Re: Fastcode MM Rule
- From: "Pierre le Riche" <pleriche@xxxxxxxxxxx>
- Date: Wed, 4 May 2005 13:45:41 +0200
Hi Robert,
> I'm curious to see whether Pierre's new MM will invalidate the previous
> analysis!
I kinda put that version on ice when everyone complained that I'm not
allowed to use the last few bytes in a block (that dword reading/writing
rule). It had 8-byte aligned blocks with 1 byte overhead between blocks. The
memory usage was good, but I still had to work on the speed. I've still got
it somewhere, so I can finish it if that rule ever gets changed.
At the moment I'm working on a different idea, but I don't get much time to
work on it, so it might take a while (and even then the idea might turn out
to be not that great). Sadly it seems the more genius a plan seems
initially, the more complicated the implementation becomes and the slower it
eventually turns out to be :-(.
Regards,
Pierre
.
- Follow-Ups:
- Re: Fastcode MM Rule
- From: Dennis
- Re: Fastcode MM Rule
- References:
- Re: Fastcode MM Rule
- From: Eric Grange
- Re: Fastcode MM Rule
- From: Robert Houdart
- Re: Fastcode MM Rule
- From: Eric Grange
- Re: Fastcode MM Rule
- From: Robert Houdart
- Re: Fastcode MM Rule
- Prev by Date: Re: New replay benchmark : XML Parser
- Next by Date: Re: New replay benchmark : XML Parser
- Previous by thread: Re: Fastcode MM Rule
- Next by thread: Re: Fastcode MM Rule
- Index(es):
Relevant Pages
|