Skip to contentSkip to conversation

ALSA 1.0.0pre1 is released

ANN Takashi Iwai 7 participants 21 messages Nov 19, 2003 – Dec 1, 2003

lore
Share & tools
lore
#0 lore raw

Hi all,

as you might have already seen on the web news page, we released 1.0.0pre1 tarball. This is intended for the wide tests before the official 1.0.0 release. We planned to release 1.0.0 (hopefully) in this month, so please everyone test this version and give bug reports now.

Especially, if you have a card or a mobo with ICH (or compatible) chip and *multiple* AC97 codecs, testing this version is appreciated. Please give (also positive) feedback whether it works. The handling of multiple codecs was changed recently, but not tested well because of lack of hardware.

Important changes / fixes since 0.9.8

- joystick is configured now (again) by module options instead of control API. see alsa-kernel/Documentation/Joystick.txt.

- the default value of cards_limit option of sound module is 1. note that this doesn't mean that the max. number of cards is limited to 1, but it restricts the number of auto-loaded cards. if you load a module manually, it's added safely.

- PCM OSS emulation was fixed, e.g. mplayer delay problem should be fixed now. Hopefully this doesn't introduce another bug...

- AC97 handler is rewritten. please test!

- some serious ISA PnP bugs are fixed (e.g. cs4236 freeze).

- some bugs in OSS-emulation library are fixed. please test!

- dmix problems with xmms ALSA plugin is solved.

- alsa-driver tree can be built with 2.6 kernel.

- alsaconf is now moved to alsa-utils tarball.

Known problems

- pdplus driver seems broken. more testers?

- korg1212 driver seems broken (somehow).

- PCM OSS emulation might cause oops when you compile the kernel without frame pointer (depending on the compiler).

- opti92x still might not work (although some bugs were fixed).

--
Takashi Iwai <[email protected]> ALSA Developer - www.alsa-project.org

-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program. Does SourceForge.net help you be more productive? Does it help you create better code? SHARE THE LOVE, and help us help YOU! Click Here: http://sourceforge.net/donate/

#1 lore raw

Was the patching described by Eric Hathaway's post done for the new version, or was it decided it wasn't really necessary? See: [Alsa-devel] aureal driver compile problem and patch The patch is copied below for your convenience.

I would also like to suggest that the INSTALL file and read me files have specific references to the pages at:
http://alsa.opensrc.org/index.php?page=AlsaDrivers
In my case the ones for Aureal
(http://alsa.opensrc.org/index.php?page=au88x0 and links from there) were invaluable.
I would also like to sugget the addition of a NewToLinuxReadThis file with a quick overview that will get newbies who have just installed a distro and can't get their soundcard to work up and running in the easiest possible way. Maybe including the reference to the
http://alsa.opensrc.org/index.php?page=AlsaDrivers will do.

Some other mods that would be good for Newbies is if alsamixergui is package independent and had an obvious Save Settings button that initiated a script that did away with the need to directly use alsactl and to modify the modules.conf file. It could also pop up a reference to the
http://alsa.opensrc.org/index.php?page=AlsaDrivers page on starting.

I would also like to suggest a gui for the install process with scripting that did the detection of card, whether kernel-source is installed, whether sound is loaded, whether oss is needed, whether tools are needed (based on soundcard), doing ./snddevices, checking for/getting, installing and setting unmuting/volumes for alsamixergui - the whole process should be able to be made a lot easier for people who just got a distro off the front cover of a computer mag. For these people ease of installation and use of sound and surfing (the winmodem problem) are likely to be a big deal right at the start.
cheers
Paul
--------------- BEGIN PATCH -----------------------
diff -ur alsa-driver-0.9.8/acore/Makefile alsa-driver-0.9.8-patched/acore/Makefile

acore/Makefile Tue Oct 21 22:21:12 2003 renamed from acore/Makefile Tue Sep 16 06:14:03 2003 +0 -0
@@ -10,6 +10,9 @@

obj-$(CONFIG_SND_SERIALMIDI) += snd-rawmidi.o snd.o obj-$(CONFIG_SND_MSND_PINNACLE) += snd-pcm.o snd-timer.o snd-page-alloc.o snd.o snd-rawmidi.o
obj-$(CONFIG_SND_USB_US428) += snd-pcm.o snd-timer.o snd-page-alloc.o snd.o snd-rawmidi.o snd-hwdep.o
+obj-$(CONFIG_SND_AU8810) += snd-pcm.o snd-timer.o snd-page-alloc.o snd.o snd-rawmidi.o
+obj-$(CONFIG_SND_AU8820) += snd-pcm.o snd-timer.o snd-page-alloc.o snd.o snd-rawmidi.o
+obj-$(CONFIG_SND_AU8830) += snd-pcm.o snd-timer.o snd-page-alloc.o snd.o snd-rawmidi.o

export-objs := sound.o pcm.o pcm_lib.o rawmidi.o timer.o hwdep.o memalloc.o

diff -ur alsa-driver-0.9.8/acore/seq/Makefile alsa-driver-0.9.8-patched/acore/seq/Makefile
--- alsa-driver-0.9.8/acore/seq/Makefile Mon Jun 2 09:03:38 2003 +++ alsa-driver-0.9.8-patched/acore/seq/Makefile Tue Oct 21 22:21:20 2003@@ -7,6 +7,9 @@

obj-$(CONFIG_SND_SERIALMIDI) += snd-seq-midi.o snd-seq.o snd-seq-device.o snd-seq-midi-event.o
obj-$(CONFIG_SND_MSND_PINNACLE) += snd-seq-midi.o snd-seq.o snd-seq-device.o snd-seq-midi-event.o
+obj-$(CONFIG_SND_AU8810) += snd-seq-midi.o snd-seq.o snd-seq-device.o snd-seq-midi-event.o snd-seq-instr.o
+obj-$(CONFIG_SND_AU8820) += snd-seq-midi.o snd-seq.o snd-seq-device.o snd-seq-midi-event.o snd-seq-instr.o
+obj-$(CONFIG_SND_AU8830) += snd-seq-midi.o snd-seq.o snd-seq-device.o snd-seq-midi-event.o snd-seq-instr.o

export-objs := seq_device.o seq.o seq_instr.o seq_midi_emul.o \
seq_midi_event.o seq_virmidi.o
diff -ur alsa-driver-0.9.8/drivers/mpu401/Makefile alsa-driver-0.9.8-patched/drivers/mpu401/Makefile
--- alsa-driver-0.9.8/drivers/mpu401/Makefile Mon Jun 2 09:03:39 2003 +++ alsa-driver-0.9.8-patched/drivers/mpu401/Makefile Tue Oct 21 22:21:36 2003@@ -3,6 +3,10 @@
include $(TOPDIR)/toplevel.config
include $(TOPDIR)/Makefile.conf

+obj-$(CONFIG_SND_AU8810) += snd-mpu401-uart.o
+obj-$(CONFIG_SND_AU8820) += snd-mpu401-uart.o
+obj-$(CONFIG_SND_AU8830) += snd-mpu401-uart.o
+
export-objs := mpu401_uart.o

TOPDIR = $(MAINSRCDIR)
diff -ur alsa-driver-0.9.8/pci/ac97/Makefile
alsa-driver-0.9.8-patched/pci/ac97/Makefile

pci/ac97/Makefile Tue Oct 21 22:21:49 2003 renamed from pci/ac97/Makefile Sun Feb 9 13:35:57 2003 +0 -0
@@ -3,6 +3,10 @@

include $(TOPDIR)/toplevel.config
include $(TOPDIR)/Makefile.conf

+obj-$(CONFIG_SND_AU8810) += snd-ac97-codec.o
+obj-$(CONFIG_SND_AU8820) += snd-ac97-codec.o
+obj-$(CONFIG_SND_AU8830) += snd-ac97-codec.o
+
export-objs := ac97_codec.o ak4531_codec.o

TOPDIR = $(MAINSRCDIR)
--------------- END PATCH -----------------------

----- Original Message -----
From: "Takashi Iwai" <[email protected]>
To: <[email protected]>
Cc: <[email protected]>
Sent: Thursday, November 20, 2003 2:46 AM
Subject: [Alsa-devel] [ANN] ALSA 1.0.0pre1 is released

Quoted 57 lines

Hi all,

as you might have already seen on the web news page, we released 1.0.0pre1 tarball. This is intended for the wide tests before the official 1.0.0 release. We planned to release 1.0.0 (hopefully) in this month, so please everyone test this version and give bug reports now.

Especially, if you have a card or a mobo with ICH (or compatible) chip and *multiple* AC97 codecs, testing this version is appreciated. Please give (also positive) feedback whether it works. The handling of multiple codecs was changed recently, but not tested well because of lack of hardware.

Important changes / fixes since 0.9.8

- joystick is configured now (again) by module options instead of control API. see alsa-kernel/Documentation/Joystick.txt.

- the default value of cards_limit option of sound module is 1. note that this doesn't mean that the max. number of cards is limited to 1, but it restricts the number of auto-loaded cards. if you load a module manually, it's added safely.

- PCM OSS emulation was fixed, e.g. mplayer delay problem should be fixed now. Hopefully this doesn't introduce another bug...

- AC97 handler is rewritten. please test!

- some serious ISA PnP bugs are fixed (e.g. cs4236 freeze).

- some bugs in OSS-emulation library are fixed. please test!

- dmix problems with xmms ALSA plugin is solved.

- alsa-driver tree can be built with 2.6 kernel.

- alsaconf is now moved to alsa-utils tarball.

Known problems

- pdplus driver seems broken. more testers?

- korg1212 driver seems broken (somehow).

- PCM OSS emulation might cause oops when you compile the kernel without frame pointer (depending on the compiler).

- opti92x still might not work (although some bugs were fixed).

--
Takashi Iwai <[email protected]> ALSA Developer - www.alsa-project.org

Quoted 10 lines

-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program. Does SourceForge.net help you be more productive? Does it help you create better code? SHARE THE LOVE, and help us help YOU! Click Here: http://sourceforge.net/donate/
_______________________________________________
Alsa-devel mailing list [email protected]
https://lists.sourceforge.net/lists/listinfo/alsa-devel

-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program. Does SourceForge.net help you be more productive? Does it help you create better code? SHARE THE LOVE, and help us help YOU! Click Here: http://sourceforge.net/donate/

#2 lore raw

At Thu, 20 Nov 2003 09:39:30 +1100,

Quoted 5 lines · Paul Hanly (Hotmail)

Was the patching described by Eric Hathaway's post done for the new version, or was it decided it wasn't really necessary? See: [Alsa-devel] aureal driver compile problem and patch The patch is copied below for your convenience.

it (the fixed one) is already in 1.0.0pre1 tarball.

Takashi

-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program. Does SourceForge.net help you be more productive? Does it help you create better code? SHARE THE LOVE, and help us help YOU! Click Here: http://sourceforge.net/donate/

#3 lore raw
In reply to Takashi Iwai
Quoted 5 lines

as you might have already seen on the web news page, we released 1.0.0pre1 tarball. This is intended for the wide tests before the official 1.0.0 release. We planned to release 1.0.0 (hopefully) in this month, so please everyone test this version and give bug reports now.

