6.19: apply dmacoherent patch for arm64 by nicholasaiello · Pull Request #6 · nicholasaiello/linux

nicholasaiello · 2026-01-05T03:30:04Z

No description provided.

Copilot

Pull request overview

This PR applies DMA coherent patches to support ARM64 architecture, specifically addressing cache coherency issues in GPU memory management for ARM64 platforms (BCM2711 and BCM2712).

Key changes:

Adds ARM64-specific DMA coherent page protection handling in TTM memory management
Disables single-page cached mapping optimization on ARM64 to avoid coherency issues
Enables AMD Radeon/AMDGPU, Intel Xe, and HDA audio driver support in ARM64 defconfigs

Reviewed changes

Copilot reviewed 6 out of 6 changed files in this pull request and generated 2 comments.

Show a summary per file

File	Description
drivers/gpu/drm/ttm/ttm_module.c	Applies pgprot_dmacoherent() for cached mappings on ARM64 to ensure proper cache coherency
drivers/gpu/drm/ttm/ttm_bo_util.c	Disables single-page kmap optimization on ARM64 to force vmap with proper page protection
drivers/gpu/drm/i915/display/intel_vga.c	Guards VGA I/O operations with CONFIG_VGA_CONSOLE to prevent issues when VGA console is disabled
drivers/gpu/drm/i915/display/intel_display.c	Guards legacy cursor update path with CONFIG_VGA_CONSOLE (incorrectly)
arch/arm64/configs/bcm2712_defconfig	Enables AMD GPU drivers (Radeon, AMDGPU), Intel Xe, and HDA audio codec support
arch/arm64/configs/bcm2711_defconfig	Enables AMD GPU drivers (Radeon, AMDGPU), Intel Xe, and HDA audio codec support

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2026-01-05T03:33:36Z

+#ifdef CONFIG_ARM64
+		return pgprot_dmacoherent(tmp);
+#else
 		return tmp;
+#endif


Inconsistent ARM64 detection method. This code uses CONFIG_ARM64 to detect ARM64 architecture, but elsewhere in the same file (line 80) the code uses aarch64 (the compiler-defined macro) for the same purpose. For consistency and to align with the existing pattern in this file, consider using aarch64 instead of CONFIG_ARM64, or ensure there's a specific reason for using the config option here.

[ Upstream commit 163e5f2 ] When using perf record with the `--overwrite` option, a segmentation fault occurs if an event fails to open. For example: perf record -e cycles-ct -F 1000 -a --overwrite Error: cycles-ct:H: PMU Hardware doesn't support sampling/overflow-interrupts. Try 'perf stat' perf: Segmentation fault #0 0x6466b6 in dump_stack debug.c:366 #1 0x646729 in sighandler_dump_stack debug.c:378 #2 0x453fd1 in sigsegv_handler builtin-record.c:722 #3 0x7f8454e65090 in __restore_rt libc-2.32.so[54090] #4 0x6c5671 in __perf_event__synthesize_id_index synthetic-events.c:1862 #5 0x6c5ac0 in perf_event__synthesize_id_index synthetic-events.c:1943 #6 0x458090 in record__synthesize builtin-record.c:2075 raspberrypi#7 0x45a85a in __cmd_record builtin-record.c:2888 raspberrypi#8 0x45deb6 in cmd_record builtin-record.c:4374 raspberrypi#9 0x4e5e33 in run_builtin perf.c:349 raspberrypi#10 0x4e60bf in handle_internal_command perf.c:401 raspberrypi#11 0x4e6215 in run_argv perf.c:448 raspberrypi#12 0x4e653a in main perf.c:555 raspberrypi#13 0x7f8454e4fa72 in __libc_start_main libc-2.32.so[3ea72] raspberrypi#14 0x43a3ee in _start ??:0 The --overwrite option implies --tail-synthesize, which collects non-sample events reflecting the system status when recording finishes. However, when evsel opening fails (e.g., unsupported event 'cycles-ct'), session->evlist is not initialized and remains NULL. The code unconditionally calls record__synthesize() in the error path, which iterates through the NULL evlist pointer and causes a segfault. To fix it, move the record__synthesize() call inside the error check block, so it's only called when there was no error during recording, ensuring that evlist is properly initialized. Fixes: 4ea648a ("perf record: Add --tail-synthesize option") Signed-off-by: Shuai Xue <xueshuai@linux.alibaba.com> Signed-off-by: Namhyung Kim <namhyung@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>

