Mirror of the gdb-patches mailing list
 help / color / mirror / Atom feed
From: Jan Kratochvil <jan.kratochvil@redhat.com>
To: Mark Kettenis <mark.kettenis@xs4all.nl>
Cc: gdb-patches@sourceware.org
Subject: Re: [patch] Fix Linux attach to signalled/stopped processes
Date: Tue, 26 Jun 2007 22:40:00 -0000	[thread overview]
Message-ID: <20070626210341.GA1406@host0.dyn.jankratochvil.net> (raw)
In-Reply-To: <200706151801.l5FI1xH6022843@brahms.sibelius.xs4all.nl>

[-- Attachment #1: Type: text/plain, Size: 1251 bytes --]

On Fri, 15 Jun 2007 20:01:59 +0200, Mark Kettenis wrote:
> > Date: Wed, 6 Jun 2007 16:34:32 +0200
> > From: Jan Kratochvil <jan.kratochvil@redhat.com>
> > 
> > The patch changes the functionality of TO_ATTACH.  The former
> > functionality was too UNIX centric.  I was reading all the TO_ATTACH
> > OS-flavor implementations and I believe the non-"inf-ptrace.c" ones
> > do not need update.  The code was tested only on Linux kernel,
> > though.
> 
> I disagree here.  You're adding all sorts of goo to work around
> Linux-specific problems.  Please put this stuff in linux-nat.c

While that "/proc/PID/status" reading part looks definitely Linux specific
I would guess that the first waitpid(2) after PTRACE_ATTACH may not return
SIGSTOP even on other ptrace(2)-using OSes (BSD?).

Attaching a testcase for a possible non-Linux kernel test which:
Should print approx. 1x '.' per second (on each caught pending SIGALRM signal).
At least Linux 2.6.22-rc5.x86_64 prints '!' which looks as a kernel bug to me.

How does currently GDB behave there while attaching to a SIGSTOPped process?
This waitpid(2)-after-ptrace(2) signals redelivery looks as ptrace(2) specific
and not Linux specific to me.  And so it would belong to inf-ptrace.c.



Regards,
Jan

[-- Attachment #2: ptrace-test.c --]
[-- Type: text/plain, Size: 2926 bytes --]

/* This testcase is part of GDB, the GNU debugger.

   Copyright 2007 Free Software Foundation, Inc.

   This program is free software; you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
   the Free Software Foundation; either version 2 of the License, or
   (at your option) any later version.

   This program is distributed in the hope that it will be useful,
   but WITHOUT ANY WARRANTY; without even the implied warranty of
   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
   GNU General Public License for more details.

   You should have received a copy of the GNU General Public License
   along with this program; if not, write to the Free Software
   Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */

#define _GNU_SOURCE
#include <signal.h>
#include <unistd.h>
#include <string.h>
#include <stdio.h>
#include <stdlib.h>
#include <errno.h>
#include <sys/wait.h>
#include <sys/ptrace.h>
#include <assert.h>
#include <fcntl.h>

static void action(int sig, siginfo_t * info, void *uc)
{
  raise (SIGALRM);
}

static void loop (void)
{
  struct sigaction act;

  memset (&act, 0, sizeof(struct sigaction));
  act.sa_sigaction = action;
  act.sa_flags = SA_RESTART;
  sigaction (SIGALRM, &act, 0);

  raise (SIGALRM);

  putchar ('!');

  for (;;)
    pause ();
  /* NOTREACHED */
  abort ();
}

static pid_t child;

static void
cleanup (void)
{
  kill (child, SIGKILL);
}

static void
handler (int signo)
{
  cleanup ();
}

int main (void)
{
  void (*handler_orig) (int signo);

  setbuf (stdout, NULL);

  child = fork ();
  switch (child)
    {
      case -1:
	abort ();
      case 0:
	loop ();
	/* NOTREACHED */
	abort ();
      default:
        break;
    }

  atexit (cleanup);
  handler_orig = signal (SIGABRT, handler);
  assert (handler_orig == SIG_DFL);

  for (;;)
    {
      errno = 0;
      ptrace (PTRACE_ATTACH, child, NULL, NULL);
      assert_perror (errno);
      unsigned long sig;

