On Tuesday 09 of August 2011 14:56:57 Herbert Poetzl wrote: > On Tue, Aug 09, 2011 at 01:35:13PM +0200, Pawel Sikora wrote: > > fresh logs from another lock: > > http://pluto.agmk.net/kernel/opteron.lock-01.txt > > could you do the address check for the Code: lines > for that one too and mark the <xx> for me? first pattern with mark <5b> is located in vmlinux file at 0x6341a3. text section starts in file at 0x200000 with logical adress ffffffff81000000, so if i caclulate correctly the code is: ffffffff81434174: 66 66 66 2e 0f 1f 84 data32 data32 nopw %cs:0x0(%rax,%rax,1) ffffffff8143417b: 00 00 00 00 00 ffffffff81434180 <down_read>: ffffffff81434180: 55 push %rbp ffffffff81434181: 48 89 e5 mov %rsp,%rbp ffffffff81434184: 53 push %rbx ffffffff81434185: 48 89 fb mov %rdi,%rbx ffffffff81434188: 48 83 ec 08 sub $0x8,%rsp ffffffff8143418c: e8 3f ed ff ff callq ffffffff81432ed0 <_cond_resched> ffffffff81434191: 48 89 d8 mov %rbx,%rax ffffffff81434194: f0 48 ff 00 lock incq (%rax) ffffffff81434198: 79 05 jns ffffffff8143419f <down_read+0x1f> ffffffff8143419a: e8 c1 bc e0 ff callq ffffffff8123fe60 <call_rwsem_down_read_failed> ffffffff8143419f: 48 83 c4 08 add $0x8,%rsp ffffffff814341a3: 5b pop %rbx <===== HERE ffffffff814341a4: c9 leaveq ffffffff814341a5: c3 retq second pattern with mark <4c> is located at 0x43482a: ffffffff81234790 <find_next_bit>: (...) ffffffff812347fc: 48 f7 c7 c0 ff ff ff test $0xffffffffffffffc0,%rdi ffffffff81234803: 75 e3 jne ffffffff812347e8 <find_next_bit+0x58> ffffffff81234805: 48 85 ff test %rdi,%rdi ffffffff81234808: 4c 89 c8 mov %r9,%rax ffffffff8123480b: 74 20 je ffffffff8123482d <find_next_bit+0x9d> ffffffff8123480d: 49 8b 00 mov (%r8),%rax ffffffff81234810: b9 40 00 00 00 mov $0x40,%ecx ffffffff81234815: 48 c7 c2 ff ff ff ff mov $0xffffffffffffffff,%rdx ffffffff8123481c: 29 f9 sub %edi,%ecx ffffffff8123481e: 48 d3 ea shr %cl,%rdx ffffffff81234821: 48 21 d0 and %rdx,%rax ffffffff81234824: 74 0a je ffffffff81234830 <find_next_bit+0xa0> ffffffff81234826: 48 0f bc c0 bsf %rax,%rax ffffffff8123482a: 4c 01 c8 add %r9,%rax <====== HERE ffffffff8123482d: c9 leaveq ffffffff8123482e: c3 retq code: fb 4c 89 65 e0 4c 89 7d f8 49 89 d5 85 f6 74 2a 41 0f b6 10 45 89 cc 89 f0 4d 89 c7 44 38 ca 74 0f eb 2d 0f 1f 40 00 49 83 c7 01 is located at 0x331db1: ffffffff81131d90 <check_bytes_and_report>: ffffffff81131d90: 55 push %rbp ffffffff81131d91: b8 01 00 00 00 mov $0x1,%eax ffffffff81131d96: 48 89 e5 mov %rsp,%rbp ffffffff81131d99: 48 83 ec 50 sub $0x50,%rsp ffffffff81131d9d: 4c 89 75 f0 mov %r14,-0x10(%rbp) ffffffff81131da1: 49 89 f6 mov %rsi,%r14 ffffffff81131da4: 8b 75 10 mov 0x10(%rbp),%esi ffffffff81131da7: 48 89 5d d8 mov %rbx,-0x28(%rbp) ffffffff81131dab: 4c 89 6d e8 mov %r13,-0x18(%rbp) ffffffff81131daf: 48 89 fb mov %rdi,%rbx <==== FROM HERE 0xfb ffffffff81131db2: 4c 89 65 e0 mov %r12,-0x20(%rbp) ffffffff81131db6: 4c 89 7d f8 mov %r15,-0x8(%rbp) ffffffff81131dba: 49 89 d5 mov %rdx,%r13 ffffffff81131dbd: 85 f6 test %esi,%esi ffffffff81131dbf: 74 2a je ffffffff81131deb <check_bytes_and_report+0x5b> ffffffff81131dc1: 41 0f b6 10 movzbl (%r8),%edx ffffffff81131dc5: 45 89 cc mov %r9d,%r12d ffffffff81131dc8: 89 f0 mov %esi,%eax ffffffff81131dca: 4d 89 c7 mov %r8,%r15 ffffffff81131dcd: 44 38 ca cmp %r9b,%dl ffffffff81131dd0: 74 0f je ffffffff81131de1 <check_bytes_and_report+0x51> ffffffff81131dd2: eb 2d jmp ffffffff81131e01 <check_bytes_and_report+0x71> ffffffff81131dd4: 0f 1f 40 00 nopl 0x0(%rax) ffffffff81131dd8: 49 83 c7 01 add $0x1,%r15 but i can't find in vmlinux this pattern: 00 00 66 66 66 66 90 65 48 8b 04 25 c8 b5 00 00 83 a0 3c e0 ff ff fb 0f ae f0 48 8b 80 38 e0 ff ff a8 08 0f 85 17 01 00 00 fb f4 maybe it comes from some loaded module?