- Alsa-devel - mailman.alsa-project.org

[alsa-devel] lockdep warning on 2.6.29-rc1 with pulseaudio
by Johannes Berg 14 Jan '09

14 Jan '09

Sorry if you've seen this already, I couldn't find it on gmane. soemthing is doing mutex_lock(&pcm->open_mutex); but this lock depends on &chip->open_mutex which in turn depends on &codec->spdif_mutex (seems a little odd) and that already depends on mmap_sem, through might_fault()? [ 49.742444] ======================================================= [ 49.742449] [ INFO: possible circular locking dependency detected ] [ 49.742453] 2.6.29-rc1-wl-11361-g269245b-dirty #15 [ 49.742455] ------------------------------------------------------- [ 49.742458] pulseaudio/3550 is trying to acquire lock: [ 49.742461] (&pcm->open_mutex){--..}, at: [<ffffffffa059f86e>] snd_pcm_release+0x3e/0xb0 [snd_pcm] [ 49.742479] [ 49.742479] but task is already holding lock: [ 49.742482] (&mm->mmap_sem){----}, at: [<ffffffff802c4a42>] sys_munmap+0x42/0x80 [ 49.742493] [ 49.742493] which lock already depends on the new lock. [ 49.742494] [ 49.742497] [ 49.742497] the existing dependency chain (in reverse order) is: [ 49.742500] [ 49.742501] -> #3 (&mm->mmap_sem){----}: [ 49.742507] [<ffffffff80273427>] check_prev_add+0x3b7/0x770 [ 49.742513] [<ffffffff80273de6>] validate_chain+0x606/0x6c0 [ 49.742518] [<ffffffff802742df>] __lock_acquire+0x43f/0xa10 [ 49.742522] [<ffffffff80274941>] lock_acquire+0x91/0xc0 [ 49.742526] [<ffffffff802bc39f>] might_fault+0x7f/0xb0 [ 49.742532] [<ffffffffa07fb15d>] snd_hda_mixer_amp_tlv+0x6d/0xf0 [snd_hda_codec] [ 49.742544] [<ffffffffa0b20bb0>] alc_cap_vol_tlv+0x70/0xa0 [snd_hda_codec_realtek] [ 49.742557] [<ffffffffa03a30be>] snd_ctl_tlv_ioctl+0x14e/0x1d0 [snd] [ 49.742571] [<ffffffffa03a527b>] snd_ctl_ioctl+0x26b/0x540 [snd] [ 49.742581] [<ffffffff802e82d1>] vfs_ioctl+0x31/0xa0 [ 49.742587] [<ffffffff802e8699>] do_vfs_ioctl+0x89/0x350 [ 49.742592] [<ffffffff802e89aa>] sys_ioctl+0x4a/0x80 [ 49.742597] [<ffffffff8020bd1b>] system_call_fastpath+0x16/0x1b [ 49.742603] [<ffffffffffffffff>] 0xffffffffffffffff [ 49.742608] [ 49.742609] -> #2 (&codec->spdif_mutex){--..}: [ 49.742615] [<ffffffff80273427>] check_prev_add+0x3b7/0x770 [ 49.742619] [<ffffffff80273de6>] validate_chain+0x606/0x6c0 [ 49.742623] [<ffffffff802742df>] __lock_acquire+0x43f/0xa10 [ 49.742628] [<ffffffff80274941>] lock_acquire+0x91/0xc0 [ 49.742632] [<ffffffff805ac5fc>] mutex_lock_nested+0xec/0x380 [ 49.742639] [<ffffffffa07f8d54>] snd_hda_multi_out_analog_open+0x84/0x130 [snd_hda_codec] [ 49.742649] [<ffffffffa0b1ac60>] alc880_playback_pcm_open+0x20/0x30 [snd_hda_codec_realtek] [ 49.742659] [<ffffffffa0897f39>] azx_pcm_open+0x209/0x280 [snd_hda_intel] [ 49.742667] [<ffffffffa05a01b8>] snd_pcm_open_substream+0x58/0xa0 [snd_pcm] [ 49.742677] [<ffffffffa05a038b>] snd_pcm_open+0x18b/0x280 [snd_pcm] [ 49.742687] [<ffffffffa05a04ec>] snd_pcm_playback_open+0x2c/0x40 [snd_pcm] [ 49.742697] [<ffffffffa039f7fc>] __snd_open+0x14c/0x270 [snd] [ 49.742707] [<ffffffffa039f947>] snd_open+0x27/0x40 [snd] [ 49.742717] [<ffffffff802dcd47>] chrdev_open+0x137/0x210 [ 49.742722] [<ffffffff802d7a5a>] __dentry_open+0x10a/0x340 [ 49.742727] [<ffffffff802d7d97>] nameidata_to_filp+0x57/0x70 [ 49.742732] [<ffffffff802e69de>] do_filp_open+0x1fe/0x970 [ 49.742737] [<ffffffff802d7868>] do_sys_open+0x78/0x110 [ 49.742742] [<ffffffff802d792b>] sys_open+0x1b/0x20 [ 49.742746] [<ffffffff8020bd1b>] system_call_fastpath+0x16/0x1b [ 49.742751] [<ffffffffffffffff>] 0xffffffffffffffff [ 49.742756] [ 49.742756] -> #1 (&chip->open_mutex){--..