Re: Mint 1.10 h7 bug

[hohmuth@inf.tu-dresden.de (Michael Hohmuth)]

Ulrich Kuehn wrote:

> After applying the h7 patch, I discovered that under MultiTOS the system
> again hangs quite often while using window gadgets. This did not happen
> with the h5 patch.

Just a wild guess: I don't know whether MultiTOS uses pipes for
anything, but... perhaps there's a bug lurking around in Juergen's
pipe optimizations patches?  Perhaps, while fixing an off-by one
error, it introduces another one?  (I have to admit that I didn't look
at that code...)

This patch was new with h7:

From: Juergen Lock <nox@jelal.north.de>
Subject: Re: MiNTlib (I think I'm on pl43): Bug in sprintf().
Message-Id: <9407172345.AA00280@jelal.north.de>

biosfs.c,pipefs.c,tty.c:
        add readb+writeb for ptys (for the slave->master pipe),
        use memcpy for all pipes
        really buffer 4K not 4K - 1 byte so the copies are aligned more often

> This behavior did also occur before, when I compiled mint with gcc 2.5.8;
> but this time, it was compiled with gcc 2.3.3, which cured the problem
> before....

One more guess: Perhaps array handling / pointer arithmetics had been
changed between 2.3.3 and 2.5.8?

Michael
-- 
Email: hohmuth@inf.tu-dresden.de
WWW:   http://www.inf.tu-dresden.de/~mh1/