May 2018 - Alsa-devel - mailman.alsa-project.org

[alsa-devel] [PATCH] ASoC: cs35l35: Add use_single_rw to regmap config
by Paul Handrigan 05 May '18

05 May '18

Add the use_single_rw flag to regmap config since the device does not support bulk transactions over i2c. Signed-off-by: Paul Handrigan <Paul.Handrigan(a)cirrus.com> --- sound/soc/codecs/cs35l35.c | 1 + 1 file changed, 1 insertion(+) diff --git a/sound/soc/codecs/cs35l35.c b/sound/soc/codecs/cs35l35.c index a4a2cb1..bd6226b 100644 --- a/sound/soc/codecs/cs35l35.c +++ b/sound/soc/codecs/cs35l35.c @@ -1105,6 +1105,7 @@ static struct regmap_config cs35l35_regmap = { .readable_reg = cs35l35_readable_register, .precious_reg = cs35l35_precious_register, .cache_type = REGCACHE_RBTREE, + .use_single_rw = true, }; static irqreturn_t cs35l35_irq(int irq, void *data) -- 2.7.4

2 1

[alsa-devel] [PATCH 1/4] ASoC: sgtl5000: Make the node name generic
by Fabio Estevam 05 May '18

05 May '18

From: Fabio Estevam <fabio.estevam(a)nxp.com> According to Devicetree Specification v0.2 document: "The name of a node should be somewhat generic, reflecting the function of the device and not its precise programming model." Do as suggested in the binding example. Signed-off-by: Fabio Estevam <fabio.estevam(a)nxp.com> --- Documentation/devicetree/bindings/sound/sgtl5000.txt | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/Documentation/devicetree/bindings/sound/sgtl5000.txt b/Documentation/devicetree/bindings/sound/sgtl5000.txt index 9a36c7e..0f21445 100644 --- a/Documentation/devicetree/bindings/sound/sgtl5000.txt +++ b/Documentation/devicetree/bindings/sound/sgtl5000.txt @@ -39,7 +39,7 @@ VDDIO 1.8V 2.5V 3.3V Example: -codec: sgtl5000@a { +sgtl5000: codec@a { compatible = "fsl,sgtl5000"; reg = <0x0a>; #sound-dai-cells = <0>; -- 2.7.4

3 11

[alsa-devel] [PATCH v3 0/7] ASoC: omap: Deprecate omap-pcm for sdma-pcm
by Peter Ujfalusi 05 May '18

05 May '18

Hi, Changes since v2: - Return -ENODEV from sdma_pcm_platform_register() in case !CONFIG_SND_SDMA_SOC Changes since v1: - Rebased on the n810 patches (Kconfig conflict) - Delete the omap-pcm as the last step instead of keeping the .c/.h file. If there is a need it can be resurrected with a simple revert. - I left out the Acked-by from Jarkko for the last patch since I have changed it omap-pcm served us well over the years, but there is no reason to keep it around as we can rely on generic dmaengine_pcm code without the need to maintain custom code. The series has been tested on: omap2 n810, omap3 beagle-xm, omap4 PandaES, omap4 Blaze (SDP), am57xx beagle-x15, dra7 evm. McBSP, McASP, McPDM, DMIC and HDMI audio was tested and I have found no issue with them. I was not able to test on OMAP1 for various reasons, but mainly because the Nokia 770 audio is not supported in mainline... Regards, Peter --- Peter Ujfalusi (7): ASoC: omap: Introduce the generic_dmaengine_pcm based sdma-pcm ASoC: davinci-mcasp: Convert to use the sdma-pcm instead of omap-pcm ASoC: omap-hdmi-audio: Convert to use the sdma-pcm instead of omap-pcm ASoC: omap-dmic: Convert to use the sdma-pcm instead of omap-pcm ASoC: omap-mcpdm: Convert to use the sdma-pcm instead of omap-pcm ASoC: omap-mcbsp: Convert to use the sdma-pcm instead of omap-pcm ASoC: omap: Delete the obsolete omap-pcm include/sound/omap-pcm.h | 30 ---- sound/soc/davinci/Kconfig | 2 +- sound/soc/davinci/davinci-mcasp.c | 8 +- sound/soc/omap/Kconfig | 33 ++-- sound/soc/omap/Makefile | 4 +- sound/soc/omap/omap-dmic.c | 4 +- sound/soc/omap/omap-hdmi-audio.c | 5 +- sound/soc/omap/omap-mcbsp.c | 4 +- sound/soc/omap/omap-mcpdm.c | 4 +- sound/soc/omap/omap-pcm.c | 262 ------------------------------ sound/soc/omap/sdma-pcm.c | 68 ++++++++ sound/soc/omap/sdma-pcm.h | 21 +++ 12 files changed, 124 insertions(+), 321 deletions(-) delete mode 100644 include/sound/omap-pcm.h delete mode 100644 sound/soc/omap/omap-pcm.c create mode 100644 sound/soc/omap/sdma-pcm.c create mode 100644 sound/soc/omap/sdma-pcm.h -- Peter Texas Instruments Finland Oy, Porkkalankatu 22, 00180 Helsinki. Y-tunnus/Business ID: 0615521-4. Kotipaikka/Domicile: Helsinki