}: [ 49.742762] [<ffffffff80273427>] check_prev_add+0x3b7/0x770 [ 49.742767] [<ffffffff80273de6>] validate_chain+0x606/0x6c0 [ 49.742771] [<ffffffff802742df>] __lock_acquire+0x43f/0xa10 [ 49.742775] [<ffffffff80274941>] lock_acquire+0x91/0xc0 [ 49.742780] [<ffffffff805ac5fc>] mutex_lock_nested+0xec/0x380 [ 49.742785] [<ffffffffa0897d78>] azx_pcm_open+0x48/0x280 [snd_hda_intel] [ 49.742792] [<ffffffffa05a01b8>] snd_pcm_open_substream+0x58/0xa0 [snd_pcm] [ 49.742802] [<ffffffffa05a038b>] snd_pcm_open+0x18b/0x280 [snd_pcm] [ 49.742812] [<ffffffffa05a04ec>] snd_pcm_playback_open+0x2c/0x40 [snd_pcm] [ 49.742822] [<ffffffffa039f7fc>] __snd_open+0x14c/0x270 [snd] [ 49.742832] [<ffffffffa039f947>] snd_open+0x27/0x40 [snd] [ 49.742841] [<ffffffff802dcd47>] chrdev_open+0x137/0x210 [ 49.742846] [<ffffffff802d7a5a>] __dentry_open+0x10a/0x340 [ 49.742850] [<ffffffff802d7d97>] nameidata_to_filp+0x57/0x70 [ 49.742855] [<ffffffff802e69de>] do_filp_open+0x1fe/0x970 [ 49.742860] [<ffffffff802d7868>] do_sys_open+0x78/0x110 [ 49.742864] [<ffffffff802d792b>] sys_open+0x1b/0x20 [ 49.742869] [<ffffffff8020bd1b>] system_call_fastpath+0x16/0x1b [ 49.742874] [<ffffffffffffffff>] 0xffffffffffffffff [ 49.742882] [ 49.742883] -> #0 (&pcm->open_mutex){--..}: [ 49.742889] [<ffffffff802730c7>] check_prev_add+0x57/0x770 [ 49.742893] [<ffffffff80273de6>] validate_chain+0x606/0x6c0 [ 49.742897] [<ffffffff802742df>] __lock_acquire+0x43f/0xa10 [ 49.742902] [<ffffffff80274941>] lock_acquire+0x91/0xc0 [ 49.742906] [<ffffffff805ac5fc>] mutex_lock_nested+0xec/0x380 [ 49.742911] [<ffffffffa059f86e>] snd_pcm_release+0x3e/0xb0 [snd_pcm] [ 49.742921] [<ffffffff802dae20>] __fput+0xd0/0x1e0 [ 49.742926] [<ffffffff802daf4d>] fput+0x1d/0x30 [ 49.742931] [<ffffffff802c377a>] remove_vma+0x4a/0x80 [ 49.742936] [<ffffffff802c49b2>] do_munmap+0x342/0x390 [ 49.742940] [<ffffffff802c4a50>] sys_munmap+0x50/0x80 [ 49.742944] [<ffffffff8020bd1b>] system_call_fastpath+0x16/0x1b [ 49.742949] [<ffffffffffffffff>] 0xffffffffffffffff [ 49.742954] [ 49.742954] other info that might help us debug this: [ 49.742955] [ 49.742958] 1 lock held by pulseaudio/3550: [ 49.742961] #0: (&mm->mmap_sem){----}, at: [<ffffffff802c4a42>] sys_munmap+0x42/0x80 [ 49.742969] [ 49.742970] stack backtrace: [ 49.742973] Pid: 3550, comm: pulseaudio Not tainted 2.6.29-rc1-wl-11361-g269245b-dirty #15 [ 49.742976] Call Trace: [ 49.742981] [<ffffffff80272b70>] print_circular_bug_tail+0xe0/0xf0 [ 49.742985] [<ffffffff802730c7>] check_prev_add+0x57/0x770 [ 49.742991] [<ffffffff8026f414>] ? get_lock_stats+0x34/0x70 [ 49.742995] [<ffffffff80273de6>] validate_chain+0x606/0x6c0 [ 49.742998] [<ffffffff802742df>] __lock_acquire+0x43f/0xa10 [ 49.743002] [<ffffffff80274941>] lock_acquire+0x91/0xc0 [ 49.743012] [<ffffffffa059f86e>] ? snd_pcm_release+0x3e/0xb0 [snd_pcm] [ 49.743017] [<ffffffff805ac5fc>] mutex_lock_nested+0xec/0x380 [ 49.743027] [<ffffffffa059f86e>] ? snd_pcm_release+0x3e/0xb0 [snd_pcm] [ 49.743036] [<ffffffffa059f86e>] ? snd_pcm_release+0x3e/0xb0 [snd_pcm] [ 49.743046] [<ffffffffa059f86e>] snd_pcm_release+0x3e/0xb0 [snd_pcm] [ 49.743051] [<ffffffff802dae20>] __fput+0xd0/0x1e0 [ 49.743055] [<ffffffff802daf4d>] fput+0x1d/0x30 [ 49.743059] [<ffffffff802c377a>] remove_vma+0x4a/0x80 [ 49.743063] [<ffffffff802c49b2>] do_munmap+0x342/0x390 [ 49.743067] [<ffffffff802c4a50>] sys_munmap+0x50/0x80 [ 49.743071] [<ffffffff8020bd1b>] system_call_fastpath+0x16/0x1b Incidentally, sound doesn't work on my macbook5,1, I get lots of controls but no output. johannes

