December 2008 - Alsa-devel - mailman.alsa-project.org

[alsa-devel] How to write a PXA CPU
by Juan Pablo Gómez 09 Dec '08

09 Dec '08

Hi, I'm looking for documentation that explains me how to write a PXA document. It's for a NIOS 2 processor in an altera De2 board. -- Regards, Juan Pablo Muchas gracias. Atentamente, Juan Pablo Gómez

2 3

[alsa-devel] [PATCH 0/9] ASoC: TWL4030: DAPM implementation for outputs (playback) - try3
by Peter Ujfalusi 09 Dec '08

09 Dec '08

Hello, The whole series again, altrough 1 and 2 has been accepted already. This series of patches (try3) adds DAPM implementation for all outputs found on TWL. The design allows one to configure all outputs independently from each other, since the internal routing of TWL actually treats them individually. The pre-DAC routings can be configured in the board specific files of needed. I have sent a patch which has added this to the codec driver. Based on the output routing selections all the corresponding DAC and analog PGA on the chain will be powered on. Remove the Headset output gain modifications from the twl4030_power_up and twl4030_power_down function, since they are not needed. Changes since try1 (comments addressed): - Removed the " Switch" from the option texts and also from the intercon. --- Peter Ujfalusi (9): ASoC: TWL4030: Correct DAPM_DAC with power control ASoC: TWL4030: Add Analog PGA control switch to DAPM ASoC: TWL4030: Add DAPM event handler for output MUX selection ASoC: TWL4030: DAPM mapping of the Earpiece output ASoC: TWL4030: DAPM mapping of the PreDriv outputs ASoC: TWL4030: DAPM mapping of the Headset outputs ASoC: TWL4030: DAPM mapping of the Carkit outputs ASoC: TWL4030: DAPM mapping of the Handsfree outputs ASoC: TWL4030: Do not alter the Headset output volume on power-up/down sound/soc/codecs/twl4030.c | 263 +++++++++++++++++++++++++++++++++++++++++--- 1 files changed, 248 insertions(+), 15 deletions(-)

2 10

[alsa-devel] [PATCH 0/9] ASoC: TWL4030: DAPM implementation for outputs (playback) - try2
by Peter Ujfalusi 09 Dec '08

09 Dec '08

Hello, This series of patches (try2) adds DAPM implementation for all outputs found on TWL. The design allows one to configure all outputs independently from each other, since the internal routing of TWL actually treats them individually. The pre-DAC routings can be configured in the board specific files of needed. I have sent a patch which has added this to the codec driver. Based on the output routing selections all the corresponding DAC and analog PGA on the chain will be powered on. Remove the Headset output gain modifications from the twl4030_power_up and twl4030_power_down function, since they are not needed. Changes since try1 (comments addressed): - erapiece typo - too fast/slow finger combination ;) - Removed the " Switch" from the option texts - rename the confusing _controls to _control - Analog PGA control reworked: - SND_SOC_DAPM_PGA for power switch - Normal control stereo (user control) for DA enable/disable (mute/unmute): - "DAC1 Analog Playback Volume" + "DAC1 Analog Playback Switch" - "DAC2 Analog Playback Volume" + "DAC2 Analog Playback Switch" --- Peter Ujfalusi (9): ASoC: TWL4030: Correct DAPM_DAC with power control ASoC: TWL4030: Add Analog PGA control switch to DAPM ASoC: TWL4030: Add DAPM event handler for output MUX selection ASoC: TWL4030: DAPM mapping of the Earpiece output ASoC: TWL4030: DAPM mapping of the PreDriv outputs ASoC: TWL4030: DAPM mapping of the Headset outputs ASoC: TWL4030: DAPM mapping of the Carkit outputs ASoC: TWL4030: DAPM mapping of the Handsfree outputs ASoC: TWL4030: Do not alter the Headset output volume on power-up/down sound/soc/codecs/twl4030.c | 263 +++++++++++++++++++++++++++++++++++++++++--- 1 files changed, 248 insertions(+), 15 deletions(-)

2 13

[alsa-devel] [PATCH 0/8] ASoC: TWL4030: DAPM implementation for outputs (playback)
by Peter Ujfalusi 09 Dec '08

09 Dec '08

Hello, This series of patches adds DAPM implementation for all outputs found on TWL. The design allows one to configure all outputs independently from each other, since the internal routing of TWL actually treats them individually. The pre-DAC routings can be configured in the board specific files of needed. I have sent a patch which has added this to the codec driver. Based on the output routing selections all the corresponding DAC and analog PGA on the chain will be powered on. --- Peter Ujfalusi (8): ASoC: TWL4030: Correct DAPM_DAC with power control ASoC: TWL4030: Add Analog PGA control switch to DAPM ASoC: TWL4030: Add DAPM event handler for output MUX selection ASoC: TWL4030: DAPM mapping of the Earpiece output ASoC: TWL4030: DAPM mapping of the PreDriv outputs ASoC: TWL4030: DAPM mapping of the Headset outputs ASoC: TWL4030: DAPM mapping of the Carkit outputs ASoC: TWL4030: DAPM mapping of the Handsfree outputs sound/soc/codecs/twl4030.c | 248 +++++++++++++++++++++++++++++++++++++++++++- 1 files changed, 243 insertions(+), 5 deletions(-)

