Big Count Collectives Test Suite #15
Conversation
jjhursey
force-pushed
the
big-count-coll-testing
branch
from
db878fb to
d73c295
Compare
| # count = TEST_PAYLOAD_SIZE / sizeof(datatype) | ||
| ###################################################################### | ||
| # INT_MAX : == 2 GB so guard will not trip | ||
| TEST_PAYLOAD_SIZE=2147483647 |
There was a problem hiding this comment.
Don't we want this TEST_PAYLOAD_SIZE to be BIGGER than TEST_UNIFORM_COUNT? say 4 GB or 4GB + a few KB?
README.md (at the top) seems to suggest that the default TEST_PAYLOAD_SIZE would be UINT_MAX ???
See: https://github.com/open-mpi/ompi-tests-public/pull/15/files#diff-26cf9c4e5070bbfe5af943597e4ee73210cb78749e416e5bcb00fb204331f490R98
There was a problem hiding this comment.
You can, but it should be relative to the datatype. So we don't want to set it to INT_MAX x 8 because then the count would be too big for an int datatype. I left it as a default of INT_MAX and folks can change it as they see fit. The TEST_UNIFORM_COUNT is probably the more useful of the two variants.
* Run collectives with a `count` parameter as close to `INT_MAX` as possible.
- This test suite often highlights cases where the underlying algorithm
assumes that the payload (roughly `count x sizeof(datatype)`) is
and `int` when it should be handled as a `size_t`.
* Includes:
- Test with `int` and `double _Complex` primitive data types
- Correctness checks
- Mechanism to control (as best as we can) the amount of memory consumed
per node.
* Assumes:
- Roughly the same amount of memory per node
- Same number of processes per node
* See `README.md` for details
Signed-off-by: Joshua Hursey <jhursey@us.ibm.com>
jjhursey
force-pushed
the
big-count-coll-testing
branch
from
d73c295 to
6332a0e
Compare
|
I just force pushed the resolved changes from the review. |
|
Discussed on March 1, 2022 teleconf. Ok to merge. |
countparameter as close toINT_MAXas possible.assumes that the payload (roughly
count x sizeof(datatype)) isand
intwhen it should be handled as asize_t.intanddouble _Complexprimitive data typesper node.
README.mdfor detailsSigned-off-by: Joshua Hursey jhursey@us.ibm.com