Menu
▾
▴
mpgraph-developers — developer discussions
You can subscribe to this list here.
| 2013 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
(3) |
Sep
|
Oct
|
Nov
|
Dec
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2014 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
(2) |
Nov
(5) |
Dec
|
| 2015 |
Jan
(13) |
Feb
(12) |
Mar
(1) |
Apr
(1) |
May
|
Jun
|
Jul
(5) |
Aug
(7) |
Sep
|
Oct
(1) |
Nov
|
Dec
|
|
From: Bryan T. <br...@sy...> - 2015-02-20 19:57:01
|
Please try checking out the code from svn and recompile. I believe this was fixed since the last release. Bryan On Friday, February 20, 2015, leon song <leo...@gm...> wrote: > > Hi, > > I have successfully compiled the Mapgraph and just start trying to run > this SSSP example. I have the following failures: > > [song@sandy mapgraph]$ which nvcc > /usr/local/cuda-6.5/bin/nvcc > [song@sandy mapgraph]$ ./Algorithms/SSSP/SSSP -g > smallRegressionGraphs/small.mtx > Running on host: sandy > Detected 1 CUDA Capable device(s) > Running on this device: > Device 0: "Tesla K20c" > CUDA Driver Version / Runtime Version 6.5 / 6.5 > CUDA Capability Major/Minor version number: 3.5 > Total amount of global memory: 4800 MBytes (5032706048 > bytes) > Reading from smallRegressionGraphs/small.mtx: > Parsing MARKET COO format (7 nodes, 11 directed edges)... Done parsing > (0s). > Converting 7 vertices, 11 directed edges (unordered tuples) to CSR > format... Done converting (0s). > Single source vertex: 0 > GPU 0 column_indices: 11 elements (176 bytes) > GPU 0 row_offsets: 8 elements (32 bytes) > CPU to GPU memory transfer time: 0.038286 ms > num_srcs=1, src=0, iter_num=2147483647 > Starting vertex: 0 > [../../GASengine/enactor_vertex_centric.cuh, 206] EnactorVertexCentric > cudaBindTexture bitmask_tex_ref failed (CUDA error 11: invalid argument) > > > Could you please help me with this? > > Thanks, > > Shuaiwen Leon Song > > Staff research Scientist > > PNNL > > -- ---- Bryan Thompson Chief Scientist & Founder SYSTAP, LLC 4501 Tower Road Greensboro, NC 27410 br...@sy... 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. |
|
From: leon s. <leo...@gm...> - 2015-02-20 19:46:26
|
Hi, I have successfully compiled the Mapgraph and just start trying to run this SSSP example. I have the following failures: [song@sandy mapgraph]$ which nvcc /usr/local/cuda-6.5/bin/nvcc [song@sandy mapgraph]$ ./Algorithms/SSSP/SSSP -g smallRegressionGraphs/small.mtx Running on host: sandy Detected 1 CUDA Capable device(s) Running on this device: Device 0: "Tesla K20c" CUDA Driver Version / Runtime Version 6.5 / 6.5 CUDA Capability Major/Minor version number: 3.5 Total amount of global memory: 4800 MBytes (5032706048 bytes) Reading from smallRegressionGraphs/small.mtx: Parsing MARKET COO format (7 nodes, 11 directed edges)... Done parsing (0s). Converting 7 vertices, 11 directed edges (unordered tuples) to CSR format... Done converting (0s). Single source vertex: 0 GPU 0 column_indices: 11 elements (176 bytes) GPU 0 row_offsets: 8 elements (32 bytes) CPU to GPU memory transfer time: 0.038286 ms num_srcs=1, src=0, iter_num=2147483647 Starting vertex: 0 [../../GASengine/enactor_vertex_centric.cuh, 206] EnactorVertexCentric cudaBindTexture bitmask_tex_ref failed (CUDA error 11: invalid argument) Could you please help me with this? Thanks, Shuaiwen Leon Song Staff research Scientist PNNL |
