[llvm-branch-commits] [clang] [SSAF][WPA] Add PointerFlowReachableAnalysis (PR #193097)

[Ziqing Luo via llvm-branch-commits]

================
@@ -0,0 +1,442 @@
+//===- UnsafeBufferReachableAnalysisTest.cpp
+//-------------------------------===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+
+#include "../TestFixture.h"
+#include "clang/ScalableStaticAnalysisFramework/Analyses/EntityPointerLevel/EntityPointerLevel.h"
+#include "clang/ScalableStaticAnalysisFramework/Analyses/PointerFlow/PointerFlow.h"
+#include "clang/ScalableStaticAnalysisFramework/Analyses/PointerFlow/PointerFlowAnalysis.h"
+#include "clang/ScalableStaticAnalysisFramework/Analyses/UnsafeBufferUsage/UnsafeBufferUsage.h"
+#include "clang/ScalableStaticAnalysisFramework/Analyses/UnsafeBufferUsage/UnsafeBufferUsageAnalysis.h"
+#include "clang/ScalableStaticAnalysisFramework/Core/EntityLinker/LUSummary.h"
+#include "clang/ScalableStaticAnalysisFramework/Core/Model/BuildNamespace.h"
+#include "clang/ScalableStaticAnalysisFramework/Core/Model/EntityId.h"
+#include "clang/ScalableStaticAnalysisFramework/Core/Model/EntityLinkage.h"
+#include "clang/ScalableStaticAnalysisFramework/Core/Model/EntityName.h"
+#include "clang/ScalableStaticAnalysisFramework/Core/WholeProgramAnalysis/AnalysisDriver.h"
+#include "clang/ScalableStaticAnalysisFramework/Core/WholeProgramAnalysis/WPASuite.h"
+#include "llvm/ADT/Sequence.h"
+#include "gtest/gtest.h"
+#include <map>
+#include <memory>
+#include <optional>
+
+using namespace clang;
+using namespace ssaf;
+
+namespace clang::ssaf {
+extern PointerFlowEntitySummary buildPointerFlowEntitySummary(EdgeSet Edges);
+extern UnsafeBufferUsageEntitySummary
+    buildUnsafeBufferUsageEntitySummary(EntityPointerLevelSet);
+} // namespace clang::ssaf
+
+namespace {
+
+/// Enumerate all possible partitions of `N` nodes into `NumGrps` groups.
----------------
ziqingluo-90 wrote:

This tests all combinations of distributing `N` graph edges into `M` contributors -- that is, different combinations of sub-graphs that form the same whole graph.

The current naive implementation does not really take much advantage of this, though. But I'd imagine the naive implementation may not scale when the graph becomes large enough. If we do any sub-graph-based optimization/parallelization later, this enumeration would be necessary.

https://github.com/llvm/llvm-project/pull/193097