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

[alsa-devel] [PATCH] fix incorrect rounding direction in snd_interval_ratnum()
by Krzysztof Helt 21 Dec '09

21 Dec '09

From: Krzysztof Helt <krzysztof.h1(a)wp.pl> The direction of rounding is incorrect in the snd_interval_ratnum() It was detected with following parameters (sb8 driver playing 8kHz stereo file): - num is always 1000000 - requested frequency rate is from 7999 to 7999 (single frequency) The first loop calculates div_down(num, freq->min) which is 125. Thus, a frequency range's minimum value is 1000000 / 125 = 8000 Hz. The second loop calculates div_up(num, freq->max) which is 126 The frequency range's maximum value is 1000000 / 126 = 7936 Hz. The range maximum is lower than the range minimum so the function fails due to empty result range. Signed-off-by: Krzysztof Helt <krzysztof.h1(a)wp.pl> --- sound/core/pcm_lib.c | 4 ++-- 1 files changed, 2 insertions(+), 2 deletions(-) diff --git a/sound/core/pcm_lib.c b/sound/core/pcm_lib.c index 30f4108..a27545b 100644 --- a/sound/core/pcm_lib.c +++ b/sound/core/pcm_lib.c @@ -758,7 +758,7 @@ int snd_interval_ratnum(struct snd_interval *i, int diff; if (q == 0) q = 1; - den = div_down(num, q); + den = div_up(num, q); if (den < rats[k].den_min) continue; if (den > rats[k].den_max) @@ -794,7 +794,7 @@ int snd_interval_ratnum(struct snd_interval *i, i->empty = 1; return -EINVAL; } - den = div_up(num, q); + den = div_down(num, q); if (den > rats[k].den_max) continue; if (den < rats[k].den_min) -- 1.6.4 ---------------------------------------------------------------------- Co Ci sie dzisiaj snilo? Czy wiesz, co oznacza Twoj sen? Sprawdz w senniku >> http://link.interia.pl/f2532

2 1

[alsa-devel] [PATCH] HDA: add powersaving hook for Realtek
by Hector Martin 21 Dec '09

21 Dec '09

The current Realtek code makes no specific provision for turning stuff off. The codec chip is placed into low-power mode generically, but this doesn't turn off any external hardware connected to it, in particular external amplifiers. This patch creates a hook function that is called by the codec suspend/resume functions. It ought to disable any external hardware in a device-specific way. I've implemented a generic ALC889 function that sets the EAPD pin properly, and used it for the Acer Aspire 8930G which can benefit from this feature. On my laptop, this results in ~0.5W extra savings. -- Hector Martin (hector(a)marcansoft.com) Public Key: http://www.marcansoft.com/marcan.asc Signed-off-by: Hector Martin <hector(a)marcansoft.com> --- linux-2.6.32/sound/pci/hda/patch_realtek_orig.c 2009-12-20 19:31:08.940900751 +0100 +++ linux-2.6.32/sound/pci/hda/patch_realtek.c 2009-12-20 22:14:11.852025021 +0100 @@ -334,6 +334,9 @@ /* hooks */ void (*init_hook)(struct hda_codec *codec); void (*unsol_event)(struct hda_codec *codec, unsigned int res); +#ifdef CONFIG_SND_HDA_POWER_SAVE + void (*power_hook)(struct hda_codec *codec, int power); +#endif /* for pin sensing */ unsigned int sense_updated: 1; @@ -385,6 +388,7 @@ void (*init_hook)(struct hda_codec *); #ifdef CONFIG_SND_HDA_POWER_SAVE struct hda_amp_list *loopbacks; + void (*power_hook)(struct hda_codec *codec, int power); #endif }; @@ -897,6 +901,7 @@ spec->unsol_event = preset->unsol_event; spec->init_hook = preset->init_hook; #ifdef CONFIG_SND_HDA_POWER_SAVE + spec->power_hook = preset->power_hook; spec->loopback.amplist = preset->loopbacks; #endif @@ -1824,6 +1845,16 @@ spec->autocfg.speaker_pins[2] = 0x1b; } +#ifdef CONFIG_SND_HDA_POWER_SAVE +static void alc889_power_eapd(struct hda_codec *codec, int power) +{ + snd_hda_codec_write(codec, 0x14, 0, + AC_VERB_SET_EAPD_BTLENABLE, power ? 2 : 0); + snd_hda_codec_write(codec, 0x15, 0, + AC_VERB_SET_EAPD_BTLENABLE, power ? 2 : 0); +} +#endif + /* * ALC880 3-stack model * @@ -3611,12 +3642,25 @@ snd_hda_detach_beep_device(codec); } +#ifdef CONFIG_SND_HDA_POWER_SAVE +static int alc_suspend(struct hda_codec *codec, pm_message_t state) +{ + struct alc_spec *spec = codec->spec; + if (spec && spec->power_hook) + spec->power_hook(codec, 0); + return 0; +} +#endif + #ifdef SND_HDA_NEEDS_RESUME static int alc_resume(struct hda_codec *codec) { + struct alc_spec *spec = codec->spec; codec->patch_ops.init(codec); snd_hda_codec_resume_amp(codec); snd_hda_codec_resume_cache(codec); + if (spec && spec->power_hook) + spec->power_hook(codec, 1); return 0; } #endif @@ -3633,6 +3677,7 @@ .resume = alc_resume, #endif #ifdef CONFIG_SND_HDA_POWER_SAVE + .suspend = alc_suspend, .check_power_status = alc_check_power_status, #endif }; @@ -9320,6 +9365,9 @@ .unsol_event = alc_automute_amp_unsol_event, .setup = alc889_acer_aspire_8930g_setup, .init_hook = alc_automute_amp, +#ifdef CONFIG_SND_HDA_POWER_SAVE + .power_hook = alc889_power_eapd, +#endif }, [ALC888_ACER_ASPIRE_7730G] = { .mixers = { alc883_3ST_6ch_mixer,

