[flang-commits] [clang] [flang] [lld] [llvm] Test entry-points for pass-plugins with Bye in clang, flang and lld (PR #172463)

[Stefan Gränitz via flang-commits]

================
@@ -0,0 +1,62 @@
+! REQUIRES: plugins, examples
+
+! Entry-points in default and -O0 pipeline
+!
+! RUN: %flang -fpass-plugin=%llvmshlibdir/Bye%pluginext \
+! RUN:        -Xflang -load -Xflang %llvmshlibdir/Bye%pluginext \
+! RUN:        -mllvm -print-ep-callbacks -o /dev/null -S %s | FileCheck --check-prefix=EP %s
+!
+! RUN: %flang -fpass-plugin=%llvmshlibdir/Bye%pluginext -flto=full -O0 \
+! RUN:        -Xflang -load -Xflang %llvmshlibdir/Bye%pluginext \
+! RUN:        -mllvm -print-ep-callbacks -o /dev/null -S %s | FileCheck --check-prefix=EP %s
+!
+! RUN: %flang -fpass-plugin=%llvmshlibdir/Bye%pluginext -flto=thin -O0 \
+! RUN:        -Xflang -load -Xflang %llvmshlibdir/Bye%pluginext \
+! RUN:        -mllvm -print-ep-callbacks -o /dev/null -S %s | FileCheck --check-prefix=EP %s
+!
+! EP:     PipelineStart
+! EP:     PipelineEarlySimplification
+! EP-NOT: Peephole
+! EP:     ScalarOptimizerLate
+! EP-NOT: Peephole
+! EP:     OptimizerEarly
+! EP:     VectorizerStart
+! EP:     VectorizerEnd
+! EP:     OptimizerLast
----------------
weliveindetail wrote:

I think it can be surprising that Flang runs optimizier and vectorizer callbacks in its default pipeline, while Clang doesn't do that.

https://github.com/llvm/llvm-project/pull/172463