July 2016 - Alsa-devel - mailman.alsa-project.org

[alsa-devel] [PATCH v2 0/3] ASoC: topology: Add support for BE DAIs
by mengdong.lin＠linux.intel.com 29 Jul '16

29 Jul '16

From: Mengdong Lin <mengdong.lin(a)linux.intel.com> There is topology ABI udpate in this series. Topology core can check size of each ABI objects to detect version mismatch between user space and kernel. This series adds support for BE (Back End) DAIs: - Define the type and ABI struct for Backend DAIs. - The topology core can configure an existing BE DAI or create new BE DAIs. The user space series "topology: Add support for BE DAIs" is also submitted. There will be 2 pairs of kernel & user space series for PCM & BE links later. History: v2: Remove unused argument to fix the compiler warning. sound/soc/soc-topology.c: In function ‘soc_tplg_be_dai_config’: sound/soc/soc-topology.c:1789:22: warning: too many arguments for format [-Wformat-extra-args] dev_err(tplg->dev, "ASoC: Invalid BE DAI name\n", Mengdong Lin (3): ASoC: topology: ABI - Add the types for BE DAI ASoC: topology: Add support for configuring existing BE DAIs ASoC: topology: Able to create BE DAIs include/sound/soc-topology.h | 1 + include/uapi/sound/asoc.h | 34 +++++++++- sound/soc/soc-core.c | 3 +- sound/soc/soc-topology.c | 155 ++++++++++++++++++++++++++++++++++++++++++- 4 files changed, 188 insertions(+), 5 deletions(-) -- 2.5.0

2 4

[alsa-devel] Neverending HDA pin quirks, how does Windows cope?
by Daniel Drake 28 Jul '16

28 Jul '16

Hi, We're constantly having to support new laptop models from vendors and the majority of them always seem to need some kind of HDA quirk adding, especially for pin usage info where the BIOS data is clearly incorrect. From the vast collection of quirks present in Linux's HDA drivers I see we are not alone. We're trying to figure out if there is a way the BIOS can be fixed for the cases that come from our partners, but I am also curious as to how this is managed on Windows and if we can learn anything from that. I had a look at the windows driver for some affected hardware. I was unable to identify any hardware-specific quirks in the .inf files, but using "strings" I found that various binary .sys, .dll and .dat files refer to hundreds of different laptop models. That might suggest that for each new laptop model, they make a modification to the windows driver, recompile, and get the new version signed again by MS? Sounds insane, anyone know for sure how this is handled? Thanks Daniel

1 0

[alsa-devel] [PATCH v4] topology: Return -EINVAL at once on failure to find a reference
by mengdong.lin＠linux.intel.com 28 Jul '16

