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

[alsa-devel] [PATCH 1/1] ASoC: fsl_asrc: use dma_transfer_direction enum when calling dmaengine_prep_dma_cyclic
by Nicolas Iooss 21 Apr '17

21 Apr '17

When fsl_asrc_dma_prepare_and_submit() calls dmaengine_prep_dma_cyclic(), it uses either DMA_TO_DEVICE or DMA_FROM_DEVICE. These values are both items of dma_data_direction enum, not of dma_data_direction enum, which is the type expected by dmaengine_prep_dma_cyclic(). Replace DMA_TO_DEVICE with DMA_MEM_TO_DEV and DMA_FROM_DEVICE with DMA_DEV_TO_MEM to fix this type mismatch issue. Signed-off-by: Nicolas Iooss <nicolas.iooss_linux(a)m4x.org> --- sound/soc/fsl/fsl_asrc_dma.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/sound/soc/fsl/fsl_asrc_dma.c b/sound/soc/fsl/fsl_asrc_dma.c index dc30d780f874..282d841840b1 100644 --- a/sound/soc/fsl/fsl_asrc_dma.c +++ b/sound/soc/fsl/fsl_asrc_dma.c @@ -76,7 +76,7 @@ static int fsl_asrc_dma_prepare_and_submit(struct snd_pcm_substream *substream) pair->dma_chan[!dir], runtime->dma_addr, snd_pcm_lib_buffer_bytes(substream), snd_pcm_lib_period_bytes(substream), - dir == OUT ? DMA_TO_DEVICE : DMA_FROM_DEVICE, flags); + dir == OUT ? DMA_MEM_TO_DEV : DMA_DEV_TO_MEM, flags); if (!pair->desc[!dir]) { dev_err(dev, "failed to prepare slave DMA for Front-End\n"); return -ENOMEM; -- 2.11.0

4 3

[alsa-devel] [PATCH] ASoC: Intel: Skylake: Fix a couple user after free bugs
by Dan Carpenter 21 Apr '17

21 Apr '17

We need to use the _safe() version of list_for_each_entry() here because of the kfree(modules). Fixes: b8c722ddd548 ("ASoC: Intel: Skylake: Add support for deferred DSP module bind") Signed-off-by: Dan Carpenter <dan.carpenter(a)oracle.com> diff --git a/sound/soc/intel/skylake/skl-pcm.c b/sound/soc/intel/skylake/skl-pcm.c index 600faad19bd4..d43d1976dd3b 100644 --- a/sound/soc/intel/skylake/skl-pcm.c +++ b/sound/soc/intel/skylake/skl-pcm.c @@ -1323,10 +1323,10 @@ int skl_platform_unregister(struct device *dev) { struct hdac_ext_bus *ebus = dev_get_drvdata(dev); struct skl *skl = ebus_to_skl(ebus); - struct skl_module_deferred_bind *modules; + struct skl_module_deferred_bind *modules, *tmp; if (!list_empty(&skl->bind_list)) { - list_for_each_entry(modules, &skl->bind_list, node) { + list_for_each_entry_safe(modules, tmp, &skl->bind_list, node) { list_del(&modules->node); kfree(modules); } diff --git a/sound/soc/intel/skylake/skl-topology.c b/sound/soc/intel/skylake/skl-topology.c index 74b3acf3bb07..aea7917e791c 100644 --- a/sound/soc/intel/skylake/skl-topology.c +++ b/sound/soc/intel/skylake/skl-topology.c @@ -1091,7 +1091,7 @@ static int skl_tplg_mixer_dapm_post_pmd_event(struct snd_soc_dapm_widget *w, struct skl_module_cfg *src_module = NULL, *dst_module; struct skl_sst *ctx = skl->skl_sst; struct skl_pipe *s_pipe = mconfig->pipe; - struct skl_module_deferred_bind *modules; + struct skl_module_deferred_bind *modules, *tmp; if (s_pipe->state == SKL_PIPE_INVALID) return -EINVAL; @@ -1105,7 +1105,7 @@ static int skl_tplg_mixer_dapm_post_pmd_event(struct snd_soc_dapm_widget *w, src_module = w_module->w->priv; - list_for_each_entry(modules, &skl->bind_list, node) { + list_for_each_entry_safe(modules, tmp, &skl->bind_list, node) { /* * When the destination module is deleted, Unbind the * modules from deferred bind list.

