On Thu, 29 Feb 2024 17:11:28 +0000 Christophe Leroy <christophe.leroy@xxxxxxxxxx> wrote: > Interesting. > > I guess 0xe0000000 is where linear RAM starts to be mapped with pages ? > Can you confirm with a dump of > /sys/kernel/debug/powerpc/block_address_translation ? # cat /sys/kernel/debug/powerpc/block_address_translation ---[ Instruction Block Address Translation ]--- 0: 0xc0000000-0xc0ffffff 0x00000000 16M Kernel x m 1: 0xc1000000-0xc13fffff 0x01000000 4M Kernel x m 2: 0xc1400000-0xc15fffff 0x01400000 2M Kernel x m 3: 0xc1600000-0xc16fffff 0x01600000 1M Kernel x m 4: - 5: - 6: - 7: - ---[ Data Block Address Translation ]--- 0: 0xc0000000-0xc0ffffff 0x00000000 16M Kernel r m 1: 0xc1000000-0xc17fffff 0x01000000 8M Kernel r m 2: 0xc1800000-0xc1bfffff 0x01800000 4M Kernel r m 3: 0xf8000000-0xfbffffff 0x7c000000 64M Kernel rw m 4: 0xfc000000-0xfdffffff 0x7a000000 32M Kernel rw m 5: - 6: - 7: - block_address_translation looks different after the page corruption: # cat /sys/kernel/debug/powerpc/block_address_translation ---[ Instruction Block Address Translation ]--- 0: 0xc0000000-0xc1ffffff 0x00000000 32M Kernel x m 1: - 2: - 3: - 4: - 5: - 6: - 7: - ---[ Data Block Address Translation ]--- 0: 0xc0000000-0xc0ffffff 0x00000000 16M Kernel rw m 1: 0xc1000000-0xc17fffff 0x01000000 8M Kernel rw m 2: 0xc1800000-0xc1bfffff 0x01800000 4M Kernel rw m 3: 0xf8000000-0xfbffffff 0x7c000000 64M Kernel rw m 4: 0xfc000000-0xfdffffff 0x7a000000 32M Kernel rw m 5: - 6: - 7: - > Do we have a problem of race with hash table ? > > Would KCSAN help with that ? KCSAN did not report any hits during "stress -m 2 --vm-bytes 965M". Options used: KCSAN_SELFTEST=y, KCSAN_REPORT_ONCE_IN_MS=12000, KCSAN_REPORT_RACE_UNKNOWN_ORIGIN=y, KCSAN_STRICT=y, KCSAN_WEAK_MEMORY=y. Regards, Erhard