28 Jul '16

From: Mengdong Lin <mengdong.lin(a)linux.intel.com> In building phase, return error -EINVAL at once if an element's references (e.g. TLV, text, control or private data) cannot be found. Move this error check right after looking up a reference, to make the code more clear. And checking the return value of tplg_copy_data() is enough, no need to check ref->elem for data references. Signed-off-by: Mengdong Lin <mengdong.lin(a)linux.intel.com> diff --git a/src/topology/ctl.c b/src/topology/ctl.c index 54f8e44..d910230 100644 --- a/src/topology/ctl.c +++ b/src/topology/ctl.c @@ -135,24 +135,23 @@ static int tplg_build_mixer_control(snd_tplg_t *tplg, if (ref->type == SND_TPLG_TYPE_TLV) { ref->elem = tplg_elem_lookup(&tplg->tlv_list, ref->id, SND_TPLG_TYPE_TLV); - if (ref->elem) - err = copy_tlv(elem, ref->elem); + if (!ref->elem) + goto ref_missing; + copy_tlv(elem, ref->elem); } else if (ref->type == SND_TPLG_TYPE_DATA) { err = tplg_copy_data(tplg, elem, ref); if (err < 0) return err; } - - if (!ref->elem) { - SNDERR("error: cannot find '%s' referenced by" - " control '%s'\n", ref->id, elem->id); - return -EINVAL; - } else if (err < 0) - return err; } return 0; + +ref_missing: + SNDERR("error: cannot find '%s' referenced by control '%s'\n", + ref->id, elem->id); + return -EINVAL; } static void copy_enum_texts(struct tplg_elem *enum_elem, @@ -183,23 +182,23 @@ static int tplg_build_enum_control(snd_tplg_t *tplg, if (ref->type == SND_TPLG_TYPE_TEXT) { ref->elem = tplg_elem_lookup(&tplg->text_list, ref->id, SND_TPLG_TYPE_TEXT); - if (ref->elem) - copy_enum_texts(elem, ref->elem); + if (!ref->elem) + goto ref_missing; + copy_enum_texts(elem, ref->elem); } else if (ref->type == SND_TPLG_TYPE_DATA) { err = tplg_copy_data(tplg, elem, ref); if (err < 0) return err; } - if (!ref->elem) { - SNDERR("error: cannot find '%s' referenced by" - " control '%s'\n", ref->id, elem->id); - return -EINVAL; - } else if (err < 0) - return err; } return 0; + +ref_missing: + SNDERR("error: cannot find '%s' referenced by control '%s'\n", + ref->id, elem->id); + return -EINVAL; } /* check referenced private data for a byte control */ diff --git a/src/topology/dapm.c b/src/topology/dapm.c index 4d343b2..988a5fc 100644 --- a/src/topology/dapm.c +++ b/src/topology/dapm.c @@ -160,7 +160,10 @@ static int tplg_build_widget(snd_tplg_t *tplg, base = &elem->ref_list; - /* for each ref in this control elem */ + /* Look up and merge each reference in this widget elem. + * For widgets added by C API, its control elements are already set + * and so no need to look up. + */ list_for_each(pos, base) { ref = list_entry(pos, struct tplg_ref, list); @@ -170,48 +173,50 @@ static int tplg_build_widget(snd_tplg_t *tplg, if (!ref->elem) ref->elem = tplg_elem_lookup(&tplg->mixer_list, ref->id, SND_TPLG_TYPE_MIXER); - if (ref->elem) - err = copy_dapm_control(elem, ref->elem); + if (!ref->elem) + goto ref_missing; + + err = copy_dapm_control(elem, ref->elem); break; case SND_TPLG_TYPE_ENUM: if (!ref->elem) ref->elem = tplg_elem_lookup(&tplg->enum_list, ref->id, SND_TPLG_TYPE_ENUM); - if (ref->elem) - err = copy_dapm_control(elem, ref->elem); + if (!ref->elem) + goto ref_missing; + + err = copy_dapm_control(elem, ref->elem); break; case SND_TPLG_TYPE_BYTES: if (!ref->elem) ref->elem = tplg_elem_lookup(&tplg->bytes_ext_list, ref->id, SND_TPLG_TYPE_BYTES); - if (ref->elem) - err = copy_dapm_control(elem, ref->elem); + if (!ref->elem) + goto ref_missing; + + err = copy_dapm_control(elem, ref->elem); break; case SND_TPLG_TYPE_DATA: err = tplg_copy_data(tplg, elem, ref); - if (err < 0) - return err; break; default: break; } - if (!ref->elem) { - SNDERR("error: cannot find control '%s'" - " referenced by widget '%s'\n", - ref->id, elem->id); - return -EINVAL; - } - if (err < 0) return err; } return 0; + +ref_missing: + SNDERR("error: cannot find '%s' referenced by widget '%s'\n", + ref->id, elem->id); + return -EINVAL; } int tplg_build_widgets(snd_tplg_t *tplg) -- 2.5.0

3 3

[alsa-devel] [PATCH] topology: Fix inaccurate comments when building a widget element
by mengdong.lin＠linux.intel.com 28 Jul '16

28 Jul '16

From: Mengdong Lin <mengdong.lin(a)linux.intel.com> The original comments use 'control' by mistake when building a widget. Also describe the difference between the referenced controls defined by C API and text conf file. Signed-off-by: Mengdong Lin <mengdong.lin(a)linux.intel.com> diff --git a/src/topology/dapm.c b/src/topology/dapm.c index e3c90d8..0141653 100644 --- a/src/topology/dapm.c +++ b/src/topology/dapm.c @@ -150,7 +150,7 @@ static int copy_dapm_control(struct tplg_elem *elem, struct tplg_elem *ref) return 0; } -/* check referenced controls for a widget */ +/* Merge referenced controls and private data for a widget */ static int tplg_build_widget(snd_tplg_t *tplg, struct tplg_elem *elem) { @@ -160,7 +160,12 @@ static int tplg_build_widget(snd_tplg_t *tplg, base = &elem->ref_list; - /* for each ref in this control elem */ + /* Look up and merge each control or data reference of this + * widget element. For a widget defined by C API, its control + * references are already bound and can be directly merged + * without lookup. Only control references defined by text + * conf file need lookup before being merged. + */ list_for_each(pos, base) { ref = list_entry(pos, struct tplg_ref, list); -- 2.5.0

1 0

[alsa-devel] [PATCH] ASoC: rsnd: Fixup SRCm_IFSVR calculate method
by Kuninori Morimoto 27 Jul '16

