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

[alsa-devel] [PATCH] ASoC: change mask in snd_soc_get/put_volsw_sx to unsigned int
by bgoswami＠codeaurora.org 03 Jan '18

03 Jan '18

From: Banajit Goswami <bgoswami(a)codeaurora.org> If the result of (min + max) is negative in functions snd_soc_get_volsw_sx() or snd_soc_put_volsw_sx(), there will be an overflow for the variable 'mask'. UBSAN: Undefined behaviour in sound/soc/soc-ops.c:382:6 signed integer overflow: -2147483648 - 1 cannot be represented in type 'int' Fix this by updating the variable type of 'mask' to unsigned int. Signed-off-by: Banajit Goswami <bgoswami(a)codeaurora.org> --- sound/soc/soc-ops.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/sound/soc/soc-ops.c b/sound/soc/soc-ops.c index 500f98c..7144a51 100644 --- a/sound/soc/soc-ops.c +++ b/sound/soc/soc-ops.c @@ -378,7 +378,7 @@ int snd_soc_get_volsw_sx(struct snd_kcontrol *kcontrol, unsigned int rshift = mc->rshift; int max = mc->max; int min = mc->min; - int mask = (1 << (fls(min + max) - 1)) - 1; + unsigned int mask = (1 << (fls(min + max) - 1)) - 1; unsigned int val; int ret; @@ -423,7 +423,7 @@ int snd_soc_put_volsw_sx(struct snd_kcontrol *kcontrol, unsigned int rshift = mc->rshift; int max = mc->max; int min = mc->min; - int mask = (1 << (fls(min + max) - 1)) - 1; + unsigned int mask = (1 << (fls(min + max) - 1)) - 1; int err = 0; unsigned int val, val_mask, val2 = 0; -- The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum, a Linux Foundation Collaborative Project

2 1

[alsa-devel] [PATCH] ASoC: skl: Fix kernel warning due to zero NHTL entry
by Takashi Iwai 03 Jan '18

03 Jan '18

I got the following kernel warning when loading snd-soc-skl module on Dell Latitude 7270 laptop: memremap attempted on mixed range 0x0000000000000000 size: 0x0 WARNING: CPU: 0 PID: 484 at kernel/memremap.c:98 memremap+0x8a/0x180 Call Trace: skl_nhlt_init+0x82/0xf0 [snd_soc_skl] skl_probe+0x2ee/0x7c0 [snd_soc_skl] .... It seems that the machine doesn't support the SKL DSP gives the empty NHLT entry, and it triggers the warning. For avoiding it, let do the zero check before calling memremap(). Cc: <stable(a)vger.kernel.org> Signed-off-by: Takashi Iwai <tiwai(a)suse.de> --- sound/soc/intel/skylake/skl-nhlt.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/sound/soc/intel/skylake/skl-nhlt.c b/sound/soc/intel/skylake/skl-nhlt.c index ba265119e0a1..9dd77383a4a7 100644 --- a/sound/soc/intel/skylake/skl-nhlt.c +++ b/sound/soc/intel/skylake/skl-nhlt.c @@ -44,7 +44,8 @@ struct nhlt_acpi_table *skl_nhlt_init(struct device *dev) obj = acpi_evaluate_dsm(handle, &osc_guid, 1, 1, NULL); if (obj && obj->type == ACPI_TYPE_BUFFER) { nhlt_ptr = (struct nhlt_resource_desc *)obj->buffer.pointer; - nhlt_table = (struct nhlt_acpi_table *) + if (nhlt_ptr->length) + nhlt_table = (struct nhlt_acpi_table *) memremap(nhlt_ptr->min_addr, nhlt_ptr->length, MEMREMAP_WB); ACPI_FREE(obj); -- 2.15.1

2 1

[alsa-devel] Fixing snd_pcm_indirect_playback_transfer to support rewinds?
by Floris Bos 03 Jan '18

