July 2023 - Alsa-devel - mailman.alsa-project.org

[PATCH v2 00/15] PCI: Define Intel PCI IDs and use them in drivers
by Amadeusz Sławiński 14 Jul '23

14 Jul '23

PCI IDs for Intel HDA are duplicated across quite a few drivers, due to various configurations and historical reasons. Currently almost all uses of HDA PCI IDs have corresponding comment telling which platform it is. Additionally there are some inconsistencies between drivers about which ID corresponds to which device. Simplify things, by adding PCI IDs to global header and make use of them in drivers. This allows for removal of comments by having IDs themselves being self explanatory. Additionally it allows for removal of existing inconsistencies by having one source of truth. Changes from v1: - Remove BXT-T PCI ID as it is not released (Andy) - Reorder macros to start from pci_match_id() (Andy) - Add comment about BXT->APL name change in commit messages (Andy) - Use SST as part of macro name for DSP only devices (Andy) - Add PCI IDs for all SST combinations (Andy) Changes from RFC: - Sort Intel PCI IDs before adding new ones - Fix ordering of new PCI IDs (Andy) - Define all used Intel IDs (Andy) - Add macros for controller type detection (Andy/Bjorn) - Add set of patches changing to use above macro (Andy/Bjorn) - Use PCI_DEVICE_DATA for Intel IDs in sound/pci/hda/hda_intel.c (Andy) - Commit message wording (Andy) - Remove unnecessary tabs (Andy) Amadeusz Sławiński (15): PCI: Sort Intel PCI IDs by number PCI: Add Intel Audio DSP devices to pci_ids.h ASoC: SOF: Remove unused Broxton PCI ID ALSA: Remove unused Broxton PCI ID ALSA: hda: Add controller matching macros ALSA: hda: Use global PCI match macro ALSA: hda/i915: Use global PCI match macro ASoC: Intel: Skylake: Use global PCI match macro ALSA: intel-dsp-config: Convert to PCI device IDs defines ALSA: hda: Convert to PCI device IDs defines ASoC: Intel: avs: Convert to PCI device IDs defines ASoC: Intel: avs: Convert to PCI device IDs defines ASoC: Intel: Skylake: Convert to PCI device IDs defines ASoC: SOF: Intel: Convert to PCI device IDs defines ASoC: Intel: sst: Convert to PCI device IDs defines include/linux/pci_ids.h | 105 +++++-- include/sound/hda_codec.h | 3 - include/sound/hdaudio.h | 26 ++ sound/hda/hdac_i915.c | 7 +- sound/hda/intel-dsp-config.c | 124 ++++---- sound/pci/hda/hda_intel.c | 373 ++++++++++--------------- sound/soc/intel/atom/sst/sst.c | 7 +- sound/soc/intel/atom/sst/sst.h | 3 - sound/soc/intel/atom/sst/sst_pci.c | 4 +- sound/soc/intel/avs/board_selection.c | 10 +- sound/soc/intel/avs/core.c | 16 +- sound/soc/intel/skylake/skl-messages.c | 16 +- sound/soc/intel/skylake/skl-pcm.c | 3 +- sound/soc/intel/skylake/skl.c | 36 +-- sound/soc/sof/intel/pci-apl.c | 8 +- sound/soc/sof/intel/pci-cnl.c | 15 +- sound/soc/sof/intel/pci-icl.c | 12 +- sound/soc/sof/intel/pci-mtl.c | 3 +- sound/soc/sof/intel/pci-skl.c | 6 +- sound/soc/sof/intel/pci-tgl.c | 45 +-- sound/soc/sof/intel/pci-tng.c | 3 +- 21 files changed, 383 insertions(+), 442 deletions(-) -- 2.34.1

3 23

[PATCH] lib: Introduce KUnit tests for minmax
by Wojciech Ziemba 14 Jul '23

14 Jul '23

This patch adds KUnit coverage for minmax_array() macros from <linux/minmax.h>. It assures that these macros compile correctly and work as expected when different types are used. The test cases exercise types: s8, u8, s16, u16, s32, u32, s64, u64 with arrays of random numbers. Signed-off-by: Wojciech Ziemba <wojciech.ziemba(a)intel.com> --- lib/Kconfig.debug | 12 +++++ lib/Makefile | 1 + lib/minmax_kunit.c | 132 +++++++++++++++++++++++++++++++++++++++++++++ 3 files changed, 145 insertions(+) create mode 100644 lib/minmax_kunit.c diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug index fbc89baf7de6..e845876482fb 100644 --- a/lib/Kconfig.debug +++ b/lib/Kconfig.debug @@ -2777,6 +2777,18 @@ config SIPHASH_KUNIT_TEST This is intended to help people writing architecture-specific optimized versions. If unsure, say N. +config MINMAX_KUNIT_TEST + tristate "KUnit test for minmax" if !KUNIT_ALL_TESTS + depends on KUNIT + default KUNIT_ALL_TESTS + help + Builds KUnit tests module for minmax. + + For more information on KUnit and unit tests in general please refer + to the KUnit documentation in Documentation/dev-tools/kunit/. + + If unsure, say N. + config TEST_UDELAY tristate "udelay test driver" help diff --git a/lib/Makefile b/lib/Makefile index 42d307ade225..950a3f388e6f 100644 --- a/lib/Makefile +++ b/lib/Makefile @@ -396,6 +396,7 @@ obj-$(CONFIG_FORTIFY_KUNIT_TEST) += fortify_kunit.o obj-$(CONFIG_STRCAT_KUNIT_TEST) += strcat_kunit.o obj-$(CONFIG_STRSCPY_KUNIT_TEST) += strscpy_kunit.o obj-$(CONFIG_SIPHASH_KUNIT_TEST) += siphash_kunit.o +obj-$(CONFIG_MINMAX_KUNIT_TEST) += minmax_kunit.o obj-$(CONFIG_GENERIC_LIB_DEVMEM_IS_ALLOWED) += devmem_is_allowed.o diff --git a/lib/minmax_kunit.c b/lib/minmax_kunit.c new file mode 100644 index 000000000000..00ec175156d6 --- /dev/null +++ b/lib/minmax_kunit.c @@ -0,0 +1,132 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * Copyright (c) 2023 Intel Corporation + * + * KUnit test cases for minmax. + */ +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + +#include <kunit/test.h> + +#include <linux/kernel.h> +#include <linux/minmax.h> +#include <linux/module.h> +#include <linux/types.h> + +#define MINMAX_ARR_TEST(type, op, expected) \ + static const type expected_##op##_##type = expected; \ +static void op##_array_##type##_test(struct kunit *test) \ +{ \ + type value = op##_array(test_vector_##type, \ + ARRAY_SIZE(test_vector_##type)); \ + KUNIT_EXPECT_EQ(test, value, expected_##op##_##type); \ +} + +/* type 's8' */ +static const s8 test_vector_s8[] = { + 1, 85, -112, -61, 6, 0, 117, -55, 68, -55, 23, 1, -100, 55, -118 +}; + +MINMAX_ARR_TEST(s8, min, -118) +MINMAX_ARR_TEST(s8, max, 117) + +/* type 'u8' */ +static const u8 test_vector_u8[] = { + 209, 145, 131, 180, 88, 97, 240, 170, 148, 158, 35, 226, 14, 244, 2 +}; + +MINMAX_ARR_TEST(u8, min, 2) +MINMAX_ARR_TEST(u8, max, 244) + +/* type 's16' */ +static const s16 test_vector_s16[] = { + 2641, 30296, -15090, 11371, 13995, 