Hello Daniel, I have another reference count issue, and have a patch for it. Although they have differenct panic stack, I think they have same root cause. https://www.redhat.com/archives/linux-cachefs/2018-February/msg00000.html I did try the patch of https://www.redhat.com/archives/linux-cachefs/2017-June/msg00002.html, but it looks like doesn't work on my environment. Thanks, -Lei 在 06/04/2018 7:16 PM, Daniel Axtens 写道: > Hi, > > We have a user reporting BUG splats and FSCache assertion failures with > FSCache and NFS. The full stack traces are below. > > We are consistently seeing the following FSCache assertion: > > [81738.886634] FS-Cache: > [81738.888281] FS-Cache: Assertion failed > [81738.889461] FS-Cache: 6 == 5 is false > [81738.890625] ------------[ cut here ]------------ > [81738.891706] kernel BUG at /build/linux-hVVhWi/linux-4.4.0/fs/fscache/operation.c:494! > > which references the following: > > ASSERTIFCMP(op->state != FSCACHE_OP_ST_INITIALISED && > op->state != FSCACHE_OP_ST_COMPLETE, > op->state, ==, FSCACHE_OP_ST_CANCELLED); > > So the operation has become DEAD when it was expected to be CANCELLED. > > fscache_put_operation() is the only thing I can find that will mark an > operation as DEAD, and then only when atomic_dec_and_test(&op->usage) > goes to 0 twice. > > This looks supsiciously like this bug reported in June 2017: > https://www.redhat.com/archives/linux-cachefs/2017-June/msg00002.html > "Re: Problem: Race in fscache operation enqueuing for > reading and copying multiple pages from cachefiles to netfs causes > kernel oops or at worst memory scribbling." > > Indeed, immediatlely prior to the assertion failure, we sometimes see a > WARN in the workqueue code from trying to insert a work that is not > pending, and the stack trace includes cachefiles_read_waiter calling > fscache_enqueue_operation. > > Was there ever any progress on that bug? Or is there anything else I > should be looking at? > > Thanks in advance! > > Regards, > Daniel > > > Full stack traces: > > [81738.885081] ------------[ cut here ]------------ > [81738.885165] WARNING: CPU: 67 PID: 70869 at /build/linux-hVVhWi/linux-4.4.0/kernel/workqueue.c:623 insert_work+0xac/0xc0() > [81738.885175] Modules linked in: binfmt_misc veth ipt_MASQUERADE nf_nat_masquerade_ipv4 nf_conntrack_netlink nfnetlink xfrm_user xfrm_algo iptable_nat nf_conntrack_ipv4 nf_defrag_ipv4 nf_nat_ipv4 xt_addrtype xt_conntrack nf_nat nf_conntrack br_netfilter bridge stp llc overlay xt_multiport iptable_filter ip_tables x_tables nfsv3 nfs_acl rpcsec_gss_krb5 auth_rpcgss nfsv4 nfs lockd grace nv_peer_mem(OE) cachefiles fscache msr rdma_ucm(OE) ib_ucm(OE) rdma_cm(OE) iw_cm(OE) configfs ib_ipoib(OE) ib_cm(OE) ib_uverbs(OE) ib_umad(OE) mlx5_ib(OE) mlx4_ib(OE) ib_core(OE) mlx4_en(OE) mlx4_core(OE) bonding ipmi_ssif intel_rapl x86_pkg_temp_thermal intel_powerclamp coretemp kvm_intel kvm irqbypass crct10dif_pclmul crc32_pclmul ghash_clmulni_intel aesni_intel aes_x86_64 lrw gf128mul glue_helper ablk_helper cryptd joydev input_leds mxm_wmi sb_edac edac_core lpc_ich mei_me mei nvidia_uvm(POE) shpchp 8250_fintek ipmi_si wmi acpi_power_meter mac_hid sunrpc ipmi_devintf ipmi_msghandler knem(OE) autofs4 > nvidia_drm(POE) nvidia_modeset(POE) ast i2c_algo_bit nvidia(POE) ttm