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

[PATCH v3 00/35] bitops: add atomic find_bit() operations
by Yury Norov 16 Dec '23

16 Dec '23

Add helpers around test_and_{set,clear}_bit() that allow to search for clear or set bits and flip them atomically. The target patterns may look like this: for (idx = 0; idx < nbits; idx++) if (test_and_clear_bit(idx, bitmap)) do_something(idx); Or like this: do { bit = find_first_bit(bitmap, nbits); if (bit >= nbits) return nbits; } while (!test_and_clear_bit(bit, bitmap)); return bit; In both cases, the opencoded loop may be converted to a single function or iterator call. Correspondingly: for_each_test_and_clear_bit(idx, bitmap, nbits) do_something(idx); Or: return find_and_clear_bit(bitmap, nbits); Obviously, the less routine code people have to write themself, the less probability to make a mistake. Those are not only handy helpers but also resolve a non-trivial issue of using non-atomic find_bit() together with atomic test_and_{set,clear)_bit(). The trick is that find_bit() implies that the bitmap is a regular non-volatile piece of memory, and compiler is allowed to use such optimization techniques like re-fetching memory instead of caching it. For example, find_first_bit() is implemented like this: for (idx = 0; idx * BITS_PER_LONG < sz; idx++) { val = addr[idx]; if (val) { sz = min(idx * BITS_PER_LONG + __ffs(val), sz); break; } } On register-memory architectures, like x86, compiler may decide to access memory twice - first time to compare against 0, and second time to fetch its value to pass it to __ffs(). When running find_first_bit() on volatile memory, the memory may get changed in-between, and for instance, it may lead to passing 0 to __ffs(), which is undefined. This is a potentially dangerous call. find_and_clear_bit() as a wrapper around test_and_clear_bit() naturally treats underlying bitmap as a volatile memory and prevents compiler from such optimizations. Now that KCSAN is catching exactly this type of situations and warns on undercover memory modifications. We can use it to reveal improper usage of find_bit(), and convert it to atomic find_and_*_bit() as appropriate. In some cases concurrent operations with plain find_bit() are acceptable. For example: - two threads running find_*_bit(): safe wrt ffs(0) and returns correct value, because underlying bitmap is unchanged; - find_next_bit() in parallel with set or clear_bit(), when modifying a bit prior to the start bit to search: safe and correct; - find_first_bit() in parallel with set_bit(): safe, but may return wrong bit number; - find_first_zero_bit() in parallel with clear_bit(): same as above. In last 2 cases find_bit() may not return a correct bit number, but it may be OK if caller requires any (not exactly the first) set or clear bit, correspondingly. In such cases, KCSAN may be safely silenced with data_race(). But in most cases where KCSAN detects concurrency people should carefully review their code and likely protect critical sections or switch to atomic find_and_bit(), as appropriate. The 1st patch of the series adds the following atomic primitives: find_and_set_bit(addr, nbits); find_and_set_next_bit(addr, nbits, start); ... Here find_and_{set,clear} part refers to the corresponding test_and_{set,clear}_bit function. Suffixes like _wrap or _lock derive their semantics from corresponding find() or test() functions. For brevity, the naming omits the fact that we search for zero bit in find_and_set, and correspondingly search for set bit in find_and_clear functions. The patch also adds iterators with atomic semantics, like for_each_test_and_set_bit(). Here, the naming rule is to simply prefix corresponding atomic operation with 'for_each'. In [1] Jan reported 2% slowdown in a single-thread search test when switching find_bit() function to treat bitmaps as volatile arrays. On the other hand, kernel robot in the same thread reported +3.7% to the performance of will-it-scale.per_thread_ops test. Assuming that our compilers are sane and generate better code against properly annotated data, the above discrepancy doesn't look weird. When running on non-volatile bitmaps, plain find_bit() outperforms atomic find_and_bit(), and vice-versa. So, all users of find_bit() API, where heavy concurrency is expected, are encouraged to switch to atomic find_and_bit() as appropriate. The 1st patch of this series adds atomic find_and_bit() API, 2nd adds a basic test for new API, and all the following patches spread it over the kernel. They can be applied separately from each other