Adam Guthrie Reporter
	Description • 17 years ago

After the patch for bug 194404 landed, there was an increase in private bytes and working set in the talos graphs. 

Graph: http://graphs.mozilla.org/graph.html#spst=range&spstart=0&spend=1186162910&bpst=cursor&bpstart=0&bpend=1186162910&m1tid=8&m1bl=0&m1avg=0

Bonsai: http://bonsai.mozilla.org/cvsquery.cgi?treeid=default&module=PhoenixTinderbox&branch=HEAD&branchtype=match&dir=&file=&filetype=match&who=&whotype=match&sortby=Date&hours=2&date=explicit&mindate=2007-07-26+10%3A30&maxdate=2007-07-26+12%3A00&cvsroot=%2Fcvsroot

	Johnny Stenback (:jst)
	Comment 1 • 17 years ago

Dan, any chance your checkin in that window could have increased our working set?

	Johnny Stenback (:jst)
	Comment 2 • 17 years ago

The reason I ask is that I'd be surprised (even amazed) if any of our tests even touch the code that I added (i.e. opens modal content windows using window.showModalDialog()), and thus I find it hard to believe that that change would have increased our working set etc. More code, sure, a bit, but since it's not used in the tests it shouldn't make any difference in the heap size.

	dwitte@gmail.com
	Comment 3 • 17 years ago

it's extremely unlikely, since that codepath only deals with etld lookups (not additions to the table), and the change is one of correctness and shouldn't have any effect in practice with the etld table we use. if you want to prove it, i recommend we just back the patch out for a cycle (what's the cycle time on that machine?) and see.

	Johnny Stenback (:jst)
	Comment 4 • 17 years ago

I hear ya. I doubt we'll know if either of our changes are the cause of this w/o backing the changes out to let that tinderbox cycle w/o the changes, and given that your change looks a lot easier to back out, I'd rather see you try first :) Can you do that and comment on when the change was backed out, and we'll check the graphs etc. If that doesn't show anything, I can try the same with my change, and then we'll know for sure.

	dwitte@gmail.com
	Comment 5 • 17 years ago

ok; will do when i get the chance.

	dwitte@gmail.com
	Comment 6 • 17 years ago

i backed out my change and saw no significant difference in private bytes for 2 cycles. jst?

	Wayne Mery (:wsmwk)
	Comment 7 • 10 years ago

(In reply to Dan Witte (:dwitte) (not reading bugmail, email to contact) from comment #6)
> i backed out my change and saw no significant difference in private bytes
> for 2 cycles. jst?

do we still care about this?

	:Gavin Sharp [email: gavin@gavinsharp.com]
	Comment 8 • 10 years ago

too old. We're going to potentially back out showModalDialog anyhow (bug 981796).