Re: glibc-2.0.93 homebrew distribution tests

[John Kennedy <jk@csuchico.edu>]

[Andreas Jaeger]
> John>   The mount thing is weird.  The binaries I compiled on the
> John> i686-pc-linux-gnu seem to be having troubles with something in
> John> or behind strtok() on a i586-pc-linux-gnu box.  Same everything
> John> mind you... just taking the chrooted environment, copying it onto
> John> a different computer's disk and booting it.  I've never had a
> John> problem with that before (using 2.0.5 -> 2.0.7pre3, in any case).
> 
> ... might John's glibc be optimzied for i686 - and therefore not
> working on i586?  The test strtok program works for my old i486
> without problems.

[Ulrich Drepper]
>  This certainly could be a reason but I cannot see how this can
>  happen since the configure script determines the processor type.

  Remember, I compiled it on the x686 and copied it onto the x586.  It
was never recompiled for the x586.  That was one of my first guesses
(which is why I mentioned it) since I see a lot of inline assembly at
times for optimized string functions, but since it was time-expensive to
test I hadn't done it yet.  I didn't have any problems doing the exact
same thing with 2.0.5 thru 2.0.7pre3 so I've been figuring that most of
the x686 optimizations didn't affect libc.

  Note that I'm using gcc-2.8.1 and binutils-2.9.1.0.4, so I'm not
using one of the really blatant optimizing compilers.

  If 2.0.93 is doing something new, it is a good time.  They just
released linux-2.1.103 so I'll be recompiling everything again anyway
(part of my kernel and software endurance test; if it survives compiling
everything (3+ hours), it is probably good).  (:
								--- john