3 9

Re: [alsa-devel] Buffer size calculation based on period size
by noman pouigt 04 May '18

04 May '18

>>noman pouigt wrote: > I am using tinycap to capture the audio and > setting below settings: > bitwidth = 16 > rate = 16000 > channel = 1 > period_size = 320 > period_count = 2 > > I am wondering how is below > period_size and buffer_size calculated? > > Shouldn't period_size of 320 bytes and > period_count of 2 fail? >From the tinyalsa source code: >>param_set_min(&params, SNDRV_PCM_HW_PARAM_PERIOD_SIZE, config->period_size); >>So it uses either the supplied value, or if that is too small, the smallest allowed value. Probably I was not clear; Why ALSA picked period_size of 2042 bytes when the minimum period bytes specified is 4084? Reason for this question is probably my wrong understanding. AFAIK: period_bytes_min specified in hw_params means the minimum bytes after which audio driver can expect an interrupt. So, total buffer size should be multiple of period_bytes_min or multiple of user provided period size if it is greater than period_bytes_min. So, if the above understanding is right then buffer size should have been 4084*2 (number of periods asked by user). Ultimately the question is how period size is calculated given min_period_bytes, max_period_bytes, buffer_bytes_max from pcm_hardware and user supplied period size and period count? I tried looking into the code but looks like there is lot of macro magic going on in pcm native layer. > cat sub0/hw_params > > access: RW_INTERLEAVED > format: S16_LE > subformat: STD > channels: 1 > rate: 16000 (16000/1) > period_size: 2042 > buffer_size: 4084 >>The actual values of period_size and period_count are returned by >>pcm_open().

2 1

[alsa-devel] [PATCH] ALSA: pcm: Hide local_irq_disable/enable() and local_irqsave/restore()
by Sebastian Andrzej Siewior 04 May '18

04 May '18

From: Anna-Maria Gleixner <anna-maria(a)linutronix.de> The snd_pcm_stream_lock_irq*() functions decouple disabling interrupts from the actual locking process. This does not work as expected if the locking primitives are replaced like on preempt-rt. Provide one function for locking which uses correct locking primitives. Signed-off-by: Anna-Maria Gleixner <anna-maria(a)linutronix.de> Signed-off-by: Sebastian Andrzej Siewior <bigeasy(a)linutronix.de> --- sound/core/pcm_native.c | 85 +++++++++++++++++++++++++++-------------- 1 file changed, 57 insertions(+), 28 deletions(-) diff --git a/sound/core/pcm_native.c b/sound/core/pcm_native.c index d18b3982548b..92c10fc838e9 100644 --- a/sound/core/pcm_native.c +++ b/sound/core/pcm_native.c @@ -99,6 +99,57 @@ static inline void down_write_nonblock(struct rw_semaphore *lock) cond_resched(); } +#define PCM_LOCK_DEFAULT 0 +#define PCM_LOCK_IRQ 1 +#define PCM_LOCK_IRQSAVE 2 + +static unsigned long __snd_pcm_stream_lock_mode(struct snd_pcm_substream *substream, + unsigned int mode) +{ + unsigned long flags = 0; + if (substream->pcm->nonatomic) { + down_read_nested(&snd_pcm_link_rwsem, SINGLE_DEPTH_NESTING); + mutex_lock(&substream->self_group.mutex); + } else { + switch (mode) { + case PCM_LOCK_DEFAULT: + read_lock(&snd_pcm_link_rwlock); + break; + case PCM_LOCK_IRQ: + read_lock_irq(&snd_pcm_link_rwlock); + break; + case PCM_LOCK_IRQSAVE: + read_lock_irqsave(&snd_pcm_link_rwlock, flags); + break; + } + spin_lock(&substream->self_group.lock); + } + return flags; +} + +static void __snd_pcm_stream_unlock_mode(struct snd_pcm_substream *substream, + unsigned int mode, unsigned long flags) +{ + if (substream->pcm->nonatomic) { + mutex_unlock(&substream->self_group.mutex); + up_read(&snd_pcm_link_rwsem); + } else { + spin_unlock(&substream->self_group.lock); + + switch (mode) { + case PCM_LOCK_DEFAULT: + read_unlock(&snd_pcm_link_rwlock); + break; + case PCM_LOCK_IRQ: + read_unlock_irq(&snd_pcm_link_rwlock); + break; + case PCM_LOCK_IRQSAVE: + read_unlock_irqrestore(&snd_pcm_link_rwlock, flags); + break; + } + } +} + /** * snd_pcm_stream_lock - Lock the PCM stream * @substream: PCM substream @@ -109,13 +160,7 @@ static inline void down_write_nonblock(struct rw_semaphore *lock) */ void snd_pcm_stream_lock(struct snd_pcm_substream *substream) { - if (substream->pcm->nonatomic) { - down_read_nested(&snd_pcm_link_rwsem, SINGLE_DEPTH_NESTING); - mutex_lock(&substream->self_group.mutex); - } else { - read_lock(&snd_pcm_link_rwlock); - spin_lock(&substream->self_group.lock); - } + __snd_pcm_stream_lock_mode(substream, PCM_LOCK_DEFAULT); } EXPORT_SYMBOL_GPL(snd_pcm_stream_lock); @@ -127,13 +172,7 @@ EXPORT_SYMBOL_GPL(snd_pcm_stream_lock); */ void snd_pcm_stream_unlock(struct snd_pcm_substream *substream) { - if (substream->pcm->nonatomic) { - mutex_unlock(&substream->self_group.mutex); - up_read(&snd_pcm_link_rwsem); - } else { - spin_unlock(&substream->self_group.lock); - read_unlock(&snd_pcm_link_rwlock); - } + __snd_pcm_stream_unlock_mode(substream, PCM_LOCK_DEFAULT, 0); } EXPORT_SYMBOL_GPL(snd_pcm_stream_unlock); @@ -147,9 +186,7 @@ EXPORT_SYMBOL_GPL(snd_pcm_stream_unlock); */ void snd_pcm_stream_lock_irq(struct snd_pcm_substream *substream) { - if (!substream->pcm->nonatomic) - local_irq_disable(); - snd_pcm_stream_lock(substream); + __snd_pcm_stream_lock_mode(substream, PCM_LOCK_IRQ); } EXPORT_SYMBOL_GPL(snd_pcm_stream_lock_irq); @@ -161,19 +198,13 @@ EXPORT_SYMBOL_GPL(snd_pcm_stream_lock_irq); */ void snd_pcm_stream_unlock_irq(struct snd_pcm_substream *substream) { - snd_pcm_stream_unlock(substream); - if (!substream->pcm->nonatomic) - local_irq_enable(); + __snd_pcm_stream_unlock_mode(substream, PCM_LOCK_IRQ, 0); } EXPORT_SYMBOL_GPL(snd_pcm_stream_unlock_irq); unsigned long _snd_pcm_stream_lock_irqsave(struct snd_pcm_substream *substream) { - unsigned long flags = 0; - if (!substream->pcm->nonatomic) - local_irq_save(flags); - snd_pcm_stream_lock(substream); - return flags; + return __snd_pcm_stream_lock_mode(substream, PCM_LOCK_IRQSAVE); } EXPORT_SYMBOL_GPL(_snd_pcm_stream_lock_irqsave); @@ -187,9 +218,7 @@ EXPORT_SYMBOL_GPL(_snd_pcm_stream_lock_irqsave); void snd_pcm_stream_unlock_irqrestore(struct snd_pcm_substream *substream, unsigned long flags) { - snd_pcm_stream_unlock(substream); - if (!substream->pcm->nonatomic) - local_irq_restore(flags); + __snd_pcm_stream_unlock_mode(substream, PCM_LOCK_IRQSAVE, flags); } EXPORT_SYMBOL_GPL(snd_pcm_stream_unlock_irqrestore); -- 2.17.0

