-
Notifications
You must be signed in to change notification settings - Fork 118
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Fix for certain modelDB CI failures #3194
Conversation
✔️ e93caaa -> Azure artifacts URL |
e93caaa
to
ea049c1
Compare
Codecov ReportAll modified and coverable lines are covered by tests ✅
Additional details and impacted files@@ Coverage Diff @@
## master #3194 +/- ##
=======================================
Coverage 67.06% 67.07%
=======================================
Files 570 569 -1
Lines 111232 111206 -26
=======================================
- Hits 74600 74592 -8
+ Misses 36632 36614 -18 ☔ View full report in Codecov by Sentry. |
✔️ ea049c1 -> Azure artifacts URL |
This reverts commit bea2ceb.
✔️ bea2ceb -> Azure artifacts URL |
Quality Gate passedIssues Measures |
✔️ 3e7434c -> Azure artifacts URL |
For context, #3135 introduced a level of indirection to
std::pow
, which seems to be problematic for some models (model IDs:149100 3800 113997 144520 126467 256388
), in particular, when using the-O2
optimization flag (see #3137 (comment), #3137 (comment), #3137 (comment)).The solution is to include
nrnconf.h
, and put the entire NOCMODL change under an#if
guard, sincenrnconf.h
has a#define NRN_ENABLE_ARCH_INDEP_EXP_POW [0/1]
, and the#if
guard appears later, so it works in either case.The azure URL of the successful run (I disabled all CI to build it faster): #3194 (comment)
The modelDB run: https://github.com/neuronsimulator/nrn-modeldb-ci/actions/runs/11782044837/job/32816199196
Note that I am using #3163 as the baseline (this azure URL: #3163 (comment)) due to #3093, which causes some other models to fail due to fact that 8.2.6 doesn't use
TABLE
forhh.mod
, while currentmaster
does, but this does not have an impact on the models mentioned above.