From owner-freebsd-current  Fri Mar 24 17:35:02 1995
Return-Path: current-owner
Received: (from majordom@localhost) by freefall.cdrom.com (8.6.10/8.6.6) id RAA27421 for current-outgoing; Fri, 24 Mar 1995 17:35:02 -0800
Received: from pelican.com (pelican.com [134.24.4.62]) by freefall.cdrom.com (8.6.10/8.6.6) with SMTP id RAA27406 for <current@freebsd.org>; Fri, 24 Mar 1995 17:34:59 -0800
Received: by pelican.com (Smail3.1.28.1 #5)
	id m0rsKkV-000K3WC; Fri, 24 Mar 95 17:34 WET
Message-Id: <m0rsKkV-000K3WC@pelican.com>
From: pete@pelican.com (Pete Carah)
Subject: xv messed up
To: current@FreeBSD.org
Date: Fri, 24 Mar 1995 17:34:42 -0800 (PST)
X-Mailer: ELM [version 2.4 PL23]
Content-Type: text
Content-Length: 674       
Sender: current-owner@FreeBSD.org
Precedence: bulk

xv seems to mess up on some PDS/VICAR files (yes, the voyager 2 pictures)
if -O is used; this seems to attack both 2.6.2 and whatever current is
(2.6.3?)

It seg-faults in the vicinity of the switch in ReadPicFile in xv.c,
with a bogus value for the fname (the pointer itself contains some
miscellaneous upper-case ascii, if I can trust gdb...)  The exact location
of the seg fault and other messages that appear differ depending on the
number and placement of fprintf(stderr, ...) statements; that is often
symptomatic of optimizer bugs.  The code itself looks very straightforward
in that area.

Works fine with -O not specified and doesn't "seem" slower either.

-- Pete