[ Upstream commit 2939203 ] In the current implementation, the enetc_xdp_xmit() always transmits redirected XDP frames even if the link is down, but the frames cannot be transmitted from TX BD rings when the link is down, so the frames are still kept in the TX BD rings. If the XDP program is uninstalled, users will see the following warning logs. fsl_enetc 0000:00:00.0 eno0: timeout for tx ring #6 clear More worse, the TX BD ring cannot work properly anymore, because the HW PIR and CIR are not equal after the re-initialization of the TX BD ring. At this point, the BDs between CIR and PIR are invalid, which will cause a hardware malfunction. Another reason is that there is internal context in the ring prefetch logic that will retain the state from the first incarnation of the ring and continue prefetching from the stale location when we re-initialize the ring. The internal context is only reset by an FLR. That is to say, for LS1028A ENETC, software cannot set the HW CIR and PIR when initializing the TX BD ring. It does not make sense to transmit redirected XDP frames when the link is down. Add a link status check to prevent transmission in this condition. This fixes part of the issue, but more complex cases remain. For example, the TX BD ring may still contain unsent frames when the link goes down. Those situations require additional patches, which will build on this one. Fixes: 9d2b68c ("net: enetc: add support for XDP_REDIRECT") Signed-off-by: Wei Fang <wei.fang@nxp.com> Reviewed-by: Frank Li <Frank.Li@nxp.com> Reviewed-by: Hariprasad Kelam <hkelam@marvell.com> Reviewed-by: Vladimir Oltean <vladimir.oltean@nxp.com> Link: https://patch.msgid.link/20251211020919.121113-1-wei.fang@nxp.com Signed-off-by: Paolo Abeni <pabeni@redhat.com> Signed-off-by: Sasha Levin <sashal@kernel.org>

commit c943bfc upstream. After a copy pair swap the block device's "device" symlink points to the secondary CCW device, but the gendisk's parent remained the primary, leaving /sys/block/<dasdx> under the wrong parent. Move the gendisk to the secondary's device with device_move(), keeping the sysfs topology consistent after the swap. Fixes: 413862c ("s390/dasd: add copy pair swap capability") Cc: stable@vger.kernel.org #6.1 Reviewed-by: Jan Hoeppner <hoeppner@linux.ibm.com> Signed-off-by: Stefan Haberland <sth@linux.ibm.com> Signed-off-by: Jens Axboe <axboe@kernel.dk> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

In the current implementation, the enetc_xdp_xmit() always transmits redirected XDP frames even if the link is down, but the frames cannot be transmitted from TX BD rings when the link is down, so the frames are still kept in the TX BD rings. If the XDP program is uninstalled, users will see the following warning logs. fsl_enetc 0000:00:00.0 eno0: timeout for tx ring #6 clear More worse, the TX BD ring cannot work properly anymore, because the HW PIR and CIR are not equal after the re-initialization of the TX BD ring. At this point, the BDs between CIR and PIR are invalid, which will cause a hardware malfunction. Another reason is that there is internal context in the ring prefetch logic that will retain the state from the first incarnation of the ring and continue prefetching from the stale location when we re-initialize the ring. The internal context is only reset by an FLR. That is to say, for LS1028A ENETC, software cannot set the HW CIR and PIR when initializing the TX BD ring. It does not make sense to transmit redirected XDP frames when the link is down. Add a link status check to prevent transmission in this condition. This fixes part of the issue, but more complex cases remain. For example, the TX BD ring may still contain unsent frames when the link goes down. Those situations require additional patches, which will build on this one. Fixes: 9d2b68c ("net: enetc: add support for XDP_REDIRECT") Signed-off-by: Wei Fang <wei.fang@nxp.com> Reviewed-by: Frank Li <Frank.Li@nxp.com> Reviewed-by: Hariprasad Kelam <hkelam@marvell.com> Reviewed-by: Vladimir Oltean <vladimir.oltean@nxp.com> Link: https://patch.msgid.link/20251211020919.121113-1-wei.fang@nxp.com Signed-off-by: Paolo Abeni <pabeni@redhat.com>