2 13

[alsa-devel] [PATCH] hda: fixed invalid value for SND_PCI_QUIRK
by Matthew Ranostay 14 Jan '09

14 Jan '09

Changed invalid STAC_92HD71BXX_REF value for reference board SND_PCI_QUIRK to correct STAC_92HD83XXX_REF. Signed-off-by: Matthew Ranostay <mranostay(a)embeddedalley.com> --- diff --git a/pci/hda/patch_sigmatel.c b/pci/hda/patch_sigmatel.c index 35b83dc..a4e2d8f 100644 --- a/pci/hda/patch_sigmatel.c +++ b/pci/hda/patch_sigmatel.c @@ -1682,7 +1682,7 @@ static const char *stac92hd83xxx_models[STAC_92HD83XXX_MODELS] = { static struct snd_pci_quirk stac92hd83xxx_cfg_tbl[] = { /* SigmaTel reference board */ SND_PCI_QUIRK(PCI_VENDOR_ID_INTEL, 0x2668, - "DFI LanParty", STAC_92HD71BXX_REF), + "DFI LanParty", STAC_92HD83XXX_REF), {} /* terminator */ };

2 1

[alsa-devel] Creative X-Fi not recognized since 1/9 build
by Susan Cragin 14 Jan '09

14 Jan '09

This is FYI. I have a creative X-fi soundcard and have been compiling developer builds for a couple of weeks, mostly with great success. But as of the 1/9 build, my X-fi card is no longer recognized. The driver it uses is CA0110. I do not re-compile kernel. Susan FYI PS -- About sound quality... I had been running Dragon NaturallySpeaking 10 on Linux through wine using the driver. The program measures accuracy for dictated sounds, and my scores have been pretty good, running from 28-31. With the hda-intel they are 25-28.

