Bug fix: Don't forget closing files#1083
Merged
ax3l merged 2 commits intoopenPMD:devfrom Aug 11, 2021
Merged
Conversation
franzpoeschel
force-pushed
the
fix-dont-forget-closing-files
branch
2 times, most recently
from
e0def7e to
d2435db
Compare
franzpoeschel
force-pushed
the
fix-dont-forget-closing-files
branch
from
d2435db to
3378311
Compare
Contributor
Author
|
I've added a failing test now and will push the fix again after the CI has run |
Contributor
Author
|
This fails less tests than I had hoped for, but it does fail some. (Also, it fails for me locally) |
ax3l
approved these changes
Aug 11, 2021
| if( IOHandler()->m_frontendAccess == Access::READ_ONLY ) | ||
| for( auto it = begin; it != end; ++it ) | ||
| { | ||
| // Phase 1 |
Member
There was a problem hiding this comment.
What does Phase mean in this context? :)
Contributor
Author
There was a problem hiding this comment.
It's just the 3 discrete steps that each iteration of the loop consists of: Flushing the frontend, closing the file, flushing the backend. I just added those comments as an optical guideline to find things faster.
Member
There was a problem hiding this comment.
Cool, thanks! Do we maybe want to add this to the comment, e.g. "Phase 1: flush frontend" etc.? Good to keep the context in such long logic chains :)
ax3l
pushed a commit
to ax3l/openPMD-api
that referenced
this pull request
Aug 18, 2021
* Failing test * Bug fix: Don't forget closing files
ax3l
added a commit
that referenced
this pull request
Aug 18, 2021
* Bug fix: Don't forget closing files (#1083) * Failing test * Bug fix: Don't forget closing files * HDF5: Fix String Vlen Attribute Reads (#1084) We inofficially try to also support HDF5 variable lengths strings in reading, just to increase compatibility. This was never working it seems. * Fix reading of vector attributes with only one contained value (#1085) * Failing test * Conversions in Attribute.hpp 1) single values to 1-value vectors 2) vectors to arrays 3) arrays to vectors * Some cleanup in Attribute.hpp 1) Simplify types in DoConvert, remove unnecessary template parameter 2) Replace a long if-then-else chain by variantSrc::visit * CoreTest: Fix std::array constructors Make more widely compile-able. * Explicit casting in some places This avoids some warnings * Intel compilers: Don't use variantSrc::visit They don't like it * Remove scattered checks for vector attributes * Generalize icpc guard As defined in CMake for compiler identification. * Doc ICC version (2021.3.0) * setAttribute: Reject Empty Strings (#1087) * setAttribute: Reject Empty Strings Some backends, especially HDF5, do not allow us to define zero-sized strings. We thus need to catch this in the frontend and forward the restriction to the user. * Test: setAttribute("key", "") throws * setAttribute Check: C++14 compatible * Don't read iterations if they have already been parsed (#1089) * Release: 0.14.2 Co-authored-by: Franz Pöschel <franz.poeschel@gmail.com>
Member
|
I already applied this to 0.14.2 :) |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Noticed while working on #949, our refactored flush logic will sometimes forget to close a file
TODO
Probablyno it does notIterationOpenedneeds a third entryCloseMeor something