July 2026 - Sound-open-firmware - mailman.alsa-project.org

[PATCH 1/2] ASoC: SOF: don't use "/**" for non-kernel-doc comments
by Randy Dunlap 28 Jul '26

28 Jul '26

Modify these errant comments to use "/*" since they are not kernel-doc comments. Warning: ../include/sound/sof/header.h:182 This comment starts with '/**', but isn't a kernel-doc comment. * OOPS header architecture specific data. Warning: ../include/sound/sof/header.h:190 This comment starts with '/**', but isn't a kernel-doc comment. * OOPS header platform specific data. Signed-off-by: Randy Dunlap <rdunlap(a)infradead.org> --- Cc: Liam Girdwood <lgirdwood(a)gmail.com> Cc: Peter Ujfalusi <peter.ujfalusi(a)linux.intel.com> Cc: Bard Liao <yung-chuan.liao(a)linux.intel.com> Cc: Daniel Baluta <daniel.baluta(a)nxp.com> Cc: Kai Vehmanen <kai.vehmanen(a)linux.intel.com> Cc: Pierre-Louis Bossart <pierre-louis.bossart(a)linux.dev> Cc: Vijendar Mukunda <Vijendar.Mukunda(a)amd.com> Cc: sound-open-firmware(a)alsa-project.org include/sound/sof/header.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) --- linext-2026-0710.orig/include/sound/sof/header.h +++ linext-2026-0710/include/sound/sof/header.h @@ -179,7 +179,7 @@ struct sof_ipc_compound_hdr { uint32_t count; /**< count of 0 means end of compound sequence */ } __packed; -/** +/* * OOPS header architecture specific data. */ struct sof_ipc_dsp_oops_arch_hdr { @@ -187,7 +187,7 @@ struct sof_ipc_dsp_oops_arch_hdr { uint32_t totalsize; /* Total size of oops message */ } __packed; -/** +/* * OOPS header platform specific data. */ struct sof_ipc_dsp_oops_plat_hdr {

2 2

[PATCH v2 0/4] ASoC: remove conditional returns with no effect
by Sang-Heon Jeon 27 Jul '26

27 Jul '26

Hello, This series removes conditional returns where both branches return the same value, so the check has no effect. As suggested by Vijendar, the ASoC patch of the treewide v1 series [1] is split per vendor and posted separately. All patches are generated by the Coccinelle script, which you can find in v1. If you would rather drop any of the patches, just let me know and I will address it in the next version. --- Changes from v1 [1] - Split patches per vendor - Change base to broonie/sound/for-next - Filter only ASoC patch [1] https://lore.kernel.org/all/20260723184538.3888637-1-ekffu200098@gmail.com/ --- Sang-Heon Jeon (4): ASoC: amd: acp: remove conditional return with no effect ASoC: Intel: atom: remove conditional return with no effect ASoC: samsung: smdk_spdif: remove conditional return with no effect ASoC: SOF: Intel: remove conditional return with no effect sound/soc/amd/acp/acp-mach-common.c | 11 +++-------- sound/soc/intel/atom/sst-mfld-platform-pcm.c | 7 +------ sound/soc/samsung/smdk_spdif.c | 8 ++------ sound/soc/sof/intel/hda-dsp.c | 6 +----- 4 files changed, 7 insertions(+), 25 deletions(-) -- 2.43.0

2 2

[PATCH 14/36] ASoC: remove conditional return with no effect
by Sang-Heon Jeon 24 Jul '26