03 Jan '18

Hi, I am been having problems using Pulse Audio on the Raspberry Pi for quite a while, in which audio stops playing if two programs want to play through Pulse simultaneously. Or if one program already has finished playing, but you start the second within 5 seconds. Thinking that was a problem in the snd_bcm2835 module, I created an issue in the github repo that houses the Raspberry Pi kernel fork about it in 2014: https://github.com/raspberrypi/linux/issues/688 However I now have the impression the problem is not in their module specific code, but that snd_pcm_indirect_playback_transfer() which the module uses, doesn't deal with rewinds correctly, which Pulse uses extensively. Is there any effort being done to fix this? I came across the "ALSA: pcm: Fix negative appl_ptr handling in pcm-indirect helpers" patch in this mailing list, so apparently the problem does is known to you guys. However looking at this fragment of the patch: == +static inline int snd_pcm_indirect_playback_transfer(struct snd_pcm_substream *substream, struct snd_pcm_indirect *rec, snd_pcm_indirect_copy_t copy) @@ -56,6 +56,8 @@ snd_pcm_indirect_playback_transfer(struct snd_pcm_substream *substream, if (diff) { if (diff < -(snd_pcm_sframes_t) (runtime->boundary / 2)) diff += runtime->boundary; + if (diff < 0) + return -EINVAL; == It seems it is simply denying rewinds instead of making them work? Isn't there any way to make them work, instead of disabling functionality userspace seems to be expecting to have? Yours sincerely, Floris Bos

2 3

[alsa-devel] [PATCH v2 1/2] ASoC: rt5645: change micbias widget type to supply.
by Hans de Goede 03 Jan '18

03 Jan '18

From: Bard Liao <bardliao(a)realtek.com> Register "micbias1" and "micbias2" to supply widgets as modern drivers do. This should not cause any (new) issues for existing users of the codec, since micbias support is broken anyways. Micbias support needs the RT5645_MICBIAS?_POW_CTRL_SEL bits in the RT5645_GEN_CTRL2 register to be updated when enabled/disabled which we currently do not do. The updating of these bits will be fixed in a follow-up commit. Signed-off-by: Bard Liao <bardliao(a)realtek.com> Signed-off-by: Hans de Goede <hdegoede(a)redhat.com> --- Changes in v2: -Update the commit message to explain why changing the type of the widget is ok / will not cause any (new) issues for existing users of the codec --- sound/soc/codecs/rt5645.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/sound/soc/codecs/rt5645.c b/sound/soc/codecs/rt5645.c index fd5c36afdbb5..fdab5bf56c61 100644 --- a/sound/soc/codecs/rt5645.c +++ b/sound/soc/codecs/rt5645.c @@ -1980,10 +1980,10 @@ static const struct snd_soc_dapm_widget rt5645_dapm_widgets[] = { /* Input Side */ /* micbias */ - SND_SOC_DAPM_MICBIAS("micbias1", RT5645_PWR_ANLG2, - RT5645_PWR_MB1_BIT, 0), - SND_SOC_DAPM_MICBIAS("micbias2", RT5645_PWR_ANLG2, - RT5645_PWR_MB2_BIT, 0), + SND_SOC_DAPM_SUPPLY("micbias1", RT5645_PWR_ANLG2, + RT5645_PWR_MB1_BIT, 0, NULL, 0), + SND_SOC_DAPM_SUPPLY("micbias2", RT5645_PWR_ANLG2, + RT5645_PWR_MB2_BIT, 0, NULL, 0), /* Input Lines */ SND_SOC_DAPM_INPUT("DMIC L1"), SND_SOC_DAPM_INPUT("DMIC R1"), -- 2.14.3

1 1

[alsa-devel] [PATCH] ASoC: Intel: Skylake: Disable clock gating during firmware and library download
by Guneshwor Singh 03 Jan '18

