[PATCH] D106588: [CSSPGO] Tweak ICP threshold in top-down inliner

[Lei Wang via Phabricator via llvm-commits]

wlei created this revision.
Herald added subscribers: ormris, hoy, wenlei, lxfind, hiraditya.
wlei requested review of this revision.
Herald added a project: LLVM.
Herald added a subscriber: llvm-commits.

Repository:
  rG LLVM Github Monorepo

https://reviews.llvm.org/D106588

Files:
  llvm/lib/Transforms/IPO/SampleProfile.cpp
  llvm/test/Transforms/SampleProfile/csspgo-inline-icall.ll


Index: llvm/test/Transforms/SampleProfile/csspgo-inline-icall.ll
===================================================================
--- llvm/test/Transforms/SampleProfile/csspgo-inline-icall.ll
+++ llvm/test/Transforms/SampleProfile/csspgo-inline-icall.ll
@@ -1,7 +1,7 @@
-; RUN: opt < %s -sample-profile -sample-profile-file=%S/Inputs/indirect-call-csspgo.prof -sample-profile-icp-threshold=100 -pass-remarks=sample-profile -S -o /dev/null 2>&1 | FileCheck -check-prefix=ICP-ALL %s
-; RUN: opt < %s -passes=sample-profile -sample-profile-file=%S/Inputs/indirect-call-csspgo.prof -sample-profile-icp-threshold=100 -pass-remarks=sample-profile -S -o /dev/null 2>&1 | FileCheck -check-prefix=ICP-ALL %s
-; RUN: opt < %s -sample-profile -sample-profile-file=%S/Inputs/indirect-call-csspgo.prof -sample-profile-icp-threshold=100 -pass-remarks=sample-profile -sample-profile-inline-size=0 -S -o /dev/null 2>&1 | FileCheck -check-prefix=ICP-HOT %s
-; RUN: opt < %s -passes=sample-profile -sample-profile-file=%S/Inputs/indirect-call-csspgo.prof -sample-profile-icp-threshold=100 -pass-remarks=sample-profile -sample-profile-inline-size=0 -S -o /dev/null 2>&1 | FileCheck -check-prefix=ICP-HOT %s
+; RUN: opt < %s -sample-profile -sample-profile-file=%S/Inputs/indirect-call-csspgo.prof -sample-profile-icp-percent-threshold=1 -pass-remarks=sample-profile -S -o /dev/null 2>&1 | FileCheck -check-prefix=ICP-ALL %s
+; RUN: opt < %s -passes=sample-profile -sample-profile-file=%S/Inputs/indirect-call-csspgo.prof -sample-profile-icp-percent-threshold=1  -pass-remarks=sample-profile -S -o /dev/null 2>&1 | FileCheck -check-prefix=ICP-ALL %s
+; RUN: opt < %s -sample-profile -sample-profile-file=%S/Inputs/indirect-call-csspgo.prof -sample-profile-icp-percent-threshold=1  -pass-remarks=sample-profile -sample-profile-inline-size=0 -S -o /dev/null 2>&1 | FileCheck -check-prefix=ICP-HOT %s
+; RUN: opt < %s -passes=sample-profile -sample-profile-file=%S/Inputs/indirect-call-csspgo.prof -sample-profile-icp-percent-threshold=1  -pass-remarks=sample-profile -sample-profile-inline-size=0 -S -o /dev/null 2>&1 | FileCheck -check-prefix=ICP-HOT %s
 
 define void @test(void ()*) #0 !dbg !3 {
 ;; Add two direct call to force top-down order for sample profile loader
Index: llvm/lib/Transforms/IPO/SampleProfile.cpp
===================================================================
--- llvm/lib/Transforms/IPO/SampleProfile.cpp
+++ llvm/lib/Transforms/IPO/SampleProfile.cpp
@@ -197,10 +197,15 @@
     "sample-profile-cold-inline-threshold", cl::Hidden, cl::init(45),
     cl::desc("Threshold for inlining cold callsites"));
 
-static cl::opt<int> ProfileICPThreshold(
-    "sample-profile-icp-threshold", cl::Hidden, cl::init(5),
+static cl::opt<unsigned> ProfileICPCountThreshold(
+    "sample-profile-icp-count-threshold", cl::Hidden, cl::init(1),
+    cl::desc("Always allow the number of promotions in proirity-based sample "
+             "profile loader inlining."));
+
+static cl::opt<unsigned> ProfileICPPercentThreshold(
+    "sample-profile-icp-percent-threshold", cl::Hidden, cl::init(25),
     cl::desc(
-        "Relative hotness threshold for indirect "
+        "Relative hotness percentage threshold for indirect "
         "call promotion in proirity-based sample profile loader inlining."));
 
 static cl::opt<bool> CallsitePrioritizedInline(
@@ -1348,6 +1353,7 @@
       auto CalleeSamples = findIndirectCallFunctionSamples(*I, Sum);
       uint64_t SumOrigin = Sum;
       Sum *= Candidate.CallsiteDistribution;
+      unsigned ICPCount = 0;
       for (const auto *FS : CalleeSamples) {
         // TODO: Consider disable pre-lTO ICP for MonoLTO as well
         if (LTOPhase == ThinOrFullLTOPhase::ThinLTOPreLink) {
@@ -1361,7 +1367,9 @@
         // ICP isn't introducing excessive speculative checks even if individual
         // target looks beneficial to promote and inline. That means we should
         // only do ICP when there's a small number dominant targets.
-        if (EntryCountDistributed < SumOrigin / ProfileICPThreshold)
+        if (ICPCount >= ProfileICPCountThreshold &&
+            EntryCountDistributed * 100 <
+                SumOrigin * ProfileICPPercentThreshold)
           break;
         // TODO: Fix CallAnalyzer to handle all indirect calls.
         // For indirect call, we don't run CallAnalyzer to get InlineCost
@@ -1383,6 +1391,7 @@
             if (getInlineCandidate(&NewCandidate, CB))
               CQueue.emplace(NewCandidate);
           }
+          ICPCount++;
           Changed = true;
         }
       }


-------------- next part --------------
A non-text attachment was scrubbed...
Name: D106588.360944.patch
Type: text/x-patch
Size: 4568 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20210722/56730465/attachment.bin>