Aaron Richton wrote:
On Sat, 24 Jan 2009, Howard Chu wrote:
I'm still finding odd problems in HEAD, so you should wait for a new call for testing.
With HEAD, I got to test050 before suffering a deadlock. Not sure if this is worth an ITS since I think it's the area actively being worked on...
Right. Thanks for the trace. It shows a writer is blocked waiting for the socket to become writable. What happened to the other servers at this point?
Just for reference, this is consumer1:
t@1 (l@1) stopped in __lwp_wait at 0x7fb1ff64 0x7fb1ff64: __lwp_wait+0x0004: ta %icc,0x00000008 current thread: t@1 [1] __lwp_wait(0x2, 0xffbff2bc, 0x7f98f9e0, 0x7f9424fc, 0x1, 0xffbff284), at 0x7fb1ff64 [2] lwp_wait(0x2, 0xffbff2bc, 0x2cf98, 0x7f984e70, 0x5, 0xffbff2b4), at 0x7f94d1cc [3] _thrp_join(0x2, 0x0, 0x0, 0x1, 0x81010100, 0xff00), at 0x7f9490c4 =>[4] ldap_pvt_thread_join(thread = 2U, thread_return = (nil)), line 197 in "thr_posix.c" [5] slapd_daemon(), line 2658 in "daemon.c" [6] main(argc = 8, argv = 0xffbff4dc), line 948 in "main.c" t@2 (l@2) stopped in _poll at 0x7fb1e238 0x7fb1e238: _poll+0x0004: ta %icc,0x00000008 current thread: t@2 [1] _poll(0x7effbb88, 0x3, 0xbb8, 0x0, 0x3, 0x7effbd91), at 0x7fb1e238 [2] select_large_fdset(0x13, 0x20, 0x7effe218, 0x0, 0x7effbd90, 0x7effbd90), at 0x7fad2b6c =>[3] slapd_daemon_task(ptr = (nil)), line 2291 in "daemon.c" t@3 (l@3) stopped in __lwp_park at 0x7f9554b0 0x7f9554b0: __lwp_park+0x0010: ta %icc,0x00000008 current thread: t@3 [1] __lwp_park(0x0, 0x0, 0x0, 0x1, 0x7f968000, 0x0), at 0x7f9554b0 [2] cond_wait_queue(0x3f3af8, 0x7f968c08, 0x0, 0x0, 0x7f870400, 0x7f968000), at 0x7f9526b8 [3] _cond_wait_cancel(0x3f3af8, 0x3f3ae0, 0x2cf0cc, 0x7e7ff920, 0x0, 0x0), at 0x7f952e74 [4] _pthread_cond_wait(0x3f3af8, 0x3f3ae0, 0x2cf0cc, 0x452e5c, 0x3, 0x0), at 0x7f952eb0 =>[5] ldap_pvt_thread_cond_wait(cond = 0x3f3af8, mutex = 0x3f3ae0), line 277 in "thr_posix.c" [6] ldap_int_thread_pool_wrapper(xpool = 0x3f3ad8), line 654 in "tpool.c" t@4 (l@4) stopped in __lwp_park at 0x7f9554b0 0x7f9554b0: __lwp_park+0x0010: ta %icc,0x00000008 current thread: t@4 [1] __lwp_park(0x0, 0x0, 0x0, 0x1, 0x7f968000, 0x0), at 0x7f9554b0 [2] cond_wait_queue(0x3f3af8, 0x7f968c08, 0x0, 0x0, 0x7f870600, 0x7f968000), at 0x7f9526b8 [3] _cond_wait_cancel(0x3f3af8, 0x3f3ae0, 0x2b91e0, 0x2b91e4, 0x4, 0x0), at 0x7f952e74 [4] _pthread_cond_wait(0x3f3af8, 0x3f3ae0, 0x1, 0x50acc4, 0x0, 0x0), at 0x7f952eb0 =>[5] ldap_pvt_thread_cond_wait(cond = 0x3f3af8, mutex = 0x3f3ae0), line 277 in "thr_posix.c" [6] ldap_int_thread_pool_wrapper(xpool = 0x3f3ad8), line 654 in "tpool.c" t@5 (l@5) stopped in __lwp_park at 0x7f9554b0 0x7f9554b0: __lwp_park+0x0010: ta %icc,0x00000008 current thread: t@5 [1] __lwp_park(0x0, 0x0, 0x0, 0x1, 0x7f968000, 0x0), at 0x7f9554b0 [2] cond_wait_queue(0x3f3af8, 0x7f968c08, 0x0, 0x0, 0x7f870800, 0x7f968000), at 0x7f9526b8 [3] _cond_wait_cancel(0x3f3af8, 0x3f3ae0, 