> journalctl -k --grep UBSAN -ojson --output-fields=MESSAGE |jq -r .MESSAGE |uniq
UBSAN: array-index-out-of-bounds in /build/linux-X4HRYL/linux-6.5.0/drivers/gpu/drm/amd/amdgpu/../pm/powerplay/hwmgr/processpptables.c:1249:61
UBSAN: array-index-out-of-bounds in /build/linux-X4HRYL/linux-6.5.0/drivers/gpu/drm/amd/amdgpu/../pm/powerplay/hwmgr/processpptables.c:1251:44
UBSAN: array-index-out-of-bounds in /build/linux-X4HRYL/linux-6.5.0/drivers/gpu/drm/amd/amdgpu/../pm/powerplay/hwmgr/processpptables.c:1249:46
UBSAN: array-index-out-of-bounds in /build/linux-X4HRYL/linux-6.5.0/drivers/gpu/drm/amd/amdgpu/../pm/powerplay/hwmgr/processpptables.c:1221:34
UBSAN: array-index-out-of-bounds in /build/linux-X4HRYL/linux-6.5.0/drivers/gpu/drm/amd/amdgpu/../pm/powerplay/hwmgr/processpptables.c:1222:44
UBSAN: array-index-out-of-bounds in /build/linux-X4HRYL/linux-6.5.0/drivers/gpu/drm/amd/amdgpu/../pm/powerplay/hwmgr/processpptables.c:1221:19
UBSAN: array-index-out-of-bounds in /build/linux-X4HRYL/linux-6.5.0/drivers/gpu/drm/amd/amdgpu/../pm/powerplay/hwmgr/processpptables.c:1278:45
UBSAN: array-index-out-of-bounds in /build/linux-X4HRYL/linux-6.5.0/drivers/gpu/drm/amd/amdgpu/../pm/powerplay/hwmgr/processpptables.c:1279:65
UBSAN: array-index-out-of-bounds in /build/linux-X4HRYL/linux-6.5.0/drivers/gpu/drm/amd/amdgpu/../pm/powerplay/hwmgr/processpptables.c:1280:9
UBSAN: array-index-out-of-bounds in /build/linux-X4HRYL/linux-6.5.0/drivers/gpu/drm/amd/amdgpu/../pm/powerplay/hwmgr/processpptables.c:1303:44
UBSAN: array-index-out-of-bounds in /build/linux-X4HRYL/linux-6.5.0/drivers/gpu/drm/amd/amdgpu/../pm/powerplay/hwmgr/processpptables.c:1304:64
UBSAN: array-index-out-of-bounds in /build/linux-X4HRYL/linux-6.5.0/drivers/gpu/drm/amd/amdgpu/../pm/powerplay/hwmgr/processpptables.c:1305:9
UBSAN: array-index-out-of-bounds in /build/linux-X4HRYL/linux-6.5.0/drivers/gpu/drm/amd/amdgpu/../pm/powerplay/hwmgr/processpptables.c:394:34
UBSAN: array-index-out-of-bounds in /build/linux-X4HRYL/linux-6.5.0/drivers/gpu/drm/amd/amdgpu/../pm/powerplay/hwmgr/processpptables.c:395:4
UBSAN: array-index-out-of-bounds in /build/linux-X4HRYL/linux-6.5.0/drivers/gpu/drm/amd/amdgpu/../pm/powerplay/hwmgr/processpptables.c:397:19
UBSAN: array-index-out-of-bounds in /build/linux-X4HRYL/linux-6.5.0/drivers/gpu/drm/amd/amdgpu/../pm/powerplay/hwmgr/processpptables.c:1500:42
UBSAN: array-index-out-of-bounds in /build/linux-X4HRYL/linux-6.5.0/drivers/gpu/drm/amd/amdgpu/../pm/powerplay/hwmgr/processpptables.c:1501:42
UBSAN: array-index-out-of-bounds in /build/linux-X4HRYL/linux-6.5.0/drivers/gpu/drm/amd/amdgpu/../pm/powerplay/hwmgr/processpptables.c:1502:42
UBSAN: array-index-out-of-bounds in /build/linux-X4HRYL/linux-6.5.0/drivers/gpu/drm/amd/amdgpu/../pm/powerplay/hwmgr/processpptables.c:929:40
I'm afraid they may be related to the kernel KSSP shenanigans landing in amdgpu, and not related at all to bcache (they just happen to be be USBSAN errors too). Should I open a separate bug report?
