diff options
Diffstat (limited to 'llvm/lib')
| -rw-r--r-- | llvm/lib/Target/X86/X86InstrSSE.td | 85 |
1 files changed, 41 insertions, 44 deletions
diff --git a/llvm/lib/Target/X86/X86InstrSSE.td b/llvm/lib/Target/X86/X86InstrSSE.td index 668ea97631a..359026c3e51 100644 --- a/llvm/lib/Target/X86/X86InstrSSE.td +++ b/llvm/lib/Target/X86/X86InstrSSE.td @@ -822,9 +822,9 @@ def MOVHLPSrr : PSI<0x12, MRMSrcReg, (outs VR128:$dst), let AddedComplexity = 20 in { def : Pat<(v4f32 (movddup VR128:$src, (undef))), - (MOVLHPSrr VR128:$src, VR128:$src)>, Requires<[HasSSE1]>; + (MOVLHPSrr VR128:$src, VR128:$src)>; def : Pat<(v2i64 (movddup VR128:$src, (undef))), - (MOVLHPSrr VR128:$src, VR128:$src)>, Requires<[HasSSE1]>; + (MOVLHPSrr VR128:$src, VR128:$src)>; } @@ -1084,13 +1084,11 @@ let isReMaterializable = 1, isAsCheapAsAMove = 1, canFoldAsLoad = 1, def V_SET0 : PSI<0x57, MRMInitReg, (outs VR128:$dst), (ins), "", [(set VR128:$dst, (v4i32 immAllZerosV))]>; -let Predicates = [HasSSE1] in { - def : Pat<(v2i64 immAllZerosV), (V_SET0)>; - def : Pat<(v8i16 immAllZerosV), (V_SET0)>; - def : Pat<(v16i8 immAllZerosV), (V_SET0)>; - def : Pat<(v2f64 immAllZerosV), (V_SET0)>; - def : Pat<(v4f32 immAllZerosV), (V_SET0)>; -} +def : Pat<(v2i64 immAllZerosV), (V_SET0)>; +def : Pat<(v8i16 immAllZerosV), (V_SET0)>; +def : Pat<(v16i8 immAllZerosV), (V_SET0)>; +def : Pat<(v2f64 immAllZerosV), (V_SET0)>; +def : Pat<(v4f32 immAllZerosV), (V_SET0)>; // FR32 to 128-bit vector conversion. let isAsCheapAsAMove = 1 in @@ -3051,13 +3049,13 @@ let Predicates = [HasSSE2] in { let AddedComplexity = 15 in { // Zeroing a VR128 then do a MOVS{S|D} to the lower bits. def : Pat<(v2f64 (X86vzmovl (v2f64 (scalar_to_vector FR64:$src)))), - (MOVLSD2PDrr (V_SET0), FR64:$src)>, Requires<[HasSSE2]>; + (MOVLSD2PDrr (V_SET0), FR64:$src)>; def : Pat<(v4f32 (X86vzmovl (v4f32 (scalar_to_vector FR32:$src)))), - (MOVLSS2PSrr (V_SET0), FR32:$src)>, Requires<[HasSSE1]>; + (MOVLSS2PSrr (V_SET0), FR32:$src)>; def : Pat<(v4f32 (X86vzmovl (v4f32 VR128:$src))), - (MOVLPSrr (V_SET0), VR128:$src)>, Requires<[HasSSE1]>; + (MOVLPSrr (V_SET0), VR128:$src)>; def : Pat<(v4i32 (X86vzmovl (v4i32 VR128:$src))), - (MOVLPSrr (V_SET0), VR128:$src)>, Requires<[HasSSE1]>; + (MOVLPSrr (V_SET0), VR128:$src)>; } // Splat v2f64 / v2i64 @@ -3075,8 +3073,7 @@ def : Pat<(unpckh (v2i64 VR128:$src), (undef)), // Special unary SHUFPSrri case. def : Pat<(v4f32 (pshufd:$src3 VR128:$src1, (undef))), (SHUFPSrri VR128:$src1, VR128:$src1, - (SHUFFLE_get_shuf_imm VR128:$src3))>, - Requires<[HasSSE1]>; + (SHUFFLE_get_shuf_imm VR128:$src3))>; let AddedComplexity = 5 in def : Pat<(v4f32 (pshufd:$src2 VR128:$src1, (undef))), (PSHUFDri VR128:$src1, (SHUFFLE_get_shuf_imm VR128:$src2))>, @@ -3122,13 +3119,13 @@ def : Pat<(v4f32 (unpckl_undef:$src2 VR128:$src, (undef))), } let AddedComplexity = 10 in { def : Pat<(v4f32 (unpckl_undef VR128:$src, (undef))), - (UNPCKLPSrr VR128:$src, VR128:$src)>, Requires<[HasSSE1]>; + (UNPCKLPSrr VR128:$src, VR128:$src)>; def : Pat<(v16i8 (unpckl_undef VR128:$src, (undef))), - (PUNPCKLBWrr VR128:$src, VR128:$src)>, Requires<[HasSSE2]>; + (PUNPCKLBWrr VR128:$src, VR128:$src)>; def : Pat<(v8i16 (unpckl_undef VR128:$src, (undef))), - (PUNPCKLWDrr VR128:$src, VR128:$src)>, Requires<[HasSSE2]>; + (PUNPCKLWDrr VR128:$src, VR128:$src)>; def : Pat<(v4i32 (unpckl_undef VR128:$src, (undef))), - (PUNPCKLDQrr VR128:$src, VR128:$src)>, Requires<[HasSSE2]>; + (PUNPCKLDQrr VR128:$src, VR128:$src)>; } // vector_shuffle v1, <undef>, <2, 2, 3, 3, ...> @@ -3142,13 +3139,13 @@ def : Pat<(v4f32 (unpckh_undef:$src2 VR128:$src, (undef))), } let AddedComplexity = 10 in { def : Pat<(v4f32 (unpckh_undef VR128:$src, (undef))), - (UNPCKHPSrr VR128:$src, VR128:$src)>, Requires<[HasSSE1]>; + (UNPCKHPSrr VR128:$src, VR128:$src)>; def : Pat<(v16i8 (unpckh_undef VR128:$src, (undef))), - (PUNPCKHBWrr VR128:$src, VR128:$src)>, Requires<[HasSSE2]>; + (PUNPCKHBWrr VR128:$src, VR128:$src)>; def : Pat<(v8i16 (unpckh_undef VR128:$src, (undef))), - (PUNPCKHWDrr VR128:$src, VR128:$src)>, Requires<[HasSSE2]>; + (PUNPCKHWDrr VR128:$src, VR128:$src)>; def : Pat<(v4i32 (unpckh_undef VR128:$src, (undef))), - (PUNPCKHDQrr VR128:$src, VR128:$src)>, Requires<[HasSSE2]>; + (PUNPCKHDQrr VR128:$src, VR128:$src)>; } let AddedComplexity = 20 in { @@ -3170,25 +3167,25 @@ def : Pat<(v4i32 (movhlps_undef VR128:$src1, (undef))), let AddedComplexity = 20 in { // vector_shuffle v1, (load v2) <4, 5, 2, 3> using MOVLPS def : Pat<(v4f32 (movlp VR128:$src1, (load addr:$src2))), - (MOVLPSrm VR128:$src1, addr:$src2)>, Requires<[HasSSE1]>; + (MOVLPSrm VR128:$src1, addr:$src2)>; def : Pat<(v2f64 (movlp VR128:$src1, (load addr:$src2))), - (MOVLPDrm VR128:$src1, addr:$src2)>, Requires<[HasSSE2]>; + (MOVLPDrm VR128:$src1, addr:$src2)>; def : Pat<(v4i32 (movlp VR128:$src1, (load addr:$src2))), - (MOVLPSrm VR128:$src1, addr:$src2)>, Requires<[HasSSE2]>; + (MOVLPSrm VR128:$src1, addr:$src2)>; def : Pat<(v2i64 (movlp VR128:$src1, (load addr:$src2))), - (MOVLPDrm VR128:$src1, addr:$src2)>, Requires<[HasSSE2]>; + (MOVLPDrm VR128:$src1, addr:$src2)>; } // (store (vector_shuffle (load addr), v2, <4, 5, 2, 3>), addr) using MOVLPS def : Pat<(store (v4f32 (movlp (load addr:$src1), VR128:$src2)), addr:$src1), - (MOVLPSmr addr:$src1, VR128:$src2)>, Requires<[HasSSE1]>; + (MOVLPSmr addr:$src1, VR128:$src2)>; def : Pat<(store (v2f64 (movlp (load addr:$src1), VR128:$src2)), addr:$src1), - (MOVLPDmr addr:$src1, VR128:$src2)>, Requires<[HasSSE2]>; + (MOVLPDmr