summaryrefslogtreecommitdiff
path: root/test/Transforms
diff options
context:
space:
mode:
Diffstat (limited to 'test/Transforms')
-rw-r--r--test/Transforms/EliminateAvailableExternally/visibility.ll11
-rw-r--r--test/Transforms/GVN/pre-new-inst.ll29
-rw-r--r--test/Transforms/IndVarSimplify/lrev-existing-umin.ll36
-rw-r--r--test/Transforms/Inline/frameescape.ll12
-rw-r--r--test/Transforms/InstCombine/align-external.ll26
-rw-r--r--test/Transforms/InstCombine/intrinsics.ll22
-rw-r--r--test/Transforms/InstCombine/load-combine-metadata.ll29
-rw-r--r--test/Transforms/InstCombine/load_combine_aa.ll15
-rw-r--r--test/Transforms/InstSimplify/2011-09-05-InsertExtractValue.ll9
-rw-r--r--test/Transforms/InstSimplify/floating-point-compare.ll15
-rw-r--r--test/Transforms/InstSimplify/undef.ll14
-rw-r--r--test/Transforms/LICM/PR24013.ll19
-rw-r--r--test/Transforms/LoopDistribute/basic-with-memchecks.ll8
-rw-r--r--test/Transforms/LoopIdiom/ctpop-multiple-users-crash.ll34
-rw-r--r--test/Transforms/LoopRotate/oz-disable.ll30
-rw-r--r--test/Transforms/LoopStrengthReduce/ephemeral.ll41
-rw-r--r--test/Transforms/LoopUnroll/unroll-pragmas.ll27
-rw-r--r--test/Transforms/LoopVectorize/X86/vectorization-remarks.ll4
-rw-r--r--test/Transforms/PlaceSafepoints/statepoint-frameescape.ll8
-rw-r--r--test/Transforms/SLPVectorizer/AMDGPU/simplebb.ll5
-rw-r--r--test/Transforms/SLPVectorizer/X86/cse.ll31
-rw-r--r--test/Transforms/SLPVectorizer/X86/gep.ll1
-rw-r--r--test/Transforms/SLPVectorizer/X86/loopinvariant.ll18
-rw-r--r--test/Transforms/SLPVectorizer/X86/pr19657.ll102
-rw-r--r--test/Transforms/SROA/basictest.ll11
25 files changed, 427 insertions, 130 deletions
diff --git a/test/Transforms/EliminateAvailableExternally/visibility.ll b/test/Transforms/EliminateAvailableExternally/visibility.ll
new file mode 100644
index 0000000000000..9966fcf30e85d
--- /dev/null
+++ b/test/Transforms/EliminateAvailableExternally/visibility.ll
@@ -0,0 +1,11 @@
+; RUN: opt -elim-avail-extern -S < %s | FileCheck %s
+
+; CHECK: declare hidden void @f()
+define available_externally hidden void @f() {
+ ret void
+}
+
+define void @g() {
+ call void @f()
+ ret void
+}
diff --git a/test/Transforms/GVN/pre-new-inst.ll b/test/Transforms/GVN/pre-new-inst.ll
new file mode 100644
index 0000000000000..238b8a687cccc
--- /dev/null
+++ b/test/Transforms/GVN/pre-new-inst.ll
@@ -0,0 +1,29 @@
+; RUN: opt -basicaa -gvn -S %s | FileCheck %s
+
+%MyStruct = type { i32, i32 }
+define i8 @foo(i64 %in, i8* %arr) {
+ %addr = alloca %MyStruct
+ %dead = trunc i64 %in to i32
+ br i1 undef, label %next, label %tmp
+
+tmp:
+ call void @bar()
+ br label %next
+
+next:
+ %addr64 = bitcast %MyStruct* %addr to i64*
+ store i64 %in, i64* %addr64
+ br label %final
+
+final:
+ %addr32 = getelementptr %MyStruct, %MyStruct* %addr, i32 0, i32 0
+ %idx32 = load i32, i32* %addr32
+
+; CHECK: %resptr = getelementptr i8, i8* %arr, i32 %dead
+ %resptr = getelementptr i8, i8* %arr, i32 %idx32
+ %res = load i8, i8* %resptr
+
+ ret i8 %res
+}
+
+declare void @bar()
diff --git a/test/Transforms/IndVarSimplify/lrev-existing-umin.ll b/test/Transforms/IndVarSimplify/lrev-existing-umin.ll
new file mode 100644
index 0000000000000..961c9fd944d93
--- /dev/null
+++ b/test/Transforms/IndVarSimplify/lrev-existing-umin.ll
@@ -0,0 +1,36 @@
+; RUN: opt -S -indvars < %s | FileCheck %s
+
+define void @f(i32 %length.i.88, i32 %length.i, i8* %tmp12, i32 %tmp10, i8* %tmp8) {
+; CHECK-LABEL: @f(
+not_zero11.preheader:
+ %tmp13 = icmp ugt i32 %length.i, %length.i.88
+ %tmp14 = select i1 %tmp13, i32 %length.i.88, i32 %length.i
+ %tmp15 = icmp sgt i32 %tmp14, 0
+ br i1 %tmp15, label %not_zero11, label %not_zero11.postloop
+
+not_zero11:
+ %v_1 = phi i32 [ %tmp22, %not_zero11 ], [ 0, %not_zero11.preheader ]
+ %tmp16 = zext i32 %v_1 to i64
+ %tmp17 = getelementptr inbounds i8, i8* %tmp8, i64 %tmp16
+ %tmp18 = load i8, i8* %tmp17, align 1
+ %tmp19 = zext i8 %tmp18 to i32
+ %tmp20 = or i32 %tmp19, %tmp10
+ %tmp21 = trunc i32 %tmp20 to i8
+ %addr22 = getelementptr inbounds i8, i8* %tmp12, i64 %tmp16
+ store i8 %tmp21, i8* %addr22, align 1
+ %tmp22 = add nuw nsw i32 %v_1, 1
+ %tmp23 = icmp slt i32 %tmp22, %tmp14
+ br i1 %tmp23, label %not_zero11, label %main.exit.selector
+
+main.exit.selector:
+; CHECK-LABEL: main.exit.selector:
+; CHECK: %tmp24 = icmp slt i32 %tmp14, %length.i
+ %tmp24 = icmp slt i32 %tmp22, %length.i
+ br i1 %tmp24, label %not_zero11.postloop, label %leave
+
+leave:
+ ret void
+
+not_zero11.postloop:
+ ret void
+}
diff --git a/test/Transforms/Inline/frameescape.ll b/test/Transforms/Inline/frameescape.ll
index fb336024f9375..6615fe9a76e4d 100644
--- a/test/Transforms/Inline/frameescape.ll
+++ b/test/Transforms/Inline/frameescape.ll
@@ -1,13 +1,13 @@
; RUN: opt -inline -S < %s | FileCheck %s
-; PR23216: We can't inline functions using llvm.frameescape.