03 Jan '18

From: Pardha Saradhi K <pardha.saradhi.kesapragada(a)intel.com> During firmware and library download, sometimes it is observed that firmware and library download is timed-out resulting into probe failure. This patch disables dynamic clock gating while firmware and library download. Signed-off-by: Pardha Saradhi K <pardha.saradhi.kesapragada(a)intel.com> Signed-off-by: Sanyog Kale <sanyog.r.kale(a)intel.com> Signed-off-by: Guneshwor Singh <guneshwor.o.singh(a)intel.com> --- sound/soc/intel/skylake/skl-messages.c | 4 ++++ sound/soc/intel/skylake/skl-pcm.c | 4 ++++ 2 files changed, 8 insertions(+) diff --git a/sound/soc/intel/skylake/skl-messages.c b/sound/soc/intel/skylake/skl-messages.c index 81ee1118ed71..8cbf080c38b3 100644 --- a/sound/soc/intel/skylake/skl-messages.c +++ b/sound/soc/intel/skylake/skl-messages.c @@ -417,7 +417,11 @@ int skl_resume_dsp(struct skl *skl) if (skl->skl_sst->is_first_boot == true) return 0; + /* disable dynamic clock gating during fw and lib download */ + ctx->enable_miscbdcge(ctx->dev, false); + ret = skl_dsp_wake(ctx->dsp); + ctx->enable_miscbdcge(ctx->dev, true); if (ret < 0) return ret; diff --git a/sound/soc/intel/skylake/skl-pcm.c b/sound/soc/intel/skylake/skl-pcm.c index 8d0895b409b5..e46828533826 100644 --- a/sound/soc/intel/skylake/skl-pcm.c +++ b/sound/soc/intel/skylake/skl-pcm.c @@ -1343,7 +1343,11 @@ static int skl_platform_soc_probe(struct snd_soc_platform *platform) return -EIO; } + /* disable dynamic clock gating during fw and lib download */ + skl->skl_sst->enable_miscbdcge(platform->dev, false); + ret = ops->init_fw(platform->dev, skl->skl_sst); + skl->skl_sst->enable_miscbdcge(platform->dev, true); if (ret < 0) { dev_err(platform->dev, "Failed to boot first fw: %d\n", ret); return ret; -- 2.15.1

3 2

[alsa-devel] [PATCH v6 0/6] ASoC: Intel: Skylake: Add a clk driver to enable ssp clks early
by Sriram Periyasamy 03 Jan '18