on per-subsystems basis, or I can pull them in bitmap tree, as appropriate. [1] https://lore.kernel.org/lkml/634f5fdf-e236-42cf-be8d-48a581c21660@alu.unizg… --- v1: https://lore.kernel.org/netdev/20231118155105.25678-29-yury.norov@gmail.com… v2: https://lore.kernel.org/all/20231204185101.ddmkvsr2xxsmoh2u@quack3/T/ v3: - collect more reviews; - align wording in commit messages @ Bjorn Helgaas; - add examples where non-atomic find_bit() may safely race @ Jan Kara; - patch #3: use if-else instead of ternary operator @ Jens Axboe; - patch #13: align coding style @ Vitaly Kuznetsov, Sean Christopherson; Yury Norov (35): lib/find: add atomic find_bit() primitives lib/find: add test for atomic find_bit() ops lib/sbitmap; optimize __sbitmap_get_word() by using find_and_set_bit() watch_queue: optimize post_one_notification() by using find_and_clear_bit() sched: add cpumask_find_and_set() and use it in __mm_cid_get() mips: sgi-ip30: optimize heart_alloc_int() by using find_and_set_bit() sparc: optimize alloc_msi() by using find_and_set_bit() perf/arm: use atomic find_bit() API drivers/perf: optimize ali_drw_get_counter_idx() by using find_and_set_bit() dmaengine: idxd: optimize perfmon_assign_event() ath10k: optimize ath10k_snoc_napi_poll() with an atomic iterator wifi: rtw88: optimize the driver by using atomic iterator KVM: x86: hyper-v: optimize and cleanup kvm_hv_process_stimers() PCI: hv: Optimize hv_get_dom_num() by using find_and_set_bit() scsi: core: optimize scsi_evt_emit() by using an atomic iterator scsi: mpi3mr: optimize the driver by using find_and_set_bit() scsi: qedi: optimize qedi_get_task_idx() by using find_and_set_bit() powerpc: optimize arch code by using atomic find_bit() API iommu: optimize subsystem by using atomic find_bit() API media: radio-shark: optimize driver by using atomic find_bit() API sfc: optimize driver by using atomic find_bit() API tty: nozomi: optimize interrupt_handler() usb: cdc-acm: optimize acm_softint() block: null_blk: replace get_tag() with a generic find_and_set_bit_lock() RDMA/rtrs: optimize __rtrs_get_permit() by using find_and_set_bit_lock() mISDN: optimize get_free_devid() media: em28xx: cx231xx: optimize drivers by using find_and_set_bit() ethernet: rocker: optimize ofdpa_port_internal_vlan_id_get() serial: sc12is7xx: optimize sc16is7xx_alloc_line() bluetooth: optimize cmtp_alloc_block_id() net: smc: optimize smc_wr_tx_get_free_slot_index() ALSA: use atomic find_bit() functions where applicable m68k: optimize get_mmu_context() microblaze: optimize get_mmu_context() sh: mach-x3proto: optimize ilsel_enable() arch/m68k/include/asm/mmu_context.h | 11 +- arch/microblaze/include/asm/mmu_context_mm.h | 11 +- arch/mips/sgi-ip30/ip30-irq.c | 12 +- arch/powerpc/mm/book3s32/mmu_context.c | 10 +- arch/powerpc/platforms/pasemi/dma_lib.c | 45 +-- arch/powerpc/platforms/powernv/pci-sriov.c | 12 +- arch/sh/boards/mach-x3proto/ilsel.c | 4 +- arch/sparc/kernel/pci_msi.c | 9 +- arch/x86/kvm/hyperv.c | 40 +-- drivers/block/null_blk/main.c | 41 +-- drivers/dma/idxd/perfmon.c | 8 +- drivers/infiniband/ulp/rtrs/rtrs-clt.c | 15 +- drivers/iommu/arm/arm-smmu/arm-smmu.h | 10 +- drivers/iommu/msm_iommu.c | 18 +- drivers/isdn/mISDN/core.c | 9 +- drivers/media/radio/radio-shark.c | 5 +- drivers/media/radio/radio-shark2.c | 5 +- drivers/media/usb/cx231xx/cx231xx-cards.c | 16 +- drivers/media/usb/em28xx/em28xx-cards.c | 37 +-- drivers/net/ethernet/rocker/rocker_ofdpa.c | 11 +- drivers/net/ethernet/sfc/rx_common.c | 4 +- drivers/net/ethernet/sfc/siena/rx_common.c | 4 +- drivers/net/ethernet/sfc/siena/siena_sriov.c | 14 +- drivers/net/wireless/ath/ath10k/snoc.c | 9 +- drivers/net/wireless/realtek/rtw88/pci.c | 5 +- drivers/net/wireless/realtek/rtw89/pci.c | 5 +- drivers/pci/controller/pci-hyperv.c | 7 +- drivers/perf/alibaba_uncore_drw_pmu.c | 10 +- drivers/perf/arm-cci.c | 24 +- drivers/perf/arm-ccn.c | 10 +- drivers/perf/arm_dmc620_pmu.c | 9 +- drivers/perf/arm_pmuv3.c | 8 +- drivers/scsi/mpi3mr/mpi3mr_os.c | 21 +- drivers/scsi/qedi/qedi_main.c | 9 +- drivers/scsi/scsi_lib.c | 7 +- drivers/tty/nozomi.c | 5 +- drivers/tty/serial/sc16is7xx.c | 8 +- drivers/usb/class/cdc-acm.c | 5 +- include/linux/cpumask.h | 12 + include/linux/find.h | 293 +++++++++++++++++++ kernel/sched/sched.h | 14 +- kernel/watch_queue.c | 6 +- lib/find_bit.c | 85 ++++++ lib/sbitmap.c | 46 +-- lib/test_bitmap.c | 61 ++++ net/bluetooth/cmtp/core.c | 10 +- net/smc/smc_wr.c | 10 +- sound/pci/hda/hda_codec.c | 7 +- sound/usb/caiaq/audio.c | 13 +- 49 files changed, 631 insertions(+), 419 deletions(-) -- 2.40.1

