diff options
| author | 2017-02-28 17:41:45 +0000 | |
|---|---|---|
| committer | 2017-03-24 14:02:54 +0000 | |
| commit | f6d4668c42933e2f85ddbc94e276c49db4e2b1dd (patch) | |
| tree | 3343addb8cde45eb4070d4876f26bef1002248dc /compiler/image_test.cc | |
| parent | 0bbb750f6ff9a52bc4649995fc78ebea87e31dd1 (diff) | |
Improvements in the Inliner.
- Change from a depth limit to a total number of HInstructions
inlined limit. Remove the dex2oat depth limit argument.
- Add more stats to diagnose reasons for not inlining.
- Clean up logging to easily parse output.
Individual Ritz benchmarks improve from 3 to 10%.
No change in other heuristics. There was already an instruction budget.
Note that the instruction budget is rarely hit in the "apps" I've tried
with.
Compile-times improve from 5 to 15%.
Code size go from 4% increase (Gms) to 1% decrease (Docs).
bug:35724239
test: test-art-host test-art-target
Change-Id: I5a35c4bd826cf21fead77859709553c5b57608d6
Diffstat (limited to 'compiler/image_test.cc')
| -rw-r--r-- | compiler/image_test.cc | 1 |
1 files changed, 0 insertions, 1 deletions
diff --git a/compiler/image_test.cc b/compiler/image_test.cc index 7ee494a131..897d81993d 100644 --- a/compiler/image_test.cc +++ b/compiler/image_test.cc @@ -363,7 +363,6 @@ void ImageTest::Compile(ImageHeader::StorageMode storage_mode, } CreateCompilerDriver(Compiler::kOptimizing, kRuntimeISA, kIsTargetBuild ? 2U : 16U); // Set inline filter values. - compiler_options_->SetInlineDepthLimit(CompilerOptions::kDefaultInlineDepthLimit); compiler_options_->SetInlineMaxCodeUnits(CompilerOptions::kDefaultInlineMaxCodeUnits); image_classes_.clear(); if (!extra_dex.empty()) { |