On 24/02/2026 11:06, Leo Yan wrote:
On Mon, Feb 23, 2026 at 01:28:03PM +0000, Suzuki K Poulose wrote:
[...]
[ 827.216924] Possible unsafe locking scenario: [ 827.216924] [ 827.222849] CPU0 CPU1 [ 827.227380] ---- ---- [ 827.231912] lock(coresight_mutex); [ 827.235491] lock(ect_mutex); [ 827.241073] lock(coresight_mutex); [ 827.247176] lock(ect_mutex); [ 827.250233] [ 827.250233] *** DEADLOCK ***
Thanks for the test, Suzuki.
I have sent a refactor patch to dismiss locking chain: https://lore.kernel.org/linux-arm-kernel/20260224-arm_coresight_refactor_cti...
For recording, I enabled the following configs to reproduce the lockdep report:
Thanks for the patch, I can confirm that it fixes the problem and is indeed a good cleanup. Hacking the device release operation wasn't really a good idea.
Cheers Suzuki
CONFIG_LOCKUP_DETECTOR CONFIG_PROVE_LOCKING
Thanks, Leo