Fix dynamic alloc #38
Conversation
…xing accidental comment associated w/this file
…dated comments relate dto functions
…s of the new functions
…ratable parameters
…type input variables.
…for printing out updates relevant to calibratable parameters.
|
Given that #37 is a big functional change that's still in draft mode, would it make sense to
|
hellkite500
force-pushed
the
fix-dynamic-alloc
branch
from
9e83f76
to
2ffd6e4
Compare
| CC = gcc -g | ||
|
|
||
| # define any compile-time flags | ||
| # for newer compiler, the right hand side is not needed... | ||
| # ... for instance, remove the flags if using gnu/10.1.0 on Cheyenne. | ||
| CFLAGS = | ||
| CFLAGS = -fsanitize=address |
There was a problem hiding this comment.
Probably best not to commit changes like this back to the shared repository.
There was a problem hiding this comment.
I put this here intentionally since this makefile is used to build the executable that is run under the CI test runner. Since this saga started with a weird memory issue, I thought it might be helpful to have this as the default used to build and test the code for future changes.
|
Incidentally, the |
|
As far as merge/integration is concerned, I'm happy to rebase this again and reorganize the commits as appropriate/needed. I'm not sure a PR against the feature branch of #37 would help significantly, based on discussions with @Ben-Choat I think that work is nearly ready to merge, and that might complicate things. As long as this doesn't go stale, I can rebase on master once the feature is complete. |
…y+num_time_delay_histo_ords, instead of just num_time_dealy_histo_ords. Also edited a print statement, and an error to provide assistance to user about how to fix error.
…e Q differently depending on if stand alon emode or not. Seeing discrepancies between master stand_alone results and current stand_alone results. Differences exist before these edits.
…ly to output when in stand_alone mode
Refactors the allocation of Q to ensure the correct size for iterating over the histogram size.
Also fixes an accumulation bug where Q was not previously accounting for the time delayed contributions.
Removes hard coded limits and issues warnings when users supply values that may violate original model assumptions.
A few code cleanup/refactors for simplifying and optmization loops.
A few other small bug fixes and comments added.
Note this PR is dependent on #37. It is branched from this commit on the PR branch: cfbdec5