addr:$src1, VR128:$src2)>; def : Pat<(store (v4i32 (movlp (bc_v4i32 (loadv2i64 addr:$src1)), VR128:$src2)), addr:$src1), - (MOVLPSmr addr:$src1, VR128:$src2)>, Requires<[HasSSE1]>; + (MOVLPSmr addr:$src1, VR128:$src2)>; def : Pat<(store (v2i64 (movlp (load addr:$src1), VR128:$src2)), addr:$src1), - (MOVLPDmr addr:$src1, VR128:$src2)>, Requires<[HasSSE2]>; + (MOVLPDmr addr:$src1, VR128:$src2)>; let AddedComplexity = 15 in { // Setting the lowest element in the vector. @@ -3208,7 +3205,7 @@ def : Pat<(v4i32 (movlp VR128:$src1, VR128:$src2)), // fall back to this for SSE1) def : Pat<(v4f32 (movlp:$src3 VR128:$src1, (v4f32 VR128:$src2))), (SHUFPSrri VR128:$src2, VR128:$src1, - (SHUFFLE_get_shuf_imm VR128:$src3))>, Requires<[HasSSE1]>; + (SHUFFLE_get_shuf_imm VR128:$src3))>; // Set lowest element and zero upper elements. let AddedComplexity = 15 in @@ -3250,30 +3247,30 @@ def : Pat<(v2i32 (fp_to_sint (v2f64 VR128:$src))), // Use movaps / movups for SSE integer load / store (one byte shorter). def : Pat<(alignedloadv4i32 addr:$src), - (MOVAPSrm addr:$src)>, Requires<[HasSSE1]>; + (MOVAPSrm addr:$src)>; def : Pat<(loadv4i32 addr:$src), - (MOVUPSrm addr:$src)>, Requires<[HasSSE1]>; + (MOVUPSrm addr:$src)>; def : Pat<(alignedloadv2i64 addr:$src), - (MOVAPSrm addr:$src)>, Requires<[HasSSE2]>; + (MOVAPSrm addr:$src)>; def : Pat<(loadv2i64 addr:$src), - (MOVUPSrm addr:$src)>, Requires<[HasSSE2]>; + (MOVUPSrm addr:$src)>; def : Pat<(alignedstore (v2i64 VR128:$src), addr:$dst), - (MOVAPSmr addr:$dst, VR128:$src)>, Requires<[HasSSE2]>; + (MOVAPSmr addr:$dst, VR128:$src)>; def : Pat<(alignedstore (v4i32 VR128:$src), addr:$dst), - (MOVAPSmr addr:$dst, VR128:$src)>, Requires<[HasSSE2]>; + (MOVAPSmr addr:$dst, VR128:$src)>; def : Pat<(alignedstore (v8i16 VR128:$src), addr:$dst), - (MOVAPSmr addr:$dst, VR128:$src)>, Requires<[HasSSE2]>; + (MOVAPSmr addr:$dst, VR128:$src)>; def : Pat<(alignedstore (v16i8 VR128:$src), addr:$dst), - (MOVAPSmr addr:$dst, VR128:$src)>, Requires<[HasSSE2]>; + (MOVAPSmr addr:$dst, VR128:$src)>; def : Pat<(store (v2i64 VR128:$src), addr:$dst), - (MOVUPSmr addr:$dst, VR128:$src)>, Requires<[HasSSE2]>; + (MOVUPSmr addr:$dst, VR128:$src)>; def : Pat<(store (v4i32 VR128:$src), addr:$dst), - (MOVUPSmr addr:$dst, VR128:$src)>, Requires<[HasSSE2]>; + (MOVUPSmr addr:$dst, VR128:$src)>; def : Pat<(store (v8i16 VR128:$src), addr:$dst), - (MOVUPSmr addr:$dst, VR128:$src)>, Requires<[HasSSE2]>; + (MOVUPSmr addr:$dst, VR128:$src)>; def : Pat<(store (v16i8 VR128:$src), addr:$dst), - (MOVUPSmr addr:$dst, VR128:$src)>, Requires<[HasSSE2]>; + (MOVUPSmr addr:$dst, VR128:$src)>; //===----------------------------------------------------------------------===// // SSE4.1 Instructions |
