Bug 19380

darktable-2.0.6-1.mga5 was pushed to Core Backports Testing, please test :)

References:
http://www.darktable.org/2016/05/darktable-2-0-4-released/
http://www.darktable.org/2016/07/darktable-2-0-5-released/
http://www.darktable.org/2016/09/darktable-2-0-6-released/

RPM and SRPM in Core Backports Testing:
darktable-2.0.6-1.mga5

Assignee: rverschelde => qa-bugs
QA Contact: (none) => rverschelde

Hi,

I install it without problem. And I use it until now without problem.

I continue testing but for me all is good.

Good job ! Thank you.

Thanks for testing. If you tested on 32-bit, you can add MGA5-32-OK to the "Whiteboard" of the bug report. If you tested on 64-bit, it's MGA5-64-OK.

I tested 65-bit on my computer

This afternoon, I will install a virtual machine 32-bit to test too.

Whiteboard: (none) => MGA5-64-OK

(In reply to Marc Mascré from comment #4)
> I tested 65-bit on my computer

This guy always have better hardware than others, +1 bit to rule them all ;-)

CC: (none) => lists.jjorge

^^ yes I know 
I saw it too late ... 
I have slidy fingers :)

have you found the time to test on 32 bits systems ?

CC: (none) => mageia

I test it but not enough in my opinion.

For now, I tried 
- Install : no problem 
- To open some raw file : no problem
- To make some minor and quick modifications : no problem

I would like to use more modules and exportation to finish the test.

But until now, all is ok.

I do not expect regressions between 2.0.3 and 2.0.6, so if you've tested extensively on x86_64 and it seems to install and run OK on i586, you can probably consider it good already.

Created attachment 8469 [details]
segfault report

I had a normal used of the x86_64 version. Every things work perfectly. So this version is ok for me.

But for the i586 I have some problems. I said I make some minor and quick modifications. I was a mistake, sorry. I would like to test more and I had a segfault error.

It appear after loading raw files in dartable. I am on the lighttable, all is ok. I clic on the darkroom, and 2 or 3 second later darktable crash with a report. It's append each time I tried.

I let the report in attached files

Does the same segfault happen with 2.0.3 (and with 1.6.9)?

Is it reproducible at all times, and if so on what platform (real hardware, virtualbox)?

It seem to be reproducible all the times. I test on virtualbox machine (I didn't have other 32bits machine)

4096Mo of ram, only one processor, 16Mo video card ... what eles ?

I tested the 32bits 2.0.3 version, and it crash too. 
I can't test older version. When I tried I had a message : database version of 'library.db' is too new for this build of darktable abording. I don't know what to do to correct. 


Juste for information. When you use the 32bits version of darktable you have a message from the dev team near this : 

"you use a 32bits version of darktable. It is very limited with memory available.
We had lot of report about problems and crash with the 32 bits version. 
You should use a 64 bits version. If not, you could be SURE to have problème who could not be corrected."

with a button : "yes I understood, let me hurt using the 32bits version of darktable"


1/ I like the dev humour :)
2/ Maybe the problem is not from your package but from darktable 32bits itself.

Hehe, nice message. Given the devs warning, and since 2.0.3 was already crashing the same way, I'd be in favour of validating the 2.0.6 backport then, as it does not introduce regressions. (The decision is up to the rest of the QA team though, as I prepared the backport myself)

can you start to prepare the adv file ? ( the time qa validates it ).

We don't use advisories for backports (so far). The package listing is in comment 1.

ah ok :D

Validating the backport

Keywords: (none) => validated_backport
CC: (none) => davidwhodgins

pushed on backports

Status: NEW => RESOLVED
Resolution: (none) => FIXED