03 Jan '18

For certain platforms, clocks (mclk/sclk/fs) are required to be up before the stream start. Example: some codecs needs the mclk/sclk/fs to be enabled early for a successful clock synchronization. Some platforms require clock to be enabled at boot and be always ON. By sending set_dma_control IPC (with the i2s blobs queried from NHLT), these clocks can be enabled early after the firmware is downloaded. With this series, a virtual clock driver is created which provides interface to send the required IPCs from machine driver to enable the clocks. NHLT is parsed during probe and the clock information is populated. The pointer to blob is cached and sent along with the set_dma_control IPC structure during the clk prepare/unprepare callback. Clocks are created for a ssp if the nhlt table has endpoint configuration for that particular ssp. Kabylake machine driver uses the clock interface to enable the clocks early as it is required by the rt5663 driver for clock synchronization. v5 -> v6 - Remove unwanted checks from clk ops - Modify recalc_rate to return just the cached rate and remove the rate calculations - Add license based on SPDX v4 -> v5 - Remove checks for clock enable status from machine driver since taken care in the framework already - Add check in the skl_clk_set_rate to avoid different rates when clock is enabled already v3 -> v4 - Add missing signed-offs v2 -> v3 - Moved the clk ops and IPCs from Skylake driver to clk driver and reordered commits accordingly - Add the support for extended I2S blob config which supports multiple mclk dividers in NHLT - Enable the clocks as well in DAPM PMU event instead of hw_params in machine drivers as confirmed by codec vendor - Do not register the clk if there is no valid clock source is avail in the I2S blob - Take care of error return in the clk driver - Address rest of the review comments and more optimization added - Fix the warning sound/soc/intel/skylake/skl.c:724:1-3: WARNING: PTR_ERR_OR_ZERO can be used reported by scripts/coccinelle/api/ptr_ret.cocci - Modified DSP replies as human readable to ease the debugging - Add firmware replies for MCLK/SCLK clocks if they are running already v1 -> v2 - Register parent clocks with skylake device. With the patch "clk: Add support for runtime PM" soon to be merged will help DSP to stay active on call to clock enable. Reference: (https://patchwork.kernel.org/patch/9911741/) - Fix the machine driver to enable clocks early for headphone playback path as well to fix a pop noise issue Harsha Priya (1): ASoC: Intel: kbl: Enable mclk and ssp sclk early Naveen M (1): ASoC: Intel: eve: Enable mclk and ssp sclk early Sriram Periyasamy (2): ASoC: Intel: Skylake: Add ssp clock driver ASoC: Intel: Skylake: Add extended I2S config blob support in Clock driver Subhransu S. Prusty (2): ASoC: Intel: Skylake: Make DSP replies more human readable ASoC: Intel: Skylake: Add FW reply for MCLK/SCLK IPC sound/soc/intel/Kconfig | 3 + sound/soc/intel/boards/Kconfig | 2 + sound/soc/intel/boards/kbl_rt5663_max98927.c | 95 ++++- .../soc/intel/boards/kbl_rt5663_rt5514_max98927.c | 94 +++++ sound/soc/intel/skylake/Makefile | 5 + sound/soc/intel/skylake/skl-i2s.h | 31 ++ sound/soc/intel/skylake/skl-messages.c | 1 + sound/soc/intel/skylake/skl-nhlt.c | 41 +- sound/soc/intel/skylake/skl-ssp-clk.c | 432 +++++++++++++++++++++ sound/soc/intel/skylake/skl-ssp-clk.h | 38 ++ sound/soc/intel/skylake/skl-sst-ipc.c | 50 ++- sound/soc/intel/skylake/skl.h | 6 + 12 files changed, 774 insertions(+), 24 deletions(-) create mode 100644 sound/soc/intel/skylake/skl-ssp-clk.c -- 2.7.4

4 10

[alsa-devel] [PATCH] slimbus: qcom: add HAS_IOMEM dependency
by srinivas.kandagatla＠linaro.org 03 Jan '18

03 Jan '18

From: Srinivas Kandagatla <srinivas.kandagatla(a)linaro.org> Below build failure was reported on UML, ERROR: "devm_ioremap_resource" [drivers/slimbus/slim-qcom-ctrl.ko] undefined! ERROR: "__ioread32_copy" [drivers/slimbus/slim-qcom-ctrl.ko] undefined! ERROR: "__iowrite32_copy" [drivers/slimbus/slim-qcom-ctrl.ko] undefined! ERROR: "devm_ioremap" [drivers/slimbus/slim-qcom-ctrl.ko] undefined! This patch fixes it by making qcom slimbus depend on HAS_IOMEM, as these are only defined when HAS_IOMEM is selected. Reported-by: Thomas Meyer <thomas(a)m3y3r.de> Signed-off-by: Srinivas Kandagatla <srinivas.kandagatla(a)linaro.org> --- drivers/slimbus/Kconfig | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/slimbus/Kconfig b/drivers/slimbus/Kconfig index 78bdd4808282..1a632fad597e 100644 --- a/drivers/slimbus/Kconfig +++ b/drivers/slimbus/Kconfig @@ -16,6 +16,7 @@ if SLIMBUS config SLIM_QCOM_CTRL tristate "Qualcomm SLIMbus Manager Component" depends on SLIMBUS + depends on HAS_IOMEM help Select driver if Qualcomm's SLIMbus Manager Component is programmed using Linux kernel. -- 2.15.0

