gdb/

	* gdbthread.h (struct thread_info): Comment on field
	`step_after_step_resume_breakpoint'.

diff --git a/gdb/ChangeLog b/gdb/ChangeLog
index 77c6c56f2a6ec834f374fb7deec7af762f8e8d88..584959d3c00988ed5b99f239cd9c2a6510313692 100644 (file)
--- a/gdb/ChangeLog
+++ b/gdb/ChangeLog
@@ -1,3 +1,8 @@
+2011-09-07  Yao Qi  <yao@codesourcery.com>
+
+       * gdbthread.h (struct thread_info): Comment on field
+       `step_after_step_resume_breakpoint'.
+
 2011-09-07  Abhijit Halder  <abhijit.k.halder@gmail.com>
 
        * remote.c (remote_console_output): Reindent.

diff --git a/gdb/gdbthread.h b/gdb/gdbthread.h
index 0b19a8dc3c5ea82f6ef49623ce88dceabe2a3af6..6baeb92c3e3f65a57afd770d820530f4d8fa0a7b 100644 (file)
--- a/gdb/gdbthread.h
+++ b/gdb/gdbthread.h
@@ -176,7 +176,14 @@ struct thread_info
   int stepping_over_breakpoint;
 
   /* Set to TRUE if we should finish single-stepping over a breakpoint
-     after hitting the current step-resume breakpoint.  */
+     after hitting the current step-resume breakpoint.  The context here
+     is that GDB is to do `next' or `step' while signal arrives.
+     When stepping over a breakpoint and signal arrives, GDB will attempt
+     to skip signal handler, so it inserts a step_resume_breakpoint at the
+     signal return address, and resume inferior.
+     step_after_step_resume_breakpoint is set to TRUE at this moment in
+     order to keep GDB in mind that there is still a breakpoint to step over
+     when GDB gets back SIGTRAP from step_resume_breakpoint.  */
   int step_after_step_resume_breakpoint;
 
   /* This is set TRUE when a catchpoint of a shared library event