Looks like somehow hdspmix and hdspconf did not make it to the alsa-tools tarball (but they are there in CVS).... -- Fernando

-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program. Does SourceForge.net help you be more productive? Does it help you create better code? SHARE THE LOVE, and help us help YOU! Click Here: http://sourceforge.net/donate/

#4 lore raw

Re: ALSA 1.0.0pre1 is released - intel8x0 is oopsing (2.6)

Just tried ALSA CVS and 2.6.0-test9 current bk.
CVS snd-intel8x0 is oopsing.

Unable to handle kernel NULL pointer dereference at virtual address 00000040
 printing eip:
e1a3a9e6
*pde = 00000000
Oops: 0000 [#1]
CPU:    0
EIP:    0060:[<e1a3a9e6>]    Not tainted
EFLAGS: 00010202

EIP is at snd_intel8x0_mixer+0x2b6/0x400 [snd_intel8x0]
eax: 00000020 ebx: dd8ecc00 ecx: 00000008 edx: 00000080
esi: dc13f4c0 edi: 00000001 ebp: 00000020 esp: dc137c74
ds: 007b es: 007b ss: 0068
Process modprobe (pid: 887, threadinfo=dc136000 task=dec572e0)

Stack: dd8ecc00 00000080 e1a3f540 00000005 df678400 dda04a00 00000000 dd8ecc00
       e1a3a710 00000000 c1549400 00000000 00000000 00000000 00000000 
00000000
       00000000 00000000 00000000 00000000 00000000 00000000 00000000 
00000000
Call Trace:
 [<e1a3a710>] snd_intel8x0_mixer_free_ac97+0x0/0x20 [snd_intel8x0]
 [<e1a3adc7>] snd_intel8x0_ich_chip_init+0x1d7/0x300 [snd_intel8x0]
 [<e1a39290>] snd_intel8x0_codec_write+0x0/0xd0 [snd_intel8x0]
 [<e1a39360>] snd_intel8x0_codec_read+0x0/0x110 [snd_intel8x0]
 [<e1a3a700>] snd_intel8x0_mixer_free_ac97_bus+0x0/0x10 [snd_intel8x0]
 [<e1a3bfea>] snd_intel8x0_probe+0x10a/0x280 [snd_intel8x0]
 [<c016c531>] dput+0x31/0x220
 [<c0202822>] pci_device_probe_static+0x52/0x70
 [<c020287c>] __pci_device_probe+0x3c/0x50
 [<c02028bc>] pci_device_probe+0x2c/0x50
 [<c02428df>] bus_match+0x3f/0x70
 [<c0242a0c>] driver_attach+0x5c/0xa0
 [<c0242cd3>] bus_add_driver+0x93/0xb0
 [<c024310f>] driver_register+0x2f/0x40
 [<c0202aac>] pci_register_driver+0x5c/0x90
 [<e195c013>] alsa_card_intel8x0_init+0x13/0x8e [snd_intel8x0]
 [<c0135a09>] sys_init_module+0x119/0x230
 [<c010b3d9>] sysenter_past_esp+0x52/0x71

Code: 0f b7 40 20 0f b6 84 83 c8 02 00 00 c0 e0 06 08 c1 eb cb 80

Anything else you need to know?

Quoted 14 lines · Fernando Pablo Lopez-Lezcano
Quoted 5 lines

as you might have already seen on the web news page, we released 1.0.0pre1 tarball. This is intended for the wide tests before the official 1.0.0 release. We planned to release 1.0.0 (hopefully) in this month, so please everyone test this version and give bug reports now.

Looks like somehow hdspmix and hdspconf did not make it to the alsa-tools tarball (but they are there in CVS).... -- Fernando

Quoted 10 lines

-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program. Does SourceForge.net help you be more productive? Does it help you create better code? SHARE THE LOVE, and help us help YOU! Click Here: http://sourceforge.net/donate/
_______________________________________________
Alsa-devel mailing list [email protected]
https://lists.sourceforge.net/lists/listinfo/alsa-devel

-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program. Does SourceForge.net help you be more productive? Does it help you create better code? SHARE THE LOVE, and help us help YOU! Click Here: http://sourceforge.net/donate/

#5 lore raw

Re: Re: ALSA 1.0.0pre1 is released - intel8x0 is oopsing (2.6)

At Thu, 20 Nov 2003 15:44:53 +0800,

Quoted 3 lines · R CHAN

Just tried ALSA CVS and 2.6.0-test9 current bk.
CVS snd-intel8x0 is oopsing.

could you try the attached patch?

Takashi

#6 lore raw

Re: Re: ALSA 1.0.0pre1 is released - intel8x0 is oopsing (2.6) - FIXED!

In reply to R CHAN

The oops has been fixed by the intel8x0 patch. Thanks.

/proc/asound/card0> cat ../cards
0 [I82801DBICH4 ]: ICH - Intel 82801DB-ICH4

                     Intel 82801DB-ICH4 at 0xdffffa00, irq 10

/proc/asound/card0> cat codec97#0/ac97#0-0
0-0/0: Realtek ALC202 rev 0

Capabilities     :
DAC resolution   : 20-bit
ADC resolution   : 18-bit
3D enhancement   : Realtek 3D Stereo Enhancement

Current setup
Mic gain : +0dB [+0dB]
POP path : pre 3D
Sim. stereo : off

3D enhancement   : off
Loudness         : off
Mono output      : MIX
Mic select       : Mic1

ADC/DAC loopback : off
Extended ID : codec=0 rev=1 AMAP DSA=0 SPDIF DRA VRA
Extended status : SPCV LDAC SDAC CDAC SPDIF=3/4 VRA
PCM front DAC : 48000Hz
PCM ADC : 48000Hz
SPDIF Control : Consumer PCM Category=0x2 Generation=1 Rate=48kHz

Quoted 77 lines · R CHAN

Just tried ALSA CVS and 2.6.0-test9 current bk.
CVS snd-intel8x0 is oopsing.

