-
Notifications
You must be signed in to change notification settings - Fork 262
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
dap4_test_test_raw not working on cmake builds? #887
Comments
Not seeing this on our end; can you share how you are configuring with cmake, and I'll try to recreate? |
Ed- |
I am seeing this in every cmake build now. Perhaps there's a server down or slow? Here's an example. I checkout the latest master and do:
And I get:
|
I am going to try this again later and see if I get the same result. I may be having some network issues here that are preventing a connection from being made in time, and so something's timing out. |
I don't think that the dap4 tests are doing any remoter server accesses at the moment |
OK, my file system is being hammered by loads of CI testing, and has been all day. I've been playing with the new valgrind testing and it's pretty freaking slow. So I will try again when all my CI tests have stopped and see if I still get this problem. |
OK, total false alarm. Shared library problem. It couldn't find the hdf5 HL library. |
No, it was not a shared library problem. It is something I don't understand. ncdump is segfaulting for the tests in test_raw.sh. This is from a fresh clone of the unidata netCDF-c repo, built like this:
When I run test_raw.sh I see:
How exactly do you guys build with cmake, specifying the directories for HDF5? That is, can you send me what you are using from the command line? |
run test_raw.sh with the -x bash flag set so we can see the exact command that |
I get this:
|
ok, I am testing a fix. I thought I fixed this, but apparently did not |
Make sure that all attempts to access the rc file triples check first for triple->host being null.
Hi Ed, typically I invoke as follows, from the root netcdf-c directory:
The path I specify via |
OK thanks! I will try that next time. |
I'm working on a presentation for COSMIC tomorrow, but I'm going to try to roll a bunch of outstanding PR's into an aggregation, this afternoon, so that we can get 4.6.1 out and move on to the next. Mentioning it here because it's as good a place as any :) |
Awesome. |
This issue should be fixed by pr #888 |
Yes, I merged #888 into HPC netCDF and this issue is now fixed. Thanks! |
When I check out current master, and do a cmake build, I am now seeing this failure:
This doesn't seem to happen with autotool builds. This seems new but I can't really say.
The text was updated successfully, but these errors were encountered: