Re: status of performance regressions on mozilla-central

L. David Baron wrote:
[...]
> Now that the performance numbers for this are all in, I've observed
> the following, assuming that the changeset printed by talos is
> accurate and I've kept track of everything I was looking at in the
> tinderbox display correctly:
[...]
>   * on Vista, however, the news is not so good.  On Vista, there were
>     significant regressions yesterday not only in Tp (Tp3, to the
>     graph server) -- around 10-15%, but also in Txul -- around 35%
>     regression.  Today, on Vista, the needle didn't move.  Not at
>     all.  We didn't get back anything from any of the things that
>     affected the other platforms.
[...]

What about the other check-ins in that timeframe? I'm looking for 
example at WINNT 6.0 talos mozilla-central nochrome qm-pvista-trunk04 
now. The build with the lower (tp: 395.55) tp was id:20080925110153 
rev:7b75a1f34434, the build following this id:20080925124346
rev:92cc665e9b84 had a higher tp (tp: 480.02). There was a security bug 
check-in, Bug 456896, how much impact could this bug have on Tp (I do 
not know that much about layout perf...)?

Frank