27 Jul '16

From: Hiroyuki Yokoyama <hiroyuki.yokoyama.vx(a)renesas.com> This patch fixes the calculation accuracy degradation of SRCm_IFSVR register value. Signed-off-by: Hiroyuki Yokoyama <hiroyuki.yokoyama.vx(a)renesas.com> Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx(a)renesas.com> --- sound/soc/sh/rcar/src.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/sound/soc/sh/rcar/src.c b/sound/soc/sh/rcar/src.c index e39f916..969a516 100644 --- a/sound/soc/sh/rcar/src.c +++ b/sound/soc/sh/rcar/src.c @@ -226,8 +226,12 @@ static void rsnd_src_set_convert_rate(struct rsnd_dai_stream *io, ifscr = 0; fsrate = 0; if (fin != fout) { + u64 n; + ifscr = 1; - fsrate = 0x0400000 / fout * fin; + n = (u64)0x0400000 * fin; + do_div(n, fout); + fsrate = n; } /* -- 1.9.1

2 1

[alsa-devel] [PATCH v2] ASoC: dapm: Add a dummy snd_pcm_runtime to avoid NULL pointer access
by Nicolin Chen 27 Jul '16

27 Jul '16

The SND_SOC_DAPM_PRE_PMU case would call startup()/hw_params() that might access substream->runtime through other functions. For example: Unable to handle kernel NULL pointer dereference at virtual address [....] PC is at snd_pcm_hw_rule_add+0x24/0x1b0 LR is at snd_pcm_hw_constraint_list+0x20/0x28 [....] Process arecord (pid: 424, stack limit = 0xffffffc1ecaf0020) Call trace: [<ffffffc00086be68>] snd_pcm_hw_rule_add+0x24/0x1b0 [<ffffffc00086c014>] snd_pcm_hw_constraint_list+0x20/0x28 [<ffffffc0008b47a4>] cs53l30_pcm_startup+0x24/0x30 [<ffffffc0008a6260>] snd_soc_dai_link_event+0x290/0x354 [<ffffffc0008a7528>] dapm_seq_check_event.isra.31+0x134/0x2c8 [<ffffffc0008a7768>] dapm_seq_run_coalesced+0x94/0x1c8 [<ffffffc0008a7940>] dapm_seq_run+0xa4/0x404 [<ffffffc0008a8bac>] dapm_power_widgets+0x524/0x984 [<ffffffc0008ab1c4>] snd_soc_dapm_stream_event+0x8c/0xa8 [<ffffffc0008ac7f4>] soc_pcm_prepare+0x10c/0x1ec [<ffffffc000865b9c>] snd_pcm_do_prepare+0x1c/0x38 [<ffffffc000865600>] snd_pcm_action_single+0x40/0x88 [<ffffffc0008656b8>] snd_pcm_action_nonatomic+0x70/0x90 [<ffffffc000868d28>] snd_pcm_common_ioctl1+0xb6c/0xdd8 [<ffffffc000869508>] snd_pcm_capture_ioctl1+0x200/0x334 [<ffffffc00086a084>] snd_pcm_ioctl_compat+0x648/0x95c [<ffffffc0001ff4b4>] compat_SyS_ioctl+0xac/0xfc4 [<ffffffc000084cf0>] el0_svc_naked+0x24/0x28 ---[ end trace 0dc4f99c2759c35c ]--- So this patch adds a dummy runtime for the original dummy substream to merely avoid the NULL pointer access. Signed-off-by: Nicolin Chen <nicoleotsuka(a)gmail.com> --- Changelog: v1->v2 * Allocate the runtime from heap instead of stack due to limitation of the stack sizes on different platforms. sound/soc/soc-dapm.c | 10 ++++++++++ 1 file changed, 10 insertions(+) diff --git a/sound/soc/soc-dapm.c b/sound/soc/soc-dapm.c index 8698c26..d908ff8 100644 --- a/sound/soc/soc-dapm.c +++ b/sound/soc/soc-dapm.c @@ -3493,6 +3493,7 @@ static int snd_soc_dai_link_event(struct snd_soc_dapm_widget *w, const struct snd_soc_pcm_stream *config = w->params + w->params_select; struct snd_pcm_substream substream; struct snd_pcm_hw_params *params = NULL; + struct snd_pcm_runtime *runtime = NULL; u64 fmt; int ret; @@ -3541,6 +3542,14 @@ static int snd_soc_dai_link_event(struct snd_soc_dapm_widget *w, memset(&substream, 0, sizeof(substream)); + /* Allocate a dummy snd_pcm_runtime for startup() and other ops() */ + runtime = kzalloc(sizeof(*runtime), GFP_KERNEL); + if (!runtime) { + ret = -ENOMEM; + goto out; + } + substream.runtime = runtime; + switch (event) { case SND_SOC_DAPM_PRE_PMU: substream.stream = SNDRV_PCM_STREAM_CAPTURE; @@ -3606,6 +3615,7 @@ static int snd_soc_dai_link_event(struct snd_soc_dapm_widget *w, } out: + kfree(runtime); kfree(params); return ret; } -- 2.1.4

