--- On Fri, 6/20/08, Benjamin Herrenschmidt <[EMAIL PROTECTED]> wrote:
> I though OS X had a trick with a CR bit that would disable
> the dcbz optimization on the first alignment fault ? Or did they
> totally remove it ?
Ah, it's coming back to me. :)
Apple added 'dcbz', removed it, and then there was the clever trick of
optimizing the code path with a boot-time perf test and/or changing the code on
the first fault...I'm not sure what's implemented in the recent builds.
If Linux can do something similar, that should allow good perf on cacheable and
cache-inhibited space as well as different CPUs (eg, if 'dcba' is available,
then you don't need the alignment fault hack).
--Sanjay
_______________________________________________
Linuxppc-dev mailing list
[email protected]
https://ozlabs.org/mailman/listinfo/linuxppc-dev
- [RFC 1/3] powerpc: __copy_tofrom_user tweaked for C... Mark Nelson
- Re: [RFC 1/3] powerpc: __copy_tofrom_user twea... Arnd Bergmann
- Re: [RFC 1/3] powerpc: __copy_tofrom_user ... Gunnar von Boehn
- Re: [RFC 1/3] powerpc: __copy_tofrom_u... Sanjay Patel
- Re: [RFC 1/3] powerpc: __copy_tofr... Gunnar von Boehn
- Re: [RFC 1/3] powerpc: __copy... Sanjay Patel
- Re: [RFC 1/3] powerpc: __... Benjamin Herrenschmidt
- Re: [RFC 1/3] powerpc... Sanjay Patel
- Re: [RFC 1/3] powerpc: __... Gunnar von Boehn
- Re: [RFC 1/3] powerpc... Geert Uytterhoeven
- Re: [RFC 1/3] powerpc... Paul Mackerras
- Re: [RFC 1/3] powerpc... Gunnar von Boehn
- Re: [Cbe-oss-dev] [RFC 1/3] powerpc: _... Paul Mackerras
- Re: [Cbe-oss-dev] [RFC 1/3] powerp... Gunnar von Boehn
- Re: [Cbe-oss-dev] [RFC 1/3] powerp... Arnd Bergmann
- Re: [Cbe-oss-dev] [RFC 1/3] p... Paul Mackerras
- Re: [Cbe-oss-dev] [RFC 1/... David Miller
- Re: [Cbe-oss-dev] [RFC 1/... Arnd Bergmann