+; PR23216: We can't inline functions using llvm.localescape.
-declare void @llvm.frameescape(...)
+declare void @llvm.localescape(...)
declare i8* @llvm.frameaddress(i32)
-declare i8* @llvm.framerecover(i8*, i8*, i32)
+declare i8* @llvm.localrecover(i8*, i8*, i32)
define internal void @foo(i8* %fp) {
- %a.i8 = call i8* @llvm.framerecover(i8* bitcast (i32 ()* @bar to i8*), i8* %fp, i32 0)
+ %a.i8 = call i8* @llvm.localrecover(i8* bitcast (i32 ()* @bar to i8*), i8* %fp, i32 0)
%a = bitcast i8* %a.i8 to i32*
store i32 42, i32* %a
ret void
@@ -16,7 +16,7 @@ define internal void @foo(i8* %fp) {
define internal i32 @bar() {
entry:
%a = alloca i32
- call void (...) @llvm.frameescape(i32* %a)
+ call void (...) @llvm.localescape(i32* %a)
%fp = call i8* @llvm.frameaddress(i32 0)
tail call void @foo(i8* %fp)
%r = load i32, i32* %a
@@ -27,7 +27,7 @@ entry:
define internal i32 @bar_alwaysinline() alwaysinline {
entry:
%a = alloca i32
- call void (...) @llvm.frameescape(i32* %a)
+ call void (...) @llvm.localescape(i32* %a)
tail call void @foo(i8* null)
ret i32 0
}
diff --git a/test/Transforms/InstCombine/align-external.ll b/test/Transforms/InstCombine/align-external.ll
index ee98a01201795..15f3096105bbd 100644
--- a/test/Transforms/InstCombine/align-external.ll
+++ b/test/Transforms/InstCombine/align-external.ll
@@ -3,16 +3,14 @@
; Don't assume that external global variables or those with weak linkage have
; their preferred alignment. They may only have the ABI minimum alignment.
-; CHECK: %s = shl i64 %a, 3
-; CHECK: %r = or i64 %s, ptrtoint (i32* @A to i64)
-; CHECK: %q = add i64 %r, 1
-; CHECK: ret i64 %q
-
target datalayout = "i32:8:32"
@A = external global i32
@B = weak_odr global i32 0
+@C = available_externally global <4 x i32> zeroinitializer, align 4
+; CHECK: @C = available_externally global <4 x i32> zeroinitializer, align 4
+
define i64 @foo(i64 %a) {
%t = ptrtoint i32* @A to i64
%s = shl i64 %a, 3
@@ -21,9 +19,23 @@ define i64 @foo(i64 %a) {
ret i64 %q
}
+; CHECK-LABEL: define i64 @foo(i64 %a)
+; CHECK: %s = shl i64 %a, 3
+; CHECK: %r = or i64 %s, ptrtoint (i32* @A to i64)
+; CHECK: %q = add i64 %r, 1
+; CHECK: ret i64 %q
+
define i32 @bar() {
-; CHECK-LABEL: @bar(
%r = load i32, i32* @B, align 1
-; CHECK: align 1
ret i32 %r
}
+
+; CHECK-LABEL: @bar()
+; CHECK: align 1
+
+define void @vec_store() {
+ store <4 x i32> <i32 0, i32 1, i32 2, i32 3>, <4 x i32>* @C, align 4
+ ret void
+}
+; CHECK: define void @vec_store()
+; CHECK: store <4 x i32> <i32 0, i32 1, i32 2, i32 3>, <4 x i32>* @C, align 4
diff --git a/test/Transforms/InstCombine/intrinsics.ll b/test/Transforms/InstCombine/intrinsics.ll
index 9767704c85cfd..bea063787a751 100644
--- a/test/Transforms/InstCombine/intrinsics.ll
+++ b/test/Transforms/InstCombine/intrinsics.ll
@@ -17,6 +17,8 @@ declare i32 @llvm.cttz.i32(i32, i1) nounwind readnone
declare i32 @llvm.ctlz.i32(i32, i1) nounwind readnone
declare i32 @llvm.ctpop.i32(i32) nounwind readnone
declare i8 @llvm.ctlz.i8(i8, i1) nounwind readnone
+declare double @llvm.cos.f64(double %Val) nounwind readonly
+declare double @llvm.sin.f64(double %Val) nounwind readonly
define i8 @uaddtest1(i8 %A, i8 %B) {
%x = call %overflow.result @llvm.uadd.with.overflow.i8(i8 %A, i8 %B)
@@ -425,3 +427,23 @@ define %ov.result.32 @never_overflows_ssub_test0(i32 %a) {
; CHECK-NEXT: %[[x:.*]] = insertvalue %ov.result.32 { i32 undef, i1 false }, i32 %a, 0
; CHECK-NEXT: ret %ov.result.32 %[[x]]
}
+
+define void @cos(double *%P) {
+entry:
+ %B = tail call double @llvm.cos.f64(double 0.0) nounwind
+ store volatile double %B, double* %P
+
+ ret void
+; CHECK-LABEL: @cos(
+; CHECK: store volatile double 1.000000e+00, double* %P
+}
+
+define void @sin(double *%P) {
+entry:
+ %B = tail call double @llvm.sin.f64(double 0.0) nounwind
+ store volatile double %B, double* %P
+
+ ret void
+; CHECK-LABEL: @sin(
+; CHECK: store volatile double 0.000000e+00, double* %P
+}
diff --git a/test/Transforms/InstCombine/load-combine-metadata.ll b/test/Transforms/InstCombine/load-combine-metadata.ll
new file mode 100644
index 0000000000000..9b9c1fe607b90
--- /dev/null
+++ b/test/Transforms/InstCombine/load-combine-metadata.ll
@@ -0,0 +1,29 @@
+; RUN: opt -instcombine -S < %s | FileCheck %s
+
+target datalayout = "e-m:e-p:64:64:64-i64:64-f80:128-n8:16:32:64-S128"
+
+; CHECK-LABEL: @test_load_load_combine_metadata(
+; Check that range and AA metadata is combined
+; CHECK: %[[V:.*]] = load i32, i32* %0
+; CHECK-SAME: !tbaa !{{[0-9]+}}
+; CHECK-SAME: !range ![[RANGE:[0-9]+]]
+; CHECK: store i32 %[[V]], i32* %1
+; CHECK: store i32 %[[V]], i32* %2
+define void @test_load_load_combine_metadata(i32*, i32*, i32*) {
+ %a = load i32, i32* %0, !tbaa !8, !range !0, !alias.scope !5, !noalias !6
+ %b = load i32, i32* %0, !tbaa !8, !range !1
+ store i32 %a, i32* %1
+ store i32 %b, i32* %2
+ ret void
+}
+
+; CHECK: ![[RANGE]] = !{i32 0, i32 1, i32 8, i32 9}
+!0 = !{ i32 0, i32 1 }
+!1 = !{ i32 8, i32 9 }
+!2 = !{!2}
+!3 = !{!3, !2}
+!4 = !{!4, !2}
+!5 = !{!3}
+!6 = !{!4}
+!7 = !{ !"tbaa root" }
+!8 = !{ !7, !7, i64 0 }
diff --git a/test/Transforms/InstCombine/load_combine_aa.ll b/test/Transforms/InstCombine/load_combine_aa.ll
new file mode 100644
index 0000000000000..b84b81ddd5d96
--- /dev/null
+++ b/test/Transforms/InstCombine/load_combine_aa.ll
@@ -0,0 +1,15 @@
+; RUN: opt -basicaa -instcombine -S < %s | FileCheck %s
+
+; CHECK-LABEL: @test_load_combine_aa(
+; CHECK: %[[V:.*]] = load i32, i32* %0
+; CHECK: store i32 0, i32* %3
+; CHECK: store i32 %[[V]], i32* %1
+; CHECK: store i32 %[[V]], i32* %2
+define void @test_load_combine_aa(i32*, i32*, i32*, i32* noalias) {
+ %a = load i32, i32* %0
+ store i32 0, i32* %3
+ %b = load i32, i32* %0
+ store i32 %a, i32* %1
+ store i32 %b, i32* %2
+ ret void
+}
diff --git a/test/Transforms/InstSimplify/2011-09-05-InsertExtractValue.ll b/test/Transforms/InstSimplify/2011-09-05-InsertExtractValue.ll
index 885cb70007e67..7e391aba30456 100644
--- a/test/Transforms/InstSimplify/2011-09-05-InsertExtractValue.ll
+++ b/test/Transforms/InstSimplify/2011-09-05-InsertExtractValue.ll
@@ -27,3 +27,12 @@ define { i8, i32 } @test2({ i8*, i32 } %x) {
ret { i8, i32 } %ins
; CHECK-LABEL: @test2(
}
+
+define i32 @test3(i32 %a, float %b) {
+ %agg1 = insertvalue {i32, float} undef, i32 %a, 0
+ %agg2 = insertvalue {i32, float} %agg1, float %b, 1
+ %ev = extractvalue {i32, float} %agg2, 0
+ ret i32 %ev
+; CHECK-LABEL: @test3(
+; CHECK: ret i32 %a
+}
diff --git a/test/Transforms/InstSimplify/floating-point-compare.ll b/test/Transforms/InstSimplify/floating-point-compare.ll
index af48d062b4f64..8174f58345332 100644
--- a/test/Transforms/InstSimplify/floating-point-compare.ll
+++ b/test/Transforms/InstSimplify/floating-point-compare.ll
@@ -58,3 +58,18 @@ define i1 @orderedLessZeroPowi(double,double) {
ret i1 %olt
}
+define i1 @nonans1(double %in1, double %in2) {
+ %cmp = fcmp nnan uno double %in1, %in2
+ ret i1 %cmp
+
+; CHECK-LABEL: @nonans1
+; CHECK-NEXT: ret i1 false
+}
+
+define i1 @nonans2(double %in1, double %in2) {
+ %cmp = fcmp nnan ord double %in1, %in2
+ ret i1 %cmp
+
+; CHECK-LABEL: @nonans2
+; CHECK-NEXT: ret i1 true
+}
diff --git a/test/Transforms/InstSimplify/undef.ll b/test/Transforms/InstSimplify/undef.ll
index f1f0b037fdbd1..d75dc364243cd 100644
--- a/test/Transforms/InstSimplify/undef.ll
+++ b/test/Transforms/InstSimplify/undef.ll
@@ -265,3 +265,17 @@ define i32 @test34(i32 %a) {
%b = lshr i32 undef, 0
ret i32 %b
}
+
+; CHECK-LABEL: @test35
+; CHECK: ret i32 undef
