Kernel Bug Tracker – Bug 7228
Disabling ACPI is the only way to enable sound on certain Conexant-codec hda-intel platforms
Last modified: 2008-04-07 00:59:01 UTC
Most recent kernel where this bug did not occur: -
Distribution: Fedora Core 5 (with hand-rolled mainline 2.6.18 kernel)
Hardware Environment: Toshiba Satellite Pro P100 laptop
Software Environment: ALSA-1.013rc3 / snd-hda-intel
This bug has already been reported to the ALSA bugtracker, but they have
referred us to the Kernel bugzilla because "Drivers are not required to know
anything about ACPI, so by definition it's an ACPI problem. It should be
reported to bugzilla.kernel.org. Once the ACPI bug is resolved we can start to
debug the sound issue." ALSA bug reference:
This laptop has the following Conexant sound chipset (HDA Intel):
[carl@localhost ~]$ head -5 /proc/asound/card0/codec#0
Codec: Conexant ID 5047
Vendor Id: 0x14f15047
Subsystem Id: 0x1179ff31
Revision Id: 0x100000
When booting normally (or with 'pci=noacpi', 'acpi_irq_balance',
'acpi_irq_nobalance', 'acpi=oldboot'), all ALSA drivers load without errors or
warnings. However, when playing sounds or music, the speakers remain silent.
When booting with 'acpi=off' or 'acpi=ht', sound can be heard but it is very
quiet and sounds distorted. Also, sound only comes from the left channel.
Steps to reproduce: Boot linux, play sound (leads to failure). Boot linux with
'acpi=off' kernel parameter, play sound (leads to quiet, bad quality sound).
Same problem with ubuntu dapper (kernel 2.6.15-26-686) on a toshiba P100 197.
Same problem with Debian Sid (kernel 2.6.18-2-686) on a Toshiba L35 notebook
with similar audio chipset (ATI SB 450 / Realtek ALC 861)
Same problem with debian etch, kernel version 2.6.18 and toshiba satellite
p100-119 (Codec: Conexant ID 5047, Address: 0, Vendor Id: 0x14f15047, Subsystem
Id: 0x1179ff31, Revision Id: 0x100000).
Are there any news on this issue ?
It's a rather serious issue for users since it's quite well hidden, while
preventing devices from working (you can't detect it at first look, you have
to look at each device and try it).
Is there any hope for a fix in the 2.6.19 kernel ?
I have reported similar problem with my notebook. See bug #6575
This is not a Linux kernel bug.
The DSDT is buggy.
The way to fix this issue is described here :
As you can see, there is a test in the DSDT. If the OS is linux, then, I don't
know why, an IRQ conflict seems to be created... If you change the DSDT
to "fake" windows vista, it works :/
Really silly DSDT issue, nothing Linux has to care about.
Anyway, this issue still shows a problem with pci=noacpi. With this setting,
there shouldn't be any IRQ conflict ...
same problem for me too -
i am using alsa & libs from from kernel 184.108.40.206 SMP of slackware 12 RC1 hugesmp.s on a toshiba p100-209 PSPA3E
with acpi=off sound works ok, a little bit distorted on high levels (i think its normal...)
with acpi didn't work -
i had to remove some lines from /etc/pcmcia/config.opts:
# These are the official ports to use from pcmcia-cs:
#include port 0x100-0x4ff, port 0x800-0x8ff, port 0xc00-0xcff
# However, ports 0x810-0x81f hurt on some DELL machines and
# ports 0x3b0-0x3df hurt on some FSC machines, so we use this port
# list instead:
include port 0x100-0x3af, port 0x820-0x8ff, port 0xc00-0xcff
# ports removed below
# port 0x3e0-0x4ff, port 0x800-0x80f
i removed these 2 ports from "include port ..." line and could boot.
is it a acpi problem + wrong irq assign ???
root@igorpc:/home/igor# head -5 /proc/asound/card0/codec#0
Codec: Conexant CX20551 (Waikiki)
Vendor Id: 0x14f15047
Subsystem Id: 0x1179ff31
Revision Id: 0x100000
root@igorpc:/home/igor# lspci -nv
00:1b.0 0403: 8086:27d8 (rev 02)
Flags: bus master, fast devsel, latency 0, IRQ 21
Memory at d2500000 (64-bit, non-prefetchable) [size=16K]
for those who have the issue fixed with DSDT override, can you please attach your acpidump? If _OSI override is needed, this means either BIOS is buggy or Linux lacks some capability. We should make sure this isn't the later case.
From the bug 6575, somebody suggests linux doesn't power on something, this might be Linux doesn't call _PS0 for sound card or something.
*** Bug 7515 has been marked as a duplicate of this bug. ***
*** Bug 6575 has been marked as a duplicate of this bug. ***
As of 220.127.116.11 bug still present (I'm form bug #6575 marked as duplicate of this one). Is there something I can do to shade some light on the issue?
(In reply to comment #11)
> As of 18.104.22.168 bug still present (I'm form bug #6575 marked as duplicate of
> this one). Is there something I can do to shade some light on the issue?
Marcin, have you tried to override the DSDT as in comment# 6? would you please help to provide what shaohua needed in comment# 8? thanks!
Yes I tried DSDT override as suggested in comment #6. There is some OS cheking in DSDT table but looks harmless to me. Anyway I have commented out whole block of the code in DSDT and hardcoded values for Windows XP SP2 (values for Vista are not in the table). This does not change anything. What type of acpidump is needed? Comment #8 is not clear in this aspect.
Toshiba has a BIOS update. I installed it and now sound works with ACPI.
please attach the output of 'acpidump' tools. If your distro hasn't it, you can get it from http://www.kernel.org/pub/linux/kernel/people/lenb/acpi/utils/pmtools-20071116.tar.bz2
For those who can workaround the issue with overrided DSDT, please attach the acpidump output here.
Ok, somebody posted the DSDT in the gentoo bugzilla. It appears the OSYS setting is different per OS string, but the DSDT seems not doing different things per OSYS, maybe SMM mode code does.
Can anyone attach the output with 'lspci -vvvxxxx' of the system in the two cases w/wo 'acpi=off'
I do not own this laptop any more.
Created attachment 14581 [details]
lspci -vvvxxxx from a toshiba satellite p100-j01 that exhibited this problem.
Created attachment 14582 [details]
lspci -vvvxxxx from a toshiba satellite p100-j01 that exhibited this problem, with acpi=off
I usually boot with acpi_osi="!Linux" on my machine (toshiba satellite p100-j01) but the above attachments were created on that machine running kernel 2.6.22-gentoo-r10 without any acpi_osi setting or dsdt patching. BIOS version 3.30 was used.
Sorry for the late response. Jack, if you use pci=nommconf. It appears there isn't different with acpi enabled except interrupt is different. Also, please attach your acpidump.
I'm (sadly) no longer running Linux on the machine any more. I intend to in the future, but I'm not sure when I'll get around to reinstalling.
When I was having this problem, this thread on forums.gentoo.org looked quite promising: http://forums.gentoo.org/viewtopic-t-606767-highlight-sound+toshiba+p100.html
This gentoo bug is also relevant, I think: https://bugs.gentoo.org/show_bug.cgi?id=201271
And it looks like it may have been fixed.
If nobody provides further info, I'll mark this track invalid.