Improve mesh batch nrm2: add OpenMP parallelization for packed batches
All threads resolved!
All threads resolved!
Description
Add OpenMP parallelization for the norm calculation (mesh_batch_nrm2).
Checklist
-
I have checked that my code follows the Octopus coding standards -
I have added tests for all the new features added in this request.
Merge request reports
Activity
added Optimization label
Codecov Report
Merging #882 into develop will decrease coverage by
0.00%
. The diff coverage is63.88%
.@@ Coverage Diff @@ ## develop #882 +/- ## =========================================== - Coverage 70.76% 70.76% -0.01% =========================================== Files 501 501 Lines 96456 96472 +16 =========================================== + Hits 68258 68268 +10 - Misses 28198 28204 +6
Impacted Files Coverage Δ src/grid/mesh_batch.F90 100.00% <ø> (ø)
src/grid/mesh_batch_inc.F90 88.44% <63.88%> (-0.75%)
Continue to review full report at Codecov.
Legend - Click here to learn more
Δ = absolute <relative> (impact)
,ø = not affected
,? = missing data
Powered by Codecov. Last update e0511a4...6072f27. Read the comment docs.Edited by Codecovadded 25 commits
-
b737afdf...e0511a41 - 23 commits from branch
develop
- e066665d - Improve nrm2: less calls to abs
- 5dc40ba3 - OpenMP version of mesh_batch_nrm2 for packed states
-
b737afdf...e0511a41 - 23 commits from branch
added 1 commit
- e03fac84 - OpenMP version of mesh_batch_nrm2 for packed states
assigned to @nicolastd
- Resolved by Nicolas Tancogne-Dejean
- Resolved by Nicolas Tancogne-Dejean
- Resolved by Nicolas Tancogne-Dejean
- Resolved by Nicolas Tancogne-Dejean
added 1 commit
- 6072f272 - Fix logic of checking if component in sum is non-zero
@nicolastd the pipeline now passes with the most recent fix.
mentioned in commit 0f2a5bad
mentioned in issue #328 (closed)
changed milestone to %10.0
added in Changelog label
Please register or sign in to reply