[Bug server/16679] New: Regression in UST static tracepoint

["qiyao at gcc dot gnu.org" <sourceware-bugzilla at sourceware dot org>]

https://sourceware.org/bugzilla/show_bug.cgi?id=16679

            Bug ID: 16679
           Summary: Regression in UST static tracepoint
           Product: gdb
           Version: HEAD
            Status: NEW
          Severity: normal
          Priority: P2
         Component: server
          Assignee: unassigned at sourceware dot org
          Reporter: qiyao at gcc dot gnu.org

I recently see some FAILs in gdb.trace/strace.exp.

Start gdb.trace/strace in GDB,

(gdb) b main
Breakpoint 1 at 0x8048a53: file
../../../../git/gdb/testsuite/gdb.trace/strace.c, line 29.
(gdb) c
Continuing.

Breakpoint 1, main () at ../../../../git/gdb/testsuite/gdb.trace/strace.c:29
29        int a = 0;
(gdb) set debug remote 1
(gdb) info static-tracepoint-markers 
Cnt     ID                                       Enb Address    What
Sending packet: $qTfSTM#1f...Packet received:
mb7fb19eb:6d657461646174612f636f72655f6d61726b65725f666f726d6174:6368616e6e656c202573206e616d6520257320666f726d6174202573
Sending packet: $qTsSTM#2c...Timed out.
Timed out.
Timed out.
Ignoring packet error, continuing...
1       metadata/core_marker_format              n   0xb7fb19eb in
marker_set_format at ../../../Software/ust-0.11/libust/marker.c:516
         Data: "channel %s name %s format %s"

In the other window, start GDBserver,
$ ./gdbserver/gdbserver :1234 ./testsuite/gdb.trace/strace 
Process ./testsuite/gdb.trace/strace created; pid = 8437
Listening on port 1234
Remote debugging from host 127.0.0.1
agent: resumed helper thread
agent: signalling helper thread
agent: waiting for helper thread's response
agent: helper thread's response received
agent: stop helper thread
agent: resumed helper thread
agent: signalling helper thread
agent: waiting for helper thread's response

Attach to GDBserver, and the backtrace is:

(gdb) bt
#0  0xb772a424 in __kernel_vsyscall ()
#1  0x4cf56803 in __read_nocancel () from /lib/libc.so.6
#2  0x0804a9e6 in agent_run_command (pid=8437, cmd=0x9ac11c8 "qTsSTM", len=7)
at ../../../../git/gdb/gdbserver/../common/agent.c:267
#3  0x0805bde5 in run_inferior_command (cmd=0x9ac11c8 "qTsSTM", len=7) at
../../../../git/gdb/gdbserver/tracepoint.c:6790
#4  0x08060f49 in cmd_qtstmat (packet=0x9ac11c8 "qTsSTM") at
../../../../git/gdb/gdbserver/tracepoint.c:3943
#5  handle_tracepoint_query (packet=<optimized out>) at
../../../../git/gdb/gdbserver/tracepoint.c:4290
#6  0x0805425e in handle_query (own_buf=<optimized out>, packet_len=6,
new_packet_len_p=0xbff1230c) at ../../../../git/gdb/gdbserver/server.c:2142
#7  0x08055a0f in process_serial_event () at
../../../../git/gdb/gdbserver/server.c:3445
#8  handle_serial_event (err=0, client_data=0x0) at
../../../../git/gdb/gdbserver/server.c:3889
#9  0x0805a588 in handle_file_event (event_file_desc=4) at
../../../../git/gdb/gdbserver/event-loop.c:434
#10 0x0805adb2 in process_event () at
../../../../git/gdb/gdbserver/event-loop.c:189
#11 start_event_loop () at ../../../../git/gdb/gdbserver/event-loop.c:552
#12 0x08049fc4 in main (argc=3, argv=0xbff12494) at
../../../../git/gdb/gdbserver/server.c:3289


'git bisect' shows this regression is caused by the this patch:

  [FYI 3/3] Teach GDBserver's Linux backend about no unwaited-for children
(TARGET_WAITKIND_NO_RESUMED).
  https://sourceware.org/ml/gdb-patches/2014-02/msg00828.html

This may affect the thread control (stop and resume) for agent.  Further
analysis is needed.

-- 
You are receiving this mail because:
You are on the CC list for the bug.