On Thu, 20 Sep 2007, Alexey Dobriyan wrote: > OK, I do clean boot, ssh to box, then sudo slabinfo -v. > > > ============================================================================= > BUG kmalloc-16: Object padding overwritten > ----------------------------------------------------------------------------- > > INFO: 0xffff810100fd8998-0xffff810100fd8999. First byte 0xa7 instead of 0x5a > INFO: Allocated in sysfs_new_dirent+0x100/0x120 age=11055 cpu=0 pid=3474 > INFO: Freed in kobject_uevent_env+0x123/0x430 age=11055 cpu=0 pid=3474 > INFO: Slab 0xffff810004837740 used=28 fp=0xffff810100fd89a0 > flags=0x8000000000000083 > INFO: Object 0xffff810100fd8948 @offset=2376 fp=0xffff810100fd89a0
Hmmm.. A corrupted sysfs object at an offset of one word from the end of the object that could never have been caught by SLAB since it does not check more than 4 bytes. Does the value 0x5ea7 tell us anything? Maybe a counter was incremented a couple of times from the initial value of 0x5a5a that was put there by SLUB? > INFO: 0xffff810101b45310-0xffff810101b45311. First byte 0xd3 instead of 0x5a > INFO: Allocated in kobject_get_path+0x57/0xc0 age=18405 cpu=1 pid=2006 > INFO: Freed in kobject_uevent_env+0x123/0x430 age=18405 cpu=1 pid=2006 > INFO: Slab 0xffff81000485f718 used=8 fp=0xffff810101b45318 > flags=0x8000000000000083 > INFO: Object 0xffff810101b452c0 @offset=704 fp=0xffff810101b45370 > > Bytes b4 0xffff810101b452b0: db f9 fb ff 00 00 00 00 5a 5a 5a 5a 5a 5a 5a 5a > Ûùûÿ....ZZZZZZZZ > Object 0xffff810101b452c0: 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b a5 > kkkkkkkkkkkkkkk¥ > Redzone 0xffff810101b452d0: bb bb bb bb bb bb bb bb > »»»»»»»» > Padding 0xffff810101b45310: d3 5e 5a 5a 5a 5a 5a 5a > Ó^ZZZZZZ Ditto.... but here we have a freed object in the above case the object is still in use. Done by different processes at different times. Eric: Anything that comes to mind in sysfs?
