December 2017 - Alsa-devel - mailman.alsa-project.org

[alsa-devel] [PATCH] ASoC: rsnd: more clear rsnd_get_dalign() for DALIGN
by Kuninori Morimoto 08 Dec '17

08 Dec '17

From: Kuninori Morimoto <kuninori.morimoto.gx(a)renesas.com> On Renesas sound device, DALIGN which exchanges channel position is needed because SW and HW are using defferent data order if 16bit data. It is not needed when 24bit data. rsnd_get_dalign() returns necessary value, but it was confusable code. This patch makes it more simple. Tested-by: Hiroyuki Yokoyama <hiroyuki.yokoyama.vx(a)renesas.com> Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx(a)renesas.com> --- sound/soc/sh/rcar/core.c | 33 ++++++++++----------------------- 1 file changed, 10 insertions(+), 23 deletions(-) diff --git a/sound/soc/sh/rcar/core.c b/sound/soc/sh/rcar/core.c index ab2cd53..e438b801 100644 --- a/sound/soc/sh/rcar/core.c +++ b/sound/soc/sh/rcar/core.c @@ -294,11 +294,12 @@ u32 rsnd_get_dalign(struct rsnd_mod *mod, struct rsnd_dai_stream *io) struct rsnd_mod *ssiu = rsnd_io_to_mod_ssiu(io); struct rsnd_mod *target; struct snd_pcm_runtime *runtime = rsnd_io_to_runtime(io); - u32 val = 0x76543210; - u32 mask = ~0; /* - * *Hardware* L/R and *Software* L/R are inverted. + * *Hardware* L/R and *Software* L/R are inverted for 16bit data. + * 31..16 15...0 + * HW: [L ch] [R ch] + * SW: [R ch] [L ch] * We need to care about inversion timing to control * Playback/Capture correctly. * The point is [DVC] needs *Hardware* L/R, [MEM] needs *Software* L/R @@ -325,27 +326,13 @@ u32 rsnd_get_dalign(struct rsnd_mod *mod, struct rsnd_dai_stream *io) target = cmd ? cmd : ssiu; } - mask <<= runtime->channels * 4; - val = val & mask; - - switch (runtime->sample_bits) { - case 16: - val |= 0x67452301 & ~mask; - break; - case 32: - val |= 0x76543210 & ~mask; - break; - } - - /* - * exchange channeles on SRC if possible, - * otherwise, R/L volume settings on DVC - * changes inverted channels - */ - if (mod == target) - return val; - else + /* Non target mod or 24bit data needs normal DALIGN */ + if ((runtime->sample_bits != 16) || + (mod != target)) return 0x76543210; + /* Target mod needs inverted DALIGN when 16bit */ + else + return 0x67452301; } u32 rsnd_get_busif_shift(struct rsnd_dai_stream *io, struct rsnd_mod *mod) -- 1.9.1

2 1

[alsa-devel] [PATCH] ASoC: atmel-classd: select correct Kconfig symbol
by Alexandre Belloni 08 Dec '17

08 Dec '17

SND_ATMEL_SOC_CLASSD selects SND_ATMEL_SOC_DMA but the driver itself handles its own DMA operations and doesn't need anything from atmel-pcm-dma.c or atmel_ssc_dai.c. Replace SND_ATMEL_SOC_DMA by SND_SOC_GENERIC_DMAENGINE_PCM which is the only one actually required. This may end up in a configuration leading to a link error: sound/soc/atmel/atmel_ssc_dai.o: In function `atmel_ssc_set_audio': atmel_ssc_dai.c:(.text+0x79c): undefined reference to `atmel_pcm_dma_platform_register' atmel_ssc_dai.c:(.text+0x79c): relocation truncated to fit: R_AARCH64_CALL26 against undefined symbol `atmel_pcm_dma_platform_register' sound/soc/atmel/atmel_ssc_dai.o: In function `atmel_ssc_put_audio': atmel_ssc_dai.c:(.text+0xf24): undefined reference to `atmel_pcm_dma_platform_unregister' atmel_ssc_dai.c:(.text+0xf24): relocation truncated to fit: R_AARCH64_CALL26 against undefined symbol `atmel_pcm_dma_platform_unregister' Tested on sama5d2 xplained with the following configuration where nothing selects SND_ATMEL_SOC_DMA: CONFIG_SND_ATMEL_SOC=y CONFIG_SND_ATMEL_SOC_CLASSD=y Reported-by: Arnd Bergmann <arnd(a)arndb.de> Tested-by: Arnd Bergmann <arnd(a)arndb.de> Fixes: e0a25b6d1862 ("ASoC: atmel-classd: add the Audio Class D Amplifier") Signed-off-by: Alexandre Belloni <alexandre.belloni(a)free-electrons.com> --- sound/soc/atmel/Kconfig | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/sound/soc/atmel/Kconfig b/sound/soc/atmel/Kconfig index 4a56f3dfba51..dcee145dd179 100644 --- a/sound/soc/atmel/Kconfig +++ b/sound/soc/atmel/Kconfig @@ -64,7 +64,7 @@ config SND_AT91_SOC_SAM9X5_WM8731 config SND_ATMEL_SOC_CLASSD tristate "Atmel ASoC driver for boards using CLASSD" depends on ARCH_AT91 || COMPILE_TEST - select SND_ATMEL_SOC_DMA + select SND_SOC_GENERIC_DMAENGINE_PCM select REGMAP_MMIO help Say Y if you want to add support for Atmel ASoC driver for boards using -- 2.15.1

2 1

[alsa-devel] [PATCH] ASoC: Change kern log level to avoid unwanted messages
by Naveen Manohar 08 Dec '17

08 Dec '17

patch suppresses the warning message "control load not supported" as this is a debug information to help debug issues in topology. Signed-off-by: Naveen Manohar <naveen.m(a)intel.com> diff --git a/sound/soc/intel/skylake/skl-topology.c b/sound/soc/intel/skylake/skl-topology.c index a072bcf..81923da 100644 --- a/sound/soc/intel/skylake/skl-topology.c +++ b/sound/soc/intel/skylake/skl-topology.c @@ -2908,7 +2908,7 @@ static int skl_tplg_control_load(struct snd_soc_component *cmpnt, break; default: - dev_warn(bus->dev, "Control load not supported %d:%d:%d\n", + dev_dbg(bus->dev, "Control load not supported %d:%d:%d\n", hdr->ops.get, hdr->ops.put, hdr->ops.info); break; } -- 1.9.1

3 2

[alsa-devel] Cherry Trail + RT5645 devices with a mono speaker ?
by Hans de Goede 08 Dec '17

08 Dec '17

Hi Pierre-Louis, Bard, I've been looking into getting some Cherry Trail + RT5645 devices with a mono speaker I have to work properly. Specifically the goal is to mix the right channel into the left output, so that sounds which are only played over the right channel do not get lost. I'm using this UCM file: https://github.com/plbossart/UCM/blob/master/byt-rt5640/HiFi Looking at the rtl5640 UCM file, getting the mono speaker to work (in a first simple attempt) should be as simple as replacing: cset "name='SPOR MIX SPKVOL R Switch' on" With cset "name='SPOL MIX SPKVOL R Switch' on" But this does not work, the speaker test sound for the right speaker is still silent (works with headphones). I've also looked into directly poking the RT5645_SPO_MIXER i2c register for testing, but when checking its value with the original unmodified UCM file: [root@localhost ~]# i2cget -y -f 1 0x1a 0x48 w 0x06c8 That is: 0xc806 as the output of i2cget "w" mode needs byteswapping. Note that both the RT5645_M_SV_L_SPM_L and RT5645_M_SV_R_SPM_L bits are already cleared, which is weird as this is before I've modified anything. Also the RT5645_M_SV_R_SPM_R bit is cleared, but that is expected. Even if I manually set reg 0x48 to 0x07c8 which AFAIK should enable output of both left and right channels on the left speaker I still only get sounds played on the left channel. Any insights / help with this would be very much welcome. Regards, Hans

2 7

[alsa-devel] [PATCH] ASoC: wm_adsp: Fix validation of firmware and coeff lengths
by Ben Hutchings 08 Dec '17

