Re: Variable confidence/urgency prefetch?
- From: Seongbae Park <Seongbae.Park@xxxxxxx>
- Date: Wed, 5 Apr 2006 16:11:48 +0000 (UTC)
Paul Aaron Clayton <Dysthymicdolt@xxxxxxx> wrote:
Seongbae Park wrote:
[snip]
US4+ has weak and strong prefetches -
the processor is allowed to drop weak prefetches on various
conditions (typically prefetch queue full condition or TLB miss),
whereas the strong prefetch are not dropped (almost) no matter what.
The intention is just as you said - one for potentially higher cost
but guaranteed prefetching, the other for guaranteed
small/no runtime cost but no guarantee of actual prefetching.
In contrast, US3 and US4 have only weak prefetches
which often forced the compiler to issue duplicate prefetches
to ensure all necessary data are really prefetched.
See 8.72.2 Weak versus Strong Prefetch of UA2005:
http://opensparc.sunsource.net/specs/UA2005-current-draft-P-EXT.pdf
A belated thank you for the pointer. (BTW, why is saving this public
document unallowed--especially as a little extra effort allows one to
save it anyway?)
No idea.
http://blogs.sun.com/roller/page/dweaver
probably can tell you why.
It seems UltraSPARC does not have an equivalent to Alpha's
Write Hint 64bytes or PPC's Data Cache Block Zero (or Allocate).
Did the guardians of the ISA judge that the block write instruction
(which writes 8 FP registers to memory) was adequate for most
uses of WH64/DCB[ZA] or that such instructions provide little
benefit in real programs or what?
It seems that some memory
allocators (especially a binary buddy allocator?) could use such to
avoid accesses to main memory for blocks that initially contain no
meaningful data.
See 5.9 Block Initiailizing Store in
http://opensparc.sunsource.net/specs/UST1-UASuppl-current-draft-P-EXT.pdf
Please note that this is T1 supplement to the UA2005 spec.
I guess "the guardians" (whoever they are) haven't decided on
what form of block init store is acceptable
for the non-implementation specific portion of ISA.
My guess is that some form of this will percolate up
to the general UA200x specification in the future.
--
#pragma ident "Seongbae Park, compiler, http://blogs.sun.com/seongbae/"
.
- References:
- Re: Variable confidence/urgency prefetch?
- From: Paul Aaron Clayton
- Re: Variable confidence/urgency prefetch?
- Prev by Date: Re: RFC - New Object-Oriented Method of Parallel Programming
- Next by Date: Re: SIMD/Vector benchmarks
- Previous by thread: Re: Variable confidence/urgency prefetch?
- Index(es):
Relevant Pages
|
Loading