-
Notifications
You must be signed in to change notification settings - Fork 60
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
CMB profile #676
base: master
Are you sure you want to change the base?
CMB profile #676
Conversation
@robertmaynard the current status is that all dependencies build on my RHEL6 system (I use gcc 4.9.2 from Hashstack). The CMB package fails with:
Which suggests that we have incompatible VTK in Hashstack. Do you know what exact version of VTK and CMB works together? I am using your original versions, and I can update to the latest versions, unless there needs to be a specific version that works. |
I believe the ParaView version CMB requires On Wed, Feb 25, 2015 at 1:53 PM, Ondřej Čertík [email protected]
|
Thanks! I have rebuilt Paraview with 55f3b0ed365b5a14a24ab642b5f5aee0483c1096, but I still get exactly the same error with CMB. I'll investigate what's going on with the |
MAXIMUM_NUMBER_OF_PIECES was removed in VTK 6.2, so it looks like you On Wed, Feb 25, 2015 at 2:20 PM, Ondřej Čertík [email protected] wrote:
|
I went into ParaView repository, checked out commit 55f3b0ed365b5a14a24ab642b5f5aee0483c1096, did
it checks out VTK commit a5daf047a2d480eb465e2dcd50c362674fefe326. |
hmmm this is pretty old than. Looks On Wed, Feb 25, 2015 at 2:39 PM, Ondřej Čertík [email protected]
|
I used e90304c0053a68ab316790748e097d810cb17ed1 (unfortunately had to put your ParaView patches back in...) and it fixed the problem. Now I am facing the following problem:
Note that I use mesa, so I use: diff --git a/pkgs/cmb.yaml b/pkgs/cmb.yaml
index 97fd653..f3f81ea 100644
--- a/pkgs/cmb.yaml
+++ b/pkgs/cmb.yaml
@@ -3,7 +3,7 @@ dependencies:
build: [boost, png, qt, zlib, gdal, hdf5,
mpi, netcdf4, netcdf4cpp, freetype,
python, numpy, matplotlib, libxml2,
- paraview, remus, vxl, kml, molequeue, smtk, zmq]
+ paraview, remus, vxl, kml, molequeue, smtk, zmq, mesa]
sources:
- key: git:b24225a52187a124f5f83bdb7212d363f9254519 I need to debug this a bit to see what's going on. ParaView builds just fine. |
Even though CMake is supplied with (via `base/cmake_package.py`) -DCMAKE_PREFIX_PATH="${BOOST_DIR};${BZIP2_DIR};..." For some reason, some of the `-I` statements do not get propagated to g++. If we add the line -DCMAKE_CXX_FLAGS:STRING="${CPPFLAGS}" Then all the necessary `-I` statements are propagated. Note the quotation marks around "${CPPFLAGS}$, those are necessary, since there are spaces in the variable.
As of cdfe644, CMB now builds (I used the following profile: https://gist.github.com/certik/570355076297d48438ce). The problem was fixed by db4a231. @robertmaynard it looks like a bug in CMake or the CMakeLists.txt, see the commit log. I haven't figured out the real cause, just a workaround. I used the latest CMB commit (b24225a52187a124f5f83bdb7212d363f9254519). @robertmaynard, how hard would it be to fix CMB, so that it builds against some newer version of ParaView, so that we don't need to ship the temporary patches again? Any other comments to this PR? |
Kitware just opensourced the latest version of CMB, the latest version of dependencies at https://gitlab.kitware.com/cmb/cmb-superbuild/blob/master/versions.cmake So now we just need to update this PR. |
I took the
cmb-profile
branch from https://github.com/robertmaynard/hashstack/tree/cmb-profile/, rebased it and fixed packages to build. As of cdfe644, everything builds nicely and seems to be working fine. The only problem is that I had to downgrade ParaView (and re-introduce all the patches that have been since merged into upstream ParaView).TODO: