Red Hat Bugzilla – Bug 206295
GFS2: bad spinlock magic, gfs2_mmap
Last modified: 2009-05-27 23:31:11 EDT
New oops, current rawhide (2.6.17-1.2630) Kernel BUG at lib/list_debug.c:26─────────────────────────────────────┘ invalid opcode: 0000 [1] SMP last sysfs file: /block/xvda/dev CPU 0 Modules linked in: dm_emc dm_round_robin dm_multipath dm_snapshot dm_mirror dm_zero dm_mod xfs jfs reiserfs lock_nolock gfs2 ext3 jbd msdos raid456 xor raid1 raid0 xenblk xennet iscsi_tcp libiscsi scsi_transport_iscsi sr_mod sd_mod scsi_mod ide_cd cdrom ipv6 squashfs pcspkr loop nfs nfs_acl fscache lockd sunrpc vfat fat cramfs Pid: 601, comm: build-locale-ar Not tainted 2.6.17-1.2630.fc6xen #1 RIP: e030:[<ffffffff80330d00>] [<ffffffff80330d00>] __list_add+0x24/0x68 RSP: e02b:ffff88000b847cd8 EFLAGS: 00010286 RAX: 0000000000000058 RBX: ffff88000b3494b0 RCX: ffff88000b846000 RDX: 0000000000000000 RSI: 0000000000000001 RDI: 0000000000000000 RBP: ffff88000b847d18 R08: ffffffff804c88b8 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000000 R12: ffff88000b847d18 R13: ffffc2000010ab40 R14: 0000000000000000 R15: ffff88000b349490 FS: 0000000000000000(0063) GS:ffffffff8064e000(0000) knlGS:0000000000000000 CS: e033 DS: 0000 ES: 0000 Process build-locale-ar (pid: 601, threadinfo ffff88000b846000, task ffff880017524820) Stack: ffff880014ab0b50 ffffc200000ba000 ffff88000b847cf8 ffffffff882daf18 ffffc200000ba000 000002db000001f6 0000000000000000 0000000000000000 ffff88000b3494b0 ffffc2000010ab60 Call Trace: [<ffffffff882daf18>] :gfs2:databuf_lo_before_commit+0xd3/0x522 [<ffffffff882d9717>] :gfs2:gfs2_log_flush+0x11e/0x2d5 [<ffffffff882eae25>] :gfs2:gfs2_trans_end+0x13f/0x176 [<ffffffff882d6f07>] :gfs2:gfs2_glock_nq_atime+0x213/0x2a2 [<ffffffff882dfaca>] :gfs2:gfs2_mmap+0x36/0x83 [<ffffffff882d6e46>] :gfs2:gfs2_glock_nq_atime+0x152/0x2a2 [<ffffffff8020de6b>] do_mmap_pgoff+0x49f/0x773 [<ffffffff80262079>] _spin_unlock_irq+0x9/0x10 [<ffffffff80223c42>] sys_mmap+0x89/0xc1 [<ffffffff8025da05>] tracesys+0xa7/0xb2 Code: 0f 0b 68 c8 d8 47 80 c2 1a 00 48 8b 55 00 48 39 da 74 1b 48 RIP [<ffffffff80330d00>] __list_add+0x24/0x68 RSP <ffff88000b847cd8> <0>BUG: spinlock cpu recursion on CPU#0, gfs2_logd/586 (Not tainted) lock: ffffc2000010aae0, .magic: dead4ead, .owner: build-locale-ar/601, .owner_cpu: 0 Call Trace: [<ffffffff8020722e>] _raw_spin_lock+0x58/0xeb [<ffffffff882d920e>] :gfs2:gfs2_ail1_empty+0x23/0x9f [<ffffffff80298f07>] keventd_create_kthread+0x0/0x66 [<ffffffff882ce250>] :gfs2:gfs2_logd+0x4b/0x10e [<ffffffff802839f3>] __wake_up_common+0x3e/0x68 [<ffffffff80298f07>] keventd_create_kthread+0x0/0x66 [<ffffffff882ce205>] :gfs2:gfs2_logd+0x0/0x10e [<ffffffff802335c8>] kthread+0xfe/0x132 [<ffffffff8025e238>] child_rip+0xa/0x12 [<ffffffff80298f07>] keventd_create_kthread+0x0/0x66 [<ffffffff8025fee7>] thread_return+0x0/0xec [<ffffffff802334ca>] kthread+0x0/0x132 [<ffffffff8025e22e>] child_rip+0x0/0x12
Installs need to work on gfs2 for GA.
Agreed. This needs to be root installable for GA even though it's tech preview.
I have been unable to reproduce this panic with the current gfs2 code in RHEL5. Quite a number of bugs have been fixed in GFS2 some of them mmap releated so it is quite possible this problem has been fixed. Marking with needinfo for now.
Yeah, this one was gotten past a while ago and then bug 215088 was hit