Summary:
In the benchmark (https://github.com/vetter/shoc) we are researching,
the duplicated load is not eliminated because MemoryDependenceAnalysis
hit the BlockScanLimit. This patch change it into a command line option
instead of a hardcoded value.
Patched by Xuetian Weng.
Test Plan: test/Analysis/MemoryDependenceAnalysis/memdep-block-scan-limit.ll
Reviewers: jingyue, reames
Subscribers: reames, llvm-commits
Differential Revision: http://reviews.llvm.org/D11366
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@242842
91177308-0d34-0410-b5e6-
96231b3b80d8
"Number of block queries that were completely cached");
// Limit for the number of instructions to scan in a block.
-static const unsigned int BlockScanLimit = 100;
+
+static cl::opt<unsigned> BlockScanLimit(
+ "memdep-block-scan-limit", cl::Hidden, cl::init(100),
+ cl::desc("The number of instructions to scan in a block in memory "
+ "dependency analysis (default = 100)"));
// Limit on the number of memdep results to process.
static const unsigned int NumResultsLimit = 100;
--- /dev/null
+; RUN: opt -S -memdep -gvn -basicaa < %s | FileCheck %s
+; RUN: opt -S -memdep -memdep-block-scan-limit=1 -gvn -basicaa < %s | FileCheck %s --check-prefix=WITH-LIMIT
+; CHECK-LABEL: @test(
+; CHECK: load
+; CHECK-NOT: load
+; WITH-LIMIT-LABEL: @test(
+; WITH-LIMIT-CHECK: load
+; WITH-LIMIT-CHECK: load
+define i32 @test(i32* %p) {
+ %1 = load i32, i32* %p
+ %2 = add i32 %1, 3
+ %3 = load i32, i32* %p
+ %4 = add i32 %2, %3
+ ret i32 %4
+}