2 1

[alsa-devel] [PATCH] ASoC: cs35l35: Allow user to configure IMON SCALE
by Charles Keepax 21 Apr '17

21 Apr '17

On the chip the IMON signal is a full 24-bits however normally only some of the bits will be sent over the bus. The chip provides a field to select which bits of the IMON will be sent back, this is the only feedback signal that has this feature. Add an additional entry to the cirrus,imon device tree property to allow the IMON scale parameter to be passed. Signed-off-by: Charles Keepax <ckeepax(a)opensource.wolfsonmicro.com> --- I realise this is changing the binding slightly, but thought we might be able to sneak it before cs35l35 ships as it hasn't shipped in a kernel yet. However, it would be relatively simple (although makes the code a little messier) to support both the 3 and 4 entry versions of the binding if people prefer? Thanks, Charles .../devicetree/bindings/sound/cs35l35.txt | 4 ++-- include/sound/cs35l35.h | 1 + sound/soc/codecs/cs35l35.c | 22 +++++++++++++++------- sound/soc/codecs/cs35l35.h | 3 +++ 4 files changed, 21 insertions(+), 9 deletions(-) diff --git a/Documentation/devicetree/bindings/sound/cs35l35.txt b/Documentation/devicetree/bindings/sound/cs35l35.txt index 457d176..016b768 100644 --- a/Documentation/devicetree/bindings/sound/cs35l35.txt +++ b/Documentation/devicetree/bindings/sound/cs35l35.txt @@ -118,8 +118,8 @@ Optional Monitor Signal Format sub-node: Sections 7.44 - 7.53 lists values for the depth, location, and frame for each monitoring signal. - - cirrus,imon : 3 8 bit values to set the depth, location, and frame - of the IMON monitor signal. + - cirrus,imon : 4 8 bit values to set the depth, location, frame and ADC + scale of the IMON monitor signal. - cirrus,vmon : 3 8 bit values to set the depth, location, and frame of the VMON monitor signal. diff --git a/include/sound/cs35l35.h b/include/sound/cs35l35.h index 88744bb..29da899 100644 --- a/include/sound/cs35l35.h +++ b/include/sound/cs35l35.h @@ -57,6 +57,7 @@ struct monitor_cfg { u8 imon_dpth; u8 imon_loc; u8 imon_frm; + u8 imon_scale; u8 vmon_dpth; u8 vmon_loc; u8 vmon_frm; diff --git a/sound/soc/codecs/cs35l35.c b/sound/soc/codecs/cs35l35.c index 6ecb7dd..88c48e2 100644 --- a/sound/soc/codecs/cs35l35.c +++ b/sound/soc/codecs/cs35l35.c @@ -918,6 +918,11 @@ static int cs35l35_codec_probe(struct snd_soc_codec *codec) CS35L35_MON_FRM_MASK, monitor_config->imon_frm << CS35L35_MON_FRM_SHIFT); + regmap_update_bits(cs35l35->regmap, + CS35L35_IMON_SCALE_CTL, + CS35L35_IMON_SCALE_MASK, + monitor_config->imon_scale << + CS35L35_IMON_SCALE_SHIFT); } if (monitor_config->vpmon_specs) { regmap_update_bits(cs35l35->regmap, @@ -1161,7 +1166,9 @@ static int cs35l35_handle_of_data(struct i2c_client *i2c_client, struct classh_cfg *classh_config = &pdata->classh_algo; struct monitor_cfg *monitor_config = &pdata->mon_cfg; unsigned int val32 = 0; - u8 