Re: [alsa-devel] Backported sbxfi driver (UNTESTED!)
At Fri, 10 Oct 2008 20:39:48 +0400, The Source wrote:
2008/10/10 Takashi Iwai tiwai@suse.de
At Fri, 10 Oct 2008 01:26:15 -0500, Ted T. Logian wrote: > > > > On Fri, 2008-10-10 at 08:01 +0200, Takashi Iwai wrote: > > At Thu, 09 Oct 2008 15:17:58 -0500, > > Ted T. Logian wrote: > > > > > > Sorry for the multiple posts, but once I rmmod snd-sbxfi, running mixer does > > > fine. It seems if I have snd-sbxfi loaded and I run alsamixer, it locks the > > > entire system. > > > > So, do you mean loading the driver itself doesn't lock up? If so, > > it's better than I expected. > > > > Did you run ever PCM playback/capture before that? This is more > > dangerous :) > > > > Also, please give your hardware details, as there are different models > > for X-Fi, and some are handled pretty differently. > > > > [BTW, please stop top-posting, and avoid HTML mails for ML. It's easy > > to switch to normal (plain) mail mode on Gmail, just a click.] > > > > > > thanks, > > > > Takashi > > > I think perhaps from uninstalling pulseaudio I got further. I can use > mixer now, and it even has a "Master" device, but nothing else... > > however, I do get a lock up later. > > I get this from running aplay, too, if this helps... Thanks! That's rather a stupid bug. The patch is below. Takashi diff --git a/sound/pci/sbxfi/sbxfi.c b/sound/pci/sbxfi/sbxfi.c index 8066bf4..8a0eea0 100644 --- a/sound/pci/sbxfi/sbxfi.c +++ b/sound/pci/sbxfi/sbxfi.c @@ -831,7 +831,7 @@ static struct sbxfi_port *sbxfi_port_alloc(struct sbxfi *chip, port->src[0] = src; port->src[1] = src + 1; spin_lock_irq(&chip->port_lock); - list_add(&chip->port_list, &port->list); + list_add(&port->list, &chip->port_list); spin_unlock_irq(&chip->port_lock); return port; } _______________________________________________ Alsa-devel mailing list Alsa-devel@alsa-project.org http://mailman.alsa-project.org/mailman/listinfo/alsa-devel
Hello, I have tried the driver with the fix you provided, but I couldn't get anything. Kernel panic on any attempt to use sound or mixer. Also kernel panic several seconds after Xorg starts. Each crash causes either hang or reboot so I couldn't get any error messages. If there are some logs left, please tell me where they are and I'll send them to you. My card is X-Fi Platinum Fatality Champion Series.
For testing, first try without X but just use Linux console, so that you have a better chance to see Oops messages. It's pretty important to see whether it's an Oops or a panic. Also, double-check whether you really installed the corrected drivers. Often it's missing in the right place by stupid reasons.
For first tests, don't use any sound server. Try with aplay like: % aplay -Dhw foo.wav Try first 96kHz samples, 48kHz samples, and then others.
