Fix PR5315, original patch by Nicolas Capens!

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86203 91177308-0d34-0410-b5e6-96231b3b80d8

diff --git a/include/llvm/IntrinsicsX86.td b/include/llvm/IntrinsicsX86.td
index 5be032bb82046663268559b999cc16a7c3e9acc6..794f4bfe816f2b6a2e7a934cfc6870e572327d2b 100644 (file)
--- a/include/llvm/IntrinsicsX86.td
+++ b/include/llvm/IntrinsicsX86.td
@@ -484,13 +484,13 @@ let TargetPrefix = "x86" in {  // All intrinsics start with "llvm.x86.".
 // Misc.
 let TargetPrefix = "x86" in {  // All intrinsics start with "llvm.x86.".
   def int_x86_sse2_packsswb_128 : GCCBuiltin<"__builtin_ia32_packsswb128">,
-              Intrinsic<[llvm_v8i16_ty], [llvm_v8i16_ty,
+              Intrinsic<[llvm_v16i8_ty], [llvm_v8i16_ty,
                          llvm_v8i16_ty], [IntrNoMem]>;
   def int_x86_sse2_packssdw_128 : GCCBuiltin<"__builtin_ia32_packssdw128">,
-              Intrinsic<[llvm_v4i32_ty], [llvm_v4i32_ty,
+              Intrinsic<[llvm_v8i16_ty], [llvm_v4i32_ty,
                          llvm_v4i32_ty], [IntrNoMem]>;
   def int_x86_sse2_packuswb_128 : GCCBuiltin<"__builtin_ia32_packuswb128">,
-              Intrinsic<[llvm_v8i16_ty], [llvm_v8i16_ty,
+              Intrinsic<[llvm_v16i8_ty], [llvm_v8i16_ty,
                          llvm_v8i16_ty], [IntrNoMem]>;
   def int_x86_sse2_movmsk_pd : GCCBuiltin<"__builtin_ia32_movmskpd">,
               Intrinsic<[llvm_i32_ty], [llvm_v2f64_ty], [IntrNoMem]>;

diff --git a/test/CodeGen/X86/2006-04-04-CrossBlockCrash.ll b/test/CodeGen/X86/2006-04-04-CrossBlockCrash.ll
index c106f57e93843a0c4741c1acdd112a4dce349796..3f67097ddc0d320e8413249714c359ca34c1facb 100644 (file)
--- a/test/CodeGen/X86/2006-04-04-CrossBlockCrash.ll
+++ b/test/CodeGen/X86/2006-04-04-CrossBlockCrash.ll
@@ -11,7 +11,7 @@ target triple = "i686-apple-darwin8.6.1"
 
 declare <4 x float> @llvm.x86.sse.cmp.ps(<4 x float>, <4 x float>, i8)
 
-declare <4 x i32> @llvm.x86.sse2.packssdw.128(<4 x i32>, <4 x i32>)
+declare <8 x i16> @llvm.x86.sse2.packssdw.128(<4 x i32>, <4 x i32>)
 
 declare i32 @llvm.x86.sse2.pmovmskb.128(<16 x i8>)
 
@@ -33,8 +33,8 @@ cond_false183:                ; preds = %cond_false, %entry
        %tmp337 = bitcast <4 x i32> %tmp336 to <4 x float>              ; <<4 x float>> [#uses=1]
        %tmp378 = tail call <4 x float> @llvm.x86.sse.cmp.ps( <4 x float> %tmp337, <4 x float> zeroinitializer, i8 1 )          ; <<4 x float>> [#uses=1]
        %tmp379 = bitcast <4 x float> %tmp378 to <4 x i32>              ; <<4 x i32>> [#uses=1]
-       %tmp388 = tail call <4 x i32> @llvm.x86.sse2.packssdw.128( <4 x i32> zeroinitializer, <4 x i32> %tmp379 )               ; <<4 x i32>> [#uses=1]
-       %tmp392 = bitcast <4 x i32> %tmp388 to <8 x i16>                ; <<8 x i16>> [#uses=1]
+       %tmp388 = tail call <8 x i16> @llvm.x86.sse2.packssdw.128( <4 x i32> zeroinitializer, <4 x i32> %tmp379 )               ; <<4 x i32>> [#uses=1]
+       %tmp392 = bitcast <8 x i16> %tmp388 to <8 x i16>                ; <<8 x i16>> [#uses=1]
        %tmp399 = extractelement <8 x i16> %tmp392, i32 7               ; <i16> [#uses=1]
        %tmp423 = insertelement <8 x i16> zeroinitializer, i16 %tmp399, i32 7           ; <<8 x i16>> [#uses=1]
        %tmp427 = bitcast <8 x i16> %tmp423 to <16 x i8>                ; <<16 x i8>> [#uses=1]

