Hi Mario,
I have just committed a patch that does your suggestion regarding
setting contexts only when a different one is used.
On 06/09/2016 12:09 AM, Mário Luzeiro wrote:
[snip]
>> In wxWidgets examples they do not need to set context, as there is only
>> one, so there is no chance OpenGL
Hi Orson,
> If I call SetCurrent() only when the current context is different than
> the requested one, I will still need to change the current context if
> another window has changed in the time period between closing window and
> calling its destructor. It means that we will still get
Hi Mario,
On 06/08/2016 08:09 PM, Mário Luzeiro wrote:
> Hi Orson,
>
> I am back on testing this openGL issues.
> I implemented your context manager in my branch.
> It looks good the implementation (still need a bit style clean up your source
> code file, eg: gl_context_mgr.cpp line 86)
>
> I
Hi Orson,
I am back on testing this openGL issues.
I implemented your context manager in my branch.
It looks good the implementation (still need a bit style clean up your source
code file, eg: gl_context_mgr.cpp line 86)
I didn't tried it properly on Windows yet.
I am debugging (on debug mode)
On 06/07/2016 03:41 PM, jp charras wrote:
> Le 07/06/2016 à 15:14, Maciej Sumiński a écrit :
>> On 06/03/2016 05:26 PM, jp charras wrote:
>>> The 3D viewer does not crash.
>>>
>>> Only Opengl crashes if I switch to cairo, or when exiting pcbnew if I am in
>>> legacy mode, and if I
>>> used the
Le 07/06/2016 à 15:14, Maciej Sumiński a écrit :
> On 06/03/2016 05:26 PM, jp charras wrote:
>> The 3D viewer does not crash.
>>
>> Only Opengl crashes if I switch to cairo, or when exiting pcbnew if I am in
>> legacy mode, and if I
>> used the opengl canvas during the session.
>> When exiting,
On 06/03/2016 05:26 PM, jp charras wrote:
> The 3D viewer does not crash.
>
> Only Opengl crashes if I switch to cairo, or when exiting pcbnew if I am in
> legacy mode, and if I
> used the opengl canvas during the session.
> When exiting, if pcbnew is in opengl mode, no crash.
Hi Jean-Pierre,
Le 03/06/2016 à 16:44, Maciej Sumiński a écrit :
> On 06/03/2016 04:40 PM, jp charras wrote:
>> Le 03/06/2016 à 16:04, Maciej Sumiński a écrit :
>>> Hi,
>>>
>>> I have just pushed the mentioned changes.
>>>
>>> @Jean-Pierre,
>>> If you experience crash on exit with OpenGL GAL again, then we can
On 06/03/2016 04:40 PM, jp charras wrote:
> Le 03/06/2016 à 16:04, Maciej Sumiński a écrit :
>> Hi,
>>
>> I have just pushed the mentioned changes.
>>
>> @Jean-Pierre,
>> If you experience crash on exit with OpenGL GAL again, then we can
>> revert commit 6881, but I do not consider this a good
Le 03/06/2016 à 16:04, Maciej Sumiński a écrit :
> Hi,
>
> I have just pushed the mentioned changes.
>
> @Jean-Pierre,
> If you experience crash on exit with OpenGL GAL again, then we can
> revert commit 6881, but I do not consider this a good solution, as it
> creates a memleak. We need to
Hi,
I have just pushed the mentioned changes.
@Jean-Pierre,
If you experience crash on exit with OpenGL GAL again, then we can
revert commit 6881, but I do not consider this a good solution, as it
creates a memleak. We need to think about another way of handling OGL
resources.
Regards,
Orson
Le 01/06/2016 à 15:01, Maciej Sumiński a écrit :
> On 05/31/2016 08:11 PM, jp charras wrote:
>> Thanks, Orson.
>>
>> This version fixes the crash on my W7, 32 bits.
>>
>> I am now trying to compile in debug mode, but I have an undefined symbol:
>>
On 05/31/2016 08:11 PM, jp charras wrote:
> Thanks, Orson.
>
> This version fixes the crash on my W7, 32 bits.
>
> I am now trying to compile in debug mode, but I have an undefined symbol:
> E:/kicad-launchpad/essais/opengl_ctx_fix/include/xnode.h:53: undefined
> reference to `vtable for XNODE'
Le 31/05/2016 à 17:53, Maciej Sumiński a écrit :
> Jean-Pierre, Mario,
>
> I have just finished another approach to the OpenGL contexts problem
> [1]. I will be grateful if you could test it. I have tried it out with 2
> window managers on Linux/64, and now I am rebuilding it on W7/64.
>
>
On 05/29/2016 08:40 PM, jp charras wrote:
[snip]
> Hi, Orson,
>
> About the crash which happens in OPENGL_GAL dtor, when executing delete
> OPENGL_GAL::glContext;, I am
> now pretty sure the crash is due to an use after delete issue.
>
> If it is called when exiting the application, no problem.
Le 09/05/2016 à 17:20, Maciej Sumiński a écrit :
> On 05/05/2016 02:26 PM, jp charras wrote:
>> Hi, Orson.
>>
>> I made a few tests:
>>
>> The crash happens in OPENGL_GAL::~OPENGL_GAL().
>>
>> Exactly when executing delete OPENGL_GAL::glContext;
>>
>> if I comment the 2 lines
>> delete
tested with r6788 after a fresh build and it is fine now!
thanks :)
On 10/05/2016 22.41, Maciej Sumiński wrote:
There was another fix for pcbnew crashing with OpenGL GAL applied in
revision 6781, so there is a chance that your problem is already
resolved. I will be grateful for feedback.
There was another fix for pcbnew crashing with OpenGL GAL applied in
revision 6781, so there is a chance that your problem is already
resolved. I will be grateful for feedback.
Regards,
Orson
On 05/10/2016 09:07 PM, easyw wrote:
> Hi,
>> Are there any particular steps I need to perform? Open
Hi,
Are there any particular steps I need to perform? Open pcbnew empty?
Load a specific board? Use the launcher?
I tried just opening pcbnew empty or with a board but when I switch from
legacy to OpenGl it crashes...The same from the kicad launcher...
I tested till rev 6776 ...
I'm going to
Do you get any error message? I am trying hard to reproduce the problem,
but none of the available PCs expose the bug. I have tried Win7 & Win10
64-bit. Both build with msys2/gcc 5.3, but I still cannot crash pcbnew.
Are there any particular steps I need to perform? Open pcbnew empty?
Load a
Hi, I'm having an issue (crash) changing from default canvas to OpenGl
Comparing my settings to the dev pre-built it seems only gcc version is
different; mine is gcc 5.3.0, pre-built is 5.2.0
is there anyone having this problem?
and here there are the settings:
Application: pcbnew
Version:
On 5/4/2016 5:20 AM, Maciej Sumiński wrote:
> On 05/03/2016 04:48 PM, Wayne Stambaugh wrote:
> [snip]
>> Not quite yet. See attached screen shot on latest msys2/mingw32. The
>> Cairo version is 1.15.2. I suppose it could be the graphics drivers on
>> my work laptop (it's one of the Intel HD
Le 04/05/2016 à 12:23, Maciej Sumiński a écrit :
> On 05/04/2016 11:45 AM, jp charras wrote:
>> Le 04/05/2016 à 11:20, Maciej Sumiński a écrit :
>>> On 05/03/2016 04:48 PM, Wayne Stambaugh wrote:
>>> [snip]
Not quite yet. See attached screen shot on latest msys2/mingw32. The
Cairo
On 05/04/2016 11:45 AM, jp charras wrote:
> Le 04/05/2016 à 11:20, Maciej Sumiński a écrit :
>> On 05/03/2016 04:48 PM, Wayne Stambaugh wrote:
>> [snip]
>>> Not quite yet. See attached screen shot on latest msys2/mingw32. The
>>> Cairo version is 1.15.2. I suppose it could be the graphics
Le 04/05/2016 à 11:20, Maciej Sumiński a écrit :
> On 05/03/2016 04:48 PM, Wayne Stambaugh wrote:
> [snip]
>> Not quite yet. See attached screen shot on latest msys2/mingw32. The
>> Cairo version is 1.15.2. I suppose it could be the graphics drivers on
>> my work laptop (it's one of the Intel
On 05/03/2016 04:48 PM, Wayne Stambaugh wrote:
[snip]
> Not quite yet. See attached screen shot on latest msys2/mingw32. The
> Cairo version is 1.15.2. I suppose it could be the graphics drivers on
> my work laptop (it's one of the Intel HD graphics chips). I'll test it
> on my home system
;
>> On this case shouldn't you not allow the begin of the update at all? (As you
>> will not be able to render it)... but I dont know exactly how it is supposed
>> to work.
>>
>> Regards,
>> Mario
>>
>>
>> From: Kicad-develope
s supposed
> to work.
>
> Regards,
> Mario
>
>
> From: Kicad-developers
> [kicad-developers-bounces+mrluzeiro=ua...@lists.launchpad.net] on behalf of
> Mário Luzeiro [mrluze...@ua.pt]
> Sent: 03 May 2016 15:27
> To: Maciej Sumiński; kicad-developers@lists.launchpad.net
>
t] on behalf of
> Mário Luzeiro [mrluze...@ua.pt]
> Sent: 03 May 2016 15:27
> To: Maciej Sumiński; kicad-developers@lists.launchpad.net
> Subject: Re: [Kicad-developers] mode openGL issues after latest updates
>
> I tested (on Windows) and it works!
> I didn't feel any issue
@lists.launchpad.net
Subject: Re: [Kicad-developers] mode openGL issues after latest updates
I tested (on Windows) and it works!
I didn't feel any issues.
Thanks,
Mario
From: Maciej Sumiński [maciej.sumin...@cern.ch]
Sent: 03 May 2016 15:20
To: Mário Luzeiro
On 05/03/2016 12:06 PM, Mário Luzeiro wrote:
> Hi Orson,
> I tested the new updates on my branch code, and now on Windows I got this
> issues after I close the 3d-viewer windows:
>
> Error: mapping vertices buffer: invalid operation
> Error: unmapping vertices buffer: invalid operation
> Error:
On 05/03/2016 03:23 PM, Wayne Stambaugh wrote:
> On 5/3/2016 6:58 AM, jp charras wrote:
>> Le 03/05/2016 à 12:06, Mário Luzeiro a écrit :
>>> Hi Orson,
>>> I tested the new updates on my branch code, and now on Windows I got this
>>> issues after I close the 3d-viewer windows:
>>>
>>> Error:
Le 03/05/2016 à 12:06, Mário Luzeiro a écrit :
> Hi Orson,
> I tested the new updates on my branch code, and now on Windows I got this
> issues after I close the 3d-viewer windows:
>
> Error: mapping vertices buffer: invalid operation
> Error: unmapping vertices buffer: invalid operation
>
Hi Orson,
I tested the new updates on my branch code, and now on Windows I got this
issues after I close the 3d-viewer windows:
Error: mapping vertices buffer: invalid operation
Error: unmapping vertices buffer: invalid operation
Error: unbinding vertices buffer: invalid operation
My guess is
34 matches
Mail list logo