2 1

[alsa-devel] [PATCH] USB quirk for Logitech Quickcam Pro 9000 name
by Peter Stokes 14 Jan '09

14 Jan '09

Hi, The Logitech QuickCam Pro 9000 does not appear to any product identification strings in its USB device descriptor. Therefore it receives a device name of "USB Device 0x46d:0x990". The attached patch below adds a USB quirk to provide a more friendly name. Also, for your reference, please find attached the output from "lsusb -vv -d 046d:0990" Best regards Peter Stokes ------------------------------------------------------------------------------- Signed-off-by: Peter Stokes <linux(a)dadeos.co.uk> diff --git a/usb/usbquirks.h b/usb/usbquirks.h index 9211575..5d8ef09 100644 --- a/usb/usbquirks.h +++ b/usb/usbquirks.h @@ -128,6 +128,14 @@ .bInterfaceClass = USB_CLASS_AUDIO, .bInterfaceSubClass = USB_SUBCLASS_AUDIO_CONTROL }, +{ + USB_DEVICE(0x046d, 0x0990), + .driver_info = (unsigned long) & (const struct snd_usb_audio_quirk) { + .vendor_name = "Logitech, Inc.", + .product_name = "QuickCam Pro 9000", + .ifnum = QUIRK_NO_INTERFACE + } +}, /* * Yamaha devices

2 1

[alsa-devel] [PATCH] hda: stac92hd8xxx amp mixers
by Matthew Ranostay 14 Jan '09

14 Jan '09

Added amp nid for stac92hd8xxx families of codecs so the input amp mixer is created. Signed-off-by: Matthew Ranostay <mranostay(a)embeddedalley.com> --- diff --git a/pci/hda/patch_sigmatel.c b/pci/hda/patch_sigmatel.c index 35b83dc..dedf8b2 100644 --- a/pci/hda/patch_sigmatel.c +++ b/pci/hda/patch_sigmatel.c @@ -331,6 +331,10 @@ static unsigned int stac92hd83xxx_pwr_mapping[4] = { 0x03, 0x0c, 0x10, 0x40, }; +static unsigned int stac92hd83xxx_amp_nids[1] = { + 0xc, +}; + static hda_nid_t stac92hd71bxx_pwr_nids[3] = { 0x0a, 0x0d, 0x0f }; @@ -4672,6 +4676,7 @@ static int patch_stac92hd83xxx(struct hda_codec *codec) spec->dmux_nids = stac92hd83xxx_dmux_nids; spec->adc_nids = stac92hd83xxx_adc_nids; spec->pwr_nids = stac92hd83xxx_pwr_nids; + spec->amp_nids = stac92hd83xxx_amp_nids; spec->pwr_mapping = stac92hd83xxx_pwr_mapping; spec->num_pwrs = ARRAY_SIZE(stac92hd83xxx_pwr_nids); spec->multiout.dac_nids = spec->dac_nids; @@ -4689,6 +4694,7 @@ static int patch_stac92hd83xxx(struct hda_codec *codec) spec->num_pins = ARRAY_SIZE(stac92hd83xxx_pin_nids); spec->num_dmuxes = ARRAY_SIZE(stac92hd83xxx_dmux_nids); spec->num_adcs = ARRAY_SIZE(stac92hd83xxx_adc_nids); + spec->num_amps = ARRAY_SIZE(stac92hd83xxx_amp_nids); spec->num_dmics = STAC92HD83XXX_NUM_DMICS; spec->dinput_mux = &stac92hd83xxx_dmux; spec->pin_nids = stac92hd83xxx_pin_nids;

2 2

Re: [alsa-devel] [PATCH] ALSA: hda - Add automatic model setting for Samsung Q45
by Luke Yelavich 14 Jan '09

14 Jan '09

Sorry to take so long to reply. This quirk was from Ubuntu bug 200210, http://launchpad.net/bugs/200210. The patch was based on a comment from Daniel T Chen, indicating the quirk needed for the bug. Luke