2 2

[alsa-devel] [PATCH] HDA: remove useless mixers on Aspire 8930G
by Hector Martin 21 Dec '09

21 Dec '09

This patch removes some extra mixers that do nothing on the Acer Aspire 8930G. The CD mixer is useless because the SATA DVD/Blu-Ray drive has no analog audio output, and the Side mixer is useless because we max out at 6ch anyway. -- Hector Martin (hector(a)marcansoft.com) Public Key: http://www.marcansoft.com/marcan.asc Signed-off-by: Hector Martin <hector(a)marcansoft.com> --- linux-2.6.32/sound/pci/hda/patch_realtek_orig.c 2009-12-20 19:31:08.940900751 +0100 +++ linux-2.6.32/sound/pci/hda/patch_realtek.c 2009-12-20 22:14:11.852025021 +0100 @@ -1780,6 +1782,25 @@ { } /* end */ }; +static struct snd_kcontrol_new alc889_acer_aspire_8930g_mixer[] = { + HDA_CODEC_VOLUME("Front Playback Volume", 0x0c, 0x0, HDA_OUTPUT), + HDA_BIND_MUTE("Front Playback Switch", 0x0c, 2, HDA_INPUT), + HDA_CODEC_VOLUME("Surround Playback Volume", 0x0d, 0x0, HDA_OUTPUT), + HDA_BIND_MUTE("Surround Playback Switch", 0x0d, 2, HDA_INPUT), + HDA_CODEC_VOLUME_MONO("Center Playback Volume", 0x0e, 1, 0x0, + HDA_OUTPUT), + HDA_CODEC_VOLUME_MONO("LFE Playback Volume", 0x0e, 2, 0x0, HDA_OUTPUT), + HDA_BIND_MUTE_MONO("Center Playback Switch", 0x0e, 1, 2, HDA_INPUT), + HDA_BIND_MUTE_MONO("LFE Playback Switch", 0x0e, 2, 2, HDA_INPUT), + HDA_CODEC_VOLUME("Line Playback Volume", 0x0b, 0x02, HDA_INPUT), + HDA_CODEC_MUTE("Line Playback Switch", 0x0b, 0x02, HDA_INPUT), + HDA_CODEC_VOLUME("Mic Playback Volume", 0x0b, 0x0, HDA_INPUT), + HDA_CODEC_VOLUME("Mic Boost", 0x18, 0, HDA_INPUT), + HDA_CODEC_MUTE("Mic Playback Switch", 0x0b, 0x0, HDA_INPUT), + { } /* end */ +}; + + static void alc888_acer_aspire_4930g_setup(struct hda_codec *codec) { struct alc_spec *spec = codec->spec; @@ -9283,7 +9283,7 @@ .init_hook = alc_automute_amp, }, [ALC888_ACER_ASPIRE_8930G] = { - .mixers = { alc888_base_mixer, + .mixers = { alc889_acer_aspire_8930g_mixer, alc883_chmode_mixer }, .init_verbs = { alc883_init_verbs, alc880_gpio1_init_verbs, alc889_acer_aspire_8930g_verbs },

2 1

[alsa-devel] [PATCH] HDA: simplify Aspire 8930G verb array
by Hector Martin 21 Dec '09