thanks,
Takashi
On Sat, 2008-10-11 at 17:47 +0200, Takashi Iwai wrote:
At Fri, 10 Oct 2008 20:39:48 +0400, The Source wrote:
2008/10/10 Takashi Iwai tiwai@suse.de
At Fri, 10 Oct 2008 01:26:15 -0500, Ted T. Logian wrote: > > > > On Fri, 2008-10-10 at 08:01 +0200, Takashi Iwai wrote: > > At Thu, 09 Oct 2008 15:17:58 -0500, > > Ted T. Logian wrote: > > > > > > Sorry for the multiple posts, but once I rmmod snd-sbxfi, running mixer does > > > fine. It seems if I have snd-sbxfi loaded and I run alsamixer, it locks the > > > entire system. > > > > So, do you mean loading the driver itself doesn't lock up? If so, > > it's better than I expected. > > > > Did you run ever PCM playback/capture before that? This is more > > dangerous :) > > > > Also, please give your hardware details, as there are different models > > for X-Fi, and some are handled pretty differently. > > > > [BTW, please stop top-posting, and avoid HTML mails for ML. It's easy > > to switch to normal (plain) mail mode on Gmail, just a click.] > > > > > > thanks, > > > > Takashi > > > I think perhaps from uninstalling pulseaudio I got further. I can use > mixer now, and it even has a "Master" device, but nothing else... > > however, I do get a lock up later. > > I get this from running aplay, too, if this helps... Thanks! That's rather a stupid bug. The patch is below. Takashi diff --git a/sound/pci/sbxfi/sbxfi.c b/sound/pci/sbxfi/sbxfi.c index 8066bf4..8a0eea0 100644 --- a/sound/pci/sbxfi/sbxfi.c +++ b/sound/pci/sbxfi/sbxfi.c @@ -831,7 +831,7 @@ static struct sbxfi_port *sbxfi_port_alloc(struct sbxfi *chip, port->src[0] = src; port->src[1] = src + 1; spin_lock_irq(&chip->port_lock); - list_add(&chip->port_list, &port->list); + list_add(&port->list, &chip->port_list); spin_unlock_irq(&chip->port_lock); return port; } _______________________________________________ Alsa-devel mailing list Alsa-devel@alsa-project.org http://mailman.alsa-project.org/mailman/listinfo/alsa-devel
Hello, I have tried the driver with the fix you provided, but I couldn't get anything. Kernel panic on any attempt to use sound or mixer. Also kernel panic several seconds after Xorg starts. Each crash causes either hang or reboot so I couldn't get any error messages. If there are some logs left, please tell me where they are and I'll send them to you. My card is X-Fi Platinum Fatality Champion Series.
For testing, first try without X but just use Linux console, so that you have a better chance to see Oops messages. It's pretty important to see whether it's an Oops or a panic. Also, double-check whether you really installed the corrected drivers. Often it's missing in the right place by stupid reasons.
For first tests, don't use any sound server. Try with aplay like: % aplay -Dhw foo.wav Try first 96kHz samples, 48kHz samples, and then others.
thanks,
Takashi
Tak,
I just wanted to make sure you knew of the different types of sbxfi cards.
For instance, I have this one: http://www.geeks.com/details.asp?invtid=SB0670-BULK&cat=SND
It's still emu20k1. However, it is impossible to use with creative's drivers, as it always causes a lockup, but it works with oss4 drivers.
I don't know if this matters to you at all, but I wanted to let you know in case it affected how the driver progressed to be compatible with the most emu20k1 cards.
At Sat, 11 Oct 2008 11:03:22 -0500, Ted T. Logian wrote:
On Sat, 2008-10-11 at 17:47 +0200, Takashi Iwai wrote:
