Mirror of the gdb-patches mailing list
 help / color / mirror / Atom feed
From: Doug Evans <dje@google.com>
To: Pedro Alves <pedro@codesourcery.com>
Cc: gdb-patches@sourceware.org
Subject: Re: [RFA] Include thread ID in target_wait debugging output.
Date: Fri, 27 Feb 2009 22:11:00 -0000	[thread overview]
Message-ID: <e394668d0902271348s32173593ye18d82ea0449fc7d@mail.gmail.com> (raw)
In-Reply-To: <e394668d0902271156g4c28676cr60fbdd754ed9c5e0@mail.gmail.com>

[-- Attachment #1: Type: text/plain, Size: 4765 bytes --]

On Fri, Feb 27, 2009 at 11:56 AM, Doug Evans <dje@google.com> wrote:
> On Fri, Feb 27, 2009 at 11:41 AM, Pedro Alves <pedro@codesourcery.com> wrote:
>> Hi Doug,
>>
>> On Thursday 26 February 2009 18:41:07, Doug Evans wrote:
>>> Hi.
>>>
>>> Printing the result of target_wait when "set debug infrun 1" is great,
>>> but I've found I've needed the thread ID.
>>>
>>> The output line was getting a bit long, so I split it up.
>>> There are other ways to split it up, but this works for me.
>>>
>>
>> The downside of that, is that "set debug timespamp" will print
>> a timestamp for each piece of the log, and any signal debug
>> output in async mode will mingle up.  Maybe do the printing to
>> a mem_fileopen buffer, and then print that to stderr in one go?
>
> Ok, I'll print to mem_fileopen first, and then print in one go.
>
>> An issue I though would be desirable to avoid here, is calling
>> target_pid_to_str, which isn't garanteed to not have side
>> effects.
>
> That's unfortunate.  Is that documented somewhere?
>
>> Introducing side effects when you enable debug
>> output is undesirable.
>
> Yep.
>
>> Maybe print the ptid in raw form, say,
>> like: (pid, lwp, tid)?  OTOH, infrun debugging output has been
>> doing that for years, so it's probably fine...
>
> Ya, there's LOTS of calls to target_pid_to_str for debugging output.
>
>>
>>> Ok to check in?
>>>
>>> 2009-02-26  Doug Evans  <dje@google.com>
>>>
>>>       Include thread ID in target_wait debugging output.
>>>       * infrun.c (print_target_wait_results): New function.
>>>       (wait_for_inferior,fetch_inferior_event): Call it.
>>>
>>> Index: infrun.c
>>> ===================================================================
>>> RCS file: /cvs/src/src/gdb/infrun.c,v
>>> retrieving revision 1.360
>>> diff -u -p -r1.360 infrun.c
>>> --- infrun.c  25 Feb 2009 02:14:22 -0000      1.360
>>> +++ infrun.c  26 Feb 2009 18:34:57 -0000
>>> @@ -1737,6 +1737,30 @@ delete_step_thread_step_resume_breakpoin
>>>    delete_step_thread_step_resume_breakpoint ();
>>>  }
>>>
>>> +/* Pretty print the results of target_wait, for debugging purposes.  */
>>> +
>>> +static void
>>> +print_target_wait_results (ptid_t waiton_ptid, ptid_t result_ptid,
>>> +                        const struct target_waitstatus *ws)
>>> +{
>>> +  char *status_string = target_waitstatus_to_string (ws);
>>> +
>>> +  fprintf_unfiltered (gdb_stdlog,
>>> +                   "infrun: target_wait (%d", PIDGET (waiton_ptid));
>>> +  if (PIDGET (waiton_ptid) != -1)
>>> +    fprintf_unfiltered (gdb_stdlog,
>>> +                     " [%s]", target_pid_to_str (waiton_ptid));
>>> +  fprintf_unfiltered (gdb_stdlog, ", status) =\n");
>>> +  fprintf_unfiltered (gdb_stdlog,
>>> +                   "infrun:   %d [%s],\n",
>>> +                   PIDGET (result_ptid), target_pid_to_str (result_ptid));
>>> +  fprintf_unfiltered (gdb_stdlog,
>>> +                   "infrun:   %s\n",
>>> +                   status_string);
>>> +
>>> +  xfree (status_string);
>>> +}
>>> +
>>>  /* Wait for control to return from inferior to debugger.
>>>
>>>     If TREAT_EXEC_AS_SIGTRAP is non-zero, then handle EXEC signals
>>> @@ -1790,14 +1814,7 @@ wait_for_inferior (int treat_exec_as_sig
>>>       ecs->ptid = target_wait (waiton_ptid, &ecs->ws);
>>>
>>>        if (debug_infrun)
>>> -     {
>>> -       char *status_string = target_waitstatus_to_string (&ecs->ws);
>>> -       fprintf_unfiltered (gdb_stdlog,
>>> -                           "infrun: target_wait (%d, status) = %d, %s\n",
>>> -                           PIDGET (waiton_ptid), PIDGET (ecs->ptid),
>>> -                           status_string);
>>> -       xfree (status_string);
>>> -     }
>>> +     print_target_wait_results (waiton_ptid, ecs->ptid, &ecs->ws);
>>>
>>>        if (treat_exec_as_sigtrap && ecs->ws.kind == TARGET_WAITKIND_EXECD)
>>>          {
>>> @@ -1875,14 +1892,7 @@ fetch_inferior_event (void *client_data)
>>>      ecs->ptid = target_wait (waiton_ptid, &ecs->ws);
>>>
>>>    if (debug_infrun)
>>> -    {
>>> -      char *status_string = target_waitstatus_to_string (&ecs->ws);
>>> -      fprintf_unfiltered (gdb_stdlog,
>>> -                       "infrun: target_wait (%d, status) = %d, %s\n",
>>> -                       PIDGET (waiton_ptid), PIDGET (ecs->ptid),
>>> -                       status_string);
>>> -      xfree (status_string);
>>> -    }
>>> +    print_target_wait_results (waiton_ptid, ecs->ptid, &ecs->ws);
>>>
>>>    if (non_stop
>>>        && ecs->ws.kind != TARGET_WAITKIND_IGNORE
>>>
>>
>>

How about this?

