User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; ru; rv:1.8.1.4) Gecko/20070515 Firefox/2.0.0.4 Build Identifier: Mozilla/5.0 (Windows; U; Windows NT 5.1; ru; rv:1.8.1.4) Gecko/20070515 Firefox/2.0.0.4 http://vflash.ru/files/tscroll/test_01.html test: speed Firefox3 < Firefox2 простите за мой английский подругому немогу Reproducible: Always Steps to Reproduce: 1. 2. 3.

There isn't enough information in this bug report for it to be useful. The testcase seems to be just writing out a DIV repeatedly by setting innerHTML, and there are existing performance bugs on that. There's no evidence that this is a security issue.

On trunk, I get appr. 11500ms, while on branch I get 4500ms. Reopening bug, I agree this might very well turn out to be a duplicate, but it isn't incomplete.

See bug 386802. Martijn, were you using the trunk build with the fix for bug 386802?

It still takes appr. 11500 ms with 2007-07-06 trunk build (which is after the fix from bug 386802).

I've tried narrowing down the regression range. On Win2k, going from trunk build 2007-04-30 to 2007-05-01 times jump from ~12400 to ~23500 ms. Here's a bonsai query: http://bonsai.mozilla.org/cvsquery.cgi?treeid=default&module=all&branch=HEAD&branchtype=match&dir=&file=&filetype=match&who=&whotype=match&sortby=Date&hours=2&date=explicit&mindate=2007-04-30+5%3A00&maxdate=2007-05-01+6%3A00&cvsroot=%2Fcvsroot

I think this is a regression from bug 368247. Just removing these border styles from "#bodyScroll .row" class, the time drops to 18000 ms. border-right-width: 1px; border-right-style: solid; border-right-color: #000000; border-bottom-width: 1px; border-bottom-style: solid; border-bottom-color: #999999;

Ok, thanks for finding out, Marius. I think there is enough info to confirm this bug.

Test case takes about 4,3s on both Chrome and FF. Perf improved a lot over the last 5 years :-)