[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [MiNT] wakegem + naes 2

> > >As far as I know it is still needed. It seems this is a kernel problem,
> > >not any AES (otherwise it would be strange, if all AESes would share same
> > >bug). But. As of kernel version 1.15.9 or something, the WakeGEM causes
> > >occasional crashes because, apparently, the kernel's interval timer does
> > >not work correctly. I have no clue what was changed and where, CHANGES
> > >file does not indicate anything suspicious (but at the other hand I
> > >discovered that CHANGES file is often not updated as it should, i.e the
> > >kernel changes are not reflected there).
> > 
> > Interesting.  Perhaps it was wakegem causing the newer betas to be
> > unstable here. It was b9 and b10 that I had problems with sudden
> > freeze ups.
>  This seems to be a problem on TT and Falcons only? I've never ever had
> n.aes lock up like that on my Hades, but it did on my TT. This might also
> be why 1.15.9b up to -current, both with and without MP enabled, is
> _extremely_ stable on my Hades.

Here too (F030). With exception, when wakegem is loaded. In this situation
the system suffers two-three crashes a day. And this is not an AES lockup,
but the kernel crash (assert inside the scheduler). And generally, not the
wakegem is guilty for that, but this will occur in presence of any program
which programs the SIGALRM to be periodically sent (which wakegem does).

Problems arose at around 1.15.8, but, as I recalled after sending the
previous mail, this may also be a fault of the compiler. Because around
that time I started compiling my kernels with gcc 2.95.2 instead of 2.8.1.

Konrad M.Kokoszkiewicz
mail: draco@atari.org

** Ea natura multitudinis est,
** aut servit humiliter, aut superbe dominatur (Liv. XXIV,25)
** Taka to juz natura pospolstwa, ze albo sluzy ono unizenie,
** albo bezczelnie sie panoszy.