uppon further inspection, the amdgpu traces are:
> journalctl -k --grep UBSAN -ojson --output- fields= MESSAGE |jq -r .MESSAGE |uniq out-of- bounds in /build/ linux-X4HRYL/ linux-6. 5.0/drivers/ gpu/drm/ amd/amdgpu/ ../pm/powerplay /hwmgr/ processpptables .c:1249: 61 out-of- bounds in /build/ linux-X4HRYL/ linux-6. 5.0/drivers/ gpu/drm/ amd/amdgpu/ ../pm/powerplay /hwmgr/ processpptables .c:1251: 44 out-of- bounds in /build/ linux-X4HRYL/ linux-6. 5.0/drivers/ gpu/drm/ amd/amdgpu/ ../pm/powerplay /hwmgr/ processpptables .c:1249: 46 out-of- bounds in /build/ linux-X4HRYL/ linux-6. 5.0/drivers/ gpu/drm/ amd/amdgpu/ ../pm/powerplay /hwmgr/ processpptables .c:1221: 34 out-of- bounds in /build/ linux-X4HRYL/ linux-6. 5.0/drivers/ gpu/drm/ amd/amdgpu/ ../pm/powerplay /hwmgr/ processpptables .c:1222: 44 out-of- bounds in /build/ linux-X4HRYL/ linux-6. 5.0/drivers/ gpu/drm/ amd/amdgpu/ ../pm/powerplay /hwmgr/ processpptables .c:1221: 19 out-of- bounds in /build/ linux-X4HRYL/ linux-6. 5.0/drivers/ gpu/drm/ amd/amdgpu/ ../pm/powerplay /hwmgr/ processpptables .c:1278: 45 out-of- bounds in /build/ linux-X4HRYL/ linux-6. 5.0/drivers/ gpu/drm/ amd/amdgpu/ ../pm/powerplay /hwmgr/ processpptables .c:1279: 65 out-of- bounds in /build/ linux-X4HRYL/ linux-6. 5.0/drivers/ gpu/drm/ amd/amdgpu/ ../pm/powerplay /hwmgr/ processpptables .c:1280: 9 out-of- bounds in /build/ linux-X4HRYL/ linux-6. 5.0/drivers/ gpu/drm/ amd/amdgpu/ ../pm/powerplay /hwmgr/ processpptables .c:1303: 44 out-of- bounds in /build/ linux-X4HRYL/ linux-6. 5.0/drivers/ gpu/drm/ amd/amdgpu/ ../pm/powerplay /hwmgr/ processpptables .c:1304: 64 out-of- bounds in /build/ linux-X4HRYL/ linux-6. 5.0/drivers/ gpu/drm/ amd/amdgpu/ ../pm/powerplay /hwmgr/ processpptables .c:1305: 9 out-of- bounds in /build/ linux-X4HRYL/ linux-6. 5.0/drivers/ gpu/drm/ amd/amdgpu/ ../pm/powerplay /hwmgr/ processpptables .c:394: 34 out-of- bounds in /build/ linux-X4HRYL/ linux-6. 5.0/drivers/ gpu/drm/ amd/amdgpu/ ../pm/powerplay /hwmgr/ processpptables .c:395: 4 out-of- bounds in /build/ linux-X4HRYL/ linux-6. 5.0/drivers/ gpu/drm/ amd/amdgpu/ ../pm/powerplay /hwmgr/ processpptables .c:397: 19 out-of- bounds in /build/ linux-X4HRYL/ linux-6. 5.0/drivers/ gpu/drm/ amd/amdgpu/ ../pm/powerplay /hwmgr/ processpptables .c:1500: 42 out-of- bounds in /build/ linux-X4HRYL/ linux-6. 5.0/drivers/ gpu/drm/ amd/amdgpu/ ../pm/powerplay /hwmgr/ processpptables .c:1501: 42 out-of- bounds in /build/ linux-X4HRYL/ linux-6. 5.0/drivers/ gpu/drm/ amd/amdgpu/ ../pm/powerplay /hwmgr/ processpptables .c:1502: 42 out-of- bounds in /build/ linux-X4HRYL/ linux-6. 5.0/drivers/ gpu/drm/ amd/amdgpu/ ../pm/powerplay /hwmgr/ processpptables .c:929: 40
UBSAN: array-index-
UBSAN: array-index-
UBSAN: array-index-
UBSAN: array-index-
UBSAN: array-index-
UBSAN: array-index-
UBSAN: array-index-
UBSAN: array-index-
UBSAN: array-index-
UBSAN: array-index-
UBSAN: array-index-
UBSAN: array-index-
UBSAN: array-index-
UBSAN: array-index-
UBSAN: array-index-
UBSAN: array-index-
UBSAN: array-index-
UBSAN: array-index-
UBSAN: array-index-
I'm afraid they may be related to the kernel KSSP shenanigans landing in amdgpu, and not related at all to bcache (they just happen to be be USBSAN errors too). Should I open a separate bug report?