24 Jul '26

Both branches of the check return the same value, so the check has no effect. Remove it and return the value directly. This is the result of running the Coccinelle script from scripts/coccinelle/misc/cond_return_no_effect.cocci. Signed-off-by: Sang-Heon Jeon <ekffu200098(a)gmail.com> --- sound/soc/amd/acp/acp-mach-common.c | 11 +++-------- sound/soc/intel/atom/sst-mfld-platform-pcm.c | 7 +------ sound/soc/samsung/smdk_spdif.c | 8 ++------ sound/soc/sof/intel/hda-dsp.c | 6 +----- 4 files changed, 7 insertions(+), 25 deletions(-) diff --git a/sound/soc/amd/acp/acp-mach-common.c b/sound/soc/amd/acp/acp-mach-common.c index ef784cca13f2..01a0aaa60246 100644 --- a/sound/soc/amd/acp/acp-mach-common.c +++ b/sound/soc/amd/acp/acp-mach-common.c @@ -938,15 +938,10 @@ static int acp_max98388_hw_params(struct snd_pcm_substream *substream, struct snd_soc_dai *codec_dai = snd_soc_card_get_codec_dai(card, MAX98388_CODEC_DAI); - int ret; - ret = snd_soc_dai_set_fmt(codec_dai, - SND_SOC_DAIFMT_CBC_CFC | SND_SOC_DAIFMT_I2S | - SND_SOC_DAIFMT_NB_NF); - if (ret < 0) - return ret; - - return ret; + return snd_soc_dai_set_fmt(codec_dai, + SND_SOC_DAIFMT_CBC_CFC | SND_SOC_DAIFMT_I2S | + SND_SOC_DAIFMT_NB_NF); } static const struct snd_soc_ops acp_max98388_ops = { diff --git a/sound/soc/intel/atom/sst-mfld-platform-pcm.c b/sound/soc/intel/atom/sst-mfld-platform-pcm.c index 9ee4d9926e06..a4c8cbfba096 100644 --- a/sound/soc/intel/atom/sst-mfld-platform-pcm.c +++ b/sound/soc/intel/atom/sst-mfld-platform-pcm.c @@ -236,12 +236,7 @@ static int sst_platform_alloc_stream(struct snd_pcm_substream *substream, stream->stream_info.str_id = str_params.stream_id; - ret_val = stream->ops->open(sst->dev, &str_params); - if (ret_val <= 0) - return ret_val; - - - return ret_val; + return stream->ops->open(sst->dev, &str_params); } static void sst_period_elapsed(void *arg) diff --git a/sound/soc/samsung/smdk_spdif.c b/sound/soc/samsung/smdk_spdif.c index 2474eb619882..515e4dfc1432 100644 --- a/sound/soc/samsung/smdk_spdif.c +++ b/sound/soc/samsung/smdk_spdif.c @@ -130,12 +130,8 @@ static int smdk_hw_params(struct snd_pcm_substream *substream, return ret; /* Set S/PDIF uses internal source clock */ - ret = snd_soc_dai_set_sysclk(cpu_dai, SND_SOC_SPDIF_INT_MCLK, - rclk_rate, SND_SOC_CLOCK_IN); - if (ret < 0) - return ret; - - return ret; + return snd_soc_dai_set_sysclk(cpu_dai, SND_SOC_SPDIF_INT_MCLK, + rclk_rate, SND_SOC_CLOCK_IN); } static const struct snd_soc_ops smdk_spdif_ops = { diff --git a/sound/soc/sof/intel/hda-dsp.c b/sound/soc/sof/intel/hda-dsp.c index e9f092f082a1..b9b2bdff4ccb 100644 --- a/sound/soc/sof/intel/hda-dsp.c +++ b/sound/soc/sof/intel/hda-dsp.c @@ -1114,11 +1114,7 @@ static int hda_dsp_s5_quirk(struct snd_sof_dev *sdev) usleep_range(500, 1000); /* Restore state for shutdown, back to reset */ - ret = hda_dsp_ctrl_link_reset(sdev, true); - if (ret < 0) - return ret; - - return ret; + return hda_dsp_ctrl_link_reset(sdev, true); } int hda_dsp_shutdown_dma_flush(struct snd_sof_dev *sdev) -- 2.43.0

1 0

[PATCH] ASoC: SOF: validate topology volume range before allocation
by Pengpeng Hou 22 Jul '26