21 Dec '09

This patch just simplifies the 8930G verb array a bit. Just use the common ALC889 EAPD verb array to make things more consistent. The file is already huge enough already. -- Hector Martin (hector(a)marcansoft.com) Public Key: http://www.marcansoft.com/marcan.asc Signed-off-by: Hector Martin <hector(a)marcansoft.com> --- linux-2.6.32/sound/pci/hda/patch_realtek_orig.c 2009-12-20 19:31:08.940900751 +0100 +++ linux-2.6.32/sound/pci/hda/patch_realtek.c 2009-12-20 22:14:11.852025021 +0100 @@ -1665,9 +1670,6 @@ /* some bit here disables the other DACs. Init=0x4900 */ {0x20, AC_VERB_SET_COEF_INDEX, 0x08}, {0x20, AC_VERB_SET_PROC_COEF, 0x0000}, -/* Enable amplifiers */ - {0x14, AC_VERB_SET_EAPD_BTLENABLE, 0x02}, - {0x15, AC_VERB_SET_EAPD_BTLENABLE, 0x02}, /* DMIC fix * This laptop has a stereo digital microphone. The mics are only 1cm apart * which makes the stereo useless. However, either the mic or the ALC889 @@ -9305,7 +9305,8 @@ .mixers = { alc889_acer_aspire_8930g_mixer, alc883_chmode_mixer }, .init_verbs = { alc883_init_verbs, alc880_gpio1_init_verbs, - alc889_acer_aspire_8930g_verbs }, + alc889_acer_aspire_8930g_verbs, + alc889_eapd_verbs}, .num_dacs = ARRAY_SIZE(alc883_dac_nids), .dac_nids = alc883_dac_nids, .num_adc_nids = ARRAY_SIZE(alc889_adc_nids),

2 1

Re: [alsa-devel] [2.6.33-rc1] [SOUND] HDA: high cpu usage, noise
by Takashi Iwai 21 Dec '09

21 Dec '09

At Sun, 20 Dec 2009 18:27:45 +0100, Maciej Rutecki wrote: > > Hi > > Affected kernel: 2.6.33-rc1 > Last known good: 2.6.32 The only fundamental change, AFAIK, is the use of MSI as default. Could you try to add enable_msi=0 option to snd-hda-intel? thanks, Takashi

1 0

[alsa-devel] [PATCH] ALSA: hda: Set Front Mic to input vref 50% for Lenovo 3000 Y410
by Daniel Chen 21 Dec '09

21 Dec '09

Hi, Attached is a patch against master HEAD setting input vref 50% for the ALC262_LENOVO_3000 model. Thanks, Dan

2 1

[alsa-devel] [PATCH] ALSA: pcm: Properly close timer when freeing slaves
by Daniel Chen 21 Dec '09

21 Dec '09

Hi, Here's a patch originally from Dan McCombs against alsa-lib.git HEAD fixing a fairly nasty fd leak for timers. Thanks, Dan

2 1

[alsa-devel] No device for DAI omap-eac-dai error
by Belisko Marek 21 Dec '09

21 Dec '09

Hi, I'm writing new omap dai (for EAC codec base omap's). My code is based on existing omap-mcbsp.c code basically. After loadingof module I got error which appear in init function when call snd_soc_register_dai(). I browse for code but can't figure out where dai->dev should be set to don't get error message and init dai properly. Thanks for any suggestions Marek

3 2

[alsa-devel] Installing a "different" version of Alsa
by Brian Welch 20 Dec '09

20 Dec '09

Hello, It's been a year since I posted to this list, but I'm running into a problem and can't figure out the best way to solve it, so I'm hoping someone here can help. I've got an SDK for an embedded platform that includes a specific version of Linux with ALSA built-in. On my platform, I want to use a different (newer) version of ALSA than the built-in version, so I want to build and load the binaries from the newer version and, more importantly, I want to compile my custom sound hardware driver against the newer alsa version and the SDK's Linux version. I don't really want to touch the kernel files that are in the kerenl/include/sound directory in the SDK. When I use kbuild to build my custom sound hardware driver against the SDK in the kernel, it always trys to pick up the older ALSA header files that are in the SDK, but I want to use the newer versions that are in the directory with the newer version of ALSA. What the best way to make sure I get the correct headers? Is what I'm trying "dangerous" and should I be going about this a different way? Thanks for any advice, Brian

2 1

[alsa-devel] ASoC updates for 2.6.33
by Mark Brown 19 Dec '09

19 Dec '09

The following changes since commit 1cf86f6f9b000e98c1b7f866f99633ae67464e2f: Kuninori Morimoto (1): ASoC: ak4642: Add default return value in ak4642_modinit 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 Guennadi Liakhovetski (1): ASoC: wm8974: fix a wrong bit definition Jon Smirl (1): ASoC: Fix disable of SPDIF on STAC9766 codec sound/soc/codecs/stac9766.c | 18 +----------------- sound/soc/codecs/wm8974.c | 2 +- 2 files changed, 2 insertions(+), 18 deletions(-)

2 3