1 4

[RESEND] ASoC: amd: acp: add pm ops support for renoir platform.
by Syed Saba Kareem 15 Dec '23

15 Dec '23

Add pm ops for renoir platform. Signed-off-by: Syed Saba Kareem <Syed.SabaKareem(a)amd.com> --- sound/soc/amd/acp/acp-renoir.c | 37 ++++++++++++++++++++++++++++++++++ 1 file changed, 37 insertions(+) diff --git a/sound/soc/amd/acp/acp-renoir.c b/sound/soc/amd/acp/acp-renoir.c index a591482a0726..b0e181c9a733 100644 --- a/sound/soc/amd/acp/acp-renoir.c +++ b/sound/soc/amd/acp/acp-renoir.c @@ -20,6 +20,7 @@ #include <sound/soc.h> #include <sound/soc-dai.h> #include <linux/dma-mapping.h> +#include <linux/pm_runtime.h> #include "amd.h" #include "acp-mach.h" @@ -196,6 +197,11 @@ static int renoir_audio_probe(struct platform_device *pdev) acp_enable_interrupts(adata); acp_platform_register(dev); + pm_runtime_set_autosuspend_delay(&pdev->dev, ACP_SUSPEND_DELAY_MS); + pm_runtime_use_autosuspend(&pdev->dev); + pm_runtime_mark_last_busy(&pdev->dev); + pm_runtime_set_active(&pdev->dev); + pm_runtime_enable(&pdev->dev); return 0; } @@ -208,11 +214,42 @@ static void renoir_audio_remove(struct platform_device *pdev) acp_platform_unregister(dev); } +static int __maybe_unused rn_pcm_resume(struct device *dev) +{ + struct acp_dev_data *adata = dev_get_drvdata(dev); + struct acp_stream *stream; + struct snd_pcm_substream *substream; + snd_pcm_uframes_t buf_in_frames; + u64 buf_size; + + spin_lock(&adata->acp_lock); + list_for_each_entry(stream, &adata->stream_list, list) { + substream = stream->substream; + if (substream && substream->runtime) { + buf_in_frames = (substream->runtime->buffer_size); + buf_size = frames_to_bytes(substream->runtime, buf_in_frames); + config_pte_for_stream(adata, stream); + config_acp_dma(adata, stream, buf_size); + if (stream->dai_id) + restore_acp_i2s_params(substream, adata, stream); + else + restore_acp_pdm_params(substream, adata); + } + } + spin_unlock(&adata->acp_lock); + return 0; +} + +static const struct dev_pm_ops rn_dma_pm_ops = { + SET_SYSTEM_SLEEP_PM_OPS(NULL, rn_pcm_resume) +}; + static struct platform_driver renoir_driver = { .probe = renoir_audio_probe, .remove_new = renoir_audio_remove, .driver = { .name = "acp_asoc_renoir", + .pm = &rn_dma_pm_ops, }, }; -- 2.25.1

3 2

[PATCH] ASoC: tas2781: check the validity of prm_no/cfg_no
by Gergo Koteles 15 Dec '23