28244, 27482, -7239, -9036, -3147, + 22065, 2097, -26854, -8633, 13366 +}; + +MINMAX_ARR_TEST(s16, min, -26854) +MINMAX_ARR_TEST(s16, max, 30296) + +/* type 'u16' */ +static const u16 test_vector_u16[] = { + 50102, 58387, 42350, 33531, 32109, 65525, 25985, 31889, 14772, 7296, + 48242, 53340, 33594, 19856, 45544 +}; + +MINMAX_ARR_TEST(u16, min, 7296) +MINMAX_ARR_TEST(u16, max, 65525) + +/* type 's32' */ +static const s32 test_vector_s32[] = { + 979885732L, 885950778L, -1450966230L, -1624192399L, -1753023734L, + 530868381L, 656215973L, 630677997L, 391000329L, 1253640357L, + 438040531L, -993355872L, 585127620L, -1068766240L, 147051568L +}; + +MINMAX_ARR_TEST(s32, min, -1753023734L) +MINMAX_ARR_TEST(s32, max, 1253640357L) + +/* type 'u32' */ +static const u32 test_vector_u32[] = { + 2161504111UL, 1183092838UL, 1728689175UL, 2309150449UL, 3735269177UL, + 1550306875UL, 2147576091UL, 4187089172UL, 545942284UL, 1969608375UL, + 2624877400UL, 3029665375UL, 1293120185UL, 3801307680UL, 1161238802UL +}; + +MINMAX_ARR_TEST(u32, min, 545942284UL) +MINMAX_ARR_TEST(u32, max, 4187089172UL) + +/* type 's64' */ +static const s64 test_vector_s64[] = { + -8438490804591620000LL, 8025377321703520000LL, 3208690154945970000LL, + -2321767394263770000LL, -1278489889635390000LL, 2190470827346570000LL, + -2022822382784580000LL, 6599789147713410000LL, 3949996236706410000LL, + -8261935734503760000LL, 661645293834217000LL, 8346383784731440000LL, + -3443819189603500000LL, 7985744187902840000LL, 5083298007916820000LL +}; + +MINMAX_ARR_TEST(s64, min, -8438490804591620000LL) +MINMAX_ARR_TEST(s64, max, 8346383784731440000LL) + +/* type 'u64' */ +static const u64 test_vector_u64[] = { + 8249683091780880000ULL, 13192984997103100000ULL, 17673970905507700000ULL, + 11101419446779000000ULL, 7059417167478820000ULL, 17488355550255500000ULL, + 1592214570100350000ULL, 4324504718429290000ULL, 9233486922226120000ULL, + 17205873265072200000ULL, 15951973546886800000ULL, 8694358874973410000ULL, + 11857168979503600000ULL, 15015693759541800000ULL, 4758196429468010000ULL +}; + +MINMAX_ARR_TEST(u64, min, 1592214570100350000ULL) +MINMAX_ARR_TEST(u64, max, 17673970905507700000ULL) +#undef MINMAX_ARR_TEST + +static struct kunit_case minmax_array_test_cases[] = { + KUNIT_CASE(min_array_s8_test), + KUNIT_CASE(max_array_s8_test), + KUNIT_CASE(min_array_u8_test), + KUNIT_CASE(max_array_u8_test), + KUNIT_CASE(min_array_s16_test), + KUNIT_CASE(max_array_s16_test), + KUNIT_CASE(min_array_u16_test), + KUNIT_CASE(max_array_u16_test), + KUNIT_CASE(min_array_s32_test), + KUNIT_CASE(max_array_s32_test), + KUNIT_CASE(min_array_u32_test), + KUNIT_CASE(max_array_u32_test), + KUNIT_CASE(min_array_s64_test), + KUNIT_CASE(max_array_s64_test), + KUNIT_CASE(min_array_u64_test), + KUNIT_CASE(max_array_u64_test), + {} +}; + +static struct kunit_suite minmax_test_suite = { + .name = "minmax_array", + .test_cases = minmax_array_test_cases, +}; +kunit_test_suite(minmax_test_suite); + +MODULE_AUTHOR("Intel"); +MODULE_LICENSE("GPL"); +MODULE_DESCRIPTION("KUnit test module for minmax"); -- 2.38.1

