Hi list, Recently I encountered a deadloop in 1 master + 1 worker mode of vpp(also happens in 1 master 0 worker mode) when I do some vpp (based on vpp 21.10) configuration through govpp.
The thread stats in gdb is: (gdb) info threads Id Target Id Frame * 1 Thread 0x7f735c3287c0 (LWP 274213) "vpp_main" internal_mallinfo (m=0x7f731c227040) at /home/fortitude/glx/vpp/src/vppinfra/dlmalloc.c:2100 2 Thread 0x7f73127d2700 (LWP 274214) "eal-intr-thread" 0x00007f735c5f449e in epoll_wait (epfd=16, events=0x7f73127d1d30, maxevents=7, timeout=-1) at ../sysdeps/unix/sysv/linux/epoll_wait.c:30 3 Thread 0x7f7311fd1700 (LWP 274215) "vpp_wk_0" 0x00007f735c5d774b in sched_yield () at ../sysdeps/unix/syscall-template.S:78 >From what I observed, it is the inner loop (segment_holds) become a dead loop, and I still have not found the reason (see bottom of the mail.) there is old mail https://lists.fd.io/mt/89947053/675661 in this list report the same issue, but no answer yet and I can't reply to that thread, so I create this one. Hope someone can give some clue on this. static struct dlmallinfo internal_mallinfo(mstate m) { struct dlmallinfo nm = { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 }; ensure_initialization(); if (!PREACTION(m)) { check_malloc_state(m); if (is_initialized(m)) { size_t nfree = SIZE_T_ONE; /* top always free */ size_t mfree = m->topsize + TOP_FOOT_SIZE; size_t sum = mfree; msegmentptr s = &m->seg; while (s != 0) { mchunkptr q = align_as_chunk(s->base); while (segment_holds(s, q) && q != m->top && q->head != FENCEPOST_HEAD) { size_t sz = chunksize(q); sum += sz; if (!is_inuse(q)) { mfree += sz; ++nfree; } q = next_chunk(q); } s = s->next; }
-=-=-=-=-=-=-=-=-=-=-=- Links: You receive all messages sent to this group. View/Reply Online (#21616): https://lists.fd.io/g/vpp-dev/message/21616 Mute This Topic: https://lists.fd.io/mt/92202054/21656 Group Owner: vpp-dev+ow...@lists.fd.io Unsubscribe: https://lists.fd.io/g/vpp-dev/leave/1480452/21656/631435203/xyzzy [arch...@mail-archive.com] -=-=-=-=-=-=-=-=-=-=-=-