15 Dec '23

Add additional checks for program/config numbers to avoid loading from invalid addresses. If prm_no/cfg_no is negative, skip uploading program/config. The tas2781-hda driver caused a NULL pointer dereference after loading module, and before first runtime_suspend. the state was: tas_priv->cur_conf = -1; tas_priv->tasdevice[i].cur_conf = 0; program = &(tas_fmw->programs[-1]); BUG: kernel NULL pointer dereference, address: 0000000000000010 Call Trace: <TASK> ? __die+0x23/0x70 ? page_fault_oops+0x171/0x4e0 ? vprintk_emit+0x175/0x2b0 ? exc_page_fault+0x7f/0x180 ? asm_exc_page_fault+0x26/0x30 ? tasdevice_load_block_kernel+0x21/0x310 [snd_soc_tas2781_fmwlib] tasdevice_select_tuningprm_cfg+0x268/0x3a0 [snd_soc_tas2781_fmwlib] tasdevice_tuning_switch+0x69/0x710 [snd_soc_tas2781_fmwlib] tas2781_hda_playback_hook+0xd4/0x110 [snd_hda_scodec_tas2781_i2c] Fixes: 915f5eadebd2 ("ASoC: tas2781: firmware lib") CC: stable(a)vger.kernel.org Signed-off-by: Gergo Koteles <soyer(a)irl.hu> --- sound/soc/codecs/tas2781-fmwlib.c | 14 ++++++++------ 1 file changed, 8 insertions(+), 6 deletions(-) diff --git a/sound/soc/codecs/tas2781-fmwlib.c b/sound/soc/codecs/tas2781-fmwlib.c index eb55abae0d7b..1dfac9b2fca2 100644 --- a/sound/soc/codecs/tas2781-fmwlib.c +++ b/sound/soc/codecs/tas2781-fmwlib.c @@ -2219,11 +2219,11 @@ int tasdevice_select_tuningprm_cfg(void *context, int prm_no, goto out; } - conf = &(tas_fmw->configs[cfg_no]); for (i = 0, prog_status = 0; i < tas_priv->ndev; i++) { if (cfg_info[rca_conf_no]->active_dev & (1 << i)) { - if (tas_priv->tasdevice[i].cur_prog != prm_no - || tas_priv->force_fwload_status) { + if (prm_no >= 0 + && (tas_priv->tasdevice[i].cur_prog != prm_no + || tas_priv->force_fwload_status)) { tas_priv->tasdevice[i].cur_conf = -1; tas_priv->tasdevice[i].is_loading = true; prog_status++; @@ -2258,7 +2258,8 @@ int tasdevice_select_tuningprm_cfg(void *context, int prm_no, } for (i = 0, status = 0; i < tas_priv->ndev; i++) { - if (tas_priv->tasdevice[i].cur_conf != cfg_no + if (cfg_no >= 0 + && tas_priv->tasdevice[i].cur_conf != cfg_no && (cfg_info[rca_conf_no]->active_dev & (1 << i)) && (tas_priv->tasdevice[i].is_loaderr == false)) { status++; @@ -2268,6 +2269,7 @@ int tasdevice_select_tuningprm_cfg(void *context, int prm_no, } if (status) { + conf = &(tas_fmw->configs[cfg_no]); status = 0; tasdevice_load_data(tas_priv, &(conf->dev_data)); for (i = 0; i < tas_priv->ndev; i++) { @@ -2311,7 +2313,7 @@ int tasdevice_prmg_load(void *context, int prm_no) } for (i = 0, prog_status = 0; i < tas_priv->ndev; i++) { - if (tas_priv->tasdevice[i].cur_prog != prm_no) { + if (prm_no >= 0 && tas_priv->tasdevice[i].cur_prog != prm_no) { tas_priv->tasdevice[i].cur_conf = -1; tas_priv->tasdevice[i].is_loading = true; prog_status++; @@ -2356,7 +2358,7 @@ int tasdevice_prmg_calibdata_load(void *context, int prm_no) } for (i = 0, prog_status = 0; i < tas_priv->ndev; i++) { - if (tas_priv->tasdevice[i].cur_prog != prm_no) { + if (prm_no >= 0 && tas_priv->tasdevice[i].cur_prog != prm_no) { tas_priv->tasdevice[i].cur_conf = -1; tas_priv->tasdevice[i].is_loading = true; prog_status++; base-commit: ffc253263a1375a65fa6c9f62a893e9767fbebfa -- 2.43.0

2 1

[PATCH] ASoC: fsl_mqs: remove duplicated including
by Wang Jinchao 15 Dec '23

15 Dec '23

rm the second \#include <linux/of.h> Signed-off-by: Wang Jinchao <wangjinchao(a)xfusion.com> --- sound/soc/fsl/fsl_mqs.c | 1 - 1 file changed, 1 deletion(-) diff --git a/sound/soc/fsl/fsl_mqs.c b/sound/soc/fsl/fsl_mqs.c index f2d74ec05cdf..c73e6c141f7e 100644 --- a/sound/soc/fsl/fsl_mqs.c +++ b/sound/soc/fsl/fsl_mqs.c @@ -12,7 +12,6 @@ #include <linux/mfd/syscon/imx6q-iomuxc-gpr.h> #include <linux/of.h> #include <linux/pm_runtime.h> -#include <linux/of.h> #include <linux/pm.h> #include <linux/slab.h> #include <sound/soc.h> -- 2.40.0

1 0

[PATCH v3 0/5] ASoC: codecs: add support for WCD939x Codec
by Neil Armstrong 15 Dec '23

15 Dec '23

Add the main WCD9390/WCD9395 Audio Codec driver to support: - 4 ADC inputs for up to 5 Analog Microphones - 4 DMIC inputs for up to 8 Digital Microphones - 4 Microphone BIAS - Stereo Headphone output - Mono EAR output - MBHC engine for Headset Detection This adds: - bindings - MBHC changes to support Type-C muc - Soundwire Slave driver - Code driver The USB Mux subsystem support will be submitted separalety since it's a functionally separate subsystem connected over I2C with it's own reset signal. Signed-off-by: Neil Armstrong <neil.armstrong(a)linaro.org> --- Changes in v3: - Fixed W=1 and smatch warnings reported by lkp & Dan Carpenter - Fixed dependency on CONFIG_TYPEC and added guards to not build type-c related code when disabled - Collected review on second bindings patch - Link to v2: https://lore.kernel.org/r/20231201-topic-sm8650-upstream-wcd939x-codec-v2-0… Changes in v2: - Bindings: - Dropped all references to "Slave" or "Host" terminology when possible - Collected review for first patch - Added wcd9395 as fallback of wcd9390 - Fixes typos errors - MBHC: - Dropped all references to "Slave" or "Host" terminology when possible - Fixed EXPORT_SYMBOL into EXPORT_SYMBOL_GPL - Fixed typo in commit message - Soundwire Devices driver - Dropped all references to "Slave" or "Host" terminology when possible - Dropped comments and unused code - Reworked wcd939x_swr_get_current_bank() - Added comments to wcd9390_interrupt_callback() - Reworked regmap's wcd939x_readonly/volatile_register checks - Added comments explaining while bind/unbind are empty - Added comment on SDW_SLAVE_ENTRY meaning - Added more register fields defines - Style fixes - Codec driver - Dropped all references to "Slave" or "Host" terminology when possible - Added MICB_BIAS_ values enum and used them in the code - Moved zdet_param to the top - Added TLV data for ear_pa_gain and used it - Defined as much as possible every bit field used on/from register - Replaced 1/0 to true/false when writing to single bit fields - Replaced for loop on all bits with ffs(), simplified code - Simplified MICB fields handling code - Reworked and simplified wcd939x_get/set_compander and other kcontrol get/setters - Reworked and simplified MHGC impedance/zdet/qval code, dropped dead code - Added comments on wcd939x_wd_handle_irq() utility - Added comment on the interrupt handling architecture - I've not moved common code yet, I'll probably do later since it would alter wcd939x code - Link to v1: https://lore.kernel.org/r/20231123-topic-sm8650-upstream-wcd939x-codec-v1-0… --- Neil Armstrong (5): ASoC: dt-bindings: qcom,wcd938x: move out common properties ASoC: dt-bindings: document WCD939x Audio Codec ASoC: codec: wcd-mbhc-v2: add support when connected behind an USB-C audio mux ASoC: codecs: Add WCD939x Soundwire devices driver ASoC: codecs: Add WCD939x Codec driver .../devicetree/bindings/sound/qcom,wcd938x.yaml | 81 +- .../bindings/sound/qcom,wcd939x-sdw.yaml | 69 + .../devicetree/bindings/sound/qcom,wcd939x.yaml | 96 + .../bindings/sound/qcom,wcd93xx-common.yaml | 95 + sound/soc/codecs/Kconfig | 19 + sound/soc/codecs/Makefile | 7 + sound/soc/codecs/wcd-clsh-v2.h | 1 + sound/soc/codecs/wcd-mbhc-v2.c | 95 +- sound/soc/codecs/wcd-mbhc-v2.h | 3 + sound/soc/codecs/wcd939x-sdw.c | 1584 +++++++++ sound/soc/codecs/wcd939x.c | 3666 ++++++++++++++++++++ sound/soc/codecs/wcd939x.h | 983 ++++++ 12 files changed, 6604 insertions(+), 95 deletions(-) --- base-commit: 07b677953b9dca02928be323e2db853511305fa9 change-id: 20231123-topic-sm8650-upstream-wcd939x-codec-c46f621b55d4 Best regards, -- Neil Armstrong <neil.armstrong(a)linaro.org>

2 9

[PATCH] ALSA: hda/tas2781: select program 0, conf 0 by default
by Gergo Koteles 15 Dec '23

15 Dec '23

Currently, cur_prog/cur_conf remains at the default value (-1), while program 0 has been loaded into the amplifiers. In the playback hook, tasdevice_tuning_switch tries to restore the cur_prog/cur_conf. In the runtime_resume/system_resume, tasdevice_prmg_load tries to load the cur_prog as well. Set cur_prog and cur_conf to 0 if available in the firmware. Fixes: 5be27f1e3ec9 ("ALSA: hda/tas2781: Add tas2781 HDA driver") CC: stable(a)vger.kernel.org Signed-off-by: Gergo Koteles <soyer(a)irl.hu> --- sound/pci/hda/tas2781_hda_i2c.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/sound/pci/hda/tas2781_hda_i2c.c b/sound/pci/hda/tas2781_hda_i2c.c index fb802802939e..e974510367e3 100644 --- a/sound/pci/hda/tas2781_hda_i2c.c +++ b/sound/pci/hda/tas2781_hda_i2c.c @@ -543,6 +543,10 @@ static void tasdev_fw_ready(const struct firmware *fmw, void *context) tas_priv->fw_state = TASDEVICE_DSP_FW_ALL_OK; tasdevice_prmg_load(tas_priv, 0); + if (tas_priv->fmw->nr_programs > 0) + tas_priv->cur_prog = 0; + if (tas_priv->fmw->nr_configurations > 0) + tas_priv->cur_conf = 0; /* If calibrated data occurs error, dsp will still works with default * calibrated data inside algo. base-commit: ffc253263a1375a65fa6c9f62a893e9767fbebfa -- 2.43.0

