6502/llvm-6502
1
0
Fork 0
mirror of https://github.com/c64scene-ar/llvm-6502.git synced 2025-03-01 01:30:36 +00:00
Code Issues Projects Releases Wiki Activity

Fix 256-bit PALIGNR comment decoding to understand that it works on independent 256-bit lanes.

Browse Source 
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@173674 91177308-0d34-0410-b5e6-96231b3b80d8
This commit is contained in:
Craig Topper 2013-01-28 07:41:18 +00:00
parent 6ab4cbc986
commit 467016e58d
2 changed files with 26 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

13
lib/Target/X86/Utils/X86ShuffleDecode.cpp
View File

@ -66,8 +66,17 @@ void DecodePALIGNRMask(MVT VT, unsigned Imm,
unsigned NumElts = VT.getVectorNumElements();
unsigned Offset = Imm * (VT.getVectorElementType().getSizeInBits() / 8);
for (unsigned i = 0; i != NumElts; ++i)
ShuffleMask.push_back((i + Offset) % (NumElts * 2));
unsigned NumLanes = VT.getSizeInBits() / 128;
unsigned NumLaneElts = NumElts / NumLanes;
for (unsigned l = 0; l != NumElts; l += NumLaneElts) {
for (unsigned i = 0; i != NumLaneElts; ++i) {
unsigned Base = i + Offset;
// if i+offset is out of this lane then we actually need the other source
if (Base >= NumLaneElts) Base += NumElts - NumLaneElts;
ShuffleMask.push_back(Base + l);
}
}
}
/// DecodePSHUFMask - This decodes the shuffle masks for pshufd, and vpermilp*.

15
test/MC/X86/shuffle-comments.s
View File

@ -29,3 +29,18 @@ vpalignr $0, %xmm0, %xmm1, %xmm2
# CHECK: xmm2 = xmm0[0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15]
vpalignr $0, (%rax), %xmm1, %xmm2
# CHECK: xmm2 = mem[0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15]
vpalignr $8, %ymm0, %ymm1, %ymm2
# CHECK: ymm2 = ymm0[8,9,10,11,12,13,14,15],ymm1[0,1,2,3,4,5,6,7],ymm0[24,25,26,27,28,29,30,31],ymm1[16,17,18,19,20,21,22,23]
vpalignr $8, (%rax), %ymm1, %ymm2
# CHECK: ymm2 = mem[8,9,10,11,12,13,14,15],ymm1[0,1,2,3,4,5,6,7],mem[24,25,26,27,28,29,30,31],ymm1[16,17,18,19,20,21,22,23]
vpalignr $16, %ymm0, %ymm1, %ymm2
# CHECK: ymm2 = ymm1[0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31]
vpalignr $16, (%rax), %ymm1, %ymm2
# CHECK: ymm2 = ymm1[0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31]
vpalignr $0, %ymm0, %ymm1, %ymm2
# CHECK: ymm2 = ymm0[0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31]
vpalignr $0, (%rax), %ymm1, %ymm2
# CHECK: ymm2 = mem[0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31]