Re: md5 lossage

[Richard Stallman]

    I previously reported that `md5' gives the wrong results on my debian
    sparc box.  I find it isn't confined to that.  It's also wrong on
    Solaris 8 and Irix 6.5 with recent sources, but OK on x86 debian and
    redhat with various gcc versions and with cc on irix.  (It's also OK
    generally on 21.3 as far as I can tell.)  Thus it looks like an endian
    issue with the current code.

The file md5.c has not changed since the start of 2001, so if it
did not have an endian problem in 21.3, I doubt it has one now.
It seems to be designed to cope with endian issues.

I have a vague memory that the Sun and MIPS are 64-bit machines at
least recently.  Could this be a 64-bit problem, perhaps?

If you could step through the code in the case that fails on both the
Sun and the x86, comparing the data at various points, I expect you
could find the bug fairly straightforwardly.  Could you please
do that?