22 Jul '26

Topology supplies mixer min and max as unsigned 32-bit values. SOF stores them in signed fields and passes max + 1 through an int size argument to the IPC-specific volume-table allocator. An inverted range is invalid, and a maximal value can wrap or become negative before allocation. Validate min <= max and require max + 1 to remain representable as an int before storing the range or allocating the table. Signed-off-by: Pengpeng Hou <pengpeng(a)iscas.ac.cn> --- sound/soc/sof/topology.c | 18 ++++++++++++------ 1 file changed, 12 insertions(+), 6 deletions(-) diff --git a/sound/soc/sof/topology.c b/sound/soc/sof/topology.c index 63d582c65891..f4700a500726 100644 --- a/sound/soc/sof/topology.c +++ b/sound/soc/sof/topology.c @@ -836,6 +836,7 @@ static int sof_control_load_volume(struct snd_soc_component *scomp, struct snd_soc_tplg_mixer_control *mc = container_of(hdr, struct snd_soc_tplg_mixer_control, hdr); int tlv[SOF_TLV_ITEMS]; + u32 min, max; unsigned int mask; int ret; @@ -843,6 +844,11 @@ static int sof_control_load_volume(struct snd_soc_component *scomp, if (le32_to_cpu(mc->num_channels) > SND_SOC_TPLG_MAX_CHAN) return -EINVAL; + min = le32_to_cpu(mc->min); + max = le32_to_cpu(mc->max); + if (min > max || max >= INT_MAX) + return -EINVAL; + /* * If control has more than 2 channels we need to override the info. This is because even if * ASoC layer has defined topology's max channel count to SND_SOC_TPLG_MAX_CHAN = 8, the @@ -853,12 +859,12 @@ static int sof_control_load_volume(struct snd_soc_component *scomp, kc->info = snd_sof_volume_info; scontrol->comp_id = sdev->next_comp_id; - scontrol->min_volume_step = le32_to_cpu(mc->min); - scontrol->max_volume_step = le32_to_cpu(mc->max); + scontrol->min_volume_step = min; + scontrol->max_volume_step = max; scontrol->num_channels = le32_to_cpu(mc->num_channels); - scontrol->max = le32_to_cpu(mc->max); - if (le32_to_cpu(mc->max) == 1) + scontrol->max = max; + if (max == 1) goto skip; /* extract tlv data */ @@ -868,7 +874,7 @@ static int sof_control_load_volume(struct snd_soc_component *scomp, } /* set up volume table */ - ret = set_up_volume_table(scontrol, tlv, le32_to_cpu(mc->max) + 1); + ret = set_up_volume_table(scontrol, tlv, max + 1); if (ret < 0) { dev_err(scomp->dev, "error: setting up volume table\n"); return ret; @@ -901,7 +907,7 @@ static int sof_control_load_volume(struct snd_soc_component *scomp, return 0; err: - if (le32_to_cpu(mc->max) > 1) + if (max > 1) kfree(scontrol->volume_table); return ret; -- 2.43.0

1 0

[PATCH v1 1/1] ASoC: SOF: Intel: Don't use "proxy" headers
by Andy Shevchenko 18 Jul '26

18 Jul '26

Update header inclusions to follow IWYU (Include What You Use) principle. Signed-off-by: Andy Shevchenko <andriy.shevchenko(a)linux.intel.com> --- Induced by the ongoing task to drop always included kconfig.h. sound/soc/sof/intel/icl.c | 10 +++++++--- 1 file changed, 7 insertions(+), 3 deletions(-) diff --git a/sound/soc/sof/intel/icl.c b/sound/soc/sof/intel/icl.c index c1018893750c..549bb4ca73e1 100644 --- a/sound/soc/sof/intel/icl.c +++ b/sound/soc/sof/intel/icl.c @@ -9,10 +9,14 @@ * Hardware interface for audio DSP on IceLake. */ -#include <linux/kernel.h> -#include <linux/kconfig.h> -#include <linux/export.h> +#include <linux/array_size.h> #include <linux/bits.h> +#include <linux/dev_printk.h> +#include <linux/errno.h> +#include <linux/slab.h> +#include <linux/string.h> +#include <linux/types.h> + #include "../ipc4-priv.h" #include "../ops.h" #include "hda.h" -- 2.50.1

2 1

[PATCH 00/42] of: reserved_mem: Introduce devres helpers and convert drivers
by Mukesh Ojha 17 Jul '26

17 Jul '26

Drivers using of_reserved_mem_device_init() and its variants must manually call of_reserved_mem_device_release() in their remove and error-unwind paths. This is repetitive boilerplate that is easy to get wrong, and several drivers have open-coded the teardown inconsistently or skipped it entirely, leading to dangling reserved memory references. This series introduces devres-managed wrappers — devm_of_reserved_mem_device_init(), devm_of_reserved_mem_device_init_by_idx(), and devm_of_reserved_mem_device_init_by_name() — that tie the reserved memory region lifetime to the device, releasing it automatically on unbind. The remaining 40 patches convert drivers across the drm, media, ASoC, remoteproc, firmware, mmc, memory and misc subsystems to use these helpers, yielding a net reduction of ~90 lines of boilerplate. This series depends on https://lore.kernel.org/lkml/20260703164457.4040457-1-mukesh.ojha@oss.qualc… Konrad Dybcio (1): of: reserved_mem: Introduce devres-managed initialization functions Mukesh Ojha (41): of: reserved_mem: Add devm_of_reserved_mem_device_init_by_name() firmware: qcom: scm: Use devm_of_reserved_mem_device_init() remoteproc: da8xx: Use devm_of_reserved_mem_device_init() remoteproc: keystone: Use devm_of_reserved_mem_device_init() media: synopsys: hdmirx: Use devm_of_reserved_mem_device_init() remoteproc: omap: Use devm_of_reserved_mem_device_init() drm: logicvc: Use devm_of_reserved_mem_device_init() drm: hdlcd: Use devm_of_reserved_mem_device_init() drm: pl111: Use devm_of_reserved_mem_device_init() remoteproc: mtk_scp: Use devm_of_reserved_mem_device_init() media: aspeed: Use devm_of_reserved_mem_device_init() media: nuvoton: npcm-video: Use devm_of_reserved_mem_device_init() memory: tegra210-emc: Use devm_of_reserved_mem_device_init_by_name() drm: komeda: Use devm_of_reserved_mem_device_init() drm: malidp: Use devm_of_reserved_mem_device_init() drm: ingenic: Use devm_of_reserved_mem_device_init() drm: kmb: Use devm_of_reserved_mem_device_init() drm: sun4i: Use devm_of_reserved_mem_device_init() drm: xlnx: zynqmp_dpsub: Use devm_of_reserved_mem_device_init() media: arm: mali-c55: Use devm_of_reserved_mem_device_init() media: mediatek: vpu: Use devm_of_reserved_mem_device_init() mmc: sdhci-of-bst: Use devm_of_reserved_mem_device_init_by_idx() remoteproc: ti_k3: Use devm_of_reserved_mem_device_init() ASoC: mediatek: mt8192: Use devm_of_reserved_mem_device_init() ASoC: mediatek: mt8196: Use devm_of_reserved_mem_device_init() ASoC: mediatek: mt8183: Use devm_of_reserved_mem_device_init() ASoC: mediatek: mt8189: Use devm_of_reserved_mem_device_init() ASoC: SOF: imx: Use devm_of_reserved_mem_device_init_by_name() staging: media: cedrus: Use devm_of_reserved_mem_device_init() ASoC: cix-ipbloq: Use