December 2009 - Alsa-devel - mailman.alsa-project.org

[alsa-devel] X-Fi notebook support
by The Source 30 Dec '09

30 Dec '09

Hello. I have some info on that card, but not much. I have discovered that this card is detected connected via USB rather than PCI-E (the slot is ExpressCard). snd_ctxfi uses PCI interfaces right? Can X-Fi support be also added to snd_usb_audio? Or can snd_ctxfi support USB?

1 0

[alsa-devel] Front jack not working with hda-intel driver SigmaTel STAC9271D codec
by David Mohr 29 Dec '09

29 Dec '09

Hi, I'm running ubuntu 9.10. I had to force model=5stack to get any audio output at all. While the jack in the back now works, I still don't get any sound on the front jack. I haven't found any options in alsamixer, and switching the output mode in pavucontrol to the headphones just makes the sound disappear from the back jack, but I still get nothing in the front. I uploaded my alsa-info.sh output here: http://www.alsa-project.org/db/?f=3ed2ab9e7732bf08d2bbfbe8adb74bf037d9e618 Can I fix this somehow, or is it a bug? And of course thanks everyone for the usually great working alsa system. ~David

2 3

[alsa-devel] ASoC updates for 2.6.33
by Mark Brown 29 Dec '09

29 Dec '09

The following changes since commit 6b7b284958d47b77d06745b36bc7f36dab769d9b: Linus Torvalds (1): Linux 2.6.33-rc2 are available in the git repository at: git://git.kernel.org/pub/scm/linux/kernel/git/broonie/sound-2.6.git for-2.6.33 The change in the branch is much bigger than the diffstat shows due to the merge up of Linus' tree (which now builds for ARM). Guennadi Liakhovetski (1): ASoC: fix params_rate() macro use in several codecs Kuninori Morimoto (1): ASoC: fsi-ak4642: Remove ak4642_add_i2c_device Mark Brown (1): Merge commit 'v2.6.33-rc2' into for-2.6.33 sound/soc/codecs/wm8510.c | 14 +++++++------- sound/soc/codecs/wm8940.c | 14 +++++++------- sound/soc/codecs/wm8974.c | 14 +++++++------- sound/soc/sh/fsi-ak4642.c | 30 ------------------------------ 4 files changed, 21 insertions(+), 51 deletions(-)

2 1

[alsa-devel] DMA and delay feedback
by James Courtier-Dutton 29 Dec '09

29 Dec '09

Hi, Would there be benefit in separating the DMA and the transfer of audio samples to the sound card hardware from the delay and hw position feedback. For example, If the playback position is 0, the hardware has probably already transferred one period of the sound card. So, one cannot write to playback position 2 and expect it to be output to the sound card if the period size is 1024. For transfer purposes, one only needs to know that DMA transfer is complete on period X so that period X can now be over-written. I believe that this is what the DMA interrupt indicates. One needs to reliably detect a missed interrupt. It one took this approach, one might be able to hide the "missed interrupt" problem. When the interrupt finally did arrive, we would know where the next free period X is, and continue writing samples to that, irrespective of what went wrong before. This would allow for auto-recovery of the audio stream without having to stop and start the ring buffers again. One then needs to determine how one can get accurate "delay" values back from all sound card types. The "delay" value is used for applications like ensuring that audio and video play in sync. The problem ones at the moment appear to be the intel8x0 and hd-audio ones. There may be benefits from making the "delay" value return in nanoseconds instead of samples, particularly if gettimeofdate() is being used to try to produce accurate "delay" values. Kind Regards James

