diff options
Diffstat (limited to 'test/CodeGen/PowerPC')
24 files changed, 280 insertions, 111 deletions
diff --git a/test/CodeGen/PowerPC/2007-09-04-AltivecDST.ll b/test/CodeGen/PowerPC/2007-09-04-AltivecDST.ll index 82ef2b8..b6feb5a 100644 --- a/test/CodeGen/PowerPC/2007-09-04-AltivecDST.ll +++ b/test/CodeGen/PowerPC/2007-09-04-AltivecDST.ll @@ -1,4 +1,4 @@ -; RUN: llc < %s -march=ppc64 | grep dst | count 4 +; RUN: llc < %s -march=ppc64 -mattr=+altivec | grep dst | count 4 define hidden void @_Z4borkPc(i8* %image) { entry: diff --git a/test/CodeGen/PowerPC/2009-08-17-inline-asm-addr-mode-breakage.ll b/test/CodeGen/PowerPC/2009-08-17-inline-asm-addr-mode-breakage.ll index 84aa40c..91253da 100644 --- a/test/CodeGen/PowerPC/2009-08-17-inline-asm-addr-mode-breakage.ll +++ b/test/CodeGen/PowerPC/2009-08-17-inline-asm-addr-mode-breakage.ll @@ -10,8 +10,8 @@ target triple = "powerpc-apple-darwin10.0" define void @foo(i32 %y) nounwind ssp { entry: ; CHECK: foo -; CHECK: add r3 -; CHECK: 0(r3) +; CHECK: add r2 +; CHECK: 0(r2) %y_addr = alloca i32 ; <i32*> [#uses=2] %"alloca point" = bitcast i32 0 to i32 ; <i32> [#uses=0] store i32 %y, i32* %y_addr diff --git a/test/CodeGen/PowerPC/2010-02-12-saveCR.ll b/test/CodeGen/PowerPC/2010-02-12-saveCR.ll index 974a99a..0da6e43 100644 --- a/test/CodeGen/PowerPC/2010-02-12-saveCR.ll +++ b/test/CodeGen/PowerPC/2010-02-12-saveCR.ll @@ -6,17 +6,17 @@ target triple = "powerpc-apple-darwin9.6" define void @foo() nounwind { entry: -;CHECK: mfcr r2 -;CHECK: lis r3, 1 -;CHECK: rlwinm r2, r2, 8, 0, 31 -;CHECK: ori r3, r3, 34524 -;CHECK: stwx r2, r1, r3 +;CHECK: mfcr r0 +;CHECK: lis r2, 1 +;CHECK: rlwinm r0, r0, 8, 0, 31 +;CHECK: ori r2, r2, 34540 +;CHECK: stwx r0, r1, r2 ; Make sure that the register scavenger returns the same temporary register. -;CHECK: mfcr r2 -;CHECK: lis r3, 1 -;CHECK: rlwinm r2, r2, 12, 0, 31 -;CHECK: ori r3, r3, 34520 -;CHECK: stwx r2, r1, r3 +;CHECK: lis r2, 1 +;CHECK: mfcr r0 +;CHECK: ori r2, r2, 34536 +;CHECK: rlwinm r0, r0, 12, 0, 31 +;CHECK: stwx r0, r1, r2 %x = alloca [100000 x i8] ; <[100000 x i8]*> [#uses=1] %"alloca point" = bitcast i32 0 to i32 ; <i32> [#uses=0] %x1 = bitcast [100000 x i8]* %x to i8* ; <i8*> [#uses=1] @@ -25,11 +25,11 @@ entry: br label %return return: ; preds = %entry -;CHECK: lis r3, 1 -;CHECK: ori r3, r3, 34524 -;CHECK: lwzx r2, r1, r3 -;CHECK: rlwinm r2, r2, 24, 0, 31 -;CHECK: mtcrf 32, r2 +;CHECK: lis r2, 1 +;CHECK: ori r2, r2, 34540 +;CHECK: lwzx r0, r1, r2 +;CHECK: rlwinm r0, r0, 24, 0, 31 +;CHECK: mtcrf 32, r0 ret void } diff --git a/test/CodeGen/PowerPC/2010-04-07-DbgValueOtherTargets.ll b/test/CodeGen/PowerPC/2010-04-07-DbgValueOtherTargets.ll deleted file mode 100644 index 