aaron/ispc
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

Add separate variants of memory built-ins for floats and doubles.

Browse Source 
Previously, we'd bitcast e.g. a vector of floats to a vector of i32s and then
use the i32 variant of masked_load/masked_store/gather/scatter.  Now, we have
separate float/double variants of each of those.
This commit is contained in:
Matt Pharr
2012-06-07 14:29:17 -07:00
parent 1ac3e03171
commit 89a2566e01
17 changed files with 593 additions and 41 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
builtins/target-avx2.ll
View File

@@ -107,4 +107,6 @@ define i16 @__float_to_half_uniform(float %v) nounwind readnone {
gen_gather(i8)
gen_gather(i16)
gen_gather(i32)
gen_gather(float)
gen_gather(i64)
gen_gather(double)