On October 2, 2014 9:50:45 AM CEST, Jakub Jelinek <ja...@redhat.com> wrote:
>On Thu, Oct 02, 2014 at 11:15:10AM +0400, Evgeny wrote:
>> Hold on. The patch has a conflict with previously approved here.
>
>Which?
>
>In any case, I've already committed the patch.
>
>That said, if what you care now about is pr52252-atom.c with -O2
>-ftree-vectorize -mavx2, let's look at all the permutations it needs
>below. The patch I've committed improved f3, from
>3x vpshufb + 2x vpermq + 2x vpor into vperm2i128 $33 + vpalignr $11,
>left other permutations as is.
>
>So, do you have suggestions on what insns we should use for the other
>permutations? Then we can figure out how to best achieve that.
>Note that with dozens of permutation instructions in the ISA the
>computational complexity of finding optimal sequence might be too
>expensive.
I always wondered if we can use some form of a compressed table to store all
permutations and compute that statically by some means (at build time or even
checked into the repo). But yes, the complexity is quite high...
Also note the recent bug I filed about using whole-vector shifts for special
permutes so we can get rid of VEC_LSHIFT_EXPR.
Richard.
>typedef unsigned char V __attribute__ ((vector_size (32)));
>V a, b, c;
>
>void
>f1 (void)
>{
>c = __builtin_shuffle (a, b, (V) { 0, 32, 1, 33, 2, 34, 3, 35, 4, 36,
>5, 37, 6, 38, 7, 39, 8, 40, 9, 41, 10, 42, 11, 43, 12, 44, 13, 45, 14,
>46, 15, 47 });
>}
>
>void
>f2 (void)
>{
>c = __builtin_shuffle (a, b, (V) { 0, 3, 6, 9, 12, 15, 18, 21, 24, 27,
>30, 1, 4, 7, 10, 13, 16, 19, 22, 25, 28, 31, 2, 5, 8, 11, 14, 17, 20,
>23, 26, 29 });
>}
>
>void
>f3 (void)
>{
>c = __builtin_shuffle (a, b, (V) { 11, 12, 13, 14, 15, 16, 17, 18, 19,
>20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37,
>38, 39, 40, 41, 42 });
>}
>
>void
>f4 (void)
>{
>c = __builtin_shuffle (a, b, (V) { 16, 48, 17, 49, 18, 50, 19, 51, 20,
>52, 21, 53, 22, 54, 23, 55, 24, 56, 25, 57, 26, 58, 27, 59, 28, 60, 29,
>61, 30, 62, 31, 63 });
>}
>
>void
>f5 (void)
>{
>c = __builtin_shuffle (a, b, (V) { 21, 22, 23, 24, 25, 26, 27, 28, 29,
>30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47,
>48, 49, 50, 51, 52 });
>}
>
>void
>f6 (void)
>{
>c = __builtin_shuffle (a, b, (V) { 22, 23, 24, 25, 26, 27, 28, 29, 30,
>31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48,
>49, 50, 51, 52, 53 });
>}
>
> Jakub
