Re: RuleProcessorData and performance

2017-10-27 Thread maguihai1920
在 2009年10月20日星期二 UTC+8上午10:07:02，Boris Zbarsky写道： > On 10/19/09 1:37 AM, Boris Zbarsky wrote: > > I'm going to try to get some numbers from our Tp set as well > > tp for non-XUL nodes histograms are below. Summary: For the "LHS" > case, 80% of RuleProcessorData only need tag/namespace/id/class

Re: RuleProcessorData and performance

2009-10-19 Thread Boris Zbarsky
On 10/19/09 10:07 PM, Boris Zbarsky wrote: Will file a bug on this. https://bugzilla.mozilla.org/show_bug.cgi?id=523288 -Boris ___ dev-tech-layout mailing list dev-tech-layout@lists.mozilla.org https://lists.mozilla.org/listinfo/dev-tech-layout

Re: RuleProcessorData and performance

2009-10-19 Thread Boris Zbarsky
On 10/19/09 10:07 PM, Boris Zbarsky wrote: tp for non-XUL nodes histograms are below. Just did this measurement for Txul including XUL nodes and the patch from bug 523047. Results: 2/3 of the LHS datas only needed tag/classes/id. Same for about 40% of the RHS datas. So I think we're saf

Re: RuleProcessorData and performance

2009-10-19 Thread Boris Zbarsky
On 10/19/09 1:37 AM, Boris Zbarsky wrote: I'm going to try to get some numbers from our Tp set as well tp for non-XUL nodes histograms are below. Summary: For the "LHS" case, 80% of RuleProcessorData only need tag/namespace/id/class to not match. Anothe 12% are only matched against one sel

RuleProcessorData and performance

2009-10-18 Thread Boris Zbarsky
The RuleProcessorData constructor pretty routinely shows up in profiles as a few % of total time. For profiles of things like querySelector or frame construction it can be closer to 10%. The fundamental design assumption with RuleProcessorData is that a single element is compared to a bunch o