Reland [AlwaysInliner] Update BFI when inlining

Reviewed By: davidxl

Differential Revision: https://reviews.llvm.org/D88324

diff --git a/clang/test/CodeGen/lto-newpm-pipeline.c b/clang/test/CodeGen/lto-newpm-pipeline.c
index 9694cef..ad3f076 100644 (file)
--- a/clang/test/CodeGen/lto-newpm-pipeline.c
+++ b/clang/test/CodeGen/lto-newpm-pipeline.c
@@ -28,6 +28,7 @@
 // CHECK-FULL-O0: Starting llvm::Module pass manager run.
 // CHECK-FULL-O0: Running pass: AlwaysInlinerPass
 // CHECK-FULL-O0-NEXT: Running analysis: InnerAnalysisManagerProxy
+// CHECK-FULL-O0-NEXT: Running analysis: ProfileSummaryAnalysis
 // CHECK-FULL-O0-NEXT: Running pass: CanonicalizeAliasesPass
 // CHECK-FULL-O0-NEXT: Running pass: NameAnonGlobalPass
 // CHECK-FULL-O0-NEXT: Running pass: BitcodeWriterPass
@@ -36,6 +37,7 @@
 // CHECK-THIN-O0: Starting llvm::Module pass manager run.
 // CHECK-THIN-O0: Running pass: AlwaysInlinerPass
 // CHECK-THIN-O0-NEXT: Running analysis: InnerAnalysisManagerProxy
+// CHECK-THIN-O0-NEXT: Running analysis: ProfileSummaryAnalysis
 // CHECK-THIN-O0-NEXT: Running pass: CanonicalizeAliasesPass
 // CHECK-THIN-O0-NEXT: Running pass: NameAnonGlobalPass
 // CHECK-THIN-O0-NEXT: Running pass: ThinLTOBitcodeWriterPass

diff --git a/llvm/lib/Transforms/IPO/AlwaysInliner.cpp b/llvm/lib/Transforms/IPO/AlwaysInliner.cpp
index a9cf363..f3b23ea 100644 (file)
--- a/llvm/lib/Transforms/IPO/AlwaysInliner.cpp
+++ b/llvm/lib/Transforms/IPO/AlwaysInliner.cpp
@@ -15,6 +15,7 @@
 #include "llvm/ADT/SetVector.h"
 #include "llvm/Analysis/AssumptionCache.h"
 #include "llvm/Analysis/InlineCost.h"
+#include "llvm/Analysis/ProfileSummaryInfo.h"
 #include "llvm/Analysis/TargetLibraryInfo.h"
 #include "llvm/IR/CallingConv.h"
 #include "llvm/IR/DataLayout.h"
@@ -39,7 +40,7 @@ PreservedAnalyses AlwaysInlinerPass::run(Module &M,
   auto GetAssumptionCache = [&](Function &F) -> AssumptionCache & {
     return FAM.getResult<AssumptionAnalysis>(F);
   };
-  InlineFunctionInfo IFI(/*cg=*/nullptr, GetAssumptionCache);
+  auto &PSI = MAM.getResult<ProfileSummaryAnalysis>(M);
 
   SmallSetVector<CallBase *, 16> Calls;
   bool Changed = false;
@@ -67,6 +68,11 @@ PreservedAnalyses AlwaysInlinerPass::run(Module &M,
         emitInlinedInto(ORE, CB->getDebugLoc(), CB->getParent(), F, *Caller,
                         *OIC, false, DEBUG_TYPE);
 
+        InlineFunctionInfo IFI(
+            /*cg=*/nullptr, GetAssumptionCache, &PSI,
+            &FAM.getResult<BlockFrequencyAnalysis>(*(CB->getCaller())),
+            &FAM.getResult<BlockFrequencyAnalysis>(F));
+
         InlineResult Res =
             InlineFunction(*CB, IFI, /*CalleeAAR=*/nullptr, InsertLifetime);
         assert(Res.isSuccess() && "unexpected failure to inline");

diff --git a/llvm/test/Transforms/Inline/prof-update-sample-alwaysinline.ll b/llvm/test/Transforms/Inline/prof-update-sample-alwaysinline.ll
new file mode 100644 (file)
index 0000000..5bb5834
--- /dev/null
+++ b/llvm/test/Transforms/Inline/prof-update-sample-alwaysinline.ll
@@ -0,0 +1,60 @@
+; RUN: opt < %s -passes=always-inline -S | FileCheck %s
+; Checks if always-inline updates branch_weights annotation for call instructions.
+
+declare void @ext();
+declare void @ext1();
+@func = global void ()* null
+
+; CHECK: define void @callee(i32 %n) #0 !prof ![[ENTRY_COUNT:[0-9]*]]
+define void  @callee(i32 %n) #0 !prof !15 {
+  %cond = icmp sle i32 %n, 10
+  br i1 %cond, label %cond_true, label %cond_false
+cond_true:
+; ext1 is optimized away, thus not updated.
+; CHECK: call void @ext1(), !prof ![[COUNT_CALLEE1:[0-9]*]]
+  call void @ext1(), !prof !16
+  ret void
+cond_false:
+; ext is cloned and updated.
+; CHECK: call void @ext(), !prof ![[COUNT_CALLEE:[0-9]*]]
+  call void @ext(), !prof !16
+  %f = load void ()*, void ()** @func
+; CHECK: call void %f(), !prof ![[COUNT_IND_CALLEE:[0-9]*]] 
+  call void %f(), !prof !18
+  ret void
+}
+
+; CHECK: define void @caller()
+define void @caller() {
+; CHECK: call void @ext(), !prof ![[COUNT_CALLER:[0-9]*]]
+; CHECK: call void %f.i(), !prof ![[COUNT_IND_CALLER:[0-9]*]]
+  call void @callee(i32 15), !prof !17
+  ret void
+}
+
+!llvm.module.flags = !{!1}
+!1 = !{i32 1, !"ProfileSummary", !2}
+!2 = !{!3, !4, !5, !6, !7, !8, !9, !10}
+!3 = !{!"ProfileFormat", !"SampleProfile"}
+!4 = !{!"TotalCount", i64 10000}
+!5 = !{!"MaxCount", i64 10}
+!6 = !{!"MaxInternalCount", i64 1}
+!7 = !{!"MaxFunctionCount", i64 2000}
+!8 = !{!"NumCounts", i64 2}
+!9 = !{!"NumFunctions", i64 2}
+!10 = !{!"DetailedSummary", !11}
+!11 = !{!12, !13, !14}
+!12 = !{i32 10000, i64 100, i32 1}
+!13 = !{i32 999000, i64 100, i32 1}
+!14 = !{i32 999999, i64 1, i32 2}
+!15 = !{!"function_entry_count", i64 1000}
+!16 = !{!"branch_weights", i64 2000}
+!17 = !{!"branch_weights", i64 400}
+!18 = !{!"VP", i32 0, i64 140, i64 111, i64 80, i64 222, i64 40, i64 333, i64 20}
+attributes #0 = { alwaysinline }
+; CHECK: ![[ENTRY_COUNT]] = !{!"function_entry_count", i64 600}
+; CHECK: ![[COUNT_CALLEE1]] = !{!"branch_weights", i64 2000}
+; CHECK: ![[COUNT_CALLEE]] = !{!"branch_weights", i64 1200}
+; CHECK: ![[COUNT_IND_CALLEE]] = !{!"VP", i32 0, i64 84, i64 111, i64 48, i64 222, i64 24, i64 333, i64 12}
+; CHECK: ![[COUNT_CALLER]] = !{!"branch_weights", i64 800}
+; CHECK: ![[COUNT_IND_CALLER]] = !{!"VP", i32 0, i64 56, i64 111, i64 32, i64 222, i64 16, i64 333, i64 8}

diff --git a/llvm/test/Transforms/Inline/prof-update-sample.ll b/llvm/test/Transforms/Inline/prof-update-sample.ll
index 4a4471e..add861b 100644 (file)
--- a/llvm/test/Transforms/Inline/prof-update-sample.ll
+++ b/llvm/test/Transforms/Inline/prof-update-sample.ll
@@ -51,7 +51,6 @@ define void @caller() {
 !16 = !{!"branch_weights", i64 2000}
 !17 = !{!"branch_weights", i64 400}
 !18 = !{!"VP", i32 0, i64 140, i64 111, i64 80, i64 222, i64 40, i64 333, i64 20}
-attributes #0 = { alwaysinline }
 ; CHECK: ![[ENTRY_COUNT]] = !{!"function_entry_count", i64 600}
 ; CHECK: ![[COUNT_CALLEE1]] = !{!"branch_weights", i64 2000}
 ; CHECK: ![[COUNT_CALLEE]] = !{!"branch_weights", i64 1200}