- Description
- How to build
- Using MemtestCL as a library
- CLI Standalone Basic Usage
- CLI Standalone Advanced Usage
- FAQ
- Licensing
MemtestCL is a program to test the memory and logic of OpenCL-enabled GPUs, CPUs, and accelerators for errors. It is an OpenCL port of our CUDA- based tester for NVIDIA GPUs, MemtestG80.
This is the open-source version of MemtestCL, implementing the same memory tests as the closed-source version. The intended usage is as a library so that other software developers can use the MemtestCL tests to validate the correct operation of GPUs or accelerators in their own code. In addition to the core memory testing libraries, this package contains the source code to a limited version of the command-line interface standalone tester included in the closed- source build; certain capabilities, such as the ability to transmit results back to Stanford, are not present in the open-source version.
Up-to-date versions of both the closed- and open-source versions of MemtestCL may be downloaded at https://simtk.org/home/memtest. The closed-source version is available as precompiled binaries; the open-source version is available only as a source package.
This document concerns the open-source version.
First, ensure that you have installed an OpenCL SDK (typically, either that in the NVIDIA CUDA toolkits (from 3.0 onwards) or the ATI Stream SDK. Binaries built using either SDK should execute on any OpenCL implementation (ie, it is not required that the binary be built using the ATI Stream SDK to run on ATI GPUs). Common paths for SDK installations are included in the Makefiles; it may be necessary to adjust such paths to correspond to your installation.
Makefiles for 32- and 64-bit Linux, Mac OS X, and 32-bit Windows are included On Linux and OSX, it should be possible to build MemtestCL by executing the following from the root of the source package:
make -f Makefiles/Makefile.OS
where OS is one of linux32, linux64, osx. On Windows, the Visual Studio C++ compiler and make system is required (tested under VS2005); execute the following command to build:
nmake -f Makefiles\Makefile.windows
The resulting executable, memtestCL, should be immediately executable on Linux and OS X platforms. On Windows, libiconv-2.dll, libintl-2.dll, and popt1.dll must be copied from the popt/win32 subdirectory into a directory in the DLL search path (most conveniently, the root of the source distribution). MemtestCL uses the MIT/X licensed popt library to handle command line arguments; precompiled static libraries are provided for Linux and OS X, but dynamic libraries for Windows.
We encourage software developers to use MemtestCL as a code library in their programs to verify the correct operation of hardware on which they execute. The code is licensed under the LGPL, so developers of both open- and closed-source software can use it - developers of closed-source software are required to link to MemtestCL via a shared library (.so, .dll) mechanism; open-source software can integrate it via static linkage.
The API for the memory tests is defined in memtestCL_core.h. There are two APIs - a low-level API defined by the memtestFunctions class (which is a thin wrapper around the underlyinh OpenCL kernel invocations), and a high-level API defined by the memtestState and memtestMultiTester classes. At the lowest level the tests are implemented by the kernels in memtestCL_kernels.cl.
The recommended interface is the memtestMultiTester class, which automatically encapsulates details such as the maximum per-buffer allocation in a particular OCL library. An example of the API's usage can be found in the standalone tester, memtestCL_cli.cu.
MemtestCL is available for Windows, Linux, and Mac OS X-based machines. In the following directions, please replace "MemtestCL" with the name of the program included in the distribution for your operating system.
MemtestCL is a command line application; to run it, start it from a command prompt (Start->Run->cmd in Windows, Terminal.app in OS X). For basic operation, just run it from the command prompt:
memtestcl
By default, MemtestCL will test 128 megabytes of memory on the first OpenCL device on the first OpenCL platform found,running 50 iterations of its tests. On typical machines, each iteration will complete in under 10 seconds with these parameters (the speed will vary both with the speed of the card tested and the amount of memory tested). The amount of memory tested and number of test iterations can be modified by adding command line parameters as follows:
memtestcl [amount of RAM in megabytes] [number of test iterations]
For example, to run MemtestCL over 256 megabytes of RAM, with 100 test iterations, execute the following command:
memtestcl 256 100
Be aware that not all of the memory on your video card can be tested, as part of it is reserved for use by the operating system, and (as of this writing) both ATI and NVIDIA OpenCL drivers severely restrict the amount of memory available to an OpenCL program running on a GPU. If too large a test region is specified, the program will print a warning and quit. Also, if the tested GPU is currently driving a graphical desktop, the driver may impose time limits on test execution such that tests over very large test regions will time out. Timeouts or other execution errors will be trapped and will cause the test to terminate early. Due to the currently immature state of OpenCL implementations, they may also cause the program to crash.
