[Linux-cluster] gfs2 kernel bug?

[fred conner]

I am running Redhat Enterprise 5.1.

On Thu, 2008-06-26 at 13:49 +0100, Steven Whitehouse wrote:
> Hi,
> 
> It would help to know what version of GFS2 you are using, but from the
> trace it looks likely that its ancient and if you choose a more recent
> version of the code, you should find that this problem has gone away,
> 
> Steve.
> 
> On Thu, 2008-06-26 at 08:49 -0400, fred conner wrote:
> > yesterday a co-worker was compiling code on a gfs2 filesystem and it
> > hung the server.  Is there a fix for this?  thank you.
> > 
> > Jun 25 14:24:54 abcc2 kernel: dlm: connecting to 1
> > Jun 25 14:24:54 abcc2 kernel: dlm: got connection from 1
> > original: gfs2_rename+0x1a9/0x610 [gfs2]
> > new: gfs2_inplace_reserve_i+0x205/0x5d0 [gfs2]
> > Kernel BUG at fs/gfs2/glock.c:1131
> > invalid opcode: 0000 [1] SMP 
> > last sysfs file: /kernel/dlm/rgmanager/control
> > CPU 3 
> > Modules linked in: autofs4 ipmi_devintf ipmi_si ipmi_msghandler hidp
> > l2cap bluetooth nfs lockd fscache nfs_acl lock_dlm gfs2 dlm configfs
> > sunrpc tun
> >   ip_gre ip_conntrack_ftp ip_conntrack_netbios_ns ipt_REJECT ipt_LOG
> > xt_state ip_conntrack nfnetlink iptable_filter ip_tables ip6t_REJECT
> > xt_tcpudp i
> > p6table_filter ip6_tables x_tables ipv6 video sbs backlight i2c_ec
> > i2c_corea bcc2 kernel: orbutton battery asus_acpi acpi_memhotplug ac
> > parport_pc l
> > p parport joydev ata_piix libata sr_mod shpchp ide_cd cdrom e1000 sg
> > pcspkr serio_raw bnx2 dm_snapshot dm_zero dm_mirror dm_mod usb_storage
> > megaraid
> > _sas megaraid_mbox sd_mod scsi_mod megaraid_mm ext3 jbd ehci_hcd
> > ohci_hcd uhci_hcd
> > Pid: 13627, comm: mv Not tainted 2.6.18-53.1.14.el5 #1
> > iginal: gfs2_renRIP: 0010:[<ffffffff8849b216>] ame+0x1a9/0x610
> > [<ffffffff8849b216>] :gfs2:gfs2_glock_nq+0x111/0x1d4
> > RSP: 0018:ffff8103e7681a98  EFLAGS: 00010286
> > [gfs2]
> > Jun 25 1RAX: 0000000000000020 RBX: ffff8103e7681cb0 RCX:
> > ffffffff80443520
> > RDX: 00000000ffffffff RSI: 0000000000000000 RDI: ffffffff802e675c
> > RBP: ffff8103f55c3eb0 R08: 00000000ffffffff R09: 0000000000000020
> > R10: 0000000000000000 R11: 0000000000000000 R12: ffff8103eadf3378
> > R13: ffff8103eadf3378 R14: 0000000000000000 R15: ffff81041d563000
> > FS:  00002aaaaaac9f20(0000) GS:ffff81010eb9c640(0000)
> > knlGS:0000000000000000
> > new: gfs2_CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> > CR2: 0000003acba5cf10 CR3: 00000003ee29e000 CR4: 00000000000006e0
> > Process mv (pid: 13627, threadinfo ffff8103e7680000, task
> > ffff810409dd20c0)
> > inplace_reserve_Stack:  0000000000000000 00000000000200cf
> > ffff81041d039dc0 ffff8103f55c3eb0
> >   ffff81041d563000 ffff8103f55c3a60 ffff8103f55c3d60 ffffffff884ac499
> >   0000000000000000 000002d000000000 ffffffff884b3568 ffff81041d563000
> > i+0x205/0x5d0 [gCall Trace:
> > fs2]
> >   [<ffffffff884ac499>] :gfs2:gfs2_inplace_reserve_i+0x20d/0x5d0
> >   [<ffffffff80143f87>] sort+0xfa/0x189
> >   [<ffffffff884a9ab1>] :gfs2:sort_qd+0x0/0x36
> >   [<ffffffff884a9cf5>] :gfs2:gfs2_quota_check+0x9a/0x182
> >   [<ffffffff884a69ce>] :gfs2:gfs2_rename+0x3b5/0x610
> >   [<ffffffff884a6715>] :gfs2:gfs2_rename+0xfc/0x610
> >   [<ffffffff884a6757>] :gfs2:gfs2_rename+0x13e/0x610
> >   [<ffffffff884a6781>] :gfs2:gfs2_rename+0x168/0x610
> >   [<ffffffff884a67c2>] :gfs2:gfs2_rename+0x1a9/0x610
> >   [<ffffffff8849b417>] :gfs2:gfs2_holder_uninit+0xd/0x1f
> >   [<ffffffff884a765e>] :gfs2:gfs2_permission+0xae/0xd4
> >   [<ffffffff8000d42e>] permission+0x81/0xc8
> >   [<ffffffff8002a515>] vfs_rename+0x2db/0x458
> >   [<ffffffff800362fa>] sys_renameat+0x180/0x1eb
> >   [<ffffffff80065a9d>] do_page_fault+0x4eb/0x81d
> >   [<ffffffff8002c886>] mntput_no_expire+0x19/0x89
> >   [<ffffffff80031f4f>] sys_faccessat+0x148/0x18d
> >   [<ffffffff800b3a88>] audit_syscall_entry+0x14d/0x180
> >   [<ffffffff8005c28d>] tracesys+0xd5/0xe0
> > 
> > 
> > Code: 0f 0b 68 2d 1c 4b 88 c2 6b 04 8b 75 18 49 8b 44 24 78 49 8d 
> > RIP  [<ffffffff8849b216>] :gfs2:gfs2_glock_nq+0x111/0x1d4
> >   RSP <ffff8103e7681a98>
> >   <0>Kernel panic - not syncing: Fatal exception
> > 
> > 
> 
> --
> Linux-cluster mailing list
> Linux-cluster at redhat.com
> https://www.redhat.com/mailman/listinfo/linux-cluster
> 
-- 
Fred Conner [Contractor]