Re: x86, ARM, PARISC, PPC, MIPS and Sparc folks please run this

[davidsen@tmr.com (bill davidsen)]

In article <20030901020203.1779efe8.davem@redhat.com>,
David S. Miller <davem@redhat.com> wrote:

| > This is my strategy:
| > 
| > 	mmap MAP_ANON without MAP_FIXED to find a free area
| > 	mmap MAP_FIXED over the anon area at same address
| > 	mmap MAP_FIXED over the anon area at larger address
| > 
| > I don't see any strategy that lets me establish this kind of circular
| > mapping on Sparc without either (a) knowing the value of SHMLBA, or
| > (b) risking clobbering another thread's mmap.
| 
| Why do you need the same piece of data mapped to multiple places
| in the first place, and why at specific addresses?  It's purely an
| optimization of some sort, right?

I think he said he was doing DSP... there's a trick of double mapping
the same memory to save one subscript calculation in FFT (or maybe DFT)
inner loop. The only reason I know this is that a friend did a master's
thesis on DSP about 20 years ago, and I absorbed some info I hope to
never need. He also coded an FFT instruction in the LCS (programmable
firmware) of a VAX.

I am only speculating, of course.
-- 
bill davidsen <davidsen@tmr.com>
  CTO, TMR Associates, Inc
Doing interesting things with little computers since 1979.