> On Jan 20, 2015, at 9:00 AM, Mkrtchyan, Tigran <tigran.mkrtchyan@xxxxxxx> wrote: > > > > Dear fellows, > > since we have enabled commit through DS code we > permanently observe kernel crashes with RHEL6/7 and ubuntu 14.04: > > > <1>BUG: unable to handle kernel paging request at 00000000dc364913 > <1>IP: [<ffffffffa02b45df>] nfs_init_commit+0x1f/0xf0 [nfs] > <4>PGD 6393ae067 PUD 0 > <4>Oops: 0000 [#1] SMP > <4>last sysfs file: /sys/devices/system/cpu/online > <4>CPU 1 > <4>Modules linked in: vfat fat usb_storage mpt3sas mpt2sas raid_class mptctl ipmi_devintf dell_rbu openafs(P)(U) autof > s4 nfs_layout_nfsv41_files nfs lockd fscache auth_rpcgss nfs_acl sunrpc bonding 8021q garp stp llc ipv6 power_meter ac > pi_ipmi ipmi_si ipmi_msghandler iTCO_wdt iTCO_vendor_support microcode dcdbas sg bnx2 lpc_ich mfd_core i7core_edac eda > c_core ext4 jbd2 mbcache sd_mod crc_t10dif wmi pata_acpi ata_generic ata_piix mptsas mptscsih mptbase scsi_transport_s > as dm_mirror dm_region_hash dm_log dm_mod [last unloaded: scsi_wait_scan] > <4> > <4>Pid: 18209, comm: flush-0:19 Tainted: P --------------- 2.6.32-504.3.3.el6.x86_64 #1 Dell Inc. PowerEdge M610/0N582M > <4>RIP: 0010:[<ffffffffa02b45df>] [<ffffffffa02b45df>] nfs_init_commit+0x1f/0xf0 [nfs] > <4>RSP: 0018:ffff88063988da30 EFLAGS: 00010246 > <4>RAX: ffff88063988db60 RBX: ffff88009c492040 RCX: ffff88063988db30 > <4>RDX: 0000000000000000 RSI: ffff88063988db60 RDI: 00000000dc364903 > <4>RBP: ffff88063988da40 R08: ffff88063988da90 R09: f9aa37faa254d404 > <4>R10: 0000000000000010 R11: 0000000000000000 R12: 0000000000000001 > <4>R13: ffff880339f33a00 R14: ffff88063988db30 R15: ffff88063988d8c8 > <4>FS: 0000000000000000(0000) GS:ffff880028200000(0000) knlGS:0000000000000000 > <4>CS: 0010 DS: 0018 ES: 0018 CR0: 000000008005003b > <4>CR2: 00000000dc364913 CR3: 0000000639fbb000 CR4: 00000000000007e0 > <4>DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 > <4>DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 > <4>Process flush-0:19 (pid: 18209, threadinfo ffff88063988c000, task ffff88063837c040) > <4>Stack: > <4> 0000000000000000 ffff88009c492040 ffff88063988dad0 ffffffffa031fdb7 > <4><d> ffff88063837c5f8 ffff88063988da90 ffff8800a6e34600 ffff880337f2a950 > <4><d> ffff880637c99488 0000000037f2a940 ffff88063988db60 0000000000000000 > <4>Call Trace: > <4> [<ffffffffa031fdb7>] filelayout_commit_pagelist+0x277/0x3c0 [nfs_layout_nfsv41_files] > <4> [<ffffffffa02b613b>] nfs_generic_commit_list+0xab/0x100 [nfs] > <4> [<ffffffffa02b627c>] nfs_commit_inode+0xec/0x150 [nfs] > <4> [<ffffffffa02b6aab>] nfs_write_inode+0xab/0x100 [nfs] > <4> [<ffffffff811baedc>] writeback_single_inode+0x20c/0x290 > <4> [<ffffffff811bb1ad>] writeback_sb_inodes+0xbd/0x170 > <4> [<ffffffff811bb30b>] writeback_inodes_wb+0xab/0x1b0 > <4> [<ffffffff811bb703>] wb_writeback+0x2f3/0x410 > <4> [<ffffffff8100b9ce>] ? common_interrupt+0xe/0x13 > <4> [<ffffffff81088062>] ? del_timer_sync+0x22/0x30 > <4> [<ffffffff811bb9c5>] wb_do_writeback+0x1a5/0x240 > <4> [<ffffffff811bbac3>] bdi_writeback_task+0x63/0x1b0 > <4> [<ffffffff8109e987>] ? bit_waitqueue+0x17/0xd0 > <4> [<ffffffff81148360>] ? bdi_start_fn+0x0/0x100 > <4> [<ffffffff811483e6>] bdi_start_fn+0x86/0x100 > <4> [<ffffffff81148360>] ? bdi_start_fn+0x0/0x100 > <4> [<ffffffff8109e66e>] kthread+0x9e/0xc0 > <4> [<ffffffff8100c20a>] child_rip+0xa/0x20 > <4> [<ffffffff8109e5d0>] ? kthread+0x0/0xc0 > <4> [<ffffffff8100c200>] ? child_rip+0x0/0x20 > <4>Code: c3 66 66 2e 0f 1f 84 00 00 00 00 00 55 48 89 e5 53 48 83 ec 08 0f 1f 44 00 00 48 8b 06 48 89 fb 48 8b 78 18 48 39 c6 48 8b 7f 40 <48> 8b 7f 10 74 2b 4c 8b 83 c8 01 00 00 4c 8b 4e 08 4c 8d 93 c8 > <1>RIP [<ffffffffa02b45df>] nfs_init_commit+0x1f/0xf0 [nfs] > <4> RSP <ffff88063988da30> > <4>CR2: 00000000dc364913 > > > I have vmcore file as well, so let me know if you need some more information. > Hi Tigran! Have you tried a recent upstream kernel? IIRC I fixed a seeming similar filelayout commit issue not too long ago. The filelayout commit path seems to have been broken for a while - mostly because all the filelayout servers (that I know of) use stable writes, so that code path went untested... -dros -- To unsubscribe from this list: send the line "unsubscribe linux-nfs" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
