whence_is_name is false, so it's not a whence-is-name FSAL.
Thanks Dang.
Nothing directly related to the content lock. However, there have been
several use-after-free races fixed, and it's not impossible that trying
to interacte with a freed lock will cause a hang. The commits in
question are:
11e0e375e40658267cbf449afacaa53a136f7097
eb98f5b855147f44e79fe08dcec8d5057b05ea30
There have been quite a few fixes to readdir since 2.7.6 (and
specifically to whence-is-name; I believe your FSAL is a whence-is-name
FSAL?) so you might want to look into updating to a newer version.
2.8.4 should come out next week, and 3.3 soon after that.
Daniel
On 4/23/20 11:17 AM, QR wrote:
> Hi Dang, I generate a core dump for this.
>
> It seems something wrong with "entry->content_lock".
> Is there a known issue for this? Thanks in advance.
>
> Ganesha server info
> ganesha version: V2.7.6
> FSAL : In house
> nfs client info
> nfs version : nfs v3
> client info : Centos 7.4
>
> =======================================================================================================================================================================
> (gdb) thread 204
> [Switching to thread 204 (Thread 0x7fa7932f2700 (LWP 348))]
> #0 0x00007faa5ba2cf4d in __lll_lock_wait () from /lib64/libpthread.so.0
> (gdb) bt
> #0 0x00007faa5ba2cf4d in __lll_lock_wait () from /lib64/libpthread.so.0
> #1 0x00007faa5ba28d02 in _L_lock_791 () from /lib64/libpthread.so.0
> #2 0x00007faa5ba28c08 in pthread_mutex_lock () from /lib64/libpthread.so.0
> #3 0x0000000000528657 in _mdcache_lru_unref_chunk
> (chunk=0x7fa7e03e31d0, func=0x59abd0 <__func__.20247>
> "mdcache_clean_dirent_chunks", line=579)
> at
> /export/jcloud-zbs/src/
jd.com/zfs/FSAL_SkyFS/nfs-ganesha/src/FSAL/Stackable_FSALs/FSAL_MDCACHE/mdcache_lru.c:2066> #4 0x000000000053894c in mdcache_clean_dirent_chunks (entry=0x7fa740035400)
> at
> /export/jcloud-zbs/src/
jd.com/zfs/FSAL_SkyFS/nfs-ganesha/src/FSAL/Stackable_FSALs/FSAL_MDCACHE/mdcache_helpers.c:578> #5 0x0000000000538a30 in mdcache_dirent_invalidate_all
> (entry=0x7fa740035400)
> at
> /export/jcloud-zbs/src/
jd.com/zfs/FSAL_SkyFS/nfs-ganesha/src/FSAL/Stackable_FSALs/FSAL_MDCACHE/mdcache_helpers.c:603> #6 0x00000000005376fc in mdc_clean_entry (entry=0x7fa740035400) at
> /export/jcloud-zbs/src/
jd.com/zfs/FSAL_SkyFS/nfs-ganesha/src/FSAL/Stackable_FSALs/FSAL_MDCACHE/mdcache_helpers.c:302> #7 0x0000000000523559 in mdcache_lru_clean (entry=0x7fa740035400) at
> /export/jcloud-zbs/src/
jd.com/zfs/FSAL_SkyFS/nfs-ganesha/src/FSAL/Stackable_FSALs/FSAL_MDCACHE/mdcache_lru.c:592> #8 0x00000000005278b1 in mdcache_lru_get (sub_handle=0x7fa7383cdf40) at
> /export/jcloud-zbs/src/
jd.com/zfs/FSAL_SkyFS/nfs-ganesha/src/FSAL/Stackable_FSALs/FSAL_MDCACHE/mdcache_lru.c:1821> #9 0x0000000000536e91 in _mdcache_alloc_handle (export=0x1b4eed0,
> sub_handle=0x7fa7383cdf40, fs=0x0, reason=MDC_REASON_DEFAULT,
> func=0x59ac10 <__func__.20274> "mdcache_new_entry", line=691)
> at
> /export/jcloud-zbs/src/
jd.com/zfs/FSAL_SkyFS/nfs-ganesha/src/FSAL/Stackable_FSALs/FSAL_MDCACHE/mdcache_helpers.c:174> #10 0x0000000000538c50 in mdcache_new_entry (export=0x1b4eed0,
> sub_handle=0x7fa7383cdf40, attrs_in=0x7fa7932f00a0,
> attrs_out=0x7fa7932f0760, new_directory=false, entry=0x7fa7932f0018,
> state=0x0, reason=MDC_REASON_DEFAULT) at
> /export/jcloud-zbs/src/
jd.com/zfs/FSAL_SkyFS/nfs-ganesha/src/FSAL/Stackable_FSALs/FSAL_MDCACHE/mdcache_helpers.c:690> #11 0x000000000052d156 in mdcache_alloc_and_check_handle
> (export=0x1b4eed0, sub_handle=0x7fa7383cdf40, new_obj=0x7fa7932f01b0,
> new_directory=false, attrs_in=0x7fa7932f00a0,
> attrs_out=0x7fa7932f0760, tag=0x5999a4 "lookup ",
> parent=0x7fa878288cb0, name=0x7fa738229270 "cer_7_0.5",
> invalidate=0x7fa7932f009f, state=0x0)
> at
> /export/jcloud-zbs/src/
jd.com/zfs/FSAL_SkyFS/nfs-ganesha/src/FSAL/Stackable_FSALs/FSAL_MDCACHE/mdcache_handle.c:100> #12 0x000000000053aea1 in mdc_lookup_uncached
> (mdc_parent=0x7fa878288cb0, name=0x7fa738229270 "cer_7_0.5",
> new_entry=0x7fa7932f02c8, attrs_out=0x7fa7932f0760)
> at
> /export/jcloud-zbs/src/
jd.com/zfs/FSAL_SkyFS/nfs-ganesha/src/FSAL/Stackable_FSALs/FSAL_MDCACHE/mdcache_helpers.c:1391> #13 0x000000000053ab27 in mdc_lookup (mdc_parent=0x7fa878288cb0,
> name=0x7fa738229270 "cer_7_0.5", uncached=true,
> new_entry=0x7fa7932f02c8, attrs_out=0x7fa7932f0760)
> at
> /export/jcloud-zbs/src/
jd.com/zfs/FSAL_SkyFS/nfs-ganesha/src/FSAL/Stackable_FSALs/FSAL_MDCACHE/mdcache_helpers.c:1325> #14 0x000000000052d5bb in mdcache_lookup (parent=0x7fa878288ce8,
> name=0x7fa738229270 "cer_7_0.5", handle=0x7fa7932f0878,
> attrs_out=0x7fa7932f0760)
> at
> /export/jcloud-zbs/src/
jd.com/zfs/FSAL_SkyFS/nfs-ganesha/src/FSAL/Stackable_FSALs/FSAL_MDCACHE/mdcache_handle.c:181> #15 0x0000000000431a60 in fsal_lookup (parent=0x7fa878288ce8,
> name=0x7fa738229270 "cer_7_0.5", obj=0x7fa7932f0878,
> attrs_out=0x7fa7932f0760)
> at
> /export/jcloud-zbs/src/
jd.com/zfs/FSAL_SkyFS/nfs-ganesha/src/FSAL/fsal_helper.c:683> #16 0x000000000048e83e in nfs3_lookup (arg=0x7fa7381bdc88,
> req=0x7fa7381bd580, res=0x7fa73839dab0)
> at
> /export/jcloud-zbs/src/
jd.com/zfs/FSAL_SkyFS/nfs-ganesha/src/Protocols/NFS/nfs3_lookup.c:104> #17 0x000000000045703b in nfs_rpc_process_request
> (reqdata=0x7fa7381bd580) at
> /export/jcloud-zbs/src/
jd.com/zfs/FSAL_SkyFS/nfs-ganesha/src/MainNFSD/nfs_worker_thread.c:1329> #18 0x000000000045781f in nfs_rpc_process_request_slowio
> (reqdata=0x7fa7381bd580, slowio_check_cb=0x553b4a <nfs3_timeout_proc>)
> at
> /export/jcloud-zbs/src/
jd.com/zfs/FSAL_SkyFS/nfs-ganesha/src/MainNFSD/nfs_worker_thread.c:1542> #19 0x0000000000457955 in nfs_rpc_valid_NFS (req=0x7fa7381bd580) at
