[Qgis-user] QGIS 2.18.2 Crashes often on Mac (Yosemite 10.10.5)
NPE NewPathEconometrics
newpatheconometrics at gmail.com
Wed Mar 29 18:39:56 PDT 2017
The relevant response from John H was:
"Your stack indicates the Qt crashed trying to draw an ellipse, as
invoked by QGIS while attempting to start rendering of a marker symbol.
This is probably a straightforward crash to resolve by someone familiar
with the code."
I'm new to using these lists and fairly new to QGIS. Should I file this as
a bug with the QGIS group?
On Wed, Mar 29, 2017 at 6:37 PM, NPE NewPathEconometrics <
newpatheconometrics at gmail.com> wrote:
> Being new to using these lists, I accidentally responded just to a person,
> and not the list, so I'm forwarding my response to the list
> ---------- Forwarded message ----------
> From: NPE NewPathEconometrics <newpatheconometrics at gmail.com>
> Date: Sun, Mar 26, 2017 at 10:16 AM
> Subject: Re: [Qgis-user] QGIS 2.18.2 Crashes often on Mac (Yosemite
> 10.10.5)
> To: John Hawkinson <jhawk at mit.edu>
>
>
> Thank you for the thoughtful suggestions. I don't have a lot of experience
> digging into these types of crash reports, and I wasn't sure if the
> information in that report would be of interest to anybody. I pasted the
> full report into pastebin as you suggested (thanks for that idea, I had not
> heard of pastebin before).
>
> http://pastebin.com/Jcb4FH04
>
> This time, it lists thread 27 as the crashed thread. I looked through the
> report for details of that, and see this information, thought I don't
> really understand what this tells me, or how to resolve this.
>
>
> 1. Thread 27 Crashed:: Thread (pooled)
> 2. 0 QtGui 0x0000000111fb946b
> QPaintEngine::drawEllipse(QRectF const&) + 75
> 3. 1 QtGui 0x0000000111fb8409
> QPaintEngine::drawEllipse(QRect const&) + 73
> 4. 2 QtGui 0x0000000111fccd0c
> QPainter::begin(QPaintDevice*) + 476
> 5. 3 org.qgis.qgis2_core 0x0000000110ea178b
> QgsSimpleMarkerSymbolLayerV2::prepareCache(QgsSymbolV2RenderContext&)
> + 405
> 6. 4 org.qgis.qgis2_core 0x0000000110ea150e
> QgsSimpleMarkerSymbolLayerV2::startRender(QgsSymbolV2RenderContext&) +
> 1276
> 7. 5 org.qgis.qgis2_core 0x0000000110f0f6fa
> QgsSymbolV2::startRender(QgsRenderContext&, QgsFields const*) + 660
> 8. 6 org.qgis.qgis2_core 0x0000000110e2ee5a
> QgsCategorizedSymbolRendererV2::startRender(QgsRenderContext&,
> QgsFields const&) + 548
> 9. 7 org.qgis.qgis2_core 0x000000011120f000
> QgsVectorLayerRenderer::render() + 194
> 10. 8 org.qgis.qgis2_core 0x00000001110c15e3
> QgsMapRendererParallelJob::renderLayerStatic(LayerRenderJob&) + 59
> 11. 9 org.qgis.qgis2_core 0x00000001110c1fb8
> QtConcurrent::MapKernel<QList<LayerRenderJob>::iterator,
> QtConcurrent::FunctionWrapper1<void, LayerRenderJob&>
> >::runIteration(QList<LayerRenderJob>::iterator, int, void*) + 16
> 12. 10 org.qgis.qgis2_core 0x00000001110c1ffc
> QtConcurrent::MapKernel<QList<LayerRenderJob>::iterator,
> QtConcurrent::FunctionWrapper1<void, LayerRenderJob&>
> >::runIterations(QList<LayerRenderJob>::iterator, int, int, void*) + 64
> 13. 11 org.qgis.qgis2_core 0x00000001110c213b
> QtConcurrent::IterateKernel<QList<LayerRenderJob>::iterator,
> void>::forThreadFunction() + 227
> 14. 12 QtCore 0x0000000111a97559
> QtConcurrent::ThreadEngineBase::run() + 57
> 15. 13 QtCore 0x0000000111a9a32d
> QThreadPoolThread::run() + 205
> 16. 14 QtCore 0x0000000111aa73f4
> QThreadPrivate::start(void*) + 356
> 17. 15 libsystem_pthread.dylib 0x00007fff8e6dd05a
> _pthread_body + 131
> 18. 16 libsystem_pthread.dylib 0x00007fff8e6dcfd7
> _pthread_start + 176
> 19. 17 libsystem_pthread.dylib 0x00007fff8e6da3ed
> thread_start + 13
> 20.
>
>
>
> On Sat, Mar 25, 2017 at 4:18 PM, John Hawkinson <jhawk at mit.edu> wrote:
>
>> I don't have a lot of QGIS experience (none prior to 2.18), and I have
>> seen
>> more crashes than I would like (mostly related to Cocoa), but I would
>> offer some general comments:
>>
>> > I did note this warning there: *WARNING: QGIS will crash if Qt4
>> > developer components are installed in the standard /Developer
>> > location. Either rename /Developer/Applications/Qt/Plugins or
>> > uninstall Qt before running QGIS.*
>> >
>> > I don't have Qt4 developer installed there, but I'm wondering why QGIS
>> is
>> > crashing so often, and this warning is the best clue I have.
>>
>> This is not related. Both because you don't have Qt4 developer stuff
>> installed, and also because I believe the warning is about a deterministic
>> crash -- that QGIS will *always* crash if those incompatible components
>> are
>> installed, not that it will be unstable, which is what you are saying.
>>
>> > I did put a post on Stack Exchange over a week ago, with some of the
>> crash
>> > information:
>> > http://gis.stackexchange.com/questions/230576/qgis-crashes-f
>> requently-since-upgrading-to-2-18-2-on-mac
>> > No responses from that post.
>>
>> There is not enough information in your post to be useful.
>> You write:
>>
>> | (I'm not pasting the entire report because it is pretty long):
>>
>> You should probably get over this, or at least get some background
>> on what information is important. It is not always the beginning
>> of the report. It might be helpful to post the report to pastebin.com
>> or gist.github.com and provide a link, though.
>>
>> Specifically, when your crash report says:
>>
>> > Crashed Thread: 18 Thread (pooled)
>>
>> Then the stack trace for Thread 18 is what's of interest. In your
>> report, you included only the first thread (Thread 0), and only the
>> first 11 (0-10) frames of that stack. To be helpful, you should include
>> the entirety of the stack of the thread in question.
>>
>> --jhawk at mit.edu
>> John Hawkinson
>>
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/qgis-user/attachments/20170329/5b13f76c/attachment.html>
More information about the Qgis-user
mailing list