2009-02-27  Doug Evans  <dje@google.com>

        Include thread ID in target_wait debugging output.
        * infrun.c (print_target_wait_results): New function.
        (wait_for_inferior,fetch_inferior_event): Call it.

[-- Attachment #2: gdb-090227-debug-target-wait-2.patch.txt --]
[-- Type: text/plain, Size: 3228 bytes --]

2009-02-27  Doug Evans  <dje@google.com>

	Include thread ID in target_wait debugging output.
	* infrun.c (print_target_wait_results): New function.
	(wait_for_inferior,fetch_inferior_event): Call it.

Index: infrun.c
===================================================================
RCS file: /cvs/src/src/gdb/infrun.c,v
retrieving revision 1.360
diff -u -p -u -p -r1.360 infrun.c
--- infrun.c	25 Feb 2009 02:14:22 -0000	1.360
+++ infrun.c	27 Feb 2009 21:42:48 -0000
@@ -1737,6 +1737,46 @@ delete_step_thread_step_resume_breakpoin
   delete_step_thread_step_resume_breakpoint ();
 }
 
+/* Pretty print the results of target_wait, for debugging purposes.  */
+
+static void
+print_target_wait_results (ptid_t waiton_ptid, ptid_t result_ptid,
+			   const struct target_waitstatus *ws)
+{
+  char *status_string = target_waitstatus_to_string (ws);
+  struct ui_file *tmp_stream = mem_fileopen ();
+  char *text;
+  long len;
+
+  /* The text is split over several lines because it was getting too long.
+     Call fprintf_unfiltered (gdb_stdlog) once so that the text is still
+     output as a unit; we want only one timestamp printed if debug_timestamp
+     is set.  */
+
+  fprintf_unfiltered (tmp_stream,
+		      "infrun: target_wait (%d", PIDGET (waiton_ptid));
+  if (PIDGET (waiton_ptid) != -1)
+    fprintf_unfiltered (tmp_stream,
+			" [%s]", target_pid_to_str (waiton_ptid));
+  fprintf_unfiltered (tmp_stream, ", status) =\n");
+  fprintf_unfiltered (tmp_stream,
+		      "infrun:   %d [%s],\n",
+		      PIDGET (result_ptid), target_pid_to_str (result_ptid));
+  fprintf_unfiltered (tmp_stream,
+		      "infrun:   %s\n",
+		      status_string);
+
+  text = ui_file_xstrdup (tmp_stream, &len);
+
+  /* This uses %s in part to handle %'s in the text, but also to avoid
+     a gcc error: the format attribute requires a string literal.  */
+  fprintf_unfiltered (gdb_stdlog, "%s", text);
+
+  xfree (status_string);
+  xfree (text);
+  ui_file_delete (tmp_stream);
+}
+
 /* Wait for control to return from inferior to debugger.
 
    If TREAT_EXEC_AS_SIGTRAP is non-zero, then handle EXEC signals
@@ -1790,14 +1830,7 @@ wait_for_inferior (int treat_exec_as_sig
 	ecs->ptid = target_wait (waiton_ptid, &ecs->ws);
 
       if (debug_infrun)
-	{
-	  char *status_string = target_waitstatus_to_string (&ecs->ws);
-	  fprintf_unfiltered (gdb_stdlog,
-			      "infrun: target_wait (%d, status) = %d, %s\n",
-			      PIDGET (waiton_ptid), PIDGET (ecs->ptid),
-			      status_string);
-	  xfree (status_string);
-	}
+	print_target_wait_results (waiton_ptid, ecs->ptid, &ecs->ws);
 
       if (treat_exec_as_sigtrap && ecs->ws.kind == TARGET_WAITKIND_EXECD)
         {
@@ -1875,14 +1908,7 @@ fetch_inferior_event (void *client_data)
     ecs->ptid = target_wait (waiton_ptid, &ecs->ws);
 
   if (debug_infrun)
-    {
-      char *status_string = target_waitstatus_to_string (&ecs->ws);
-      fprintf_unfiltered (gdb_stdlog,
-			  "infrun: target_wait (%d, status) = %d, %s\n",
-			  PIDGET (waiton_ptid), PIDGET (ecs->ptid),
-			  status_string);
-      xfree (status_string);
-    }
+    print_target_wait_results (waiton_ptid, ecs->ptid, &ecs->ws);
 
   if (non_stop
       && ecs->ws.kind != TARGET_WAITKIND_IGNORE

  reply	other threads:[~2009-02-27 21:49 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2009-02-26 19:22 Doug Evans
2009-02-27 19:56 ` Pedro Alves
2009-02-27 21:13   ` Doug Evans
2009-02-27 22:11     ` Doug Evans [this message]
2009-02-27 22:14       ` Pedro Alves

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=e394668d0902271348s32173593ye18d82ea0449fc7d@mail.gmail.com \
    --to=dje@google.com \
    --cc=gdb-patches@sourceware.org \
    --cc=pedro@codesourcery.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox