Bug 112746 - Impress leaks on multiple places with multiple invocations when opening Impress with Alizarin template
Summary: Impress leaks on multiple places with multiple invocations when opening Impre...
Status: RESOLVED WORKSFORME
Alias: None
Product: LibreOffice
Classification: Unclassified
Component: Impress (show other bugs)
Version:
(earliest affected)
6.0.0.0.alpha0+
Hardware: All macOS (All)
: medium normal
Assignee: Not Assigned
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2017-09-29 08:16 UTC by Telesto
Modified: 2017-12-11 15:24 UTC (History)
2 users (show)

See Also:
Crash report or crash signature:


Attachments
BT & Few screenshots (5.71 MB, application/x-zip-compressed)
2017-09-29 08:17 UTC, Telesto
Details

Note You need to log in before you can comment on or make changes to this bug.
Description Telesto 2017-09-29 08:16:49 UTC
Description:
Impress leaks on multiple places with multiple invocations when opening Impress with Alizarin template

Steps to Reproduce:
1. Start Instruments.app
2. Choose Memory Leak profile tool
3. Select LibreOffice.app in instdir as target process
4. Click on the record button, LODev is started by the profiling tool
5. Wait for the StartCenter to load.
6. Click on the new Impress document icon to open a Impress document.
7. Choose the Alizarin template in the template selector by double clicking it
8. Wait until finished
9. Stop recording.
10. Analyse the profile trace.

It happens 1 out of 3 or so

Actual Results:  
Impress leaks on multiple places with multiple invocations

Expected Results:
Probably not


Reproducible: Sometimes

User Profile Reset: No

Additional Info:
Used a Lode build from 28 September around this commit 49564323e25d742ef11a9c93f5d7324459fdb95d


User-Agent: Mozilla/5.0 (Windows NT 6.3; WOW64; rv:52.0) Gecko/20100101 Firefox/52.0
Comment 1 Telesto 2017-09-29 08:17:20 UTC
Created attachment 136605 [details]
BT & Few screenshots
Comment 2 Alex Thurgood 2017-10-02 14:48:31 UTC
Version: 6.0.0.0.alpha0+
Build ID: 36535c2bef619b3a4090f6f880b975cb6d4dcb11
CPU threads: 4; OS: Mac OS X 10.13; UI render: default; 
Locale: fr-FR (fr_FR.UTF-8); Calc: group

Hmm, not really seeing it here with the above build...
Comment 3 Telesto 2017-10-02 15:16:14 UTC
(In reply to Alex Thurgood from comment #2)
> Version: 6.0.0.0.alpha0+
> Build ID: 36535c2bef619b3a4090f6f880b975cb6d4dcb11
> CPU threads: 4; OS: Mac OS X 10.13; UI render: default; 
> Locale: fr-FR (fr_FR.UTF-8); Calc: group
> 
> Hmm, not really seeing it here with the above build...

This one is a bit hard to track (1 out of 3 runs). And I can't figure out why. It's easier to reproduce with a regular daily, than with a debug build for sure
Comment 4 Julien Nabet 2017-10-04 20:09:30 UTC
With macOS 10.12 + master sources updated today, I'm not sure about the results I get.
First time, I only got few leak (most of them concern HID) + 1 for MacSpellChecking but on macSpell var. I submitted this https://gerrit.libreoffice.org/#/c/43140/ for this one.

When relaunching Instruments after having patched locally my sources, I got lots of leaks. (mutex one for example)

Then I closed and relaunched, again few leaks.

The only difference is the second time, I play around with other apps because it's quite boring to wait for Instruments. Could it be related?

In brief, it seems we must care about results provided by Instruments.
Comment 5 Xisco Faulí 2017-11-08 07:40:32 UTC
Hi Telesto,
IS the issue still reproducible after cfce9b418072ff930d400890dcc0781e919fce24 ?
Comment 6 Alex Thurgood 2017-12-08 08:16:26 UTC
I still can't see any leaks (other than the HID ones already reported elsewhere) on my latest master build:
Version: 6.1.0.0.alpha0+
Build ID: e6f38aed7c4d6f9a2fd20a3b8eba9083afcc67a2
CPU threads: 4; OS: Mac OS X 10.13.1; UI render: default; 
Locale: fr-FR (fr_FR.UTF-8); Calc: group threaded
Comment 7 Xisco Faulí 2017-12-11 13:38:21 UTC
(In reply to Xisco Faulí from comment #5)
> Hi Telesto,
> IS the issue still reproducible after
> cfce9b418072ff930d400890dcc0781e919fce24 ?

Putting into NEEDINFO until Telesto retest it in a master build...
Comment 8 Telesto 2017-12-11 15:24:59 UTC
I'm not seeing any leaks either, closing WFM