2 1

[alsa-devel] Buffer size calculation based on period size
by noman pouigt 04 May '18

04 May '18

I am at my wits end trying to figure out how period_size is being calculated in pcm_native layer based on period_size and period_count being set by the userspace. I am using tinycap to capture the audio and setting below settings: bitwidth = 16 rate = 16000 channel = 1 period_size = 320 period_count = 2 I am wondering how is below period_size and buffer_size calculated? Shouldn't period_size of 320 bytes and period_count of 2 fail? cat sub0/hw_params access: RW_INTERLEAVED format: S16_LE subformat: STD channels: 1 rate: 16000 (16000/1) period_size: 2042 buffer_size: 4084 And this is my settings in pcm_hardware structure: static struct snd_pcm_hardware s1000_pcm_hw = { .buffer_bytes_max = 1024*1024, .period_bytes_min = 4084, .period_bytes_max = (1024*1024)/2, .periods_min = 2, .periods_max = 32, };

2 1

[alsa-devel] [PATCH] ALSA: dice: support transmission mode at 176.4/192.0 kHz for M-Audio 610/2626
by Takashi Sakamoto 04 May '18

04 May '18

At highest sampling transmission frequency, ALSA dice driver detects jumbo payload in tx packets from M-Audio 610/2626. $ dmesg snd_dice fw1.0: Detect jumbo payload: 0388 0348 Usual dice devices have a quirk called as dual-wire incompliant to IEC 61883-6 at highest sampling transmission frequency. On the other hand, M-Audio 610/2626 is compliant. This is a reason of the detection of jumbo payload. This commit adds support compliant transmission mode. For these devices, tx packets are correctly handled by this driver, while rx packets can't make these devices generate sound, unfortunately. The other factor affects it. Signed-off-by: Takashi Sakamoto <o-takashi(a)sakamocchi.jp> --- sound/firewire/dice/dice-stream.c | 6 +++++- sound/firewire/dice/dice.c | 9 +++++++++ sound/firewire/dice/dice.h | 2 ++ 3 files changed, 16 insertions(+), 1 deletion(-) diff --git a/sound/firewire/dice/dice-stream.c b/sound/firewire/dice/dice-stream.c index 422a1ab1ddf5..e3dbe357b4fa 100644 --- a/sound/firewire/dice/dice-stream.c +++ b/sound/firewire/dice/dice-stream.c @@ -204,7 +204,11 @@ static int keep_resources(struct snd_dice *dice, * For this quirk, blocking mode is required and PCM buffer size should * be aligned to SYT_INTERVAL. */ - double_pcm_frames = rate > 96000; + if (rate < 176400 || dice->no_dualwire) + double_pcm_frames = false; + else + double_pcm_frames = true; + if (double_pcm_frames) { rate /= 2; pcm_chs *= 2; diff --git a/sound/firewire/dice/dice.c b/sound/firewire/dice/dice.c index 40f7a32e4893..8baad7704bf0 100644 --- a/sound/firewire/dice/dice.c +++ b/sound/firewire/dice/dice.c @@ -236,6 +236,15 @@ static int dice_probe(struct fw_unit *unit, (snd_dice_detect_formats_t)entry->driver_data; } + /* + * M-Audio units are compliant to IEC 61883-6 at high sampling + * transmission frequency thus don't support 'dualwire' quirk. These + * units have 0x0100d1 in its version field in config ROM, against + * the value of DICE_INTERFACE. + */ + if (entry->vendor_id == OUI_MAUDIO) + dice->no_dualwire = true; + spin_lock_init(&dice->lock); mutex_init(&dice->mutex); init_completion(&dice->clock_accepted); diff --git a/sound/firewire/dice/dice.h b/sound/firewire/dice/dice.h index 505b79fea6d9..66e7a277ab68 100644 --- a/sound/firewire/dice/dice.h +++ b/sound/firewire/dice/dice.h @@ -113,6 +113,8 @@ struct snd_dice { bool global_enabled; struct completion clock_accepted; unsigned int substreams_counter; + + bool no_dualwire; }; enum snd_dice_addr_type { -- 2.14.1