2 20

[alsa-devel] ASoC: hook for codec control updates and clock controls
by Daniel Mack 08 Dec '08

08 Dec '08

Hi, Codecs like the tlv320aic3x family have soft volume controls to smoothly drive up and down mixer elements in order to prevent clicks and pops. In fact, most of the controls on this chip are implemented as such. For the mixers to work, however, the I2S input clock is needed which is currently only true when the DAC/ACD elements are active but not when the chip is just operating as an analog mixer. To allow this functionality, the I2S clock must be present for a short period after each volume control update to the codec and disabled again after some milliseconds. It should, however, not be enabled if it's already running and of course, it must not be disabled if the ADC/DAC is still running. Adding a hook to struct snd_soc_codec which is called from snd_soc_update_bits() isn't a big deal, but the problem is the abstraction layers in this case - for good reasons, the cpu_dai is not accessible from the codec dai. Or did I miss a link? Is there a sane way to address this problem? Thanks and best regards, Daniel

3 9

[alsa-devel] [patch] alsa-utils fix for intl support
by Hans-Christian Egtvedt 08 Dec '08

08 Dec '08

Hi, Feel free to comment on the included patch. I had to add it because aseqnet would not link without -lintl. -- Best regards, Hans-Christian Egtvedt

2 5

[alsa-devel] [patch] alsa-utils check if __USE_BSD is defined before compiling "BSD functions"
by Hans-Christian Egtvedt 08 Dec '08

08 Dec '08

Hi, Another bug/issue I tripped over when compiling alsa-utils in an environment using uClibc to supply the C library functions. Here I have enabled some old BSD style functions. The attached patch will honor them if they are enabled. Without this patch I get a redefined error during compile. -- Best regards, Hans-Christian Egtvedt

2 3

[alsa-devel] [PATCH 0/7] ASoC updates
by Mark Brown 08 Dec '08

08 Dec '08

The following changes since commit 682d5874f3d654b5d13d9b8dd56b9e05cfadd01b: Takashi Iwai (1): ALSA: ASoC: Fix old-style trigger callback in s3c2443-ac97.c are available in the git repository at: git://opensource.wolfsonmicro.com/linux-2.6-asoc for-tiwai Mark Brown (4): ASoC: Rename snd_soc_register_card() to snd_soc_init_card() ASoC: Annotate core removal function ASoC: Remove in-code changelog from AD73311 driver ASoC: Don't free static data in WM9713 Peter Ujfalusi (3): ASoC: TWL4030: Change the capture volume control to TLV ASoC: TWL4030: Change the common playback volume controls ASoC: TWL4030: Add volume controls for outputs include/sound/soc.h | 2 +- sound/soc/codecs/ac97.c | 2 +- sound/soc/codecs/ad1980.c | 2 +- sound/soc/codecs/ad73311.c | 5 +-- sound/soc/codecs/ak4535.c | 2 +- sound/soc/codecs/cs4270.c | 2 +- sound/soc/codecs/pcm3008.c | 2 +- sound/soc/codecs/ssm2602.c | 2 +- sound/soc/codecs/tlv320aic23.c | 2 +- sound/soc/codecs/tlv320aic26.c | 2 +- sound/soc/codecs/tlv320aic3x.c | 2 +- sound/soc/codecs/twl4030.c | 74 +++++++++++++++++++++++++++++++++------ sound/soc/codecs/uda134x.c | 2 +- sound/soc/codecs/uda1380.c | 2 +- sound/soc/codecs/wm8510.c | 2 +- sound/soc/codecs/wm8580.c | 2 +- sound/soc/codecs/wm8728.c | 2 +- sound/soc/codecs/wm8731.c | 2 +- sound/soc/codecs/wm8750.c | 2 +- sound/soc/codecs/wm8753.c | 2 +- sound/soc/codecs/wm8900.c | 2 +- sound/soc/codecs/wm8903.c | 2 +- sound/soc/codecs/wm8971.c | 2 +- sound/soc/codecs/wm8990.c | 2 +- sound/soc/codecs/wm9712.c | 2 +- sound/soc/codecs/wm9713.c | 3 +- sound/soc/soc-core.c | 8 ++-- 27 files changed, 91 insertions(+), 45 deletions(-)

4 12

[alsa-devel] HDA: CLFE DAC on ALC889 (Acer Aspire 8930G)
by Hector Martin 07 Dec '08