Unable to handle kernel NULL pointer dereference at virtual address 00000040
 printing eip:
e1a3a9e6
*pde = 00000000
Oops: 0000 [#1]
CPU:    0
EIP:    0060:[<e1a3a9e6>]    Not tainted
EFLAGS: 00010202

EIP is at snd_intel8x0_mixer+0x2b6/0x400 [snd_intel8x0]
eax: 00000020 ebx: dd8ecc00 ecx: 00000008 edx: 00000080
esi: dc13f4c0 edi: 00000001 ebp: 00000020 esp: dc137c74
ds: 007b es: 007b ss: 0068
Process modprobe (pid: 887, threadinfo=dc136000 task=dec572e0)

Stack: dd8ecc00 00000080 e1a3f540 00000005 df678400 dda04a00 00000000 dd8ecc00
       e1a3a710 00000000 c1549400 00000000 00000000 00000000 00000000 
00000000
       00000000 00000000 00000000 00000000 00000000 00000000 00000000 
00000000
Call Trace:
 [<e1a3a710>] snd_intel8x0_mixer_free_ac97+0x0/0x20 [snd_intel8x0]
 [<e1a3adc7>] snd_intel8x0_ich_chip_init+0x1d7/0x300 [snd_intel8x0]
 [<e1a39290>] snd_intel8x0_codec_write+0x0/0xd0 [snd_intel8x0]
 [<e1a39360>] snd_intel8x0_codec_read+0x0/0x110 [snd_intel8x0]
 [<e1a3a700>] snd_intel8x0_mixer_free_ac97_bus+0x0/0x10 [snd_intel8x0]
 [<e1a3bfea>] snd_intel8x0_probe+0x10a/0x280 [snd_intel8x0]
 [<c016c531>] dput+0x31/0x220
 [<c0202822>] pci_device_probe_static+0x52/0x70
 [<c020287c>] __pci_device_probe+0x3c/0x50
 [<c02028bc>] pci_device_probe+0x2c/0x50
 [<c02428df>] bus_match+0x3f/0x70
 [<c0242a0c>] driver_attach+0x5c/0xa0
 [<c0242cd3>] bus_add_driver+0x93/0xb0
 [<c024310f>] driver_register+0x2f/0x40
 [<c0202aac>] pci_register_driver+0x5c/0x90
 [<e195c013>] alsa_card_intel8x0_init+0x13/0x8e [snd_intel8x0]
 [<c0135a09>] sys_init_module+0x119/0x230
 [<c010b3d9>] sysenter_past_esp+0x52/0x71

Code: 0f b7 40 20 0f b6 84 83 c8 02 00 00 c0 e0 06 08 c1 eb cb 80

Anything else you need to know?

Quoted 15 lines · Fernando Pablo Lopez-Lezcano
Quoted 5 lines

as you might have already seen on the web news page, we released 1.0.0pre1 tarball. This is intended for the wide tests before the official 1.0.0 release. We planned to release 1.0.0 (hopefully) in this month, so please everyone test this version and give bug reports now.

Looks like somehow hdspmix and hdspconf did not make it to the alsa-tools tarball (but they are there in CVS).... -- Fernando

Quoted 10 lines

-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program. Does SourceForge.net help you be more productive? Does it help you create better code? SHARE THE LOVE, and help us help YOU! Click Here: http://sourceforge.net/donate/
_______________________________________________
Alsa-devel mailing list [email protected]
https://lists.sourceforge.net/lists/listinfo/alsa-devel

Quoted 10 lines

-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program. Does SourceForge.net help you be more productive? Does it help you create better code? SHARE THE LOVE, and help us help YOU! Click Here: http://sourceforge.net/donate/
_______________________________________________
Alsa-devel mailing list [email protected]
https://lists.sourceforge.net/lists/listinfo/alsa-devel

-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program. Does SourceForge.net help you be more productive? Does it help you create better code? SHARE THE LOVE, and help us help YOU! Click Here: http://sourceforge.net/donate/

#7 lore raw
In reply to Fernando Pablo Lopez-Lezcano

At 19 Nov 2003 17:30:29 -0800,

Quoted 9 lines · Fernando Pablo Lopez-Lezcano
Quoted 6 lines

as you might have already seen on the web news page, we released 1.0.0pre1 tarball. This is intended for the wide tests before the official 1.0.0 release. We planned to release 1.0.0 (hopefully) in this month, so please everyone test this version and give bug reports now.

Looks like somehow hdspmix and hdspconf did not make it to the alsa-tools tarball (but they are there in CVS)....

oh, that happens again.

i guess it's because there is no FLTK installed on Jaroslav's machine, and configure failed to create Makefile for them to pack.

Jaroslav, could you install fltk to put them all in?

Takashi

-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program. Does SourceForge.net help you be more productive? Does it help you create better code? SHARE THE LOVE, and help us help YOU! Click Here: http://sourceforge.net/donate/

#8 lore raw

1.0.0pre1 / ice1712 / mplayer-oss-delay

In reply to Takashi Iwai
Quoted 3 lines · Takashi Iwai

- PCM OSS emulation was fixed, e.g. mplayer delay problem should be fixed now. Hopefully this doesn't introduce another bug...

Strange, the new version doesn't seem to behave any different than the last cvs version I tested (Nov-02) (is there maybe any reason why make install wouldn't properly update everything?). Skipping in mplayer or xmms sill takes half a second, mpg123 needs >30 seconds for starting playback as well as java-apps do. I still have a DMX6fire and am running gentoo with kernel 2.4.20-gentoo-r5. If you need debugging infos please ask, I'm still not sure if the strace-info I sent you once was useful or not.

Quoted 2 lines

- some bugs in OSS-emulation library are fixed. please test!

