Use PTRACE_SINGLESTEP_ONE when single-stepping one thread.

Currently, when we receive a request to single-step one single thread
(Eg, when single-stepping out of a breakpoint), we use the
PTRACE_SINGLESTEP pthread request, which does single-step
the corresponding thread, but also resumes execution of all
other threads in the inferior.

This causes problems when debugging programs where another thread
receives multiple debug events while trying to single-step a specific
thread out of a breakpoint (with infrun traces turned on):

    (gdb) continue
    Continuing.
    infrun: clear_proceed_status_thread (Thread 126)
    [...]
    infrun: clear_proceed_status_thread (Thread 142)
    [...]
    infrun: clear_proceed_status_thread (Thread 146)
    infrun: clear_proceed_status_thread (Thread 125)
    infrun: proceed (addr=0xffffffff, signal=GDB_SIGNAL_DEFAULT, step=0)
    infrun: resume (step=1, signal=GDB_SIGNAL_0), trap_expected=1, current thread [Thread 142] at 0x10684838
    infrun: wait_for_inferior ()
    infrun: target_wait (-1, status) =
    infrun:   42000 [Thread 146],
    infrun:   status->kind = stopped, signal = GDB_SIGNAL_REALTIME_34
    infrun: infwait_normal_state
    infrun: TARGET_WAITKIND_STOPPED
    infrun: stop_pc = 0x10a187f4
    infrun: context switch
    infrun: Switching context from Thread 142 to Thread 146
    infrun: random signal (GDB_SIGNAL_REALTIME_34)
    infrun: switching back to stepped thread
    infrun: Switching context from Thread 146 to Thread 142
    infrun: resume (step=1, signal=GDB_SIGNAL_0), trap_expected=1, current thread [Thread 142] at 0x10684838
    infrun: prepare_to_wait
    [...handling of similar events for threads 145, 144 and 143 snipped...]
    infrun: prepare_to_wait
    infrun: target_wait (-1, status) =
    infrun:   42000 [Thread 146],
    infrun:   status->kind = stopped, signal = GDB_SIGNAL_REALTIME_34
    infrun: infwait_normal_state
    infrun: TARGET_WAITKIND_STOPPED
    infrun: stop_pc = 0x10a187f4
    infrun: context switch
    infrun: Switching context from Thread 142 to Thread 146
    ../../src/gdb/inline-frame.c:339: internal-error: skip_inline_frames: Assertion `find_inline_frame_state (ptid) == NULL' failed.

What happens is that GDB keeps sending requests to resume one specific
thread, and keeps receiving debugging events for other threads.
Things break down when the one of the other threads receives a debug
event for the second time (thread 146 in the example above).

This patch fixes the problem by making sure that only one thread
gets resumed, thus preventing the other threads from generating
an unexpected event.

gdb/gdbserver/ChangeLog:

        * lynx-low.c (lynx_resume): Use PTRACE_SINGLESTEP_ONE if N == 1.
        Remove FIXME comment about assumption about N.

diff --git a/gdb/gdbserver/ChangeLog b/gdb/gdbserver/ChangeLog
index 6a5005e0a0157eb4ad24f6739cbb379cc8eafbed..92b213a3f1f8075a83bac5360a83ec9a7311fda3 100644 (file)
--- a/gdb/gdbserver/ChangeLog
+++ b/gdb/gdbserver/ChangeLog
@@ -1,3 +1,8 @@
+2014-12-15  Joel Brobecker  <brobecker@adacore.com>
+
+       * lynx-low.c (lynx_resume): Use PTRACE_SINGLESTEP_ONE if N == 1.
+       Remove FIXME comment about assumption about N.
+
 2014-12-13  Joel Brobecker  <brobecker@adacore.com>
 
        * configure.ac: If large-file support is disabled in GDBserver,

diff --git a/gdb/gdbserver/lynx-low.c b/gdb/gdbserver/lynx-low.c
index 6178e038a1bd81939f484f289895b2a75cd56152..3b8366953ecdc4be75e237721f2bc3b6265163a7 100644 (file)
--- a/gdb/gdbserver/lynx-low.c
+++ b/gdb/gdbserver/lynx-low.c
@@ -320,10 +320,11 @@ lynx_attach (unsigned long pid)
 static void
 lynx_resume (struct thread_resume *resume_info, size_t n)
 {
-  /* FIXME: Assume for now that n == 1.  */
   ptid_t ptid = resume_info[0].thread;
-  const int request = (resume_info[0].kind == resume_step
-                       ? PTRACE_SINGLESTEP : PTRACE_CONT);
+  const int request
+    = (resume_info[0].kind == resume_step
+       ? (n == 1 ? PTRACE_SINGLESTEP_ONE : PTRACE_SINGLESTEP)
+       : PTRACE_CONT);
   const int signal = resume_info[0].sig;
 
   /* If given a minus_one_ptid, then try using the current_process'