October 2021 - Alsa-devel - mailman.alsa-project.org

[PATCH] ALSA: usb-audio: allow -EPIPE errors for some v2 messages
by Greg Kroah-Hartman 14 Oct '21

14 Oct '21

The Schiit Hel device does not like to respond to all get_ctl_value_v2() requests for some reason. This used to work in older kernels, but now with more strict checking, this failure causes the device to fail to work. Cc: Jaroslav Kysela <perex(a)perex.cz> Cc: Takashi Iwai <tiwai(a)suse.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- This fixes the Shiit Hel device that I have. It used to work on older kernels (a year ago?), but stopped working for some reason and I didn't take the time to track it down. This change fixes the issue for me, but feels wrong for some reason. At least now I can use the device as a headphone driver, much better than the built-in one for my current machine... If needed, I can take the time to do bisection to track down the real issue here, it might be due to stricter endpoint checking in the USB core, but that feels wrong somehow. Here's the debugfs USB output for this device, showing the endpoints: T: Bus=07 Lev=02 Prnt=02 Port=01 Cnt=01 Dev#= 5 Spd=480 MxCh= 0 D: Ver= 2.00 Cls=ef(misc ) Sub=02 Prot=01 MxPS=64 #Cfgs= 1 P: Vendor=30be ProdID=0101 Rev=01.02 S: Manufacturer=Schiit Audio S: Product=Schiit Hel C: #Ifs= 4 Cfg#= 1 Atr=c0 MxPwr=0mA I: If#= 0 Alt= 0 #EPs= 1 Cls=01(audio) Sub=01 Prot=20 Driver=snd-usb-audio E: Ad=8f(I) Atr=03(Int.) MxPS= 6 Ivl=1ms I: If#= 1 Alt= 1 #EPs= 2 Cls=01(audio) Sub=02 Prot=20 Driver=snd-usb-audio E: Ad=05(O) Atr=05(Isoc) MxPS= 104 Ivl=125us E: Ad=85(I) Atr=11(Isoc) MxPS= 4 Ivl=1ms I: If#= 2 Alt= 1 #EPs= 1 Cls=01(audio) Sub=02 Prot=20 Driver=snd-usb-audio E: Ad=88(I) Atr=05(Isoc) MxPS= 156 Ivl=125us I: If#= 3 Alt= 0 #EPs= 1 Cls=03(HID ) Sub=00 Prot=00 Driver=usbhid E: Ad=84(I) Atr=03(Int.) MxPS= 16 Ivl=1ms Any other suggestions to fix this are welcome. sound/usb/mixer.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/sound/usb/mixer.c b/sound/usb/mixer.c index a2ce535df14b..37d3d697776b 100644 --- a/sound/usb/mixer.c +++ b/sound/usb/mixer.c @@ -371,11 +371,11 @@ static int get_ctl_value_v2(struct usb_mixer_elem_info *cval, int request, validx, idx, buf, size); snd_usb_unlock_shutdown(chip); - if (ret < 0) { + if ((ret < 0) && (ret != -EPIPE)) { error: usb_audio_err(chip, - "cannot get ctl value: req = %#x, wValue = %#x, wIndex = %#x, type = %d\n", - request, validx, idx, cval->val_type); + "cannot get ctl value: req = %#x, wValue = %#x, wIndex = %#x, type = %d, ret = %d\n", + request, validx, idx, cval->val_type, ret); return ret; } -- 2.33.0

2 11

[PATCH] ASoC: soc-compress: prevent the potentially use of null pointer
by Jiasheng Jiang 14 Oct '21