1 0

[PATCH v2 1/2] ALSA: pcmtest: Add 'open' PCM callback error injection
by Ivan Orlov 14 Jul '23

14 Jul '23

Extend 'pcmtest' virtual driver with 'open' callback error injection functionality, as it already can inject errors into other PCM callbacks. Add module parameter which enables EBUSY error injection in the 'open' PCM callback. Signed-off-by: Ivan Orlov <ivan.orlov0322(a)gmail.com> --- V1 -> V2: - Split changes in the patch into two different patches sound/drivers/pcmtest.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/sound/drivers/pcmtest.c b/sound/drivers/pcmtest.c index 291e7fe47893..e74c523e49eb 100644 --- a/sound/drivers/pcmtest.c +++ b/sound/drivers/pcmtest.c @@ -65,6 +65,7 @@ static int inject_delay; static bool inject_hwpars_err; static bool inject_prepare_err; static bool inject_trigger_err; +static bool inject_open_err; static short fill_mode = FILL_MODE_PAT; @@ -88,6 +89,9 @@ module_param(inject_prepare_err, bool, 0600); MODULE_PARM_DESC(inject_prepare_err, "Inject EINVAL error in the 'prepare' callback"); module_param(inject_trigger_err, bool, 0600); MODULE_PARM_DESC(inject_trigger_err, "Inject EINVAL error in the 'trigger' callback"); +module_param(inject_open_err, bool, 0600); +MODULE_PARM_DESC(inject_open_err, "Inject EBUSY error in the 'open' callback"); + struct pcmtst { struct snd_pcm *pcm; @@ -364,6 +368,9 @@ static int snd_pcmtst_pcm_open(struct snd_pcm_substream *substream) struct snd_pcm_runtime *runtime = substream->runtime; struct pcmtst_buf_iter *v_iter; + if (inject_open_err) + return -EBUSY; + v_iter = kzalloc(sizeof(*v_iter), GFP_KERNEL); if (!v_iter) return -ENOMEM; -- 2.34.1

2 2

Re: [PATCH 2/4] ASoC: codecs: es8326: Fix power-up sequence
by Zhu Ning 14 Jul '23

14 Jul '23

> > > This looks wrong, you're resyncing the cache and then start resetting > > > registers? It feels like the ordering is off here, and some of this > > > reset sequence might want to be done with the cache bypassed. Are you > > > sure that there's no corruption of user visible state resulting from the > > > power up sequence, especially around the HP driver? > > > Basically the chip runs through the start-up sequence every time in > > the resume function. > > Will move the reset sequence to es8326_suspend > > That's not going to overwrite any user visible settings? Registers written here are internal bias/vmid which should be fixed. User settings like alc and volume won't be overwritten here.

1 0

[PATCH] ASoC: simple-card: add trigger-stop entry parser
by Maxim Kochetkov 14 Jul '23

14 Jul '23

It may be useful to specify trigger-stop for some DMA-based simple audio card. So add this "trigger-stop" device tree entry parser. Signed-off-by: Maxim Kochetkov <fido_max(a)inbox.ru> --- sound/soc/generic/simple-card.c | 19 +++++++++++++++++++ 1 file changed, 19 insertions(+) diff --git a/sound/soc/generic/simple-card.c b/sound/soc/generic/simple-card.c index a78babf44f38..86df7d326068 100644 --- a/sound/soc/generic/simple-card.c +++ b/sound/soc/generic/simple-card.c @@ -145,8 +145,27 @@ static int simple_parse_node(struct asoc_simple_priv *priv, struct snd_soc_dai_link_component *dlc; struct asoc_simple_dai *dai; int ret; + const char *str; + struct { + char *name; + unsigned int val; + } of_trigger_table[] = { + { "default", SND_SOC_TRIGGER_ORDER_DEFAULT }, + { "ldc", SND_SOC_TRIGGER_ORDER_LDC }, + }; + if (cpu) { + ret = of_property_read_string(np, "trigger-stop", &str); + if (ret == 0) { + for (int i = 0; i < ARRAY_SIZE(of_trigger_table); i++) { + if (strcmp(str, of_trigger_table[i].name) == 0) { + dai_link->trigger_stop = of_trigger_table[i].val; + break; + } + } + } + dlc = asoc_link_to_cpu(dai_link, 0); dai = simple_props_to_dai_cpu(dai_props, 0); } else { -- 2.40.1

2 1

[PATCH 0/2] kselftest/alsa: Decrease pcm-test duration to avoid timeouts
by Nícolas F. R. A. Prado 13 Jul '23

13 Jul '23

This series decreases the pcm-test duration in order to avoid timeouts by first moving the audio stream duration to a variable and subsequently decreasing it. Nícolas F. R. A. Prado (2): kselftest/alsa: pcm-test: Move stream duration and margin to variables kselftest/alsa: pcm-test: Decrease stream duration from 4 to 2 seconds tools/testing/selftests/alsa/pcm-test.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) -- 2.41.0

5 15

[PATCH] ASoC: q6dsp: q6apm: make g_apm static
by Min-Hua Chen 13 Jul '23

13 Jul '23

This patch fixes the following sprse warning: sound/soc/qcom/qdsp6/q6apm.c:30:14: sparse: warning: symbol 'g_apm' was not declared. Should it be static? No functional change intended Signed-off-by: Min-Hua Chen <minhuadotchen(a)gmail.com> --- sound/soc/qcom/qdsp6/q6apm.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/sound/soc/qcom/qdsp6/q6apm.c b/sound/soc/qcom/qdsp6/q6apm.c index 7bfac9492ab5..d5ad9cbd3c21 100644 --- a/sound/soc/qcom/qdsp6/q6apm.c +++ b/sound/soc/qcom/qdsp6/q6apm.c @@ -27,7 +27,7 @@ struct apm_graph_mgmt_cmd { #define APM_GRAPH_MGMT_PSIZE(p, n) ALIGN(struct_size(p, sub_graph_id_list, n), 8) -struct q6apm *g_apm; +static struct q6apm *g_apm; int q6apm_send_cmd_sync(struct q6apm *apm, struct gpr_pkt *pkt, uint32_t rsp_opcode) { -- 2.34.1

2 1

[PATCH 00/10] ASoC: ad: Update Analog Devices drivers to maple tree register cache
by Mark Brown 13 Jul '23

13 Jul '23

The maple tree register cache has now got to feature parity with the rbtree cache, there are some different tradeoffs made and it should be a better choice for most modern systems. Convert the Analog Devices drivers to use the more modern data structure. Signed-off-by: Mark Brown <broonie(a)kernel.org> --- Mark Brown (10): ASoC: ad1836: Update to use maple tree register cache ASoC: ad1980: Update to use maple tree register cache ASoC: adau1372: Update to use maple tree register cache ASoC: adau1373: Update to use maple tree register cache ASoC: adau1701: Update to use maple tree register cache ASoC: adau1761: Update to use maple tree register cache ASoC: adau1781: Update to use maple tree register cache ASoC: adau1977: Update to use maple tree register cache ASoC: adau7118: Update to use maple tree register cache ASoC: adav80x: Update to use maple tree register cache sound/soc/codecs/ad1836.c | 2 +- sound/soc/codecs/ad1980.c | 2 +- sound/soc/codecs/adau1372.c | 2 +- sound/soc/codecs/adau1373.c | 2 +- sound/soc/codecs/adau1701.c | 2 +- sound/soc/codecs/adau1761.c | 2 +- sound/soc/codecs/adau1781.c | 2 +- sound/soc/codecs/adau1977.c | 2 +- sound/soc/codecs/adau7118-i2c.c | 2 +- sound/soc/codecs/adav80x.c | 2 +- 10 files changed, 10 insertions(+), 10 deletions(-) --- base-commit: 06c2afb862f9da8dc5efa4b6076a0e48c3fbaaa5 change-id: 20230701-asoc-ad-maple-170068cf0c39 Best regards, -- Mark Brown <broonie(a)kernel.org>

2 12

[PATCH 2/2] ALSA: hda/realtek: Add quirks for Lenovo 16IAH7, 16IAX7 and 16ARHA7
by David Xu 13 Jul '23