devm_of_reserved_mem_device_init() drm: aspeed: Use devm_of_reserved_mem_device_init() drm: arcpgu: Use devm_of_reserved_mem_device_init() ASoC: mediatek: mt8173: Use devm_of_reserved_mem_device_init() ASoC: mediatek: mt8186: Use devm_of_reserved_mem_device_init() ASoC: mediatek: mt8188: Use devm_of_reserved_mem_device_init() ASoC: mediatek: mt8195: Use devm_of_reserved_mem_device_init() ASoC: SOF: mediatek: mt8186: Use devm_of_reserved_mem_device_init() ASoC: SOF: mediatek: mt8195: Use devm_of_reserved_mem_device_init() misc: fastrpc: Use devm_of_reserved_mem_device_init() ASoC: fsl: imx-rpmsg: Use devm_of_reserved_mem_device_init_by_idx() ASoC: sprd: Use devm_of_reserved_mem_device_init() drivers/firmware/qcom/qcom_scm.c | 22 +++------ .../gpu/drm/arm/display/komeda/komeda_dev.c | 4 +- drivers/gpu/drm/arm/hdlcd_drv.c | 6 +-- drivers/gpu/drm/arm/malidp_drv.c | 4 +- drivers/gpu/drm/aspeed/aspeed_gfx_drv.c | 2 +- drivers/gpu/drm/ingenic/ingenic-drm-drv.c | 13 +----- drivers/gpu/drm/kmb/kmb_drv.c | 12 +---- drivers/gpu/drm/logicvc/logicvc_drm.c | 21 ++++----- drivers/gpu/drm/pl111/pl111_drv.c | 4 +- drivers/gpu/drm/sun4i/sun4i_drv.c | 4 +- drivers/gpu/drm/tiny/arcpgu.c | 2 +- drivers/gpu/drm/xlnx/zynqmp_dpsub.c | 4 +- .../platform/arm/mali-c55/mali-c55-core.c | 12 ++--- drivers/media/platform/aspeed/aspeed-video.c | 12 ++--- drivers/media/platform/mediatek/vpu/mtk_vpu.c | 3 +- drivers/media/platform/nuvoton/npcm-video.c | 9 +--- .../platform/synopsys/hdmirx/snps_hdmirx.c | 16 +------ drivers/memory/tegra/tegra210-emc-core.c | 21 ++++----- drivers/misc/fastrpc.c | 2 +- drivers/mmc/host/sdhci-of-bst.c | 7 +-- drivers/of/of_reserved_mem.c | 41 +++++++++++++++++ drivers/remoteproc/da8xx_remoteproc.c | 10 +--- drivers/remoteproc/keystone_remoteproc.c | 16 +------ drivers/remoteproc/mtk_scp.c | 3 +- drivers/remoteproc/omap_remoteproc.c | 13 +----- drivers/remoteproc/ti_k3_common.c | 13 +----- drivers/remoteproc/ti_k3_common.h | 1 - .../staging/media/sunxi/cedrus/cedrus_hw.c | 6 +-- include/linux/of_reserved_mem.h | 46 +++++++++++++++++++ sound/hda/controllers/cix-ipbloq.c | 2 +- sound/soc/fsl/imx-rpmsg.c | 2 +- sound/soc/mediatek/mt8173/mt8173-afe-pcm.c | 2 +- sound/soc/mediatek/mt8183/mt8183-afe-pcm.c | 13 +----- sound/soc/mediatek/mt8186/mt8186-afe-pcm.c | 2 +- sound/soc/mediatek/mt8188/mt8188-afe-pcm.c | 2 +- sound/soc/mediatek/mt8189/mt8189-afe-pcm.c | 16 +------ sound/soc/mediatek/mt8192/mt8192-afe-pcm.c | 11 +---- sound/soc/mediatek/mt8195/mt8195-afe-pcm.c | 2 +- sound/soc/mediatek/mt8196/mt8196-afe-pcm.c | 14 +----- sound/soc/sof/imx/imx-common.c | 9 ++-- sound/soc/sof/mediatek/mt8186/mt8186.c | 2 +- sound/soc/sof/mediatek/mt8195/mt8195.c | 2 +- sound/soc/sprd/sprd-pcm-dma.c | 3 +- 43 files changed, 162 insertions(+), 249 deletions(-) -- 2.53.0

