412953 - trace_malloc_leaks regression on fxdbug-win32-tb (10.7KB --> 13.8KB)

Status: RESOLVED INCOMPLETE

(Firefox :: Bookmarks & History, defect, P2)

Description

[Steve England [:stevee]]

http://build-graphs.mozilla.org/graph/query.cgi?tbox=fxdbug-win32-tb&testname=trace_malloc_leaks&autoscale=1&size=&units=bytes&ltype=&points=&showpoint=2008%253A01%253A18%253A05%253A47%253A49%252C14433&avg=1&days=2

Lk has jumped from ~10.7KB to ~13.8KB on WINNT 5.2 fxdbug-win32-tb Dep Debug + Leak Test.

Build that started 2008/01/17 @ 22:35:00 looks normal.
Build that started 2008/01/17 @ 23:32:25 has increased leak stats.

Looks like it coincides with the landing of bug 382711.

Comment 1

[Mike Schroepfer]

--> Detrich since bug 382711 is implicated

Comment 2

[Dietrich Ayala (:dietrich)]

Attachment: backout pre-emptive leak changes from bug 382711

this patch backs out the changes in the patch for bug 382711 that were not related to the bug itself, but were pre-emptive anti-leak changes :p

Comment 3

[Robert Sayre]

does this fix the leak?

Comment 4

[Dietrich Ayala (:dietrich)]

i haven't tested on windows yet. on mac, it didn't seem to make a difference, nor did backing out the patch entirely.

Comment 5

[Robert Sayre]

The code you are changing shouldn't leak. There should be trace-malloc data in the full log for those builds that will tell you what new objects it thinks are leaking. What are they?

Comment 6

[Dietrich Ayala (:dietrich)]

Bloat/Leak Delta Report
--------------------------------------------------------------------------------------
Current file:  /e/builds/tinderbox/Fx-Trunk-Memtest/WINNT_5.2_Depend/bloat-cur.log
Previous file: /e/builds/tinderbox/Fx-Trunk-Memtest/WINNT_5.2_Depend/bloat-prev.log
----------------------------------------------leaks------leaks%------bloat------bloat%

ERROR - unable to calculate bloat/leak data.

^^ Is this the right place, or would new objects be denoted in the "BloatView: ALL (cumulative) LEAK AND BLOAT STATISTICS" table somewhere?

Comment 7

[Robert Sayre]

That's the right place. How frustrating!

Comment 8

[Jonas Sicking (:sicking) No longer reading bugmail consistently]

Upping this for now.

What is the status here? In comment 4 sounds like backing out bug 382711 doesn't help at all, wouldn't that mean that this is a regression from something else?

Comment 9

[Dietrich Ayala (:dietrich)]

(In reply to comment #8)
> Upping this for now.
> 
> In comment 4 sounds like backing out bug 382711
> doesn't help at all, 

no, the leak didn't occur on mac, so testing on mac at all is pointless. i'll try to get some time to set up a windows vm to test this, but probably not going to happen for b5.

Comment 10

[Mike Connor [:mconnor]]

given that lk is now at 77k, I'm not really thinking we should block on 3.1k right now.  Time to ship is now, but we want to drill down on this as soon as the dust starts to settle a little.

Comment 11

[Gervase Markham [:gerv]]

Bug 451915 - move Firefox/Places bugs to Firefox/Bookmarks and History. Remove all bugspam from this move by filtering for the string "places-to-b-and-h".

In Thunderbird 3.0b, you do that as follows:
Tools | Message Filters
Make sure the correct account is selected. Click "New"
Conditions: Body   contains   places-to-b-and-h
Change the action to "Delete Message".
Select "Manually Run" from the dropdown at the top.
Click OK.

Select the filter in the list, make sure "Inbox" is selected at the bottom, and click "Run Now". This should delete all the bugspam. You can then delete the filter.

Gerv

Comment 12

[Marco Bonardo [:mak]]

This is no more actionable