================
@@ -592,10 +599,15 @@ void preprocessUnreachableBlocks(FlowFunction &Func) {
/// Decide if stale profile matching can be applied for a given function.
/// Currently we skip inference for (very) large instances and for instances
/// having "unexpected" control flow (e.g., having no sink basic blocks).
-bool canApplyInference(const FlowFunction &Func) {
+bool canApplyInference(const FlowFunction &Func,
+ const yaml::bolt::BinaryFunctionProfile &YamlBF) {
if (Func.Blocks.size() > opts::StaleMatchingMaxFuncSize)
return false;
+ if ((double)Func.MatchedExecCount / YamlBF.ExecCount >=
+ opts::MatchedProfileThreshold / 100.0)
+ return false;
----------------
WenleiHe wrote:
> For block-based matching, the threshold should be higher than 5%, perhaps
> closer to a half?
Yes. Threshold of course need to be tuned based on the heuristic chosen. I just
feel that block count based threshold could be a better proxy of how confident
we are about the graph match and whether stale profile matching should
proceed.. But I don't have very strong opinion.
https://github.com/llvm/llvm-project/pull/95156
_______________________________________________
llvm-branch-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