diff --git a/test/CodeGen/X86/2006-05-01-SchedCausingSpills.ll b/test/CodeGen/X86/2006-05-01-SchedCausingSpills.ll
index 49f3a95705ad032ec4497cb9dd97010d1e2a3362..b0453299669e2680f90528dfe151789109bfa472 100644 (file)
--- a/test/CodeGen/X86/2006-05-01-SchedCausingSpills.ll
+++ b/test/CodeGen/X86/2006-05-01-SchedCausingSpills.ll
@@ -17,8 +17,8 @@ define i32 @foo(<4 x float>* %a, <4 x float>* %b, <4 x float>* %c, <4 x float>*
        %tmp75 = bitcast <4 x float> %tmp74 to <4 x i32>                ; <<4 x i32>> [#uses=1]
        %tmp88 = tail call <4 x float> @llvm.x86.sse.cmp.ps( <4 x float> %tmp44, <4 x float> %tmp61, i8 1 )             ; <<4 x float>> [#uses=1]
        %tmp89 = bitcast <4 x float> %tmp88 to <4 x i32>                ; <<4 x i32>> [#uses=1]
-       %tmp98 = tail call <4 x i32> @llvm.x86.sse2.packssdw.128( <4 x i32> %tmp75, <4 x i32> %tmp89 )          ; <<4 x i32>> [#uses=1]
-       %tmp102 = bitcast <4 x i32> %tmp98 to <8 x i16>         ; <<8 x i16>> [#uses=1]
+       %tmp98 = tail call <8 x i16> @llvm.x86.sse2.packssdw.128( <4 x i32> %tmp75, <4 x i32> %tmp89 )          ; <<4 x i32>> [#uses=1]
+       %tmp102 = bitcast <8 x i16> %tmp98 to <8 x i16>         ; <<8 x i16>> [#uses=1]
        %tmp.upgrd.1 = shufflevector <8 x i16> %tmp102, <8 x i16> undef, <8 x i32> < i32 0, i32 1, i32 2, i32 3, i32 6, i32 5, i32 4, i32 7 >           ; <<8 x i16>> [#uses=1]
        %tmp105 = shufflevector <8 x i16> %tmp.upgrd.1, <8 x i16> undef, <8 x i32> < i32 2, i32 1, i32 0, i32 3, i32 4, i32 5, i32 6, i32 7 >           ; <<8 x i16>> [#uses=1]
        %tmp105.upgrd.2 = bitcast <8 x i16> %tmp105 to <4 x float>              ; <<4 x float>> [#uses=1]
@@ -32,8 +32,8 @@ define i32 @foo(<4 x float>* %a, <4 x float>* %b, <4 x float>* %c, <4 x float>*
        %tmp134 = bitcast <4 x float> %tmp133 to <4 x i32>              ; <<4 x i32>> [#uses=1]
        %tmp147 = tail call <4 x float> @llvm.x86.sse.cmp.ps( <4 x float> %tmp44, <4 x float> %tmp120, i8 1 )           ; <<4 x float>> [#uses=1]
        %tmp148 = bitcast <4 x float> %tmp147 to <4 x i32>              ; <<4 x i32>> [#uses=1]
-       %tmp159 = tail call <4 x i32> @llvm.x86.sse2.packssdw.128( <4 x i32> %tmp134, <4 x i32> %tmp148 )               ; <<4 x i32>> [#uses=1]
-       %tmp163 = bitcast <4 x i32> %tmp159 to <8 x i16>                ; <<8 x i16>> [#uses=1]
+       %tmp159 = tail call <8 x i16> @llvm.x86.sse2.packssdw.128( <4 x i32> %tmp134, <4 x i32> %tmp148 )               ; <<4 x i32>> [#uses=1]
+       %tmp163 = bitcast <8 x i16> %tmp159 to <8 x i16>                ; <<8 x i16>> [#uses=1]
        %tmp164 = shufflevector <8 x i16> %tmp163, <8 x i16> undef, <8 x i32> < i32 0, i32 1, i32 2, i32 3, i32 6, i32 5, i32 4, i32 7 >                ; <<8 x i16>> [#uses=1]
        %tmp166 = shufflevector <8 x i16> %tmp164, <8 x i16> undef, <8 x i32> < i32 2, i32 1, i32 0, i32 3, i32 4, i32 5, i32 6, i32 7 >                ; <<8 x i16>> [#uses=1]
