Package: release.debian.org Severity: important Hello. I received bug report that my package python-pyopencl-doc contains non-free file in examples (#722014). Licence reads: "Copyright NVIDIA. ALl rights reserved. NVIDIA Corporation and its licensors retain all intellectual property and proprietary rights in and to this software and related documentation. Any use, reproduction, disclosure, or distribution of this software and related documentation without an express license agreement from NVIDIA Corporation is strictly prohibited." I have checked and non-free file examples/matrix-multiply.py exists in all versions of PyOpenCL - in oldstable, stable, testing, and unstable. Upstream has been notified and removed this file from archive, and I'll soon upload fixed version to unstable. After asking on debian-mentors I was advised that I should also upload DFSG-free versions of packages to stable and oldstable: http://lists.debian.org/debian-mentors/2013/09/msg00104.html I attach debdiffs of proposed changes - please review those. I am DM, not DD, so I will need sponsor for those uploads. Best regards. -- System Information: Debian Release: jessie/sid APT prefers unstable APT policy: (500, 'unstable'), (500, 'testing'), (1, 'experimental') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 3.10-2-amd64 (SMP w/2 CPU cores) Locale: LANG=pl_PL.UTF-8, LC_CTYPE=pl_PL.UTF-8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash -- Tomasz Rybak GPG/PGP key ID: 2AD5 9860 Fingerprint A481 824E 7DD3 9C0E C40A 488E C654 FB33 2AD5 9860 http://member.acm.org/~tomaszrybak
diff -Nru pyopencl-0.92/debian/changelog pyopencl-0.92.dfsg/debian/changelog
--- pyopencl-0.92/debian/changelog 2010-11-11 23:10:57.000000000 +0100
+++ pyopencl-0.92.dfsg/debian/changelog 2013-09-17 17:12:20.000000000 +0200
@@ -1,3 +1,9 @@
+pyopencl (0.92.dfsg-1) oldstable; urgency=low
+
+ * Remove non-free file from examples (#722014).
+
+ -- Tomasz Rybak <tomasz.rybak@post.pl> Tue, 17 Sep 2013 17:03:46 +0200
+
pyopencl (0.92-1) unstable; urgency=high
* New upstream release
diff -Nru pyopencl-0.92/debian/patches/python-versions.patch pyopencl-0.92.dfsg/debian/patches/python-versions.patch
--- pyopencl-0.92/debian/patches/python-versions.patch 2010-11-11 13:30:27.000000000 +0100
+++ pyopencl-0.92.dfsg/debian/patches/python-versions.patch 2013-09-17 17:12:20.000000000 +0200
@@ -4,11 +4,11 @@
Python version and include it into name of linked library.
Forwarded: not-needed
Author: Tomasz Rybak <bogomips@post.pl>
-Last-Update: 2010-11-11
-Index: pyopencl-0.92/setup.py
+Last-Update: 2010-06-19
+Index: pyopencl-0.92~beta+git20100618/setup.py
===================================================================
---- pyopencl-0.92.orig/setup.py 2010-11-11 13:22:42.000000000 +0100
-+++ pyopencl-0.92/setup.py 2010-11-11 13:22:55.000000000 +0100
+--- pyopencl-0.92~beta+git20100618.orig/setup.py 2010-06-19 20:30:40.000000000 +0200
++++ pyopencl-0.92~beta+git20100618/setup.py 2010-06-19 20:31:07.000000000 +0200
@@ -42,6 +42,7 @@
def main():
@@ -17,7 +17,7 @@
from aksetup_helper import (hack_distutils, get_config, setup,
NumpyExtension, set_up_shipped_boost_if_requested)
-@@ -51,6 +52,7 @@
+@@ -53,6 +54,7 @@
EXTRA_OBJECTS, EXTRA_DEFINES = set_up_shipped_boost_if_requested(conf)
LIBRARY_DIRS = conf["BOOST_LIB_DIR"]
diff -Nru pyopencl-0.92/debian/patches/replace-setuptools.patch pyopencl-0.92.dfsg/debian/patches/replace-setuptools.patch
--- pyopencl-0.92/debian/patches/replace-setuptools.patch 2010-11-11 13:30:27.000000000 +0100
+++ pyopencl-0.92.dfsg/debian/patches/replace-setuptools.patch 2013-09-17 17:12:20.000000000 +0200
@@ -3,11 +3,11 @@
contains all necessary files and we avoid network traffic from build machine.