4 6

[alsa-devel] frequency analysis with alsa and fftw
by Markus Vogl 28 Dec '09

28 Dec '09

hello, i´m absolute newbie in programming alsa, but for my school project i have to do so. my project is to analyse the mic input frequency from my asus eepc 1008ha with ubuntu 9.10 and show it in a terminal window. i would use my acoustic 6 string guitar to play in front of my netbook. i´ve testet the howto from linuxjornal.com .. http://www.linuxjournal.com/article/6735 ...but i´ve no plan to use the mic input and then analyse it with fftw. regards markus -- Jetzt kostenlos herunterladen: Internet Explorer 8 und Mozilla Firefox 3.5 - sicherer, schneller und einfacher! http://portal.gmx.net/de/go/atbrowser

2 4

[alsa-devel] [PATCH - usb/usbaudio.c 1/1] sound/usb: Relax urb data align. restriction for HVR-950Q and HVR-850 only
by John S Gruber 28 Dec '09

28 Dec '09

Addressing audio quality problem. In sound/usb/usbaudio.c, for the Hauppage HVR-950Q only, change retire_capture_urb to copy the entire byte stream while still counting entire audio frames. urbs unaligned on channel sample boundaries are still truncated to the next lowest stride (audio slot) size to try to retain channel alignment in cases of data loss over usb. With the HVR950Q the left and right channel samples can be split between two different urbs. Throwing away extra channel samples causes a sound quality problem for stereo streams as the left and right channels are swapped repeatedly. BugLink: https://bugs.launchpad.net/ubuntu/+bug/495745 Signed-off-by: John S. Gruber <JohnSGruber(a)gmail.com> --- usb/usbaudio.c | 53 ++++++++++++++++++++++++++++++++++++++++------------- 1 files changed, 40 insertions(+), 13 deletions(-) diff --git a/usb/usbaudio.c b/usb/usbaudio.c index b074a59..d450b23 100644 --- a/usb/usbaudio.c +++ b/usb/usbaudio.c @@ -108,6 +108,7 @@ MODULE_PARM_DESC(ignore_ctl_error, #define MAX_URBS 8 #define SYNC_URBS 4 /* always four urbs for sync */ #define MAX_QUEUE 24 /* try not to exceed this queue length, in ms */ +#define ALLOW_SUBSLOT_BOUNDARIES 0x01 /* quirk */ struct audioformat { struct list_head list; @@ -127,6 +128,7 @@ struct audioformat { unsigned int rate_min, rate_max; /* min/max rates */ unsigned int nr_rates; /* number of rate table entries */ unsigned int *rate_table; /* rate table */ + unsigned int txfr_quirks; /* transfer quirks */ }; struct snd_usb_substream; @@ -175,6 +177,8 @@ struct snd_usb_substream { unsigned int running: 1; /* running status */ unsigned int hwptr_done; /* processed frame position in the buffer */ + unsigned int byteptr; /* position, in bytes, of next move */ + unsigned int txfr_quirks; /* substream transfer quirks */ unsigned int transfer_done; /* processed frames since last period update */ unsigned long active_mask; /* bitmask of active urbs */ unsigned long unlink_mask; /* bitmask of unlinked urbs */ @@ -343,7 +347,7 @@ static int retire_capture_urb(struct snd_usb_substream *subs, unsigned long flags; unsigned char *cp; int i; - unsigned int stride, len, oldptr; + unsigned int stride, len, bytelen, oldbyteptr; int period_elapsed = 0; stride = runtime->frame_bits >> 3; @@ -354,29 +358,39 @@ static int retire_capture_urb(struct snd_usb_substream *subs, snd_printd(KERN_ERR "frame %d active: %d\n", i, urb->iso_frame_desc[i].status); // continue; } - len = urb->iso_frame_desc[i].actual_length / stride; - if (! len) + bytelen = (urb->iso_frame_desc[i].actual_length); + if (!bytelen) continue; + if (bytelen%(runtime->sample_bits>>3) != 0) { + int oldbytelen = bytelen; + bytelen = ((bytelen/stride)*stride); + snd_printd(KERN_ERR "Corrected urb data len. %d->%d\n", + oldbytelen, bytelen); + } + if (!(subs->txfr_quirks & ALLOW_SUBSLOT_BOUNDARIES)) + bytelen = (bytelen/stride)*stride; /* update the current pointer */ spin_lock_irqsave(&subs->lock, flags); - oldptr = subs->hwptr_done; - subs->hwptr_done += len; - if (subs->hwptr_done >= runtime->buffer_size) - subs->hwptr_done -= runtime->buffer_size; + len = (bytelen + (subs->byteptr % stride)) / stride; subs->transfer_done += len; if (subs->transfer_done >= runtime->period_size) { subs->transfer_done -= runtime->period_size; period_elapsed = 1; } + oldbyteptr = subs->byteptr; + subs->byteptr += bytelen; + if (subs->byteptr >= runtime->buffer_size*stride) + subs->byteptr -= runtime->buffer_size*stride; + subs->hwptr_done = subs->byteptr/stride; spin_unlock_irqrestore(&subs->lock, flags); /* copy a data chunk */ - if (oldptr + len > runtime->buffer_size) { - unsigned int cnt = runtime->buffer_size - oldptr; - unsigned int blen = cnt * stride; - memcpy(runtime->dma_area + oldptr * stride, cp, blen); - memcpy(runtime->dma_area, cp + blen, len * stride - blen); + if ((oldbyteptr + bytelen) > (runtime->buffer_size*stride)) { + unsigned int blen; + blen = (runtime->buffer_size*stride) - oldbyteptr; + memcpy(runtime->dma_area + oldbyteptr, cp, blen); + memcpy(runtime->dma_area, cp + blen, bytelen - blen); } else { - memcpy(runtime->dma_area + oldptr * stride, cp, len * stride); + memcpy(runtime->dma_area + oldbyteptr, cp, bytelen); } } if (period_elapsed) @@ -1363,6 +1377,7 @@ static int set_format(struct snd_usb_substream *subs, struct audioformat *fmt) subs->syncpipe = subs->syncinterval = 0; subs->maxpacksize = fmt->maxpacksize; subs->fill_max = 0; + subs->txfr_quirks = fmt->txfr_quirks; /* we need a sync pipe in async OUT or adaptive IN mode */ /* check the number of EP, since some devices have broken @@ -1532,6 +1547,7 @@ static int snd_usb_pcm_prepare(struct snd_pcm_substream *substream) /* reset the pointer */ subs->hwptr_done = 0; subs->transfer_done = 0; + subs->byteptr = 0; subs->phase = 0; runtime->delay = 0; @@ -2776,6 +2792,7 @@ static int parse_audio_endpoints(struct snd_usb_audio *chip, int iface_no) fp->maxpacksize = (((fp->maxpacksize >> 11) & 3) + 1) * (fp->maxpacksize & 0x7ff); fp->attributes = csep ? csep[3] : 0; + fp->txfr_quirks = 0; /* some quirks for attributes here */ @@ -2804,6 +2821,16 @@ static int parse_audio_endpoints(struct snd_usb_audio *chip, int iface_no) else fp->ep_attr |= EP_ATTR_SYNC; break; + case USB_ID(0x2040, 0x7200): /* Hauppage hvr950Q */ + case USB_ID(0x2040, 0x7221): /* Hauppage hvr950Q */ + case USB_ID(0x2040, 0x7222): /* Hauppage hvr950Q */ + case USB_ID(0x2040, 0x7223): /* Hauppage hvr950Q */ + case USB_ID(0x2040, 0x7224): /* Hauppage hvr950Q */ + case USB_ID(0x2040, 0x7225): /* Hauppage hvr950Q */ + case USB_ID(0x2040, 0x7230): /* Hauppage hvr850 */ + case USB_ID(0x2040, 0x7250): /* Hauppage hvr950Q */ + fp->txfr_quirks |= ALLOW_SUBSLOT_BOUNDARIES; + break; } /* ok, let's parse further... */ -- 1.6.3.3 I'm new to changing the kernel. Normally the driver should transfer only whole audio frames, disregarding fragments, in order to keep channel and sample synchronization in the face of data loss on the bus. That's how the current driver works. The patch relaxes this (only) for devices with the quirk. An earlier version of this patch was posted recently to the linux-media list. No one there suggested a way to make the au0828 chip in this equipment behave better. They kindly forwarded me to this mailing list for review and consideration of the patch. To test the patch with a non-quirk device run it on a busy or problematic bus and listen for audio switching channels. To test with a HVR-950Q use composite input and disconnect either the left or right channel. Without the patch the audio will alternate between channels quickly, most easily heard with headphones or earphones, producing a flutter sound, With the patch the remaining test audio channel appears only on one channel and is steady.