1 0

[alsa-devel] [PATCH 1/9] R3Di and SBZ quirk entires + alt firmware loading
by Connor McAdams 04 May '18

04 May '18

This patch adds PCI quirk ID's for the Sound Blaster Z and Recon3Di. Only the currently tested ID's have been added. This patch also adds the ability to load alternative firmwares for each card, the firmwares can be obtained from within the Windows driver. The Recon3Di uses "ctefx-r3di.bin" and the Sound Blaster Z uses "ctefx-sbz.bin". If the alternative firmware for the given quirk is not found, the original ctefx.bin will be used. This has been confirmed to work for both the R3Di and the SBZ. This patch also makes the character array *dirstr a const. Signed-off-by: Connor McAdams <conmanx360(a)gmail.com> --- sound/pci/hda/patch_ca0132.c | 61 +++++++++++++++++++++++++++++++++++++++----- 1 file changed, 55 insertions(+), 6 deletions(-) diff --git a/sound/pci/hda/patch_ca0132.c b/sound/pci/hda/patch_ca0132.c index 768ea86..8346100 100644 --- a/sound/pci/hda/patch_ca0132.c +++ b/sound/pci/hda/patch_ca0132.c @@ -72,12 +72,16 @@ #define SCP_GET 1 #define EFX_FILE "ctefx.bin" +#define SBZ_EFX_FILE "ctefx-sbz.bin" +#define R3DI_EFX_FILE "ctefx-r3di.bin" #ifdef CONFIG_SND_HDA_CODEC_CA0132_DSP MODULE_FIRMWARE(EFX_FILE); +MODULE_FIRMWARE(SBZ_EFX_FILE); +MODULE_FIRMWARE(R3DI_EFX_FILE); #endif -static char *dirstr[2] = { "Playback", "Capture" }; +static const char *dirstr[2] = { "Playback", "Capture" }; enum { SPEAKER_OUT, @@ -734,6 +738,7 @@ struct ca0132_spec { unsigned int scp_resp_header; unsigned int scp_resp_data[4]; unsigned int scp_resp_count; + bool alt_firmware_present; /* mixer and effects related */ unsigned char dmic_ctl; @@ -762,6 +767,8 @@ struct ca0132_spec { enum { QUIRK_NONE, QUIRK_ALIENWARE, + QUIRK_SBZ, + QUIRK_R3DI, }; static const struct hda_pintbl alienware_pincfgs[] = { @@ -782,6 +789,10 @@ static const struct snd_pci_quirk ca0132_quirks[] = { SND_PCI_QUIRK(0x1028, 0x0685, "Alienware 15 2015", QUIRK_ALIENWARE), SND_PCI_QUIRK(0x1028, 0x0688, "Alienware 17 2015", QUIRK_ALIENWARE), SND_PCI_QUIRK(0x1028, 0x0708, "Alienware 15 R2 2016", QUIRK_ALIENWARE), + SND_PCI_QUIRK(0x1102, 0x0010, "Sound Blaster Z", QUIRK_SBZ), + SND_PCI_QUIRK(0x1102, 0x0023, "Sound Blaster Z", QUIRK_SBZ), + SND_PCI_QUIRK(0x1458, 0xA016, "Recon3Di", QUIRK_R3DI), + SND_PCI_QUIRK(0x1458, 