monitor_array[3]; + u8 monitor_array[4]; + const int imon_array_size = ARRAY_SIZE(monitor_array); + const int mon_array_size = imon_array_size - 1; int ret = 0; if (!np) @@ -1302,15 +1309,16 @@ static int cs35l35_handle_of_data(struct i2c_client *i2c_client, monitor_config->is_present = signal_format ? true : false; if (monitor_config->is_present) { ret = of_property_read_u8_array(signal_format, "cirrus,imon", - monitor_array, ARRAY_SIZE(monitor_array)); + monitor_array, imon_array_size); if (!ret) { monitor_config->imon_specs = true; monitor_config->imon_dpth = monitor_array[0]; monitor_config->imon_loc = monitor_array[1]; monitor_config->imon_frm = monitor_array[2]; + monitor_config->imon_scale = monitor_array[3]; } ret = of_property_read_u8_array(signal_format, "cirrus,vmon", - monitor_array, ARRAY_SIZE(monitor_array)); + monitor_array, mon_array_size); if (!ret) { monitor_config->vmon_specs = true; monitor_config->vmon_dpth = monitor_array[0]; @@ -1318,7 +1326,7 @@ static int cs35l35_handle_of_data(struct i2c_client *i2c_client, monitor_config->vmon_frm = monitor_array[2]; } ret = of_property_read_u8_array(signal_format, "cirrus,vpmon", - monitor_array, ARRAY_SIZE(monitor_array)); + monitor_array, mon_array_size); if (!ret) { monitor_config->vpmon_specs = true; monitor_config->vpmon_dpth = monitor_array[0]; @@ -1326,7 +1334,7 @@ static int cs35l35_handle_of_data(struct i2c_client *i2c_client, monitor_config->vpmon_frm = monitor_array[2]; } ret = of_property_read_u8_array(signal_format, "cirrus,vbstmon", - monitor_array, ARRAY_SIZE(monitor_array)); + monitor_array, mon_array_size); if (!ret) { monitor_config->vbstmon_specs = true; monitor_config->vbstmon_dpth = monitor_array[0]; @@ -1334,7 +1342,7 @@ static int cs35l35_handle_of_data(struct i2c_client *i2c_client, monitor_config->vbstmon_frm = monitor_array[2]; } ret = of_property_read_u8_array(signal_format, "cirrus,vpbrstat", - monitor_array, ARRAY_SIZE(monitor_array)); + monitor_array, mon_array_size); if (!ret) { monitor_config->vpbrstat_specs = true; monitor_config->vpbrstat_dpth = monitor_array[0]; @@ -1342,7 +1350,7 @@ static int cs35l35_handle_of_data(struct i2c_client *i2c_client, monitor_config->vpbrstat_frm = monitor_array[2]; } ret = of_property_read_u8_array(signal_format, "cirrus,zerofill", - monitor_array, ARRAY_SIZE(monitor_array)); + monitor_array, mon_array_size); if (!ret) { monitor_config->zerofill_specs = true; monitor_config->zerofill_dpth = monitor_array[0]; diff --git a/sound/soc/codecs/cs35l35.h b/sound/soc/codecs/cs35l35.h index 54e9ac5..5a6e43a 100644 --- a/sound/soc/codecs/cs35l35.h +++ b/sound/soc/codecs/cs35l35.h @@ -148,6 +148,9 @@ #define CS35L35_MON_FRM_MASK 0x80 #define CS35L35_MON_FRM_SHIFT 7 +#define CS35L35_IMON_SCALE_MASK 0xF8 +#define CS35L35_IMON_SCALE_SHIFT 3 + #define CS35L35_MS_MASK 0x80 #define CS35L35_MS_SHIFT 7 #define CS35L35_SPMODE_MASK 0x40 -- 2.1.4

