From 154a2abe274b59833a86b96eb45a075647ba1434 Mon Sep 17 00:00:00 2001 From: tradke Date: Fri, 2 Jun 2000 14:52:55 +0000 Subject: Usable version of thorn Hyperslab which extracts M-dimensional orthogonal subspaces from an N-dimensional data space. The Hyperslab API consists of 2 routines: - Hyperslab_GetLocalHyperslab() which returns a hyperslab from the processor's local data. This will probably be used for IO methods that want to output the hyperslab chunks in parallel (as is done currently by IOHDF5 and IOFlexIO). - Hyperslab_GetHyperslab() which collects hyperslab chunks from all processors using MPI and returns the global hyperslab data. This is used in IOASCII where the output is done by only one processor. Open questions: - How to specify the directions in which to span the hyperslab. - Putting the parameters passed to Hyperslab routines into structures rather than passing them individually. git-svn-id: http://svn.cactuscode.org/arrangements/CactusPUGH/PUGHSlab/trunk@22 10716dce-81a3-4424-a2c8-48026a0d3035 --- schedule.ccl | 14 -------------- 1 file changed, 14 deletions(-) (limited to 'schedule.ccl') diff --git a/schedule.ccl b/schedule.ccl index b97bb96..77309fe 100644 --- a/schedule.ccl +++ b/schedule.ccl @@ -1,16 +1,2 @@ # Schedule definitions for thorn Hyperslab # $Header$ - -STORAGE: testgf - -schedule InitTestGF at BASEGRID -{ - LANG:C -} "Init test GFs" - -schedule TestSlab2D at INITIAL -{ - LANG:C -} "Testing slabbing" - - -- cgit v1.2.3