Increasing the inline limit from (overly conservative) 200 to 300. Given each BB costs 20 and each instruction costs 5, 200 means a 4 BB function + 24 instructions (actually less because caller's size also contributes to it).
Furthermore, double the limit when more than 10% of the callee instructions are vector instructions. Multimedia kernels tend to love inlining.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@48725 91177308-0d34-0410-b5e6-96231b3b80d8
diff --git a/lib/Transforms/IPO/InlineSimple.cpp b/lib/Transforms/IPO/InlineSimple.cpp
index fe0ea5a..a007103 100644
--- a/lib/Transforms/IPO/InlineSimple.cpp
+++ b/lib/Transforms/IPO/InlineSimple.cpp
@@ -40,6 +40,9 @@
int getInlineCost(CallSite CS) {
return CA.getInlineCost(CS, NeverInline);
}
+ float getInlineFudgeFactor(CallSite CS) {
+ return CA.getInlineFudgeFactor(CS);
+ }
virtual bool doInitialization(CallGraph &CG);
};
char SimpleInliner::ID = 0;