5 8

[alsa-devel] [PATCH] ALSA: test off by one in setsamplerate()
by Roel Kluin 28 Dec '09

28 Dec '09

With `while (i++ < MAX_WRITE_RETRY)' i reaches MAX_WRITE_RETRY + 1 after the loop Signed-off-by: Roel Kluin <roel.kluin(a)gmail.com> --- sound/pci/riptide/riptide.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/sound/pci/riptide/riptide.c b/sound/pci/riptide/riptide.c index b5ca02e..e66ef2b 100644 --- a/sound/pci/riptide/riptide.c +++ b/sound/pci/riptide/riptide.c @@ -1058,7 +1058,7 @@ setsamplerate(struct cmdif *cif, unsigned char *intdec, unsigned int rate) rptr.retwords[2] != M && rptr.retwords[3] != N && i++ < MAX_WRITE_RETRY); - if (i == MAX_WRITE_RETRY) { + if (i > MAX_WRITE_RETRY) { snd_printdd("sent samplerate %d: %d failed\n", *intdec, rate); return -EIO;

2 1

[alsa-devel] [PATCH] ALSA: atiixp: Specify codec for Foxconn RC4107MA-RS2
by Daniel Chen 28 Dec '09

28 Dec '09

Hi, Attached is a patch against sound-2.6.git and also targeted for stable to make sound audible by default for this mainboard. Thanks, Dan

2 1

[alsa-devel] [PATCH 1/1] Added functionality for E-mu 0404USB/0202USB/TrackerPre
by Zerg Cannibal 28 Dec '09

28 Dec '09

Added functionality: 1) Extension Units support (all XU settings now available at alsamixer, kmix, etc): - "AnalogueIn soft limiter" switch; - "Sample rate" selector (values 0,1,2,3,4,5 corresponds to 44.1 48 ... 192 kHz); - "DigitalIn CLK source" selector (internal/external) (**); - "DigitalOut format SPDIF/AC3" switch (**); (**)E-mu-0404usb only. 2) Automatic device sample rate adjustment depending on substream samplerate for both capture and playback substream. Patch is based on Mac OS-X driver (http://sourceforge.net/projects/zaudiodrivermac/) and successfully tested by Ubuntu community users for 0404usb and 0202usb. Signed-off-by: CannibalZerg <cnb_zerg(a)yahoo.com> --- sound/usb/usbaudio.c | 44 ++++++++++++++++++++++++++++++ sound/usb/usbaudio.h | 2 + sound/usb/usbmixer.c | 73 +++++++++++++++++++++++++++++++++++++++++++++++--- 3 files changed, 115 insertions(+), 4 deletions(-) diff --git a/sound/usb/usbaudio.c b/sound/usb/usbaudio.c index 31b63ea..d8eb029 100644 --- a/sound/usb/usbaudio.c +++ b/sound/usb/usbaudio.c @@ -1271,6 +1271,42 @@ static int init_usb_sample_rate(struct usb_device *dev, int iface, } /* + * For E-Mu 0404USB/0202USB/TrackerPre sample rate should be set for device, + * not for interface. +*/ +static void set_format_emu_quirk(struct snd_usb_substream *subs, struct audioformat *fmt) +{ + unsigned char emu_samplerate_id = 0; + + /* When capture is active sample rate shouldn't be changed by playback substream*/ + if (subs->direction == SNDRV_PCM_STREAM_PLAYBACK) { + if (subs->stream->substream[SNDRV_PCM_STREAM_CAPTURE].interface != -1) + return; + } + + switch (fmt->rate_min) { + case 48000: + emu_samplerate_id = 1; + break; + case 88200: + emu_samplerate_id = 2; + break; + case 96000: + emu_samplerate_id = 3; + break; + case 176400: + emu_samplerate_id = 4; + break; + case 192000: + emu_samplerate_id = 5; + break; + default: + emu_samplerate_id = 0; + } + snd_emuusb_set_samplerate(subs->stream->chip, emu_samplerate_id); +} + +/* * find a matching format and set up the interface */ static int set_format(struct snd_usb_substream *subs, struct audioformat *fmt) @@ -1383,6 +1419,14 @@ static int set_format(struct snd_usb_substream *subs, struct audioformat *fmt) subs->cur_audiofmt = fmt; + switch (subs->stream->chip->usb_id) { + case USB_ID(0x041e, 0x3f02): /* E-Mu 0202 USB */ + case USB_ID(0x041e, 0x3f04): /* E-Mu 0404 USB */ + case USB_ID(0x041e, 0x3f0a): /* E-Mu Tracker Pre */ + set_format_emu_quirk(subs, fmt); + break; + } + #if 0 printk(KERN_DEBUG "setting done: format = %d, rate = %d..%d, channels = %d\n", diff --git a/sound/usb/usbaudio.h b/sound/usb/usbaudio.h index 9826337..8558773 100644 --- a/sound/usb/usbaudio.h +++ b/sound/usb/usbaudio.h @@ -233,6 +233,8 @@ void snd_usbmidi_input_stop(struct list_head* p); void snd_usbmidi_input_start(struct list_head* p); void snd_usbmidi_disconnect(struct list_head *p); +void snd_emuusb_set_samplerate(struct snd_usb_audio *chip, unsigned char samplerate_id); + /* * retrieve usb_interface descriptor from the host interface * (conditional for compatibility with the older API) diff --git a/sound/usb/usbmixer.c b/sound/usb/usbmixer.c index c998220..400208c 100644 --- a/sound/usb/usbmixer.c +++ b/sound/usb/usbmixer.c @@ -186,6 +186,21 @@ enum { USB_PROC_DCR_RELEASE = 6, }; +/*E-mu 0202(0404) eXtension Unit(XU) control*/ +enum { + USB_XU_CLOCK_RATE = 0xe301, + USB_XU_CLOCK_SOURCE = 0xe302, + USB_XU_DIGITAL_IO_STATUS = 0xe303, + USB_XU_DEVICE_OPTIONS = 0xe304, + USB_XU_DIRECT_MONITORING = 0xe305, + USB_XU_METERING = 0xe306 +}; +enum { + USB_XU_CLOCK_SOURCE_SELECTOR = 0x02, /* clock source*/ + USB_XU_CLOCK_RATE_SELECTOR = 0x03, /* clock rate */ + USB_XU_DIGITAL_FORMAT_SELECTOR = 0x01, /* the