At Fri, 10 Oct 2008 20:39:48 +0400, The Source wrote: > > 2008/10/10 Takashi Iwai <tiwai@suse.de> > > At Fri, 10 Oct 2008 01:26:15 -0500, > Ted T. Logian wrote: > > > > > > > > On Fri, 2008-10-10 at 08:01 +0200, Takashi Iwai wrote: > > > At Thu, 09 Oct 2008 15:17:58 -0500, > > > Ted T. Logian wrote: > > > > > > > > Sorry for the multiple posts, but once I rmmod snd-sbxfi, running > mixer does > > > > fine. It seems if I have snd-sbxfi loaded and I run alsamixer, it > locks the > > > > entire system. > > > > > > So, do you mean loading the driver itself doesn't lock up? If so, > > > it's better than I expected. > > > > > > Did you run ever PCM playback/capture before that? This is more > > > dangerous :) > > > > > > Also, please give your hardware details, as there are different models > > > for X-Fi, and some are handled pretty differently. > > > > > > [BTW, please stop top-posting, and avoid HTML mails for ML. It's easy > > > to switch to normal (plain) mail mode on Gmail, just a click.] > > > > > > > > > thanks, > > > > > > Takashi > > > > > > I think perhaps from uninstalling pulseaudio I got further. I can use > > mixer now, and it even has a "Master" device, but nothing else... > > > > however, I do get a lock up later. > > > > I get this from running aplay, too, if this helps... > > Thanks! That's rather a stupid bug. > The patch is below. > > Takashi > > diff --git a/sound/pci/sbxfi/sbxfi.c b/sound/pci/sbxfi/sbxfi.c > index 8066bf4..8a0eea0 100644 > --- a/sound/pci/sbxfi/sbxfi.c > +++ b/sound/pci/sbxfi/sbxfi.c > @@ -831,7 +831,7 @@ static struct sbxfi_port *sbxfi_port_alloc(struct > sbxfi *chip, > port->src[0] = src; > port->src[1] = src + 1; > spin_lock_irq(&chip->port_lock); > - list_add(&chip->port_list, &port->list); > + list_add(&port->list, &chip->port_list); > spin_unlock_irq(&chip->port_lock); > return port; > } > _______________________________________________ > Alsa-devel mailing list > Alsa-devel@alsa-project.org > http://mailman.alsa-project.org/mailman/listinfo/alsa-devel > > Hello, I have tried the driver with the fix you provided, but I couldn't get > anything. Kernel panic on any attempt to use sound or mixer. Also kernel panic > several seconds after Xorg starts. Each crash causes either hang or reboot so > I couldn't get any error messages. If there are some logs left, please tell me > where they are and I'll send them to you. > My card is X-Fi Platinum Fatality Champion Series. For testing, first try without X but just use Linux console, so that you have a better chance to see Oops messages. It's pretty important to see whether it's an Oops or a panic. Also, double-check whether you really installed the corrected drivers. Often it's missing in the right place by stupid reasons. For first tests, don't use any sound server. Try with aplay like: % aplay -Dhw foo.wav Try first 96kHz samples, 48kHz samples, and then others. thanks, Takashi
Tak,
I just wanted to make sure you knew of the different types of sbxfi cards.
Yeah, especially Vista-compatible cards require a very strange initialization sequence, as you find in the source code...
For instance, I have this one: http://www.geeks.com/details.asp?invtid=SB0670-BULK&cat=SND
It's still emu20k1. However, it is impossible to use with creative's drivers, as it always causes a lockup, but it works with oss4 drivers.
Does it lock after loading the driver? If not, what does /proc/asound/cards shows? Also, show the output of "lspci -nv" for the device, and "lspci -xvvv" before and after loading the driver.
I'm afraid it's UAA (vista-compatible) type. If so, maybe something is wrong in the initialization, or it takes a wrong port address (BAR), etc...
I don't know if this matters to you at all, but I wanted to let you know in case it affected how the driver progressed to be compatible with the most emu20k1 cards.
Indeed, there are really different models with emu20k1...
thanks,
Takashi
What does this mean? That there are X-Fi cards which aren't Vista-compatible? Also, shouldn't UAA cards work with hda-intel driver with specific patch for Creative?