08 Dec '17

The checks for whether another region/block header could be present are subtracting the size from the current offset. Obviously we should instead subtract the offset from the size. The checks for whether the region/block data fit in the file are adding the data size to the current offset and header size, without checking for integer overflow. Rearrange these so that overflow is impossible. Cc: stable(a)vger.kernel.org Signed-off-by: Ben Hutchings <ben.hutchings(a)codethink.co.uk> --- sound/soc/codecs/wm_adsp.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/sound/soc/codecs/wm_adsp.c b/sound/soc/codecs/wm_adsp.c index 65c059b5ffd7..66e32f5d2917 100644 --- a/sound/soc/codecs/wm_adsp.c +++ b/sound/soc/codecs/wm_adsp.c @@ -1733,7 +1733,7 @@ static int wm_adsp_load(struct wm_adsp *dsp) le64_to_cpu(footer->timestamp)); while (pos < firmware->size && - pos - firmware->size > sizeof(*region)) { + sizeof(*region) < firmware->size - pos) { region = (void *)&(firmware->data[pos]); region_name = "Unknown"; reg = 0; @@ -1782,8 +1782,8 @@ static int wm_adsp_load(struct wm_adsp *dsp) regions, le32_to_cpu(region->len), offset, region_name); - if ((pos + le32_to_cpu(region->len) + sizeof(*region)) > - firmware->size) { + if (le32_to_cpu(region->len) > + firmware->size - pos - sizeof(*region)) { adsp_err(dsp, "%s.%d: %s region len %d bytes exceeds file length %zu\n", file, regions, region_name, @@ -2253,7 +2253,7 @@ static int wm_adsp_load_coeff(struct wm_adsp *dsp) blocks = 0; while (pos < firmware->size && - pos - firmware->size > sizeof(*blk)) { + sizeof(*blk) < firmware->size - pos) { blk = (void *)(&firmware->data[pos]); type = le16_to_cpu(blk->type); @@ -2327,8 +2327,8 @@ static int wm_adsp_load_coeff(struct wm_adsp *dsp) } if (reg) { - if ((pos + le32_to_cpu(blk->len) + sizeof(*blk)) > - firmware->size) { + if (le32_to_cpu(blk->len) > + firmware->size - pos - sizeof(*blk)) { adsp_err(dsp, "%s.%d: %s region len %d bytes exceeds file length %zu\n", file, blocks, region_name, -- 2.15.0.rc0

3 2

[alsa-devel] [PATCH RESEND] ASoC: wm_adsp: Fix validation of firmware and coeff lengths
by Ben Hutchings 08 Dec '17

08 Dec '17

The checks for whether another region/block header could be present are subtracting the size from the current offset. Obviously we should instead subtract the offset from the size. The checks for whether the region/block data fit in the file are adding the data size to the current offset and header size, without checking for integer overflow. Rearrange these so that overflow is impossible. Cc: stable(a)vger.kernel.org Signed-off-by: Ben Hutchings <ben.hutchings(a)codethink.co.uk> Acked-by: Charles Keepax <ckeepax(a)opensource.cirrus.com> Tested-by: Charles Keepax <ckeepax(a)opensource.cirrus.com> --- sound/soc/codecs/wm_adsp.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/sound/soc/codecs/wm_adsp.c b/sound/soc/codecs/wm_adsp.c index 65c059b5ffd7..66e32f5d2917 100644 --- a/sound/soc/codecs/wm_adsp.c +++ b/sound/soc/codecs/wm_adsp.c @@ -1733,7 +1733,7 @@ static int wm_adsp_load(struct wm_adsp *dsp) le64_to_cpu(footer->timestamp)); while (pos < firmware->size && - pos - firmware->size > sizeof(*region)) { + sizeof(*region) < firmware->size - pos) { region = (void *)&(firmware->data[pos]); region_name = "Unknown"; reg = 0; @@ -1782,8 +1782,8 @@ static int wm_adsp_load(struct wm_adsp *dsp) regions, le32_to_cpu(region->len), offset, region_name); - if ((pos + le32_to_cpu(region->len) + sizeof(*region)) > - firmware->size) { + if (le32_to_cpu(region->len) > + firmware->size - pos - sizeof(*region)) { adsp_err(dsp, "%s.%d: %s region len %d bytes exceeds file length %zu\n", file, regions, region_name, @@ -2253,7 +2253,7 @@ static int wm_adsp_load_coeff(struct wm_adsp *dsp) blocks = 0; while (pos < firmware->size && - pos - firmware->size > sizeof(*blk)) { + sizeof(*blk) < firmware->size - pos) { blk = (void *)(&firmware->data[pos]); type = le16_to_cpu(blk->type); @@ -2327,8 +2327,8 @@ static int wm_adsp_load_coeff(struct wm_adsp *dsp) } if (reg) { - if ((pos + le32_to_cpu(blk->len) + sizeof(*blk)) > - firmware->size) { + if (le32_to_cpu(blk->len) > + firmware->size - pos - sizeof(*blk)) { adsp_err(dsp, "%s.%d: %s region len %d bytes exceeds file length %zu\n", file, blocks, region_name, -- 2.15.0.rc0

2 1

[alsa-devel] [PATCH] alsa-lib: pcm: segmentation fault snd_pcm_open
by Markus Seeber 08 Dec '17

