[rfc] linux-nat: Never PTRACE_CONT a stepping thread

[Jan Kratochvil <jan dot kratochvil at redhat dot com>]

Hi,

as referenced in:
	[patch 3/4]#3 linux-nat: Do not respawn signals
	http://sourceware.org/ml/gdb-patches/2010-09/msg00360.html

If multiple signals happen besides SIGTRAP GDB still may switch from thread A
away (as not considering it stepping) to thread B for SIGUSR1 and accidentally
PTRACE_CONT thread A while resignalling SIGUSR1.

It probably could have its own testcase.
I will code one depending on the resolution of the #3 series above.

No regressions on {x86_64,x86_64-m32,i686}-fedora13-linux-gnu.


Thanks,
Jan


gdb/
2010-09-20  Jan Kratochvil  <jan.kratochvil@redhat.com>

	* gdbthread.h (currently_stepping): New declaration.
	* infrun.c (currently_stepping): Remove the forward declaration.
	(currently_stepping): Make it global.
	* linux-nat.c (resume_callback) <lp->stopped && lp->status == 0>: New
	variables tp and step, initialized them.  Pass STEP to to_resume.
	Print also possibly "PTRACE_SINGLESTEP" if STEP.  Initialize LP->STEP.

diff --git a/gdb/gdbthread.h b/gdb/gdbthread.h
index cd24eaf..e89e88d 100644
--- a/gdb/gdbthread.h
+++ b/gdb/gdbthread.h
@@ -352,4 +352,6 @@ extern struct thread_info* inferior_thread (void);
 
 extern void update_thread_list (void);
 
+extern int currently_stepping (struct thread_info *tp);
+
 #endif /* GDBTHREAD_H */
diff --git a/gdb/infrun.c b/gdb/infrun.c
index 0720b31..8afdd77 100644
--- a/gdb/infrun.c
+++ b/gdb/infrun.c
@@ -74,8 +74,6 @@ static int follow_fork (void);
 static void set_schedlock_func (char *args, int from_tty,
 				struct cmd_list_element *c);
 
-static int currently_stepping (struct thread_info *tp);
-
 static int currently_stepping_or_nexting_callback (struct thread_info *tp,
 						   void *data);
 
@@ -4842,7 +4840,7 @@ infrun: not switching back to stepped thread, it has vanished\n");
 
 /* Is thread TP in the middle of single-stepping?  */
 
-static int
+int
 currently_stepping (struct thread_info *tp)
 {
   return ((tp->step_range_end && tp->step_resume_breakpoint == NULL)
diff --git a/gdb/linux-nat.c b/gdb/linux-nat.c
index a0eca0e..69b9a5f 100644
--- a/gdb/linux-nat.c
+++ b/gdb/linux-nat.c
@@ -1820,20 +1820,26 @@ resume_callback (struct lwp_info *lp, void *data)
     }
   else if (lp->stopped && lp->status == 0)
     {
+      struct thread_info *tp = find_thread_ptid (lp->ptid);
+      /* lp->step may already contain a stale value.  */
+      int step = tp ? currently_stepping (tp) : 0;
+
       if (debug_linux_nat)
 	fprintf_unfiltered (gdb_stdlog,
-			    "RC:  PTRACE_CONT %s, 0, 0 (resuming sibling)\n",
+			    "RC:  %s %s, 0, 0 (resuming sibling)\n",
+			    step ? "PTRACE_SINGLESTEP" : "PTRACE_CONT",
 			    target_pid_to_str (lp->ptid));
 
       linux_ops->to_resume (linux_ops,
 			    pid_to_ptid (GET_LWP (lp->ptid)),
-			    0, TARGET_SIGNAL_0);
+			    step, TARGET_SIGNAL_0);
       if (debug_linux_nat)
 	fprintf_unfiltered (gdb_stdlog,
-			    "RC:  PTRACE_CONT %s, 0, 0 (resume sibling)\n",
+			    "RC:  %s %s, 0, 0 (resume sibling)\n",
+			    step ? "PTRACE_SINGLESTEP" : "PTRACE_CONT",
 			    target_pid_to_str (lp->ptid));
       lp->stopped = 0;
-      lp->step = 0;
+      lp->step = step;
       memset (&lp->siginfo, 0, sizeof (lp->siginfo));
       lp->stopped_by_watchpoint = 0;
     }