13 Jul '23

Add quirks for Lenovo 16IAH7, 16IAX7 and 16ARHA7 that have dual cs35l41 amplifiers that drives their speakers. Please note that this patch depends on patch 1, for the added quirks to work for the specified models, proper CSC3551 configurtion fixups have to be applied. Signed-off-by: David Xu <xuwd1(a)hotmail.com> --- sound/pci/hda/patch_realtek.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/sound/pci/hda/patch_realtek.c b/sound/pci/hda/patch_realtek.c index e2f8b608de82..cc10bb8b75d1 100644 --- a/sound/pci/hda/patch_realtek.c +++ b/sound/pci/hda/patch_realtek.c @@ -9831,6 +9831,7 @@ static const struct snd_pci_quirk alc269_fixup_tbl[] = { SND_PCI_QUIRK(0x17aa, 0x31af, "ThinkCentre Station", ALC623_FIXUP_LENOVO_THINKSTATION_P340), SND_PCI_QUIRK(0x17aa, 0x3801, "Lenovo Yoga9 14IAP7", ALC287_FIXUP_YOGA9_14IAP7_BASS_SPK_PIN), SND_PCI_QUIRK(0x17aa, 0x3802, "Lenovo Yoga DuetITL 2021", ALC287_FIXUP_YOGA7_14ITL_SPEAKERS), + SND_PCI_QUIRK(0x17aa, 0x3803, "Legion Slim 7i 16IAH7", ALC287_FIXUP_CS35L41_I2C_2), SND_PCI_QUIRK(0x17aa, 0x3813, "Legion 7i 15IMHG05", ALC287_FIXUP_LEGION_15IMHG05_SPEAKERS), SND_PCI_QUIRK(0x17aa, 0x3818, "Lenovo C940 / Yoga Duet 7", ALC298_FIXUP_LENOVO_C940_DUET7), SND_PCI_QUIRK(0x17aa, 0x3819, "Lenovo 13s Gen2 ITL", ALC287_FIXUP_13S_GEN2_SPEAKERS), @@ -9846,6 +9847,10 @@ static const struct snd_pci_quirk alc269_fixup_tbl[] = { SND_PCI_QUIRK(0x17aa, 0x3853, "Lenovo Yoga 7 15ITL5", ALC287_FIXUP_YOGA7_14ITL_SPEAKERS), SND_PCI_QUIRK(0x17aa, 0x3855, "Legion 7 16ITHG6", ALC287_FIXUP_LEGION_16ITHG6), SND_PCI_QUIRK(0x17aa, 0x3869, "Lenovo Yoga7 14IAL7", ALC287_FIXUP_YOGA9_14IAP7_BASS_SPK_PIN), + SND_PCI_QUIRK(0x17aa, 0x386e, "Legion Slim 7i 16IAH7", ALC287_FIXUP_CS35L41_I2C_2), + SND_PCI_QUIRK(0x17aa, 0x386f, "Legion 7i 16IAX7", ALC287_FIXUP_CS35L41_I2C_2), + SND_PCI_QUIRK(0x17aa, 0x3874, "Legion 7i 16IAX7", ALC287_FIXUP_CS35L41_I2C_2), + SND_PCI_QUIRK(0x17aa, 0x3877, "Legion 7 16ARHA7", ALC287_FIXUP_CS35L41_I2C_2), SND_PCI_QUIRK(0x17aa, 0x3902, "Lenovo E50-80", ALC269_FIXUP_DMIC_THINKPAD_ACPI), SND_PCI_QUIRK(0x17aa, 0x3977, "IdeaPad S210", ALC283_FIXUP_INT_MIC), SND_PCI_QUIRK(0x17aa, 0x3978, "Lenovo B50-70", ALC269_FIXUP_DMIC_THINKPAD_ACPI), -- 2.41.0

1 0

Re: [PATCH 2/4] ASoC: codecs: es8326: Fix power-up sequence
by Zhu Ning 13 Jul '23

13 Jul '23

> > Old power-up sequence causes large pop noise during start-up. > > Using a new sequence instead. > > > static const struct regmap_range es8326_volatile_ranges[] = { > > - regmap_reg_range(ES8326_HP_DETECT, ES8326_HP_DETECT), > > + regmap_reg_range(ES8326_HPDET_TYPE, ES8326_HPDET_TYPE), > > }; > > As well as not seeming to correspond to the patch description this seems > obviously buggy - even if _HPDET_TYPE was missed from the list of > volatile registers no software change should be able to make a volatile > register non-volatile, that's a property of the hardware. The changes > to read from HPDET_TYPE also seem unrelated to the main change here in > the resume function, this should probably be multiple commits. Just a naming space change, will create a new patch for it. > > static int es8326_resume(struct snd_soc_component *component) > > { > > struct es8326_priv *es8326 = snd_soc_component_get_drvdata(component); > > - unsigned int reg; > > > > regcache_cache_only(es8326->regmap, false); > > regcache_sync(es8326->regmap); > > > > + /* reset register value to default */ > > + regmap_write(es8326->regmap, ES8326_CSM_I2C_STA, 0x01); > > + usleep_range(1000, 3000); > > + regmap_write(es8326->regmap, ES8326_CSM_I2C_STA, 0x00); > > This looks wrong, you're resyncing the cache and then start resetting > registers? It feels like the ordering is off here, and some of this > reset sequence might want to be done with the cache bypassed. Are you > sure that there's no corruption of user visible state resulting from the > power up sequence, especially around the HP driver? Basically the chip runs through the start-up sequence every time in the resume function. Will move the reset sequence to es8326_suspend

2 1