2 1

Getting a strange error code with Behringer UMC1820
by Sebastien Alaiwan 15 Dec '23

15 Dec '23

Hi, I've been running some experiments on the Behringer UMC1820, which is a USB audio interface, supposedly class-compliant. I wrote a tiny program "alsa-sine" that simply plays silence on a given ALSA device ( please find the source for it at the bottom of this message ). This tiny program allows to specify the buffer size, and the period size ( Here, I'll always be using a buffer size of 256 ). Here, i'm testing 5 values for the period size (the last command line argument). All period sizes work, except ... 127. See below: $ ./alsa_sine 'hw:CARD=UMC1820,DEV=0' 256 125 buffer_size=256 (asked 256), period_size=125 (asked 125), channels=12, format=S24_3LE (24 bpp) OK $ ./alsa_sine 'hw:CARD=UMC1820,DEV=0' 256 126 buffer_size=256 (asked 256), period_size=126 (asked 126), channels=12, format=S24_3LE (24 bpp) OK $ ./alsa_sine 'hw:CARD=UMC1820,DEV=0' 256 127 buffer_size=256 (asked 256), period_size=127 (asked 127), channels=12, format=S24_3LE (24 bpp) [alsa_sine.c:66] error: snd_pcm_drain(pcm) : Input/output error Aborting. $ ./alsa_sine 'hw:CARD=UMC1820,DEV=0' 256 128 buffer_size=256 (asked 256), period_size=128 (asked 128), channels=12, format=S24_3LE (24 bpp) OK $ ./alsa_sine 'hw:CARD=UMC1820,DEV=0' 256 129 buffer_size=256 (asked 256), period_size=128 (asked 129), channels=12, format=S24_3LE (24 bpp) OK I also own a Virus TI, which also implements an USB audio interface, also supposedly class-compliant. So I ran the same test on it, and interestingly, there's no error: $ ./alsa_sine 'hw:CARD=TI,DEV=0' 256 125 buffer_size=256 (asked 256), period_size=125 (asked 125), channels=2, format=S16_LE (16 bpp) OK $ ./alsa_sine 'hw:CARD=TI,DEV=0' 256 126 buffer_size=256 (asked 256), period_size=126 (asked 126), channels=2, format=S16_LE (16 bpp) OK $ ./alsa_sine 'hw:CARD=TI,DEV=0' 256 127 buffer_size=256 (asked 256), period_size=127 (asked 127), channels=2, format=S16_LE (16 bpp) OK $ ./alsa_sine 'hw:CARD=TI,DEV=0' 256 128 buffer_size=256 (asked 256), period_size=128 (asked 128), channels=2, format=S16_LE (16 bpp) OK $ ./alsa_sine 'hw:CARD=TI,DEV=0' 256 129 buffer_size=256 (asked 256), period_size=128 (asked 129), channels=2, format=S16_LE (16 bpp) OK So at this point, I'm wondering if I might be doing something wrong. I'm not trying to achieve anything here other than trying to understand what's going on. Maybe the UMC1820 doesn't support a period size of 127 ? ( does it even make sense to say this? ) If so, how is it possible for an application to know about this limitation? Maybe my test program is simply incorrect, and it's behavior is undefined? But then, how can we explain the difference in behavior between the UMC1820 and the Virus TI ? My kernel version is: Linux ANTEC 6.5.0-5-amd64 #1 SMP PREEMPT_DYNAMIC Debian 6.5.13-1 (2023-11-29) x86_64 GNU/Linux Also, Debian tells me I'm using the package: libasound 1.2.10-1 BTW, please let me know if this is not the proper mailing list to talk about such things! Thanks, Sebastien Alaiwan //--------------------------------------------------------------- // gcc alsa_sine.c -lasound #include <alsa/asoundlib.h> #include <stdio.h> const int SampleRate = 48000; void checkCall(int ret, const char* expr, const char* file, int line) { if(ret < 0) { fprintf(stderr, "[%s:%d] error: %s : %s\nAborting.\n", file, line, expr, snd_strerror(ret)); exit(1); } } #define CHECK_CALL(a) checkCall(a, # a, __FILE__, __LINE__) int main(int argc, char* argv[]) { if(argc != 4) { fprintf(stderr, "Usage: %s <device> <buffer_size> <period_size>\n", argv[0]); return 1; } const char* devName = argv[1]; const snd_pcm_uframes_t asked_buffer_size = atoi(argv[2]); const snd_pcm_uframes_t asked_period_size = atoi(argv[3]); snd_pcm_t* pcm = NULL; CHECK_CALL(snd_pcm_open(&pcm, devName, SND_PCM_STREAM_PLAYBACK, 0)); snd_pcm_hw_params_t* hw_params = NULL; CHECK_CALL(snd_pcm_hw_params_malloc(&hw_params)); CHECK_CALL(snd_pcm_hw_params_any(pcm, hw_params)); snd_pcm_format_t fmt; unsigned int channels = 0; CHECK_CALL(snd_pcm_hw_params_get_format(hw_params, &fmt)); CHECK_CALL(snd_pcm_hw_params_get_channels_min(hw_params, &channels)); snd_pcm_uframes_t buffer_size = asked_buffer_size; snd_pcm_uframes_t period_size = asked_period_size; CHECK_CALL(snd_pcm_hw_params_set_rate(pcm, hw_params, SampleRate, 0)); CHECK_CALL(snd_pcm_hw_params_set_access(pcm, hw_params, SND_PCM_ACCESS_RW_INTERLEAVED)); CHECK_CALL(snd_pcm_hw_params_set_buffer_size_near(pcm, hw_params, &buffer_size)); CHECK_CALL(snd_pcm_hw_params_set_period_size_near(pcm, hw_params, &period_size, NULL)); CHECK_CALL(snd_pcm_hw_params(pcm, hw_params)); snd_pcm_hw_params_free(hw_params); printf("buffer_size=%d (asked %d), ", (int)buffer_size, (int)asked_buffer_size); printf("period_size=%d (asked %d), ", (int)period_size, (int)asked_period_size); printf("channels=%d, ", channels); printf("format=%s (%d bpp)\n", snd_pcm_format_name(fmt), snd_pcm_format_width(fmt)); const int totalSamples = SampleRate * 1.0; int bitsPerSample = snd_pcm_format_width(fmt); assert(bitsPerSample % 8 == 0); int size = totalSamples * channels * bitsPerSample / 8; uint8_t* buffer = (uint8_t*)malloc(size); memset(buffer, 0, size); CHECK_CALL(snd_pcm_writei(pcm, buffer, totalSamples)); free(buffer); CHECK_CALL(snd_pcm_drain(pcm)); snd_pcm_close(pcm); printf("OK\n"); return 0; } //---------------------------------------------------------------

