| Summary: | Optimize calls to TRACER in BufferOutputStream | ||||||
|---|---|---|---|---|---|---|---|
| Product: | [Modeling] EMF | Reporter: | Egidijus Vaisnora <vaisegid> | ||||
| Component: | cdo.core | Assignee: | Egidijus Vaisnora <vaisegid> | ||||
| Status: | CLOSED FIXED | QA Contact: | Eike Stepper <stepper> | ||||
| Severity: | enhancement | ||||||
| Priority: | P3 | Flags: | stepper:
review+
|
||||
| Version: | 4.0 | ||||||
| Target Milestone: | --- | ||||||
| Hardware: | PC | ||||||
| OS: | Mac OS X - Carbon (unsup.) | ||||||
| Whiteboard: | |||||||
| Attachments: |
|
||||||
|
Description
Egidijus Vaisnora
Created attachment 189193 [details]
patch v1
Have you measured then performance gain? This way or that way I'm fine with your changes. Maybe we want to make ensureBuffer() private so that the compiler has a chance to inline it? Similarly there's inlining potential by making the public flush() method call a new private method which is then also called by flushIfFilled(). If all this helps you may want to look at BufferInputStream, too... Method took about 12% of total call time (was visible after applying 337397 patch). Not such big overhead, but fix by itself is trivial. Added fix to the BufferInputStream too. However with your suggestion about inline I am not sure. I left it untouch. fixed in trunk, revision 7177 Committed revision 7183: - trunk/plugins/org.eclipse.net4j Available in R20110608-1407 |