Re: [avr-libc-dev] [ Bug #1720 ] memcmp bug ?

[Joerg Wunsch]

As Joerg Wunsch wrote:

> > Is this really a bug?
> 
> Yes, it is:

I should have started thinking before writing...

Of course, as Marek already pointed out, the code is absolutely
correct, and handles the unsigned comparision pretty well.  After
thinking more about it, at least i came to this conclusion myself,
but Marek was faster in sending his message.

No, i don't think we need to make everything and all -mint8 proof.
So this is just a typical case where the documentation then needs
to document that it isn't, and we're done.  Still, the existing
implementation might be useful even for -mint8 users, in case all
they want to do is comparing for equality.
-- 
J"org Wunsch                                           Unix support engineer
address@hidden        http://www.interface-systems.de/~j/