      /* Deliver one SIGSTOP just for sure.
	 If the process was already stopped AND some other process (like shell)
	 has already waited for it we would get stuck in waitpid ().  */
      sig = SIGSTOP;
      do
	{
	  pid_t got_pid;
	  int status;

	  errno = 0;
	  ptrace (PT_CONTINUE, child, (void *) 1UL, (void *) sig);
	  /* For unstopped processes the preventive signal may ESRCH.  */
	  if (sig != SIGSTOP)
	    {
	      assert_perror (errno);
	      putchar ('.');
	    }

	  got_pid = waitpid (child, &status, 0);
	  assert (got_pid == child);

	  /* Check if the thread has exited.  */
	  assert (!WIFEXITED (status));
	  assert (!WIFSIGNALED (status));

	  assert (WIFSTOPPED (status));
	  sig = WSTOPSIG (status);
	  assert (sig != 0);
	}
      while (sig != SIGSTOP);

      errno = 0;
      ptrace (PTRACE_DETACH, child, (void *) 1UL, (void *) 0UL);
      assert_perror (errno);
    }

  return 0;
}

  reply	other threads:[~2007-06-26 21:04 UTC|newest]

Thread overview: 52+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2007-06-06 14:34 Jan Kratochvil
2007-06-11 13:44 ` Jan Kratochvil
2007-06-15 18:02 ` Mark Kettenis
2007-06-26 22:40   ` Jan Kratochvil [this message]
2007-06-27  0:13     ` Mark Kettenis
2007-06-27 11:59       ` Jan Kratochvil
2007-06-27 18:30         ` Mark Kettenis
2007-06-30 11:45         ` Jan Kratochvil
2007-06-30 11:57           ` Eli Zaretskii
2007-06-30 17:15             ` Jan Kratochvil
2007-06-30 18:52               ` Eli Zaretskii
     [not found]               ` <200706301852.l5UIq8ek010536@brahms.sibelius.xs4all.nl>
2007-07-01  3:17                 ` Eli Zaretskii
2007-07-01  9:34                   ` Mark Kettenis
2007-07-01 10:03                     ` Jan Kratochvil
2008-03-31 22:07 Doug Evans
2008-04-02  0:01 ` Jan Kratochvil
2008-04-02  0:07   ` Roland McGrath
2008-04-10 15:30     ` Daniel Jacobowitz
2008-04-10 15:37       ` Jan Kratochvil
2008-04-10 15:49         ` Daniel Jacobowitz
2008-04-10 16:00           ` Jan Kratochvil
2008-04-10 19:59             ` Daniel Jacobowitz
2008-04-10 15:39   ` Daniel Jacobowitz
2008-04-10 16:00     ` Jan Kratochvil
2008-04-10 19:48       ` Daniel Jacobowitz
2008-04-11  8:46       ` Roland McGrath
2008-04-11 17:46         ` Jan Kratochvil
2008-04-11 19:01           ` Daniel Jacobowitz
2008-04-12  7:58           ` Roland McGrath
2008-04-14 15:09             ` Daniel Jacobowitz
2008-04-14 15:31               ` Daniel Jacobowitz
2008-04-15 22:14                 ` Jan Kratochvil
2008-05-01 18:50                   ` Daniel Jacobowitz
2008-07-05  8:48                     ` Jan Kratochvil
2008-07-05 13:48                       ` Daniel Jacobowitz
2008-09-24 12:46                 ` Andreas Schwab
2008-09-26  3:52                   ` Jan Kratochvil
2008-09-26 13:00                     ` Daniel Jacobowitz
2008-09-28 11:43                       ` Jan Kratochvil
2008-09-28 14:58                         ` Daniel Jacobowitz
2008-04-15  8:14               ` Roland McGrath
2008-04-15 13:02                 ` Daniel Jacobowitz
2008-04-16  7:01                   ` Roland McGrath
2008-04-11 22:20         ` Daniel Jacobowitz
2008-04-11 22:21           ` Pedro Alves
2008-04-11 22:25             ` Daniel Jacobowitz
2008-04-12  0:02               ` Pedro Alves
2008-04-12  0:19                 ` Pedro Alves
2008-04-13  9:35                   ` Pedro Alves
2008-04-13 13:40                     ` Pedro Alves
2008-04-12 16:38           ` Roland McGrath
2008-04-12 16:43           ` Eli Zaretskii

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=20070626210341.GA1406@host0.dyn.jankratochvil.net \
    --to=jan.kratochvil@redhat.com \
    --cc=gdb-patches@sourceware.org \
    --cc=mark.kettenis@xs4all.nl \
    /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