14 Oct '21

The codec_dai and the cpu_dai are from rtd. When the rtd->dais[0] or rtd->dais[rtd->num_cpus] is None, it could lead to the use of null pointer in the snd_soc_dai_stream_valid(). So it might be better to add the check before the use of snd_soc_dai_stream_valid(). Fixes: 467fece ("ASoC: soc-dai: move snd_soc_dai_stream_valid() to soc-dai.c") Signed-off-by: Jiasheng Jiang <jiasheng(a)iscas.ac.cn> --- sound/soc/soc-compress.c | 14 ++++++++------ 1 file changed, 8 insertions(+), 6 deletions(-) diff --git a/sound/soc/soc-compress.c b/sound/soc/soc-compress.c index b4f5935..67c3df1 100644 --- a/sound/soc/soc-compress.c +++ b/sound/soc/soc-compress.c @@ -535,12 +535,14 @@ int snd_soc_new_compress(struct snd_soc_pcm_runtime *rtd, int num) } /* check client and interface hw capabilities */ - if (snd_soc_dai_stream_valid(codec_dai, SNDRV_PCM_STREAM_PLAYBACK) && - snd_soc_dai_stream_valid(cpu_dai, SNDRV_PCM_STREAM_PLAYBACK)) - playback = 1; - if (snd_soc_dai_stream_valid(codec_dai, SNDRV_PCM_STREAM_CAPTURE) && - snd_soc_dai_stream_valid(cpu_dai, SNDRV_PCM_STREAM_CAPTURE)) - capture = 1; + if (codec_dai && cpu_dai) { + if (snd_soc_dai_stream_valid(codec_dai, SNDRV_PCM_STREAM_PLAYBACK) && + snd_soc_dai_stream_valid(cpu_dai, SNDRV_PCM_STREAM_PLAYBACK)) + playback = 1; + if (snd_soc_dai_stream_valid(codec_dai, SNDRV_PCM_STREAM_CAPTURE) && + snd_soc_dai_stream_valid(cpu_dai, SNDRV_PCM_STREAM_CAPTURE)) + capture = 1; + } /* * Compress devices are unidirectional so only one of the directions -- 2.7.4

2 1

[PATCH] ASoC: dt-bindings: rockchip: i2s-tdm: Fix rockchip, i2s-[rt]x-route
by Geert Uytterhoeven 13 Oct '21

13 Oct '21

make dt_binding_check: Documentation/devicetree/bindings/sound/rockchip,i2s-tdm.yaml: properties:rockchip,i2s-rx-route: {'required': ['maxItems']} is not allowed for {'$ref': '/schemas/types.yaml#/definitions/uint32-array', 'description': 'Defines the mapping of I2S RX sdis to I2S data bus lines. By default, they are mapped one-to-one. rockchip,i2s-rx-route = <3> would mean sdi3 is receiving from data0.', 'maxItems': 4, 'items': [{'enum': [0, 1, 2, 3]}]} hint: "maxItems" is not needed with an "items" list from schema $id: http://devicetree.org/meta-schemas/items.yaml# Documentation/devicetree/bindings/sound/rockchip,i2s-tdm.yaml: properties:rockchip,i2s-tx-route: {'required': ['maxItems']} is not allowed for {'$ref': '/schemas/types.yaml#/definitions/uint32-array', 'description': 'Defines the mapping of I2S TX sdos to I2S data bus lines. By default, they are mapped one-to-one. rockchip,i2s-tx-route = <3> would mean sdo3 is sending to data0.', 'maxItems': 4, 'items': [{'enum': [0, 1, 2, 3]}]} hint: "maxItems" is not needed with an "items" list from schema $id: http://devicetree.org/meta-schemas/items.yaml# Drop the "-" from the enums to fix this. Fixes: 510f1c133aedcf69 ("ASoC: dt-bindings: rockchip: add i2s-tdm bindings") Signed-off-by: Geert Uytterhoeven <geert+renesas(a)glider.be> --- v4/v5 differed slightly from what Rob suggested in his review of v3 https://lore.kernel.org/all/YS6kHkeOKCJMOOIf@robh.at.kernel.org/ --- Documentation/devicetree/bindings/sound/rockchip,i2s-tdm.yaml | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/Documentation/devicetree/bindings/sound/rockchip,i2s-tdm.yaml b/Documentation/devicetree/bindings/sound/rockchip,i2s-tdm.yaml index dce8b4136ec6db4e..ce3e18b50230dc9b 100644 --- a/Documentation/devicetree/bindings/sound/rockchip,i2s-tdm.yaml +++ b/Documentation/devicetree/bindings/sound/rockchip,i2s-tdm.yaml @@ -112,7 +112,7 @@ properties: rockchip,i2s-rx-route = <3> would mean sdi3 is receiving from data0. maxItems: 4 items: - - enum: [0, 1, 2, 3] + enum: [0, 1, 2, 3] rockchip,i2s-tx-route: $ref: /schemas/types.yaml#/definitions/uint32-array @@ -122,7 +122,7 @@ properties: rockchip,i2s-tx-route = <3> would mean sdo3 is sending to data0. maxItems: 4 items: - - enum: [0, 1, 2, 3] + enum: [0, 1, 2, 3] rockchip,io-multiplex: description: -- 2.25.1