If you suspect that your graphics card is having issues (for example, it fails running Folding@home work units), we strongly recommend that you test as large a memory region as is practical, and run thousands of test iterations. In our testing, we have found that even "problematic" cards may only fail sporadically (e.g., once every 50,000 test iterations). Like other stress testing tools, to properly verify stability MemtestCL should be run for an extended period of time.
MemtestCL supports the use of various command line flags to enable advanced functionality. Flags may be issued in any order, and may precede or follow the memory size and iteration count parameters (but the memory size must always precede the iteration count).
To run MemtestCL on an OpenCL platform other than the first (e.g., if you have both the AMD and NVIDIA OpenCL implementations installed), use the --platform or -p flags, passing the index of the platform to test (starting at zero). If you do not know the index of the OpenCL platform you want, just run MemtestCL with no parameters - a list of all platforms found will be printed immediately following the usage summary. For example, to run MemtestCL on the second platform in a system:
memtestcl --platform 1
To run MemtestCL on an OpenCL device (e.g., GPU) other than the first one on the selected platform, use the --gpu or -g flags, passing the index of the device to test (starting at zero). MemtestCL prints a list of all devices on the selected platform (and their indices) before running tests. For example, to run MemtestCL on the third OpenCL device on the default platform (platform index 0):
memtestcl --gpu 2
The --platform and --gpu flags may be combined, to select a different platform and device. This may be necessary in multi-vendor or multi-GPU configurations. Refer to the list of platforms and devices on the current platform printed at program start to determine the right combination. To select the third GPU on the second platform:
memtestcl --platform 1 --gpu 2
Finally, to display the license agreement for MemtestCL, provide the --license or -l options:
memtestcl -l
To check lots of memory on AMD cards extra environment variables can be useful, depending on the driver version. For example with the below settings testing 7672MB was succesfully done on a 8GB AMD gpu w/o display on 64-bit windows.
set GPU_MAX_HEAP_SIZE=100
set GPU_SINGLE_ALLOC_PERCENT=100
set GPU_ENABLE_LARGE_ALLOCATION=1
-
I have an {ATI 2xxx/3xxx ,NVIDIA 5/6/7-series} video card and it doesn't work!
- Only OpenCL-capable hardware is supported. As of this writing, this includes any NVIDIA GeForce 8-series or newer graphics cards (e.g., GeForce 8-, 9-, GT- GTS-, GTX-series; Quadro FX, and Tesla boards), and the ATI Radeon 4xxx and 5xxx-series graphics cards. MemtestCL has also been tested on Intel and AMD CPUs using the AMD OpenCL implementation. OpenCL implementations exist for the Cell processor, but this has not been tested.
-
I have an OpenCL-enabled card, but it still doesn't work!
- You must have an OpenCL-enabled graphics driver and OpenCL runtime installed. For NVIDIA hardware, this requires a version 195 ForceWare driver or newer; ATI requires BOTH a v9.12 or newer Catalyst driver, as well as (at the time of writing) the ATI Stream SDK.
-
I get an error complaining about a missing "OpenCL.dll" on Windows!
- You must install OpenCL. NVIDIA bundles OpenCL.dll with recent (v195 or newer) drivers. For ATI GPUs, you must install an OpenCL- capable video driver, and also install the ATI Stream SDK (http://developer.amd.com/gpu/atistreamsdk/). For CPU support, installing only the ATI Stream SDK is sufficient.
The source code to the open-source edition of MemtestCL is Copyright 2010, Stanford University, and is licensed under the terms of the GNU Lesser General Public License, version 3, reproduced below:
GNU LESSER GENERAL PUBLIC LICENSE
Version 3, 29 June 2007
Copyright (C) 2007 Free Software Foundation, Inc. <http://fsf.org/>
Everyone is permitted to copy and distribute verbatim copies
of this license document, but changing it is not allowed.
This version of the GNU Lesser General Public License incorporates
the terms and conditions of version 3 of the GNU General Public
License, supplemented by the additional permissions listed below.
0. Additional Definitions.
As used herein, "this License" refers to version 3 of the GNU Lesser
General Public License, and the "GNU GPL" refers to version 3 of the GNU
General Public License.
"The Library" refers to a covered work governed by this License,
other than an Application or a Combined Work as defined below.
An "Application" is any work that makes use of an interface provided
by the Library, but which is not otherwise based on the Library.
Defining a subclass of a class defined by the Library is deemed a mode
of using an interface provided by the Library.
A "Combined Work" is a work produced by combining or linking an
Application with the Library. The particular version of the Library
with which the Combined Work was made is also called the "Linked
Version".