0xA036, "Recon3Di", QUIRK_R3DI), {} }; @@ -3207,7 +3218,7 @@ static int ca0132_select_out(struct hda_codec *codec) pin_ctl & ~PIN_HP); /* enable speaker node */ pin_ctl = snd_hda_codec_read(codec, spec->out_pins[0], 0, - AC_VERB_GET_PIN_WIDGET_CONTROL, 0); + AC_VERB_GET_PIN_WIDGET_CONTROL, 0); snd_hda_set_pin_ctl(codec, spec->out_pins[0], pin_ctl | PIN_OUT); } else { @@ -4370,11 +4381,49 @@ static void ca0132_set_dsp_msr(struct hda_codec *codec, bool is96k) static bool ca0132_download_dsp_images(struct hda_codec *codec) { bool dsp_loaded = false; + struct ca0132_spec *spec = codec->spec; const struct dsp_image_seg *dsp_os_image; const struct firmware *fw_entry; - - if (request_firmware(&fw_entry, EFX_FILE, codec->card->dev) != 0) - return false; + /* + * Alternate firmwares for different variants. The Recon3Di apparently + * can use the default firmware, but I'll leave the option in case + * it needs it again. + */ + switch (spec->quirk) { + case QUIRK_SBZ: + if (request_firmware(&fw_entry, SBZ_EFX_FILE, + codec->card->dev) != 0) { + codec_dbg(codec, "SBZ alt firmware not detected. "); + spec->alt_firmware_present = false; + } else { + codec_dbg(codec, "Sound Blaster Z firmware selected."); + spec->alt_firmware_present = true; + } + break; + case QUIRK_R3DI: + if (request_firmware(&fw_entry, R3DI_EFX_FILE, + codec->card->dev) != 0) { + codec_dbg(codec, "Recon3Di alt firmware not detected."); + spec->alt_firmware_present = false; + } else { + codec_dbg(codec, "Recon3Di firmware selected."); + spec->alt_firmware_present = true; + } + break; + default: + spec->alt_firmware_present = false; + break; + } + /* + * Use default ctefx.bin if no alt firmware is detected, or if none + * exists for your particular codec. + */ + if (!spec->alt_firmware_present) { + codec_dbg(codec, "Default firmware selected."); + if (request_firmware(&fw_entry, EFX_FILE, + codec->card->dev) != 0) + return false; + } dsp_os_image = (struct dsp_image_seg *)(fw_entry->data); if (dspload_image(codec, dsp_os_image, 0, 0, true, 0)) { @@ -4476,7 +4525,7 @@ static struct hda_verb ca0132_base_exit_verbs[] = { {} }; -/* Other verbs tables. Sends after DSP download. */ +/* Other verbs tables. Sends after DSP download. */ static struct hda_verb ca0132_init_verbs0[] = { /* chip init verbs */ {0x15, 0x70D, 0xF0}, -- 2.7.4