2 1

[PATCH v3 1/4] ASoC: qcom: common: Add qcom_snd_tdm_hw_params function
by Jianhua Lu 14 Dec '23

14 Dec '23

Add qcom TDM setup function to support TDM ports for qcom platform. Signed-off-by: Jianhua Lu <lujianhua000(a)gmail.com> --- Changes in v3: 1. new patch 2. split qcom_snd_tdm_hw_params function from [Patch v2 1/2] to here sound/soc/qcom/common.c | 69 +++++++++++++++++++++++++++++++++++++++++ sound/soc/qcom/common.h | 2 ++ 2 files changed, 71 insertions(+) diff --git a/sound/soc/qcom/common.c b/sound/soc/qcom/common.c index 483bbf53a541..c0ab201416ef 100644 --- a/sound/soc/qcom/common.c +++ b/sound/soc/qcom/common.c @@ -5,6 +5,7 @@ #include <dt-bindings/sound/qcom,q6afe.h> #include <linux/module.h> #include <sound/jack.h> +#include <sound/pcm_params.h> #include <linux/input-event-codes.h> #include "common.h" @@ -13,6 +14,8 @@ static const struct snd_soc_dapm_widget qcom_jack_snd_widgets[] = { SND_SOC_DAPM_MIC("Mic Jack", NULL), }; +static unsigned int tdm_slot_offset[8] = { 0, 4, 8, 12, 16, 20, 24, 28 }; + int qcom_snd_parse_of(struct snd_soc_card *card) { struct device_node *np; @@ -239,4 +242,70 @@ int qcom_snd_wcd_jack_setup(struct snd_soc_pcm_runtime *rtd, return 0; } EXPORT_SYMBOL_GPL(qcom_snd_wcd_jack_setup); + +int qcom_snd_tdm_hw_params(struct snd_pcm_substream *substream, + struct snd_pcm_hw_params *params) +{ + struct snd_soc_pcm_runtime *rtd = snd_soc_substream_to_rtd(substream); + struct snd_soc_dai *cpu_dai = snd_soc_rtd_to_cpu(rtd, 0); + int slots = ARRAY_SIZE(tdm_slot_offset); + int channels, slot_width, tx_mask, rx_mask; + int ret; + + switch (params_format(params)) { + case SNDRV_PCM_FORMAT_S16_LE: + slot_width = 16; + break; + case SNDRV_PCM_FORMAT_S24_LE: + slot_width = 24; + break; + case SNDRV_PCM_FORMAT_S32_LE: + slot_width = 32; + break; + default: + dev_err(rtd->dev, "%s: invalid param format 0x%x\n", __func__, + params_format(params)); + return -EINVAL; + } + + channels = params_channels(params); + if (substream->stream == SNDRV_PCM_STREAM_PLAYBACK) { + tx_mask = 0; + rx_mask = BIT(channels) - 1; + ret = snd_soc_dai_set_tdm_slot(cpu_dai, tx_mask, rx_mask, slots, slot_width); + if (ret < 0) { + dev_err(rtd->dev, "%s: failed to set tdm slot, err:%d\n", + __func__, ret); + return ret; + } + + ret = snd_soc_dai_set_channel_map(cpu_dai, 0, NULL, channels, + tdm_slot_offset); + if (ret < 0) { + dev_err(rtd->dev, "%s: failed to set channel map, err:%d\n", + __func__, ret); + return ret; + } + } else { + tx_mask = 0xf; + rx_mask = 0; + ret = snd_soc_dai_set_tdm_slot(cpu_dai, tx_mask, rx_mask, slots, slot_width); + if (ret < 0) { + dev_err(rtd->dev, "%s: failed to set tdm slot, err:%d\n", + __func__, ret); + return ret; + } + + ret = snd_soc_dai_set_channel_map(cpu_dai, channels, + tdm_slot_offset, 0, NULL); + if (ret < 0) { + dev_err(rtd->dev, "%s: failed to set channel map, err:%d\n", + __func__, ret); + return ret; + } + } + + return 0; +} +EXPORT_SYMBOL_GPL(qcom_snd_tdm_hw_params); MODULE_LICENSE("GPL v2"); diff --git a/sound/soc/qcom/common.h b/sound/soc/qcom/common.h index d7f80ee5ae26..b583110f556e 100644 --- a/sound/soc/qcom/common.h +++ b/sound/soc/qcom/common.h @@ -9,5 +9,7 @@ int qcom_snd_parse_of(struct snd_soc_card *card); int qcom_snd_wcd_jack_setup(struct snd_soc_pcm_runtime *rtd, struct snd_soc_jack *jack, bool *jack_setup); +int qcom_snd_tdm_hw_params(struct snd_pcm_substream *substream, + struct snd_pcm_hw_params *params); #endif -- 2.41.0