13 55

[PATCH] ASoC: SOF: amd: add missing blank line between Kconfig entries
by Vijendar Mukunda 07 Jul '26

07 Jul '26

Add a missing blank line between SND_SOC_SOF_AMD_ACP70 and SND_SOC_SOF_AMD_ACP7X config entries to conform to Kconfig formatting convention. Signed-off-by: Vijendar Mukunda <Vijendar.Mukunda(a)amd.com> --- sound/soc/sof/amd/Kconfig | 1 + 1 file changed, 1 insertion(+) diff --git a/sound/soc/sof/amd/Kconfig b/sound/soc/sof/amd/Kconfig index 7413bfdb0943..903e7ec3b3ba 100644 --- a/sound/soc/sof/amd/Kconfig +++ b/sound/soc/sof/amd/Kconfig @@ -103,6 +103,7 @@ config SND_SOC_SOF_AMD_ACP70 Select this option for SOF support on AMD ACP7.0/ACP7.1 version based platforms. Say Y if you want to enable SOF on ACP7.0/ACP7.1 based platforms. + config SND_SOC_SOF_AMD_ACP7X tristate "SOF support for ACP7.B/7.F platforms" depends on SND_SOC_SOF_PCI -- 2.48.1

2 1

Re: [PATCH 00/42] of: reserved_mem: Introduce devres helpers and convert drivers
by Mukesh Ojha 06 Jul '26

06 Jul '26

On Mon, Jul 06, 2026 at 01:14:13PM +0100, Mark Brown wrote: > On Sat, Jul 04, 2026 at 01:08:13AM +0530, Mukesh Ojha wrote: > > Drivers using of_reserved_mem_device_init() and its variants must > > manually call of_reserved_mem_device_release() in their remove and > > error-unwind paths. This is repetitive boilerplate that is easy to > > get wrong, and several drivers have open-coded the teardown > > inconsistently or skipped it entirely, leading to dangling reserved > > memory references. > > Just as a general thing, rather than sending a huge cross subsystem > series for something like this it's probably better to just send a > couple of examples, then once the new API is introduced and either has a > pullable tag or has made it to mainline go and do the conversion. This > avoids clogging everyone's inbox with very big serieses. Apologies for this; this is my first series touching these many subsystems, and I also learned the hard way when the entire series was not sent in one shot due to the daily sending limit via git send-email[1], and I had to cut many cc lists for patches 33/42 to 42/42. [1] 4.5.3 Your message has too many recipients. For more information regarding 4.5.3 Google's sending limits, go to 4.5.3 https://support.google.com/mail/?p=TooManyRecipientsError 5a478bee46e88-30f0bbd2362sm49597158eec.20 - gsmtp Thanks for the suggestion; point noted. -- -Mukesh Ojha

1 0

[PATCH 00/16] ASoC: SOF: amd: add support for ACP7.B/7.F platforms
by Vijendar Mukunda 04 Jul '26

04 Jul '26

This patch series adds Sound Open Firmware (SOF) enablement for AMD ACP7.B/ACP7.F platforms by introducing a dedicated ACP7x PCI driver/ops layer built on the existing ACP SOF infrastructure, plus required IPC3 topology/token extensions for ACP7x I2S/TDM. Changes: - Add new ACP7x PCI driver + DSP ops, including suspend/resume and IRQ handling paths. - Extend AMD ACP firmware loading logic for signed images and SRAM-based data loading on ACP7x. - Introduce a new SOF DAI type for AMD I2S/TDM and extend IPC3 topology tokens to carry an ACP TDM/I2S format field. Link: https://github.com/thesofproject/linux/pull/5813 Vijendar Mukunda (16): ASoC: amd: acp: add ACPI machine table for ACP7.B/7.F SOF driver ASoC: SOF: amd: add base platform support for ACP7.B/7.F ASoC: SOF: amd: mask ACP7x PGFSM status poll ASoC: SOF: amd: refactor SW1 I2S error reason clear in acp_irq_handler ASoC: SOF: amd: add ACP7x probe and remove ASoC: SOF: amd: add ACP7x IRQ handler for DSP IPC ASoC: SOF: amd: extend signed firmware pre-run for ACP7x ASoC: SOF: amd: require full ACP header for ACP7 signed firmware ASoC: SOF: amd: validate SizeFWSigned before signed FW length on ACP7x ASoC: SOF: amd: add post-firmware-run delay for ACP7x ASoC: SOF: amd: extend configure_and_run_sha_dma for ACPI signed FW flag ASoC: SOF: amd: wire signed firmware load callback for ACP7x via ACPI ASoC: SOF: amd: load ACP7.B/7.F signed data firmware to SRAM ASoC: SOF: amd: add ACP I2S format field and topology token ASoC: SOF: amd: add ACP7x I2S DAI type and topology support ASoC: SOF: amd: add system and runtime PM ops for ACP7x include/sound/sof/dai-amd.h | 1 + include/sound/sof/dai.h | 2 + include/uapi/sound/sof/tokens.h | 1 + sound/soc/amd/acp-config.c | 12 ++ sound/soc/amd/mach-config.h | 1 + sound/soc/sof/amd/Kconfig | 10 ++ sound/soc/sof/amd/Makefile | 2 + sound/soc/sof/amd/acp-dsp-offset.h | 17 ++ sound/soc/sof/amd/acp-ipc.c | 15 +- sound/soc/sof/amd/acp-loader.c | 37 +++- sound/soc/sof/amd/acp.c | 267 ++++++++++++++++++++++++++++- sound/soc/sof/amd/acp.h | 20 +++ sound/soc/sof/amd/acp7x.c | 185 ++++++++++++++++++++ sound/soc/sof/amd/pci-acp7x.c | 116 +++++++++++++ sound/soc/sof/ipc3-pcm.c | 6 + sound/soc/sof/ipc3-topology.c | 39 ++++- sound/soc/sof/topology.c | 3 +- 17 files changed, 720 insertions(+), 14 deletions(-) create mode 100644 sound/soc/sof/amd/acp7x.c create mode 100644 sound/soc/sof/amd/pci-acp7x.c -- 2.45.2