        %tmp166.upgrd.4 = bitcast <8 x i16> %tmp166 to <4 x float>              ; <<4 x float>> [#uses=1]
@@ -47,8 +47,8 @@ define i32 @foo(<4 x float>* %a, <4 x float>* %b, <4 x float>* %c, <4 x float>*
        %tmp195 = bitcast <4 x float> %tmp194 to <4 x i32>              ; <<4 x i32>> [#uses=1]
        %tmp208 = tail call <4 x float> @llvm.x86.sse.cmp.ps( <4 x float> %tmp44, <4 x float> %tmp181, i8 1 )           ; <<4 x float>> [#uses=1]
        %tmp209 = bitcast <4 x float> %tmp208 to <4 x i32>              ; <<4 x i32>> [#uses=1]
-       %tmp220 = tail call <4 x i32> @llvm.x86.sse2.packssdw.128( <4 x i32> %tmp195, <4 x i32> %tmp209 )               ; <<4 x i32>> [#uses=1]
-       %tmp224 = bitcast <4 x i32> %tmp220 to <8 x i16>                ; <<8 x i16>> [#uses=1]
+       %tmp220 = tail call <8 x i16> @llvm.x86.sse2.packssdw.128( <4 x i32> %tmp195, <4 x i32> %tmp209 )               ; <<4 x i32>> [#uses=1]
+       %tmp224 = bitcast <8 x i16> %tmp220 to <8 x i16>                ; <<8 x i16>> [#uses=1]
        %tmp225 = shufflevector <8 x i16> %tmp224, <8 x i16> undef, <8 x i32> < i32 0, i32 1, i32 2, i32 3, i32 6, i32 5, i32 4, i32 7 >                ; <<8 x i16>> [#uses=1]
        %tmp227 = shufflevector <8 x i16> %tmp225, <8 x i16> undef, <8 x i32> < i32 2, i32 1, i32 0, i32 3, i32 4, i32 5, i32 6, i32 7 >                ; <<8 x i16>> [#uses=1]
        %tmp227.upgrd.6 = bitcast <8 x i16> %tmp227 to <4 x float>              ; <<4 x float>> [#uses=1]
@@ -62,8 +62,8 @@ define i32 @foo(<4 x float>* %a, <4 x float>* %b, <4 x float>* %c, <4 x float>*
        %tmp256 = bitcast <4 x float> %tmp255 to <4 x i32>              ; <<4 x i32>> [#uses=1]
        %tmp269 = tail call <4 x float> @llvm.x86.sse.cmp.ps( <4 x float> %tmp44, <4 x float> %tmp242, i8 1 )           ; <<4 x float>> [#uses=1]
        %tmp270 = bitcast <4 x float> %tmp269 to <4 x i32>              ; <<4 x i32>> [#uses=1]
-       %tmp281 = tail call <4 x i32> @llvm.x86.sse2.packssdw.128( <4 x i32> %tmp256, <4 x i32> %tmp270 )               ; <<4 x i32>> [#uses=1]
-       %tmp285 = bitcast <4 x i32> %tmp281 to <8 x i16>                ; <<8 x i16>> [#uses=1]
+       %tmp281 = tail call <8 x i16> @llvm.x86.sse2.packssdw.128( <4 x i32> %tmp256, <4 x i32> %tmp270 )               ; <<4 x i32>> [#uses=1]
+       %tmp285 = bitcast <8 x i16> %tmp281 to <8 x i16>                ; <<8 x i16>> [#uses=1]
        %tmp286 = shufflevector <8 x i16> %tmp285, <8 x i16> undef, <8 x i32> < i32 0, i32 1, i32 2, i32 3, i32 6, i32 5, i32 4, i32 7 >                ; <<8 x i16>> [#uses=1]
        %tmp288 = shufflevector <8 x i16> %tmp286, <8 x i16> undef, <8 x i32> < i32 2, i32 1, i32 0, i32 3, i32 4, i32 5, i32 6, i32 7 >                ; <<8 x i16>> [#uses=1]
        %tmp288.upgrd.8 = bitcast <8 x i16> %tmp288 to <4 x float>              ; <<4 x float>> [#uses=1]
@@ -73,4 +73,4 @@ define i32 @foo(<4 x float>* %a, <4 x float>* %b, <4 x float>* %c, <4 x float>*
 
 declare <4 x float> @llvm.x86.sse.cmp.ps(<4 x float>, <4 x float>, i8)
 
-declare <4 x i32> @llvm.x86.sse2.packssdw.128(<4 x i32>, <4 x i32>)
+declare <8 x i16> @llvm.x86.sse2.packssdw.128(<4 x i32>, <4 x i32>)

diff --git a/test/CodeGen/X86/2007-05-17-ShuffleISelBug.ll b/test/CodeGen/X86/2007-05-17-ShuffleISelBug.ll
index 989dfc5bdb2cd530aa0a48e0992eb5c9f0df7999..b27ef836960fcaca563e9c3655342072f83c3b1b 100644 (file)
--- a/test/CodeGen/X86/2007-05-17-ShuffleISelBug.ll
+++ b/test/CodeGen/X86/2007-05-17-ShuffleISelBug.ll
@@ -1,7 +1,7 @@
 ; RUN: llc < %s -march=x86 -mattr=+sse2
 ; RUN: llc < %s -march=x86 -mattr=+sse2 | not grep punpckhwd
 
-declare <8 x i16> @llvm.x86.sse2.packuswb.128(<8 x i16>, <8 x i16>)
+declare <16 x i8> @llvm.x86.sse2.packuswb.128(<8 x i16>, <8 x i16>)
 
 declare <8 x i16> @llvm.x86.sse2.psrl.w(<8 x i16>, <8 x i16>)
 
@@ -13,8 +13,8 @@ define fastcc void @test(i32* %src, i32 %sbpr, i32* %dst, i32 %dbpr, i32 %w, i32
        %tmp805 = add <4 x i32> %tmp777, zeroinitializer
        %tmp832 = bitcast <4 x i32> %tmp805 to <8 x i16>
        %tmp838 = tail call <8 x i16> @llvm.x86.sse2.psrl.w( <8 x i16> %tmp832, <8 x i16> < i16 8, i16 undef, i16 undef, i16 undef, i16 undef, i16 undef, i16 undef, i16 undef > )
-       %tmp1020 = tail call <8 x i16> @llvm.x86.sse2.packuswb.128( <8 x i16> zeroinitializer, <8 x i16> %tmp838 )
-       %tmp1030 = bitcast <8 x i16> %tmp1020 to <4 x i32>
+       %tmp1020 = tail call <16 x i8> @llvm.x86.sse2.packuswb.128( <8 x i16> zeroinitializer, <8 x i16> %tmp838 )
+       %tmp1030 = bitcast <16 x i8> %tmp1020 to <4 x i32>
        %tmp1033 = add <4 x i32> zeroinitializer, %tmp1030
        %tmp1048 = bitcast <4 x i32> %tmp1033 to <2 x i64>
        %tmp1049 = or <2 x i64> %tmp1048, zeroinitializer

diff --git a/test/CodeGen/X86/vec_zero-2.ll b/test/CodeGen/X86/vec_zero-2.ll
index e42b5384b08175786a22b3efad596133b329347f..cdb030eb38bfebf6e057b4b44167c9a1feb7a00c 100644 (file)
--- a/test/CodeGen/X86/vec_zero-2.ll
+++ b/test/CodeGen/X86/vec_zero-2.ll
@@ -12,8 +12,8 @@ bb4743:               ; preds = %bb1664
        %tmp5257 = sub <8 x i16> %tmp5256, zeroinitializer              ; <<8 x i16>> [#uses=1]
        %tmp5258 = bitcast <8 x i16> %tmp5257 to <2 x i64>              ; <<2 x i64>> [#uses=1]
        %tmp5265 = bitcast <2 x i64> %tmp5258 to <8 x i16>              ; <<8 x i16>> [#uses=1]
-       %tmp5266 = call <8 x i16> @llvm.x86.sse2.packuswb.128( <8 x i16> %tmp5265, <8 x i16> zeroinitializer ) nounwind readnone                ; <<8 x i16>> [#uses=1]
-       %tmp5267 = bitcast <8 x i16> %tmp5266 to <2 x i64>              ; <<2 x i64>> [#uses=1]
+       %tmp5266 = call <16 x i8> @llvm.x86.sse2.packuswb.128( <8 x i16> %tmp5265, <8 x i16> zeroinitializer ) nounwind readnone                ; <<8 x i16>> [#uses=1]
+       %tmp5267 = bitcast <16 x i8> %tmp5266 to <2 x i64>              ; <<2 x i64>> [#uses=1]
        %tmp5294 = and <2 x i64> zeroinitializer, %tmp5267              ; <<2 x i64>> [#uses=1]
        br label %bb5310
 bb5310:                ; preds = %bb4743, %bb1664
@@ -21,4 +21,4 @@ bb5310:               ; preds = %bb4743, %bb1664
        ret i32 0
 }
 
-declare <8 x i16> @llvm.x86.sse2.packuswb.128(<8 x i16>, <8 x i16>) nounwind readnone 
+declare <16 x i8> @llvm.x86.sse2.packuswb.128(<8 x i16>, <8 x i16>) nounwind readnone