2008/10/11 Takashi Iwai tiwai@suse.de:
At Sat, 11 Oct 2008 11:03:22 -0500, Ted T. Logian wrote:
On Sat, 2008-10-11 at 17:47 +0200, Takashi Iwai wrote:
At Fri, 10 Oct 2008 20:39:48 +0400, The Source wrote: > > 2008/10/10 Takashi Iwai <tiwai@suse.de> > > At Fri, 10 Oct 2008 01:26:15 -0500, > Ted T. Logian wrote: > > > > > > > > On Fri, 2008-10-10 at 08:01 +0200, Takashi Iwai wrote: > > > At Thu, 09 Oct 2008 15:17:58 -0500, > > > Ted T. Logian wrote: > > > > > > > > Sorry for the multiple posts, but once I rmmod snd-sbxfi, running > mixer does > > > > fine. It seems if I have snd-sbxfi loaded and I run alsamixer, it > locks the > > > > entire system. > > > > > > So, do you mean loading the driver itself doesn't lock up? If so, > > > it's better than I expected. > > > > > > Did you run ever PCM playback/capture before that? This is more > > > dangerous :) > > > > > > Also, please give your hardware details, as there are different models > > > for X-Fi, and some are handled pretty differently. > > > > > > [BTW, please stop top-posting, and avoid HTML mails for ML. It's easy > > > to switch to normal (plain) mail mode on Gmail, just a click.] > > > > > > > > > thanks, > > > > > > Takashi > > > > > > I think perhaps from uninstalling pulseaudio I got further. I can use > > mixer now, and it even has a "Master" device, but nothing else... > > > > however, I do get a lock up later. > > > > I get this from running aplay, too, if this helps... > > Thanks! That's rather a stupid bug. > The patch is below. > > Takashi > > diff --git a/sound/pci/sbxfi/sbxfi.c b/sound/pci/sbxfi/sbxfi.c > index 8066bf4..8a0eea0 100644 > --- a/sound/pci/sbxfi/sbxfi.c > +++ b/sound/pci/sbxfi/sbxfi.c > @@ -831,7 +831,7 @@ static struct sbxfi_port *sbxfi_port_alloc(struct > sbxfi *chip, > port->src[0] = src; > port->src[1] = src + 1; > spin_lock_irq(&chip->port_lock); > - list_add(&chip->port_list, &port->list); > + list_add(&port->list, &chip->port_list); > spin_unlock_irq(&chip->port_lock); > return port; > } > _______________________________________________ > Alsa-devel mailing list > Alsa-devel@alsa-project.org > http://mailman.alsa-project.org/mailman/listinfo/alsa-devel > > Hello, I have tried the driver with the fix you provided, but I couldn't get > anything. Kernel panic on any attempt to use sound or mixer. Also kernel panic > several seconds after Xorg starts. Each crash causes either hang or reboot so > I couldn't get any error messages. If there are some logs left, please tell me > where they are and I'll send them to you. > My card is X-Fi Platinum Fatality Champion Series. For testing, first try without X but just use Linux console, so that you have a better chance to see Oops messages. It's pretty important to see whether it's an Oops or a panic. Also, double-check whether you really installed the corrected drivers. Often it's missing in the right place by stupid reasons. For first tests, don't use any sound server. Try with aplay like: % aplay -Dhw foo.wav Try first 96kHz samples, 48kHz samples, and then others. thanks, Takashi
Tak,
I just wanted to make sure you knew of the different types of sbxfi cards.
Yeah, especially Vista-compatible cards require a very strange initialization sequence, as you find in the source code...
For instance, I have this one: http://www.geeks.com/details.asp?invtid=SB0670-BULK&cat=SND
It's still emu20k1. However, it is impossible to use with creative's drivers, as it always causes a lockup, but it works with oss4 drivers.
Does it lock after loading the driver? If not, what does /proc/asound/cards shows? Also, show the output of "lspci -nv" for the device, and "lspci -xvvv" before and after loading the driver.
I'm afraid it's UAA (vista-compatible) type. If so, maybe something is wrong in the initialization, or it takes a wrong port address (BAR), etc...
I don't know if this matters to you at all, but I wanted to let you know in case it affected how the driver progressed to be compatible with the most emu20k1 cards.
Indeed, there are really different models with emu20k1...
thanks,
Takashi _______________________________________________ Alsa-devel mailing list Alsa-devel@alsa-project.org http://mailman.alsa-project.org/mailman/listinfo/alsa-devel
Vedran Miletić wrote:
What does this mean? That there are X-Fi cards which aren't Vista-compatible? Also, shouldn't UAA cards work with hda-intel driver with specific patch for Creative?
X-Fi cards that are Vista compatible boot up in hda-intel mode, so yes, we could provide support in the hda-intel driver for them. The features would be limited, but the user would get some sound.
Kind Regards
James
James Courtier-Dutton пишет:
Vedran Miletić wrote:
What does this mean? That there are X-Fi cards which aren't Vista-compatible? Also, shouldn't UAA cards work with hda-intel driver with specific patch for Creative?
