#820: close all HDF5 objects in EOS Omni table reader -------------------+-------------------------------------------------------- Reporter: rhaas | Type: defect Status: new | Priority: minor Milestone: | Component: EinsteinToolkit thorn Version: | Keywords: -------------------+-------------------------------------------------------- hdf5 does not automatically close all object handles of objects in a file if one closes the (last) file handle. This could potentially lead hdf5 to leave metadata in memory (and it is just not tidy to leave open objects around)
#820: close all HDF5 objects in EOS Omni table reader ------------------------------------+--------------------------------------- Reporter: rhaas | Owner: Type: defect | Status: review Priority: minor | Milestone: Component: EinsteinToolkit thorn | Version: Resolution: | Keywords: ------------------------------------+--------------------------------------- Changes (by rhaas):
* status: new => review
#820: close all HDF5 objects in EOS Omni table reader ------------------------------------+--------------------------------------- Reporter: rhaas | Owner: Type: defect | Status: review Priority: minor | Milestone: Component: EinsteinToolkit thorn | Version: Resolution: | Keywords: ------------------------------------+---------------------------------------
Comment (by knarf):
This patch also declares a function static and removes another function declaration. Is this intentional and related to the problem? It doesn't seem to be.
#820: close all HDF5 objects in EOS Omni table reader ------------------------------------+--------------------------------------- Reporter: rhaas | Owner: Type: defect | Status: review Priority: minor | Milestone: Component: EinsteinToolkit thorn | Version: Resolution: | Keywords: ------------------------------------+---------------------------------------
Comment (by rhaas):
It's intentional. It does not relate to the problem. It should be done though, since pushing a function with the rather common name "file_is_readable" into the global namespace seems dangerous. It removes a function declaration that Cactus now provides on its own. You are certainly correct though. Lines 133 and 134 are the only changes that are actually mentioned in the ticket title.
#820: close all HDF5 objects in EOS Omni table reader ------------------------------------+--------------------------------------- Reporter: rhaas | Owner: Type: defect | Status: reviewed_ok Priority: minor | Milestone: Component: EinsteinToolkit thorn | Version: Resolution: | Keywords: ------------------------------------+--------------------------------------- Changes (by knarf):
* status: review => reviewed_ok
Comment:
Was just checking - ok to commit the whole patch, thanks!
#820: close all HDF5 objects in EOS Omni table reader ------------------------------------+--------------------------------------- Reporter: rhaas | Owner: Type: defect | Status: reviewed_ok Priority: minor | Milestone: Component: EinsteinToolkit thorn | Version: Resolution: | Keywords: ------------------------------------+---------------------------------------
Comment (by rhaas):
Hello Frank. Sorry to sound so grumpy. That was not intended. I had written the comment in the two minutes before leaving for a talk. I'll wait until after the release with since it is not critical in any way.
#820: close all HDF5 objects in EOS Omni table reader ------------------------------------+--------------------------------------- Reporter: rhaas | Owner: Type: defect | Status: closed Priority: minor | Milestone: Component: EinsteinToolkit thorn | Version: Resolution: fixed | Keywords: ------------------------------------+--------------------------------------- Changes (by rhaas):
* status: reviewed_ok => closed * resolution: => fixed
Comment:
Sorry. I had confused #821 and this one. I applied this patch since it is so trivial. Thanks for the review.
trac@lists.einsteintoolkit.org
-
Einstein Toolkit