[PATCH] D26097: Don't claim the udiv created in BypassSlowDivision is exact.

Justin Lebar via llvm-commits llvm-commits at lists.llvm.org
Fri Oct 28 14:53:27 PDT 2016


This revision was automatically updated to reflect the committed changes.
Closed by commit rL285459: Don't claim the udiv created in BypassSlowDivision is exact. (authored by jlebar).

Changed prior to commit:
  https://reviews.llvm.org/D26097?vs=76242&id=76251#toc

Repository:
  rL LLVM

https://reviews.llvm.org/D26097

Files:
  llvm/trunk/lib/Transforms/Utils/BypassSlowDivision.cpp
  llvm/trunk/test/Transforms/CodeGenPrepare/NVPTX/bypass-slow-div-not-exact.ll


Index: llvm/trunk/lib/Transforms/Utils/BypassSlowDivision.cpp
===================================================================
--- llvm/trunk/lib/Transforms/Utils/BypassSlowDivision.cpp
+++ llvm/trunk/lib/Transforms/Utils/BypassSlowDivision.cpp
@@ -120,8 +120,7 @@
                                                  BypassType);
 
   // udiv/urem because optimization only handles positive numbers
-  Value *ShortQuotientV = FastBuilder.CreateExactUDiv(ShortDividendV,
-                                                      ShortDivisorV);
+  Value *ShortQuotientV = FastBuilder.CreateUDiv(ShortDividendV, ShortDivisorV);
   Value *ShortRemainderV = FastBuilder.CreateURem(ShortDividendV,
                                                   ShortDivisorV);
   Value *FastQuotientV = FastBuilder.CreateCast(Instruction::ZExt,
Index: llvm/trunk/test/Transforms/CodeGenPrepare/NVPTX/bypass-slow-div-not-exact.ll
===================================================================
--- llvm/trunk/test/Transforms/CodeGenPrepare/NVPTX/bypass-slow-div-not-exact.ll
+++ llvm/trunk/test/Transforms/CodeGenPrepare/NVPTX/bypass-slow-div-not-exact.ll
@@ -0,0 +1,16 @@
+; RUN: opt -S -codegenprepare < %s | FileCheck %s
+
+target datalayout = "e-i64:64-v16:16-v32:32-n16:32:64"
+target triple = "nvptx64-nvidia-cuda"
+
+; Check that the smaller-width division that the BypassSlowDivision pass
+; creates is not marked as "exact" (that is, it doesn't claim that the
+; numerator is a multiple of the denominator).
+;
+; CHECK-LABEL: @test
+define void @test(i64 %a, i64 %b, i64* %retptr) {
+  ; CHECK: udiv i32
+  %d = sdiv i64 %a, %b
+  store i64 %d, i64* %retptr
+  ret void
+}


-------------- next part --------------
A non-text attachment was scrubbed...
Name: D26097.76251.patch
Type: text/x-patch
Size: 1667 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20161028/2b219531/attachment.bin>


More information about the llvm-commits mailing list