…itives The "valid" readout delay between the two reads of the watchdog is larger than the valid delta between the resulting watchdog and clocksource intervals, which results in false positive watchdog results. Assume TSC is the clocksource and HPET is the watchdog and both have a uncertainty margin of 250us (default). The watchdog readout does: 1) wdnow = read(HPET); 2) csnow = read(TSC); 3) wdend = read(HPET); The valid window for the delta between #1 and #3 is calculated by the uncertainty margins of the watchdog and the clocksource: m = 2 * watchdog.uncertainty_margin + cs.uncertainty margin; which results in 750us for the TSC/HPET case. The actual interval comparison uses a smaller margin: m = watchdog.uncertainty_margin + cs.uncertainty margin; which results in 500us for the TSC/HPET case. That means the following scenario will trigger the watchdog: Watchdog cycle N: 1) wdnow[N] = read(HPET); 2) csnow[N] = read(TSC); 3) wdend[N] = read(HPET); Assume the delay between #1 and #2 is 100us and the delay between #1 and Watchdog cycle N + 1: 4) wdnow[N + 1] = read(HPET); 5) csnow[N + 1] = read(TSC); 6) wdend[N + 1] = read(HPET); If the delay between #4 and #6 is within the 750us margin then any delay between #4 and #5 which is larger than 600us will fail the interval check and mark the TSC unstable because the intervals are calculated against the previous value: wd_int = wdnow[N + 1] - wdnow[N]; cs_int = csnow[N + 1] - csnow[N]; Putting the above delays in place this results in: cs_int = (wdnow[N + 1] + 610us) - (wdnow[N] + 100us); -> cs_int = wd_int + 510us; which is obviously larger than the allowed 500us margin and results in marking TSC unstable. Fix this by using the same margin as the interval comparison. If the delay between two watchdog reads is larger than that, then the readout was either disturbed by interconnect congestion, NMIs or SMIs. Fixes: 4ac1dd3 ("clocksource: Set cs_watchdog_read() checks based on .uncertainty_margin") Reported-by: Daniel J Blueman <daniel@quora.org> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Reviewed-by: Paul E. McKenney <paulmck@kernel.org> Tested-by: Paul E. McKenney <paulmck@kernel.org> Link: https://lore.kernel.org/lkml/20250602223251.496591-1-daniel@quora.org/ Link: https://patch.msgid.link/87bjjxc9dq.ffs@tglx

…itives [ Upstream commit c06343b ] The "valid" readout delay between the two reads of the watchdog is larger than the valid delta between the resulting watchdog and clocksource intervals, which results in false positive watchdog results. Assume TSC is the clocksource and HPET is the watchdog and both have a uncertainty margin of 250us (default). The watchdog readout does: 1) wdnow = read(HPET); 2) csnow = read(TSC); 3) wdend = read(HPET); The valid window for the delta between #1 and #3 is calculated by the uncertainty margins of the watchdog and the clocksource: m = 2 * watchdog.uncertainty_margin + cs.uncertainty margin; which results in 750us for the TSC/HPET case. The actual interval comparison uses a smaller margin: m = watchdog.uncertainty_margin + cs.uncertainty margin; which results in 500us for the TSC/HPET case. That means the following scenario will trigger the watchdog: Watchdog cycle N: 1) wdnow[N] = read(HPET); 2) csnow[N] = read(TSC); 3) wdend[N] = read(HPET); Assume the delay between #1 and #2 is 100us and the delay between #1 and Watchdog cycle N + 1: 4) wdnow[N + 1] = read(HPET); 5) csnow[N + 1] = read(TSC); 6) wdend[N + 1] = read(HPET); If the delay between #4 and #6 is within the 750us margin then any delay between #4 and #5 which is larger than 600us will fail the interval check and mark the TSC unstable because the intervals are calculated against the previous value: wd_int = wdnow[N + 1] - wdnow[N]; cs_int = csnow[N + 1] - csnow[N]; Putting the above delays in place this results in: cs_int = (wdnow[N + 1] + 610us) - (wdnow[N] + 100us); -> cs_int = wd_int + 510us; which is obviously larger than the allowed 500us margin and results in marking TSC unstable. Fix this by using the same margin as the interval comparison. If the delay between two watchdog reads is larger than that, then the readout was either disturbed by interconnect congestion, NMIs or SMIs. Fixes: 4ac1dd3 ("clocksource: Set cs_watchdog_read() checks based on .uncertainty_margin") Reported-by: Daniel J Blueman <daniel@quora.org> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Reviewed-by: Paul E. McKenney <paulmck@kernel.org> Tested-by: Paul E. McKenney <paulmck@kernel.org> Link: https://lore.kernel.org/lkml/20250602223251.496591-1-daniel@quora.org/ Link: https://patch.msgid.link/87bjjxc9dq.ffs@tglx Signed-off-by: Sasha Levin <sashal@kernel.org>

[ Upstream commit a70493e ] The ETM decoder incorrectly assumed that auxtrace queue indices were equivalent to CPU number. This assumption is used for inserting records into the queue, and for fetching queues when given a CPU number. This assumption held when Perf always opened a dummy event on every CPU, even if the user provided a subset of CPUs on the commandline, resulting in the indices aligning. For example: # event : name = cs_etm//u, , id = { 2451, 2452 }, type = 11 (cs_etm), size = 136, config = 0x4010, { sample_period, samp> # event : name = dummy:u, , id = { 2453, 2454, 2455, 2456 }, type = 1 (PERF_TYPE_SOFTWARE), size = 136, config = 0x9 (PER> 0 0 0x200 [0xd0]: PERF_RECORD_ID_INDEX nr: 6 ... id: 2451 idx: 2 cpu: 2 tid: -1 ... id: 2452 idx: 3 cpu: 3 tid: -1 ... id: 2453 idx: 0 cpu: 0 tid: -1 ... id: 2454 idx: 1 cpu: 1 tid: -1 ... id: 2455 idx: 2 cpu: 2 tid: -1 ... id: 2456 idx: 3 cpu: 3 tid: -1 Since commit 811082e ("perf parse-events: Support user CPUs mixed with threads/processes") the dummy event no longer behaves in this way, making the ETM event indices start from 0 on the first CPU recorded regardless of its ID: # event : name = cs_etm//u, , id = { 771, 772 }, type = 11 (cs_etm), size = 144, config = 0x4010, { sample_period, sample> # event : name = dummy:u, , id = { 773, 774 }, type = 1 (PERF_TYPE_SOFTWARE), size = 144, config = 0x9 (PERF_COUNT_SW_DUM> 0 0 0x200 [0x90]: PERF_RECORD_ID_INDEX nr: 4 ... id: 771 idx: 0 cpu: 2 tid: -1 ... id: 772 idx: 1 cpu: 3 tid: -1 ... id: 773 idx: 0 cpu: 2 tid: -1 ... id: 774 idx: 1 cpu: 3 tid: -1 This causes the following segfault when decoding: $ perf record -e cs_etm//u -C 2,3 -- true $ perf report perf: Segmentation fault -------- backtrace -------- #0 0xaaaabf9fd020 in ui__signal_backtrace setup.c:110 #1 0xffffab5c7930 in __kernel_rt_sigreturn [vdso][930] #2 0xaaaabfb68d30 in cs_etm_decoder__reset cs-etm-decoder.c:85 #3 0xaaaabfb65930 in cs_etm__get_data_block cs-etm.c:2032 #4 0xaaaabfb666fc in cs_etm__run_per_cpu_timeless_decoder cs-etm.c:2551 #5 0xaaaabfb6692c in (cs_etm__process_timeless_queues cs-etm.c:2612 #6 0xaaaabfb63390 in cs_etm__flush_events cs-etm.c:921 raspberrypi#7 0xaaaabfb324c0 in auxtrace__flush_events auxtrace.c:2915 raspberrypi#8 0xaaaabfaac378 in __perf_session__process_events session.c:2285 raspberrypi#9 0xaaaabfaacc9c in perf_session__process_events session.c:2442 raspberrypi#10 0xaaaabf8d3d90 in __cmd_report builtin-report.c:1085 raspberrypi#11 0xaaaabf8d6944 in cmd_report builtin-report.c:1866 raspberrypi#12 0xaaaabf95ebfc in run_builtin perf.c:351 raspberrypi#13 0xaaaabf95eeb0 in handle_internal_command perf.c:404 raspberrypi#14 0xaaaabf95f068 in run_argv perf.c:451 raspberrypi#15 0xaaaabf95f390 in main perf.c:558 raspberrypi#16 0xffffaab97400 in __libc_start_call_main libc_start_call_main.h:74 raspberrypi#17 0xffffaab974d8 in __libc_start_main@@GLIBC_2.34 libc-start.c:128 raspberrypi#18 0xaaaabf8aa8f0 in _start perf[7a8f0] Fix it by inserting into the queues based on CPU number, rather than using the index. Fixes: 811082e ("perf parse-events: Support user CPUs mixed with threads/processes") Signed-off-by: James Clark <james.clark@linaro.org> Tested-by: Leo Yan <leo.yan@arm.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: coresight@lists.linaro.org Cc: Ian Rogers <irogers@google.com> Cc: Ingo Molnar <mingo@redhat.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: John Garry <john.g.garry@oracle.com> Cc: Mark Rutland <mark.rutland@arm.com> Cc: Mike Leach <mike.leach@linaro.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Suzuki Poulouse <suzuki.poulose@arm.com> Cc: Thomas Falcon <thomas.falcon@intel.com> Cc: Will Deacon <will@kernel.org> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: Sasha Levin <sashal@kernel.org>

[ Upstream commit d935187 ] A potential circular locking dependency (ABBA deadlock) exists between `ec_dev->lock` and the clock framework's `prepare_lock`. The first order (A -> B) occurs when scp_ipi_send() is called while `ec_dev->lock` is held (e.g., within cros_ec_cmd_xfer()): 1. cros_ec_cmd_xfer() acquires `ec_dev->lock` and calls scp_ipi_send(). 2. scp_ipi_send() calls clk_prepare_enable(), which acquires `prepare_lock`. See #0 in the following example calling trace. (Lock Order: `ec_dev->lock` -> `prepare_lock`) The reverse order (B -> A) is more complex and has been observed (learned) by lockdep. It involves the clock prepare operation triggering power domain changes, which then propagates through sysfs and power supply uevents, eventually calling back into the ChromeOS EC driver and attempting to acquire `ec_dev->lock`: 1. Something calls clk_prepare(), which acquires `prepare_lock`. It then triggers genpd operations like genpd_runtime_resume(), which takes `&genpd->mlock`. 2. Power domain changes can trigger regulator changes; regulator changes can then trigger device link changes; device link changes can then trigger sysfs changes. Eventually, power_supply_uevent() is called. 3. This leads to calls like cros_usbpd_charger_get_prop(), which calls cros_ec_cmd_xfer_status(), which then attempts to acquire `ec_dev->lock`. See #1 ~ #6 in the following example calling trace. (Lock Order: `prepare_lock` -> `&genpd->mlock` -> ... -> `&ec_dev->lock`) Move the clk_prepare()/clk_unprepare() operations for `scp->clk` to the remoteproc prepare()/unprepare() callbacks. This ensures `prepare_lock` is only acquired in prepare()/unprepare() callbacks. Since `ec_dev->lock` is not involved in the callbacks, the dependency loop is broken. This means the clock is always "prepared" when the SCP is running. The prolonged "prepared time" for the clock should be acceptable as SCP is designed to be a very power efficient processor. The power consumption impact can be negligible. A simplified calling trace reported by lockdep: > -> #6 (&ec_dev->lock) > cros_ec_cmd_xfer > cros_ec_cmd_xfer_status > cros_usbpd_charger_get_port_status > cros_usbpd_charger_get_prop > power_supply_get_property > power_supply_show_property > power_supply_uevent > dev_uevent > uevent_show > dev_attr_show > sysfs_kf_seq_show > kernfs_seq_show > -> #5 (kn->active#2) > kernfs_drain > __kernfs_remove > kernfs_remove_by_name_ns > sysfs_remove_file_ns > device_del > __device_link_del > device_links_driver_bound > -> #4 (device_links_lock) > device_link_remove > _regulator_put > regulator_put > -> #3 (regulator_list_mutex) > regulator_lock_dependent > regulator_disable > scpsys_power_off > _genpd_power_off > genpd_power_off > -> #2 (&genpd->mlock/1) > genpd_add_subdomain > pm_genpd_add_subdomain > scpsys_add_subdomain > scpsys_probe > -> #1 (&genpd->mlock) > genpd_runtime_resume > __rpm_callback > rpm_callback > rpm_resume > __pm_runtime_resume > clk_core_prepare > clk_prepare > -> #0 (prepare_lock) > clk_prepare > scp_ipi_send > scp_send_ipi > mtk_rpmsg_send > rpmsg_send > cros_ec_pkt_xfer_rpmsg Signed-off-by: Tzung-Bi Shih <tzungbi@kernel.org> Reviewed-by: Chen-Yu Tsai <wenst@chromium.org> Tested-by: Chen-Yu Tsai <wenst@chromium.org> Link: https://lore.kernel.org/r/20260112110755.2435899-1-tzungbi@kernel.org Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org> Signed-off-by: Sasha Levin <sashal@kernel.org>