2 1

[PATCH v2] ASoC: rt1011: Fix 'I2S Reference' enum control
by Peter Ujfalusi 13 Oct '21

13 Oct '21

There are several things the patch adding the support for 'I2S Reference' got wrong: - "None" selection is in fact equals to last selected reference - The custom put overrides RX/TX len, TDM slot sizes, etc - the enum is useless in most part for the reference tracking - there is no need for EXT control as there is a single bit in RT1011_TDM1_SET_1 register (bit 7) which selects the reference - it was using ucontrol->value.integer.value[0] in the put/get callbacks which causesed access to 'I2S Reference' enum with alsamixer to fail Complements: c3de683c4d1d ("ASoC: rt1011: Fix 'I2S Reference' enum control caused error") Signed-off-by: Peter Ujfalusi <peter.ujfalusi(a)linux.intel.com> --- Hi, Changes since v1: - Rebased on top of next - changed the tag from Fixes to Complements to make sure that the crash fix and the fix for the implementation stays together. This patch is an iteration on https://lore.kernel.org/alsa-devel/20211011144518.2518-1-peter.ujfalusi@lin… to fix 87f40af26c262 ("ASoC: rt1011: add i2s reference control for rt1011"). In essence it is reverting the original patch to use SOC_ENUM_SINGLE_DECL/SOC_ENUM to handle the bit to select the I2S reference. Regards, Peter sound/soc/codecs/rt1011.c | 53 ++++----------------------------------- sound/soc/codecs/rt1011.h | 7 ------ 2 files changed, 5 insertions(+), 55 deletions(-) diff --git a/sound/soc/codecs/rt1011.c b/sound/soc/codecs/rt1011.c index bdfcbb81fa19..297af7ff824c 100644 --- a/sound/soc/codecs/rt1011.c +++ b/sound/soc/codecs/rt1011.c @@ -1312,54 +1312,12 @@ static int rt1011_r0_load_info(struct snd_kcontrol *kcontrol, } static const char * const rt1011_i2s_ref_texts[] = { - "None", "Left Channel", "Right Channel" + "Left Channel", "Right Channel" }; -static SOC_ENUM_SINGLE_EXT_DECL(rt1011_i2s_ref_enum, rt1011_i2s_ref_texts); - -static int rt1011_i2s_ref_put(struct snd_kcontrol *kcontrol, - struct snd_ctl_elem_value *ucontrol) -{ - struct snd_soc_component *component = - snd_soc_kcontrol_component(kcontrol); - struct rt1011_priv *rt1011 = - snd_soc_component_get_drvdata(component); - int i2s_ref_ch = ucontrol->value.enumerated.item[0]; - - switch (i2s_ref_ch) { - case RT1011_I2S_REF_LEFT_CH: - regmap_write(rt1011->regmap, RT1011_TDM_TOTAL_SET, 0x0240); - regmap_write(rt1011->regmap, RT1011_TDM1_SET_2, 0x8); - regmap_write(rt1011->regmap, RT1011_TDM1_SET_1, 0x1022); - regmap_write(rt1011->regmap, RT1011_ADCDAT_OUT_SOURCE, 0x4); - break; - case RT1011_I2S_REF_RIGHT_CH: - regmap_write(rt1011->regmap, RT1011_TDM_TOTAL_SET, 0x0240); - regmap_write(rt1011->regmap, RT1011_TDM1_SET_2, 0x8); - regmap_write(rt1011->regmap, RT1011_TDM1_SET_1, 0x10a2); - regmap_write(rt1011->regmap, RT1011_ADCDAT_OUT_SOURCE, 0x4); - break; - default: - dev_info(component->dev, "I2S Reference: Do nothing\n"); - } - - rt1011->i2s_ref = ucontrol->value.enumerated.item[0]; - - return 0; -} - -static int rt1011_i2s_ref_get(struct snd_kcontrol *kcontrol, - struct snd_ctl_elem_value *ucontrol) -{ - struct snd_soc_component *component = - snd_soc_kcontrol_component(kcontrol); - struct rt1011_priv *rt1011 = - snd_soc_component_get_drvdata(component); - - ucontrol->value.enumerated.item[0] = rt1011->i2s_ref; - - return 0; -} +static SOC_ENUM_SINGLE_DECL(rt1011_i2s_ref_enum, + RT1011_TDM1_SET_1, 7, + rt1011_i2s_ref_texts); static const struct snd_kcontrol_new rt1011_snd_controls[] = { /* I2S Data In Selection */ @@ -1400,8 +1358,7 @@ static const struct snd_kcontrol_new rt1011_snd_controls[] = { SOC_SINGLE("R0 Temperature", RT1011_STP_INITIAL_RESISTANCE_TEMP, 2, 255, 0), /* I2S Reference */ - SOC_ENUM_EXT("I2S Reference", rt1011_i2s_ref_enum, - rt1011_i2s_ref_get, rt1011_i2s_ref_put), + SOC_ENUM("I2S Reference", rt1011_i2s_ref_enum), }; static int rt1011_is_sys_clk_from_pll(struct snd_soc_dapm_widget *source, diff --git a/sound/soc/codecs/rt1011.h b/sound/soc/codecs/rt1011.h index afb2fad94216..68fadc15fa8c 100644 --- a/sound/soc/codecs/rt1011.h +++ b/sound/soc/codecs/rt1011.h @@ -654,12 +654,6 @@ enum { RT1011_AIFS }; -enum { - RT1011_I2S_REF_NONE, - RT1011_I2S_REF_LEFT_CH, - RT1011_I2S_REF_RIGHT_CH, -}; - /* BiQual & DRC related settings */ #define RT1011_BQ_DRC_NUM 128 struct rt1011_bq_drc_params { @@ -698,7 +692,6 @@ struct rt1011_priv { unsigned int r0_reg, cali_done; unsigned int r0_calib, temperature_calib; int recv_spk_mode; - unsigned int i2s_ref; }; #endif /* end of _RT1011_H_ */ -- 2.33.0