1 0

[alsa-devel] [RFC v5 0/9] Enable HDA Codec support on Intel Platforms (Series2)
by Rakesh Ughreja 03 Jan '18

03 Jan '18

Many Intel platforms (SKL, KBL) etc. in the market supports enhanced audio capabilities which also includes DSP processing. This patch carry forwards the works that is done in the previous series to enable HD Audio codecs on such platforms. This patch series adds ASoC HDA codec driver for Intel platforms. It is written by reusing the legacy HDA ALSA codec driver. Intention is to maximize the reuse and minimize the changes in the legacy HDA codec driver. I would like to receive feedback before proceeding further on this direction. INFO: - This series is tested on KBL based product (Dell XPS 13). - Basic playback is working with headset and speakers. - HDMI playback is working. - Capture operation is not tested. - Playback using legacy drivers is tested. - More platforms and use cases coverage can be added once we have basic agreement in terms of the overall approach. - With this rebase I observed that if HDMI device is not connected to the machine, the iDisp codec is not detected and so the card is not created. work around is to make sure that HDMI device is connected at driver loading time. Changes in v5: - Rebased to Mark's tree, 4.15-rc5 - Fixed issue related to hdac_hda module loading order. Changes in v4: - new field bus type is added to distinguish bus type allocated by the controller or platform drivers. - extended routines are provided in hdac_bus which are called by legacy codec driver when the bus type is ext. - hdac_hda driver is converted into a passive library. Changes in v3: - snd_hda_codec_new API is split into two, so that it can be used by legacy as well as ASoC codec drivers. API is split into two to avoid bus registration when called by ASoC drivers. - Moved the __hda_codec_driver_register and hda_codec_driver_unregister APIs into generic driver and it performs registration for ASoC codec drivers also. So now there is only one kernel image for ASoC as well as legacy codec drivers. Registration functions had to moved to generic driver to resolve the circular dependency between hdac_hda and hda_codec drivers. - There is a Kconfig entry added to compile the ASoC version of HDA codec driver. For now only the Realtek codec entry is added. Same thing needs to be done for rest of the codec drivers also. Changes in v2: - Using Topology framework to create DAIs and DAI Links - Moved most commonly used functions into a separate file for machine driver. - Implemented separate Realtek ASoC HDA Driver driver for Realtek HDA codecs. It is just a driver registration wrapper which reuses the legacy HDA codec driver. This allows the removes the limitation that was present Rakesh Ughreja (9): ASoC: Intel: Boards: Machine driver for Intel platforms ASoC: Intel: Skylake: Add entry in sst_acpi_mach for HDA codecs ASoC: Intel: Skylake: add HDA BE DAIs ASoC: Intel: Skylake: use hda_bus instead of hdac_bus ALSA: hda - split snd_hda_codec_new function ALSA: hdac: remove memory allocation from snd_hdac_ext_bus_device_init ALSA: hdac: add bus type and extended ops support in hdac_bus ASoC: hdac_hda: add asoc extension for legacy HDA codec drivers ASoC: Intel: Boards: add support for HDA codecs include/sound/hdaudio.h | 24 ++ include/sound/hdaudio_ext.h | 6 +- sound/hda/ext/hdac_ext_bus.c | 14 +- sound/hda/hdac_bus.c | 1 + sound/pci/hda/hda_bind.c | 15 + sound/pci/hda/hda_codec.c | 67 +++- sound/pci/hda/hda_codec.h | 2 + sound/soc/codecs/Kconfig | 5 + sound/soc/codecs/Makefile | 2 + sound/soc/codecs/hdac_hda.c | 517 +++++++++++++++++++++++++++ sound/soc/codecs/hdac_hda.h | 20 ++ sound/soc/intel/Kconfig | 1 + sound/soc/intel/boards/Kconfig | 10 + sound/soc/intel/boards/Makefile | 2 + sound/soc/intel/boards/skl_hda_dsp_common.c | 163 +++++++++ sound/soc/intel/boards/skl_hda_dsp_common.h | 34 ++ sound/soc/intel/boards/skl_hda_dsp_generic.c | 132 +++++++ sound/soc/intel/skylake/skl-pcm.c | 32 +- sound/soc/intel/skylake/skl.c | 87 ++++- sound/soc/intel/skylake/skl.h | 10 +- 20 files changed, 1101 insertions(+), 43 deletions(-) create mode 100644 sound/soc/codecs/hdac_hda.c create mode 100644 sound/soc/codecs/hdac_hda.h create mode 100644 sound/soc/intel/boards/skl_hda_dsp_common.c create mode 100644 sound/soc/intel/boards/skl_hda_dsp_common.h create mode 100644 sound/soc/intel/boards/skl_hda_dsp_generic.c -- 2.7.4