[ Upstream commit a70493e ] The ETM decoder incorrectly assumed that auxtrace queue indices were equivalent to CPU number. This assumption is used for inserting records into the queue, and for fetching queues when given a CPU number. This assumption held when Perf always opened a dummy event on every CPU, even if the user provided a subset of CPUs on the commandline, resulting in the indices aligning. For example: # event : name = cs_etm//u, , id = { 2451, 2452 }, type = 11 (cs_etm), size = 136, config = 0x4010, { sample_period, samp> # event : name = dummy:u, , id = { 2453, 2454, 2455, 2456 }, type = 1 (PERF_TYPE_SOFTWARE), size = 136, config = 0x9 (PER> 0 0 0x200 [0xd0]: PERF_RECORD_ID_INDEX nr: 6 ... id: 2451 idx: 2 cpu: 2 tid: -1 ... id: 2452 idx: 3 cpu: 3 tid: -1 ... id: 2453 idx: 0 cpu: 0 tid: -1 ... id: 2454 idx: 1 cpu: 1 tid: -1 ... id: 2455 idx: 2 cpu: 2 tid: -1 ... id: 2456 idx: 3 cpu: 3 tid: -1 Since commit 811082e ("perf parse-events: Support user CPUs mixed with threads/processes") the dummy event no longer behaves in this way, making the ETM event indices start from 0 on the first CPU recorded regardless of its ID: # event : name = cs_etm//u, , id = { 771, 772 }, type = 11 (cs_etm), size = 144, config = 0x4010, { sample_period, sample> # event : name = dummy:u, , id = { 773, 774 }, type = 1 (PERF_TYPE_SOFTWARE), size = 144, config = 0x9 (PERF_COUNT_SW_DUM> 0 0 0x200 [0x90]: PERF_RECORD_ID_INDEX nr: 4 ... id: 771 idx: 0 cpu: 2 tid: -1 ... id: 772 idx: 1 cpu: 3 tid: -1 ... id: 773 idx: 0 cpu: 2 tid: -1 ... id: 774 idx: 1 cpu: 3 tid: -1 This causes the following segfault when decoding: $ perf record -e cs_etm//u -C 2,3 -- true $ perf report perf: Segmentation fault -------- backtrace -------- #0 0xaaaabf9fd020 in ui__signal_backtrace setup.c:110 #1 0xffffab5c7930 in __kernel_rt_sigreturn [vdso][930] #2 0xaaaabfb68d30 in cs_etm_decoder__reset cs-etm-decoder.c:85 #3 0xaaaabfb65930 in cs_etm__get_data_block cs-etm.c:2032 #4 0xaaaabfb666fc in cs_etm__run_per_cpu_timeless_decoder cs-etm.c:2551 #5 0xaaaabfb6692c in (cs_etm__process_timeless_queues cs-etm.c:2612 #6 0xaaaabfb63390 in cs_etm__flush_events cs-etm.c:921 raspberrypi#7 0xaaaabfb324c0 in auxtrace__flush_events auxtrace.c:2915 raspberrypi#8 0xaaaabfaac378 in __perf_session__process_events session.c:2285 raspberrypi#9 0xaaaabfaacc9c in perf_session__process_events session.c:2442 raspberrypi#10 0xaaaabf8d3d90 in __cmd_report builtin-report.c:1085 raspberrypi#11 0xaaaabf8d6944 in cmd_report builtin-report.c:1866 raspberrypi#12 0xaaaabf95ebfc in run_builtin perf.c:351 raspberrypi#13 0xaaaabf95eeb0 in handle_internal_command perf.c:404 raspberrypi#14 0xaaaabf95f068 in run_argv perf.c:451 raspberrypi#15 0xaaaabf95f390 in main perf.c:558 raspberrypi#16 0xffffaab97400 in __libc_start_call_main libc_start_call_main.h:74 raspberrypi#17 0xffffaab974d8 in __libc_start_main@@GLIBC_2.34 libc-start.c:128 raspberrypi#18 0xaaaabf8aa8f0 in _start perf[7a8f0] Fix it by inserting into the queues based on CPU number, rather than using the index. Fixes: 811082e ("perf parse-events: Support user CPUs mixed with threads/processes") Signed-off-by: James Clark <james.clark@linaro.org> Tested-by: Leo Yan <leo.yan@arm.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: coresight@lists.linaro.org Cc: Ian Rogers <irogers@google.com> Cc: Ingo Molnar <mingo@redhat.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: John Garry <john.g.garry@oracle.com> Cc: Mark Rutland <mark.rutland@arm.com> Cc: Mike Leach <mike.leach@linaro.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Suzuki Poulouse <suzuki.poulose@arm.com> Cc: Thomas Falcon <thomas.falcon@intel.com> Cc: Will Deacon <will@kernel.org> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: Sasha Levin <sashal@kernel.org>