> /export/jcloud-zbs/src/
jd.com/zfs/FSAL_SkyFS/nfs-ganesha/src/MainNFSD/nfs_worker_thread.c:1586> #20 0x00007faa5c901435 in svc_vc_decode (req=0x7fa7381bd580) at
> /export/jcloud-zbs/src/
jd.com/zfs/FSAL_SkyFS/nfs-ganesha/src/libntirpc/src/svc_vc.c:829> #21 0x000000000044a497 in nfs_rpc_decode_request (xprt=0x7fa8840027d0,
> xdrs=0x7fa7384befa0)
> at
> /export/jcloud-zbs/src/
jd.com/zfs/FSAL_SkyFS/nfs-ganesha/src/MainNFSD/nfs_rpc_dispatcher_thread.c:1345> #22 0x00007faa5c901346 in svc_vc_recv (xprt=0x7fa8840027d0) at
> /export/jcloud-zbs/src/
jd.com/zfs/FSAL_SkyFS/nfs-ganesha/src/libntirpc/src/svc_vc.c:802> #23 0x00007faa5c8fda92 in svc_rqst_xprt_task (wpe=0x7fa8840029e8) at
> /export/jcloud-zbs/src/
jd.com/zfs/FSAL_SkyFS/nfs-ganesha/src/libntirpc/src/svc_rqst.c:769> #24 0x00007faa5c8fdeea in svc_rqst_epoll_events (sr_rec=0x1b62c60,
> n_events=1) at
> /export/jcloud-zbs/src/
jd.com/zfs/FSAL_SkyFS/nfs-ganesha/src/libntirpc/src/svc_rqst.c:941> #25 0x00007faa5c8fe17f in svc_rqst_epoll_loop (sr_rec=0x1b62c60) at
> /export/jcloud-zbs/src/
jd.com/zfs/FSAL_SkyFS/nfs-ganesha/src/libntirpc/src/svc_rqst.c:1014> #26 0x00007faa5c8fe232 in svc_rqst_run_task (wpe=0x1b62c60) at
> /export/jcloud-zbs/src/
jd.com/zfs/FSAL_SkyFS/nfs-ganesha/src/libntirpc/src/svc_rqst.c:1050> #27 0x00007faa5c906dbb in work_pool_thread (arg=0x7fa7540008c0) at
> /export/jcloud-zbs/src/
jd.com/zfs/FSAL_SkyFS/nfs-ganesha/src/libntirpc/src/work_pool.c:181> #28 0x00007faa5ba26dc5 in start_thread () from /lib64/libpthread.so.0
> #29 0x00007faa5b33321d in clone () from /lib64/libc.so.6
> (gdb) frame 3
> #3 0x0000000000528657 in _mdcache_lru_unref_chunk
> (chunk=0x7fa7e03e31d0, func=0x59abd0 <__func__.20247>
> "mdcache_clean_dirent_chunks", line=579)
> at
> /export/jcloud-zbs/src/
jd.com/zfs/FSAL_SkyFS/nfs-ganesha/src/FSAL/Stackable_FSALs/FSAL_MDCACHE/mdcache_lru.c:2066> 2066in
> /export/jcloud-zbs/src/
jd.com/zfs/FSAL_SkyFS/nfs-ganesha/src/FSAL/Stackable_FSALs/FSAL_MDCACHE/mdcache_lru.c> (gdb) p qlane->mtx
> $42 = {__data = {__lock = 2, __count = 0, *__owner = 225*, __nusers = 1,
> __kind = 0, __spins = 0, __elision = 0, __list = {__prev = 0x0, __next =
> 0x0}},
> __size = "\002\000\000\000\000\000\000\000\341\000\000\000\001",
> '\000' <repeats 26 times>, __align = 2}
> =======================================================================================================================================================================
> (gdb) info threads
> 84 Thread 0x7fa9a68e8700 (LWP 228) 0x00007faa5ba2a03e in
> pthread_rwlock_wrlock () from /lib64/libpthread.so.0
> 83 Thread 0x7fa9a69e9700 (LWP 227) 0x00007faa5ba29e24 in
> pthread_rwlock_rdlock () from /lib64/libpthread.so.0
> *82 Thread 0x7fa9a6aea700 (LWP 225)* 0x00007faa5ba2cf4d in
> __lll_lock_wait () from /lib64/libpthread.so.0
> 81 Thread 0x7fa9a6beb700 (LWP 226) 0x00007faa5ba2a03e in
> pthread_rwlock_wrlock () from /lib64/libpthread.so.0
> 80 Thread 0x7fa9a6cec700 (LWP 224) 0x00007faa5ba2a03e in
> pthread_rwlock_wrlock () from /lib64/libpthread.so.0
> 79 Thread 0x7fa9a6ded700 (LWP 223) 0x00007faa5ba29e24 in
> pthread_rwlock_rdlock () from /lib64/libpthread.so.0
> 78 Thread 0x7fa9a6eee700 (LWP 222) 0x00007faa5ba2a03e in
> pthread_rwlock_wrlock () from /lib64/libpthread.so.0
> 77 Thread 0x7fa9a6fef700 (LWP 221) 0x00007faa5ba29e24 in
> pthread_rwlock_rdlock () from /lib64/libpthread.so.0
> 76 Thread 0x7fa9a70f0700 (LWP 220) 0x00007faa5ba2a03e in
> pthread_rwlock_wrlock () from /lib64/libpthread.so.0
> 75 Thread 0x7fa9a71f1700 (LWP 219) 0x00007faa5ba2a03e in
> pthread_rwlock_wrlock () from /lib64/libpthread.so.0
> 74 Thread 0x7fa9a72f2700 (LWP 218) 0x00007faa5ba29e24 in
> pthread_rwlock_rdlock () from /lib64/libpthread.so.0
> 73 Thread 0x7fa9a73f3700 (LWP 217) 0x00007faa5ba2a03e in
> pthread_rwlock_wrlock () from /lib64/libpthread.so.0
> =======================================================================================================================================================================
> (gdb) *thread 82*
> [Switching to thread 82 (Thread 0x7fa9a6aea700 (LWP 225))]
> #0 0x00007faa5ba2cf4d in __lll_lock_wait () from /lib64/libpthread.so.0
> (gdb) bt
> #0 0x00007faa5ba2cf4d in __lll_lock_wait () from /lib64/libpthread.so.0
> #1 0x00007faa5ba2a307 in _L_lock_14 () from /lib64/libpthread.so.0
> #2 0x00007faa5ba2a2b3 in pthread_rwlock_trywrlock () from
> /lib64/libpthread.so.0
> #3 0x0000000000524547 in *lru_reap_chunk_impl* (qid=LRU_ENTRY_L2,
> parent=0x7fa7f8122810)
> at
> /export/jcloud-zbs/src/
jd.com/zfs/FSAL_SkyFS/nfs-ganesha/src/FSAL/Stackable_FSALs/FSAL_MDCACHE/mdcache_lru.c:809> #4 0x0000000000524a01 in mdcache_get_chunk (parent=0x7fa7f8122810,
> prev_chunk=0x0, whence=0)
> at
> /export/jcloud-zbs/src/
jd.com/zfs/FSAL_SkyFS/nfs-ganesha/src/FSAL/Stackable_FSALs/FSAL_MDCACHE/mdcache_lru.c:886> #5 0x000000000053f0b6 in mdcache_populate_dir_chunk
> (directory=0x7fa7f8122810, whence=0, dirent=0x7fa9a6ae7f80,
> prev_chunk=0x0, eod_met=0x7fa9a6ae7f7f)
> at
> /export/jcloud-zbs/src/
jd.com/zfs/FSAL_SkyFS/nfs-ganesha/src/FSAL/Stackable_FSALs/FSAL_MDCACHE/mdcache_helpers.c:2554> #6 0x0000000000540aff in mdcache_readdir_chunked
> (directory=0x7fa7f8122810, whence=0, dir_state=0x7fa9a6ae8130,
> cb=0x43215e <populate_dirent>, attrmask=122830, eod_met=0x7fa9a6ae884b)
> at
> /export/jcloud-zbs/src/
jd.com/zfs/FSAL_SkyFS/nfs-ganesha/src/FSAL/Stackable_FSALs/FSAL_MDCACHE/mdcache_helpers.c:2991> #7 0x000000000052ee71 in mdcache_readdir (dir_hdl=0x7fa7f8122848,
> whence=0x7fa9a6ae8110, dir_state=0x7fa9a6ae8130, cb=0x43215e
> <populate_dirent>, attrmask=122830, eod_met=0x7fa9a6ae884b)