3 18

Re: [PATCH 16/16] ASoC: SOF: amd: add system and runtime PM ops for ACP7x
by Mukunda,Vijendar 04 Jul '26

04 Jul '26

On 7/3/26 22:39, Mark Brown wrote: > On Fri, Jul 03, 2026 at 10:21:25PM +0530, Mukunda,Vijendar wrote: >> On 7/3/26 21:58, Mark Brown wrote: >>> On Wed, Jul 01, 2026 at 03:25:17PM +0530, Vijendar Mukunda wrote: >>> The interrupt is requested with IRQF_SHARED so probably worth double >>> checking if the interrupt handler is safe to run while the suspend and >>> resume callbacks are running, I see we reset the device during suspend >>> so there might be some risk of corrupted register reads? >> There won't be any corrupted register reads when acp reset sequence >> is executed, all the acp registers are set to default values. i.e Interrupt >> control registers are disabled. This sequence ensures that till interrupt > Including whatever the threaded handler is doing? I didn't actually > check properly, just saw the indirection through the SOF I/O functions > and the threaded handler. |The threaded handler (acp_sof_ipc_irq_thread) can only be scheduled when acp7x_irq_handler returns IRQ_WAKE_THREAD, which only happens when ACP_DSP_TO_HOST_IRQ is set in ACP_DSP_SW_INTR_STAT. By the time the platform suspend callback runs, the SOF core has already quiesced the DSP (ctx_save IPC, pipeline teardown), so no new DSP-to-host interrupt can arrive. The subsequent acp_reset() then zeroes all interrupt-enable registers, preventing any further IRQ_WAKE_THREAD from acp7x_irq_handler. If a thread was already queued before reset completes, it runs acp_sof_ipc_irq_thread which reads the scratch SRAM mailbox registers (dsp_msg_write, dsp_ack_write). After a soft reset those registers read 0, so the handler finds nothing to do and returns IRQ_HANDLED cleanly.| >> masks are enabled during resume sequence, no interrupt will be asserted >> for ACP IP. > Right, but IRQF_SHARED means it might be an interrupt for something > else. Correct. If another device asserts the shared INTx line during suspend, the kernel calls acp7x_irq_handler. In that case both ACP_DSP_SW_INTR_STAT and ACP_EXT_INTR_STAT read 0 (device is reset/idle), so the handler returns IRQ_NONE immediately without scheduling the thread or touching any other registers. No corrupted reads or harmful side effects occur. In practice the ACP PCI device on ACP7.B/7.F has the INTx line assigned exclusively to itself — we verified on target that no other device shares it. IRQF_SHARED is set solely because the kernel's PCI layer requires it for INTx interrupts; omitting it causes request_threaded_irq to fail even when the line is unshared. We can drop IRQF_SHARED flag, if that is preferred. Please let us know.

1 0