Date: Mon, 14 Jun 1999 07:15:14 +0200 From: dirk.meyer@dinoex.sub.org (Dirk Meyer) To: freebsd-bugs@freebsd.org Subject: Re: gnu/12175: gdb crashes with pids > 32736 Message-ID: <GvZ4qF/fhu@dmeyer.dinoex.sub.org> References: <199906121941.VAA83110@home.dinoex.sub.org><37641D27.40AA635D@bigfoot.com>
index | next in thread | previous in thread | raw e-mail
> > >Synopsis: gdb crashes with pids > 32736
>
> I've tried to reproduce this my i386 -current machine, but have been unable to.
> I've successfully run gdb on crashed programs with pids over 70000. However,
> after looking through the gdb code, I've noticed that the pid is declared as an
> int; shouldn't it be declared as a pid_t?
FreeBSD_3.2 RELASE, USA_RESIDENT_NO, no des/crypto/kerberos.
make world, ports/bash2, make install
I would suggest these, I found a similar problem in the bash.
The corefile generated by bash2 can't be read without aborting gdb.
If you ike to run any tests let me know.
I have the corefiles (bash.core and gdb.core) here.
And no i can't call gdb.core with gdb either.
A small test-programm with SIG11 fatal get debugged,
I have to check when my pids are high again.
kind regards Dirk
-- Dirk Meyer, Im Grund 4, 34317 Habichtswald, Germany
-- Tel. +49-5606-6512
It has the internal identification `ports/12174'.
>Submitter-Id: current-users
>Originator: Dirk Meyer
>Organization: privat
>Confidential: no
>Synopsis: bash terminate with corefile if pid > 32000
>Severity: serious
>Priority: medium
>Category: ports
>Release: FreeBSD 3.2-RELEASE i386
>Class: sw-bug
>Environment:
FreeBSD 3.2-RELEASE i386
running bash
>Description:
in comand completition bash terminates:
/kernel: pid 2111 (bash), uid 1000: exited on signal 11 (core dumped)
/kernel: pid 81266 (bash), uid 1000: exited on signal 11 (core dumped)
/kernel: pid 81276 (bash), uid 1000: exited on signal 11 (core dumped)
/kernel: pid 81301 (bash), uid 1000: exited on signal 11 (core dumped)
/kernel: pid 404 (bash), uid 0: exited on signal 11 (core dumped)
/kernel: pid 81322 (bash), uid 0: exited on signal 11 (core dumped)
/kernel: pid 1272 (bash), uid 1000: exited on signal 11
/kernel: pid 81494 (bash), uid 0: exited on signal 11 (core dumped)
>How-To-Repeat:
the new pid of the system has reached a value > 32736.
The shell itsel could still have an old pid in range.
using <TAB> key to complete a path on the command line,
pressing <TAB> twice to get a list that would create a
question "Display all xxx possibilities? (y or n)"
>Fix:
this patch seem to fix the problem
--- execute_cmd.c.orig Tue Jan 26 22:23:49 1999
+++ execute_cmd.c Sat Jun 12 20:58:59 1999
@@ -2207,7 +2207,8 @@
static int
execute_null_command (redirects, pipe_in, pipe_out, async, old_last_command_subst_pid)
REDIRECT *redirects;
- int pipe_in, pipe_out, async, old_last_command_subst_pid;
+ int pipe_in, pipe_out, async;
+ pid_t old_last_command_subst_pid;
{
if (pipe_in != NO_PIPE || pipe_out != NO_PIPE || async)
{
@@ -2955,7 +2956,7 @@
{
char *pathname, *command, **args;
int nofork;
- int pid;
+ pid_t pid;
nofork = (cmdflags & CMD_NO_FORK); /* Don't fork, just exec, if no pipes */
pathname = words->word->word;
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-bugs" in the body of the message
help
Want to link to this message? Use this
URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?GvZ4qF/fhu>