[ Upstream commit d935187 ] A potential circular locking dependency (ABBA deadlock) exists between `ec_dev->lock` and the clock framework's `prepare_lock`. The first order (A -> B) occurs when scp_ipi_send() is called while `ec_dev->lock` is held (e.g., within cros_ec_cmd_xfer()): 1. cros_ec_cmd_xfer() acquires `ec_dev->lock` and calls scp_ipi_send(). 2. scp_ipi_send() calls clk_prepare_enable(), which acquires `prepare_lock`. See #0 in the following example calling trace. (Lock Order: `ec_dev->lock` -> `prepare_lock`) The reverse order (B -> A) is more complex and has been observed (learned) by lockdep. It involves the clock prepare operation triggering power domain changes, which then propagates through sysfs and power supply uevents, eventually calling back into the ChromeOS EC driver and attempting to acquire `ec_dev->lock`: 1. Something calls clk_prepare(), which acquires `prepare_lock`. It then triggers genpd operations like genpd_runtime_resume(), which takes `&genpd->mlock`. 2. Power domain changes can trigger regulator changes; regulator changes can then trigger device link changes; device link changes can then trigger sysfs changes. Eventually, power_supply_uevent() is called. 3. This leads to calls like cros_usbpd_charger_get_prop(), which calls cros_ec_cmd_xfer_status(), which then attempts to acquire `ec_dev->lock`. See #1 ~ #6 in the following example calling trace. (Lock Order: `prepare_lock` -> `&genpd->mlock` -> ... -> `&ec_dev->lock`) Move the clk_prepare()/clk_unprepare() operations for `scp->clk` to the remoteproc prepare()/unprepare() callbacks. This ensures `prepare_lock` is only acquired in prepare()/unprepare() callbacks. Since `ec_dev->lock` is not involved in the callbacks, the dependency loop is broken. This means the clock is always "prepared" when the SCP is running. The prolonged "prepared time" for the clock should be acceptable as SCP is designed to be a very power efficient processor. The power consumption impact can be negligible. A simplified calling trace reported by lockdep: > -> #6 (&ec_dev->lock) > cros_ec_cmd_xfer > cros_ec_cmd_xfer_status > cros_usbpd_charger_get_port_status > cros_usbpd_charger_get_prop > power_supply_get_property > power_supply_show_property > power_supply_uevent > dev_uevent > uevent_show > dev_attr_show > sysfs_kf_seq_show > kernfs_seq_show > -> #5 (kn->active#2) > kernfs_drain > __kernfs_remove > kernfs_remove_by_name_ns > sysfs_remove_file_ns > device_del > __device_link_del > device_links_driver_bound > -> #4 (device_links_lock) > device_link_remove > _regulator_put > regulator_put > -> #3 (regulator_list_mutex) > regulator_lock_dependent > regulator_disable > scpsys_power_off > _genpd_power_off > genpd_power_off > -> #2 (&genpd->mlock/1) > genpd_add_subdomain > pm_genpd_add_subdomain > scpsys_add_subdomain > scpsys_probe > -> #1 (&genpd->mlock) > genpd_runtime_resume > __rpm_callback > rpm_callback > rpm_resume > __pm_runtime_resume > clk_core_prepare > clk_prepare > -> #0 (prepare_lock) > clk_prepare > scp_ipi_send > scp_send_ipi > mtk_rpmsg_send > rpmsg_send > cros_ec_pkt_xfer_rpmsg Signed-off-by: Tzung-Bi Shih <tzungbi@kernel.org> Reviewed-by: Chen-Yu Tsai <wenst@chromium.org> Tested-by: Chen-Yu Tsai <wenst@chromium.org> Link: https://lore.kernel.org/r/20260112110755.2435899-1-tzungbi@kernel.org Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org> Signed-off-by: Sasha Levin <sashal@kernel.org>