0x2cf0cc, 0x7d7ff920, 0xb, 0x7d7ff3da), at 0x7f952e74 [4] _pthread_cond_wait(0x3f3af8, 0x3f3ae0, 0x2cf0cc, 0x4fe89c, 0x4, 0x0), at 0x7f952eb0 =>[5] ldap_pvt_thread_cond_wait(cond = 0x3f3af8, mutex = 0x3f3ae0), line 277 in "thr_posix.c" [6] ldap_int_thread_pool_wrapper(xpool = 0x3f3ad8), line 654 in "tpool.c" t@6 (l@6) stopped in __lwp_park at 0x7f9554b0 0x7f9554b0: __lwp_park+0x0010: ta %icc,0x00000008 current thread: t@6 [1] __lwp_park(0x0, 0x0, 0x0, 0x1, 0x7f968000, 0xf0000000), at 0x7f9554b0 [2] cond_wait_queue(0x438200, 0x7f968c08, 0x0, 0x0, 0x7f870a00, 0x7f968000), at 0x7f9526b8 [3] _cond_wait_cancel(0x438200, 0x4381e8, 0x2d004, 0x7f983bf0, 0x5, 0x4), at 0x7f952e74 [4] _pthread_cond_wait(0x438200, 0x4381e8, 0x1, 0x7fb1881c, 0x9, 0x7fa75088), at 0x7f952eb0 =>[5] ldap_pvt_thread_cond_wait(cond = 0x438200, mutex = 0x4381e8), line 277 in "thr_posix.c" [6] send_ldap_ber(conn = 0x438108, ber = 0x7cfff6b8), line 217 in "result.c" [7] slap_send_search_entry(op = 0x7cfffaf4, rs = 0x7cfff89c), line 1246 in "result.c" [8] syncprov_sendresp(op = 0x7cfffaf4, opc = 0x7cfff948, so = 0x426d88, e = 0x7cfff974, mode = 1), line 817 in "syncprov.c" [9] syncprov_qplay(op = 0x7cfffaf4, rtask = 0x50a7b0), line 888 in "syncprov.c" [10] syncprov_qtask(ctx = 0x7cfffe0c, arg = 0x50a7b0), line 951 in "syncprov.c" [11] ldap_int_thread_pool_wrapper(xpool = 0x3f3ad8), line 663 in "tpool.c" t@7 (l@7) stopped in __lwp_park at 0x7f9554b0 0x7f9554b0: __lwp_park+0x0010: ta %icc,0x00000008 current thread: t@7 [1] __lwp_park(0x0, 0x0, 0x0, 0x1, 0x7f968000, 0x0), at 0x7f9554b0 [2] cond_wait_queue(0x3f3b08, 0x7f968c08, 0x0, 0x0, 0x7f870c00, 0x7f968000), at 0x7f9526b8 [3] _cond_wait_cancel(0x3f3b08, 0x3f3ae0, 0x0, 0x0, 0x0, 0x0), at 0x7f952e74 [4] _pthread_cond_wait(0x3f3b08, 0x3f3ae0, 0x0, 0x0, 0x0, 0x0), at 0x7f952eb0 =>[5] ldap_pvt_thread_cond_wait(cond = 0x3f3b08, mutex = 0x3f3ae0), line 277 in "thr_posix.c" [6] handle_pause(tpool = 0x3a6d3c, do_pause = 1), line 738 in "tpool.c" [7] ldap_pvt_thread_pool_pause(tpool = 0x3a6d3c), line 761 in "tpool.c" [8] config_back_add(op = 0x7c7ff8d8, rs = 0x7c7ff548), line 4711 in "bconfig.c" [9] overlay_op_walk(op = 0x7c7ff8d8, rs = 0x7c7ff548, which = op_add, oi = 0x452518, on = (nil)), line 670 in "backover.c" [10] over_op_func(op = 0x7c7ff8d8, rs = 0x7c7ff548, which = op_add), line 722 in "backover.c" [11] over_op_add(op = 0x7c7ff8d8, rs = 0x7c7ff548), line 768 in "backover.c" [12] syncrepl_entry(si = 0x452b58, op = 0x7c7ff8d8, entry = 0x523a9c, modlist = 0x7c7ff6c4, syncstate = 1, syncUUID = 0x7c7ff720, syncCSN = 0x453438), line 2166 in "syncrepl.c" [13] do_syncrep2(op = 0x7c7ff8d8, si = 0x452b58), line 892 in "syncrepl.c" [14] do_syncrepl(ctx = 0x7c7ffe0c, arg = 0x457ad0), line 1333 in "syncrepl.c" [15] connection_read_thread(ctx = 0x7c7ffe0c, argv = 0x8), line 1228 in "connection.c" [16] ldap_int_thread_pool_wrapper(xpool = 0x3f3ad8), line 663 in "tpool.c"