Don't narrow the load and store in a load+twiddle+store sequence unless

there are clearly no stores between the load and the store. This fixes
this miscompile reported as PR7833.

This breaks the test/CodeGen/X86/narrow_op-2.ll optimization, which is
safe, but awkward to prove safe. Move it to X86's README.txt.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@112861 91177308-0d34-0410-b5e6-96231b3b80d8

diff --git a/lib/CodeGen/SelectionDAG/DAGCombiner.cpp b/lib/CodeGen/SelectionDAG/DAGCombiner.cpp
index 7ffbf8d791af81b1a9aa5f40e51e10b39e282445..c9c4d91e973610487a8149e843d45f2e5438c474 100644 (file)
--- a/lib/CodeGen/SelectionDAG/DAGCombiner.cpp
+++ b/lib/CodeGen/SelectionDAG/DAGCombiner.cpp
@@ -5798,7 +5798,8 @@ SDValue DAGCombiner::ReduceLoadOpStoreWidth(SDNode *N) {
     return SDValue();
 
   SDValue N0 = Value.getOperand(0);
-  if (ISD::isNormalLoad(N0.getNode()) && N0.hasOneUse()) {
+  if (ISD::isNormalLoad(N0.getNode()) && N0.hasOneUse() &&
+      Chain == SDValue(N0.getNode(), 1)) {
     LoadSDNode *LD = cast<LoadSDNode>(N0);
     if (LD->getBasePtr() != Ptr)
       return SDValue();

diff --git a/lib/Target/X86/README.txt b/lib/Target/X86/README.txt
index 69c6d33ae5e28d1f840cdcb8c79e28180ac32be3..a305ae6ec5505eea1d01814a6ae8b49f7c8734d6 100644 (file)
--- a/lib/Target/X86/README.txt
+++ b/lib/Target/X86/README.txt
@@ -1915,3 +1915,48 @@ And the following x86 code:
 It should be possible to eliminate the sign extensions.
 
 //===---------------------------------------------------------------------===//
+
+LLVM misses a load+store narrowing opportunity in this code:
+
+%struct.bf = type { i64, i16, i16, i32 }
+
+@bfi = external global %struct.bf*                ; <%struct.bf**> [#uses=2]
+
+define void @t1() nounwind ssp {
+entry:
+  %0 = load %struct.bf** @bfi, align 8            ; <%struct.bf*> [#uses=1]
+  %1 = getelementptr %struct.bf* %0, i64 0, i32 1 ; <i16*> [#uses=1]
+  %2 = bitcast i16* %1 to i32*                    ; <i32*> [#uses=2]
+  %3 = load i32* %2, align 1                      ; <i32> [#uses=1]
+  %4 = and i32 %3, -65537                         ; <i32> [#uses=1]
+  store i32 %4, i32* %2, align 1
+  %5 = load %struct.bf** @bfi, align 8            ; <%struct.bf*> [#uses=1]
+  %6 = getelementptr %struct.bf* %5, i64 0, i32 1 ; <i16*> [#uses=1]
+  %7 = bitcast i16* %6 to i32*                    ; <i32*> [#uses=2]
+  %8 = load i32* %7, align 1                      ; <i32> [#uses=1]
+  %9 = and i32 %8, -131073                        ; <i32> [#uses=1]
+  store i32 %9, i32* %7, align 1
+  ret void
+}
+
+LLVM currently emits this:
+
+  movq  bfi(%rip), %rax
+  andl  $-65537, 8(%rax)
+  movq  bfi(%rip), %rax
+  andl  $-131073, 8(%rax)
+  ret
+
+It could narrow the loads and stores to emit this:
+
+  movq  bfi(%rip), %rax
+  andb  $-2, 10(%rax)
+  movq  bfi(%rip), %rax
+  andb  $-3, 10(%rax)
+  ret
+
+The trouble is that there is a TokenFactor between the store and the
+load, making it non-trivial to determine if there's anything between
+the load and the store which would prohibit narrowing.
+
+//===---------------------------------------------------------------------===//

diff --git a/test/CodeGen/X86/narrow_op-2.ll b/test/CodeGen/X86/narrow_op-2.ll
deleted file mode 100644 (file)
index 796ef7a..0000000
--- a/test/CodeGen/X86/narrow_op-2.ll
+++ /dev/null
@@ -1,25 +0,0 @@
-; RUN: llc < %s -march=x86-64 | FileCheck %s
-
-       %struct.bf = type { i64, i16, i16, i32 }
-@bfi = external global %struct.bf*
-
-define void @t1() nounwind ssp {
-entry:
-
-; CHECK: andb  $-2, 10(
-; CHECK: andb  $-3, 10(
-
-       %0 = load %struct.bf** @bfi, align 8
-       %1 = getelementptr %struct.bf* %0, i64 0, i32 1
-       %2 = bitcast i16* %1 to i32*
-       %3 = load i32* %2, align 1
-       %4 = and i32 %3, -65537
-       store i32 %4, i32* %2, align 1
-       %5 = load %struct.bf** @bfi, align 8
-       %6 = getelementptr %struct.bf* %5, i64 0, i32 1
-       %7 = bitcast i16* %6 to i32*
-       %8 = load i32* %7, align 1
-       %9 = and i32 %8, -131073
-       store i32 %9, i32* %7, align 1
-       ret void
-}

diff --git a/test/CodeGen/X86/store-narrow.ll b/test/CodeGen/X86/store-narrow.ll
index 5682e7caf8bd55fb789b2f8211807246f5bddc8f..abc5174c98de7064efcad1a85a3beeddf00deb61 100644 (file)
--- a/test/CodeGen/X86/store-narrow.ll
+++ b/test/CodeGen/X86/store-narrow.ll
@@ -1,6 +1,6 @@
 ; rdar://7860110
-; RUN: llc < %s | FileCheck %s -check-prefix=X64
-; RUN: llc -march=x86 < %s | FileCheck %s -check-prefix=X32
+; RUN: llc -asm-verbose=false < %s | FileCheck %s -check-prefix=X64
+; RUN: llc -march=x86 -asm-verbose=false < %s | FileCheck %s -check-prefix=X32
 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"
 target triple = "x86_64-apple-darwin10.2"
 
@@ -125,3 +125,30 @@ entry:
 ; X32: movb    %cl, 5(%{{.*}})
 }
 
+; PR7833
+
+@g_16 = internal global i32 -1
+
+; X64: test8:
+; X64-NEXT: movl _g_16(%rip), %eax
+; X64-NEXT: movl $0, _g_16(%rip)
+; X64-NEXT: orl  $1, %eax
+; X64-NEXT: movl %eax, _g_16(%rip)
+; X64-NEXT: ret
+define void @test8() nounwind {
+  %tmp = load i32* @g_16
+  store i32 0, i32* @g_16
+  %or = or i32 %tmp, 1
+  store i32 %or, i32* @g_16
+  ret void
+}
+
+; X64: test9:
+; X64-NEXT: orb $1, _g_16(%rip)
+; X64-NEXT: ret
+define void @test9() nounwind {
+  %tmp = load i32* @g_16
+  %or = or i32 %tmp, 1
+  store i32 %or, i32* @g_16
+  ret void
+}