2 1

[PATCH] ASoC: rt1011: Fix 'I2S Reference' enum control
by Peter Ujfalusi 13 Oct '21

13 Oct '21

There are several things the patch adding the support for 'I2S Reference' got wrong: - "None" selection is in fact equals to last selected reference - The custom put overrides RX/TX len, TDM slot sizes, etc - the enum is useless in most part for the reference tracking - there is no need for EXT control as there is a single bit in RT1011_TDM1_SET_1 register (bit 7) which selects the reference - it was using ucontrol->value.integer.value[0] in the put/get callbacks which causes access to 'I2S Reference' enum with alsamixer to fail: $ alsamixer cannot load mixer controls: Invalid argument cml_rt1011_rt5682 cml_rt1011_rt5682: control 2:0:0:TL I2S Reference:0: access overflow It should have used ucontrol->value.enumerated.item[0], but since there is no need for the custom code, it does not really matter. Fixes: 87f40af26c262 ("ASoC: rt1011: add i2s reference control for rt1011") Reported-by: Seppo Ingalsuo <seppo.ingalsuo(a)linux.intel.com> Signed-off-by: Peter Ujfalusi <peter.ujfalusi(a)linux.intel.com> --- Hi, This patch is an iteration on https://lore.kernel.org/alsa-devel/20211011144518.2518-1-peter.ujfalusi@lin… to fix 87f40af26c262 ("ASoC: rt1011: add i2s reference control for rt1011"). In essence it is reverting the original patch to use SOC_ENUM_SINGLE_DECL/SOC_ENUM to handle the bit to select the I2S reference. Regards, Peter sound/soc/codecs/rt1011.c | 56 +++++---------------------------------- sound/soc/codecs/rt1011.h | 7 ----- 2 files changed, 6 insertions(+), 57 deletions(-) diff --git a/sound/soc/codecs/rt1011.c b/sound/soc/codecs/rt1011.c index 508597866dff..297af7ff824c 100644 --- a/sound/soc/codecs/rt1011.c +++ b/sound/soc/codecs/rt1011.c @@ -1311,56 +1311,13 @@ static int rt1011_r0_load_info(struct snd_kcontrol *kcontrol, .put = rt1011_r0_load_mode_put \ } -static const char * const rt1011_i2s_ref[] = { - "None", "Left Channel", "Right Channel" +static const char * const rt1011_i2s_ref_texts[] = { + "Left Channel", "Right Channel" }; -static SOC_ENUM_SINGLE_DECL(rt1011_i2s_ref_enum, 0, 0, - rt1011_i2s_ref); - -static int rt1011_i2s_ref_put(struct snd_kcontrol *kcontrol, - struct snd_ctl_elem_value *ucontrol) -{ - struct snd_soc_component *component = - snd_soc_kcontrol_component(kcontrol); - struct rt1011_priv *rt1011 = - snd_soc_component_get_drvdata(component); - int i2s_ref_ch = ucontrol->value.integer.value[0]; - - switch (i2s_ref_ch) { - case RT1011_I2S_REF_LEFT_CH: - regmap_write(rt1011->regmap, RT1011_TDM_TOTAL_SET, 0x0240); - regmap_write(rt1011->regmap, RT1011_TDM1_SET_2, 0x8); - regmap_write(rt1011->regmap, RT1011_TDM1_SET_1, 0x1022); - regmap_write(rt1011->regmap, RT1011_ADCDAT_OUT_SOURCE, 0x4); - break; - case RT1011_I2S_REF_RIGHT_CH: - regmap_write(rt1011->regmap, RT1011_TDM_TOTAL_SET, 0x0240); - regmap_write(rt1011->regmap, RT1011_TDM1_SET_2, 0x8); - regmap_write(rt1011->regmap, RT1011_TDM1_SET_1, 0x10a2); - regmap_write(rt1011->regmap, RT1011_ADCDAT_OUT_SOURCE, 0x4); - break; - default: - dev_info(component->dev, "I2S Reference: Do nothing\n"); - } - - rt1011->i2s_ref = ucontrol->value.integer.value[0]; - - return 0; -} - -static int rt1011_i2s_ref_get(struct snd_kcontrol *kcontrol, - struct snd_ctl_elem_value *ucontrol) -{ - struct snd_soc_component *component = - snd_soc_kcontrol_component(kcontrol); - struct rt1011_priv *rt1011 = - snd_soc_component_get_drvdata(component); - - ucontrol->value.integer.value[0] = rt1011->i2s_ref; - - return 0; -} +static SOC_ENUM_SINGLE_DECL(rt1011_i2s_ref_enum, + RT1011_TDM1_SET_1, 7, + rt1011_i2s_ref_texts); static const struct snd_kcontrol_new rt1011_snd_controls[] = { /* I2S Data In Selection */ @@ -1401,8 +1358,7 @@ static const struct snd_kcontrol_new rt1011_snd_controls[] = { SOC_SINGLE("R0 Temperature", RT1011_STP_INITIAL_RESISTANCE_TEMP, 2, 255, 0), /* I2S Reference */ - SOC_ENUM_EXT("I2S Reference", rt1011_i2s_ref_enum, - rt1011_i2s_ref_get, rt1011_i2s_ref_put), + SOC_ENUM("I2S Reference", rt1011_i2s_ref_enum), }; static int rt1011_is_sys_clk_from_pll(struct snd_soc_dapm_widget *source, diff --git a/sound/soc/codecs/rt1011.h b/sound/soc/codecs/rt1011.h index afb2fad94216..68fadc15fa8c 100644 --- a/sound/soc/codecs/rt1011.h +++ b/sound/soc/codecs/rt1011.h @@ -654,12 +654,6 @@ enum { RT1011_AIFS }; -enum { - RT1011_I2S_REF_NONE, - RT1011_I2S_REF_LEFT_CH, - RT1011_I2S_REF_RIGHT_CH, -}; - /* BiQual & DRC related settings */ #define RT1011_BQ_DRC_NUM 128 struct rt1011_bq_drc_params { @@ -698,7 +692,6 @@ struct rt1011_priv { unsigned int r0_reg, cali_done; unsigned int r0_calib, temperature_calib; int recv_spk_mode; - unsigned int i2s_ref; }; #endif /* end of _RT1011_H_ */ -- 2.33.0

2 1

[PATCH v4 0/2] ASoC: rt9120: Add Richtek RT9120 supprot
by cy_huang 13 Oct '21

13 Oct '21

From: ChiYuan Huang <cy_huang(a)richtek.com> This patch series Add the Richtek RT9120 support. In v4: - Add 'classd_tlv' for 'SPK Gain Volume' control item. - Unify the tlv declaration to the postfix '_tlv'. - Fix 'digital_tlv' mute as 1 to declare the minimum is muted. In v3: - Add dvdd regulator binding to check the dvdd voltage domain. - Refine sdo_select_text. - Use switch case in 'internal_power_event' function. - Remove the volume and mute initially write in component probe. - Remove the mute API. It's no need by HW design. In v2: - Add missing #sound-dai-cells property. ChiYuan Huang (2): ASoC: dt-bindings: rt9120: Add initial bindings ASoC: rt9120: Add rt9210 audio amplifier support .../devicetree/bindings/sound/richtek,rt9120.yaml | 59 +++ sound/soc/codecs/Kconfig | 10 + sound/soc/codecs/Makefile | 2 + sound/soc/codecs/rt9120.c | 495 +++++++++++++++++++++ 4 files changed, 566 insertions(+) create mode 100644 Documentation/devicetree/bindings/sound/richtek,rt9120.yaml create mode 100644 sound/soc/codecs/rt9120.c -- 2.7.4

2 3

[PATCH v2] ASoC: wm8960: Fix clock configuration on slave mode
by Shengjiu Wang 13 Oct '21

13 Oct '21

There is a noise issue for 8kHz sample rate on slave mode. Compared with master mode, the difference is the DACDIV setting, after correcting the DACDIV, the noise is gone. There is no noise issue for 48kHz sample rate, because the default value of DACDIV is correct for 48kHz. So wm8960_configure_clocking() should be functional for ADC and DAC function even if it is slave mode. In order to be compatible for old use case, just add condition for checking that sysclk is zero with slave mode. Fixes: 0e50b51aa22f ("ASoC: wm8960: Let wm8960 driver configure its bit clock and frame clock") Signed-off-by: Shengjiu Wang <shengjiu.wang(a)nxp.com> --- changes in v2: - change error message to warning - add more comments sound/soc/codecs/wm8960.c | 13 ++++++++++--- 1 file changed, 10 insertions(+), 3 deletions(-) diff --git a/sound/soc/codecs/wm8960.c b/sound/soc/codecs/wm8960.c index 9e621a254392..499604f1e178 100644 --- a/sound/soc/codecs/wm8960.c +++ b/sound/soc/codecs/wm8960.c @@ -742,9 +742,16 @@ static int wm8960_configure_clocking(struct snd_soc_component *component) int i, j, k; int ret; - if (!(iface1 & (1<<6))) { - dev_dbg(component->dev, - "Codec is slave mode, no need to configure clock\n"); + /* + * For Slave mode clocking should still be configured, + * so this if statement should be removed, but some platform + * may not work if the sysclk is not configured, to avoid such + * compatible issue, just add '!wm8960->sysclk' condition in + * this if statement. + */ + if (!(iface1 & (1 << 6)) && !wm8960->sysclk) { + dev_warn(component->dev, + "slave mode, but proceeding with no clock configuration\n"); return 0; } -- 2.17.1

3 2

[PATCH] ASoC: amd: vangogh: constify static struct snd_soc_dai_ops
by Rikard Falkeborn 13 Oct '21

