Pycuda 中的分段错误使用 NVIDIA 的 cuSolver 库

Question

我正在尝试制作一个受scikits-cuda 库启发的 pycuda 包装器，对于 Nvidia 的新 cuSolver 库中提供的一些操作，首先我需要通过 cusolverDnSgetrf() 操作执行 LU 分解。但在此之前，我需要“工作空间”参数，cuSolver 提供的用于获取该参数的工具名为 cusolverDnSgetrf_bufferSize();但是当我使用它时，只是崩溃并返回一个分段错误。我做错了什么？

注意：我已经使用 scikits-cuda 处理了这个操作，但是 cuSolver 库使用了很多此类参数，我想比较 scikits-cuda 和我的实现与新库之间的用法。

---

import numpy as np
import pycuda.gpuarray
import ctypes
import ctypes.util

libcusolver = ctypes.cdll.LoadLibrary('libcusolver.so')

class _types:
  handle = ctypes.c_void_p

libcusolver.cusolverDnCreate.restype = int
libcusolver.cusolverDnCreate.argtypes = [_types.handle]

def cusolverCreate():
    handle = _types.handle()
    libcusolver.cusolverDnCreate(ctypes.byref(handle))
    return handle.value

libcusolver.cusolverDnDestroy.restype = int
libcusolver.cusolverDnDestroy.argtypes = [_types.handle]

def cusolverDestroy(handle):
    libcusolver.cusolverDnDestroy(handle)


libcusolver.cusolverDnSgetrf_bufferSize.restype = int
libcusolver.cusolverDnSgetrf_bufferSize.argtypes =[_types.handle,
                                       ctypes.c_int,
                                       ctypes.c_int,
                                       ctypes.c_void_p,
                                       ctypes.c_int,
                                       ctypes.c_void_p]

def cusolverLUFactorization(handle, matrix):
    m,n=matrix.shape
    mtx_gpu = gpuarray.to_gpu(matrix.astype('float32'))
    work=gpuarray.zeros(1, np.float32)
    status=libcusolver.cusolverDnSgetrf_bufferSize(
                          handle, m, n,
                          int(mtx_gpu.gpudata),
                          n, int(work.gpudata))
    print status


x = np.asarray(np.random.rand(3, 3), np.float32)
handle_solver=cusolverCreate()
cusolverLUFactorization(handle_solver,x)
cusolverDestroy(handle_solver)

Answer 1

cusolverDnSgetrf_bufferSize的最后一个参数应该是常规指针，而不是GPU内存指针。尝试修改cusolverLUFactorization()函数如下：

def cusolverLUFactorization(handle, matrix):
    m,n=matrix.shape
    mtx_gpu = gpuarray.to_gpu(matrix.astype('float32'))

    work = ctypes.c_int()
    status = libcusolver.cusolverDnSgetrf_bufferSize(
                         handle, m, n,
                         int(mtx_gpu.gpudata),
                         n, ctypes.pointer(work))
    print status
    print work.value