From 448790d566daf0179ca3cbf25eb3d0b916147418 Mon Sep 17 00:00:00 2001
From: Craig Topper <craig.topper@gmail.com>
Date: Thu, 12 Apr 2012 07:23:00 +0000
Subject: [PATCH] Fix 128-bit ptest intrinsics to take v2i64 instead of v4f32
 since these are integer instructions.

llvm-svn: 154580
---
 include/llvm/IntrinsicsX86.td          |  6 +++---
 lib/Target/X86/X86InstrSSE.td          |  8 ++++----
 test/CodeGen/X86/avx-intrinsics-x86.ll | 18 +++++++++---------
 test/CodeGen/X86/sse41.ll              | 18 +++++++++---------
 4 files changed, 25 insertions(+), 25 deletions(-)

diff --git a/include/llvm/IntrinsicsX86.td b/include/llvm/IntrinsicsX86.td
index f4abba98c08..a6fda4a3afc 100644
--- a/include/llvm/IntrinsicsX86.td
+++ b/include/llvm/IntrinsicsX86.td
@@ -904,13 +904,13 @@ let TargetPrefix = "x86" in {  // All intrinsics start with "llvm.x86.".
 // Test instruction with bitwise comparison.
 let TargetPrefix = "x86" in { // All intrinsics start with "llvm.x86.".
   def int_x86_sse41_ptestz          : GCCBuiltin<"__builtin_ia32_ptestz128">,
-          Intrinsic<[llvm_i32_ty], [llvm_v4f32_ty, llvm_v4f32_ty],
+          Intrinsic<[llvm_i32_ty], [llvm_v2i64_ty, llvm_v2i64_ty],
                     [IntrNoMem]>;
   def int_x86_sse41_ptestc          : GCCBuiltin<"__builtin_ia32_ptestc128">,
-          Intrinsic<[llvm_i32_ty], [llvm_v4f32_ty, llvm_v4f32_ty],
+          Intrinsic<[llvm_i32_ty], [llvm_v2i64_ty, llvm_v2i64_ty],
                     [IntrNoMem]>;
   def int_x86_sse41_ptestnzc        : GCCBuiltin<"__builtin_ia32_ptestnzc128">,
-          Intrinsic<[llvm_i32_ty], [llvm_v4f32_ty, llvm_v4f32_ty],
+          Intrinsic<[llvm_i32_ty], [llvm_v2i64_ty, llvm_v2i64_ty],
                     [IntrNoMem]>;
 }
 
diff --git a/lib/Target/X86/X86InstrSSE.td b/lib/Target/X86/X86InstrSSE.td
index 7741f409db0..408ab16778d 100644
--- a/lib/Target/X86/X86InstrSSE.td
+++ b/lib/Target/X86/X86InstrSSE.td
@@ -6331,11 +6331,11 @@ def : Pat<(f64 (ftrunc FR64:$src)),
 let Defs = [EFLAGS], Predicates = [HasAVX] in {
 def VPTESTrr  : SS48I<0x17, MRMSrcReg, (outs), (ins VR128:$src1, VR128:$src2),
                 "vptest\t{$src2, $src1|$src1, $src2}",
-                [(set EFLAGS, (X86ptest VR128:$src1, (v4f32 VR128:$src2)))]>,
+                [(set EFLAGS, (X86ptest VR128:$src1, (v2i64 VR128:$src2)))]>,
                 OpSize, VEX;
 def VPTESTrm  : SS48I<0x17, MRMSrcMem, (outs), (ins VR128:$src1, f128mem:$src2),
                 "vptest\t{$src2, $src1|$src1, $src2}",
-                [(set EFLAGS,(X86ptest VR128:$src1, (memopv4f32 addr:$src2)))]>,
+                [(set EFLAGS,(X86ptest VR128:$src1, (memopv2i64 addr:$src2)))]>,
                 OpSize, VEX;
 
 def VPTESTYrr : SS48I<0x17, MRMSrcReg, (outs), (ins VR256:$src1, VR256:$src2),
@@ -6351,11 +6351,11 @@ def VPTESTYrm : SS48I<0x17, MRMSrcMem, (outs), (ins VR256:$src1, i256mem:$src2),
 let Defs = [EFLAGS] in {
 def PTESTrr : SS48I<0x17, MRMSrcReg, (outs), (ins VR128:$src1, VR128:$src2),
               "ptest\t{$src2, $src1|$src1, $src2}",
-              [(set EFLAGS, (X86ptest VR128:$src1, (v4f32 VR128:$src2)))]>,
+              [(set EFLAGS, (X86ptest VR128:$src1, (v2i64 VR128:$src2)))]>,
               OpSize;
 def PTESTrm : SS48I<0x17, MRMSrcMem, (outs), (ins VR128:$src1, f128mem:$src2),
               "ptest\t{$src2, $src1|$src1, $src2}",
-              [(set EFLAGS, (X86ptest VR128:$src1, (memopv4f32 addr:$src2)))]>,
+              [(set EFLAGS, (X86ptest VR128:$src1, (memopv2i64 addr:$src2)))]>,
               OpSize;
 }
 
diff --git a/test/CodeGen/X86/avx-intrinsics-x86.ll b/test/CodeGen/X86/avx-intrinsics-x86.ll
index 616601a9438..b33493252a5 100644
--- a/test/CodeGen/X86/avx-intrinsics-x86.ll
+++ b/test/CodeGen/X86/avx-intrinsics-x86.ll
@@ -1078,33 +1078,33 @@ define <2 x i64> @test_x86_sse41_pmuldq(<4 x i32> %a0, <4 x i32> %a1) {
 declare <2 x i64> @llvm.x86.sse41.pmuldq(<4 x i32>, <4 x i32>) nounwind readnone
 
 
-define i32 @test_x86_sse41_ptestc(<4 x float> %a0, <4 x float> %a1) {
+define i32 @test_x86_sse41_ptestc(<2 x i64> %a0, <2 x i64> %a1) {
   ; CHECK: vptest 
   ; CHECK: sbbl
-  %res = call i32 @llvm.x86.sse41.ptestc(<4 x float> %a0, <4 x float> %a1) ; <i32> [#uses=1]
+  %res = call i32 @llvm.x86.sse41.ptestc(<2 x i64> %a0, <2 x i64> %a1) ; <i32> [#uses=1]
   ret i32 %res
 }
-declare i32 @llvm.x86.sse41.ptestc(<4 x float>, <4 x float>) nounwind readnone
+declare i32 @llvm.x86.sse41.ptestc(<2 x i64>, <2 x i64>) nounwind readnone
 
 
-define i32 @test_x86_sse41_ptestnzc(<4 x float> %a0, <4 x float> %a1) {
+define i32 @test_x86_sse41_ptestnzc(<2 x i64> %a0, <2 x i64> %a1) {
   ; CHECK: vptest 
   ; CHECK: seta
   ; CHECK: movzbl
-  %res = call i32 @llvm.x86.sse41.ptestnzc(<4 x float> %a0, <4 x float> %a1) ; <i32> [#uses=1]
+  %res = call i32 @llvm.x86.sse41.ptestnzc(<2 x i64> %a0, <2 x i64> %a1) ; <i32> [#uses=1]
   ret i32 %res
 }
-declare i32 @llvm.x86.sse41.ptestnzc(<4 x float>, <4 x float>) nounwind readnone
+declare i32 @llvm.x86.sse41.ptestnzc(<2 x i64>, <2 x i64>) nounwind readnone
 
 
-define i32 @test_x86_sse41_ptestz(<4 x float> %a0, <4 x float> %a1) {
+define i32 @test_x86_sse41_ptestz(<2 x i64> %a0, <2 x i64> %a1) {
   ; CHECK: vptest 
   ; CHECK: sete
   ; CHECK: movzbl
-  %res = call i32 @llvm.x86.sse41.ptestz(<4 x float> %a0, <4 x float> %a1) ; <i32> [#uses=1]
+  %res = call i32 @llvm.x86.sse41.ptestz(<2 x i64> %a0, <2 x i64> %a1) ; <i32> [#uses=1]
   ret i32 %res
 }
-declare i32 @llvm.x86.sse41.ptestz(<4 x float>, <4 x float>) nounwind readnone
+declare i32 @llvm.x86.sse41.ptestz(<2 x i64>, <2 x i64>) nounwind readnone
 
 
 define <2 x double> @test_x86_sse41_round_pd(<2 x double> %a0) {
diff --git a/test/CodeGen/X86/sse41.ll b/test/CodeGen/X86/sse41.ll
index 2ac4cb435a7..54264b16aea 100644
--- a/test/CodeGen/X86/sse41.ll
+++ b/test/CodeGen/X86/sse41.ll
@@ -183,8 +183,8 @@ define <4 x float> @insertps_3(<4 x float> %t1, <4 x float> %t2) nounwind {
 ; X64:    insertps  $0, %xmm1, %xmm0        
 }
 
-define i32 @ptestz_1(<4 x float> %t1, <4 x float> %t2) nounwind {
-        %tmp1 = call i32 @llvm.x86.sse41.ptestz(<4 x float> %t1, <4 x float> %t2) nounwind readnone
+define i32 @ptestz_1(<2 x i64> %t1, <2 x i64> %t2) nounwind {
+        %tmp1 = call i32 @llvm.x86.sse41.ptestz(<2 x i64> %t1, <2 x i64> %t2) nounwind readnone
         ret i32 %tmp1
 ; X32: _ptestz_1:
 ; X32:    ptest 	%xmm1, %xmm0
@@ -195,8 +195,8 @@ define i32 @ptestz_1(<4 x float> %t1, <4 x float> %t2) nounwind {
 ; X64:    sete	%al
 }
 
-define i32 @ptestz_2(<4 x float> %t1, <4 x float> %t2) nounwind {
-        %tmp1 = call i32 @llvm.x86.sse41.ptestc(<4 x float> %t1, <4 x float> %t2) nounwind readnone
+define i32 @ptestz_2(<2 x i64> %t1, <2 x i64> %t2) nounwind {
+        %tmp1 = call i32 @llvm.x86.sse41.ptestc(<2 x i64> %t1, <2 x i64> %t2) nounwind readnone
         ret i32 %tmp1
 ; X32: _ptestz_2:
 ; X32:    ptest 	%xmm1, %xmm0
@@ -207,8 +207,8 @@ define i32 @ptestz_2(<4 x float> %t1, <4 x float> %t2) nounwind {
 ; X64:    sbbl	%eax
 }
 
-define i32 @ptestz_3(<4 x float> %t1, <4 x float> %t2) nounwind {
-        %tmp1 = call i32 @llvm.x86.sse41.ptestnzc(<4 x float> %t1, <4 x float> %t2) nounwind readnone
+define i32 @ptestz_3(<2 x i64> %t1, <2 x i64> %t2) nounwind {
+        %tmp1 = call i32 @llvm.x86.sse41.ptestnzc(<2 x i64> %t1, <2 x i64> %t2) nounwind readnone
         ret i32 %tmp1
 ; X32: _ptestz_3:
 ; X32:    ptest 	%xmm1, %xmm0
@@ -220,9 +220,9 @@ define i32 @ptestz_3(<4 x float> %t1, <4 x float> %t2) nounwind {
 }
 
 
-declare i32 @llvm.x86.sse41.ptestz(<4 x float>, <4 x float>) nounwind readnone
-declare i32 @llvm.x86.sse41.ptestc(<4 x float>, <4 x float>) nounwind readnone
-declare i32 @llvm.x86.sse41.ptestnzc(<4 x float>, <4 x float>) nounwind readnone
+declare i32 @llvm.x86.sse41.ptestz(<2 x i64>, <2 x i64>) nounwind readnone
+declare i32 @llvm.x86.sse41.ptestc(<2 x i64>, <2 x i64>) nounwind readnone
+declare i32 @llvm.x86.sse41.ptestnzc(<2 x i64>, <2 x i64>) nounwind readnone
 
 ; This used to compile to insertps $0  + insertps $16.  insertps $0 is always
 ; pointless.