Are any of those bugfixes related to the problems mentioned above? If not I don't really know where to start testing...

Quoted 3 lines

- dmix problems with xmms ALSA plugin is solved.

I think I can confirm that, I noticed some problems with sample rate conversion before that seem to be gone.

Again thanks a lot for your effort on these issues, Steffen

-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program. Does SourceForge.net help you be more productive? Does it help you create better code? SHARE THE LOVE, and help us help YOU! Click Here: http://sourceforge.net/donate/

#9 lore raw

Re: 1.0.0pre1 / ice1712 / mplayer-oss-delay

At Thu, 20 Nov 2003 11:01:24 +0100,

Quoted 6 lines · Steffen Sauder
Quoted 3 lines · Takashi Iwai

- PCM OSS emulation was fixed, e.g. mplayer delay problem should be fixed now. Hopefully this doesn't introduce another bug...

Quoted 8 lines

Strange, the new version doesn't seem to behave any different than the last cvs version I tested (Nov-02) (is there maybe any reason why make install wouldn't properly update everything?). Skipping in mplayer or xmms sill takes half a second, mpg123 needs >30 seconds for starting playback as well as java-apps do. I still have a DMX6fire and am running gentoo with kernel 2.4.20-gentoo-r5. If you need debugging infos please ask, I'm still not sure if the strace-info I sent you once was useful or not.

ah, that's a different bug which we got fixed.
i couldn't reproduce this bug yet on my machine, though.
i just checked your last post, and it looks like a problem specific to DMX6fire rather than the general kernel oss-emulation.
the app issues so many SET_FORMAT and RESET ioctls, and each takes a half second or so. it's likely a long i2c transfer or something like that.

Quoted 5 lines
Quoted 3 lines

- some bugs in OSS-emulation library are fixed. please test!

Are any of those bugfixes related to the problems mentioned above? If not I don't really know where to start testing...

i don't think so.

Takashi

-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program. Does SourceForge.net help you be more productive? Does it help you create better code? SHARE THE LOVE, and help us help YOU! Click Here: http://sourceforge.net/donate/

#10 lore raw
In reply to Takashi Iwai

At Wed, 19 Nov 2003 16:46:04 +0100,

Quoted 9 lines · I wrote:

Hi all,

as you might have already seen on the web news page, we released 1.0.0pre1 tarball. This is intended for the wide tests before the official 1.0.0 release. We planned to release 1.0.0 (hopefully) in this month, so please everyone test this version and give bug reports now.

Quoted 8 lines

Especially, if you have a card or a mobo with ICH (or compatible) chip and *multiple* AC97 codecs, testing this version is appreciated. Please give (also positive) feedback whether it works. The handling of multiple codecs was changed recently, but not tested well because of lack of hardware.

Important changes / fixes since 0.9.8

A couple of more important thing I forgot to mention about ALSA library update:

- the default PCM API is in the new style.

If your app still uses the PCM API in the old style (i.e. without defining ALSA_PCM_OLD_HW_PARAM_API), the compilation with the new library may fail. In such a case, fix the API call with the new style, or simply add

	#define ALSA_PCM_OLD_HW_PARAMS_API
	#define ALSA_PCM_OLD_SW_PARAMS_API

before inclusion of <alsa/asoundlib.h>. See alsa-lib/NOTES for details.

The binary compatibility is kept with the new alsa-lib, so the programs will run as well even after installing the new alsa-lib.

- SND_LIB_MAJOR is jumped to 1, and SND_LIB_MINOR is 0.

If your application checks the ALSA version with only SND_LIB_MINOR macro, you'll need to fix it.

--
Takashi Iwai <[email protected]> ALSA Developer - www.alsa-project.org

-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program. Does SourceForge.net help you be more productive? Does it help you create better code? SHARE THE LOVE, and help us help YOU! Click Here: http://sourceforge.net/donate/

#11 lore raw
In reply to Takashi Iwai
Quoted 5 lines · --- Takashi Iwai

Hi all,

- PCM OSS emulation might cause oops when you compile the kernel
without frame pointer (depending on the compiler).

Hi,

As a Gentoo user whose make.conf includes -fomit-frame-pointer (gcc 3.3.2), I'd love it if you could elaborate on that a little bit:

a) is this code also in the mainline kernel
(test9-bk*)?

b) any idea what would be required to fix it to allow no frame pointer?

I just started experiencing oopsen yesterday after installing -test9-bk24, using the older libs/utils (0.98, I think). But I also put new hardware in (IEEE 1394 card) and added USB mouse support. I'm seeing weird things on my earlier kernel version, too so it could certainly be unrelated...

Dave

-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program. Does SourceForge.net help you be more productive? Does it help you create better code? SHARE THE LOVE, and help us help YOU! Click Here: http://sourceforge.net/donate/

#12 lore raw

At Thu, 20 Nov 2003 10:30:32 -0800 (PST),

Quoted 8 lines · Dave Neuer
Quoted 5 lines · --- Takashi Iwai

Hi all,

- PCM OSS emulation might cause oops when you compile the kernel
without frame pointer (depending on the compiler).

Quoted 8 lines

Hi,

As a Gentoo user whose make.conf includes -fomit-frame-pointer (gcc 3.3.2), I'd love it if you could elaborate on that a little bit:

a) is this code also in the mainline kernel (test9-bk*)?

yes.

Quoted 2 lines

b) any idea what would be required to fix it to allow no frame pointer?

i'm not sure which gcc version causes. at least, i can *not* reproduce this on my system (gcc-3.3.1).

a workaround is to add a local CFLAGS for rate.c in Makefile. it's included in alsa-driver's Makefile, but not put in alsa-kernel, i.e. the tree to be merged to the main 2.6.0 tree.