X-Fi cards that are Vista compatible boot up in hda-intel mode, so yes, we could provide support in the hda-intel driver for them. The features would be limited, but the user would get some sound.
Kind Regards
James
Alsa-devel mailing list Alsa-devel@alsa-project.org http://mailman.alsa-project.org/mailman/listinfo/alsa-devel
Attaching files you requested. I don't have 96000Hz samples, tried 22KHz - reboot. Attempt to load module while Xorg is running causes reboot. Note that you should use latest mercurial oss to port - earlier versions have conflict with drivers like nvidia. alsamixer shows only master but that probably because asound.conf deosn't contain needed info (I usually use system-config-soundcard to write all that needed to asound.conf but I can't do it since Xorg with sbxfi don't work).
Takashi Iwai wrote:
I'm afraid it's UAA (vista-compatible) type. If so, maybe something is wrong in the initialization, or it takes a wrong port address (BAR), etc...
The BARs change when one switches from UAA (the default on some cards) to XFi mode. You will probably have to provide some method for correcting this after the change.
Kind Regards
James
At Sun, 12 Oct 2008 09:44:37 +0100, James Courtier-Dutton wrote:
Takashi Iwai wrote:
I'm afraid it's UAA (vista-compatible) type. If so, maybe something is wrong in the initialization, or it takes a wrong port address (BAR), etc...
The BARs change when one switches from UAA (the default on some cards) to XFi mode. You will probably have to provide some method for correcting this after the change.
The OSS driver seems to overwrite the old values of BAR1-5...
Takashi
Takashi Iwai ?????:
At Fri, 10 Oct 2008 20:39:48 +0400, The Source wrote:
2008/10/10 Takashi Iwai tiwai@suse.de
At Fri, 10 Oct 2008 01:26:15 -0500, Ted T. Logian wrote: > > > > On Fri, 2008-10-10 at 08:01 +0200, Takashi Iwai wrote: > > At Thu, 09 Oct 2008 15:17:58 -0500, > > Ted T. Logian wrote: > > > > > > Sorry for the multiple posts, but once I rmmod snd-sbxfi, running mixer does > > > fine. It seems if I have snd-sbxfi loaded and I run alsamixer, it locks the > > > entire system. > > > > So, do you mean loading the driver itself doesn't lock up? If so, > > it's better than I expected. > > > > Did you run ever PCM playback/capture before that? This is more > > dangerous :) > > > > Also, please give your hardware details, as there are different models > > for X-Fi, and some are handled pretty differently. > > > > [BTW, please stop top-posting, and avoid HTML mails for ML. It's easy > > to switch to normal (plain) mail mode on Gmail, just a click.] > > > > > > thanks, > > > > Takashi > > > I think perhaps from uninstalling pulseaudio I got further. I can use > mixer now, and it even has a "Master" device, but nothing else... > > however, I do get a lock up later. > > I get this from running aplay, too, if this helps... Thanks! That's rather a stupid bug. The patch is below. Takashi diff --git a/sound/pci/sbxfi/sbxfi.c b/sound/pci/sbxfi/sbxfi.c index 8066bf4..8a0eea0 100644 --- a/sound/pci/sbxfi/sbxfi.c +++ b/sound/pci/sbxfi/sbxfi.c @@ -831,7 +831,7 @@ static struct sbxfi_port *sbxfi_port_alloc(struct sbxfi *chip, port->src[0] = src; port->src[1] = src + 1; spin_lock_irq(&chip->port_lock); - list_add(&chip->port_list, &port->list); + list_add(&port->list, &chip->port_list); spin_unlock_irq(&chip->port_lock); return port; } _______________________________________________ Alsa-devel mailing list Alsa-devel@alsa-project.org http://mailman.alsa-project.org/mailman/listinfo/alsa-devel
Hello, I have tried the driver with the fix you provided, but I couldn't get anything. Kernel panic on any attempt to use sound or mixer. Also kernel panic several seconds after Xorg starts. Each crash causes either hang or reboot so I couldn't get any error messages. If there are some logs left, please tell me where they are and I'll send them to you. My card is X-Fi Platinum Fatality Champion Series.
For testing, first try without X but just use Linux console, so that you have a better chance to see Oops messages. It's pretty important to see whether it's an Oops or a panic. Also, double-check whether you really installed the corrected drivers. Often it's missing in the right place by stupid reasons.
For first tests, don't use any sound server. Try with aplay like: % aplay -Dhw foo.wav Try first 96kHz samples, 48kHz samples, and then others.
thanks,
Takashi
aplay in console causes reboot, sorry.
At Sat, 11 Oct 2008 20:28:23 +0400, The Source wrote:
Takashi Iwai ?????:
At Fri, 10 Oct 2008 20:39:48 +0400, The Source wrote:
2008/10/10 Takashi Iwai tiwai@suse.de
At Fri, 10 Oct 2008 01:26:15 -0500, Ted T. Logian wrote: > > > > On Fri, 2008-10-10 at 08:01 +0200, Takashi Iwai wrote: > > At Thu, 09 Oct 2008 15:17:58 -0500, > > Ted T. Logian wrote: > > > > > > Sorry for the multiple posts, but once I rmmod snd-sbxfi, running mixer does > > > fine. It seems if I have snd-sbxfi loaded and I run alsamixer, it locks the > > > entire system. > > > > So, do you mean loading the driver itself doesn't lock up? If so, > > it's better than I expected. > > > > Did you run ever PCM playback/capture before that? This is more > > dangerous :) > > > > Also, please give your hardware details, as there are different models > > for X-Fi, and some are handled pretty differently. > > > > [BTW, please stop top-posting, and avoid HTML mails for ML. It's easy > > to switch to normal (plain) mail mode on Gmail, just a click.] > > > > > > thanks, > > > > Takashi > > > I think perhaps from uninstalling pulseaudio I got further. I can use > mixer now, and it even has a "Master" device, but nothing else... > > however, I do get a lock up later. > > I get this from running aplay, too, if this helps... Thanks! That's rather a stupid bug. The patch is below. Takashi diff --git a/sound/pci/sbxfi/sbxfi.c b/sound/pci/sbxfi/sbxfi.c index 8066bf4..8a0eea0 100644 --- a/sound/pci/sbxfi/sbxfi.c +++ b/sound/pci/sbxfi/sbxfi.c @@ -831,7 +831,7 @@ static struct sbxfi_port *sbxfi_port_alloc(struct sbxfi *chip, port->src[0] = src; port->src[1] = src + 1; spin_lock_irq(&chip->port_lock); - list_add(&chip->port_list, &port->list); + list_add(&port->list, &chip->port_list); spin_unlock_irq(&chip->port_lock); return port; } _______________________________________________ Alsa-devel mailing list Alsa-devel@alsa-project.org http://mailman.alsa-project.org/mailman/listinfo/alsa-devel
Hello, I have tried the driver with the fix you provided, but I couldn't get anything. Kernel panic on any attempt to use sound or mixer. Also kernel panic several seconds after Xorg starts. Each crash causes either hang or reboot so I couldn't get any error messages. If there are some logs left, please tell me where they are and I'll send them to you. My card is X-Fi Platinum Fatality Champion Series.
For testing, first try without X but just use Linux console, so that you have a better chance to see Oops messages. It's pretty important to see whether it's an Oops or a panic. Also, double-check whether you really installed the corrected drivers. Often it's missing in the right place by stupid reasons.
For first tests, don't use any sound server. Try with aplay like: % aplay -Dhw foo.wav Try first 96kHz samples, 48kHz samples, and then others.
thanks,
Takashi
aplay in console causes reboot, sorry.
OK, then this sounds like the same symptom as Ted's. Could you give the information described in my last followup to Ted's post?
Takashi
participants (5)
-
James Courtier-Dutton
-
Takashi Iwai
-
Ted T. Logian
-
The Source
-
Vedran Miletić