3 11

[alsa-devel] [PATCH 1/9] R3Di and SBZ quirk entires + alt firmware loading
by Connor McAdams 04 May '18

04 May '18

This patch adds PCI quirk ID's for the Sound Blaster Z and Recon3Di. Only the currently tested ID's have been added. This patch also adds the ability to load alternative firmwares for each card, the firmwares can be obtained from within the Windows driver. The Recon3Di uses "ctefx-r3di.bin" and the Sound Blaster Z uses "ctefx-sbz.bin". If the alternative firmware for the given quirk is not found, the original ctefx.bin will be used. This has been confirmed to work for both the R3Di and the SBZ. This patch also makes the character array *dirstr a const. Signed-off-by: Connor McAdams <conmanx360(a)gmail.com> --- sound/pci/hda/patch_ca0132.c | 55 ++++++++++++++++++++++++++++++++++++++++---- 1 file changed, 51 insertions(+), 4 deletions(-) diff --git a/sound/pci/hda/patch_ca0132.c b/sound/pci/hda/patch_ca0132.c index 768ea86..9a0286e 100644 --- a/sound/pci/hda/patch_ca0132.c +++ b/sound/pci/hda/patch_ca0132.c @@ -72,12 +72,16 @@ #define SCP_GET 1 #define EFX_FILE "ctefx.bin" +#define SBZ_EFX_FILE "ctefx-sbz.bin" +#define R3DI_EFX_FILE "ctefx-r3di.bin" #ifdef CONFIG_SND_HDA_CODEC_CA0132_DSP MODULE_FIRMWARE(EFX_FILE); +MODULE_FIRMWARE(SBZ_EFX_FILE); +MODULE_FIRMWARE(R3DI_EFX_FILE); #endif -static char *dirstr[2] = { "Playback", "Capture" }; +static const char *dirstr[2] = { "Playback", "Capture" }; enum { SPEAKER_OUT, @@ -734,6 +738,7 @@ struct ca0132_spec { unsigned int scp_resp_header; unsigned int scp_resp_data[4]; unsigned int scp_resp_count; + bool alt_firmware_present; /* mixer and effects related */ unsigned char dmic_ctl; @@ -762,6 +767,8 @@ struct ca0132_spec { enum { QUIRK_NONE, QUIRK_ALIENWARE, + QUIRK_SBZ, + QUIRK_R3DI, }; static const struct hda_pintbl alienware_pincfgs[] = { @@ -782,6 +789,10 @@ static const struct snd_pci_quirk ca0132_quirks[] = { SND_PCI_QUIRK(0x1028, 0x0685, "Alienware 15 2015", QUIRK_ALIENWARE), SND_PCI_QUIRK(0x1028, 0x0688, "Alienware 17 2015", QUIRK_ALIENWARE), SND_PCI_QUIRK(0x1028, 0x0708, "Alienware 15 R2 2016", QUIRK_ALIENWARE), + SND_PCI_QUIRK(0x1102, 0x0010, "Sound Blaster Z", QUIRK_SBZ), + SND_PCI_QUIRK(0x1102, 0x0023, "Sound Blaster Z", QUIRK_SBZ), + SND_PCI_QUIRK(0x1458, 0xA016, "Recon3Di", QUIRK_R3DI), + SND_PCI_QUIRK(0x1458, 0xA036, "Recon3Di", QUIRK_R3DI), {} }; @@ -4370,11 +4381,47 @@ static void ca0132_set_dsp_msr(struct hda_codec *codec, bool is96k) static bool ca0132_download_dsp_images(struct hda_codec *codec) { bool dsp_loaded = false; + struct ca0132_spec *spec = codec->spec; const struct dsp_image_seg *dsp_os_image; const struct firmware *fw_entry; - - if (request_firmware(&fw_entry, EFX_FILE, codec->card->dev) != 0) - return false; + /* + * Alternate firmwares for different variants. The Recon3Di apparently + * can use the default firmware, but I'll leave the option in case + * it needs it again. + */ + switch (spec->quirk) { + case QUIRK_SBZ: + if (request_firmware(&fw_entry, SBZ_EFX_FILE, + codec->card->dev) != 0) { + codec_dbg(codec, "SBZ alt firmware not detected. "); + spec->alt_firmware_present = false; + } else { + codec_dbg(codec, "Sound Blaster Z firmware selected."); + spec->alt_firmware_present = true; + } + break; + case QUIRK_R3DI: + if (request_firmware(&fw_entry, R3DI_EFX_FILE, + codec->card->dev) != 0) { + codec_dbg(codec, "Recon3Di alt firmware not detected."); + spec->alt_firmware_present = false; + } else { + codec_dbg(codec, "Recon3Di firmware selected."); + spec->alt_firmware_present = true; + } + break; + default: + spec->alt_firmware_present = false; + break; + } + /* Use default ctefx.bin if no alt firmware is detected, or if none + * exists for your particular codec. + */ + if (spec->alt_firmware_present == false) { + codec_dbg(codec, "Default firmware selected."); + if (request_firmware(&fw_entry, EFX_FILE, + codec->card->dev) != 0) + return false; + } dsp_os_image = (struct dsp_image_seg *)(fw_entry->data); if (dspload_image(codec, dsp_os_image, 0, 0, true, 0)) { -- 2.7.4

2 9

[alsa-devel] Segmentation fault in libasound.so.2.0.0 (alsa-lib-1.1.6)
by Paul Bordukoff 04 May '18