nicholasaiello self-assigned this Jan 5, 2026

Copilot AI review requested due to automatic review settings January 5, 2026 03:30

Copilot started reviewing on behalf of nicholasaiello January 5, 2026 03:31 View session

Copilot AI reviewed Jan 5, 2026

View reviewed changes

nicholasaiello force-pushed the rpi-6.19.y-drm-ttm branch from 29e03d8 to 269047e Compare January 10, 2026 18:04

nicholasaiello force-pushed the rpi-6.19.y branch from 6d7283b to 92e7cca Compare January 10, 2026 18:04

nicholasaiello force-pushed the rpi-6.19.y-drm-ttm branch from 269047e to 6902be0 Compare January 12, 2026 11:51

nicholasaiello force-pushed the rpi-6.19.y branch from 92e7cca to 12e23f2 Compare January 12, 2026 11:51

nicholasaiello force-pushed the rpi-6.19.y-drm-ttm branch 2 times, most recently from 547c3ec to 79e8531 Compare January 23, 2026 15:21

nicholasaiello force-pushed the rpi-6.19.y branch from c61a972 to 808defe Compare January 23, 2026 15:21

nicholasaiello force-pushed the rpi-6.19.y-drm-ttm branch from 79e8531 to bcf93f0 Compare February 1, 2026 23:06

nicholasaiello force-pushed the rpi-6.19.y branch from 808defe to 7bf0b49 Compare February 1, 2026 23:06

nicholasaiello force-pushed the rpi-6.19.y-drm-ttm branch from bcf93f0 to 0181f31 Compare February 6, 2026 13:19

nicholasaiello force-pushed the rpi-6.19.y branch from 7bf0b49 to 266a5aa Compare February 6, 2026 13:19

nicholasaiello force-pushed the rpi-6.19.y-drm-ttm branch from 0181f31 to 94377e7 Compare February 21, 2026 17:03

nicholasaiello force-pushed the rpi-6.19.y branch from 266a5aa to a92e0e5 Compare February 21, 2026 17:03

nicholasaiello force-pushed the rpi-6.19.y-drm-ttm branch from 94377e7 to f86fb7b Compare March 4, 2026 13:30

nicholasaiello force-pushed the rpi-6.19.y branch from a92e0e5 to a697f1a Compare March 4, 2026 13:30

apply dmacoherent patch for arm64

7a01bbf

nicholasaiello force-pushed the rpi-6.19.y-drm-ttm branch from f86fb7b to 7a01bbf Compare March 7, 2026 15:24

nicholasaiello force-pushed the rpi-6.19.y branch from a697f1a to dce4fc9 Compare March 7, 2026 15:24

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

6.19: apply dmacoherent patch for arm64#6

6.19: apply dmacoherent patch for arm64#6
nicholasaiello wants to merge 1 commit into
rpi-6.19.yfrom
rpi-6.19.y-drm-ttm

nicholasaiello commented Jan 5, 2026

Uh oh!

Copilot AI left a comment

Uh oh!

Copilot AI Jan 5, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Conversation

nicholasaiello commented Jan 5, 2026

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Copilot AI Jan 5, 2026

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants