openmp/omplocal

Search:
Group by:

All Macro for LocalTensorField CPU-only loops

This module provides the all iterator for parallelized loops on LocalTensorField objects. LocalTensorField operations run on host memory and use CPU thread parallelization.

Usage: for n in all 0..<local.numSites(): localCn = localA.getSite(n) + localB.getSite(n)

The all macro is exported by parallel.nim for general use. For the corresponding GPU iterator, see each in ompdisp.nim.

Procs

proc hasEchoStatement(node: NimNode): bool {....raises: [], tags: [], forbids: [].}
Check if the AST contains echo statements

Macros

macro all(forLoop: ForLoopStmt): untyped

OpenMP parallel all loop for LocalTensorField (CPU threads)

Uses ompParallelFor template for actual parallelization.

Usage: for n in all 0..<local.numSites(): localCn = localA.getSite(n) + localB.getSite(n)