Ticket #11592 (closed: worksforme)
ParaView 4.3 testing Linux
Reported by: | Owen Arnold | Owned by: | Federico M Pouzols |
---|---|---|---|
Priority: | blocker | Milestone: | Release 3.4 |
Component: | Framework | Keywords: | vates |
Cc: | Blocked By: | ||
Blocking: | Tester: |
Description (last modified by Owen Arnold) (diff)
- Install and uninstall
- Cube axes
- Re-binning
- Menus
- View switching.
- Default views and backgrounds via Preferences
- Overplotting Peaks Workspaces, zooming and selecting
- Drag and drop
- Launching the slice viewer from the multislice mode
Please report what flavour of linux you used in your test section for this ticket.
Change History
comment:1 Changed 5 years ago by Owen Arnold
- Status changed from new to assigned
- Owner changed from Nick Draper to Federico M Pouzols
- Description modified (diff)
- Summary changed from ParaView 4.3 testing Windows 8 (cloned) to ParaView 4.3 testing Linux
comment:2 Changed 5 years ago by Federico M Pouzols
I'm testing on debian and Windows 7. For now I checked install/uninstall and basic usage with small datasets. Everything seems to work well. Menus, MD workspaces, peaks, different views, etc.
I really tried to fool it with another ParaView version that I had installed and it behaved solid like a rock.
The only extremely nit picky thing that I noticed is that if you try to save a screenshot when there's nothing being displayed it will show you an "it failed" error, but will not tell you why - or prevent you from saving nothing.
comment:3 Changed 5 years ago by Owen Arnold
Thanks for this. Can you open a ticket about that screenshot issue and assign to Anton/cc Steve Hahn. We ought to try to fix that.
comment:4 Changed 5 years ago by Federico M Pouzols
Tested nightly binaries 3.3.20150417.2020 on debian and windows 7. I also tested binaries built from the source distribution on debian.
I'd say with very high confidence this is all very robust, with the only exception of the shift-click slice viewer issue. It seems to work well with histo and event workspaces, peaks data, and file-backed workspaces. There doesn't seem to be any platform specific issue as far as I could see. I'm less confident about the multiple Rebin options. They all seem to behave but I don't know well enough their usage/behavior to tell with confidence whether the results are fine.
I tried to make it crash and/or misbehave with all the points in the list and anything else that I could perversely find in the GUI. Other functionality that seems ok: the time/frames play controls, help, using and abusing the tree of sources. All the tests were done with smallish files from the test data, although I also did a bit of torturing with a very large SQW file.
comment:5 Changed 5 years ago by Federico M Pouzols
Well, in the end I managed to get a couple of crashes. One on windows and one on linux. But they are hardly reproducible and could well be related to my graphic card woes. Just noting them here in case they ring any bell:
- On windows, crash when switching from "Scatter plot mode" to "Three slice mode" while visualizing as surface an MD workspace coming from SEQ_MDEW that had been rebinned and sliced a bit randomly.
- On debian, crash while visualizing around ten different sources simultaneously, it seem to crash after a few rebins and perspective changes. This is the trace that I got:
#0 0x00007fffc5cfbb30 in pqProxy::getProxy() const () from /home/fedemp/build-mantidnightly/build-paraview/lib/libvtkpqCore-pv4.3.so.1 #1 0x00007fffc5d09751 in pqScalarsToColors::getScalarRange() const () from /home/fedemp/build-mantidnightly/build-paraview/lib/libvtkpqCore-pv4.3.so.1 #2 0x00007fffc70d47ef in Mantid::Vates::SimpleGui::AutoScaleRangeGenerator::setMinBufferAndMaxBuffer(pqPipelineSource*, double&, double&) () from /home/fedemp/build-mantidnightly/build-mantid/bin//libMantidVatesSimpleGuiViewWidgets.so #3 0x00007fffc70d4e47 in Mantid::Vates::SimpleGui::AutoScaleRangeGenerator::getStandardColorScale() () from /home/fedemp/build-mantidnightly/build-mantid/bin//libMantidVatesSimpleGuiViewWidgets.so #4 0x00007fffc70d5078 in Mantid::Vates::SimpleGui::AutoScaleRangeGenerator::getColorScale() () from /home/fedemp/build-mantidnightly/build-mantid/bin//libMantidVatesSimpleGuiViewWidgets.so #5 0x00007fffc70dae05 in Mantid::Vates::SimpleGui::ColorUpdater::autoScale() () from /home/fedemp/build-mantidnightly/build-mantid/bin//libMantidVatesSimpleGuiViewWidgets.so #6 0x00007fffc7109c84 in Mantid::Vates::SimpleGui::ViewBase::setAutoColorScale() () from /home/fedemp/build-mantidnightly/build-mantid/bin//libMantidVatesSimpleGuiViewWidgets.so #7 0x00007fffc710a968 in Mantid::Vates::SimpleGui::ViewBase::onColorMapChange(pqColorMapModel const*) () from /home/fedemp/build-mantidnightly/build-mantid/bin//libMantidVatesSimpleGuiViewWidgets.so #8 0x00007fffc710ad35 in Mantid::Vates::SimpleGui::ViewBase::onVisibilityChanged(pqPipelineSource*, pqDataRepresentation*) () from /home/fedemp/build-mantidnightly/build-mantid/bin//libMantidVatesSimpleGuiViewWidgets.so #9 0x00007fffc710ecef in Mantid::Vates::SimpleGui::ViewBase::qt_static_metacall(QObject*, QMetaObject::Call, int, void**) () from /home/fedemp/build-mantidnightly/build-mantid/bin//libMantidVatesSimpleGuiViewWidgets.so #10 0x00007ffff39e4f4c in QMetaObject::activate(QObject*, QMetaObject const*, int, void**) () from /usr/lib/x86_64-linux-gnu/libQtCore.so.4 #11 0x00007fffc5d39e0b in pqPipelineSource::visibilityChanged(pqPipelineSource*, pqDataRepresentation*) () from /home/fedemp/build-mantidnightly/build-paraview/lib/libvtkpqCore-pv4.3.so.1 #12 0x00007ffff39e4f4c in QMetaObject::activate(QObject*, QMetaObject const*, int, void**) () from /usr/lib/x86_64-linux-gnu/libQtCore.so.4 #13 0x00007fffc5d38f4b in pqOutputPort::visibilityChanged(pqOutputPort*, pqDataRepresentation*) () from /home/fedemp/build-mantidnightly/build-paraview/lib/libvtkpqCore-pv4.3.so.1 #14 0x00007ffff39e4f4c in QMetaObject::activate(QObject*, QMetaObject const*, int, void**) () from /usr/lib/x86_64-linux-gnu/libQtCore.so.4 #15 0x00007fffc5d3bacf in pqRepresentation::visibilityChanged(bool) () from /home/fedemp/build-mantidnightly/build-paraview/lib/libvtkpqCore-pv4.3.so.1 #16 0x00007fffc5d08984 in pqRepresentation::onVisibilityChanged() () from /home/fedemp/build-mantidnightly/build-paraview/lib/libvtkpqCore-pv4.3.so.1 #17 0x00007ffff39e9641 in QObject::event(QEvent*) () from /usr/lib/x86_64-linux-gnu/libQtCore.so.4 #18 0x00007ffff46d929c in QApplicationPrivate::notify_helper(QObject*, QEvent*) () from /usr/lib/x86_64-linux-gnu/libQtGui.so.4 #19 0x00007ffff46dfda8 in QApplication::notify(QObject*, QEvent*) () from /usr/lib/x86_64-linux-gnu/libQtGui.so.4 #20 0x000000000086ff19 in MantidApplication::notify(QObject*, QEvent*) () #21 0x00007ffff39cff8d in QCoreApplication::notifyInternal(QObject*, QEvent*) () from /usr/lib/x86_64-linux-gnu/libQtCore.so.4 #22 0x00007ffff39d3821 in QCoreApplicationPrivate::sendPostedEvents(QObject*, int, QThreadData*) () from /usr/lib/x86_64-linux-gnu/libQtCore.so.4
comment:6 Changed 5 years ago by Federico M Pouzols
Note on the status of this ticket:
- As far as I can tell all the issues found here, and related ones, were fixed last week. It overall looked solid and well behaved.
I'd add that I used to experience many random crashes with the older ParaView (no acceleration graphic card on Windows because of hardware issues, and virtual graphic card on Linux). And I think that 4.1 was still a bit unreliable. This issue seems to have gone away with the newer ParaView version used now in the VSI.
comment:7 Changed 5 years ago by Federico M Pouzols
- Status changed from assigned to verify
- Resolution set to fixed
Ask Owen Arnold for data if you need it.