einsteintoolkit / tickets / issues / #1023 - Improve OpenMP parallelisation of SummationByParts — Bitbucket

Issue #1023 closed

Erik Schnetter created an issue 2012-08-08

The Intel compiler does not handle workshare constructs well. The attached patch replaces them by explicit loops, which execute faster. This makes a measurable difference on Hopper with 24 OpenMP threads.

This only modifies one operator; other operators could be treated in the same way.

Keyword:

sbp-openmp.diff

Comments (4)

Erik Schnetter reporter
- changed status to open
- removed comment
- 2012-08-08T15:48:17+00:00
Frank Löffler
- changed status to open
- removed comment
The patch looks ok. I didn't check all the indices really carefully (due to the length of the patch) and didn't run testsuites. Assuming tests show no difference between both versions using multiple threads I think it is ok to commit this. I'll leave testing to Erik. :)
- 2012-09-12T15:25:47+00:00
Erik Schnetter reporter
- changed status to resolved
- removed comment
Applied.
- 2012-09-13T14:39:12+00:00
Roland Haas
- edited description
- changed status to closed
- 2019-02-21T20:13:35+00:00
Log in to comment

Assignee: –

Type: enhancement

Priority: minor

Status: closed

Component: EinsteinToolkit thorn

Milestone: –

Version: –

Votes: 0

Watchers: 0

Jira Software: the preferred issue tracker for Bitbucket. Join the team!