#470: Check is wrong inside prolongate_3d_o1_rf2.cc ---------------------+------------------------------------------------------ Reporter: sbrandt | Owner: Type: defect | Status: new Priority: major | Milestone: Component: Other | Version: Keywords: | ---------------------+------------------------------------------------------ The problem materialized when creating wave equation tutorials for the Teragrid conference. I was unable to run an FMR application in parallel. The check comprises lines 76-80 of the file:
if (not regbbox.expand(offsetlo, offsethi).is_contained_in(srcbbox) or not regbbox .is_contained_in(dstbbox)) { CCTK_WARN (0, "Internal error: region extent is not contained in array extent"); }
#470: Check is wrong inside prolongate_3d_o1_rf2.cc ----------------------+----------------------------------------------------- Reporter: sbrandt | Owner: eschnett Type: defect | Status: new Priority: major | Milestone: Component: Carpet | Version: Resolution: | Keywords: ----------------------+----------------------------------------------------- Changes (by sbrandt):
* owner: => eschnett * component: Other => Carpet
#470: Check is wrong inside prolongate_3d_o1_rf2.cc ----------------------+----------------------------------------------------- Reporter: sbrandt | Owner: eschnett Type: defect | Status: new Priority: major | Milestone: Component: Carpet | Version: Resolution: | Keywords: ----------------------+-----------------------------------------------------
Comment (by eschnett):
It seems that offsethi is set to 1 even if the interpolation falls onto a grid point, where no offset is required. Other parts of Carpet handle this correctly and pass in an interpolation region that is large enough, but smaller than expected here.
#470: Check is wrong inside prolongate_3d_o1_rf2.cc ----------------------+----------------------------------------------------- Reporter: sbrandt | Owner: eschnett Type: defect | Status: new Priority: major | Milestone: Component: Carpet | Version: Resolution: | Keywords: ----------------------+-----------------------------------------------------
Comment (by rhaas):
This is fixed in the mercurial version of Carpet which has become the official verion in the ET Maxwell release. Should we close the ticket since it only affects an "old" version of Carpet?
#470: Check is wrong inside prolongate_3d_o1_rf2.cc ----------------------+----------------------------------------------------- Reporter: sbrandt | Owner: eschnett Type: defect | Status: closed Priority: major | Milestone: Component: Carpet | Version: Resolution: fixed | Keywords: ----------------------+----------------------------------------------------- Changes (by eschnett):
* status: new => closed * resolution: => fixed
Comment:
Yes, this is corrected in the current release.
trac@lists.einsteintoolkit.org
-
Einstein Toolkit