+define i32 @test35(<4 x i32> %V) {
+ %b = extractelement <4 x i32> %V, i32 4
+ ret i32 %b
+}
+
+; CHECK-LABEL: @test36
+; CHECK: ret i32 undef
+define i32 @test36(i32 %V) {
+ %b = extractelement <4 x i32> undef, i32 %V
+ ret i32 %b
+}
diff --git a/test/Transforms/LICM/PR24013.ll b/test/Transforms/LICM/PR24013.ll
new file mode 100644
index 0000000000000..4557bfcfd1226
--- /dev/null
+++ b/test/Transforms/LICM/PR24013.ll
@@ -0,0 +1,19 @@
+; RUN: opt -licm -S < %s | FileCheck %s
+
+define void @f(i1 zeroext %p1) {
+; CHECK-LABEL: @f(
+entry:
+ br label %lbl
+
+lbl.loopexit: ; No predecessors!
+ br label %lbl
+
+lbl: ; preds = %lbl.loopexit, %entry
+ %phi = phi i32 [ %conv, %lbl.loopexit ], [ undef, %entry ]
+; CHECK: phi i32 [ undef, {{.*}} ], [ undef
+ br label %if.then.5
+
+if.then.5: ; preds = %if.then.5, %lbl
+ %conv = zext i1 undef to i32
+ br label %if.then.5
+}
diff --git a/test/Transforms/LoopDistribute/basic-with-memchecks.ll b/test/Transforms/LoopDistribute/basic-with-memchecks.ll
index fde06d33c5a5c..3aced48504111 100644
--- a/test/Transforms/LoopDistribute/basic-with-memchecks.ll
+++ b/test/Transforms/LoopDistribute/basic-with-memchecks.ll
@@ -32,8 +32,9 @@ entry:
%e = load i32*, i32** @E, align 8
br label %for.body
-; We have two compares for each array overlap check which is a total of 10
-; compares.
+; We have two compares for each array overlap check.
+; Since the checks to A and A + 4 get merged, this will give us a
+; total of 8 compares.
;
; CHECK: for.body.lver.memcheck:
; CHECK: = icmp
@@ -48,9 +49,6 @@ entry:
; CHECK: = icmp
; CHECK: = icmp
-; CHECK: = icmp
-; CHECK: = icmp
-
; CHECK-NOT: = icmp
; CHECK: br i1 %memcheck.conflict, label %for.body.ph.lver.orig, label %for.body.ph.ldist1
diff --git a/test/Transforms/LoopIdiom/ctpop-multiple-users-crash.ll b/test/Transforms/LoopIdiom/ctpop-multiple-users-crash.ll
new file mode 100644
index 0000000000000..ddb7bdbe7d191
--- /dev/null
+++ b/test/Transforms/LoopIdiom/ctpop-multiple-users-crash.ll
@@ -0,0 +1,34 @@
+; RUN: opt -loop-idiom -S < %s | FileCheck %s
+
+target datalayout = "e-m:o-i64:64-i128:128-n32:64-S128"
+target triple = "arm64-apple-ios8.0.0"
+
+; When we replace the precondition with a ctpop, we need to ensure
+; that only the first branch reads the ctpop. The store prior
+; to that should continue to read from the original compare.
+
+; CHECK: %tobool.5 = icmp ne i32 %num, 0
+; CHECK: store i1 %tobool.5, i1* %ptr
+
+define internal fastcc i32 @num_bits_set(i32 %num, i1* %ptr) #1 {
+entry:
+ %tobool.5 = icmp ne i32 %num, 0
+ store i1 %tobool.5, i1* %ptr
+ br i1 %tobool.5, label %for.body.lr.ph, label %for.end
+
+for.body.lr.ph: ; preds = %entry
+ br label %for.body
+
+for.body: ; preds = %for.body.lr.ph, %for.body
+ %count.07 = phi i32 [ 0, %for.body.lr.ph ], [ %inc, %for.body ]
+ %num.addr.06 = phi i32 [ %num, %for.body.lr.ph ], [ %and, %for.body ]
+ %sub = add i32 %num.addr.06, -1
+ %and = and i32 %sub, %num.addr.06
+ %inc = add nsw i32 %count.07, 1
+ %tobool = icmp ne i32 %and, 0
+ br i1 %tobool, label %for.body, label %for.end
+
+for.end: ; preds = %for.cond.for.end_crit_edge, %entry
+ %count.0.lcssa = phi i32 [ %inc, %for.body ], [ 0, %entry ]
+ ret i32 %count.0.lcssa
+} \ No newline at end of file
diff --git a/test/Transforms/LoopRotate/oz-disable.ll b/test/Transforms/LoopRotate/oz-disable.ll
new file mode 100644
index 0000000000000..7a6a9bf33a188
--- /dev/null
+++ b/test/Transforms/LoopRotate/oz-disable.ll
@@ -0,0 +1,30 @@
+; REQUIRES: asserts
+; RUN: opt < %s -S -Os -debug -debug-only=loop-rotate 2>&1 | FileCheck %s -check-prefix=OS
+; RUN: opt < %s -S -Oz -debug -debug-only=loop-rotate 2>&1 | FileCheck %s -check-prefix=OZ
+
+; Loop should be rotated for -Os but not for -Oz.
+; OS: rotating Loop at depth 1
+; OZ-NOT: rotating Loop at depth 1
+
+@e = global i32 10
+
+declare void @use(i32)
+
+define void @test() {
+entry:
+ %end = load i32, i32* @e
+ br label %loop
+
+loop:
+ %n.phi = phi i32 [ %n, %loop.fin ], [ 0, %entry ]
+ %cond = icmp eq i32 %n.phi, %end
+ br i1 %cond, label %exit, label %loop.fin
+
+loop.fin:
+ %n = add i32 %n.phi, 1
+ call void @use(i32 %n)
+ br label %loop
+
+exit:
+ ret void
+}
diff --git a/test/Transforms/LoopStrengthReduce/ephemeral.ll b/test/Transforms/LoopStrengthReduce/ephemeral.ll
new file mode 100644
index 0000000000000..a0d1d44b1bac4
--- /dev/null
+++ b/test/Transforms/LoopStrengthReduce/ephemeral.ll
@@ -0,0 +1,41 @@
+; RUN: opt < %s -loop-reduce -S | FileCheck %s
+
+target datalayout = "e-i64:64-v16:16-v32:32-n16:32:64"
+
+; for (int i = 0; i < n; ++i) {
+; use(i * 5 + 3);
+; // i * a + b is ephemeral and shouldn't be promoted by LSR
+; __builtin_assume(i * a + b >= 0);
+; }
+define void @ephemeral(i32 %a, i32 %b, i32 %n) {
+; CHECK-LABEL: @ephemeral(
+entry:
+ br label %loop
+
+loop:
+ %i = phi i32 [ 0, %entry ], [ %inc, %loop ]
+ ; Only i and i * 5 + 3 should be indvars, not i * a + b.
+; CHECK: phi i32
+; CHECK: phi i32
+; CHECK-NOT: phi i32
+ %inc = add nsw i32 %i, 1
+ %exitcond = icmp eq i32 %inc, %n
+
+ %0 = mul nsw i32 %i, 5
+ %1 = add nsw i32 %0, 3
+ call void @use(i32 %1)
+
+ %2 = mul nsw i32 %i, %a
+ %3 = add nsw i32 %2, %b
+ %4 = icmp sgt i32 %3, -1
+ call void @llvm.assume(i1 %4)
+
+ br i1 %exitcond, label %exit, label %loop
+
+exit:
+ ret void
+}
+
+declare void @use(i32)
+
+declare void @llvm.assume(i1)
diff --git a/test/Transforms/LoopUnroll/unroll-pragmas.ll b/test/Transforms/LoopUnroll/unroll-pragmas.ll
index 1354181becd0c..8e0d77513cc10 100644
--- a/test/Transforms/LoopUnroll/unroll-pragmas.ll
+++ b/test/Transforms/LoopUnroll/unroll-pragmas.ll
@@ -86,9 +86,9 @@ for.end: ; preds = %for.body
; #pragma clang loop unroll(full)
; Loop should be fully unrolled.
;
-; CHECK-LABEL: @loop64_with_enable(
+; CHECK-LABEL: @loop64_with_full(
; CHECK-NOT: br i1
-define void @loop64_with_enable(i32* nocapture %a) {
+define void @loop64_with_full(i32* nocapture %a) {
entry:
br label %for.body
@@ -139,14 +139,13 @@ for.end: ; preds = %for.body
!6 = !{!"llvm.loop.unroll.count", i32 4}
; #pragma clang loop unroll(full)
-; Full unrolling is requested, but loop has a dynamic trip count so
+; Full unrolling is requested, but loop has a runtime trip count so
; no unrolling should occur.
;
-; CHECK-LABEL: @dynamic_loop_with_enable(
+; CHECK-LABEL: @runtime_loop_with_full(
; CHECK: store i32
; CHECK-NOT: store i32
-; CHECK: br i1
-define void @dynamic_loop_with_enable(i32* nocapture %a, i32 %b) {
+define void @runtime_loop_with_full(i32* nocapture %a, i32 %b) {
entry:
%cmp3 = icmp sgt i32 %b, 0
br i1 %cmp3, label %for.body, label %for.end, !llvm.loop !8
@@ -168,22 +167,22 @@ for.end: ; preds = %for.body, %entry
!8 = !{!8, !4}
; #pragma clang loop unroll_count(4)
-; Loop has a dynamic trip count. Unrolling should occur, but no
-; conditional branches can be removed.
+; Loop has a runtime trip count. Runtime unrolling should occur and loop
+; should be duplicated (original and 4x unrolled).
;
-; CHECK-LABEL: @dynamic_loop_with_count4(
+; CHECK-LABEL: @runtime_loop_with_count4(
+; CHECK: for.body.prol:
+; CHECK: store
; CHECK-NOT: store
; CHECK: br i1
+; CHECK: for.body
; CHECK: store
-; CHECK: br i1
; CHECK: store
-; CHECK: br i1
; CHECK: store
-; CHECK: br i1
; CHECK: store
+; CHECK-NOT: store
; CHECK: br i1
-; CHECK-NOT: br i1
-define void @dynamic_loop_with_count4(i32* nocapture %a, i32 %b) {
+define void @runtime_loop_with_count4(i32* nocapture %a, i32 %b) {
entry:
%cmp3 = icmp sgt i32 %b, 0
br i1 %cmp3, label %for.body, label %for.end, !llvm.loop !9
diff --git a/test/Transforms/LoopVectorize/X86/vectorization-remarks.ll b/test/Transforms/LoopVectorize/X86/vectorization-remarks.ll
index 1c21748d8bdd1..8640950be32e5 100644
--- a/test/Transforms/LoopVectorize/X86/vectorization-remarks.ll
+++ b/test/Transforms/LoopVectorize/X86/vectorization-remarks.ll
@@ -9,9 +9,9 @@
; DEBUG-OUTPUT-NOT: .loc
; DEBUG-OUTPUT-NOT: {{.*}}.debug_info
-; VECTORIZED: remark: vectorization-remarks.c:17:8: vectorized loop (vectorization factor: 4, unrolling interleave factor: 1)
-; UNROLLED: remark: vectorization-remarks.c:17:8: unrolled with interleaving factor 4 (vectorization not beneficial)
; NONE: remark: vectorization-remarks.c:17:8: loop not vectorized: vector width and interleave count are explicitly set to 1
+; VECTORIZED: remark: vectorization-remarks.c:17:8: vectorized loop (vectorization width: 4, interleaved count: 1)
+; UNROLLED: remark: vectorization-remarks.c:17:8: interleaved by 4 (vectorization not beneficial)
target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128"
diff --git a/test/Transforms/PlaceSafepoints/statepoint-frameescape.ll b/test/Transforms/PlaceSafepoints/statepoint-frameescape.ll
index a6ee5ee078fba..c4e250957a8f6 100644
--- a/test/Transforms/PlaceSafepoints/statepoint-frameescape.ll
+++ b/test/Transforms/PlaceSafepoints/statepoint-frameescape.ll
@@ -1,17 +1,17 @@
; RUN: opt %s -S -place-safepoints | FileCheck %s
-declare void @llvm.frameescape(...)
+declare void @llvm.localescape(...)
-; Do we insert the entry safepoint after the frameescape intrinsic?
+; Do we insert the entry safepoint after the localescape intrinsic?
define void @parent() gc "statepoint-example" {
; CHECK-LABEL: @parent
entry:
; CHECK-LABEL: entry
; CHECK-NEXT: alloca
-; CHECK-NEXT: frameescape
+; CHECK-NEXT: localescape
; CHECK-NEXT: statepoint
%ptr = alloca i32
- call void (...) @llvm.frameescape(i32* %ptr)
+ call void (...) @llvm.localescape(i32* %ptr)
ret void
}
diff --git a/test/Transforms/SLPVectorizer/AMDGPU/simplebb.ll b/test/Transforms/SLPVectorizer/AMDGPU/simplebb.ll
index 9ed86f8814739..35763953911b5 100644
--- a/test/Transforms/SLPVectorizer/AMDGPU/simplebb.ll
+++ b/test/Transforms/SLPVectorizer/AMDGPU/simplebb.ll
@@ -1,4 +1,9 @@
; RUN: opt -S -march=r600 -mcpu=cayman -basicaa -slp-vectorizer -dce < %s | FileCheck %s
+; XFAIL: *
+;
+; FIXME: If this test expects to be vectorized, the TTI must indicate that the target
+; has vector registers of the expected width.
+; Currently, it says there are 8 vector registers that are 32-bits wide.
target datalayout = "e-p:32:32:32-p3:16:16:16-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-v16:16:16-v24:32:32-v32:32:32-v48:64:64-v64:64:64-v96:128:128-v128:128:128-v192:256:256-v256:256:256-v512:512:512-v1024:1024:1024-v2048:2048:2048-n32:64"
diff --git a/test/Transforms/SLPVectorizer/X86/cse.ll b/test/Transforms/SLPVectorizer/X86/cse.ll
index 9f56e2195991e..8d25b3661dc37 100644
--- a/test/Transforms/SLPVectorizer/X86/cse.ll
+++ b/test/Transforms/SLPVectorizer/X86/cse.ll
@@ -12,11 +12,8 @@ target triple = "i386-apple-macosx10.8.0"
;CHECK-LABEL: @test(
;CHECK: load <2 x double>
-;CHECK: fadd <2 x double>
-;CHECK: store <2 x double>
-;CHECK: insertelement <2 x double>
-;CHECK: fadd <2 x double>
-;CHECK: store <2 x double>
+;CHECK: fadd <4 x double>
+;CHECK: store <4 x double>
;CHECK: ret i32
define i32 @test(double* nocapture %G) {
@@ -48,11 +45,12 @@ entry:
; A[2] = A[2] * 7.6 * n + 3.0;
; A[3] = A[3] * 7.4 * n + 4.0;
;}
-;CHECK-LABEL: @foo(
-;CHECK: insertelement <2 x double>
-;CHECK: insertelement <2 x double>
-;CHECK-NOT: insertelement <2 x double>
-;CHECK: ret
+; CHECK-LABEL: @foo(
+; CHECK: load <4 x double>
+; CHECK: fmul <4 x double>
+; CHECK: fmul <4 x double>
+; CHECK: fadd <4 x double>
+; CHECK: store <4 x double>
define i32 @foo(double* nocapture %A, i32 %n) {
entry:
%0 = load double, double* %A, align 8
@@ -93,7 +91,7 @@ entry:
; }
; We can't merge the gather sequences because one does not dominate the other.
-; CHECK: test2
+; CHECK-LABEL: @test2(
; CHECK: insertelement
; CHECK: insertelement
; CHECK: insertelement
@@ -140,11 +138,12 @@ define i32 @test2(double* nocapture %G, i32 %k) {
; A[2] = A[2] * 7.9 * n + 6.0;
; A[3] = A[3] * 7.9 * n + 6.0;
;}
-;CHECK-LABEL: @foo4(
-;CHECK: insertelement <2 x double>
-;CHECK: insertelement <2 x double>
-;CHECK-NOT: insertelement <2 x double>
-;CHECK: ret
+; CHECK-LABEL: @foo4(
+; CHECK: load <4 x double>
+; CHECK: fmul <4 x double>
+; CHECK: fmul <4 x double>
+; CHECK: fadd <4 x double>
+; CHECK: store <4 x double>
define i32 @foo4(double* nocapture %A, i32 %n) {
entry:
%0 = load double, double* %A, align 8
diff --git a/test/Transforms/SLPVectorizer/X86/gep.ll b/test/Transforms/SLPVectorizer/X86/gep.ll
index 3f952d7b242b2..d10f2b6015d4a 100644
--- a/test/Transforms/SLPVectorizer/X86/gep.ll
+++ b/test/Transforms/SLPVectorizer/X86/gep.ll
@@ -1,5 +1,6 @@
; RUN: opt < %s -basicaa -slp-vectorizer -S |FileCheck %s
target datalayout = "e-m:o-i64:64-f80:128-n8:16:32:64-S128"
+target triple = "x86_64-unknown-unknown"
; Test if SLP can handle GEP expressions.
; The test perform the following action:
diff --git a/test/Transforms/SLPVectorizer/X86/loopinvariant.ll b/test/Transforms/SLPVectorizer/X86/loopinvariant.ll
index 0c16c34a18882..dace4b35b8711 100644
--- a/test/Transforms/SLPVectorizer/X86/loopinvariant.ll
+++ b/test/Transforms/SLPVectorizer/X86/loopinvariant.ll
@@ -1,22 +1,19 @@
-; RUN: opt < %s -basicaa -slp-vectorizer -dce -S -mtriple=x86_64-apple-macosx10.8.0 -mcpu=corei7-avx | FileCheck %s
+; RUN: opt < %s -basicaa -slp-vectorizer -S -mcpu=corei7-avx | FileCheck %s
target datalayout = "e-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-v64:64:64-v128:128:128-a0:0:64-s0:64:64-f80:128:128-n8:16:32:64-S128"
target triple = "x86_64-apple-macosx10.8.0"
;CHECK-LABEL: @foo(
-;CHECK: load <4 x i32>
-;CHECK: add nsw <4 x i32>
-;CHECK: store <4 x i32>
-;CHECK: load <4 x i32>
-;CHECK: add nsw <4 x i32>
-;CHECK: store <4 x i32>
+;CHECK: load <8 x i32>
+;CHECK: add nsw <8 x i32>
+;CHECK: store <8 x i32>
;CHECK: ret
-define i32 @foo(i32* nocapture %A, i32 %n) #0 {
+define i32 @foo(i32* nocapture %A, i32 %n) {
entry:
%cmp62 = icmp sgt i32 %n, 0
br i1 %cmp62, label %for.body, label %for.end
-for.body: ; preds = %entry, %for.body
+for.body:
%indvars.iv = phi i64 [ %indvars.iv.next, %for.body ], [ 0, %entry ]
%arrayidx = getelementptr inbounds i32, i32* %A, i64 %indvars.iv
%0 = load i32, i32* %arrayidx, align 4
@@ -62,8 +59,7 @@ for.body: ; preds = %entry, %for.body
%cmp = icmp slt i32 %15, %n
br i1 %cmp, label %for.body, label %for.end
-for.end: ; preds = %for.body, %entry
+for.end:
ret i32 undef
}
-attributes #0 = { nounwind ssp uwtable "less-precise-fpmad"="false" "no-frame-pointer-elim"="true" "no-frame-pointer-elim-non-leaf"="true" "no-infs-fp-math"="false" "no-nans-fp-math"="false" "unsafe-fp-math"="false" "use-soft-float"="false" }
diff --git a/test/Transforms/SLPVectorizer/X86/pr19657.ll b/test/Transforms/SLPVectorizer/X86/pr19657.ll
index a687aec76103c..32f8da4c7ee03 100644
--- a/test/Transforms/SLPVectorizer/X86/pr19657.ll
+++ b/test/Transforms/SLPVectorizer/X86/pr19657.ll
@@ -1,73 +1,45 @@
-; RUN: opt < %s -O1 -basicaa -slp-vectorizer -dce -S -mtriple=x86_64-unknown-linux-gnu -mcpu=corei7-avx | FileCheck %s
+; RUN: opt < %s -basicaa -slp-vectorizer -S -mcpu=corei7-avx | FileCheck %s
+; RUN: opt < %s -basicaa -slp-vectorizer -slp-max-reg-size=128 -S -mcpu=corei7-avx | FileCheck %s --check-prefix=V128
target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128"
target triple = "x86_64-unknown-linux-gnu"
-;CHECK: load <2 x double>, <2 x double>*
-;CHECK: fadd <2 x double>
-;CHECK: store <2 x double>
+; CHECK-LABEL: @foo(
+; CHECK: load <4 x double>
+; CHECK: fadd <4 x double>
+; CHECK: fadd <4 x double>
+; CHECK: store <4 x double>
-; Function Attrs: nounwind uwtable
-define void @foo(double* %x) #0 {
- %1 = alloca double*, align 8
- store double* %x, double** %1, align 8
- %2 = load double*, double** %1, align 8
- %3 = getelementptr inbounds double, double* %2, i64 0
- %4 = load double, double* %3, align 8
- %5 = load double*, double** %1, align 8
- %6 = getelementptr inbounds double, double* %5, i64 0
- %7 = load double, double* %6, align 8
- %8 = fadd double %4, %7
- %9 = load double*, double** %1, align 8
- %10 = getelementptr inbounds double, double* %9, i64 0
- %11 = load double, double* %10, align 8
- %12 = fadd double %8, %11
- %13 = load double*, double** %1, align 8
- %14 = getelementptr inbounds double, double* %13, i64 0
- store double %12, double* %14, align 8
- %15 = load double*, double** %1, align 8
- %16 = getelementptr inbounds double, double* %15, i64 1
- %17 = load double, double* %16, align 8
- %18 = load double*, double** %1, align 8
- %19 = getelementptr inbounds double, double* %18, i64 1
- %20 = load double, double* %19, align 8
- %21 = fadd double %17, %20
- %22 = load double*, double** %1, align 8
- %23 = getelementptr inbounds double, double* %22, i64 1
- %24 = load double, double* %23, align 8
- %25 = fadd double %21, %24
- %26 = load double*, double** %1, align 8
- %27 = getelementptr inbounds double, double* %26, i64 1
- store double %25, double* %27, align 8
- %28 = load double*, double** %1, align 8
- %29 = getelementptr inbounds double, double* %28, i64 2
- %30 = load double, double* %29, align 8
- %31 = load double*, double** %1, align 8
- %32 = getelementptr inbounds double, double* %31, i64 2
- %33 = load double, double* %32, align 8
- %34 = fadd double %30, %33
- %35 = load double*, double** %1, align 8
- %36 = getelementptr inbounds double, double* %35, i64 2
- %37 = load double, double* %36, align 8
- %38 = fadd double %34, %37
- %39 = load double*, double** %1, align 8
- %40 = getelementptr inbounds double, double* %39, i64 2
- store double %38, double* %40, align 8
- %41 = load double*, double** %1, align 8
- %42 = getelementptr inbounds double, double* %41, i64 3
- %43 = load double, double* %42, align 8
- %44 = load double*, double** %1, align 8
- %45 = getelementptr inbounds double, double* %44, i64 3
- %46 = load double, double* %45, align 8
- %47 = fadd double %43, %46
- %48 = load double*, double** %1, align 8
- %49 = getelementptr inbounds double, double* %48, i64 3
- %50 = load double, double* %49, align 8
- %51 = fadd double %47, %50
- %52 = load double*, double** %1, align 8
- %53 = getelementptr inbounds double, double* %52, i64 3
- store double %51, double* %53, align 8
+; V128-LABEL: @foo(
+; V128: load <2 x double>
+; V128: fadd <2 x double>
+; V128: fadd <2 x double>
+; V128: store <2 x double>
+; V128: load <2 x double>
+; V128: fadd <2 x double>
+; V128: fadd <2 x double>
+; V128: store <2 x double>
+
+define void @foo(double* %x) {
+ %1 = load double, double* %x, align 8
+ %2 = fadd double %1, %1
+ %3 = fadd double %2, %1
+ store double %3, double* %x, align 8
+ %4 = getelementptr inbounds double, double* %x, i64 1
+ %5 = load double, double* %4, align 8
+ %6 = fadd double %5, %5
+ %7 = fadd double %6, %5
+ store double %7, double* %4, align 8
+ %8 = getelementptr inbounds double, double* %x, i64 2
+ %9 = load double, double* %8, align 8
+ %10 = fadd double %9, %9
+ %11 = fadd double %10, %9
+ store double %11, double* %8, align 8
+ %12 = getelementptr inbounds double, double* %x, i64 3
+ %13 = load double, double* %12, align 8
+ %14 = fadd double %13, %13
+ %15 = fadd double %14, %13
+ store double %15, double* %12, align 8
ret void
}
-attributes #0 = { nounwind }
diff --git a/test/Transforms/SROA/basictest.ll b/test/Transforms/SROA/basictest.ll
index a59192d718cf4..7c8955b28fa2d 100644
--- a/test/Transforms/SROA/basictest.ll
+++ b/test/Transforms/SROA/basictest.ll
@@ -1595,3 +1595,14 @@ entry:
store i32 %load, i32* %a.gep1
ret void
}
+
+define void @PR23737() {
+; CHECK-LABEL: @PR23737(
+; CHECK: store atomic volatile {{.*}} seq_cst
+; CHECK: load atomic volatile {{.*}} seq_cst
+entry:
+ %ptr = alloca i64, align 8
+ store atomic volatile i64 0, i64* %ptr seq_cst, align 8
+ %load = load atomic volatile i64, i64* %ptr seq_cst, align 8
+ ret void
+}
generated by cgit v1.2.3 (git 2.25.1) at 2025-10-08 18:27:36 +0000