kernel/linux.git/drivers/soundwire/qcom.c, branch v7.0.11

Convert remaining multi-line kmalloc_obj/flex GFP_KERNEL uses

2026-02-22T16:26:33+00:00

Conversion performed via this Coccinelle script: // SPDX-License-Identifier: GPL-2.0-only // Options: --include-headers-for-types --all-includes --include-headers --keep-comments virtual patch @gfp depends on patch && !(file in "tools") && !(file in "samples")@ identifier ALLOC = {kmalloc_obj,kmalloc_objs,kmalloc_flex, kzalloc_obj,kzalloc_objs,kzalloc_flex, kvmalloc_obj,kvmalloc_objs,kvmalloc_flex, kvzalloc_obj,kvzalloc_objs,kvzalloc_flex}; @@ ALLOC(... - , GFP_KERNEL ) $ make coccicheck MODE=patch COCCI=gfp.cocci Build and boot tested x86_64 with Fedora 42's GCC and Clang: Linux version 6.19.0+ (user@host) (gcc (GCC) 15.2.1 20260123 (Red Hat 15.2.1-7), GNU ld version 2.44-12.fc42) #1 SMP PREEMPT_DYNAMIC 1970-01-01 Linux version 6.19.0+ (user@host) (clang version 20.1.8 (Fedora 20.1.8-4.fc42), LLD 20.1.8) #1 SMP PREEMPT_DYNAMIC 1970-01-01 Signed-off-by: Kees Cook Signed-off-by: Linus Torvalds

treewide: Replace kmalloc with kmalloc_obj for non-scalar types

2026-02-21T09:02:28+00:00

This is the result of running the Coccinelle script from scripts/coccinelle/api/kmalloc_objs.cocci. The script is designed to avoid scalar types (which need careful case-by-case checking), and instead replace kmalloc-family calls that allocate struct or union object instances: Single allocations: kmalloc(sizeof(TYPE), ...) are replaced with: kmalloc_obj(TYPE, ...) Array allocations: kmalloc_array(COUNT, sizeof(TYPE), ...) are replaced with: kmalloc_objs(TYPE, COUNT, ...) Flex array allocations: kmalloc(struct_size(PTR, FAM, COUNT), ...) are replaced with: kmalloc_flex(*PTR, FAM, COUNT, ...) (where TYPE may also be *VAR) The resulting allocations no longer return "void *", instead returning "TYPE *". Signed-off-by: Kees Cook

soundwire: qcom: Use guard to avoid mixing cleanup and goto

2025-12-16T09:10:00+00:00

qcom_swrm_stream_alloc_ports() already uses cleanup.h but also has goto. Such combination is error-prone and discouraged: "... and that the "goto" statement can jump between scopes, the expectation is that usage of "goto" and cleanup helpers is never mixed in the same function." Actually simplify the code with a guard which allows to fix the discouraged style by removing the goto. Signed-off-by: Krzysztof Kozlowski Reviewed-by: Konrad Dybcio Reviewed-by: Srinivas Kandagatla Link: https://patch.msgid.link/20251201102627.146182-2-krzysztof.kozlowski@oss.qualcomm.com Signed-off-by: Vinod Koul

soundwire: qcom: adding support for v3.1.0

2025-12-08T07:07:26+00:00

Add support for controller version v3.1.0, which has changes in register layout and some register fields compared to v2.0. Signed-off-by: Srinivas Kandagatla Tested-by: Alexey Klimov # sm8550 Link: https://patch.msgid.link/20250912083225.228778-8-srinivas.kandagatla@oss.qualcomm.com Signed-off-by: Vinod Koul

soundwire: qcom: prepare for v3.x

2025-12-08T07:07:26+00:00

cleanup the register layout structs to prepare for adding new 3.x controller support. Signed-off-by: Srinivas Kandagatla Tested-by: Alexey Klimov # sm8550 Link: https://patch.msgid.link/20250912083225.228778-6-srinivas.kandagatla@oss.qualcomm.com Signed-off-by: Vinod Koul

soundwire: qcom: deprecate qcom,din/out-ports

2025-12-08T07:07:26+00:00

Number of input and output ports can be dynamically read from the controller registers, getting this value from Device Tree is redundant and potentially lead to bugs. Remove the code parsing this property along with marking this as deprecated in device tree bindings. Signed-off-by: Srinivas Kandagatla Tested-by: Alexey Klimov # sm8550 Link: https://patch.msgid.link/20250912083225.228778-5-srinivas.kandagatla@oss.qualcomm.com Signed-off-by: Vinod Koul

