On 13 May 2015, at 11:38, Barry Wardell barry.wardell@gmail.com wrote:
On Wed, May 13, 2015 at 10:28 AM, Bruno Giacomazzo bruno.giacomazzo@unitn.it wrote:
On 13 May 2015, at 11:17, Ian Hinder ian.hinder@aei.mpg.de wrote:
Allow the user to set CCTK_STRICT_MATH in the optionlist. This would indicate that strict ordering of floating point operations, and IEEE compliance, is required in this build. This could potentially then be used to select compiler optimisation settings automatically. We could then introduce additional Cactus tests, only enabled when CCTK_STRICT_MATH was set, which checked that certain operations give exactly the correct results.
I like this idea.
Assuming this is too big a change so close to the release, would it be a good idea to disable the test for now and then implement this after the release?
I would prefer that option. If tests needed to be reviewed, I would have been very unhappy about reviewing the test as ok, given that it is very likely to fail with most current ET optionlists. As such, I would prefer that we remove it for the release so that new users don't get worried about the test failure, and then implement something similar to the above scheme after the release.