Re: Patch: fix gdb_gcore.sh failure in ash/dash

[Daniel Jacobowitz <drow@false.org>]

On Sun, Sep 11, 2005 at 11:18:56PM +0200, Mark Kettenis wrote:
> > Date: Fri, 09 Sep 2005 19:11:13 -0700
> > From: Qunying Pan <qpan@mvista.com>
> > 
> > Hi,
> > 
> > Running gdb_gcore.sh in ash/dash fails with message "Hangup detected on 
> > fd 0". The following patch fixes the problem. Acceptable?
> 
> This seems like a bug in ash/dash (whatever that is).  Why isn't your
> /bin/sh not a proper bourne shell?
> 
> I'm not inclined to apply this patch just because people link /bin/sh
> to some random piece of crap.  So I'd like to see somewhat more
> motivation.

It's perfectly legitimate behavior, as far as I can tell - it depends
how long the shell keeps the pipe open and whether it uses a tempfile.
IIRC bash will prefer to use a tempfile and ash will pipe?  Not sure. 
GDB detects the SIGHUP before it writes out the coredump.

Try this, in either ash or bash:
  printf 'printf "hi\\n"\nquit\n' | gdb

Neither shell's going to give you a printf, just a hangup message.

-- 
Daniel Jacobowitz
CodeSourcery, LLC