[X86] Correct the execution domain on ROUND/VROUND instructions.

llvm-svn: 317968

diff --git a/llvm/lib/Target/X86/X86InstrSSE.td b/llvm/lib/Target/X86/X86InstrSSE.td
index e7986bf..a9bae9b 100644 (file)
--- a/llvm/lib/Target/X86/X86InstrSSE.td
+++ b/llvm/lib/Target/X86/X86InstrSSE.td
@@ -5829,7 +5829,7 @@ let ExeDomain = SSEPackedDouble in {
 
 multiclass avx_fp_unop_rm<bits<8> opcss, bits<8> opcsd,
                           string OpcodeStr> {
-let ExeDomain = GenericDomain, hasSideEffects = 0 in {
+let ExeDomain = SSEPackedSingle, hasSideEffects = 0 in {
   def SSr : SS4AIi8<opcss, MRMSrcReg,
         (outs FR32:$dst), (ins FR32:$src1, FR32:$src2, i32u8imm:$src3),
         !strconcat(OpcodeStr,
@@ -5842,7 +5842,9 @@ let ExeDomain = GenericDomain, hasSideEffects = 0 in {
         !strconcat(OpcodeStr,
              "ss\t{$src3, $src2, $src1, $dst|$dst, $src1, $src2, $src3}"),
         []>, Sched<[WriteFAddLd, ReadAfterLd]>;
+} // ExeDomain = SSEPackedSingle, hasSideEffects = 0
 
+let ExeDomain = SSEPackedDouble, hasSideEffects = 0 in {
   def SDr : SS4AIi8<opcsd, MRMSrcReg,
         (outs FR64:$dst), (ins FR64:$src1, FR64:$src2, i32u8imm:$src3),
         !strconcat(OpcodeStr,
@@ -5855,12 +5857,12 @@ let ExeDomain = GenericDomain, hasSideEffects = 0 in {
         !strconcat(OpcodeStr,
              "sd\t{$src3, $src2, $src1, $dst|$dst, $src1, $src2, $src3}"),
         []>, Sched<[WriteFAddLd, ReadAfterLd]>;
-} // ExeDomain = GenericDomain, hasSideEffects = 0
+} // ExeDomain = SSEPackedDouble, hasSideEffects = 0
 }
 
 multiclass sse41_fp_unop_s<bits<8> opcss, bits<8> opcsd,
                            string OpcodeStr> {
-let ExeDomain = GenericDomain, hasSideEffects = 0 in {
+let ExeDomain = SSEPackedSingle, hasSideEffects = 0 in {
   def SSr : SS4AIi8<opcss, MRMSrcReg,
                     (outs FR32:$dst), (ins FR32:$src1, i32u8imm:$src2),
                     !strconcat(OpcodeStr,
@@ -5873,7 +5875,9 @@ let ExeDomain = GenericDomain, hasSideEffects = 0 in {
                     !strconcat(OpcodeStr,
                                "ss\t{$src2, $src1, $dst|$dst, $src1, $src2}"),
                     []>, Sched<[WriteFAddLd, ReadAfterLd]>;
+} // ExeDomain = SSEPackedSingle, hasSideEffects = 0
 
+let ExeDomain = SSEPackedDouble, hasSideEffects = 0 in {
   def SDr : SS4AIi8<opcsd, MRMSrcReg,
                     (outs FR64:$dst), (ins FR64:$src1, i32u8imm:$src2),
                     !strconcat(OpcodeStr,
@@ -5886,14 +5890,14 @@ let ExeDomain = GenericDomain, hasSideEffects = 0 in {
                     !strconcat(OpcodeStr,
                                "sd\t{$src2, $src1, $dst|$dst, $src1, $src2}"),
                     []>, Sched<[WriteFAddLd, ReadAfterLd]>;
-} // ExeDomain = GenericDomain, hasSideEffects = 0
+} // ExeDomain = SSEPackedDouble, hasSideEffects = 0
 }
 
 multiclass sse41_fp_binop_s<bits<8> opcss, bits<8> opcsd,
                             string OpcodeStr,
                             Intrinsic F32Int,
                             Intrinsic F64Int, bit Is2Addr = 1> {
-let ExeDomain = GenericDomain, isCodeGenOnly = 1 in {
+let ExeDomain = SSEPackedSingle, isCodeGenOnly = 1 in {
   def SSr_Int : SS4AIi8<opcss, MRMSrcReg,
         (outs VR128:$dst), (ins VR128:$src1, VR128:$src2, i32u8imm:$src3),
         !if(Is2Addr,
@@ -5914,7 +5918,9 @@ let ExeDomain = GenericDomain, isCodeGenOnly = 1 in {
         [(set VR128:$dst,
              (F32Int VR128:$src1, sse_load_f32:$src2, imm:$src3))]>,
         Sched<[WriteFAddLd, ReadAfterLd]>;
+} // ExeDomain = SSEPackedSingle, isCodeGenOnly = 1
 
+let ExeDomain = SSEPackedDouble, isCodeGenOnly = 1 in {
   def SDr_Int : SS4AIi8<opcsd, MRMSrcReg,
         (outs VR128:$dst), (ins VR128:$src1, VR128:$src2, i32u8imm:$src3),
         !if(Is2Addr,
@@ -5935,7 +5941,7 @@ let ExeDomain = GenericDomain, isCodeGenOnly = 1 in {
         [(set VR128:$dst,
               (F64Int VR128:$src1, sse_load_f64:$src2, imm:$src3))]>,
         Sched<[WriteFAddLd, ReadAfterLd]>;
-} // ExeDomain = GenericDomain, isCodeGenOnly = 1
+} // ExeDomain = SSEPackedDouble, isCodeGenOnly = 1
 }
 
 // FP round - roundss, roundps, roundsd, roundpd

diff --git a/llvm/test/CodeGen/X86/sse41-schedule.ll b/llvm/test/CodeGen/X86/sse41-schedule.ll
index 76242c8..f86388f 100644 (file)
--- a/llvm/test/CodeGen/X86/sse41-schedule.ll
+++ b/llvm/test/CodeGen/X86/sse41-schedule.ll
@@ -3107,7 +3107,7 @@ declare <4 x float> @llvm.x86.sse41.round.ps(<4 x float>, i32) nounwind readnone
 define <2 x double> @test_roundsd(<2 x double> %a0, <2 x double> %a1, <2 x double> *%a2) {
 ; GENERIC-LABEL: test_roundsd:
 ; GENERIC:       # BB#0:
-; GENERIC-NEXT:    movaps %xmm0, %xmm2 # sched: [1:1.00]
+; GENERIC-NEXT:    movapd %xmm0, %xmm2 # sched: [1:1.00]
 ; GENERIC-NEXT:    roundsd $7, %xmm1, %xmm2 # sched: [3:1.00]
 ; GENERIC-NEXT:    roundsd $7, (%rdi), %xmm0 # sched: [9:1.00]
 ; GENERIC-NEXT:    addpd %xmm2, %xmm0 # sched: [3:1.00]
@@ -3115,7 +3115,7 @@ define <2 x double> @test_roundsd(<2 x double> %a0, <2 x double> %a1, <2 x doubl
 ;
 ; SLM-LABEL: test_roundsd:
 ; SLM:       # BB#0:
-; SLM-NEXT:    movaps %xmm0, %xmm2 # sched: [1:1.00]
+; SLM-NEXT:    movapd %xmm0, %xmm2 # sched: [1:1.00]
 ; SLM-NEXT:    roundsd $7, (%rdi), %xmm0 # sched: [6:1.00]
 ; SLM-NEXT:    roundsd $7, %xmm1, %xmm2 # sched: [3:1.00]
 ; SLM-NEXT:    addpd %xmm2, %xmm0 # sched: [3:1.00]