If CONFIG_ARM64_SW_TTBR0_PAN is disabled in the kernel we dont see these random crashes anymore. I ran /usr/lib/plainbox-provider-checkbox/bin/disk_stress_ng and /usr/lib/plainbox-provider-checkbox/bin/memory_stress_ng tests on CRB (seuss) the last couple of days and did not see any crash. However with disk_stress_ng I did see the following message on the console, which is a different issue than the one tracked here in this bug.
[ 4163.900212] EXT4-fs error (device sda2) in ext4_free_inode:344: Corrupt filesystem
Next step would be to root cause why disabling CONFIG_ARM64_SW_TTBR0_PAN fixes the RCU stalls and random crashes of stress-ng.
If CONFIG_ ARM64_SW_ TTBR0_PAN is disabled in the kernel we dont see these random crashes anymore. I ran /usr/lib/ plainbox- provider- checkbox/ bin/disk_ stress_ ng and /usr/lib/ plainbox- provider- checkbox/ bin/memory_ stress_ ng tests on CRB (seuss) the last couple of days and did not see any crash. However with disk_stress_ng I did see the following message on the console, which is a different issue than the one tracked here in this bug.
[ 4163.900212] EXT4-fs error (device sda2) in ext4_free_ inode:344: Corrupt filesystem
Next step would be to root cause why disabling CONFIG_ ARM64_SW_ TTBR0_PAN fixes the RCU stalls and random crashes of stress-ng.