spdif format */ + USB_XU_SOFT_LIMIT_SELECTOR = 0x03 /* soft limiter */ +}; /* * manual mapping of mixer names @@ -1330,7 +1345,32 @@ static struct procunit_info procunits[] = { { USB_PROC_DCR, "DCR", dcr_proc_info }, { 0 }, }; - +/* + * predefined data for extension units + */ +static struct procunit_value_info clock_rate_xu_info[] = { + { USB_XU_CLOCK_RATE_SELECTOR, "Selector", USB_MIXER_U8, 0}, + { 0 } +}; +static struct procunit_value_info clock_source_xu_info[] = { + { USB_XU_CLOCK_SOURCE_SELECTOR, "External", USB_MIXER_BOOLEAN}, + { 0 } +}; +static struct procunit_value_info spdif_format_xu_info[] = { + { USB_XU_DIGITAL_FORMAT_SELECTOR, "SPDIF/AC3", USB_MIXER_BOOLEAN}, + { 0 } +}; +static struct procunit_value_info soft_limit_xu_info[] = { + { USB_XU_SOFT_LIMIT_SELECTOR, " ", USB_MIXER_BOOLEAN}, + { 0 } +}; +static struct procunit_info extunits[] = { + { USB_XU_CLOCK_RATE, "Clock rate", clock_rate_xu_info }, + { USB_XU_CLOCK_SOURCE, "DigitalIn CLK source", clock_source_xu_info }, + { USB_XU_DIGITAL_IO_STATUS, "DigitalOut format:", spdif_format_xu_info }, + { USB_XU_DEVICE_OPTIONS, "AnalogueIn Soft Limit", soft_limit_xu_info }, + { 0 } +}; /* * build a processing/extension unit */ @@ -1391,8 +1431,16 @@ static int build_audio_procunit(struct mixer_build *state, int unitid, unsigned cval->max = dsc[15]; cval->res = 1; cval->initialized = 1; - } else - get_min_max(cval, valinfo->min_value); + } else { + if (type == USB_XU_CLOCK_RATE) { + /*E-Mu USB 0404/0202/TrackerPre samplerate control quirk*/ + cval->min = 0; + cval->max = 5; + cval->res = 1; + cval->initialized = 1; + } else + get_min_max(cval, valinfo->min_value); + } kctl = snd_ctl_new1(&mixer_procunit_ctl, cval); if (! kctl) { @@ -1433,7 +1481,7 @@ static int parse_audio_processing_unit(struct mixer_build *state, int unitid, un static int parse_audio_extension_unit(struct mixer_build *state, int unitid, unsigned char *desc) { - return build_audio_procunit(state, unitid, desc, NULL, "Extension Unit"); + return build_audio_procunit(state, unitid, desc, extunits, "Extension Unit"); } @@ -2109,6 +2157,23 @@ static int snd_xonar_u1_controls_create(struct usb_mixer_interface *mixer) return 0; } +void snd_emuusb_set_samplerate(struct snd_usb_audio *chip, unsigned char samplerate_id) +{ + struct usb_mixer_interface *mixer; + struct list_head *p; + struct usb_mixer_elem_info *cval; + int unitid = 12; /*SampleRate ExtensionUnit ID*/ + list_for_each(p, &chip->mixer_list) { + mixer = list_entry(p, struct usb_mixer_interface, list); + cval = mixer->id_elems[unitid]; + if (cval) { + set_cur_ctl_value(cval, cval->control << 8, samplerate_id); + snd_usb_mixer_notify_id(mixer, unitid); + } + break; + } +} + int snd_usb_create_mixer(struct snd_usb_audio *chip, int ctrlif, int ignore_error) { -- 1.6.5.6

4 7

[alsa-devel] Use of default device with snd_pcm_link()
by Devin Heitmueller 28 Dec '09

28 Dec '09

I am working on some code which captures from one PCM device and plays to another, based off of the latency.c found in alsa-lib. I noticed something strange: The code works if I set the capture device to hw:1,0 and playback device to hw:0,0. However, if I change the playback device to "default:0", the snd_pcm_link() call fails with -ENOSYS (error 38). If I dump out the properties, I can see that the the default device *is* mapped to device 0,0, so it's not clear why the behaviour would differ. My goal was just to playback to whatever the default device is, and I was under the impression that default:0 should be equivalent to "hw:0,0". Can anyone offer any brief explanation as to why the snd_pcm_link() call would fail? Thanks in advance, Devin -- Devin J. Heitmueller - Kernel Labs http://www.kernellabs.com

2 1