GNU bug report logs - #19780
Bootstrap bash binary segfaults on x86_64

Previous Next

Full log

Message #26 received at 19780 <at> debbugs.gnu.org (full text, mbox):

From: ludo <at> gnu.org (Ludovic Courtès)
To: Ricardo Wurmus <ricardo.wurmus <at> mdc-berlin.de>
Cc: 19780 <at> debbugs.gnu.org
Subject: Re: bug#19780: 13 failing tests
Date: Tue, 10 Feb 2015 16:49:56 +0100

[Message part 1 (text/plain, inline)]

Ricardo Wurmus <ricardo.wurmus <at> mdc-berlin.de> skribis:

> Oddly, I cannot seem to get a core dump, even after setting ulimit -c
> unlimited.  Is there a way to run the failing builder manually?

No.

Could you try something like this:

[Message part 2 (text/x-patch, inline)]

diff --git a/test-env.in b/test-env.in
index f66a0db..9cdd479 100644
--- a/test-env.in
+++ b/test-env.in
@@ -90,7 +90,8 @@ then
 	--substitute-urls="$GUIX_BINARY_SUBSTITUTE_URL" &
 
     daemon_pid=$!
-    trap "kill $daemon_pid ; rm -rf $NIX_STATE_DIR" EXIT
+    echo "please attach gdb to $daemon_pid and hit enter"
+    read
 fi
 
 # Avoid issues that could stem from l10n, such as language/encoding

[Message part 3 (text/plain, inline)]

When you see the message, run this from another terminal:

  $ gdb
  (gdb) attach 1234
  (gdb) set detach-on-fork off
  (gdb) continue

Normally GDB will stop when the child process receives SIGSEGV.  At that
point, you should be able to type “bt” to get the backtrace.

Could you try that?

Thanks in advance,
Ludo’.

Previous Next

GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.