04 May '18

Hello everyone, I am using ALSA (alsa-lib-1.1.6) on Gentoo laptop with x32 ABI (linux-4.16.7). My mpv (0.27.2) gets segfault each time I try to play video with sound: pashgan@light ~/temp $ mpv ./SampleVideo_1280x720_10mb.mp4 Playing: ./SampleVideo_1280x720_10mb.mp4 (+) Video --vid=1 (*) (h264 1280x720 25.000fps) (+) Audio --aid=1 --alang=und (*) (aac 6ch 48000Hz) AO: [alsa] 48000Hz stereo 2ch float VO: [opengl] 1280x720 yuv420p Segmentation fault dmesg shows: [12526.319909] traps: mpv/ao[13621] general protection ip:f28ce4fa sp:e46f2d60 error:0 in libasound.so.2.0.0[f2843000+fe000] gdb session: (gdb) bt #0 0xf29b14fa in mix_areas_32_smp (size=3935624896, dst=0x56a6e7f0, src=0x56a6e7f0, sum=0x1, dst_step=4, src_step=4, sum_step=4) at /tmp/portage/media-libs/alsa-lib-1.1.6-r1/work/alsa-lib-1.1.6/src/pcm/pcm_dmix_x86_64.h:143 #1 0xf29b2361 in mix_areas (size=3935493824, dst_ofs=4056, src_ofs=0, dst_areas=0x56a6cd90, src_areas=0x56a6dc30, dmix=0x56a6c900) at /tmp/portage/media-libs/alsa-lib-1.1.6-r1/work/alsa-lib-1.1.6/src/pcm/pcm_dmix.c:197 #2 snd_pcm_dmix_sync_area (pcm=pcm@entry=0x56a6d270) at /tmp/portage/media-libs/alsa-lib-1.1.6-r1/work/alsa-lib-1.1.6/src/pcm/pcm_dmix.c:379 #3 0xf29b267f in snd_pcm_dmix_start (pcm=<optimized out>) at /tmp/portage/media-libs/alsa-lib-1.1.6-r1/work/alsa-lib-1.1.6/src/pcm/pcm_dmix.c:611 #4 0xf2972349 in __snd_pcm_start (pcm=0x56a6d270, pcm=0x56a6d270) at /tmp/portage/media-libs/alsa-lib-1.1.6-r1/work/alsa-lib-1.1.6/src/pcm/pcm_local.h:434 #5 snd_pcm_start (pcm=<optimized out>) at /tmp/portage/media-libs/alsa-lib-1.1.6-r1/work/alsa-lib-1.1.6/src/pcm/pcm.c:1226 #6 0xf2972349 in __snd_pcm_start (pcm=0x56a6d8b0, pcm=0x56a6d8b0) at /tmp/portage/media-libs/alsa-lib-1.1.6-r1/work/alsa-lib-1.1.6/src/pcm/pcm_local.h:434 #7 snd_pcm_start (pcm=<optimized out>) at /tmp/portage/media-libs/alsa-lib-1.1.6-r1/work/alsa-lib-1.1.6/src/pcm/pcm.c:1226 #8 0xf2979a78 in __snd_pcm_start (pcm=0x56a52e60, pcm=0x56a52e60) at /tmp/portage/media-libs/alsa-lib-1.1.6-r1/work/alsa-lib-1.1.6/src/pcm/pcm_local.h:434 #9 snd1_pcm_write_areas (pcm=pcm@entry=0x56a52e60, areas=areas@entry=0xe47f2f00, offset=offset@entry=0, size=<optimized out>, size@entry=12288, func=func@entry=0xf2988770 <snd_pcm_plugin_write_areas>) at /tmp/portage/media-libs/alsa-lib-1.1.6-r1/work/alsa-lib-1.1.6/src/pcm/pcm.c:7421 #10 0xf29890e7 in snd_pcm_plugin_writei (pcm=<optimized out>, buffer=<optimized out>, size=12288) at /tmp/portage/media-libs/alsa-lib-1.1.6-r1/work/alsa-lib-1.1.6/src/pcm/pcm_plugin.c:358 #11 0x56590280 in ?? () #12 0x56594846 in ?? () #13 0xf1b9808b in start_thread () from /libx32/libpthread.so.0 #14 0xf18d4e2f in clone () from /libx32/libc.so.6 (gdb) Same bug in Gentoo bug tracker: https://bugs.gentoo.org/654754 I need help to investigate pcm_dmix behaviour, I see asm code there (pcm_dmix_x86_64.h:143) but have no idea how can I debug it (my asm experience is far from ok). Can somebody guide me?

1 0