Quoted 6 lines

I just started experiencing oopsen yesterday after installing -test9-bk24, using the older libs/utils (0.98, I think). But I also put new hardware in (IEEE 1394 card) and added USB mouse support. I'm seeing weird things on my earlier kernel version, too so it could certainly be unrelated...

this bug hits only in the sample rate conversion of OSS emulation, so it doesn't occur in most cases.
the typical case is via82xx driver without specifying dxs_support option, or some mobo with a cheap ac97 codec supporting only 48kHz. when you play mp3 on such a machine over OSS API, the sample rate is converted from 44.1 to 48khz.

Takashi

-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program. Does SourceForge.net help you be more productive? Does it help you create better code? SHARE THE LOVE, and help us help YOU! Click Here: http://sourceforge.net/donate/

#13 lore raw
In reply to Takashi Iwai
Quoted 2 lines · Takashi Iwai

Especially, if you have a card or a mobo with ICH (or compatible) chip and *multiple* AC97 codecs, testing this version is appreciated.

Some of my apps have stopped working after 1.0.0pre1 on a T23 Thinkpad (intel8x0, 82801).

Simple things like "arecord -f cd - |aplay -f cd -" work, but apps which relay on select/poll before doing any i/o do not work. I didn't have time to test, but I suspect jackd would not have worked. :)

On the same machine, alsa-driver-0.9.8 works fine, but not with 1.0.0pre1 and 1.0.0pre3.

PS The same intel8x0 ac97-chip can be found on many other
thinkpad models: at least T24, X30, X31 and T40.

Signature · 10 lines
 http://www.eca.cx
 Audio software for Linux!



-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program.
Does SourceForge.net help you be more productive?  Does it
help you create better code?  SHARE THE LOVE, and help us help
YOU!  Click Here: http://sourceforge.net/donate/
#14 lore raw

At Wed, 26 Nov 2003 19:31:39 +0200 (EET),

Quoted 9 lines · Kai Vehmanen
Quoted 2 lines · Takashi Iwai

Especially, if you have a card or a mobo with ICH (or compatible) chip and *multiple* AC97 codecs, testing this version is appreciated.

Some of my apps have stopped working after 1.0.0pre1 on a T23 Thinkpad (intel8x0, 82801).

Quoted 6 lines

Simple things like "arecord -f cd - |aplay -f cd -" work, but apps which relay on select/poll before doing any i/o do not work. I didn't have time to test, but I suspect jackd would not have worked. :)

On the same machine, alsa-driver-0.9.8 works fine, but not with 1.0.0pre1 and 1.0.0pre3.

1.0.0pre1 is broken for intel8x0, so use pre3 only :)

what do you mean exactly "don't work" in the above? the operation is blocked, or returns an error?
if it's blocked, check /proc/asound/card0/pcm0*/sub0/* files.

does this laptop have multiple codecs? you can find them in /proc/asound/card0/codec97#* directories.

thanks,

Takashi

-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program. Does SourceForge.net help you be more productive? Does it help you create better code? SHARE THE LOVE, and help us help YOU! Click Here: http://sourceforge.net/donate/

#15 lore raw
Quoted 2 lines · Takashi Iwai
Quoted 2 lines

Some of my apps have stopped working after 1.0.0pre1 on a T23 Thinkpad (intel8x0, 82801).

[...]

Quoted 5 lines
Quoted 2 lines

On the same machine, alsa-driver-0.9.8 works fine, but not with 1.0.0pre1 and 1.0.0pre3.

what do you mean exactly "don't work" in the above? the operation is blocked, or returns an error?
if it's blocked, check /proc/asound/card0/pcm0*/sub0/* files.

I didn't have time to fully analyze the problem, so I'm a little short of details, but few observations:

- playback direction never wakes up as ready
- capture direction wakes up but snd_pcm_delay() returns odd values

Quoted 2 lines

does this laptop have multiple codecs? you can find them in /proc/asound/card0/codec97#* directories.

I'm pretty sure it doesn't. Anyways, I'll check this tomorrow.

Signature · 10 lines
 http://www.eca.cx
 Audio software for Linux!



-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program.
Does SourceForge.net help you be more productive?  Does it
help you create better code?  SHARE THE LOVE, and help us help
YOU!  Click Here: http://sourceforge.net/donate/
#16 lore raw

problems with intel8x0 and 1.0.0pre3 (was: Re: [ANN] ALSA 1.0.0pre1 is released)

In reply to Takashi Iwai
Quoted 5 lines · Takashi Iwai
Quoted 2 lines

On the same machine, alsa-driver-0.9.8 works fine, but not with 1.0.0pre1 and 1.0.0pre3.

what do you mean exactly "don't work" in the above? the operation is blocked, or returns an error?
if it's blocked, check /proc/asound/card0/pcm0*/sub0/* files.

I think I found the problem. In 1.0.0preX, the supported sampling rate is limited to 48000Hz and this causes all kinds of problems if running with lower rates (I've mostly tested running at 8000Hz) - i.e. when using alsa-lib for sr-conversion.

With 1.0.0pre3, for example "arecord -r 8000 - |aplay -r 8000 -" produces constant xruns. A similar app that relies on poll()/select(), fails even more severely. When looking at...

/proc/asound/card0/pcm0p/sub0/hw_params and
/proc/asound/card0/pcm0c/sub0/hw_params

.. the rate is locked to 48000 with 1.0.0pre3. On the same machine with 0.9.8, the hw-rate is 8000. I've used the 'default' PCM device in all my tests.

Quoted 2 lines

does this laptop have multiple codecs? you can find them in /proc/asound/card0/codec97#* directories.

