[vox-tech] gzip bug?

[Peter Jay Salzman]

jeez.  i sort of wished you hadn't posted this; i was happier not
knowing.  lol.  now i'm going to be thinking about my non-optimized
"ready for i386" version of gzip running on my 1.4GHz athlon all day
now.   ;-)

one of these days i'll have to follow the "how to recompile your entire
debian system for i586" tutorial.  :-)

still doesn't explain the zlib thing, although maybe it's linked
statically because any compression scheme, in its heart of hearts is
number cruncher.

is there be any performance penalty for a dynamically linked executable
above and beyond ld.so's import of external objects at load time?

pete


begin Charles Polisher <cpolish@attbi.com> 
> On Fri, Mar 15, 2002 at 08:03:22AM -0800, Peter Jay Salzman wrote:
> > begin Charles Polisher <cpolish@attbi.com> 
> > > On Thu, Mar 14, 2002 at 11:14:08PM -0800, Kevin Dawson wrote:
> >  
> > yeah, before i replied to this message, i checked what libraries gzip is
> > linked against -- it's just linked against libc, and not zlib (which is
> > odd.  i was expecting zlib.)
> 
> *nods* Still scratching for an answer, I installed the
> sources for gzip, and found this tidbit in the README:
> 
> 
> 	WARNING: on several systems, compiler bugs cause gzip to fail, in
> 	particular when optimization options are on.  See the section "Special
> 	targets" at the end of the INSTALL file for a list of known problems.
> 	For all machines, use "make check" to check that gzip was compiled
> 	correctly.  Try compiling gzip without any optimization if you have a
> 	problem.
> 
> >From the md5sum signature, it appears the installed binary giving trouble
> is the 586 optimized version. There's a small chance that installing the
> i386 version will fix the problem...
> 
> ftp://ftp.redhat.com/pub/redhat/linux/7.1
>       /en/os/i386/RedHat/RPMS/gzip-1.3-12.i386.rpm
> 
> (url broken in half to fit the screen)