Forwarded: not-needed
Author: Tomasz Rybak <bogomips@post.pl>
-Last-Update: 2010-11-11
-Index: pyopencl-0.92/MANIFEST.in
+Last-Update: 2010-06-02
+Index: pyopencl-0.92~beta+git20100806/MANIFEST.in
===================================================================
---- pyopencl-0.92.orig/MANIFEST.in 2010-11-11 13:24:53.000000000 +0100
-+++ pyopencl-0.92/MANIFEST.in 2010-11-11 13:25:37.000000000 +0100
+--- pyopencl-0.92~beta+git20100806.orig/MANIFEST.in 2010-08-07 13:49:58.000000000 +0200
++++ pyopencl-0.92~beta+git20100806/MANIFEST.in 2010-08-07 15:13:27.000000000 +0200
@@ -7,7 +7,6 @@
include doc/*.py
include doc/source/conf.py
@@ -16,10 +16,10 @@
include configure.py
include Makefile.in
include aksetup_helper.py
-Index: pyopencl-0.92/aksetup_helper.py
+Index: pyopencl-0.92~beta+git20100806/aksetup_helper.py
===================================================================
---- pyopencl-0.92.orig/aksetup_helper.py 2010-11-11 13:25:02.000000000 +0100
-+++ pyopencl-0.92/aksetup_helper.py 2010-11-11 13:25:37.000000000 +0100
+--- pyopencl-0.92~beta+git20100806.orig/aksetup_helper.py 2010-08-07 13:49:58.000000000 +0200
++++ pyopencl-0.92~beta+git20100806/aksetup_helper.py 2010-08-07 15:13:52.000000000 +0200
@@ -1,35 +1,6 @@
-# dealings with ez_setup ------------------------------------------------------
-import distribute_setup
@@ -90,8 +90,7 @@
return schema.read_config()
-@@ -516,34 +460,7 @@
- import sys
+@@ -517,34 +461,7 @@
if conf["USE_SHIPPED_BOOST"]:
- if not exists("bpl-subset/bpl_subset/boost/version.hpp"):
@@ -112,7 +111,8 @@
- print("'USE_SHIPPED_BOOST' off.")
- print("------------------------------------------------------------------------")
- conf["USE_SHIPPED_BOOST"] = False
--
++ conf["USE_SHIPPED_BOOST"] = False
+
- delay = 10
-
- from time import sleep
@@ -122,7 +122,7 @@
- sys.stdout.flush()
- delay -= 1
- sleep(1)
-+ conf["USE_SHIPPED_BOOST"] = False
-
+-
if conf["USE_SHIPPED_BOOST"]:
conf["BOOST_INC_DIR"] = ["bpl-subset/bpl_subset"]
+ conf["BOOST_LIB_DIR"] = []
diff -Nru pyopencl-0.92/debian/rules pyopencl-0.92.dfsg/debian/rules
--- pyopencl-0.92/debian/rules 2010-11-11 13:30:27.000000000 +0100
+++ pyopencl-0.92.dfsg/debian/rules 2013-09-17 17:12:20.000000000 +0200
@@ -36,6 +36,7 @@
git clone $(GIT_URL) $(MODULE_NAME)-$(DEB_UPSTREAM_VERSION)
cd $(MODULE_NAME)-$(DEB_UPSTREAM_VERSION) && git checkout $(GIT_REVISION)
rm -rf $(MODULE_NAME)-$(DEB_UPSTREAM_VERSION)/.git $(MODULE_NAME)-$(DEB_UPSTREAM_VERSION)/.gitignore $(MODULE_NAME)-$(DEB_UPSTREAM_VERSION)/.gitmodules
+ rm -rf $(MODULE_NAME)-$(DEB_UPSTREAM_VERSION)/examples/matrix-multiply.py
tar czf $(MODULE_NAME)_$(DEB_UPSTREAM_VERSION).orig.tar.gz $(MODULE_NAME)-$(DEB_UPSTREAM_VERSION)
rm -rf $(MODULE_NAME)-$(DEB_UPSTREAM_VERSION)
diff -Nru pyopencl-0.92/examples/matrix-multiply.py pyopencl-0.92.dfsg/examples/matrix-multiply.py
--- pyopencl-0.92/examples/matrix-multiply.py 2010-10-21 19:10:19.000000000 +0200
+++ pyopencl-0.92.dfsg/examples/matrix-multiply.py 1970-01-01 01:00:00.000000000 +0100
@@ -1,241 +0,0 @@
-# example provided by Eilif Muller
-
-from __future__ import division
-
-KERNEL_CODE = """
-
-// Thread block size
-#define BLOCK_SIZE %(block_size)d
-
-// Matrix dimensions
-// (chosen as multiples of the thread block size for simplicity)
-#define WA %(w_a)d // Matrix A width
-#define HA %(h_a)d // Matrix A height
-#define WB %(w_b)d // Matrix B width
-#define HB WA // Matrix B height
-#define WC WB // Matrix C width
-#define HC HA // Matrix C height
-
-
-/*
- * Copyright 1993-2009 NVIDIA Corporation. All rights reserved.
- *
- * NVIDIA Corporation and its licensors retain all intellectual property and
- * proprietary rights in and to this software and related documentation.
- * Any use, reproduction, disclosure, or distribution of this software
- * and related documentation without an express license agreement from
- * NVIDIA Corporation is strictly prohibited.
- *
- * Please refer to the applicable NVIDIA end user license agreement (EULA)
- * associated with this source code for terms and conditions that govern
- * your use of this NVIDIA software.
- *
- */
-
-/* Matrix multiplication: C = A * B.
- * Device code.
- */
-
-#define AS(j, i) As[i + j * BLOCK_SIZE]
-#define BS(j, i) Bs[i + j * BLOCK_SIZE]
-
-////////////////////////////////////////////////////////////////////////////////
-//! Matrix multiplication on the device: C = A * B
-//! WA is A's width and WB is B's width
-////////////////////////////////////////////////////////////////////////////////
-__kernel __attribute__((reqd_work_group_size(BLOCK_SIZE,BLOCK_SIZE,1)))
-void
-matrixMul( __global float* C, __global float* A, __global float* B)
-{
- __local float As[BLOCK_SIZE*BLOCK_SIZE];
- __local float Bs[BLOCK_SIZE*BLOCK_SIZE];
-
- // Block index
- int bx = get_group_id(0);
- int by = get_group_id(1);
-
- // Thread index
- int tx = get_local_id(0);
- int ty = get_local_id(1);
-
- // Index of the first sub-matrix of A processed by the block
- int aBegin = WA * BLOCK_SIZE * by;
-
- // Index of the last sub-matrix of A processed by the block
- int aEnd = aBegin + WA - 1;
-
- // Step size used to iterate through the sub-matrices of A
- int aStep = BLOCK_SIZE;
-
- // Index of the first sub-matrix of B processed by the block
- int bBegin = BLOCK_SIZE * bx;
-
- // Step size used to iterate through the sub-matrices of B
- int bStep = BLOCK_SIZE * WB;
-
- // Csub is used to store the element of the block sub-matrix
- // that is computed by the thread
- float Csub = 0.0f;
-
- // Loop over all the sub-matrices of A and B
- // required to compute the block sub-matrix
- for (int a = aBegin, b = bBegin;
- a <= aEnd;
- a += aStep, b += bStep) {
-
- // Load the matrices from device memory
- // to shared memory; each thread loads
- // one element of each matrix
- AS(ty, tx) = A[a + WA * ty + tx];
- BS(ty, tx) = B[b + WB * ty + tx];
-
- // Synchronize to make sure the matrices are loaded
- barrier(CLK_LOCAL_MEM_FENCE);
-
- // Multiply the two matrices together;
- // each thread computes one element
- // of the block sub-matrix
- for (int k = 0; k < BLOCK_SIZE; ++k)
- Csub += AS(ty, k) * BS(k, tx);
-
- // Synchronize to make sure that the preceding
- // computation is done before loading two new
- // sub-matrices of A and B in the next iteration
- barrier(CLK_LOCAL_MEM_FENCE);
- }
-
- // Write the block sub-matrix to device memory;
- // each thread writes one element
- C[get_global_id(1) * get_global_size(0) + get_global_id(0)] = Csub;
-
-}
-
-"""
-
-import pyopencl as cl
-from time import time
-import numpy
-
-block_size = 16
-
-ctx = cl.create_some_context()
-
-for dev in ctx.devices:
- assert dev.local_mem_size > 0
-
-queue = cl.CommandQueue(ctx,
- properties=cl.command_queue_properties.PROFILING_ENABLE)
-
-#queue = cl.CommandQueue(ctx)
-
-if False:
- a_height = 4096
- #a_height = 1024
- a_width = 2048
- #a_width = 256
- #b_height == a_width
- b_width = a_height
-
-elif False:
- # like PyCUDA
- a_height = 2516
- a_width = 1472
- b_height = a_width
- b_width = 2144
-
-else:
- # CL SDK
- a_width = 50*block_size
- a_height = 100*block_size
- b_width = 50*block_size
- b_height = a_width
-
-c_width = b_width
-c_height = a_height
-
-h_a = numpy.random.rand(a_height, a_width).astype(numpy.float32)
-h_b = numpy.random.rand(b_height, b_width).astype(numpy.float32)
-h_c = numpy.empty((c_height, c_width)).astype(numpy.float32)
-
-
-kernel_params = {"block_size": block_size,
- "w_a":a_width, "h_a":a_height, "w_b":b_width}
-
-if "NVIDIA" in queue.device.vendor:
- options = "-cl-mad-enable -cl-fast-relaxed-math"
-else:
- options = None
-prg = cl.Program(ctx, KERNEL_CODE % kernel_params,
- ).build(options=options)
-kernel = prg.matrixMul
-#print prg.binaries[0]
-
-assert a_width % block_size == 0
-assert a_height % block_size == 0
-assert b_width % block_size == 0
-
-# transfer host -> device -----------------------------------------------------
-mf = cl.mem_flags
-
-t1 = time()
-
-d_a_buf = cl.Buffer(ctx, mf.READ_ONLY | mf.COPY_HOST_PTR, hostbuf=h_a)
-d_b_buf = cl.Buffer(ctx, mf.READ_ONLY | mf.COPY_HOST_PTR, hostbuf=h_b)
-d_c_buf = cl.Buffer(ctx, mf.WRITE_ONLY, size=h_c.nbytes)
-
-push_time = time()-t1
-
-# warmup ----------------------------------------------------------------------
-for i in range(5):
- event = kernel(queue, h_c.shape, (block_size, block_size),
- d_c_buf, d_a_buf, d_b_buf)
- event.wait()
-
-queue.finish()
-
-# actual benchmark ------------------------------------------------------------
-t1 = time()
-
-count = 20
-for i in range(count):
- event = kernel(queue, h_c.shape, (block_size, block_size),
- d_c_buf, d_a_buf, d_b_buf)
-
-event.wait()
-
-gpu_time = (time()-t1)/count
-
-# transfer device -> host -----------------------------------------------------
-t1 = time()
-cl.enqueue_read_buffer(queue, d_c_buf, h_c).wait()
-pull_time = time()-t1
-
-# timing output ---------------------------------------------------------------
-gpu_total_time = gpu_time+push_time+pull_time
-
-print "GPU push+compute+pull total [s]:", gpu_total_time
-print "GPU push [s]:", push_time
-print "GPU pull [s]:", pull_time
-print "GPU compute (host-timed) [s]:", gpu_time
-print "GPU compute (event-timed) [s]: ", (event.profile.end-event.profile.start)*1e-9
-
-gflop = h_c.size * (a_width * 2.) / (1000**3.)
-gflops = gflop / gpu_time
-
-print
-print "GFlops/s:", gflops
-
-# cpu comparison --------------------------------------------------------------
-t1 = time()
-h_c_cpu = numpy.dot(h_a,h_b)
-cpu_time = time()-t1
-
-print
-print "GPU==CPU:",numpy.allclose(h_c, h_c_cpu)
-print
-print "CPU time (s)", cpu_time
-print
-
-print "GPU speedup (with transfer): ", cpu_time/gpu_total_time
-print "GPU speedup (without transfer): ", cpu_time/gpu_time
-
diff -Nru pyopencl-2012.1/debian/changelog pyopencl-2012.1.dfsg/debian/changelog
--- pyopencl-2012.1/debian/changelog 2012-06-21 22:22:52.000000000 +0200
+++ pyopencl-2012.1.dfsg/debian/changelog 2013-09-17 17:29:52.000000000 +0200
@@ -1,3 +1,9 @@
+pyopencl (2012.1.dfsg-1) stable; urgency=low
+
+ * Remove non-free file from examples (#722014).
+
+ -- Tomasz Rybak <tomasz.rybak@post.pl> Tue, 17 Sep 2013 17:25:16 +0200
+
pyopencl (2012.1-1) unstable; urgency=low
* New upstream release.
diff -Nru pyopencl-2012.1/debian/rules pyopencl-2012.1.dfsg/debian/rules
--- pyopencl-2012.1/debian/rules 2012-06-21 22:22:52.000000000 +0200
+++ pyopencl-2012.1.dfsg/debian/rules 2013-09-17 17:29:52.000000000 +0200
@@ -48,6 +48,7 @@
rm -rf $(MODULE_NAME)-$(DEB_UPSTREAM_VERSION)/.git $(MODULE_NAME)-$(DEB_UPSTREAM_VERSION)/.gitignore $(MODULE_NAME)-$(DEB_UPSTREAM_VERSION)/.gitmodules
rm -rf $(MODULE_NAME)-$(DEB_UPSTREAM_VERSION)/doc/source/.gitignore
rm -rf $(MODULE_NAME)-$(DEB_UPSTREAM_VERSION)/$(GIT_SUBMODULES)/.git $(MODULE_NAME)-$(DEB_UPSTREAM_VERSION)/$(GIT_SUBMODULES)/.gitignore
+ rm -rf $(MODULE_NAME)-$(DEB_UPSTREAM_VERSION)/examples/matrix-multiply.py
tar czf $(MODULE_NAME)_$(DEB_UPSTREAM_VERSION).orig.tar.gz $(MODULE_NAME)-$(DEB_UPSTREAM_VERSION)
rm -rf $(MODULE_NAME)-$(DEB_UPSTREAM_VERSION)
diff -Nru pyopencl-2012.1/examples/matrix-multiply.py pyopencl-2012.1.dfsg/examples/matrix-multiply.py
--- pyopencl-2012.1/examples/matrix-multiply.py 2012-06-21 22:22:19.000000000 +0200
+++ pyopencl-2012.1.dfsg/examples/matrix-multiply.py 1970-01-01 01:00:00.000000000 +0100
@@ -1,241 +0,0 @@
-# example provided by Eilif Muller
-
-from __future__ import division
-
-KERNEL_CODE = """
-
-// Thread block size
-#define BLOCK_SIZE %(block_size)d
-
-// Matrix dimensions
-// (chosen as multiples of the thread block size for simplicity)
-#define WA %(w_a)d // Matrix A width
-#define HA %(h_a)d // Matrix A height
-#define WB %(w_b)d // Matrix B width
-#define HB WA // Matrix B height
-#define WC WB // Matrix C width
-#define HC HA // Matrix C height
-
-
-/*
- * Copyright 1993-2009 NVIDIA Corporation. All rights reserved.
- *
- * NVIDIA Corporation and its licensors retain all intellectual property and
- * proprietary rights in and to this software and related documentation.
- * Any use, reproduction, disclosure, or distribution of this software
- * and related documentation without an express license agreement from
- * NVIDIA Corporation is strictly prohibited.
- *
- * Please refer to the applicable NVIDIA end user license agreement (EULA)
- * associated with this source code for terms and conditions that govern
- * your use of this NVIDIA software.
- *
- */
-
-/* Matrix multiplication: C = A * B.
- * Device code.
- */
-
-#define AS(j, i) As[i + j * BLOCK_SIZE]
-#define BS(j, i) Bs[i + j * BLOCK_SIZE]
-
-////////////////////////////////////////////////////////////////////////////////
-//! Matrix multiplication on the device: C = A * B
-//! WA is A's width and WB is B's width
-////////////////////////////////////////////////////////////////////////////////
-__kernel __attribute__((reqd_work_group_size(BLOCK_SIZE,BLOCK_SIZE,1)))
-void
-matrixMul( __global float* C, __global float* A, __global float* B)
-{
- __local float As[BLOCK_SIZE*BLOCK_SIZE];
- __local float Bs[BLOCK_SIZE*BLOCK_SIZE];
-
- // Block index
- int bx = get_group_id(0);
- int by = get_group_id(1);
-
- // Thread index
- int tx = get_local_id(0);
- int ty = get_local_id(1);
-
- // Index of the first sub-matrix of A processed by the block
- int aBegin = WA * BLOCK_SIZE * by;
-
- // Index of the last sub-matrix of A processed by the block
- int aEnd = aBegin + WA - 1;
-
- // Step size used to iterate through the sub-matrices of A
- int aStep = BLOCK_SIZE;
-
- // Index of the first sub-matrix of B processed by the block
- int bBegin = BLOCK_SIZE * bx;
-
- // Step size used to iterate through the sub-matrices of B
- int bStep = BLOCK_SIZE * WB;
-
- // Csub is used to store the element of the block sub-matrix
- // that is computed by the thread
- float Csub = 0.0f;
-
- // Loop over all the sub-matrices of A and B
- // required to compute the block sub-matrix
- for (int a = aBegin, b = bBegin;
- a <= aEnd;
- a += aStep, b += bStep) {
-
- // Load the matrices from device memory
- // to shared memory; each thread loads
- // one element of each matrix
- AS(ty, tx) = A[a + WA * ty + tx];
- BS(ty, tx) = B[b + WB * ty + tx];
-
- // Synchronize to make sure the matrices are loaded
- barrier(CLK_LOCAL_MEM_FENCE);
-
- // Multiply the two matrices together;
- // each thread computes one element
- // of the block sub-matrix
- for (int k = 0; k < BLOCK_SIZE; ++k)
- Csub += AS(ty, k) * BS(k, tx);
-
- // Synchronize to make sure that the preceding
- // computation is done before loading two new
- // sub-matrices of A and B in the next iteration
- barrier(CLK_LOCAL_MEM_FENCE);
- }
-
- // Write the block sub-matrix to device memory;
- // each thread writes one element
- C[get_global_id(1) * get_global_size(0) + get_global_id(0)] = Csub;
-
-}
-
-"""
-
-import pyopencl as cl
-from time import time
-import numpy
-
-block_size = 16
-
-ctx = cl.create_some_context()
-
-for dev in ctx.devices:
- assert dev.local_mem_size > 0
-
-queue = cl.CommandQueue(ctx,
- properties=cl.command_queue_properties.PROFILING_ENABLE)
-
-#queue = cl.CommandQueue(ctx)
-
-if False:
- a_height = 4096
- #a_height = 1024
- a_width = 2048
- #a_width = 256
- #b_height == a_width
- b_width = a_height
-
-elif False:
- # like PyCUDA
- a_height = 2516
- a_width = 1472
- b_height = a_width
- b_width = 2144
-
-else:
- # CL SDK
- a_width = 50*block_size
- a_height = 100*block_size
- b_width = 50*block_size
- b_height = a_width
-
-c_width = b_width
-c_height = a_height
-
-h_a = numpy.random.rand(a_height, a_width).astype(numpy.float32)
-h_b = numpy.random.rand(b_height, b_width).astype(numpy.float32)
-h_c = numpy.empty((c_height, c_width)).astype(numpy.float32)
-
-
-kernel_params = {"block_size": block_size,
- "w_a":a_width, "h_a":a_height, "w_b":b_width}
-
-if "NVIDIA" in queue.device.vendor:
- options = "-cl-mad-enable -cl-fast-relaxed-math"
-else:
- options = ""
-prg = cl.Program(ctx, KERNEL_CODE % kernel_params,
- ).build(options=options)
-kernel = prg.matrixMul
-#print prg.binaries[0]
-
-assert a_width % block_size == 0
-assert a_height % block_size == 0
-assert b_width % block_size == 0
-
-# transfer host -> device -----------------------------------------------------
-mf = cl.mem_flags
-
-t1 = time()
-
-d_a_buf = cl.Buffer(ctx, mf.READ_ONLY | mf.COPY_HOST_PTR, hostbuf=h_a)
-d_b_buf = cl.Buffer(ctx, mf.READ_ONLY | mf.COPY_HOST_PTR, hostbuf=h_b)
-d_c_buf = cl.Buffer(ctx, mf.WRITE_ONLY, size=h_c.nbytes)
-
-push_time = time()-t1
-
-# warmup ----------------------------------------------------------------------
-for i in range(5):
- event = kernel(queue, h_c.shape[::-1], (block_size, block_size),
- d_c_buf, d_a_buf, d_b_buf)
- event.wait()
-
-queue.finish()
-
-# actual benchmark ------------------------------------------------------------
-t1 = time()
-
-count = 20
-for i in range(count):
- event = kernel(queue, h_c.shape[::-1], (block_size, block_size),
- d_c_buf, d_a_buf, d_b_buf)
-
-event.wait()
-
-gpu_time = (time()-t1)/count
-
-# transfer device -> host -----------------------------------------------------
-t1 = time()
-cl.enqueue_copy(queue, h_c, d_c_buf)
-pull_time = time()-t1
-
-# timing output ---------------------------------------------------------------
-gpu_total_time = gpu_time+push_time+pull_time
-
-print "GPU push+compute+pull total [s]:", gpu_total_time
-print "GPU push [s]:", push_time
-print "GPU pull [s]:", pull_time
-print "GPU compute (host-timed) [s]:", gpu_time
-print "GPU compute (event-timed) [s]: ", (event.profile.end-event.profile.start)*1e-9
-
-gflop = h_c.size * (a_width * 2.) / (1000**3.)
-gflops = gflop / gpu_time
-
-print
-print "GFlops/s:", gflops
-
-# cpu comparison --------------------------------------------------------------
-t1 = time()
-h_c_cpu = numpy.dot(h_a,h_b)
-cpu_time = time()-t1
-
-print
-print "GPU==CPU:",numpy.allclose(h_c, h_c_cpu)
-print
-print "CPU time (s)", cpu_time
-print
-
-print "GPU speedup (with transfer): ", cpu_time/gpu_total_time
-print "GPU speedup (without transfer): ", cpu_time/gpu_time
-
Attachment:
signature.asc
Description: This is a digitally signed message part