07 Dec '08

I've been trying to get ALSA working with this laptop. So far the attached patch adds rear support to the existing 4930g model, and tries to add CLFE support. However, I haven't been able to get CLFE to work. I don't seem to get any audio through DACs/mixers other than the first two (front and rear). I can assign the CLFE output (pin 0x16) to the front or rear DAC and hear the front audio through the CLFE speakers. I can also force the front DAC (0x02) or rear DAC (0x03) to use the CLFE stream and hear it through the front or rear speakers, so the audio is getting to the codec. But either DACs 0x04 and 0x05 aren't working, or their associated mixers (0x0e and 0x0f) aren't passing through the audio, or the OUT pins don't like selections above 1. I've checked that they're unmuted and that the volumes are turned up on both. I've also tried the 0x25->0x26 DAC/mixer auxiliary pathway, which doesn't work either. Here's what the relevant NIDs correspond to: 0x02 - (normally) Front DAC 0x03 - (normally) Rear DAC 0x04 - (normally) CLFE DAC 0x05 - (presumably) side DAC (no side speakers) 0x0c - (normally) Front mixer 0x0d - (normally) Rear mixer 0x0e - (normally) CLFE mixer 0x0f - (presumably) side mixer 0x14 - Front speakers pin 0x15 - Headphones pin (also EAPD for speaker amp) 0x16 - CLFE speakers pin 0x1b - Rear speakers pin Here's what the codec graph looks like: http://marcansoft.com/transf/codec.png Any ideas? I'm pretty stumped here. Are there any other magic bits to enable the remaining DACs or mixers? -- Hector Martin (hector(a)marcansoft.com) Public Key: http://www.marcansoft.com/marcan.asc

2 3

Re: [alsa-devel] [Swfdec] Can't use Mozilla plugin or play video with 0.8.2 / 0.9.2
by Benjamin Otte 07 Dec '08

07 Dec '08

I think I know what this is. But I spent the best part of my breakfast wondering about it. I think using Pulse Audio might work around this problem, but I think it's a driver, configuration issue on the ALSA side or a misunderstanding of the API on my part, which is why I'm sending this to the ALSA list. So here's the problem: Swfdec uses ALSA in non-blocking mode and polls its file descriptor(s). Whenever the audio queue becomes empty enough that it's useful to render more audio, the file descriptor wakes up from the poll. This setting is apparently configurable using various hw params, however, Swfdec does not configure it and has been relying on the fact that the sound device picks proper defaults, thinking that a sound card with a small buffer will likely want to wake up more often than something like Pulse Audio's new design, which keeps a 2+ second buffer for the audio. It seems though that your sound card wakes up roughly whenever a new sample can be rendered, which is 44100 times per second (or more, as from your profile it looks like ALSA resamples the audio). This causes the whole rendering code inside Swfdec to run for every sample, which is so much overhead that it kills all your performance. My question to you ALSA people now is: Who's responsible for ensuring that the number of wakeups is sane? Should I do some configuring of the audio device and if so, what kind of configuration, or should ALSA be able to pick reasonable defaults for me? Cheers, Benjamin On Sun, Dec 7, 2008 at 2:46 AM, Jesse Zbikowski <embeddedlinuxguy(a)gmail.com> wrote: > Hi Riccardo, > > On Fri, Dec 5, 2008 at 2:32 PM, Riccardo Magliocchetti > <riccardo.magliocchetti(a)gmail.com> wrote: >> Give it a run with oprofile, this is my opreport > > Thanks for this data. It looks like I may have an audio bottleneck. > opreport -t3 shows > > 31.5811 no-vmlinux > 14.4701 libswfdec-0.9.so.3.0.0 > 8.3210 libasound_module_rate_speexrate.so > 6.0088 libc-2.8.90.so > 4.7060 libpixman-1.so.0.12.0 > 3.6615 python2.5 > 3.2831 libcairo.so.2.10800.0 > > If I just look at the libraries I see the top libswfdec routines are > all audio related: > > 14.1497 libswfdec-0.9.so.3.0.0 swfdec_sound_buffer_get_n_samples > 2.1666 libswfdec-0.9.so.3.0.0 swfdec_audio_format_get_channels > 1.9369 libswfdec-0.9.so.3.0.0 swfdec_audio_format_new > 1.7904 libswfdec-0.9.so.3.0.0 swfdec_audio_flv_render > 1.3205 libswfdec-0.9.so.3.0.0 __i686.get_pc_thunk.bx > 1.2073 libswfdec-0.9.so.3.0.0 swfdec_audio_format_get_granularity > > I could try using the PulseAudio backend instead of ALSA. > >> You can try to improve your performance by using latest stable version of >> cairo and pixman > > OK, I am currently using the latest Ubuntu versions which are a bit > behind latest stable. I will see if this makes a difference. > > Package Ibex Latest > xorg-xserver 1.5.2 1.5.3 > cairo 1.8.0 1.8.4 > pixman 0.12 0.13.2 > >> play a bit with your X video driver config ie. exa vs xaa. > > I did not see libexa.so in my oprofile output although it is > installed, so I will make sure this is enabled. > > Thanks again, > Jesse Z. > _______________________________________________ > Swfdec mailing list > Swfdec(a)lists.freedesktop.org > http://lists.freedesktop.org/mailman/listinfo/swfdec >

1 0