 11 86 06 87 00 10 02 40 00 01 06 00 00 80 00
> 10: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
> 20: 00 00 00 00 00 00 00 00 00 00 00 00 43 10 e7 80
> 30: 00 00 00 00 c0 00 00 00 00 00 00 00 00 00 00 00
> 40: 08 01 00 00 00 80 62 e6 01 01 44 00 00 00 f0 f3
> 50: 0e 76 34 00 00 b0 5a 90 00 04 ff 08 d0 00 00 00
> 60: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
> 70: 00 00 00 00 21 88 42 02 04 e4 00 00 00 00 00 00
> ^^
> 80: 01 00 00 00 00 09 00 00 00 60 00 00 00 00 00 00
> 90: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
> a0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
> b0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
> c0: 01 00 02 00 00 00 00 00 00 00 00 00 00 00 00 00
> d0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
> e0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
> f0: 00 00 00 00 00 00 41 00 00 00 00 00 00 00 00 00
>
> So, bit 4 is not set... I added busarb to the printk above and it prints
> 0xd2. Why is it reset to 0x42 again later??? Found nothing in
> drivers/ide/pci/via82cxxx.c... I could read it back to check it did get
> written.
Ok, my mistake. The quirk checks the southbridge revision, but configures
byte 0x76 on the host (north) bridge, where it does get successfully set
and remains set. But it doesn't seem to help. Still, at least the comment
is wrong - it contradicts the code. Who is the author of that quirk? Any
comments whether my problem seems similar to what others observed with
this chipset?
And my audio still doesn't work properly...
Thanks
Guennadi
---
Guennadi Liakhovetski
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: [Alsa-user] arecord under 2.6.15.4-rt17 ->overruns...
2006-03-07 18:30 ` Guennadi Liakhovetski
@ 2006-03-07 18:48 ` Lee Revell
2006-03-07 19:00 ` Guennadi Liakhovetski
0 siblings, 1 reply; 11+ messages in thread
From: Lee Revell @ 2006-03-07 18:48 UTC (permalink / raw)
To: Guennadi Liakhovetski; +Cc: Sergei Steshenko, alsa-user, linux-kernel, gregkh
On Tue, 2006-03-07 at 19:30 +0100, Guennadi Liakhovetski wrote:
> Ok, my mistake. The quirk checks the southbridge revision, but configures
> byte 0x76 on the host (north) bridge, where it does get successfully set
> and remains set. But it doesn't seem to help. Still, at least the comment
> is wrong - it contradicts the code. Who is the author of that quirk? Any
> comments whether my problem seems similar to what others observed with
> this chipset?
>
I doubt that issue has anything to do with your problem.
> And my audio still doesn't work properly...
Yes you've mentioned that several times.
Lee
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: [Alsa-user] arecord under 2.6.15.4-rt17 ->overruns...
2006-03-07 18:48 ` Lee Revell
@ 2006-03-07 19:00 ` Guennadi Liakhovetski
2006-03-07 19:15 ` Lee Revell
0 siblings, 1 reply; 11+ messages in thread
From: Guennadi Liakhovetski @ 2006-03-07 19:00 UTC (permalink / raw)
To: Lee Revell; +Cc: Sergei Steshenko, alsa-user, linux-kernel, gregkh
On Tue, 7 Mar 2006, Lee Revell wrote:
> On Tue, 2006-03-07 at 19:30 +0100, Guennadi Liakhovetski wrote:
>
> > And my audio still doesn't work properly...
>
> Yes you've mentioned that several times.
Ok, sorry, I just wanted to come to some resolution - either a fix or an
agreement that it's unfixable. So, if there are no further ideas, I'll
just open an entry on bugzilla and be done with it.
Thanks for trying to help me, everyone, sorry, if I abused your help in
any way:-)
Guennadi
---
Guennadi Liakhovetski
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: [Alsa-user] arecord under 2.6.15.4-rt17 ->overruns...
2006-03-07 19:00 ` Guennadi Liakhovetski
@ 2006-03-07 19:15 ` Lee Revell
2006-03-09 22:41 ` Guennadi Liakhovetski
0 siblings, 1 reply; 11+ messages in thread
From: Lee Revell @ 2006-03-07 19:15 UTC (permalink / raw)
To: Guennadi Liakhovetski; +Cc: Sergei Steshenko, alsa-user, linux-kernel, gregkh
On Tue, 2006-03-07 at 20:00 +0100, Guennadi Liakhovetski wrote:
> On Tue, 7 Mar 2006, Lee Revell wrote:
>
> > On Tue, 2006-03-07 at 19:30 +0100, Guennadi Liakhovetski wrote:
> >
> > > And my audio still doesn't work properly...
> >
> > Yes you've mentioned that several times.
>
> Ok, sorry, I just wanted to come to some resolution - either a fix or an
> agreement that it's unfixable. So, if there are no further ideas, I'll
> just open an entry on bugzilla and be done with it.
>
> Thanks for trying to help me, everyone, sorry, if I abused your help in
> any way:-)
Does the OSS driver have the same problem?