13 Oct '21

The only usage of acp5x_i2s_dai_ops is to assign its address to the ops field in the snd_soc_dai_driver struct, which is a pointer to const. Make it const to allow the compiler to put it in read-only memory. Signed-off-by: Rikard Falkeborn <rikard.falkeborn(a)gmail.com> --- sound/soc/amd/vangogh/acp5x-i2s.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/sound/soc/amd/vangogh/acp5x-i2s.c b/sound/soc/amd/vangogh/acp5x-i2s.c index 2705e57eb713..002db3971ca9 100644 --- a/sound/soc/amd/vangogh/acp5x-i2s.c +++ b/sound/soc/amd/vangogh/acp5x-i2s.c @@ -348,7 +348,7 @@ static int acp5x_i2s_trigger(struct snd_pcm_substream *substream, return ret; } -static struct snd_soc_dai_ops acp5x_i2s_dai_ops = { +static const struct snd_soc_dai_ops acp5x_i2s_dai_ops = { .hw_params = acp5x_i2s_hwparams, .trigger = acp5x_i2s_trigger, .set_fmt = acp5x_i2s_set_fmt, -- 2.33.0

2 1

[PATCH] ASoC: mediatek: Constify static snd_soc_ops
by Rikard Falkeborn 13 Oct '21

13 Oct '21

These are only assigned to the ops field in the snd_soc_dai_link struct which is a pointer to const struct snd_soc_ops. Make them const to allow the compiler to put them in read-only memory. Signed-off-by: Rikard Falkeborn <rikard.falkeborn(a)gmail.com> --- sound/soc/mediatek/mt2701/mt2701-cs42448.c | 2 +- sound/soc/mediatek/mt2701/mt2701-wm8960.c | 2 +- sound/soc/mediatek/mt8183/mt8183-mt6358-ts3a227-max98357.c | 2 +- 3 files changed, 3 insertions(+), 3 deletions(-) diff --git a/sound/soc/mediatek/mt2701/mt2701-cs42448.c b/sound/soc/mediatek/mt2701/mt2701-cs42448.c index 44a8d5cfb0aa..d9fd6eb786aa 100644 --- a/sound/soc/mediatek/mt2701/mt2701-cs42448.c +++ b/sound/soc/mediatek/mt2701/mt2701-cs42448.c @@ -146,7 +146,7 @@ static int mt2701_cs42448_be_ops_hw_params(struct snd_pcm_substream *substream, return 0; } -static struct snd_soc_ops mt2701_cs42448_be_ops = { +static const struct snd_soc_ops mt2701_cs42448_be_ops = { .hw_params = mt2701_cs42448_be_ops_hw_params }; diff --git a/sound/soc/mediatek/mt2701/mt2701-wm8960.c b/sound/soc/mediatek/mt2701/mt2701-wm8960.c index 414e422c0eba..f56de1b918bf 100644 --- a/sound/soc/mediatek/mt2701/mt2701-wm8960.c +++ b/sound/soc/mediatek/mt2701/mt2701-wm8960.c @@ -40,7 +40,7 @@ static int mt2701_wm8960_be_ops_hw_params(struct snd_pcm_substream *substream, return 0; } -static struct snd_soc_ops mt2701_wm8960_be_ops = { +static const struct snd_soc_ops mt2701_wm8960_be_ops = { .hw_params = mt2701_wm8960_be_ops_hw_params }; diff --git a/sound/soc/mediatek/mt8183/mt8183-mt6358-ts3a227-max98357.c b/sound/soc/mediatek/mt8183/mt8183-mt6358-ts3a227-max98357.c index 94dcbd36c869..aeb1af86047e 100644 --- a/sound/soc/mediatek/mt8183/mt8183-mt6358-ts3a227-max98357.c +++ b/sound/soc/mediatek/mt8183/mt8183-mt6358-ts3a227-max98357.c @@ -335,7 +335,7 @@ static void mt8183_mt6358_tdm_shutdown(struct snd_pcm_substream *substream) __func__, ret); } -static struct snd_soc_ops mt8183_mt6358_tdm_ops = { +static const struct snd_soc_ops mt8183_mt6358_tdm_ops = { .startup = mt8183_mt6358_tdm_startup, .shutdown = mt8183_mt6358_tdm_shutdown, }; -- 2.33.0

2 1

[PATCH v3] component: do not leave master devres group open after bind
by Kai Vehmanen 13 Oct '21