08 Dec '17

Previously, calling snd_pcm_open with a null pointer argument for "name" caused a segmentation fault in snd_config_search_definition. The segmentation fault resulted from invoking undefined behavior by calling strchr with a null pointer argument. The issue can be reproduced by passing a null pointer for "name" to snd_pcm_open. To prevent this instance of strchr to be invoked with a null pointer argument, snd_config_search_definition now returns EINVAL if argument "name" is a null pointer. Signed-off-by: Markus Seeber <markus.seeber(a)spectralbird.de> --- src/conf.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/src/conf.c b/src/conf.c index e71bb972..e04b575b 100644 --- a/src/conf.c +++ b/src/conf.c @@ -5078,6 +5078,8 @@ int snd_config_search_definition(snd_config_t *config, { snd_config_t *conf; char *key; + if(name == NULL) + return -EINVAL; const char *args = strchr(name, ':'); int err; if (args) { -- 2.15.1

2 1

[alsa-devel] [PATCH] ASoC: atmel-classd: select correct Kconfig symbol
by Arnd Bergmann 08 Dec '17

08 Dec '17

Commit 061981ff8cc8 ("ASoC: atmel: properly select dma driver state") changed the way that the dependencies are handled, but then the Class D amplifier support got merged, which used the old method. This seems to have triggered a very rare randconfig condition for me now, leading to a link error: sound/soc/atmel/atmel_ssc_dai.o: In function `atmel_ssc_set_audio': atmel_ssc_dai.c:(.text+0x79c): undefined reference to `atmel_pcm_dma_platform_register' atmel_ssc_dai.c:(.text+0x79c): relocation truncated to fit: R_AARCH64_CALL26 against undefined symbol `atmel_pcm_dma_platform_register' sound/soc/atmel/atmel_ssc_dai.o: In function `atmel_ssc_put_audio': atmel_ssc_dai.c:(.text+0xf24): undefined reference to `atmel_pcm_dma_platform_unregister' atmel_ssc_dai.c:(.text+0xf24): relocation truncated to fit: R_AARCH64_CALL26 against undefined symbol `atmel_pcm_dma_platform_unregister' Changing it to select SND_ATMEL_SOC_SSC_DMA as intended rather than SND_ATMEL_SOC_DMA directly makes it work again. Fixes: e0a25b6d1862 ("ASoC: atmel-classd: add the Audio Class D Amplifier") Signed-off-by: Arnd Bergmann <arnd(a)arndb.de> --- sound/soc/atmel/Kconfig | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/sound/soc/atmel/Kconfig b/sound/soc/atmel/Kconfig index 4a56f3dfba51..2d998e5c4fb5 100644 --- a/sound/soc/atmel/Kconfig +++ b/sound/soc/atmel/Kconfig @@ -64,7 +64,7 @@ config SND_AT91_SOC_SAM9X5_WM8731 config SND_ATMEL_SOC_CLASSD tristate "Atmel ASoC driver for boards using CLASSD" depends on ARCH_AT91 || COMPILE_TEST - select SND_ATMEL_SOC_DMA + select SND_ATMEL_SOC_SSC_DMA select REGMAP_MMIO help Say Y if you want to add support for Atmel ASoC driver for boards using -- 2.9.0

4 11

Re: [alsa-devel] [Intel-gfx] [PATCH] drm/i915: Remove unused IRQ chip data of HDMI LPE audio
by Ville Syrjälä 08 Dec '17

08 Dec '17

On Fri, Dec 08, 2017 at 08:34:16AM +0000, Chen, Augustine wrote: > Dear Sirs, > Please kindly help reviewing this patch for the Bugzilla below. > https://bugs.freedesktop.org/show_bug.cgi?id=103731 Resend the patch using git send-email please. -- Ville Syrjälä Intel OTC

1 0

[alsa-devel] [PATCH v2 0/2] fix race condition in rsnd_ssi_pointer_update
by jiada_wang＠mentor.com 08 Dec '17

08 Dec '17

From: Jiada Wang <jiada_wang(a)mentor.com> This patch set aims to fix the race condition in rsnd_ssi_pointer_update, between set of byte_pos and wrap it around when new buffer starts. Jiada Wang (2): ASoC: rsnd: ssi: fix race condition in rsnd_ssi_pointer_update ASoC: rsnd: ssi: remove unnesessary period_pos Changes for v2: * Updated commit message sound/soc/sh/rcar/ssi.c | 25 +++++++++++++------------ 1 file changed, 13 insertions(+), 12 deletions(-) -- 1.7.9.5

2 3