4 3

[alsa-devel] [PATCH] ASoC: qcom: move clock names into LPASS variant struct
by Linus Walleij 21 Apr '17

21 Apr '17

The clock names for the two supported codecs are either "mi2s-*" name variants generated by code. This naming scheme does not work for platforms like MSM8660 which has I2S channels named CODEC_I2S_SPKR (rather than just "MI2S tertiary" and other repetitive names) and consequently have clocks named "codec-i2s-spkr-osr-clk" and similar. Skip the runtime generation of clock names and replace it with name lookup tables encoded into the variant data. Signed-off-by: Linus Walleij <linus.walleij(a)linaro.org> --- sound/soc/qcom/lpass-apq8016.c | 12 ++++++++++++ sound/soc/qcom/lpass-cpu.c | 22 +++++++--------------- sound/soc/qcom/lpass-ipq806x.c | 6 ++++++ sound/soc/qcom/lpass.h | 2 ++ 4 files changed, 27 insertions(+), 15 deletions(-) diff --git a/sound/soc/qcom/lpass-apq8016.c b/sound/soc/qcom/lpass-apq8016.c index 8aed72be3224..8a74844d99e2 100644 --- a/sound/soc/qcom/lpass-apq8016.c +++ b/sound/soc/qcom/lpass-apq8016.c @@ -231,6 +231,18 @@ static struct lpass_variant apq8016_data = { .wrdma_channels = 2, .dai_driver = apq8016_lpass_cpu_dai_driver, .num_dai = ARRAY_SIZE(apq8016_lpass_cpu_dai_driver), + .dai_osr_clk_names = (const char *[]) { + "mi2s-osr-clk0", + "mi2s-osr-clk1", + "mi2s-osr-clk2", + "mi2s-osr-clk3", + }, + .dai_bit_clk_names = (const char *[]) { + "mi2s-bit-clk0", + "mi2s-bit-clk1", + "mi2s-bit-clk2", + "mi2s-bit-clk3", + }, .init = apq8016_lpass_init, .exit = apq8016_lpass_exit, .alloc_dma_channel = apq8016_lpass_alloc_dma_channel, diff --git a/sound/soc/qcom/lpass-cpu.c b/sound/soc/qcom/lpass-cpu.c index 5202a584e0c6..292b103abada 100644 --- a/sound/soc/qcom/lpass-cpu.c +++ b/sound/soc/qcom/lpass-cpu.c @@ -429,7 +429,6 @@ int asoc_qcom_lpass_cpu_platform_probe(struct platform_device *pdev) struct lpass_variant *variant; struct device *dev = &pdev->dev; const struct of_device_id *match; - char clk_name[16]; int ret, i, dai_id; dsp_of_node = of_parse_phandle(pdev->dev.of_node, "qcom,adsp", 0); @@ -477,31 +476,24 @@ int asoc_qcom_lpass_cpu_platform_probe(struct platform_device *pdev) for (i = 0; i < variant->num_dai; i++) { dai_id = variant->dai_driver[i].id; - if (variant->num_dai > 1) - sprintf(clk_name, "mi2s-osr-clk%d", i); - else - sprintf(clk_name, "mi2s-osr-clk"); - drvdata->mi2s_osr_clk[dai_id] = devm_clk_get(&pdev->dev, - clk_name); + variant->dai_osr_clk_names[i]); if (IS_ERR(drvdata->mi2s_osr_clk[dai_id])) { dev_warn(&pdev->dev, - "error getting optional mi2s-osr-clk: %ld\n", + "%s() error getting optional %s: %ld\n", + __func__, + variant->dai_osr_clk_names[i], PTR_ERR(drvdata->mi2s_osr_clk[dai_id])); drvdata->mi2s_osr_clk[dai_id] = NULL; } - if (variant->num_dai > 1) - sprintf(clk_name, "mi2s-bit-clk%d", i); - else - sprintf(clk_name, "mi2s-bit-clk"); - drvdata->mi2s_bit_clk[dai_id] = devm_clk_get(&pdev->dev, - clk_name); + variant->dai_bit_clk_names[i]); if (IS_ERR(drvdata->mi2s_bit_clk[dai_id])) { dev_err(&pdev->dev, - "error getting mi2s-bit-clk: %ld\n", + "error getting %s: %ld\n", + variant->dai_bit_clk_names[i], PTR_ERR(drvdata->mi2s_bit_clk[dai_id])); return PTR_ERR(drvdata->mi2s_bit_clk[dai_id]); } diff --git a/sound/soc/qcom/lpass-ipq806x.c b/sound/soc/qcom/lpass-ipq806x.c index 608c1a92af8a..ca1e1f2d2787 100644 --- a/sound/soc/qcom/lpass-ipq806x.c +++ b/sound/soc/qcom/lpass-ipq806x.c @@ -92,6 +92,12 @@ static struct lpass_variant ipq806x_data = { .wrdma_channels = 4, .dai_driver = &ipq806x_lpass_cpu_dai_driver, .num_dai = 1, + .dai_osr_clk_names = (const char *[]) { + "mi2s-osr-clk", + }, + .dai_bit_clk_names = (const char *[]) { + "mi2s-bit-clk", + }, .alloc_dma_channel = ipq806x_lpass_alloc_dma_channel, .free_dma_channel = ipq806x_lpass_free_dma_channel, }; diff --git a/sound/soc/qcom/lpass.h b/sound/soc/qcom/lpass.h index 9b031352ea3c..b848db2d6c3d 100644 --- a/sound/soc/qcom/lpass.h +++ b/sound/soc/qcom/lpass.h @@ -91,6 +91,8 @@ struct lpass_variant { /* SOC specific dais */ struct snd_soc_dai_driver *dai_driver; int num_dai; + const char * const *dai_osr_clk_names; + const char * const *dai_bit_clk_names; }; /* register the platform driver from the CPU DAI driver */ -- 2.9.3

2 1

[alsa-devel] [alsa-lib][PATCH v2 0/3] timer/hwdep: update included headers with UAPI headers in recent kernel
by Takashi Sakamoto 21 Apr '17

21 Apr '17

Hi, In this subsystem, interfaces between kernel/userspace are described in UAPI headers of kernel source. User space library (alsa-lib) includes this header as its code base. When developers update UAPI headers in development cycle of kernel, corresponding headers in alsa-lib should be updated synchronously. The purpose of this patchset is to catch up recent UAPI header for userspace library. There're missing changes added at v4.7, v4.9 and v4.12 development period. v2 changes: - Update commit description. - Reorder commits so that aligned according to date of corresponding code change. Takashi Sakamoto (3): timer: obsolete legacy rtctimer instance hwdep: add Line6 USB series support hwdep: add support for MOTU FireWire series and RME Fireface series include/hwdep.h | 5 ++++- include/sound/asound.h | 7 +++++-- include/timer.h | 2 +- 3 files changed, 10 insertions(+), 4 deletions(-) -- 2.9.3

2 4

[alsa-devel] [alsa-lib][PATCH 0/3] hwdep: update included headers with UAPI headers in recent kernel
by Takashi Sakamoto 21 Apr '17

21 Apr '17

Hi, In this subsystem, interfaces between kernel/userspace are described in UAPI headers of kernel source. User space library (alsa-lib) includes this header as its code base. When developers update UAPI headers in development cycle of kernel, corresponding headers in alsa-lib should be updated. The purpose of this patchset is to catch up recent UAPI header for userspace library. There're missing changes added at v4.7, v4.9 and v4.12 development period. Takashi Sakamoto (3): hwdep: add Line6 USB series support hwdep: add support for MOTU FireWire series and RME Fireface series timer: obsolete legacy rtctimer instance include/hwdep.h | 5 ++++- include/sound/asound.h | 7 +++++-- include/timer.h | 2 +- 3 files changed, 10 insertions(+), 4 deletions(-) -- 2.9.3

1 4

[alsa-devel] [PATCH 0/4] topology: Support using "index" to group objects when generating binary for kernel
by fuweix.tang＠intel.com 21 Apr '17

21 Apr '17

From: Fuwei Tang <fuweix.tang(a)intel.com> The index value of an topology object can be used to group objects for different use cases. And so the audio kernel driver can ask topology core to load all objects of the specified index value(use case). The function is already supported in kernel. So in the user space, for topology objects of the same type but with different index values, they should be written into different blocks in the binary for kernel. The index value is stored in the block header. Fuwei Tang (4): topology: Insert new element based on its index value topology: Remove code parsing index value in paring each object topology: Look up references for an object based on its index topology: Group elements with the same index value into a block include/topology.h | 10 +++-- src/topology/builder.c | 94 +++++++++++++++++++++++++---------------------- src/topology/ctl.c | 31 +--------------- src/topology/dapm.c | 27 +++++--------- src/topology/data.c | 15 ++------ src/topology/elem.c | 74 +++++++++++++++++++++++++++---------- src/topology/pcm.c | 58 ++++++----------------------- src/topology/tplg_local.h | 3 +- 8 files changed, 141 insertions(+), 171 deletions(-) -- 2.7.4

2 5

[alsa-devel] [PATCH] ALSA: hda - Fix headset microphone detection for ASUS N551 and N751
by Takashi Iwai 21 Apr '17

21 Apr '17

From: Mikhail Paulyshka <me(a)mixaill.tk> Headset microphone does not work out of the box on ASUS Nx51 laptops. This patch fixes it. Patch tested on Asus N551 laptop. Asus N751 part is not tested, but according to [1] this laptop uses the same audiosystem. 1. https://bugzilla.kernel.org/show_bug.cgi?id=117781 Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=195437 Signed-off-by: Mikhail Paulyshka <me(a)mixaill.tk> Signed-off-by: Takashi Iwai <tiwai(a)suse.de> --- sound/pci/hda/patch_realtek.c | 12 ++++++++++-- 1 file changed, 10 insertions(+), 2 deletions(-) diff --git a/sound/pci/hda/patch_realtek.c b/sound/pci/hda/patch_realtek.c index 9371f1a95b33..05a26fba5ef2 100644 --- a/sound/pci/hda/patch_realtek.c +++ b/sound/pci/hda/patch_realtek.c @@ -6815,6 +6815,7 @@ enum { ALC668_FIXUP_DELL_DISABLE_AAMIX, ALC668_FIXUP_DELL_XPS13, ALC662_FIXUP_ASUS_Nx50, + ALC668_FIXUP_ASUS_Nx51_HEADSET_MODE, ALC668_FIXUP_ASUS_Nx51, ALC891_FIXUP_HEADSET_MODE, ALC891_FIXUP_DELL_MIC_NO_PRESENCE, @@ -7068,14 +7069,21 @@ static const struct hda_fixup alc662_fixups[] = { .chained = true, .chain_id = ALC662_FIXUP_BASS_1A }, + [ALC668_FIXUP_ASUS_Nx51_HEADSET_MODE] = { + .type = HDA_FIXUP_FUNC, + .v.func = alc_fixup_headset_mode_alc668, + .chain_id = ALC662_FIXUP_BASS_CHMAP + }, [ALC668_FIXUP_ASUS_Nx51] = { .type = HDA_FIXUP_PINS, .v.pins = (const struct hda_pintbl[]) { - {0x1a, 0x90170151}, /* bass speaker */ + { 0x19, 0x03a1913d }, /* use as headphone mic, without its own jack detect */ + { 0x1a, 0x90170151 }, /* bass speaker */ + { 0x1b, 0x03a1113c }, /* use as headset mic, without its own jack detect */ {} }, .chained = true, - .chain_id = ALC662_FIXUP_BASS_CHMAP, + .chain_id = ALC668_FIXUP_ASUS_Nx51_HEADSET_MODE, }, [ALC891_FIXUP_HEADSET_MODE] = { .type = HDA_FIXUP_FUNC, -- 2.11.1

1 0

[alsa-devel] [PATCH v2] ASoC: rsnd: Fix possible NULL pointer dereference
by Kuninori Morimoto 21 Apr '17

21 Apr '17

From: Marek Vasut <marek.vasut+renesas(a)gmail.com> 25165f79adc76b812bfb4d8f2ab120aafb28d0e6 ("ASoC: rsnd: enable clock-frequency for both 44.1kHz/48kHz") supports both 44.1kHz/48kHz clock-frequency settings for ADG which will be used for AUDIO_OLKOUTn. But some board doesn't need it, thus, it is not mandatory. But, above patch didn't care about the case of "clock-frequency" DT property was not present. This patch ignores ADG settings if AUDIO_OLKOUTn was not used. Signed-off-by: Marek Vasut <marek.vasut+renesas(a)gmail.com> [Kuninori: tidyup not to break non AUDIO_OLKOUTn case] Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx(a)renesas.com> --- > Marek I think this patch solves your issue sound/soc/sh/rcar/adg.c | 18 +++++++++++------- 1 file changed, 11 insertions(+), 7 deletions(-) diff --git a/sound/soc/sh/rcar/adg.c b/sound/soc/sh/rcar/adg.c index 96fef91..277774c 100644 --- a/sound/soc/sh/rcar/adg.c +++ b/sound/soc/sh/rcar/adg.c @@ -453,13 +453,18 @@ static void rsnd_adg_get_clkout(struct rsnd_priv *priv, [CLKI] = 0x2, }; - of_property_read_u32(np, "#clock-cells", &count); + ckr = 0; + rbga = 2; /* default 1/6 */ + rbgb = 2; /* default 1/6 */ /* * ADG supports BRRA/BRRB output only * this means all clkout0/1/2/3 will be same rate */ prop = of_find_property(np, "clock-frequency", NULL);; + if (!prop) + goto rsnd_adg_get_clkout_end; + req_size = prop->length / sizeof(u32); of_property_read_u32_array(np, "clock-frequency", req_rate, req_size); @@ -472,6 +477,9 @@ static void rsnd_adg_get_clkout(struct rsnd_priv *priv, req_48kHz_rate = req_rate[i]; } + if (req_rate[0] % 48000 == 0) + adg->flags = AUDIO_OUT_48; + /* * This driver is assuming that AUDIO_CLKA/AUDIO_CLKB/AUDIO_CLKC * have 44.1kHz or 48kHz base clocks for now. @@ -481,9 +489,6 @@ static void rsnd_adg_get_clkout(struct rsnd_priv *priv, * rsnd_adg_ssi_clk_try_start() * rsnd_ssi_master_clk_start() */ - ckr = 0; - rbga = 2; /* default 1/6 */ - rbgb = 2; /* default 1/6 */ adg->rbga_rate_for_441khz = 0; adg->rbgb_rate_for_48khz = 0; for_each_rsnd_clk(clk, adg, i) { @@ -528,6 +533,7 @@ static void rsnd_adg_get_clkout(struct rsnd_priv *priv, * this means all clkout0/1/2/3 will be * same rate */ + of_property_read_u32(np, "#clock-cells", &count); /* * for clkout */ @@ -557,13 +563,11 @@ static void rsnd_adg_get_clkout(struct rsnd_priv *priv, &adg->onecell); } +rsnd_adg_get_clkout_end: adg->ckr = ckr; adg->rbga = rbga; adg->rbgb = rbgb; - if (req_rate[0] % 48000 == 0) - adg->flags = AUDIO_OUT_48; - for_each_rsnd_clkout(clk, adg, i) dev_dbg(dev, "clkout %d : %p : %ld\n", i, clk, clk_get_rate(clk)); dev_dbg(dev, "BRGCKR = 0x%08x, BRRA/BRRB = 0x%x/0x%x\n", -- 1.9.1

2 1

[alsa-devel] [PATCH] ASoC: rsnd: Fix possible NULL pointer dereference
by Marek Vasut 21 Apr '17

21 Apr '17

In case the "clock-frequency" DT property is not present, the of_find_property(np, "clock-frequency", NULL) will return NULL and the subsequent req_size = prop->length / sizeof(u32); will trigger a NULL pointer dereference. This patch adds check for the NULL return value and propagates the error into the caller of the function. Signed-off-by: Marek Vasut <marek.vasut+renesas(a)gmail.com> Cc: Geert Uytterhoeven <geert+renesas(a)glider.be> Cc: Hiroyuki Yokoyama <hiroyuki.yokoyama.vx(a)renesas.com> Cc: Kuninori Morimoto <kuninori.morimoto.gx(a)renesas.com> Cc: Mark Brown <broonie(a)kernel.org> Cc: linux-renesas-soc(a)vger.kernel.org --- sound/soc/sh/rcar/adg.c | 13 +++++++++---- 1 file changed, 9 insertions(+), 4 deletions(-) diff --git a/sound/soc/sh/rcar/adg.c b/sound/soc/sh/rcar/adg.c index 96fef91b480c..a7cbe7e89dfc 100644 --- a/sound/soc/sh/rcar/adg.c +++ b/sound/soc/sh/rcar/adg.c @@ -425,8 +425,7 @@ static void rsnd_adg_get_clkin(struct rsnd_priv *priv, dev_dbg(dev, "clk %d : %p : %ld\n", i, clk, clk_get_rate(clk)); } -static void rsnd_adg_get_clkout(struct rsnd_priv *priv, - struct rsnd_adg *adg) +static int rsnd_adg_get_clkout(struct rsnd_priv *priv, struct rsnd_adg *adg) { struct clk *clk; struct device *dev = rsnd_priv_to_dev(priv); @@ -459,7 +458,10 @@ static void rsnd_adg_get_clkout(struct rsnd_priv *priv, * ADG supports BRRA/BRRB output only * this means all clkout0/1/2/3 will be same rate */ - prop = of_find_property(np, "clock-frequency", NULL);; + prop = of_find_property(np, "clock-frequency", NULL); + if (!prop) + return -EINVAL; + req_size = prop->length / sizeof(u32); of_property_read_u32_array(np, "clock-frequency", req_rate, req_size); @@ -568,6 +570,7 @@ static void rsnd_adg_get_clkout(struct rsnd_priv *priv, dev_dbg(dev, "clkout %d : %p : %ld\n", i, clk, clk_get_rate(clk)); dev_dbg(dev, "BRGCKR = 0x%08x, BRRA/BRRB = 0x%x/0x%x\n", ckr, rbga, rbgb); + return 0; } int rsnd_adg_probe(struct rsnd_priv *priv) @@ -589,7 +592,9 @@ int rsnd_adg_probe(struct rsnd_priv *priv) return ret; rsnd_adg_get_clkin(priv, adg); - rsnd_adg_get_clkout(priv, adg); + ret = rsnd_adg_get_clkout(priv, adg); + if (ret) + return ret; if (of_get_property(np, "clkout-lr-asynchronous", NULL)) adg->flags = LRCLK_ASYNC; -- 2.11.0

2 2