13 Oct '21

In current code, the devres group for aggregate master is left open after call to component_master_add_*(). This leads to problems when the master does further managed allocations on its own. When any participating driver calls component_del(), this leads to immediate release of resources. This came up when investigating a page fault occurring with i915 DRM driver unbind with 5.15-rc1 kernel. The following sequence occurs: i915_pci_remove() -> intel_display_driver_unregister() -> i915_audio_component_cleanup() -> component_del() -> component.c:take_down_master() -> hdac_component_master_unbind() [via master->ops->unbind()] -> devres_release_group(master->parent, NULL) With older kernels this has not caused issues, but with audio driver moving to use managed interfaces for more of its allocations, this no longer works. Devres log shows following to occur: component_master_add_with_match() [ 126.886032] snd_hda_intel 0000:00:1f.3: DEVRES ADD 00000000323ccdc5 devm_component_match_release (24 bytes) [ 126.886045] snd_hda_intel 0000:00:1f.3: DEVRES ADD 00000000865cdb29 grp< (0 bytes) [ 126.886049] snd_hda_intel 0000:00:1f.3: DEVRES ADD 000000001b480725 grp< (0 bytes) audio driver completes its PCI probe() [ 126.892238] snd_hda_intel 0000:00:1f.3: DEVRES ADD 000000001b480725 pcim_iomap_release (48 bytes) component_del() called() at DRM/i915 unbind() [ 137.579422] i915 0000:00:02.0: DEVRES REL 00000000ef44c293 grp< (0 bytes) [ 137.579445] snd_hda_intel 0000:00:1f.3: DEVRES REL 00000000865cdb29 grp< (0 bytes) [ 137.579458] snd_hda_intel 0000:00:1f.3: DEVRES REL 000000001b480725 pcim_iomap_release (48 bytes) So the "devres_release_group(master->parent, NULL)" ends up freeing the pcim_iomap allocation. Upon next runtime resume, the audio driver will cause a page fault as the iomap alloc was released without the driver knowing about it. Fix this issue by using the "struct master" pointer as identifier for the devres group, and by closing the devres group after the master->ops->bind() call is done. This allows devres allocations done by the driver acting as master to be isolated from the binding state of the aggregate driver. This modifies the logic originally introduced in commit 9e1ccb4a7700 ("drivers/base: fix devres handling for master device") Cc: stable(a)vger.kernel.org BugLink: https://gitlab.freedesktop.org/drm/intel/-/issues/4136 Fixes: 9e1ccb4a7700 ("drivers/base: fix devres handling for master device") Signed-off-by: Kai Vehmanen <kai.vehmanen(a)linux.intel.com> Acked-by: Imre Deak <imre.deak(a)intel.com> Acked-by: Russell King (Oracle) <rmk+kernel(a)armlinux.org.uk> --- drivers/base/component.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) V3 changes: - address feedback from Greg KH, add a Fixes tag and cc stable V2 changes: - after review form Imre and Russell, removing RFC tag - rebased on top of 5.15-rc2 (V1 was on drm-tip) - CI test results for V1 show that this patch fixes multiple failures in i915 unbind and module reload tests: https://patchwork.freedesktop.org/series/94889/ diff --git a/drivers/base/component.c b/drivers/base/component.c index 5e79299f6c3f..870485cbbb87 100644 --- a/drivers/base/component.c +++ b/drivers/base/component.c @@ -246,7 +246,7 @@ static int try_to_bring_up_master(struct master *master, return 0; } - if (!devres_open_group(master->parent, NULL, GFP_KERNEL)) + if (!devres_open_group(master->parent, master, GFP_KERNEL)) return -ENOMEM; /* Found all components */ @@ -258,6 +258,7 @@ static int try_to_bring_up_master(struct master *master, return ret; } + devres_close_group(master->parent, NULL); master->bound = true; return 1; } @@ -282,7 +283,7 @@ static void take_down_master(struct master *master) { if (master->bound) { master->ops->unbind(master->parent); - devres_release_group(master->parent, NULL); + devres_release_group(master->parent, master); master->bound = false; } } base-commit: 9e1ff307c779ce1f0f810c7ecce3d95bbae40896 -- 2.33.0

1 0