Subject: Re: bin/28590: binaries in NetBSD 2.0 should be stripped
To: None <firstname.lastname@example.org>
From: James Chacon <jmc@NetBSD.org>
Date: 12/09/2004 10:21:36
On Thu, Dec 09, 2004 at 07:21:00AM +0000, email@example.com wrote:
> >Number: 28590
> >Category: bin
> >Synopsis: binaries in NetBSD 2.0 should be stripped
> >Confidential: no
> >Severity: non-critical
> >Priority: medium
> >Responsible: bin-bug-people
> >State: open
> >Class: sw-bug
> >Submitter-Id: net
> >Arrival-Date: Thu Dec 09 07:21:00 +0000 2004
> >Originator: Igor Sobrado
> >Release: NetBSD 2.0 RC5 and NetBSD 2.0 (current ISO images)
> University of Oviedo
> Sorry, the machine is not networked yet. It is using the standard
> GENERIC_LAPTOP kernel in both 2.0_RC5 and 2.0.
> Binaries in NetBSD 2.0 RC5 and NetBSD 2.0 are not stripped. I am
> not sure if it is a requirement for debugging SMP and native threads,
> but it is a bit annoying running not stripped binaries in a
> production system. Debugging information should not be required
> in production binaries.
> The file(1) command provides required feedback about the state of
> each binary in the NetBSD operating system.
> Rebuilding the binaries on the NetBSD operating system without
> debugging information.
The file don't have debugging information in them (they aren't built w. -g).
All stripping will do at this point is remove the rest of the symbol
information which makes getting stack traces from potential core dumps fairly
useless. It's adding no real overhead to the system and retaining that
symbol data is still useful. It's not the same as "debugging information".