Jim Radford
2001-Nov-21 18:46 UTC
Assertion failure in journal_bmap() at journal.c:636: "ret != 0"
A scsi error (caused possibly by a loose cable) has left the processes accessing my ext3 file system hung in an unkillable state after it triggered an assert in ext3's journaling layer. I assume my only recourse at this point is to reboot. Please correct me if I'm wrong. Log messages appended. Thanks, -Jim Nov 21 04:04:04 attila kernel: SCSI disk error : host 0 channel 0 id 5 lun 0 return code = 10000 Nov 21 04:04:04 attila kernel: I/O error: dev 08:01, sector 4136 Nov 21 04:04:04 attila kernel: EXT3-fs error (device sd(8,1)): ext3_readdir: directory #2 contains a hole at offset 0 Nov 21 04:04:34 attila kernel: scsi0:0:5:0: Attempting to queue an ABORT message [ lots of scsi dumping ... ] Nov 21 04:05:10 attila kernel: DevQ(0:5:0): 0 waiting Nov 21 04:05:10 attila kernel: scsi0:0:5:0: Cmd aborted from QINFIFO Nov 21 04:05:11 attila kernel: aic7xxx_abort returns 0x2002 Nov 21 04:05:11 attila kernel: scsi: device set offline - not ready or command retry failed after bus reset: host 0 channel 0 id 5 lun 0 Nov 21 04:05:11 attila kernel: SCSI disk error : host 0 channel 0 id 5 lun 0 return code = 50000 Nov 21 04:05:11 attila kernel: I/O error: dev 08:01, sector 0 Nov 21 04:05:14 attila kernel: I/O error: dev 08:01, sector 12472 Nov 21 04:05:14 attila kernel: Assertion failure in journal_bmap() at journal.c:636: "ret != 0" Nov 21 04:05:14 attila kernel: invalid operand: 0000 Nov 21 04:05:14 attila kernel: CPU: 0 Nov 21 04:05:18 attila kernel: EIP: 0010:[journal_bmap+70/96] Not tainted Nov 21 04:05:18 attila kernel: EIP: 0010:[<c0167996>] Not tainted Nov 21 04:05:18 attila kernel: EFLAGS: 00010296 Nov 21 04:05:18 attila kernel: eax: 00000044 ebx: 00000000 ecx: f78c4000 edx: f78c5f64 Nov 21 04:05:18 attila kernel: esi: f7e34a00 edi: cde5e460 ebp: 00000000 esp: f7a17e48 Nov 21 04:05:18 attila kernel: ds: 0018 es: 0018 ss: 0018 Nov 21 04:05:18 attila kernel: Process kjournald (pid: 131, stackpage=f7a17000) Nov 21 04:05:18 attila kernel: Stack: c0304a80 c0300d12 c0300c03 0000027c c0300d09 f7e34a00 c0167947 f7e34a00 Nov 21 04:05:18 attila kernel: 00000a79 f7e34a00 c01679bb f7e34a00 e8001f40 c016472f f7e34a00 00000000 Nov 21 04:05:19 attila kernel: 00000000 00000000 00000000 e8001f40 cb7d1900 cac6fc40 00000011 c02ab327 Nov 21 04:05:19 attila kernel: Call Trace: [journal_next_log_block+103/112] [journal_get_descriptor_buffer+11/64] [journal_commit_transaction+1919/4656] [ip_local_deliver+231/352] [schedule+969/1504] Nov 21 04:05:19 attila kernel: Call Trace: [<c0167947>] [<c01679bb>] [<c016472f>] [<c02ab327>] [<c01159f9>] Nov 21 04:05:19 attila kernel: [kjournald+419/736] [commit_timeout+0/16] [kernel_thread+38/48] [kjournald+0/736] Nov 21 04:05:19 attila kernel: [<c01671f3>] [<c0167030>] [<c0105616>] [<c0167050>] Nov 21 04:05:19 attila kernel: Nov 21 04:05:19 attila kernel: Code: 0f 0b 83 c4 14 eb 05 8d 76 00 89 c3 89 d8 5b c3 8d 76 00 8d
Jim Radford
2001-Nov-21 19:06 UTC
Assertion failure in journal_bmap() at journal.c:636: "ret != 0"
A scsi error (caused possibly by a loose cable) has left the processes accessing my ext3 file system hung in an unkillable state after it triggered an assert in ext3's journaling layer. I assume my only recourse at this point is to reboot. Please correct me if I'm wrong. Log messages appended. Thanks, -Jim Nov 21 04:04:04 attila kernel: SCSI disk error : host 0 channel 0 id 5 lun 0 return code = 10000 Nov 21 04:04:04 attila kernel: I/O error: dev 08:01, sector 4136 Nov 21 04:04:04 attila kernel: EXT3-fs error (device sd(8,1)): ext3_readdir: directory #2 contains a hole at offset 0 Nov 21 04:04:34 attila kernel: scsi0:0:5:0: Attempting to queue an ABORT message [ lots of scsi dumping ... ] Nov 21 04:05:10 attila kernel: DevQ(0:5:0): 0 waiting Nov 21 04:05:10 attila kernel: scsi0:0:5:0: Cmd aborted from QINFIFO Nov 21 04:05:11 attila kernel: aic7xxx_abort returns 0x2002 Nov 21 04:05:11 attila kernel: scsi: device set offline - not ready or command retry failed after bus reset: host 0 channel 0 id 5 lun 0 Nov 21 04:05:11 attila kernel: SCSI disk error : host 0 channel 0 id 5 lun 0 return code = 50000 Nov 21 04:05:11 attila kernel: I/O error: dev 08:01, sector 0 Nov 21 04:05:14 attila kernel: I/O error: dev 08:01, sector 12472 Nov 21 04:05:14 attila kernel: Assertion failure in journal_bmap() at journal.c:636: "ret != 0" Nov 21 04:05:14 attila kernel: invalid operand: 0000 Nov 21 04:05:14 attila kernel: CPU: 0 Nov 21 04:05:18 attila kernel: EIP: 0010:[journal_bmap+70/96] Not tainted Nov 21 04:05:18 attila kernel: EIP: 0010:[<c0167996>] Not tainted Nov 21 04:05:18 attila kernel: EFLAGS: 00010296 Nov 21 04:05:18 attila kernel: eax: 00000044 ebx: 00000000 ecx: f78c4000 edx: f78c5f64 Nov 21 04:05:18 attila kernel: esi: f7e34a00 edi: cde5e460 ebp: 00000000 esp: f7a17e48 Nov 21 04:05:18 attila kernel: ds: 0018 es: 0018 ss: 0018 Nov 21 04:05:18 attila kernel: Process kjournald (pid: 131, stackpage=f7a17000) Nov 21 04:05:18 attila kernel: Stack: c0304a80 c0300d12 c0300c03 0000027c c0300d09 f7e34a00 c0167947 f7e34a00 Nov 21 04:05:18 attila kernel: 00000a79 f7e34a00 c01679bb f7e34a00 e8001f40 c016472f f7e34a00 00000000 Nov 21 04:05:19 attila kernel: 00000000 00000000 00000000 e8001f40 cb7d1900 cac6fc40 00000011 c02ab327 Nov 21 04:05:19 attila kernel: Call Trace: [journal_next_log_block+103/112] [journal_get_descriptor_buffer+11/64] [journal_commit_transaction+1919/4656] [ip_local_deliver+231/352] [schedule+969/1504] Nov 21 04:05:19 attila kernel: Call Trace: [<c0167947>] [<c01679bb>] [<c016472f>] [<c02ab327>] [<c01159f9>] Nov 21 04:05:19 attila kernel: [kjournald+419/736] [commit_timeout+0/16] [kernel_thread+38/48] [kjournald+0/736] Nov 21 04:05:19 attila kernel: [<c01671f3>] [<c0167030>] [<c0105616>] [<c0167050>] Nov 21 04:05:19 attila kernel: Nov 21 04:05:19 attila kernel: Code: 0f 0b 83 c4 14 eb 05 8d 76 00 89 c3 89 d8 5b c3 8d 76 00 8d
Andrew Morton
2001-Nov-22 05:26 UTC
Re: Assertion failure in journal_bmap() at journal.c:636: "ret != 0"
Jim Radford wrote:> > A scsi error (caused possibly by a loose cable) has left the processes > accessing my ext3 file system hung in an unkillable state after it > triggered an assert in ext3's journaling layer. I assume my only > recourse at this point is to reboot. Please correct me if I'm wrong. > Log messages appended.Thanks, Jim. ext3's philosophy is "I'd rather crash your computer than scribble on your disk". Stephen recently went through the filesystem and changed things so that the filesystem is turned readonly when things go wrong, rather than killing the machine. But the situation where we get a physical I/O error when locating journal blocks slipped through. I'll add it to the ttd list.