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
MPI tests fail at test_close_multiple_mpio_driver #1285
Comments
The double mpi4py gives a pattern to use at https://bitbucket.org/mpi4py/mpi4py/src/master/test/test_io.py Here is a patch. For good measure, the patch replaces mktemp (which is deprecated, possibly unsafe) with mkstemp.
|
Actually my patch to mktemp() may cause other problems, e.g.
So handle any update from Best to apply the MPI patch for test_file.py separately from handling the mkstemp patch, i.e. just
|
There were a couple of other tests using mpio which needed fixing also. This patch seems to catch them all:
|
Do you want to turn the patches into a pull request? |
Can do. |
When h5py is built with mpi, run_tests() fail:
The log shows that the first processor closed the mpio file successfully, while the second processor emitted
error in writev:Bad file descriptor
.I expect this is caused by the double
f.close()
intest_close_multiple_mpio_driver
. Is there a reason for runningf.close()
twice? It doesn't seem to be doing the mpio test any good.This is testing on Debian,
Summary of the h5py configuration
h5py 2.9.0
HDF5 1.10.4
Python 3.7.4 (default, Jul 11 2019, 10:43:21)
[GCC 8.3.0]
sys.platform linux
sys.maxsize 9223372036854775807
numpy 1.16.2
The text was updated successfully, but these errors were encountered: