Fix a bug in Subfiling VFD vector I/O setup #4821
Merged
Conversation
This file contains 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
jhendersonHDF
added
Merge - To 1.14
Priority - 0. Blocker ⛔
jhendersonHDF
requested review from
lrknox,
derobins,
byrnHDF,
fortnern,
qkoziol,
vchoi-hdfgroup,
bmribler,
glennsong09,
mattjala and
brtnfld
as code owners
jhendersonHDF
force-pushed
the
subfiling_vector_io_bugfix
branch
from
6d87425
to
9a3ccd2
Compare
Fixes a bug where the vector I/O sizes weren't being extended when one of the entries in the array is 0. This caused an over-read of the I/O sizes buffer and on some machines would cause a memory allocation failure due to the calculated I/O vector size being too large.
|
Will add a release note after checking testing on an HPC machine. The current testing in t_vfd.c wasn't generally catching this as the I/O sizes array was allocated on the stack and passed in. The buffer over-read was then usually reading valid memory for the process. Adding a test with an array allocated on the heap has a much better chance of triggering the issue. |
|
MPI_TEST_t_vfd and 2930 other tests all passed on frontier, testing the subfiling_vector_io_bugfix branch with gcc-native/13.2 compiler: https://my.cdash.org/builds/2656168. The test has been failing consistently in this configuration. |
qkoziol
approved these changes
Sep 11, 2024
lrknox
approved these changes
Sep 12, 2024
lrknox
pushed a commit
to lrknox/hdf5
that referenced
this pull request
Sep 13, 2024
Fixes a bug where the vector I/O sizes weren't being extended when one of the entries in the array is 0. This caused an over-read of the I/O sizes buffer and on some machines would cause a memory allocation failure due to the calculated I/O vector size being too large.
derobins
pushed a commit
that referenced
this pull request
Sep 16, 2024
* Convert exec_program to execute_process (#4819) * Fix a bug in Subfiling VFD vector I/O setup (#4821) Fixes a bug where the vector I/O sizes weren't being extended when one of the entries in the array is 0. This caused an over-read of the I/O sizes buffer and on some machines would cause a memory allocation failure due to the calculated I/O vector size being too large. * Add release note for signed binaries (#4826) * Draw attention to 3rd step in process to update so numbers for release. (#4825) * Update RESULT_VARIABLE
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
Fixes a bug where the vector I/O sizes weren't being extended when one of the entries in the array is 0. This caused an overread of the I/O sizes buffer and on some machines would cause a memory allocation failure due to the calculated I/O vector size being too large.