The "Minimal Corresponding Source" for a Combined Work means the
Corresponding Source for the Combined Work, excluding any source code
for portions of the Combined Work that, considered in isolation, are
based on the Application, and not on the Linked Version.
The "Corresponding Application Code" for a Combined Work means the
object code and/or source code for the Application, including any data
and utility programs needed for reproducing the Combined Work from the
Application, but excluding the System Libraries of the Combined Work.
1. Exception to Section 3 of the GNU GPL.
You may convey a covered work under sections 3 and 4 of this License
without being bound by section 3 of the GNU GPL.
2. Conveying Modified Versions.
If you modify a copy of the Library, and, in your modifications, a
facility refers to a function or data to be supplied by an Application
that uses the facility (other than as an argument passed when the
facility is invoked), then you may convey a copy of the modified
version:
a) under this License, provided that you make a good faith effort to
ensure that, in the event an Application does not supply the
function or data, the facility still operates, and performs
whatever part of its purpose remains meaningful, or
b) under the GNU GPL, with none of the additional permissions of
this License applicable to that copy.
3. Object Code Incorporating Material from Library Header Files.
The object code form of an Application may incorporate material from
a header file that is part of the Library. You may convey such object
code under terms of your choice, provided that, if the incorporated
material is not limited to numerical parameters, data structure
layouts and accessors, or small macros, inline functions and templates
(ten or fewer lines in length), you do both of the following:
a) Give prominent notice with each copy of the object code that the
Library is used in it and that the Library and its use are
covered by this License.
b) Accompany the object code with a copy of the GNU GPL and this license
document.
4. Combined Works.
You may convey a Combined Work under terms of your choice that,
taken together, effectively do not restrict modification of the
portions of the Library contained in the Combined Work and reverse
engineering for debugging such modifications, if you also do each of
the following:
a) Give prominent notice with each copy of the Combined Work that
the Library is used in it and that the Library and its use are
covered by this License.
b) Accompany the Combined Work with a copy of the GNU GPL and this license
document.
c) For a Combined Work that displays copyright notices during
execution, include the copyright notice for the Library among
these notices, as well as a reference directing the user to the
copies of the GNU GPL and this license document.
d) Do one of the following:
0) Convey the Minimal Corresponding Source under the terms of this
License, and the Corresponding Application Code in a form
suitable for, and under terms that permit, the user to
recombine or relink the Application with a modified version of
the Linked Version to produce a modified Combined Work, in the
manner specified by section 6 of the GNU GPL for conveying
Corresponding Source.
1) Use a suitable shared library mechanism for linking with the
Library. A suitable mechanism is one that (a) uses at run time
a copy of the Library already present on the user's computer
system, and (b) will operate properly with a modified version
of the Library that is interface-compatible with the Linked
Version.
e) Provide Installation Information, but only if you would otherwise
be required to provide such information under section 6 of the
GNU GPL, and only to the extent that such information is
necessary to install and execute a modified version of the
Combined Work produced by recombining or relinking the
Application with a modified version of the Linked Version. (If
you use option 4d0, the Installation Information must accompany
the Minimal Corresponding Source and Corresponding Application
Code. If you use option 4d1, you must provide the Installation
Information in the manner specified by section 6 of the GNU GPL
for conveying Corresponding Source.)
5. Combined Libraries.
You may place library facilities that are a work based on the
Library side by side in a single library together with other library
facilities that are not Applications and are not covered by this
License, and convey such a combined library under terms of your
choice, if you do both of the following:
a) Accompany the combined library with a copy of the same work based
on the Library, uncombined with any other library facilities,
conveyed under the terms of this License.
b) Give prominent notice with the combined library that part of it
is a work based on the Library, and explaining where to find the
accompanying uncombined form of the same work.
6. Revised Versions of the GNU Lesser General Public License.
The Free Software Foundation may publish revised and/or new versions
of the GNU Lesser General Public License from time to time. Such new
versions will be similar in spirit to the present version, but may
differ in detail to address new problems or concerns.
Each version is given a distinguishing version number. If the
Library as you received it specifies that a certain numbered version
of the GNU Lesser General Public License "or any later version"
applies to it, you have the option of following the terms and
conditions either of that published version or of any later version
published by the Free Software Foundation. If the Library as you
received it does not specify a version number of the GNU Lesser
General Public License, you may choose any version of the GNU Lesser
General Public License ever published by the Free Software Foundation.
If the Library as you received it specifies that a proxy can decide
whether future versions of the GNU Lesser General Public License shall
apply, that proxy's public statement of acceptance of any version is
permanent authorization for you to choose that version for the
Library.
7. The Command Line Application uses ezOptionParser with the MIT License:
Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in
all copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
THE SOFTWARE.