RPCS3/llvm-mirror
1
0
Fork 0
mirror of https://github.com/RPCS3/llvm-mirror.git synced 2025-02-11 15:26:07 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

AMDGPU: s_waitcnt field should be treated as unsigned

Browse Source 
Also make it an ImmLeaf, so it should work with global isel as well,
which was part of the point of moving it in the first place.

llvm-svn: 365842
This commit is contained in:
Matt Arsenault 2019-07-11 23:42:57 +00:00
parent a33a1a19cb
commit 74012b6127
3 changed files with 19 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
lib/Target/AMDGPU/SIInstrInfo.td
View File

@ -622,8 +622,12 @@ class bitextract_imm<int bitnum> : SDNodeXForm<imm, [{
return CurDAG->getTargetConstant(Bit, SDLoc(N), MVT::i1);
}]>;
def SIMM16bit : PatLeaf <(imm),
[{return isInt<16>(N->getSExtValue());}]
def SIMM16bit : ImmLeaf <i32,
[{return isInt<16>(Imm);}]
>;
def UIMM16bit : ImmLeaf <i32,
[{return isUInt<16>(Imm); }]
>;
class InlineImm <ValueType vt> : PatLeaf <(vt imm), [{

2
lib/Target/AMDGPU/SOPInstructions.td
View File

@ -1090,7 +1090,7 @@ def S_WAKEUP : SOPP <0x00000003, (ins), "s_wakeup"> {
let mayLoad = 1, mayStore = 1, hasSideEffects = 1 in
def S_WAITCNT : SOPP <0x0000000c, (ins WAIT_FLAG:$simm16), "s_waitcnt $simm16",
[(int_amdgcn_s_waitcnt SIMM16bit:$simm16)]>;
[(int_amdgcn_s_waitcnt UIMM16bit:$simm16)]>;
def S_SETHALT : SOPP <0x0000000d, (ins i16imm:$simm16), "s_sethalt $simm16">;
def S_SETKILL : SOPP <0x0000000b, (ins i16imm:$simm16), "s_setkill $simm16">;

12
test/CodeGen/AMDGPU/llvm.amdgcn.s.waitcnt.ll
View File

@ -31,6 +31,18 @@ define amdgpu_ps void @test2(<8 x i32> inreg %rsrc, i32 %c) {
ret void
}
; CHECK-LABEL: {{^}}test3:
; CHECK: image_load
; CHECK: s_waitcnt vmcnt(0) lgkmcnt(0)
; CHECK: image_store
define amdgpu_ps void @test3(<8 x i32> inreg %rsrc, i32 %c) {
%t = call <4 x float> @llvm.amdgcn.image.load.1d.v4f32.i32(i32 15, i32 %c, <8 x i32> %rsrc, i32 0, i32 0)
call void @llvm.amdgcn.s.waitcnt(i32 49279) ; not isInt<16>, but isUInt<16>
%c.1 = mul i32 %c, 2
call void @llvm.amdgcn.image.store.1d.v4f32.i32(<4 x float> %t, i32 15, i32 %c.1, <8 x i32> %rsrc, i32 0, i32 0)
ret void
}
declare void @llvm.amdgcn.s.waitcnt(i32) #0
declare <4 x float> @llvm.amdgcn.image.load.1d.v4f32.i32(i32, i32, <8 x i32>, i32, i32) #1