On Sat, 19 Feb 2000, Andrea Arcangeli wrote:
[ ... ]
> >glad to hear DEC/COMPAQ ingenieers about the status of the
> >__delay()/__udelay() stuff for alpha in Linuxes. For now, Pamela has to
>
> There's just an attempt to fix it. It's a patch from Jay Estabrook, I have
> uploaded it here but I don't need it with my hardware:
>
> ftp://ftp.*.kernel.org/pub/linux/kernel/people/andrea/patches/v2.2/2.2.14/alpha_delay-fix_jay-2.2.14.gz
>
> Maybe Jay can add some comment about this. The patch is strightforward, it
> make sure that the udelay code is allocated in a misaligned address.
I will add my comments too: ;-)
Why sticking with inlining for this stuff? Getting rid of inlining does
insure this stuff be aligned in everything the same way and would likely
fix the problem forever. Fixing it in memory would not affect precision
for micro-second even on early alpha machines, in my opinion, given clock
frequencies used on alpha since years.
The udelay stuff is designed to waste deterministic amounts of time and
some minimal precision requirement should be specified. Otherwise it is
just usable at nothing good.
Gérard.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Wed Feb 23 2000 - 21:00:23 EST