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

Add SSE4-16 target.

Browse Source 
Along the lines of sse4-8, this is an 8-wide target for SSE4, using
16-bit elements for the mask.  It's thus (in principle) the best
target for SIMD computation with 16-bit datatypes.
This commit is contained in:
Matt Pharr
2013-07-24 09:40:50 -07:00
parent 04d61afa23
commit 780b0dfe47
5 changed files with 463 additions and 7 deletions
Download Patch File Download Diff File Expand all files Collapse all files

14
ispc.cpp
View File

@@ -318,6 +318,14 @@ Target::Target(const char *arch, const char *cpu, const char *isa, bool pic) :
this->m_maskingIsFree = false;
this->m_maskBitCount = 8;
}
else if (!strcasecmp(isa, "sse4-16")) {
this->m_isa = Target::SSE4;
this->m_nativeVectorWidth = 8;
this->m_vectorWidth = 8;
this->m_attributes = "+sse,+sse2,+sse3,+sse41,-sse42,-sse4a,+ssse3,-popcnt,+cmov";
this->m_maskingIsFree = false;
this->m_maskBitCount = 16;
}
else if (!strcasecmp(isa, "generic-4")) {
this->m_isa = Target::GENERIC;
this->m_nativeVectorWidth = 4;
@@ -575,9 +583,9 @@ Target::SupportedTargetArchs() {
const char *
Target::SupportedTargetISAs() {
return "neon, sse2, sse2-x2, sse4, sse4-x2, avx, avx-x2"
", avx1.1, avx1.1-x2, avx2, avx2-x2"
", generic-1, generic-4, generic-8, generic-16, generic-32";
return "neon, sse2, sse2-x2, sse4, sse4-8, sse4-16, sse4-x2, "
"avx, avx-x2, avx1.1, avx1.1-x2, avx2, avx2-x2,"
"generic-1, generic-4, generic-8, generic-16, generic-32";
}