Lee
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: [Alsa-user] arecord under 2.6.15.4-rt17 ->overruns...
2006-03-07 19:15 ` Lee Revell
@ 2006-03-09 22:41 ` Guennadi Liakhovetski
2006-03-09 22:46 ` Lee Revell
0 siblings, 1 reply; 11+ messages in thread
From: Guennadi Liakhovetski @ 2006-03-09 22:41 UTC (permalink / raw)
To: Lee Revell; +Cc: Sergei Steshenko, alsa-user, linux-kernel, gregkh
On Tue, 7 Mar 2006, Lee Revell wrote:
> Does the OSS driver have the same problem?
Surprise - I was not able to reproduce the problem with oss. Whereas with
alsa I was able to lock my PC again. What I do - just load respective
drivers and either "jackd -v -d alsa" or "jackd -v -d oss". And then just
put some load in the background + try to start ardour... With alsa I
wasn't even able to start it. With oss it did run, and no xruns reported
from jackd. Normal non-rt kernel. jackd started without --realtime.
Ouch
Thanks
Guennadi
---
Guennadi Liakhovetski
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: [Alsa-user] arecord under 2.6.15.4-rt17 ->overruns...
2006-03-09 22:41 ` Guennadi Liakhovetski
@ 2006-03-09 22:46 ` Lee Revell
2006-03-09 22:56 ` Adrian Bunk
2006-03-09 23:15 ` Guennadi Liakhovetski
0 siblings, 2 replies; 11+ messages in thread
From: Lee Revell @ 2006-03-09 22:46 UTC (permalink / raw)
To: Guennadi Liakhovetski
Cc: Sergei Steshenko, alsa-user, linux-kernel, gregkh, Adrian Bunk
On Thu, 2006-03-09 at 23:41 +0100, Guennadi Liakhovetski wrote:
> On Tue, 7 Mar 2006, Lee Revell wrote:
>
> > Does the OSS driver have the same problem?
>
> Surprise - I was not able to reproduce the problem with oss. Whereas with
> alsa I was able to lock my PC again. What I do - just load respective
> drivers and either "jackd -v -d alsa" or "jackd -v -d oss". And then just
> put some load in the background + try to start ardour... With alsa I
> wasn't even able to start it. With oss it did run, and no xruns reported
> from jackd. Normal non-rt kernel. jackd started without --realtime.
>
> Ouch
>
OK, please file a report in the ALSA bug tracker against this driver.
Adrian, please add VIA to your list of OSS drivers that need to remain in the kernel.
Lee
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: [Alsa-user] arecord under 2.6.15.4-rt17 ->overruns...
2006-03-09 22:46 ` Lee Revell
@ 2006-03-09 22:56 ` Adrian Bunk
2006-03-10 19:36 ` Guennadi Liakhovetski
2006-03-09 23:15 ` Guennadi Liakhovetski
1 sibling, 1 reply; 11+ messages in thread
From: Adrian Bunk @ 2006-03-09 22:56 UTC (permalink / raw)
To: Lee Revell
Cc: Guennadi Liakhovetski, Sergei Steshenko, alsa-user, linux-kernel,
gregkh
On Thu, Mar 09, 2006 at 05:46:56PM -0500, Lee Revell wrote:
> On Thu, 2006-03-09 at 23:41 +0100, Guennadi Liakhovetski wrote:
> > On Tue, 7 Mar 2006, Lee Revell wrote:
> >
> > > Does the OSS driver have the same problem?
> >
> > Surprise - I was not able to reproduce the problem with oss. Whereas with
> > alsa I was able to lock my PC again. What I do - just load respective
> > drivers and either "jackd -v -d alsa" or "jackd -v -d oss". And then just
> > put some load in the background + try to start ardour... With alsa I
> > wasn't even able to start it. With oss it did run, and no xruns reported
> > from jackd. Normal non-rt kernel. jackd started without --realtime.
> >
> > Ouch
> >
>
> OK, please file a report in the ALSA bug tracker against this driver.
>
> Adrian, please add VIA to your list of OSS drivers that need to remain in the kernel.
As soon as I get a bug number from the ALSA bug tracker I'll add
SOUND_VIA82CXXX to my list of OSS drivers that should stay.
> Lee
cu
Adrian
--
"Is there not promise of rain?" Ling Tan asked suddenly out
of the darkness. There had been need of rain for many days.
"Only a promise," Lao Er said.
Pearl S. Buck - Dragon Seed
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: [Alsa-user] arecord under 2.6.15.4-rt17 ->overruns...
2006-03-09 22:46 ` Lee Revell
2006-03-09 22:56 ` Adrian Bunk
@ 2006-03-09 23:15 ` Guennadi Liakhovetski
1 sibling, 0 replies; 11+ messages in thread
From: Guennadi Liakhovetski @ 2006-03-09 23:15 UTC (permalink / raw)
To: Lee Revell; +Cc: Sergei Steshenko, alsa-user, linux-kernel, gregkh, Adrian Bunk
On Thu, 9 Mar 2006, Lee Revell wrote:
> OK, please file a report in the ALSA bug tracker against this driver.
Yep, I will. I am afraid, I lied to you at one place - as I said that
2.4.32 didn't work either. I tested 2.4.32, but used drivers from
alsa-driver-1.0.3. I wasn't able to compile any recent version of
alsa-library against 2.4.x native alsa drivers. I might try some older
version of alsa-lib. I'll try to put as much information as possible in
the bug-report.
Thanks
Guennadi
---
Guennadi Liakhovetski
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: [Alsa-user] arecord under 2.6.15.4-rt17 ->overruns...
2006-03-09 22:56 ` Adrian Bunk
@ 2006-03-10 19:36 ` Guennadi Liakhovetski
2006-03-10 23:35 ` Adrian Bunk
0 siblings, 1 reply; 11+ messages in thread
From: Guennadi Liakhovetski @ 2006-03-10 19:36 UTC (permalink / raw)
To: Adrian Bunk; +Cc: Lee Revell, Sergei Steshenko, alsa-user, linux-kernel, gregkh
On Thu, 9 Mar 2006, Adrian Bunk wrote:
> As soon as I get a bug number from the ALSA bug tracker I'll add
> SOUND_VIA82CXXX to my list of OSS drivers that should stay.
Bug-ID 0001906:
https://bugtrack.alsa-project.org/alsa-bug/view.php?id=1906 (requires
login)
Thanks
Guennadi
---
Guennadi Liakhovetski
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: [Alsa-user] arecord under 2.6.15.4-rt17 ->overruns...
2006-03-10 19:36 ` Guennadi Liakhovetski
@ 2006-03-10 23:35 ` Adrian Bunk
0 siblings, 0 replies; 11+ messages in thread
From: Adrian Bunk @ 2006-03-10 23:35 UTC (permalink / raw)
To: Guennadi Liakhovetski
Cc: Lee Revell, Sergei Steshenko, alsa-user, linux-kernel, gregkh
On Fri, Mar 10, 2006 at 08:36:43PM +0100, Guennadi Liakhovetski wrote:
> On Thu, 9 Mar 2006, Adrian Bunk wrote:
>
> > As soon as I get a bug number from the ALSA bug tracker I'll add
> > SOUND_VIA82CXXX to my list of OSS drivers that should stay.
>
> Bug-ID 0001906:
>...
Thanks, noted.
> Thanks
> Guennadi
cu
Adrian
--
"Is there not promise of rain?" Ling Tan asked suddenly out
of the darkness. There had been need of rain for many days.
"Only a promise," Lao Er said.
Pearl S. Buck - Dragon Seed
^ permalink raw reply [flat|nested] 11+ messages in thread
end of thread, other threads:[~2006-03-10 23:35 UTC | newest]
Thread overview: 11+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
[not found] <Pine.LNX.4.60.0603022032040.4969@poirot.grange>
[not found] ` <1141331113.3042.5.camel@mindpipe>
[not found] ` <Pine.LNX.4.60.0603022132160.4969@poirot.grange>
[not found] ` <1141333305.3042.14.camel@mindpipe>
[not found] ` <Pine.LNX.4.60.0603022207160.3033@poirot.grange>
[not found] ` <1141334604.3042.17.camel@mindpipe>
[not found] ` <Pine.LNX.4.60.0603022226130.3033@poirot.grange>
[not found] ` <1141335418.3042.25.camel@mindpipe>
[not found] ` <Pine.LNX.4.60.0603030012070.3397@poirot.grange>
[not found] ` <1141342018.3042.40.camel@mindpipe>
[not found] ` <Pine.LNX.4.60.0603030707270.2959@poirot.grange>
[not found] ` <1141410043.3042.116.camel@mindpipe>
[not found] ` <Pine.LNX.4.60.0603041429340.3283@poirot.grange>
[not found] ` <20060304154357.74f74cac@localhost>
[not found] ` <Pine.LNX.4.60.0603041823560.3601@poirot.grange>
[not found] ` <1141495123.3042.181.camel@mindpipe>
[not found] ` <Pine.LNX.4.60.0603042046450.3135@poirot.grange>
[not found] ` <1141509605.14714.11.camel@mindpipe>
2006-03-06 17:19 ` [Alsa-user] arecord under 2.6.15.4-rt17 ->overruns Guennadi Liakhovetski
2006-03-07 18:30 ` Guennadi Liakhovetski
2006-03-07 18:48 ` Lee Revell
2006-03-07 19:00 ` Guennadi Liakhovetski
2006-03-07 19:15 ` Lee Revell
2006-03-09 22:41 ` Guennadi Liakhovetski
2006-03-09 22:46 ` Lee Revell
2006-03-09 22:56 ` Adrian Bunk
2006-03-10 19:36 ` Guennadi Liakhovetski
2006-03-10 23:35 ` Adrian Bunk
2006-03-09 23:15 ` Guennadi Liakhovetski
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox