diff options
author | Lars Knoll <lars.knoll@qt.io> | 2017-05-09 08:58:28 +0200 |
---|---|---|
committer | Lars Knoll <lars.knoll@qt.io> | 2017-05-12 13:26:12 +0000 |
commit | defa2512a9128449e30d3b2c2fc42eaab201418b (patch) | |
tree | 1522ba02e6b2f3b07ba9a4760791f7f33b9c2cf1 /src/qml | |
parent | 095e7d378ee17838d9ba5d7a503da0608bb2fd0c (diff) |
Disable loop peeling
Loop peeling does in our current JIT not give us any
measurable performance benefits (no measurable diff
in any of the v8 benchmarks), and significantly
increases the size of the generated JIT code.
Change-Id: Icab7887300f9c1cd5891983cbfe5885fc2b4db91
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>
Diffstat (limited to 'src/qml')
-rw-r--r-- | src/qml/compiler/qv4ssa.cpp | 6 |
1 files changed, 5 insertions, 1 deletions
diff --git a/src/qml/compiler/qv4ssa.cpp b/src/qml/compiler/qv4ssa.cpp index 731c6ad38f..fc136b09ff 100644 --- a/src/qml/compiler/qv4ssa.cpp +++ b/src/qml/compiler/qv4ssa.cpp @@ -5417,7 +5417,11 @@ void Optimizer::run(QQmlEnginePrivate *qmlEngine, bool doTypeInference, bool pee showMeTheCode(function, "After loop detection"); // cfg2dot(function, loopDetection.allLoops()); - if (peelLoops) { + // ### disable loop peeling for now. It doesn't give any measurable performance + // improvements at this time, but significantly increases the size of the + // JIT generated code + Q_UNUSED(peelLoops); + if (0 && peelLoops) { QVector<LoopDetection::LoopInfo *> innerLoops = loopDetection.innermostLoops(); LoopPeeling(df).run(innerLoops); |