[SPDK] Re: Print backtrace in SPDK

Thanks Andrey for suggestion. The backtrace(3) seems to be the right way, but is not sufficient.

There is a function rte_dump_stack which calls backtrace and backtrace_symbols. I copied it into nvmf_tcp_close_qpair and modified a little bit as below:
+       {
+               void *func[BACKTRACE_SIZE];
+               char **symb = NULL;
+               int size;
+       
+               size = backtrace(func, BACKTRACE_SIZE);
+               symb = backtrace_symbols(func, size);
+       
+               if (symb == NULL)
+                       return;
+       
+               while (size > 0) {
+                       SPDK_ERRLOG("%d: [%s]\n", size, symb[size - 1]);
+                       size --;
+               }
+       
+               free(symb);
+       }

The output I got is like:
[2020-08-22 21:44:34.404823] tcp.c:2395:nvmf_tcp_close_qpair: *ERROR*: 13: [build/bin/nvmf_tgt(+0x126fe) [0x5627642236fe]]
[2020-08-22 21:44:34.404857] tcp.c:2395:nvmf_tcp_close_qpair: *ERROR*: 12: [/lib/x86_64-linux-gnu/libc.so.6(__libc_start_main+0xf3) [0x7fed1369d1e3]]
[2020-08-22 21:44:34.404868] tcp.c:2395:nvmf_tcp_close_qpair: *ERROR*: 11: [build/bin/nvmf_tgt(+0x128e9) [0x5627642238e9]]
[2020-08-22 21:44:34.404876] tcp.c:2395:nvmf_tcp_close_qpair: *ERROR*: 10: [build/bin/nvmf_tgt(+0xd9f32) [0x5627642eaf32]]
[2020-08-22 21:44:34.404882] tcp.c:2395:nvmf_tcp_close_qpair: *ERROR*: 9: [build/bin/nvmf_tgt(+0xdc096) [0x5627642ed096]]
[2020-08-22 21:44:34.404889] tcp.c:2395:nvmf_tcp_close_qpair: *ERROR*: 8: [build/bin/nvmf_tgt(+0xdbc73) [0x5627642ecc73]]
[2020-08-22 21:44:34.404895] tcp.c:2395:nvmf_tcp_close_qpair: *ERROR*: 7: [build/bin/nvmf_tgt(+0xdbb5d) [0x5627642ecb5d]]
[2020-08-22 21:44:34.404902] tcp.c:2395:nvmf_tcp_close_qpair: *ERROR*: 6: [build/bin/nvmf_tgt(+0xe35da) [0x5627642f45da]]
[2020-08-22 21:44:34.404908] tcp.c:2395:nvmf_tcp_close_qpair: *ERROR*: 5: [build/bin/nvmf_tgt(+0xe3054) [0x5627642f4054]]
[2020-08-22 21:44:34.404915] tcp.c:2395:nvmf_tcp_close_qpair: *ERROR*: 4: [build/bin/nvmf_tgt(+0xe2d94) [0x5627642f3d94]]
[2020-08-22 21:44:34.404922] tcp.c:2395:nvmf_tcp_close_qpair: *ERROR*: 3: [build/bin/nvmf_tgt(+0xca3b7) [0x5627642db3b7]]
[2020-08-22 21:44:34.404928] tcp.c:2395:nvmf_tcp_close_qpair: *ERROR*: 2: [build/bin/nvmf_tgt(+0xd09e8) [0x5627642e19e8]]
[2020-08-22 21:44:34.404935] tcp.c:2395:nvmf_tcp_close_qpair: *ERROR*: 1: [build/bin/nvmf_tgt(+0xd8158) [0x5627642e9158]]

This is not very helpful.

Thanks,
-Wenhua


On 8/22/20, 12:58 AM, "Andrey Kuzmin" <andrey.v.kuzmin(a)gmail.com> wrote:

    On Sat, Aug 22, 2020, 10:14 Wenhua Liu <liuw(a)vmware.com> wrote:

    > Hi,
    >
    > Does anyone know if there is a function in SPDK that prints the backtrace?
    >

    backtrace(3) lets you do that.

    Regards,
    Andrey


    > I run into a “Connection Reset by Peer” issue on host side when testing
    > NVMe/TCP. I identified it’s because some queue pairs are closed
    > unexpectedly by calling nvmf_tcp_close_qpair, but I could not figure out
    > how/why this function is called. I thought if the backtrace can be printed
    > when calling this function, it might be helpful to me to find the root
    > cause.
    >
    > Thanks,
    > -Wenhua
    > _______________________________________________
    > SPDK mailing list -- spdk(a)lists.01.org
    > To unsubscribe send an email to spdk-leave(a)lists.01.org
    >
    _______________________________________________
    SPDK mailing list -- spdk(a)lists.01.org
    To unsubscribe send an email to spdk-leave(a)lists.01.org