diff options
| author | James Morris <jmorris@namei.org> | 2026-09-28 19:18:04 -0400 |
|---|---|---|
| committer | Yixun Lan <dlan@kernel.org> | 2026-10-02 11:04:59 +0000 |
| commit | f314c98807c12272a1902aeda9c648d81b39bad2 (patch) | |
| tree | 78bb954606f00cdaf7b8625ba5766dbd412be59a /tools/perf/scripts/python/bin/stackcollapse-report | |
| parent | 2f3b7240a4c27745ec327f3f1a2f28d897e6aab6 (diff) | |
riscv: dts: spacemit: k3-com260: keep dldo4 enabled
dldo4 on the P1 PMIC is marked regulator-boot-on, but has no consumer
described in the device tree. The regulator core can therefore disable
the firmware-enabled rail during unused-regulator cleanup.
Enabling pcie4_rc on a K3 CoM260 Dev Kit without this change while the
board's included RTL8852BE module is connected to the M.2 E-key slot
results in CPU hard-lock, as seen in kernel log:
[ 32.521536] dldo4: disabling
[ 48.125823] watchdog: CPU3: Watchdog detected hard LOCKUP on cpu 4
[ 64.125829] watchdog: BUG: soft lockup - CPU#1 stuck for 26s!
[ 64.125926] Hardware name: SpacemiT K3 CoM260 IFX (DT)
[ 64.125928] epc : smp_call_function_many_cond+0x158/0x740
[ 64.125938] ra : smp_call_function_many_cond+0x11a/0x740
GitHub Copilot assisted with diagnosis and build and test automation, and
changelog drafting.
Fixes: cfe5c91cb73c ("riscv: dts: spacemit: k3: Initial support for CoM260-IFX board")
Assisted-by: GitHub-Copilot:gpt-sol-5.6
Signed-off-by: James Morris <jmorris@namei.org>
Signed-off-by: Jennifer Berringer <jberring@redhat.com>
Link: https://patch.msgid.link/20260928231807.1757995-2-jberring@redhat.com
Signed-off-by: Yixun Lan <dlan@kernel.org>
Diffstat (limited to 'tools/perf/scripts/python/bin/stackcollapse-report')
0 files changed, 0 insertions, 0 deletions