maleadt

using Test

using CUDA

const MAX_DEPTH = 16
const SELECTION_SORT = 32

function selection_sort(data, left, right)
    @inbounds for i in left:right
        min_val = data[i]

maleadt

; ModuleID = 'permutedims!'
source_filename = "permutedims!"
target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128-ni:10:11:12:13"
target triple = "x86_64-pc-linux-gnu"

%jl_value_t = type opaque
%jl_array_t = type { i8 addrspace(13)*, i64, i16, i16, i32 }

declare %jl_value_t addrspace(10)* @japi1_checkdims_perm(%jl_value_t addrspace(10)*, %jl_value_t addrspace(10)**, i32)

maleadt

# experimentation with batched tridiagonal solvers on the GPU for Oceananigans.jl
#
# - reference serial CPU implementation
# - batched GPU implementation using cuSPARSE (fastest)
# - batched GPU implementation based on the serial CPU implementation (slow but flexible)
# - parallel GPU implementation (potentially fast and flexible)
#
# see `test_batched` and `bench_batched`

using LinearAlgebra

maleadt

# experimentation with batched tridiagonal solvers on the GPU for Oceananigans.jl
#
# - reference serial CPU implementation
# - batched GPU implementation using cuSPARSE (fastest)
# - batched GPU implementation based on the serial CPU implementation (slow but flexible)
# - parallel GPU implementation (potentially fast and flexible)
#
# see `test_batched` and `bench_batched`

using CUDAdrv

maleadt

#!/opt/fritzconnection/bin/python2

import sys
from fritzconnection import FritzConnection

if len(sys.argv)>1:
    if sys.argv[1]=="config":
        print '''
host_name TODO

maleadt

#include "cuda.h"
#include "cudadebugger.h"

#include <iostream>
#include <pthread.h>
#include <signal.h>

// helpers

CUDBGAPI cudbgAPI;

maleadt

if(DEFINED ENV{LOCAL_NAMESPACE})
    # NOTE: if localizing the namespace, by rewriting just before linking, we
    #       really do need to use static linking because we can't (easily)
    #       rewrite dynamic symbols

    # NOTE: the directive below is copied from the Linux rules in
    #       CMakeDefaultMakeRuleVariables.cmake, and may have to be adjusted for
    #       your platform
    set(CMAKE_CXX_CREATE_SHARED_LIBRARY "${CMAKE_SOURCE_DIR}/rewrite_wrapper <CMAKE_SHARED_LIBRARY_CXX_FLAGS> <LINK_FLAGS> <CMAKE_SHARED_LIBRARY_CREATE_CXX_FLAGS> <CMAKE_SHARED_LIBRARY_SONAME_CXX_FLAG><TARGET_SONAME> -o <TARGET> <OBJECTS> <LINK_LIBRARIES>")
endif()

maleadt

#
# AFL dictionary for Julia
# ------------------------
#
# Contains basic keywords, operators, and some syntax building blocks.
#
# Created by Tim Besard <tim.besard@gmail.com>
#

maleadt

# The Computer Language Benchmarks Game
# binary-trees benchmark
# http://shootout.alioth.debian.org/u32/performance.php?test=binarytrees
#
# Ported from an OCaml version

abstract type BTree end

mutable struct Empty <: BTree
end

maleadt

using LinearAlgebra, Statistics
using StaticArrays, Parameters
using CUDAdrv, CUDAnative

struct Lattice{D,Q,T}
    δt  :: T
    δx  :: T
    τ   :: T
    e   :: Vector{SVector{D,T}}
    w   :: Vector{T}