Nopes, '/proc/asound/card0/codec97#0/ac97#0-0' is the only one, and contains:
--cut--
0-0/0: Cirrus Logic CS4299 rev 6

Capabilities : -headphone out-
DAC resolution : 20-bit
ADC resolution : 18-bit
3D enhancement : Crystal Semi 3D Stereo Enhancement

Current setup
Mic gain : +0dB [+0dB]
POP path : pre 3D
Sim. stereo : off

3D enhancement   : off
Loudness         : off
Mono output      : MIX
Mic select       : Mic1

ADC/DAC loopback : off
Extended ID : codec=0 rev=0 AMAP DSA=0 VRA
Extended status : VRA
PCM front DAC : 48000Hz
PCM ADC : 48000Hz
SPDIF Control : Consumer PCM Category=0x2 Generation=1 Rate=48kHz
--cut--

Signature · 10 lines
 http://www.eca.cx
 Audio software for Linux!



-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program.
Does SourceForge.net help you be more productive?  Does it
help you create better code?  SHARE THE LOVE, and help us help
YOU!  Click Here: http://sourceforge.net/donate/
#17 lore raw

Re: problems with intel8x0 and 1.0.0pre3 (was: Re: [ANN] ALSA 1.0.0pre1 is released)

At Thu, 27 Nov 2003 18:16:05 +0200 (EET),

Quoted 21 lines · Kai Vehmanen
Quoted 5 lines · Takashi Iwai
Quoted 2 lines

On the same machine, alsa-driver-0.9.8 works fine, but not with 1.0.0pre1 and 1.0.0pre3.

what do you mean exactly "don't work" in the above? the operation is blocked, or returns an error?
if it's blocked, check /proc/asound/card0/pcm0*/sub0/* files.

I think I found the problem. In 1.0.0preX, the supported sampling rate is limited to 48000Hz and this causes all kinds of problems if running with lower rates (I've mostly tested running at 8000Hz) - i.e. when using alsa-lib for sr-conversion.

With 1.0.0pre3, for example "arecord -r 8000 - |aplay -r 8000 -" produces constant xruns. A similar app that relies on poll()/select(), fails even more severely. When looking at...

/proc/asound/card0/pcm0p/sub0/hw_params and
/proc/asound/card0/pcm0c/sub0/hw_params

Quoted 8 lines

.. the rate is locked to 48000 with 1.0.0pre3. On the same machine with 0.9.8, the hw-rate is 8000. I've used the 'default' PCM device in all my tests.

Quoted 2 lines

does this laptop have multiple codecs? you can find them in /proc/asound/card0/codec97#* directories.

Nopes, '/proc/asound/card0/codec97#0/ac97#0-0' is the only one,

ok, then the problem is not the multiple codec but the detection of available sample rates...

Takashi

Quoted 17 lines

and contains:
--cut--
0-0/0: Cirrus Logic CS4299 rev 6

Capabilities : -headphone out-
DAC resolution : 20-bit
ADC resolution : 18-bit
3D enhancement : Crystal Semi 3D Stereo Enhancement

Current setup
Mic gain : +0dB [+0dB]
POP path : pre 3D
Sim. stereo : off

3D enhancement   : off
Loudness         : off
Mono output      : MIX
Mic select       : Mic1
Quoted 12 lines

ADC/DAC loopback : off
Extended ID : codec=0 rev=0 AMAP DSA=0 VRA
Extended status : VRA
PCM front DAC : 48000Hz
PCM ADC : 48000Hz
SPDIF Control : Consumer PCM Category=0x2 Generation=1 Rate=48kHz
--cut--

Signature · 2 lines
 http://www.eca.cx
 Audio software for Linux!

-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program. Does SourceForge.net help you be more productive? Does it help you create better code? SHARE THE LOVE, and help us help YOU! Click Here: http://sourceforge.net/donate/

#18 lore raw

Re: Re: problems with intel8x0 and 1.0.0pre3 (was: Re: [ANN] ALSA 1.0.0pre1 is released)

At Thu, 27 Nov 2003 17:23:43 +0100,

Quoted 25 lines · I wrote:

At Thu, 27 Nov 2003 18:16:05 +0200 (EET),

Quoted 22 lines · Kai Vehmanen
Quoted 5 lines · Takashi Iwai
Quoted 2 lines

On the same machine, alsa-driver-0.9.8 works fine, but not with 1.0.0pre1 and 1.0.0pre3.

what do you mean exactly "don't work" in the above? the operation is blocked, or returns an error?
if it's blocked, check /proc/asound/card0/pcm0*/sub0/* files.

I think I found the problem. In 1.0.0preX, the supported sampling rate is limited to 48000Hz and this causes all kinds of problems if running with lower rates (I've mostly tested running at 8000Hz) - i.e. when using alsa-lib for sr-conversion.

With 1.0.0pre3, for example "arecord -r 8000 - |aplay -r 8000 -" produces constant xruns. A similar app that relies on poll()/select(), fails even more severely. When looking at...

/proc/asound/card0/pcm0p/sub0/hw_params and
/proc/asound/card0/pcm0c/sub0/hw_params

.. the rate is locked to 48000 with 1.0.0pre3.

do you mean both for playback and capture, or only on capture side? if the playback rate isn't set correctly (i.e. aplay -Dhw with 44.1kHz doesn't work), it's a serious problem.
but, if it's a capture side, right now i found and fixed a bug. could you try and check whether it works for you?