2 7

[PATCH v4 0/2] ASoC: audio-graph-card2: Introduce playback-only/capture-only DAI link flags
by Daniel Baluta 14 Dec '23

14 Dec '23

From: Daniel Baluta <daniel.baluta(a)nxp.com> This patchseries allows users to specify a link only direction with audio-graph-card2. Changes since v1 - https://lore.kernel.org/lkml/20230801082433.548206-1-daniel.baluta@oss.nxp.… - used audio-graph-card2 instead of simple-card as we are deprecting simple-card. Changes since v2: - Fix device tree binding error reported by Rob Changes since v3: - Add signed-off-by line to patch 1/2 - Enhance commit messsage for patch 1/2 to explain why the patch is needed. Daniel Baluta (2): ASoC: audio-graph-card2: Introduce playback-only/capture-only DAI link flags ASoC: dt-bindings: audio-graph-port: Document new DAI link flags playback-only/capture-only .../bindings/sound/audio-graph-port.yaml | 6 ++++++ include/sound/simple_card_utils.h | 3 +++ sound/soc/generic/audio-graph-card2.c | 6 ++++++ sound/soc/generic/simple-card-utils.c | 19 +++++++++++++++++++ 4 files changed, 34 insertions(+) -- 2.25.1

3 4

[PATCH] ALSA: hda/tas2781: reset the amp before component_add
by Gergo Koteles 14 Dec '23

14 Dec '23

Calling component_add starts loading the firmware, the callback function writes the program to the amplifiers. If the module resets the amplifiers after component_add, it happens that one of the amplifiers does not work because the reset and program writing are interleaving. Call tas2781_reset before component_add to ensure reliable initialization. Fixes: 5be27f1e3ec9 ("ALSA: hda/tas2781: Add tas2781 HDA driver") CC: stable(a)vger.kernel.org Signed-off-by: Gergo Koteles <soyer(a)irl.hu> --- sound/pci/hda/tas2781_hda_i2c.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/sound/pci/hda/tas2781_hda_i2c.c b/sound/pci/hda/tas2781_hda_i2c.c index fb802802939e..0baaaff94c6f 100644 --- a/sound/pci/hda/tas2781_hda_i2c.c +++ b/sound/pci/hda/tas2781_hda_i2c.c @@ -675,14 +675,14 @@ static int tas2781_hda_i2c_probe(struct i2c_client *clt) pm_runtime_put_autosuspend(tas_priv->dev); + tas2781_reset(tas_priv); + ret = component_add(tas_priv->dev, &tas2781_hda_comp_ops); if (ret) { dev_err(tas_priv->dev, "Register component failed: %d\n", ret); pm_runtime_disable(tas_priv->dev); - goto err; } - tas2781_reset(tas_priv); err: if (ret) tas2781_hda_remove(&clt->dev); base-commit: ffc253263a1375a65fa6c9f62a893e9767fbebfa -- 2.43.0

2 1