1 9

[alsa-devel] [PATCH -next] soundwire: Fix typo in return value check of sdw_read()
by Wei Yongjun 03 Jan '18

03 Jan '18

Fix the typo, 'status' should be instead of 'status2'. Fixes: b0a9c37b0178 ("soundwire: Add slave status handling") Signed-off-by: Wei Yongjun <weiyongjun1(a)huawei.com> --- drivers/soundwire/bus.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/drivers/soundwire/bus.c b/drivers/soundwire/bus.c index 4c34519..266d2b3 100644 --- a/drivers/soundwire/bus.c +++ b/drivers/soundwire/bus.c @@ -671,8 +671,8 @@ static int sdw_handle_dp0_interrupt(struct sdw_slave *slave, u8 *slave_status) status2 = sdw_read(slave, SDW_DP0_INT); if (status2 < 0) { dev_err(slave->bus->dev, - "SDW_DP0_INT read failed:%d", status); - return status; + "SDW_DP0_INT read failed:%d", status2); + return status2; } status &= status2; @@ -741,10 +741,10 @@ static int sdw_handle_port_interrupt(struct sdw_slave *slave, /* Read DPN interrupt again */ status2 = sdw_read(slave, addr); - if (status < 0) { + if (status2 < 0) { dev_err(slave->bus->dev, - "SDW_DPN_INT read failed:%d", status); - return status; + "SDW_DPN_INT read failed:%d", status2); + return status2; } status &= status2;

3 2

[alsa-devel] Cherry Trail + RT5645 devices analog microphone not working.
by Hans de Goede 02 Jan '18

02 Jan '18

Hi Bard, I've been looking into getting the microphone to work on some Cherry Trail + RT5645 devices with what appears to be a simple analog microphone. You can see the microphone on one of the models I'm trying to get it to work on soldered onto the right side of the PCB here: https://i.imgur.com/bJzBQWe.jpg I'm using this ucm config: https://fedorapeople.org/~jwrdegoede/chtrt5645/ Which is basically Pierre-Louis Bossart's config from: https://github.com/plbossart/UCM Adjusted because the devices all have a mono speaker. I've been trying to get this to work myself and I can get some sound while recording by adding: cset "name='RECMIXL BST2 Switch' on" cset "name='RECMIXR BST2 Switch' on" But then I mainly get noise, I do get some sound from the microphone if I talk really really loudly, but it is barely above the noise-floor. I've the feeling that BST2 is only picking up the microphone via cross-talk and that BST1 indeed is the correct input, but for some reason I get complete silence when recording when only BST1 is selected in the RECMIXR, and yes I've checked / set the "IN1 boost volume". TL;DR: I'm trying to get an analog mic to work on 3 different models Cherry Trail + RT5645 devices and I've been unable to get this to work. I've put a registerdump of the rt5645 with gnome-sound-recorder running here: https://fedorapeople.org/~jwrdegoede/rt5645-mic-red.dump Any insights / help with this would be very much welcome. Regards, Hans

2 10