2 1

[alsa-devel] [PATCH 0/3] ALSA: usb-us122l: compatible ioctl support and fix sparse warnings
by Takashi Sakamoto 27 Jul '16

27 Jul '16

Hi, This patchset is to fix some sparse warnings in ALSA USB US122L module. Additionally, compatible ioctl is supported. Takashi Sakamoto (3): usb-us122l: replace 'unsigned' with 'unsigned int' usb-us122l: add support for compatible ioctl usb-us122l: add qualifier into pointer to userspace sound/usb/usx2y/us122l.c | 19 ++++++++++++++++--- 1 file changed, 16 insertions(+), 3 deletions(-) -- 2.7.4

2 5

[alsa-devel] da7213 questions
by Peter Meerwald-Stadler 27 Jul '16

27 Jul '16

Hello, I have a couple of questions/comments regarding the da7213 driver in mainline: (1) the da7213 driver enables mclk in BIAS_SUSPEND, so essentially the clock is always running (since idle_bias_off is not set); at least on my platform, clk_set_rate() fails when the clock is already enabled (2) da7213's srm_en variable is redundant and can be removed (3) for 32 kHz PLL mode (DAI master), section 13.28 of the datasheet indicates that PLL_SRM_EN should be asserted; the da7213 drivers goesn't (4) is only S16_LE actually supported? #define DA7213_FORMATS (SNDRV_PCM_FMTBIT_S16_LE | SNDRV_PCM_FMTBIT_S20_3LE |\ SNDRV_PCM_FMTBIT_S24_LE | SNDRV_PCM_FMTBIT_S32_LE) but in da7213_set_dai_fmt() we have /* By default only 32 BCLK per WCLK is supported */ dai_clk_mode |= DA7213_DAI_BCLKS_PER_WCLK_32; thanks, regards, p. -- Peter Meerwald-Stadler +43-664-2444418 (mobile)

2 1

[alsa-devel] Is this bug ? dev_set/get_drvdata() of rtd
by Kuninori Morimoto 27 Jul '16

27 Jul '16

Hi ALSA SoC I'm now checking ALSA SoC framework, and I noticed strange operation. Is this bug ? I think device data of "rtd->dev" is set as "rtd" on soc_post_component_init() by using dev_set_drvdata(), and many soc-core functions are based on it (= they are using dev_get_drvdata()) But, soc.h has snd_soc_pcm_set_drvdata(). device data of rtd->dev will be overwrite by this function ? But is this OK ? or bug ? It seems below drivers are using this sound/soc/intel/haswell/sst-haswell-pcm.c sound/soc/qcom/lpass-platform.c linux/sound/soc/soc-core.c static int soc_post_component_init(xxx) { ... dev_set_drvdata(rtd->dev, rtd); ... } static ssize_t codec_reg_show(xxx) { struct snd_soc_pcm_runtime *rtd = dev_get_drvdata(dev); ... } static ssize_t pmdown_time_show(xxx) { struct snd_soc_pcm_runtime *rtd = dev_get_drvdata(dev); ... } static ssize_t pmdown_time_set(xxx) { struct snd_soc_pcm_runtime *rtd = dev_get_drvdata(dev); ... } static umode_t soc_dev_attr_is_visible(xxx) { struct snd_soc_pcm_runtime *rtd = dev_get_drvdata(dev); ... } linux/include/sound/soc.h static inline void snd_soc_pcm_set_drvdata(struct snd_soc_pcm_runtime *rtd, void *data) { dev_set_drvdata(rtd->dev, data); } Best regards --- Kuninori Morimoto

2 1

[alsa-devel] [PATCH v2 0/2] topology: Fix for handling widgets' references
by mengdong.lin＠linux.intel.com 27 Jul '16

27 Jul '16

From: Mengdong Lin <mengdong.lin(a)linux.intel.com> This series fixes an error when merging a widget's private data and an inaccurate output message. History: v2: Split v1 into 2 patches. And set the valid reference element value only on success. Mengdong Lin (2): topology: Fix inaccurate message on failure to find a widgets's reference topology: Fix the missing referenced elem ptr when merging private data src/topology/dapm.c | 2 +- src/topology/data.c | 2 ++ 2 files changed, 3 insertions(+), 1 deletion(-) -- 2.5.0

4 6