Displaying 2 results from an estimated 2 matches for "lomask".
Did you mean:
nomask
2016 Feb 01
2
Question about store with unaligned memory address
Hi Bruce,
Thanks for response.
I also think it is not good way. Do you have the other ways to legalize it?
Thanks,
JinGu Kang
2016-02-01 13:11 GMT+00:00 Bruce Hoult <bruce at hoult.org>:
> In fact this is a pretty bad legalizing/lowering because you only need to
> load and edit for the first and last values in the vector. The other words
> are completely replaced and don't
2011 Mar 28
2
[LLVMdev] [Discussion] Portable SIMD programming using LLVM?
...on algorithm. For example, simd<4>::add can be implemented based on
simd<8>::add in the following manner.
===strategy_add_doubling===
1. hiMask = simd<8>::himask() //this is a constant with high 4 bits of each
8-bits field are 1111 and low 4 bits of each 8-bits field are 0000.
2. loMask = simd<8>::lowmask() //this is a constant with high 4 bits of each
8-bits field are 0000 and low 4 bits of each 8-bits field are 1111.
3. c = simd<8>::add(simd_and(loMask, a), simd_and(loMask, b))
4. d = simd<8>::add(simd_and(hiMask, a), simd_and(hiMask, b))
5. return simd_or(d, s...