On 09/11/2012 04:01 PM, Ondrej Oprala wrote:
Hi,
I've written a tiny patch that should fix this issue with su:
https://bugzilla.redhat.com/show_bug.cgi?id=747592.
None of the shells' output propagates back
to the shell where su was called (except TCSH, which reports segfault
to stdout; I've tested this with sh,ksh,zsh,bash).
So, with this patch, su should print a warning if
WIFSIGNALED(status) && WCOREDUMP(status) is true.
Cheers,
Ondrej
DIFF
From d4776f487e283649d65be3f3865141a0171a80e7 Mon Sep 17 00:00:00 2001
From: Ondrej Oprala<ooprala@xxxxxxxxxx>
Date: Tue, 11 Sep 2012 16:39:17 +0200
Subject: [PATCH] su: add segmentation fault reporting of the child process
Child processes that ended with segmentation fault previously
indicated this with return status only. The report is now more
verbose if core dump is allowed.
---
login-utils/su-common.c | 6 +++++-
1 file changed, 5 insertions(+), 1 deletion(-)
diff --git a/login-utils/su-common.c b/login-utils/su-common.c
index 23ad57d..b0dbed2 100644
--- a/login-utils/su-common.c
+++ b/login-utils/su-common.c
@@ -300,7 +300,11 @@ create_watching_parent (void)
}
if (pid != (pid_t)-1)
if (WIFSIGNALED (status))
- status = WTERMSIG (status) + 128;
+ {
+ status = WTERMSIG (status) + 128;
+ if (WCOREDUMP (status))
+ fprintf (stderr, _("Segmentation fault (core dumped)\n"));
+ }
else
status = WEXITSTATUS (status);
else
-- 1.7.11.4
You'll need to key on WTERMSIG (status).
"Segmentation fault" will not always be appropriate.
You'll get a core just with Ctrl-\ (SIGQUIT) too for example.
Something along the lines of:
"%s (core dumped)", strsignal (WTERMSIG (status))
cheers,
Pádraig.
--
To unsubscribe from this list: send the line "unsubscribe util-linux" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