> at
> /export/jcloud-zbs/src/
jd.com/zfs/FSAL_SkyFS/nfs-ganesha/src/FSAL/Stackable_FSALs/FSAL_MDCACHE/mdcache_handle.c:559> #8 0x0000000000432a85 in fsal_readdir (directory=0x7fa7f8122848,
> cookie=0, nbfound=0x7fa9a6ae884c, eod_met=0x7fa9a6ae884b,
> attrmask=122830, cb=0x491b0f <nfs3_readdirplus_callback>,
> opaque=0x7fa9a6ae8800) at
> /export/jcloud-zbs/src/
jd.com/zfs/FSAL_SkyFS/nfs-ganesha/src/FSAL/fsal_helper.c:1164> #9 0x0000000000491968 in nfs3_readdirplus (arg=0x7fa91470b9f8,
> req=0x7fa91470b2f0, res=0x7fa91407b8f0)
> at
> /export/jcloud-zbs/src/
jd.com/zfs/FSAL_SkyFS/nfs-ganesha/src/Protocols/NFS/nfs3_readdirplus.c:310> #10 0x000000000045703b in nfs_rpc_process_request
> (reqdata=0x7fa91470b2f0) at
> /export/jcloud-zbs/src/
jd.com/zfs/FSAL_SkyFS/nfs-ganesha/src/MainNFSD/nfs_worker_thread.c:1329> #11 0x000000000045781f in nfs_rpc_process_request_slowio
> (reqdata=0x7fa91470b2f0, slowio_check_cb=0x553b4a <nfs3_timeout_proc>)
> at
> /export/jcloud-zbs/src/
jd.com/zfs/FSAL_SkyFS/nfs-ganesha/src/MainNFSD/nfs_worker_thread.c:1542> #12 0x0000000000457955 in nfs_rpc_valid_NFS (req=0x7fa91470b2f0) at
> /export/jcloud-zbs/src/
jd.com/zfs/FSAL_SkyFS/nfs-ganesha/src/MainNFSD/nfs_worker_thread.c:1586> #13 0x00007faa5c901435 in svc_vc_decode (req=0x7fa91470b2f0) at
> /export/jcloud-zbs/src/
jd.com/zfs/FSAL_SkyFS/nfs-ganesha/src/libntirpc/src/svc_vc.c:829> #14 0x000000000044a497 in nfs_rpc_decode_request (xprt=0x7fa8840027d0,
> xdrs=0x7fa9142a1370)
> at
> /export/jcloud-zbs/src/
jd.com/zfs/FSAL_SkyFS/nfs-ganesha/src/MainNFSD/nfs_rpc_dispatcher_thread.c:1345> #15 0x00007faa5c901346 in svc_vc_recv (xprt=0x7fa8840027d0) at
> /export/jcloud-zbs/src/
jd.com/zfs/FSAL_SkyFS/nfs-ganesha/src/libntirpc/src/svc_vc.c:802> #16 0x00007faa5c8fda92 in svc_rqst_xprt_task (wpe=0x7fa8840029e8) at
> /export/jcloud-zbs/src/
jd.com/zfs/FSAL_SkyFS/nfs-ganesha/src/libntirpc/src/svc_rqst.c:769> #17 0x00007faa5c8fdeea in svc_rqst_epoll_events (sr_rec=0x1b62c60,
> n_events=1) at
> /export/jcloud-zbs/src/
jd.com/zfs/FSAL_SkyFS/nfs-ganesha/src/libntirpc/src/svc_rqst.c:941> #18 0x00007faa5c8fe17f in svc_rqst_epoll_loop (sr_rec=0x1b62c60) at
> /export/jcloud-zbs/src/
jd.com/zfs/FSAL_SkyFS/nfs-ganesha/src/libntirpc/src/svc_rqst.c:1014> #19 0x00007faa5c8fe232 in svc_rqst_run_task (wpe=0x1b62c60) at
> /export/jcloud-zbs/src/
jd.com/zfs/FSAL_SkyFS/nfs-ganesha/src/libntirpc/src/svc_rqst.c:1050> #20 0x00007faa5c906dbb in work_pool_thread (arg=0x7fa9380008c0) at
> /export/jcloud-zbs/src/
jd.com/zfs/FSAL_SkyFS/nfs-ganesha/src/libntirpc/src/work_pool.c:181> #21 0x00007faa5ba26dc5 in start_thread () from /lib64/libpthread.so.0
> #22 0x00007faa5b33321d in clone () from /lib64/libc.so.6
> (gdb) frame 3
> #3 0x0000000000524547 in lru_reap_chunk_impl (qid=LRU_ENTRY_L2,
> parent=0x7fa7f8122810)
> at
