I am trying to "make" algorithms but I get this error:
"../../moderngpu2/include/kernels/../device/loadstore.cuh(209): error: calling a host function("thrust::experimental::iterator_facade<thrust::permutation_iterator\<int *,="" int="" *=""> , int *, int, thrust::host_space_tag, thrust::random_access_traversal_tag, int &, long> ::operator []") from a device function("mgpu::DeviceMemToMem4<(int)128, (int)4, thrust::permutation_iterator<int *,="" int="" *=""> , int *> ") is not allowed"</int></thrust::permutation_iterator\<int>
Why? What's wrong?
Vinz,
What environment are you using? OS? CUDA version? etc.
Have you modified anything in the source code?
Can you verify that you are attempting to build from the trunk
(mapgraph-code/trunk)?
Thanks,
Bryan
Bryan Thompson
Chief Scientist & Founder
SYSTAP, LLC
4501 Tower Road
Greensboro, NC 27410
bryan@systap.com
http://bigdata.com
http://mapgraph.io
CONFIDENTIALITY NOTICE: This email and its contents and attachments are
for the sole use of the intended recipient(s) and are confidential or
proprietary to SYSTAP. Any unauthorized review, use, disclosure,
dissemination or copying of this email or its contents or attachments is
prohibited. If you have received this communication in error, please notify
the sender by reply email and permanently delete all copies of the email
and its contents and attachments.
On Tue, Sep 30, 2014 at 7:15 AM, Vinz gannicus15@users.sf.net wrote:
Related
Tickets: #10
Hi Bryan,
Thanks for the reply.
Result of "deviceQuery":
http://i.stack.imgur.com/YXUs6.png
Result of "uname -a":
Linux beowulf 3.2.0-29-generic #46-Ubuntu SMP Fri Jul 27 17:03:23 UTC 2012 x86_64 x86_64 x86_64 GNU/Linux
Result of "nvcc --verions":
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2012 NVIDIA Corporation
Built on Fri_Sep_21_17:28:58_PDT_2012
Cuda compilation tools, release 5.0, V0.2.1221
I have not changed anything in source code.
As README file suggests, I "make" at the top-level (mapgraph.0.3.2 directory).
I solved this issue using a previous version of mapgraph (0.3.1) and adding some missing library (thrust library).
For my thesis of Master's degree in Computer Science, I am going to develop some new strategies. I would be very grateful if you allow me to post more questions in the future, if any.
Thanks a lot for reply, Bryan.
Glad you got that worked out. Yes, thrust is required. Yes, you are
welcome to post more questions. Zhisong (Cc) is the lead CUDA researcher.
If you look at http://MapGraph.io you can find a copy to our most recent
paper on the multi-Gpu version of MapGraph. There is also a publication on
the single Gpu version of MapGraph which may be more relevant if your focus
is Gpu strategies for parallel graph algorithms.
Thanks,
Bryan
On Tuesday, September 30, 2014, Vinz gannicus15@users.sf.net wrote:
--
Bryan Thompson
Chief Scientist & Founder
SYSTAP, LLC
4501 Tower Road
Greensboro, NC 27410
bryan@systap.com
http://bigdata.com
http://mapgraph.io
CONFIDENTIALITY NOTICE: This email and its contents and attachments are
for the sole use of the intended recipient(s) and are confidential or
proprietary to SYSTAP. Any unauthorized review, use, disclosure,
dissemination or copying of this email or its contents or attachments is
prohibited. If you have received this communication in error, please notify
the sender by reply email and permanently delete all copies of the email
and its contents and attachments.
Related
Tickets: #10