Takashi

Index: alsa-kernel/pci/ac97/ac97_pcm.c
===================================================================
RCS file: /suse/tiwai/cvs/alsa/alsa-kernel/pci/ac97/ac97_pcm.c,v retrieving revision 1.4

pci/ac97/ac97_pcm.c +19 -6
diff -u -r1.4 ac97_pcm.c
@@ -145,6 +145,19 @@
   }
 }};
 
+/* FIXME: more various mappings for ADC? */
+static unsigned char rate_cregs[9] = {
+	AC97_PCM_LR_ADC_RATE,	/* 3 */
+	AC97_PCM_LR_ADC_RATE,	/* 4 */
+	0xff,			/* 5 */
+	AC97_PCM_MIC_ADC_RATE,	/* 6 */
+	0xff,			/* 7 */
+	0xff,			/* 8 */
+	0xff,			/* 9 */
+	0xff,			/* 10 */
+	0xff,			/* 11 */
+};
+
 static unsigned char get_slot_reg(struct ac97_pcm *pcm, unsigned short cidx,
 				  unsigned short slot, int dbl)
 {
@@ -152,7 +165,10 @@
 		return 0xff;
 	if (slot > 11)
 		return 0xff;
-	return rate_reg_tables[dbl][pcm->r[dbl].rate_table[cidx]][slot - 3];
+	if (pcm->stream == SNDRV_PCM_STREAM_PLAYBACK)
+		return rate_preg_tables[dbl][pcm->r[dbl].rate_table[cidx]][slot - 3];
+	else
+		return rate_cregs[slot - 3];
 }
 
 static int set_spdif_rate(ac97_t *ac97, unsigned short rate)
@@ -380,7 +396,6 @@
 	const struct ac97_pcm *pcm;
 	struct ac97_pcm *rpcms, *rpcm;
 	unsigned short avail_slots[2][4];
-	unsigned short all_slots[2][4];
 	unsigned char rate_table[2][4];
 	unsigned short tmp, slots;
 	unsigned int rates;
@@ -390,7 +405,7 @@
 	if (rpcms == NULL)
 		return -ENOMEM;
 	memset(avail_slots, 0, sizeof(avail_slots));
-	memset(all_slots, 0, sizeof(all_slots));
+	memset(rate_table, 0, sizeof(rate_table));
 	for (i = 0; i < 4; i++) {
 		codec = bus->codec[i];
 		if (! codec)
@@ -401,8 +416,6 @@
 			for (j = 0; j < i; j++)
 				avail_slots[1][i] &= ~avail_slots[1][j];
 		}
-		all_slots[0][i] = avail_slots[0][i];
-		all_slots[1][i] = avail_slots[1][i];
 	}
 	/* FIXME: add double rate allocation */
 	/* first step - exclusive devices */
@@ -430,7 +443,7 @@
 				if (tmp) {
 					rpcm->r[0].rslots[j] = tmp;
 					rpcm->r[0].codec[j] = bus->codec[j];
-					rpcm->r[0].rate_table[j] = rate_table[0][j];
+					rpcm->r[0].rate_table[j] = rate_table[pcm->stream][j];
 					rates = get_rates(rpcm, j, tmp, 0);
 					avail_slots[pcm->stream][j] &= ~tmp;
 				}
#19 lore raw

Re: Re: problems with intel8x0 and 1.0.0pre3 (was: Re: [ANN] ALSA 1.0.0pre1 is released)

Hi,

Quoted 5 lines · Takashi Iwai
Quoted 2 lines
Quoted 2 lines

/proc/asound/card0/pcm0p/sub0/hw_params and
.. the rate is locked to 48000 with 1.0.0pre3.

do you mean both for playback and capture, or only on capture side? if the playback rate isn't set correctly (i.e. aplay -Dhw with 44.1kHz doesn't work), it's a serious problem.

that seems to be the case. I get this error:

--cut--
[kaiv@hed040-169 ad]$ aplay -Dhw:0 ~/mp3/foo2.wav
Playing WAVE '/home/kaiv/mp3/foo2.wav' : Signed 16 bit Little Endian, Rate 44100 Hz, Stereo
Warning: rate is not accurate (requested = 44100Hz, got = 48000Hz)
please, try the plug plugin (-Dplug:hw:0)
--cut--

With an app that opens playback and capture with 8000kHz, the hw_params output looks like this:

--cut--
[root@hed040-169 pub]# cat /proc/asound/card0/pcm0p/sub0/hw_params
access: MMAP_INTERLEAVED
format: S16_LE
subformat: STD
channels: 2
rate: 48000 (48000/1)
period_size: 960
buffer_size: 1920
tick_time: 10000
[root@hed040-169 pub]# cat /proc/asound/card0/pcm0c/sub0/hw_params
access: MMAP_INTERLEAVED
format: S16_LE
subformat: STD
channels: 2
rate: 8000 (8000/1)
period_size: 160
buffer_size: 320
tick_time: 10000
--cut--

Quoted 2 lines

but, if it's a capture side, right now i found and fixed a bug. could you try and check whether it works for you?

The above results are with your patch applied. The ADC rate seems to be correct.

Signature · 10 lines
 http://www.eca.cx
 Audio software for Linux!



-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program.
Does SourceForge.net help you be more productive?  Does it
help you create better code?  SHARE THE LOVE, and help us help
YOU!  Click Here: http://sourceforge.net/donate/
#20 lore raw

Re: problems with intel8x0 and 1.0.0pre3

In reply to Kai Vehmanen

Hello,

Quoted 3 lines · Kai Vehmanen

With 1.0.0pre3, for example "arecord -r 8000 - |aplay -r 8000 -" produces constant xruns. A similar app that relies on poll()/select(), fails even more severely. When looking at...

this problem was fixed in 1.0.0rc1! It seems that the final commit to ac97_pcm.c (rev1.7) took care of the problem (the added 'rpcm->stream == rpcms[k].stream' check).

Signature · 10 lines
 http://www.eca.cx
 Audio software for Linux!



-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program.
Does SourceForge.net help you be more productive?  Does it
help you create better code?  SHARE THE LOVE, and help us help
YOU!  Click Here: http://sourceforge.net/donate/