ixgbe drm_kms_helper syscopyarea ses dca sysfillrect mlx5_core(OE) enclosure sysimgblt fb_sys_fops vxlan hid_generic ip6_udp_tunnel usbhid udp_tunnel mlx_compat(OE) hid ahci ptp megaraid_sas drm libahci pps_core mdio fjes > [81738.885422] CPU: 67 PID: 70869 Comm: xxxxxx Tainted: P OE 4.4.0-92-generic #115-Ubuntu > [81738.885428] Hardware name: xxxxx > [81738.885431] 0000000000000086 6bb3b89857e8c7df ffff887f7f2c3b60 ffffffff813f9c83 > [81738.885436] 0000000000000000 ffffffff81cbb498 ffff887f7f2c3b98 ffffffff81081312 > [81738.885439] ffff887f22fdd200 ffff887f626bc540 ffff887f656e8000 0000000000000005 > [81738.885446] Call Trace: > [81738.885452] <IRQ> [<ffffffff813f9c83>] dump_stack+0x63/0x90 > [81738.885512] [<ffffffff81081312>] warn_slowpath_common+0x82/0xc0 > [81738.885535] [<ffffffff8108145a>] warn_slowpath_null+0x1a/0x20 > [81738.885558] [<ffffffff8109869c>] insert_work+0xac/0xc0 > [81738.885571] [<ffffffff810987e9>] __queue_work+0x139/0x3b0 > [81738.885579] [<ffffffff81098c77>] queue_work_on+0x27/0x40 > [81738.885625] [<ffffffffc054b064>] fscache_enqueue_operation+0x94/0x150 [fscache] > [81738.885640] [<ffffffffc03b68ec>] cachefiles_read_waiter+0xbc/0x100 [cachefiles] > [81738.885666] [<ffffffff810c3d12>] __wake_up_common+0x52/0x90 > [81738.885694] [<ffffffff810c3d89>] __wake_up+0x39/0x50 > [81738.885715] [<ffffffff810c4040>] __wake_up_bit+0x50/0x70 > [81738.885726] [<ffffffff8118f739>] unlock_page+0x69/0x70 > [81738.885746] [<ffffffff812e8073>] mpage_end_io+0x33/0xc0 > [81738.885770] [<ffffffff813c28af>] bio_endio+0x3f/0x60 > [81738.885782] [<ffffffff813ca257>] blk_update_request+0x87/0x310 > [81738.885801] [<ffffffff815c40b3>] scsi_end_request+0x33/0x1d0 > [81738.885815] [<ffffffff815c77e6>] scsi_io_completion+0x1b6/0x690 > [81738.885841] [<ffffffff815be40f>] scsi_finish_command+0xcf/0x120 > [81738.885858] [<ffffffff815c6f74>] scsi_softirq_done+0x124/0x150 > [81738.885897] [<ffffffff813d2147>] blk_done_softirq+0x87/0xb0 > [81738.885916] [<ffffffff81085df1>] __do_softirq+0x101/0x290 > [81738.885933] [<ffffffff810860f3>] irq_exit+0xa3/0xb0 > [81738.885950] [<ffffffff818449e4>] do_IRQ+0x54/0xd0 > [81738.885975] [<ffffffff81842ac2>] common_interrupt+0x82/0x82 > [81738.885982] <EOI> [<ffffffff811fe87c>] ? try_charge+0x3c/0x660 > [81738.886035] [<ffffffff810c4125>] ? wake_up_bit+0x25/0x30 > [81738.886057] [<ffffffff81202ceb>] mem_cgroup_try_charge+0x6b/0x1e0 > [81738.886074] [<ffffffff8118fc40>] __add_to_page_cache_locked+0x60/0x1f0 > [81738.886194] [<ffffffffc141cd83>] ? nfs_pageio_add_request+0xa3/0x1b0 [nfs] > [81738.886284] [<ffffffffc141d760>] ? nfs_read_completion+0x210/0x210 [nfs] > [81738.886299] [<ffffffff8118fe27>] add_to_page_cache_lru+0x37/0x90 > [81738.886316] [<ffffffff8119df5b>] read_cache_pages+0x7b/0x120 > [81738.886401] [<ffffffffc141de79>] nfs_readpages+0x129/0x200 [nfs] > [81738.886429] [<ffffffff8119e1e6>] __do_page_cache_readahead+0x196/0x230 > [81738.886454] [<ffffffff8119e3bd>] ondemand_readahead+0x13d/0x250 > [81738.886469] [<ffffffff8119e53b>] page_cache_async_readahead+0x6b/0x70 > [81738.886478] [<ffffffff81190dd8>] generic_file_read_iter+0x458/0x670 > [81738.886529] [<ffffffffc1415292>] ? __nfs_revalidate_mapping+0xc2/0x280 [nfs] > [81738.886564] [<ffffffffc1410c72>] nfs_file_read+0x52/0xa0 [nfs] > [81738.886577] [<ffffffff8120eec4>] new_sync_read+0x94/0xd0 > [81738.886584] [<ffffffff8120ef26>] __vfs_read+0x26/0x40 > [81738.886589] [<ffffffff8120f4e6>] vfs_read+0x86/0x130 > [81738.886595] [<ffffffff81210235>] SyS_read+0x55/0xc0 > [81738.886606] [<ffffffff81841ff2>] entry_SYSCALL_64_fastpath+0x16/0x71 > [81738.886612] ---[ end trace dbbc7c60b0e5f2a3 ]--- > > [81738.886634] FS-Cache: > [81738.888281] FS-Cache: Assertion failed > [81738.889461] FS-Cache: 6 == 5 is false > [81738.890625] ------------[ cut here ]------------ > [81738.891706] kernel BUG at /build/linux-hVVhWi/linux-4.4.0/fs/fscache/operation.c:494! > [81738.892821] invalid opcode: 0000 [#1] SMP > [81738.893930] Modules linked in: binfmt_misc veth ipt_MASQUERADE nf_nat_masquerade_ipv4 nf_conntrack_netlink nfnetlink xfrm_user xfrm_algo iptable_nat nf_conntrack_ipv4 nf_defrag_ipv4 nf_nat_ipv4 xt_addrtype xt_conntrack nf_nat nf_conntrack br_netfilter bridge stp llc overlay xt_multiport iptable_filter ip_tables x_tables nfsv3 nfs_acl rpcsec_gss_krb5 auth_rpcgss nfsv4 nfs lockd grace nv_peer_mem(OE) cachefiles fscache msr rdma_ucm(OE) ib_ucm(OE) rdma_cm(OE) iw_cm(OE) configfs ib_ipoib(OE) ib_cm(OE) ib_uverbs(OE) ib_umad(OE) mlx5_ib(OE) mlx4_ib(OE) ib_core(OE) mlx4_en(OE) mlx4_core(OE) bonding ipmi_ssif intel_rapl x86_pkg_temp_thermal intel_powerclamp coretemp kvm_intel kvm irqbypass crct10dif_pclmul crc32_pclmul ghash_clmulni_intel aesni_intel aes_x86_64 lrw gf128mul glue_helper ablk_helper cryptd joydev input_leds mxm_wmi sb_edac edac_core lpc_ich mei_me mei nvidia_uvm(POE) shpchp 8250_fintek ipmi_si wmi acpi_power_meter mac_hid sunrpc ipmi_devintf ipmi_msghandler knem(OE) autofs4 > nvidia_drm(POE) nvidia_modeset(POE) ast i2c_algo_bit nvidia(POE) ttm ixgbe drm_kms_helper syscopyarea ses dca sysfillrect mlx5_core(OE) enclosure sysimgblt fb_sys_fops vxlan hid_generic ip6_udp_tunnel usbhid udp_tunnel mlx_compat(OE) hid ahci ptp megaraid_sas drm libahci pps_core mdio fjes > [81738.908228] CPU: 23 PID: 36247 Comm: kworker/u162:0 Tainted: P W OE 4.4.0-92-generic #115-Ubuntu > [81738.910190] Hardware name: xxxxx > [81738.912691] Workqueue: fscache_operation fscache_op_work_func [fscache] > [81738.914032] task: ffff8846c0f23800 ti: ffff886cf2f0c000 task.ti: ffff886cf2f0c000 > [81738.914880] RIP: 0010:[<ffffffffc054af96>] [<ffffffffc054af96>] fscache_put_operation+0x1d6/0x210 [fscache] > [81738.915788] RSP: 0000:ffff886cf2f0fde0 EFLAGS: 00010282 > [81738.916700] RAX: 0000000000000019 RBX: ffff887f626bc540 RCX: 0000000000000006 > [81738.917635] RDX: 0000000000000000 RSI: 0000000000000246 RDI: ffff887f7eccdd50 > [81738.918571] RBP: ffff886cf2f0fdf8 R08: 0000000000000000 R09: 000000000000079e > [81738.919539] R10: 0000000000000002 R11: 000000000000079e R12: ffff887f656e8000 > [81738.920454] R13: ffff887f22fdd200 R14: 0000000000001440 R15: ffff887f626bc540 > [81738.921275] FS: 0000000000000000(0000) GS:ffff887f7ecc0000(0000) knlGS:0000000000000000 > [81738.922089] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 > [81738.922922] CR2: 00007ff2cce7e880 CR3: 0000000002e0a000 CR4: 00000000003406e0 > [81738.923765] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 > [81738.924650] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 > [81738.925517] Stack: > [81738.926320] ffff887f626bc540 ffff887f656e8000 ffff887f22fdd200 ffff886cf2f0fe10 > [81738.927155] ffffffffc054b1ca ffff887f608c3ec0 ffff886cf2f0fe50 ffffffff8109a625 > [81738.928059] 0000000000000000 ffff887f608c3ef0 ffff887f656e8000 0000000000000088 > [81738.928961] Call Trace: > [81738.929825] [<ffffffffc054b1ca>] fscache_op_work_func+0x2a/0x50 [fscache] > [81738.930680] [<ffffffff8109a625>] process_one_work+0x165/0x480 > [81738.931547] [<ffffffff8109a98b>] worker_thread+0x4b/0x4c0 > [81738.932427] [<ffffffff8109a940>] ? process_one_work+0x480/0x480 > [81738.933506] [<ffffffff810a0cc5>] kthread+0xe5/0x100 > [81738.934498] [<ffffffff810a0be0>] ? kthread_create_on_node+0x1e0/0x1e0 > [81738.935551] [<ffffffff8184238f>] ret_from_fork+0x3f/0x70 > [81738.936519] [<ffffffff810a0be0>] ? kthread_create_on_node+0x1e0/0x1e0 > [81738.937372] Code: c7 c7 fb 02 55 c0 e8 e1 2b c4 c0 48 c7 c7 09 03 55 c0 e8 d5 2b c4 c0 8b 73 40 ba 05 00 00 00 48 c7 c7 b0 ee 54 c0 e8 c1 2b c4 c0 <0f> 0b 48 c7 c7 fb 02 55 c0 e8 b3 2b c4 c0 48 c7 c7 09 03 55 c0 > [81738.939258] RIP [<ffffffffc054af96>] fscache_put_operation+0x1d6/0x210 [fscache] > [81738.940185] RSP <ffff886cf2f0fde0> > [81738.944238] ---[ end trace dbbc7c60b0e5f2a4 ]--- > > (this usually follows, it just looks like a cascading failure from > messing with the work queue data structures) > > [81738.946803] BUG: unable to handle kernel paging request at ffffffffffffffd8 > [81738.947768] IP: [<ffffffff810a1370>] kthread_data+0x10/0x20 > [81738.948624] PGD 2e0d067 PUD 2e0f067 PMD 0 > [81738.949445] Oops: 0000 [#2] SMP > [81738.950237] Modules linked in: binfmt_misc veth ipt_MASQUERADE nf_nat_masquerade_ipv4 nf_conntrack_netlink nfnetlink xfrm_user xfrm_algo iptable_nat nf_conntrack_ipv4 nf_defrag_ipv4 nf_nat_ipv4 xt_addrtype xt_conntrack nf_nat nf_conntrack br_netfilter bridge stp llc overlay xt_multiport iptable_filter ip_tables x_tables nfsv3 nfs_acl rpcsec_gss_krb5 auth_rpcgss nfsv4 nfs lockd grace nv_peer_mem(OE) cachefiles fscache msr rdma_ucm(OE) ib_ucm(OE) rdma_cm(OE) iw_cm(OE) configfs ib_ipoib(OE) ib_cm(OE) ib_uverbs(OE) ib_umad(OE) mlx5_ib(OE) mlx4_ib(OE) ib_core(OE) mlx4_en(OE) mlx4_core(OE) bonding ipmi_ssif intel_rapl x86_pkg_temp_thermal intel_powerclamp coretemp kvm_intel kvm irqbypass crct10dif_pclmul crc32_pclmul ghash_clmulni_intel aesni_intel aes_x86_64 lrw gf128mul glue_helper ablk_helper cryptd joydev input_leds mxm_wmi sb_edac edac_core lpc_ich mei_me mei nvidia_uvm(POE) shpchp 8250_fintek ipmi_si wmi acpi_power_meter mac_hid sunrpc ipmi_devintf ipmi_msghandler knem(OE) autofs4 > nvidia_drm(POE) nvidia_modeset(POE) ast i2c_algo_bit nvidia(POE) ttm ixgbe drm_kms_helper syscopyarea ses dca sysfillrect mlx5_core(OE) enclosure sysimgblt fb_sys_fops vxlan hid_generic ip6_udp_tunnel usbhid udp_tunnel mlx_compat(OE) hid ahci ptp megaraid_sas drm libahci pps_core mdio fjes > [81738.959336] CPU: 23 PID: 36247 Comm: kworker/u162:0 Tainted: P D W OE 4.4.0-92-generic #115-Ubuntu > [81738.960164] Hardware name: xxxxx > [81738.960993] task: ffff8846c0f23800 ti: ffff886cf2f0c000 task.ti: ffff886cf2f0c000 > [81738.961808] RIP: 0010:[<ffffffff810a1370>] [<ffffffff810a1370>] kthread_data+0x10/0x20 > [81738.962644] RSP: 0000:ffff886cf2f0fab8 EFLAGS: 00010002 > [81738.963475] RAX: 0000000000000000 RBX: 0000000000000017 RCX: ffffffff82109e80 > [81738.964312] RDX: 0000000000000017 RSI: 0000000000000017 RDI: ffff8846c0f23800 > [81738.965142] RBP: ffff886cf2f0fab8 R08: 00000000ffffffff R09: 0000000000000000 > [81738.965996] R10: ffff8846c0f23860 R11: 0000000000009800 R12: 0000000000000000 > [81738.966825] R13: 0000000000016dc0 R14: ffff8846c0f23800 R15: ffff887f7ecd6dc0 > [81738.967676] FS: 0000000000000000(0000) GS:ffff887f7ecc0000(0000) knlGS:0000000000000000 > [81738.968534] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 > [81738.969369] CR2: 0000000000000028 CR3: 0000000002e0a000 CR4: 00000000003406e0 > [81738.970202] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 > [81738.971053] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 > [81738.971893] Stack: > [81738.972728] ffff886cf2f0fad0 ffffffff8109b9c1 ffff887f7ecd6dc0 ffff886cf2f0fb20 > [81738.973583] ffffffff8183dab0 ffff887f5f808aa0 ffff886c00000017 ffff8846c0f23800 > [81738.974448] ffff886cf2f10000 ffff8846c0f23ed0 ffff886cf2f0f6d0 0000000000000000 > [81738.975289] Call Trace: > [81738.976164] [<ffffffff8109b9c1>] wq_worker_sleeping+0x11/0x90 > [81738.977019] [<ffffffff8183dab0>] __schedule+0x650/0xa30 > [81738.977844] [<ffffffff8183dec5>] schedule+0x35/0x80 > [81738.978668] [<ffffffff81084465>] do_exit+0x775/0xb00 > [81738.979484] [<ffffffff81031c41>] oops_end+0xa1/0xd0 > [81738.980294] [<ffffffff810320fb>] die+0x4b/0x70 > [81738.981091] [<ffffffff8102f121>] do_trap+0xb1/0x140 > [81738.981882] [<ffffffff8102f4a9>] do_error_trap+0x89/0x110 > [81738.982656] [<ffffffffc054af96>] ? fscache_put_operation+0x1d6/0x210 [fscache] > [81738.983428] [<ffffffff810ca782>] ? up+0x32/0x50 > [81738.984189] [<ffffffff810d8656>] ? console_unlock+0x216/0x550 > [81738.984934] [<ffffffff8102fa10>] do_invalid_op+0x20/0x30 > [81738.985677] [<ffffffff81843b0e>] invalid_op+0x1e/0x30 > [81738.986416] [<ffffffffc054af96>] ? fscache_put_operation+0x1d6/0x210 [fscache] > [81738.987155] [<ffffffffc054b1ca>] fscache_op_work_func+0x2a/0x50 [fscache] > [81738.987890] [<ffffffff8109a625>] process_one_work+0x165/0x480 > [81738.988631] [<ffffffff8109a98b>] worker_thread+0x4b/0x4c0 > [81738.989349] [<ffffffff8109a940>] ? process_one_work+0x480/0x480 > [81738.990049] [<ffffffff810a0cc5>] kthread+0xe5/0x100 > [81738.990724] [<ffffffff810a0be0>] ? kthread_create_on_node+0x1e0/0x1e0 > [81738.991390] [<ffffffff8184238f>] ret_from_fork+0x3f/0x70 > [81738.992042] [<ffffffff810a0be0>] ? kthread_create_on_node+0x1e0/0x1e0 > [81738.992671] Code: ff ff ff be 49 02 00 00 48 c7 c7 98 b9 cb 81 e8 e7 00 fe ff e9 a6 fe ff ff 66 90 0f 1f 44 00 00 48 8b 87 18 05 00 00 55 48 89 e5 <48> 8b 40 d8 5d c3 66 2e 0f 1f 84 00 00 00 00 00 0f 1f 44 00 00 > [81738.994013] RIP [<ffffffff810a1370>] kthread_data+0x10/0x20 > [81738.994627] RSP <ffff886cf2f0fab8> > [81738.995216] CR2: ffffffffffffffd8 > [81738.995785] ---[ end trace dbbc7c60b0e5f2a5 ]--- > [81738.997887] Fixing recursive fault but reboot is needed! > > (hardlockups seem to follow) > > -- > Linux-cachefs mailing list > Linux-cachefs@xxxxxxxxxx > https://www.redhat.com/mailman/listinfo/linux-cachefs -- Linux-cachefs mailing list Linux-cachefs@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/linux-cachefs