2 1

[alsa-devel] set_params:1073: not enough memory
by Michael Hirsch 13 Jan '09

13 Jan '09

Hello all, I'm trying to get alsa 1.0.18 working on a ARM PXA processor and I'm stuck with the following error message: aplay /Sinus.raw -fcd Playing raw data '/Sinus.raw' : Signed 16 bit Little Endian, Rate 44100 Hz, Stereo aplay: set_params:1073: not enough memory alsa kernel was compiled with alsa 1.0.16, alsa.conf, located in /usr/share/alsa is the basic: pcm.!default { type hw card 0 } ctl.!default { type hw card 0 } also aplay -l and alsaconf store show the message: ALSA lib control.c:909:(snd_ctl_open_noupdate) Invalid CTL hw:0 However the card shows up /proc/asound/cards and alsamixer is able to access it. Any ideas? Best regards Michael

1 0

Re: [alsa-devel] Fixing the AICA driver
by Takashi Iwai 13 Jan '09

13 Jan '09

At Thu, 08 Jan 2009 18:50:46 +0000, Adrian McMenamin wrote: > > On Wed, 2009-01-07 at 12:02 +0100, Takashi Iwai wrote: > > > > > Could you try the git version? I thought I fixed the relevant bug recently. > > > > > > Takashi > > Hi Takashi - I assume you mean in the alssa-driver git, as I am using > the bleeding edge kernel stuff already? No, I mean alsa-lib. There is mmap_emul plugin in alsa-lib but it didn't work well. Takashi

2 1

[alsa-devel] Recorded frames sign 'error' after switchign to EABI kernel?
by Edwin van der Horst 13 Jan '09

13 Jan '09

Hello all, I'm using the ALSA API for quite some time now in my application, all worked fine until I switched to EABI kernel/compiler. My application reads signed 16 bit frames (S16_LE) using readi call: long FCAudio::readRecordedPCM(char * dest,int len) { .... frames = snd_pcm_readi(recHandle, dest, (len / asRec.bytesPS)); //read data hw .... .... } This worked fine on the non EABi kernel. After I switched to the EABI enabled kernel and compiler, I heard massive 'hiss/noise' on recorded audio. I took a dump of the recorded frames, and this came out: sample data: | ff | 58 | ff | e1 | ff | e2 | 0 | 0 | 0 | 48 | ff | e0 | ff | e3 | ff | e1 | 0 | 13 | ff | ca | ff | ae | 0 | 3 | ff | ee | ff | be| at first this seemed like correct data to me , but on a closer look the MSB(yte) is always 0x00 or 0xff I can't figure out why this is: I'm expecting 2 bytes of sample data, and in the current case I get 1 byte of correct data, and the other byte is only used for sign indication (0x00 or 0xFF). I can't figure out why this is, can anyone point me in the right direction? It's worth noting that arecord and aplay both work correctly, and using snd_pcm_dump(); I get settings as expedted and are correct. Thanks in advance... Edwin van der Horst

1 0

[alsa-devel] [PATCH - SB0500 2/2] rename "Device" to "Toshiba SB-0500" via quirks
by Andrea Borgia 13 Jan '09

13 Jan '09

Signed-off-by: Andrea Borgia <andrea(a)borgia.bo.it> diff --git a/usb/usbquirks.h b/usb/usbquirks.h index 9211575..d59323e 100644 --- a/usb/usbquirks.h +++ b/usb/usbquirks.h @@ -39,6 +39,16 @@ .idProduct = prod, \ .bInterfaceClass = USB_CLASS_VENDOR_SPEC +/* Creative/Toshiba Multimedia Center SB-0500 */ +{ + USB_DEVICE(0x041e, 0x3048), + .driver_info = (unsigned long) & (const struct snd_usb_audio_quirk) { + .vendor_name = "Toshiba", + .product_name = "SB-0500", + .ifnum = QUIRK_NO_INTERFACE + } +}, + /* Creative/E-Mu devices */ { USB_DEVICE(0x041e, 0x3010), -- 1.5.4.3

2 2