> /export/jcloud-zbs/src/
jd.com/zfs/FSAL_SkyFS/nfs-ganesha/src/FSAL/Stackable_FSALs/FSAL_MDCACHE/mdcache_lru.c:809> 809in
> /export/jcloud-zbs/src/
jd.com/zfs/FSAL_SkyFS/nfs-ganesha/src/FSAL/Stackable_FSALs/FSAL_MDCACHE/mdcache_lru.c> (gdb) p entry->content_lock
> $43 = {__data = {__lock = 2, __nr_readers = *32681*, __readers_wakeup =
> 0, __writer_wakeup = 0, __nr_readers_queued = 0, __nr_writers_queued =
> 0, __writer = -598294472, __shared = 32680,
> __pad1 = 140366914447067, __pad2 = 0, __flags = 0},
> __size = "\002\000\000\000\251\177", '\000' <repeats 18 times>,
> "\070\300V\334\250\177\000\000\333\346\022\270\251\177", '\000' <repeats
> 17 times>, __align = 140363826200578}
> =======================================================================================================================================================================
>
>
> --------------------------------
>
>
> ----- 原始邮件 -----
> 发件人："QR" <
zhbingyin@sina.com>
> 收件人："Daniel Gryniewicz" <
dgryniew@redhat.com>, "ganesha-devel"
> <
devel@lists.nfs-ganesha.org>,
> 主题：[NFS-Ganesha-Devel]回复：Re:_ganesha_hang_more_than_five_hours
> 日期：2020年02月26日 18点21分
>
> Not yet. Because the docker did not enable core dump.
>
> Will try to create a full backtrace for this, thanks.
>
>
> --------------------------------
>
>
> ----- 原始邮件 -----
> 发件人：Daniel Gryniewicz <
dgryniew@redhat.com>
> 收件人：
devel@lists.nfs-ganesha.org> 主题：[NFS-Ganesha-Devel] Re: ganesha hang more than five hours
> 日期：2020年02月25日 21点20分
>
> No, definitely not a known issue. Do you have a full backtrace of one
> (or several, if they're different) hung threads?
> Daniel
> On 2/24/20 9:28 PM, QR wrote:
> > Hi Dang,
> >
> > Ganesha hangs more than five hours. It seems that 198 svc threads hang
> > on nfs3_readdirplus.
> > Is there a known issue about this? Thanks in advance.
> >
> > Ganesha server info
> > ganesha version: V2.7.6
> > FSAL : In house
> > nfs client info
> > nfs version : nfs v3
> > client info : CentOS 7.4
> >
> > _______________________________________________
> > Devel mailing list --
devel@lists.nfs-ganesha.org> > To unsubscribe send an email to
devel-leave@lists.nfs-ganesha.org> _______________________________________________
> Devel mailing list --
devel@lists.nfs-ganesha.org> To unsubscribe send an email to
devel-leave@lists.nfs-ganesha.org> _______________________________________________
> Devel mailing list --
devel@lists.nfs-ganesha.org> To unsubscribe send an email to
devel-leave@lists.nfs-ganesha.org>
> _______________________________________________
> Devel mailing list --
devel@lists.nfs-ganesha.org> To unsubscribe send an email to
devel-leave@lists.nfs-ganesha.org>
_______________________________________________
Devel mailing list --
devel@lists.nfs-ganesha.orgTo unsubscribe send an email to
devel-leave@lists.nfs-ganesha.org