[commit] Really remove through-sigtramp breakpoints

classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view
|

[commit] Really remove through-sigtramp breakpoints

Daniel Jacobowitz-2
I removed an unused breakpoint type a few weeks ago, but failed to
remove the associated BPSTAT_WHAT constant.  This finishes the job.
It's important because BPSTAT_WHAT_LAST and the other constants are
used to size the table in stop_bpstat - without this patch I had a
number of problems on m68k-uclinux, where the shared library
breakpoint and call dummy breakpoint are both set at _start.

Tested x86_64-linux and committed.

--
Daniel Jacobowitz
CodeSourcery

2007-05-23  Daniel Jacobowitz  <[hidden email]>

        * breakpoint.h (enum bpstat_what_main_action): Remove
        BPSTAT_WHAT_THROUGH_SIGTRAMP.
        * infrun.c (process_event_stop_test): Do not check for it.

Index: breakpoint.h
===================================================================
RCS file: /cvs/src/src/gdb/breakpoint.h,v
retrieving revision 1.41
diff -u -p -r1.41 breakpoint.h
--- breakpoint.h 13 Apr 2007 13:50:32 -0000 1.41
+++ breakpoint.h 22 May 2007 21:10:30 -0000
@@ -486,10 +486,6 @@ enum bpstat_what_main_action
     /* Clear step resume breakpoint, and keep checking.  */
     BPSTAT_WHAT_STEP_RESUME,
 
-    /* Clear through_sigtramp breakpoint, muck with trap_expected, and keep
-       checking.  */
-    BPSTAT_WHAT_THROUGH_SIGTRAMP,
-
     /* Check the dynamic linker's data structures for new libraries, then
        keep checking.  */
     BPSTAT_WHAT_CHECK_SHLIBS,
Index: infrun.c
===================================================================
RCS file: /cvs/src/src/gdb/infrun.c,v
retrieving revision 1.229
diff -u -p -r1.229 infrun.c
--- infrun.c 7 May 2007 18:20:18 -0000 1.229
+++ infrun.c 22 May 2007 21:10:31 -0000
@@ -2178,15 +2178,6 @@ process_event_stop_test:
   }
  break;
 
-      case BPSTAT_WHAT_THROUGH_SIGTRAMP:
-        if (debug_infrun)
-  fprintf_unfiltered (gdb_stdlog, "infrun: BPSTAT_WHAT_THROUGH_SIGTRAMP\n");
- /* If were waiting for a trap, hitting the step_resume_break
-   doesn't count as getting it.  */
- if (trap_expected)
-  ecs->another_trap = 1;
- break;
-
       case BPSTAT_WHAT_CHECK_SHLIBS:
       case BPSTAT_WHAT_CHECK_SHLIBS_RESUME_FROM_HOOK:
  {