4a85098..0000000 --- a/test/CodeGen/PowerPC/2010-04-07-DbgValueOtherTargets.ll +++ /dev/null @@ -1,28 +0,0 @@ -; RUN: llc -O0 -march=ppc32 -asm-verbose < %s | FileCheck %s -; Check that DEBUG_VALUE comments come through on a variety of targets. - -define i32 @main() nounwind ssp { -entry: -; CHECK: DEBUG_VALUE - call void @llvm.dbg.value(metadata !6, i64 0, metadata !7), !dbg !9 - ret i32 0, !dbg !10 -} - -declare void @llvm.dbg.declare(metadata, metadata) nounwind readnone - -declare void @llvm.dbg.value(metadata, i64, metadata) nounwind readnone - -!llvm.dbg.sp = !{!0} - -!0 = metadata !{i32 589870, i32 0, metadata !1, metadata !"main", metadata !"main", metadata !"", metadata !1, i32 2, metadata !3, i1 false, i1 true, i32 0, i32 0, null, i32 0, i1 false, i32 ()* @main} ; [ DW_TAG_subprogram ] -!1 = metadata !{i32 589865, metadata !"/tmp/x.c", metadata !"/Users/manav", metadata !2} ; [ DW_TAG_file_type ] -!2 = metadata !{i32 589841, i32 0, i32 12, metadata !"/tmp/x.c", metadata !"/Users/manav", metadata !"clang version 2.9 (trunk 120996)", i1 true, i1 false, metadata !"", i32 0} ; [ DW_TAG_compile_unit ] -!3 = metadata !{i32 589845, metadata !1, metadata !"", metadata !1, i32 0, i64 0, i64 0, i64 0, i32 0, null, metadata !4, i32 0, null} ; [ DW_TAG_subroutine_type ] -!4 = metadata !{metadata !5} -!5 = metadata !{i32 589860, metadata !2, metadata !"int", metadata !1, i32 0, i64 32, i64 32, i64 0, i32 0, i32 5} ; [ DW_TAG_base_type ] -!6 = metadata !{i32 0} -!7 = metadata !{i32 590080, metadata !8, metadata !"i", metadata !1, i32 3, metadata !5, i32 0} ; [ DW_TAG_auto_variable ] -!8 = metadata !{i32 589835, metadata !0, i32 2, i32 12, metadata !1, i32 0} ; [ DW_TAG_lexical_block ] -!9 = metadata !{i32 3, i32 11, metadata !8, null} -!10 = metadata !{i32 4, i32 2, metadata !8, null} - diff --git a/test/CodeGen/PowerPC/2010-05-03-retaddr1.ll b/test/CodeGen/PowerPC/2010-05-03-retaddr1.ll index 72ae9d6..0dbc2d0 100644 --- a/test/CodeGen/PowerPC/2010-05-03-retaddr1.ll +++ b/test/CodeGen/PowerPC/2010-05-03-retaddr1.ll @@ -18,8 +18,8 @@ entry: ; CHECK: _g: ; CHECK: mflr r0 ; CHECK: stw r0, 8(r1) -; CHECK: lwz r3, 0(r1) -; CHECK: lwz r3, 8(r3) +; CHECK: lwz r2, 0(r1) +; CHECK: lwz r3, 8(r2) %0 = tail call i8* @llvm.returnaddress(i32 1) ; <i8*> [#uses=1] ret i8* %0 } diff --git a/test/CodeGen/PowerPC/2010-12-18-PPCStackRefs.ll b/test/CodeGen/PowerPC/2010-12-18-PPCStackRefs.ll index bf3d577..d1a3c9f 100644 --- a/test/CodeGen/PowerPC/2010-12-18-PPCStackRefs.ll +++ b/test/CodeGen/PowerPC/2010-12-18-PPCStackRefs.ll @@ -7,7 +7,7 @@ define i32 @main() nounwind { entry: ; Make sure we're generating references using the red zone ; CHECK: main: -; CHECK: stw r3, -12(r1) +; CHECK: stw r2, -12(r1) %retval = alloca i32 %0 = alloca i32 %"alloca point" = bitcast i32 0 to i32 diff --git a/test/CodeGen/PowerPC/DbgValueOtherTargets.test b/test/CodeGen/PowerPC/DbgValueOtherTargets.test new file mode 100644 index 0000000..9702934 --- /dev/null +++ b/test/CodeGen/PowerPC/DbgValueOtherTargets.test @@ -0,0 +1 @@ +RUN: llc -O0 -march=ppc32 -asm-verbose < %S/../Inputs/DbgValueOtherTargets.ll | FileCheck %S/../Inputs/DbgValueOtherTargets.ll diff --git a/test/CodeGen/PowerPC/LargeAbsoluteAddr.ll b/test/CodeGen/PowerPC/LargeAbsoluteAddr.ll index 6f985c8..e8765de 100644 --- a/test/CodeGen/PowerPC/LargeAbsoluteAddr.ll +++ b/test/CodeGen/PowerPC/LargeAbsoluteAddr.ll @@ -1,9 +1,9 @@ ; RUN: llc < %s -march=ppc32 -mtriple=powerpc-apple-darwin | \ -; RUN: grep "stw r4, 32751" +; RUN: grep "stw r3, 32751" ; RUN: llc < %s -march=ppc64 -mtriple=powerpc-apple-darwin | \ -; RUN: grep "stw r4, 32751" +; RUN: grep "stw r3, 32751" ; RUN: llc < %s -march=ppc64 -mtriple=powerpc-apple-darwin | \ -; RUN: grep "std r4, 9024" +; RUN: grep "std r3, 9024" define void @test() nounwind { store i32 0, i32* inttoptr (i64 48725999 to i32*) diff --git a/test/CodeGen/PowerPC/ctr-cleanup.ll b/test/CodeGen/PowerPC/ctr-cleanup.ll new file mode 100644 index 0000000..04e4ffb --- /dev/null +++ b/test/CodeGen/PowerPC/ctr-cleanup.ll @@ -0,0 +1,25 @@ +; RUN: llc < %s -mcpu=a2 | 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-f128:128:128-v128:128:128-n32:64" +target triple = "powerpc64-unknown-linux-gnu" + +define void @main() #0 { +entry: + br i1 undef, label %for.end, label %for.body + +for.body: ; preds = %for.body, %entry + %indvars.iv = phi i64 [ %indvars.iv.next, %for.body ], [ 1, %entry ] + %indvars.iv.next = add i64 %indvars.iv, 1 + %lftr.wideiv = trunc i64 %indvars.iv.next to i32 + %exitcond = icmp eq i32 %lftr.wideiv, 5 + br i1 %exitcond, label %for.end, label %for.body + +; CHECK: @main +; CHECK: li {{[0-9]+}}, 4 +; CHECK-NOT: li {{[0-9]+}}, 4 +; CHECK: bdnz + +for.end: ; preds = %for.body, %entry + ret void +} + +attributes #0 = { nounwind "less-precise-fpmad"="false" "no-frame-pointer-elim"="false" "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/CodeGen/PowerPC/dbg.ll b/test/CodeGen/PowerPC/dbg.ll index 8d87cf7..220e7ed 100644 --- a/test/CodeGen/PowerPC/dbg.ll +++ b/test/CodeGen/PowerPC/dbg.ll @@ -16,7 +16,7 @@ declare void @llvm.dbg.value(metadata, i64, metadata) nounwind readnone !llvm.dbg.cu = !{!0} -!0 = metadata !{i32 720913, i32 0, i32 12, metadata !"dbg.c", metadata !"/src", metadata !"clang version 3.1", i1 true, i1 true, metadata !"", i32 0, metadata !1, metadata !1, metadata !3, metadata !1} ; [ DW_TAG_compile_unit ] +!0 = metadata !{i32 720913, i32 0, i32 12, metadata !6, metadata !"clang version 3.1", i1 true, metadata !"", i32 0, metadata !1, metadata !1, metadata !3, metadata !1, metadata !""} ; [ DW_TAG_compile_unit ] !1 = metadata !{i32 0} !3 = metadata !{metadata !5} !5 = metadata !{i32 720942, i32 0, metadata !6, metadata !"main", metadata !"main", metadata !"", metadata !6, i32 1, metadata !7, i1 false, i1 true, i32 0, i32 0, null, i32 256, i1 true, i32 (i32, i8**)* @main, null, null, metadata !13} ; [ DW_TAG_subprogram ] diff --git a/test/CodeGen/PowerPC/frame-size.ll b/test/CodeGen/PowerPC/frame-size.ll new file mode 100644 index 0000000..0e569a4 --- /dev/null +++ b/test/CodeGen/PowerPC/frame-size.ll @@ -0,0 +1,32 @@ +; RUN: llc < %s -mtriple=powerpc64-unknown-linux-gnu -mcpu=a2 | FileCheck %s +target datalayout = "E-p:32:32:32-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:32:64-f32:32:32-f64:32:64-v64:64:64-v128:128:128-a0:0:64-f128:64:128-n32" + +define i64 @foo() nounwind { +entry: + %x = alloca [32568 x i8] + %"alloca point" = bitcast i32 0 to i32 + %x1 = bitcast [32568 x i8]* %x to i8* + +; Check that the RS spill slot has been allocated (because the estimate +; will fail the small-frame-size check and the function has spills). +; CHECK: @foo +; CHECK: stdu 1, -32768(1) + + %s1 = call i64 @bar(i8* %x1) nounwind + %s2 = call i64 @bar(i8* %x1) nounwind + %s3 = call i64 @bar(i8* %x1) nounwind + %s4 = call i64 @bar(i8* %x1) nounwind + %s5 = call i64 @bar(i8* %x1) nounwind + %s6 = call i64 @bar(i8* %x1) nounwind + %s7 = call i64 @bar(i8* %x1) nounwind + %s8 = call i64 @bar(i8* %x1) nounwind + %r = call i64 @can(i64 %s1, i64 %s2, i64 %s3, i64 %s4, i64 %s5, i64 %s6, i64 %s7, i64 %s8) nounwind + br label %return + +return: + ret i64 %r +} + +declare i64 @bar(i8*) +declare i64 @can(i64, i64, i64, i64, i64, i64, i64, i64) + diff --git a/test/CodeGen/PowerPC/iabs.ll b/test/CodeGen/PowerPC/iabs.ll index 7d089bb..f683238 100644 --- a/test/CodeGen/PowerPC/iabs.ll +++ b/test/CodeGen/PowerPC/iabs.ll @@ -1,3 +1,4 @@ +; REQUIRES: asserts ; RUN: llc < %s -march=ppc32 -stats 2>&1 | \ ; RUN: grep "4 .*Number of machine instrs printed" diff --git a/test/CodeGen/PowerPC/lbzux.ll b/test/CodeGen/PowerPC/lbzux.ll index 12f1d1f..9895130 100644 --- a/test/CodeGen/PowerPC/lbzux.ll +++ b/test/CodeGen/PowerPC/lbzux.ll @@ -1,6 +1,6 @@ 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-v128:128:128-n32:64" target triple = "powerpc64-unknown-linux-gnu" -; RUN: llc < %s | FileCheck %s +; RUN: llc -disable-ppc-unaligned < %s | FileCheck %s define fastcc void @allocateSpace(i1 %cond1, i1 %cond2) nounwind { entry: diff --git a/test/CodeGen/PowerPC/lit.local.cfg b/test/CodeGen/PowerPC/lit.local.cfg index 4019eca..aaa31d9 100644 --- a/test/CodeGen/PowerPC/lit.local.cfg +++ b/test/CodeGen/PowerPC/lit.local.cfg @@ -1,4 +1,4 @@ -config.suffixes = ['.ll', '.c', '.cpp'] +config.suffixes = ['.ll', '.c', '.cpp', '.test'] targets = set(config.root.targets_to_build.split()) if not 'PowerPC' in targets: diff --git a/test/CodeGen/PowerPC/negctr.ll b/test/CodeGen/PowerPC/negctr.ll new file mode 100644 index 0000000..9b438a1 --- /dev/null +++ b/test/CodeGen/PowerPC/negctr.ll @@ -0,0 +1,63 @@ +; RUN: llc < %s -mcpu=a2 | 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-f128:128:128-v128:128:128-n32:64" +target triple = "powerpc64-unknown-linux-gnu" + +define void @main() #0 { +entry: + br i1 undef, label %for.end, label %for.body + +for.body: ; preds = %for.body, %entry + %indvars.iv = phi i64 [ %indvars.iv.next, %for.body ], [ 1, %entry ] + %indvars.iv.next = add i64 %indvars.iv, 1 + %lftr.wideiv = trunc i64 %indvars.iv.next to i32 + %exitcond = icmp eq i32 %lftr.wideiv, 0 + br i1 %exitcond, label %for.end, label %for.body + +; FIXME: We currently can't form the 32-bit unsigned trip count necessary here! +; CHECK: @main +; CHECK-NOT: bdnz + +for.end: ; preds = %for.body, %entry + ret void +} + +define void @main1() #0 { +entry: + br i1 undef, label %for.end, label %for.body + +for.body: ; preds = %for.body, %entry + %indvars.iv = phi i64 [ %indvars.iv.next, %for.body ], [ 1, %entry ] + %indvars.iv.next = add i64 %indvars.iv, 1 + %exitcond = icmp eq i64 %indvars.iv.next, 0 + br i1 %exitcond, label %for.end, label %for.body + +; CHECK: @main1 +; CHECK: li [[REG:[0-9]+]], -1 +; CHECK: mtctr [[REG]] +; CHECK: bdnz + +for.end: ; preds = %for.body, %entry + ret void +} + +define void @main2() #0 { +entry: + br i1 undef, label %for.end, label %for.body + +for.body: ; preds = %for.body, %entry + %indvars.iv = phi i64 [ %indvars.iv.next, %for.body ], [ 1, %entry ] + %indvars.iv.next = add i64 %indvars.iv, 1 + %exitcond = icmp eq i64 %indvars.iv.next, -100000 + br i1 %exitcond, label %for.end, label %for.body + +; CHECK: @main2 +; CHECK: lis [[REG:[0-9]+]], -2 +; CHECK: ori [[REG2:[0-9]+]], [[REG]], 31071 +; CHECK: mtctr [[REG2]] +; CHECK: bdnz + +for.end: ; preds = %for.body, %entry + ret void +} + +attributes #0 = { nounwind "less-precise-fpmad"="false" "no-frame-pointer-elim"="false" "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/CodeGen/PowerPC/rlwimi3.ll b/test/CodeGen/PowerPC/rlwimi3.ll index 7efdbe9..31b6d4a 100644 --- a/test/CodeGen/PowerPC/rlwimi3.ll +++ b/test/CodeGen/PowerPC/rlwimi3.ll @@ -1,3 +1,4 @@ +; REQUIRES: asserts ; RUN: llc < %s -march=ppc32 -stats 2>&1 | \ ; RUN: grep "Number of machine instrs printed" | grep 12 diff --git a/test/CodeGen/PowerPC/structsinmem.ll b/test/CodeGen/PowerPC/structsinmem.ll index 884d3a8..8dbe63d 100644 --- a/test/CodeGen/PowerPC/structsinmem.ll +++ b/test/CodeGen/PowerPC/structsinmem.ll @@ -1,9 +1,5 @@ ; RUN: llc -mcpu=pwr7 -O0 -disable-fp-elim < %s | FileCheck %s -; FIXME: The code generation for packed structs is very poor because the -; PowerPC target wrongly rejects all unaligned loads. This test case will -; need to be revised when that is fixed. - 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-v128:128:128-n32:64" target triple = "powerpc64-unknown-linux-gnu" @@ -209,19 +205,11 @@ entry: %add13 = add nsw i32 %add11, %6 ret i32 %add13 -; CHECK: lbz {{[0-9]+}}, 149(1) -; CHECK: lbz {{[0-9]+}}, 150(1) -; CHECK: lbz {{[0-9]+}}, 147(1) -; CHECK: lbz {{[0-9]+}}, 148(1) -; CHECK: lbz {{[0-9]+}}, 133(1) -; CHECK: lbz {{[0-9]+}}, 134(1) ; CHECK: lha {{[0-9]+}}, 126(1) ; CHECK: lbz {{[0-9]+}}, 119(1) +; CHECK: lha {{[0-9]+}}, 133(1) ; CHECK: lwz {{[0-9]+}}, 140(1) -; CHECK: lhz {{[0-9]+}}, 154(1) -; CHECK: lhz {{[0-9]+}}, 156(1) -; CHECK: lbz {{[0-9]+}}, 163(1) -; CHECK: lbz {{[0-9]+}}, 164(1) -; CHECK: lbz {{[0-9]+}}, 161(1) -; CHECK: lbz {{[0-9]+}}, 162(1) +; CHECK: lwz {{[0-9]+}}, 147(1) +; CHECK: lwz {{[0-9]+}}, 154(1) +; CHECK: lwz {{[0-9]+}}, 161(1) } diff --git a/test/CodeGen/PowerPC/structsinregs.ll b/test/CodeGen/PowerPC/structsinregs.ll index ef706af..6005614 100644 --- a/test/CodeGen/PowerPC/structsinregs.ll +++ b/test/CodeGen/PowerPC/structsinregs.ll @@ -1,9 +1,5 @@ ; RUN: llc -mcpu=pwr7 -O0 -disable-fp-elim < %s | FileCheck %s -; FIXME: The code generation for packed structs is very poor because the -; PowerPC target wrongly rejects all unaligned loads. This test case will -; need to be revised when that is fixed. - 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-v128:128:128-n32:64" target triple = "powerpc64-unknown-linux-gnu" @@ -63,13 +59,13 @@ entry: %call = call i32 @callee1(%struct.s1* byval %p1, %struct.s2* byval %p2, %struct.s3* byval %p3, %struct.s4* byval %p4, %struct.s5* byval %p5, %struct.s6* byval %p6, %struct.s7* byval %p7) ret i32 %call -; CHECK: ld 9, 128(31) -; CHECK: ld 8, 136(31) -; CHECK: ld 7, 144(31) -; CHECK: lwz 6, 152(31) -; CHECK: lwz 5, 160(31) -; CHECK: lhz 4, 168(31) -; CHECK: lbz 3, 176(31) +; CHECK: ld 9, 112(31) +; CHECK: ld 8, 120(31) +; CHECK: ld 7, 128(31) +; CHECK: lwz 6, 136(31) +; CHECK: lwz 5, 144(31) +; CHECK: lhz 4, 152(31) +; CHECK: lbz 3, 160(31) } declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i32, i1) nounwind @@ -155,10 +151,10 @@ entry: ; CHECK: ld 9, 96(1) ; CHECK: ld 8, 88(1) ; CHECK: ld 7, 80(1) -; CHECK: lwz 6, 152(31) +; CHECK: lwz 6, 136(31) ; CHECK: ld 5, 64(1) -; CHECK: lhz 4, 168(31) -; CHECK: lbz 3, 176(31) +; CHECK: lhz 4, 152(31) +; CHECK: lbz 3, 160(31) } define internal i32 @callee2(%struct.t1* byval %v1, %struct.t2* byval %v2, %struct.t3* byval %v3, %struct.t4* byval %v4, %struct.t5* byval %v5, %struct.t6* byval %v6, %struct.t7* byval %v7) nounwind { @@ -195,19 +191,11 @@ entry: ; CHECK: std 5, 64(1) ; CHECK: sth 4, 62(1) ; CHECK: stb 3, 55(1) -; CHECK: lbz {{[0-9]+}}, 85(1) -; CHECK: lbz {{[0-9]+}}, 86(1) -; CHECK: lbz {{[0-9]+}}, 83(1) -; CHECK: lbz {{[0-9]+}}, 84(1) -; CHECK: lbz {{[0-9]+}}, 69(1) -; CHECK: lbz {{[0-9]+}}, 70(1) ; CHECK: lha {{[0-9]+}}, 62(1) ; CHECK: lbz {{[0-9]+}}, 55(1) +; CHECK: lha {{[0-9]+}}, 69(1) ; CHECK: lwz {{[0-9]+}}, 76(1) -; CHECK: lhz {{[0-9]+}}, 90(1) -; CHECK: lhz {{[0-9]+}}, 92(1) -; CHECK: lbz {{[0-9]+}}, 99(1) -; CHECK: lbz {{[0-9]+}}, 100(1) -; CHECK: lbz {{[0-9]+}}, 97(1) -; CHECK: lbz {{[0-9]+}}, 98(1) +; CHECK: lwz {{[0-9]+}}, 83(1) +; CHECK: lwz {{[0-9]+}}, 90(1) +; CHECK: lwz {{[0-9]+}}, 97(1) } diff --git a/test/CodeGen/PowerPC/stwu8.ll b/test/CodeGen/PowerPC/stwu8.ll index 897bfc6..e0bd043 100644 --- a/test/CodeGen/PowerPC/stwu8.ll +++ b/test/CodeGen/PowerPC/stwu8.ll @@ -1,4 +1,4 @@ -; RUN: llc < %s | FileCheck %s +; RUN: llc -disable-ppc-unaligned < %s | 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-v128:128:128-n32:64" target triple = "powerpc64-unknown-linux-gnu" diff --git a/test/CodeGen/PowerPC/unal4-std.ll b/test/CodeGen/PowerPC/unal4-std.ll new file mode 100644 index 0000000..dc4c20c --- /dev/null +++ b/test/CodeGen/PowerPC/unal4-std.ll @@ -0,0 +1,24 @@ +; RUN: llc < %s -mcpu=pwr7 | 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-f128:128:128-v128:128:128-n32:64" +target triple = "powerpc64-unknown-linux-gnu" + +define fastcc void @copy_to_conceal() #0 { +entry: + br i1 undef, label %if.then, label %if.end210 + +if.then: ; preds = %entry + br label %vector.body.i + +vector.body.i: ; preds = %vector.body.i, %if.then + %index.i = phi i64 [ 0, %vector.body.i ], [ 0, %if.then ] + store <8 x i16> zeroinitializer, <8 x i16>* undef, align 2 + br label %vector.body.i + +if.end210: ; preds = %entry + ret void + +; CHECK: @copy_to_conceal +; CHECK: stdx {{[0-9]+}}, 0, +} + +attributes #0 = { nounwind "less-precise-fpmad"="false" "no-frame-pointer-elim"="false" "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/CodeGen/PowerPC/unaligned.ll b/test/CodeGen/PowerPC/unaligned.ll new file mode 100644 index 0000000..d050803 --- /dev/null +++ b/test/CodeGen/PowerPC/unaligned.ll @@ -0,0 +1,73 @@ +; RUN: llc < %s -mtriple=powerpc64-unknown-linux-gnu -mcpu=pwr7 | FileCheck %s +target datalayout = "E-p:32:32:32-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:32:64-f32:32:32-f64:32:64-v64:64:64-v128:128:128-a0:0:64-f128:64:128-n32" + +define void @foo1(i16* %p, i16* %r) nounwind { +entry: + %v = load i16* %p, align 1 + store i16 %v, i16* %r, align 1 + ret void + +; CHECK: @foo1 +; CHECK: lhz +; CHECK: sth +} + +define void @foo2(i32* %p, i32* %r) nounwind { +entry: + %v = load i32* %p, align 1 + store i32 %v, i32* %r, align 1 + ret void + +; CHECK: @foo2 +; CHECK: lwz +; CHECK: stw +} + +define void @foo3(i64* %p, i64* %r) nounwind { +entry: + %v = load i64* %p, align 1 + store i64 %v, i64* %r, align 1 + ret void + +; CHECK: @foo3 +; CHECK: ld +; CHECK: std +} + +define void @foo4(float* %p, float* %r) nounwind { +entry: + %v = load float* %p, align 1 + store float %v, float* %r, align 1 + ret void + +; CHECK: @foo4 +; CHECK: lfs +; CHECK: stfs +} + +define void @foo5(double* %p, double* %r) nounwind { +entry: + %v = load double* %p, align 1 + store double %v, double* %r, align 1 + ret void + +; CHECK: @foo5 +; CHECK: lfd +; CHECK: stfd +} + +define void @foo6(<4 x float>* %p, <4 x float>* %r) nounwind { +entry: + %v = load <4 x float>* %p, align 1 + store <4 x float> %v, <4 x float>* %r, align 1 + ret void + +; These loads and stores are legalized into aligned loads and stores +; using aligned stack slots. +; CHECK: @foo6 +; CHECK: ld +; CHECK: ld +; CHECK: std +; CHECK: std +} + diff --git a/test/CodeGen/PowerPC/varargs.ll b/test/CodeGen/PowerPC/varargs.ll index 1769be9..90f0480 100644 --- a/test/CodeGen/PowerPC/varargs.ll +++ b/test/CodeGen/PowerPC/varargs.ll @@ -8,15 +8,16 @@ define i8* @test1(i8** %foo) nounwind { } ; P32: test1: -; P32: lwz r4, 0(r3) -; P32: addi r5, r4, 4 -; P32: stw r5, 0(r3) -; P32: lwz r3, 0(r4) -; P32: blr +; P32: lwz r2, 0(r3) +; P32: addi r4, r2, 4 +; P32: stw r4, 0(r3) +; P32: lwz r3, 0(r2) +; P32: blr ; P64: test1: -; P64: ld r4, 0(r3) -; P64: addi r5, r4, 8 -; P64: std r5, 0(r3) -; P64: ld r3, 0(r4) -; P64: blr +; P64: ld r2, 0(r3) +; P64: addi r4, r2, 8 +; P64: std r4, 0(r3) +; P64: ld r3, 0(r2) +; P64: blr + diff --git a/test/CodeGen/PowerPC/vec_cmp.ll b/test/CodeGen/PowerPC/vec_cmp.ll index 3180f46..eb41667 100644 --- a/test/CodeGen/PowerPC/vec_cmp.ll +++ b/test/CodeGen/PowerPC/vec_cmp.ll @@ -54,7 +54,7 @@ entry: } ; CHECK: v16si8_cmp_ne: ; CHECK: vcmpequb [[RET:[0-9]+]], 2, 3 -; CHECK-NOR: vnor 2, [[RET]], [[RET]] +; CHECK-NEXT: vnor 2, [[RET]], [[RET]] define <16 x i8> @v16si8_cmp_le(<16 x i8> %x, <16 x i8> %y) nounwind readnone { entry: diff --git a/test/CodeGen/PowerPC/vrspill.ll b/test/CodeGen/PowerPC/vrspill.ll index 7641017..9fb3d03 100644 --- a/test/CodeGen/PowerPC/vrspill.ll +++ b/test/CodeGen/PowerPC/vrspill.ll @@ -13,7 +13,7 @@ entry: ret void } -; CHECK: stvx 2, 0, 0 -; CHECK: lvx 2, 0, 0 +; CHECK: stvx 2, 1, +; CHECK: lvx 2, 1, declare void @foo(i32*) |