soundwire: qcom: remove unused rd_fifo_depth

2025-12-08T07:07:26+00:00

remove read fifo depth field parsing logic, as rd_fifo_depth is never used in the driver. Cleaning this up would benefit when adding new variant support which includes adding variant fields for rd_fifo_depth. ex: Glymur has this rd_fifo_depth register fields changed from v2.x Signed-off-by: Srinivas Kandagatla Reviewed-by: Konrad Dybcio Tested-by: Alexey Klimov # sm8550 Link: https://patch.msgid.link/20250912083225.228778-3-srinivas.kandagatla@oss.qualcomm.com Signed-off-by: Vinod Koul

soundwire: Use min() to improve code

2025-09-01T17:18:22+00:00

Use min() to reduce the code in qcom_swrm_xfer_msg() and improve its readability. Signed-off-by: Qianfeng Rong Link: https://lore.kernel.org/r/20250814142428.282639-1-rongqianfeng@vivo.com Signed-off-by: Vinod Koul

Merge tag 'soundwire-6.17-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/vkoul/soundwire

2025-08-01T18:09:27+00:00

Pull soundwire updates from Vinod Koul: "A couple of small core changes and driver updates: - Core: handling of nesting irqs to outside the lock, stream parameters handing on port prep failures. - AMD driver support for ACP 7.2 platforms and improved handing of slave alerts and resume sequences - Qualcomm updating driver debug spew - Intel BPT message length limitations, rt721 codec as wake capable etc" * tag 'soundwire-6.17-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/vkoul/soundwire: soundwire: amd: Add support for acp7.2 platform soundwire: stream: restore params when prepare ports fail soundwire: debugfs: move debug statement outside of error handling soundwire: amd: add check for status update registers soundwire: intel_auxdevice: add rt721 codec to wake_capable_list soundwire: Correct some property names soundwire: update Intel BPT message length limitation soundwire: intel_ace2.x: Use str_read_write() helper soundwire: amd: cancel pending slave status handling workqueue during remove sequence soundwire: amd: serialize amd manager resume sequence during pm_prepare soundwire: qcom: demote probe registration printk ASoC: cs42l43: Remove unnecessary work functions soundwire: Move handle_nested_irq outside of sdw_dev_lock MAINTAINERS: Remove Sanyog Kale as reviewer on SoundWire

soundwire: Revert "soundwire: qcom: Add set_channel_map api support"

2025-07-15T12:46:27+00:00

This reverts commit 7796c97df6b1b2206681a07f3c80f6023a6593d5. This patch broke Dragonboard 845c (sdm845). I see: Unexpected kernel BRK exception at EL1 Internal error: BRK handler: 00000000f20003e8 [#1] SMP pc : qcom_swrm_set_channel_map+0x7c/0x80 [soundwire_qcom] lr : snd_soc_dai_set_channel_map+0x34/0x78 Call trace: qcom_swrm_set_channel_map+0x7c/0x80 [soundwire_qcom] (P) sdm845_dai_init+0x18c/0x2e0 [snd_soc_sdm845] snd_soc_link_init+0x28/0x6c snd_soc_bind_card+0x5f4/0xb0c snd_soc_register_card+0x148/0x1a4 devm_snd_soc_register_card+0x50/0xb0 sdm845_snd_platform_probe+0x124/0x148 [snd_soc_sdm845] platform_probe+0x6c/0xd0 really_probe+0xc0/0x2a4 __driver_probe_device+0x7c/0x130 driver_probe_device+0x40/0x118 __device_attach_driver+0xc4/0x108 bus_for_each_drv+0x8c/0xf0 __device_attach+0xa4/0x198 device_initial_probe+0x18/0x28 bus_probe_device+0xb8/0xbc deferred_probe_work_func+0xac/0xfc process_one_work+0x244/0x658 worker_thread+0x1b4/0x360 kthread+0x148/0x228 ret_from_fork+0x10/0x20 Kernel panic - not syncing: BRK handler: Fatal exception Dan has also reported following issues with the original patch https://lore.kernel.org/all/33fe8fe7-719a-405a-9ed2-d9f816ce1d57@sabinyo.mountain/ Bug #1: The zeroeth element of ctrl->pconfig[] is supposed to be unused. We start counting at 1. However this code sets ctrl->pconfig[0].ch_mask = 128. Bug #2: There are SLIM_MAX_TX_PORTS (16) elements in tx_ch[] array but only QCOM_SDW_MAX_PORTS + 1 (15) in the ctrl->pconfig[] array so it corrupts memory like Yongqin Liu pointed out. Bug 3: Like Jie Gan pointed out, it erases all the tx information with the rx information. Cc: stable@vger.kernel.org # v6.15+ Signed-off-by: Amit Pundir Acked-by: Srinivas Kandagatla Link: https://lore.kernel.org/r/20250709174949.8541-1-amit.pundir@linaro.org Signed-off-by: Vinod Koul