mirror of
https://github.com/c64scene-ar/llvm-6502.git
synced 2024-12-23 15:29:51 +00:00
5f0d9dbdf4
This matters for example in following matrix multiply:
int **mmult(int rows, int cols, int **m1, int **m2, int **m3) {
int i, j, k, val;
for (i=0; i<rows; i++) {
for (j=0; j<cols; j++) {
val = 0;
for (k=0; k<cols; k++) {
val += m1[i][k] * m2[k][j];
}
m3[i][j] = val;
}
}
return(m3);
}
Taken from the test-suite benchmark Shootout.
We estimate the cost of the multiply to be 2 while we generate 9 instructions
for it and end up being quite a bit slower than the scalar version (48% on my
machine).
Also, properly differentiate between avx1 and avx2. On avx-1 we still split the
vector into 2 128bits and handle the subvector muls like above with 9
instructions.
Only on avx-2 will we have a cost of 9 for v4i64.
I changed the test case in test/Transforms/LoopVectorize/X86/avx1.ll to use an
add instead of a mul because with a mul we now no longer vectorize. I did
verify that the mul would be indeed more expensive when vectorized with 3
kernels:
for (i ...)
r += a[i] * 3;
for (i ...)
m1[i] = m1[i] * 3; // This matches the test case in avx1.ll
and a matrix multiply.
In each case the vectorized version was considerably slower.
radar://13304919
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@176403 91177308-0d34-0410-b5e6-96231b3b80d8
|
||
|---|---|---|
| .. | ||
| avx1.ll | ||
| conversion-cost.ll | ||
| cost-model.ll | ||
| gcc-examples.ll | ||
| lit.local.cfg | ||
| min-trip-count-switch.ll | ||
| no-vector.ll | ||
| parallel-loops-after-reg2mem.ll | ||
| parallel-loops.ll | ||
| small-size.ll | ||
| struct-store.ll | ||
| unroll_selection.ll | ||
| unroll-small-loops.ll | ||
| vector_ptr_load_store.ll | ||