|
From: Bryan T. <br...@sy...> - 2015-01-25 19:42:51
|
Not sure what happened. Try this link: https://www.dropbox.com/s/745bceva9t4ra4i/bitcoin.mtx?dl=0 Bryan ---- Bryan Thompson Chief Scientist & Founder SYSTAP, LLC 4501 Tower Road Greensboro, NC 27410 br...@sy... 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 Sun, Jan 25, 2015 at 2:30 PM, Kapil Agarwal <ka...@ga...> wrote: > Hi > > The link to the Bitcoin dataset in the paper doesn't work anymore and I > require that dataset. Could you please send it to me or a link from where I > can download it ? > > Thanks > ᐧ > > Best, Kapil > > On Thu, Jan 22, 2015 at 9:00 AM, Bryan Thompson <br...@sy...> wrote: > >> I just checked the SVN code on SourceForge. This fix was applied to SVN >> several months ago. I assume that you were using the download rather than >> the SVN trunk. The SVN trunk should be Ok for CUDA 6. >> >> Thanks, >> Bryan >> >> ---- >> Bryan Thompson >> Chief Scientist & Founder >> SYSTAP, LLC >> 4501 Tower Road >> Greensboro, NC 27410 >> br...@sy... >> 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 Thu, Jan 22, 2015 at 8:43 AM, Kapil Agarwal <ka...@ga...> wrote: >> >>> Thanks Bryan. This worked. >>> ᐧ >>> >>> Best, Kapil >>> >>> On Thu, Jan 22, 2015 at 7:00 AM, Bryan Thompson <br...@sy...> >>> wrote: >>> >>>> Kapil, >>>> >>>> Please try the following patch for CUDA 6 support. Let me know if this >>>> makes it possible for you to compile and execute correctly under CUDA 6 and >>>> I will apply this patch to the SF repository. >>>> >>>> Thanks, >>>> Bryan >>>> >>>> GASengine/enactor_vertex_centric.cuh >>>> >>>> cudaBindTexture(0, vertex_centric::contract_atomic::BitmaskTex< >>>> VisitedMask>::ref,- graph_slice->d_visited_mask, bitmask_desc,+ >>>> graph_slice->d_visited_mask, + //bitmask_desc, bytes), "EnactorVertexCentric >>>> cudaBindTexture bitmask_tex_ref failed", __FILE__, __LINE__)) >>>> >>>> >>>> ---- >>>> Bryan Thompson >>>> Chief Scientist & Founder >>>> SYSTAP, LLC >>>> 4501 Tower Road >>>> Greensboro, NC 27410 >>>> br...@sy... >>>> 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 Wed, Jan 21, 2015 at 8:22 AM, Kapil Agarwal <ka...@ga...> >>>> wrote: >>>> >>>>> CUDA 6.0 >>>>> ᐧ >>>>> >>>>> Best, Kapil >>>>> >>>>> On Wed, Jan 21, 2015 at 8:18 AM, Bryan Thompson <br...@sy...> >>>>> wrote: >>>>> >>>>>> What CUDA version? >>>>>> >>>>>> Bryan >>>>>> >>>>>> ---- >>>>>> Bryan Thompson >>>>>> Chief Scientist & Founder >>>>>> SYSTAP, LLC >>>>>> 4501 Tower Road >>>>>> Greensboro, NC 27410 >>>>>> br...@sy... >>>>>> 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 Wed, Jan 21, 2015 at 8:17 AM, Kapil Agarwal <ka...@ga...> >>>>>> wrote: >>>>>> >>>>>>> I added the following line to all the algorithm files >>>>>>> #include <unistd.h> >>>>>>> as gethostname() was not found >>>>>>> >>>>>>> I still get the same error >>>>>>> ᐧ >>>>>>> >>>>>>> Best, Kapil >>>>>>> >>>>>>> On Wed, Jan 21, 2015 at 8:01 AM, Bryan Thompson <br...@sy...> >>>>>>> wrote: >>>>>>> >>>>>>>> Can you try "make clean" and then recompile? Let me know what you >>>>>>>> have modified (if anything) over the version that is in SVN. >>>>>>>> >>>>>>>> Thanks, >>>>>>>> Bryan >>>>>>>> >>>>>>>> ---- >>>>>>>> Bryan Thompson >>>>>>>> Chief Scientist & Founder >>>>>>>> SYSTAP, LLC >>>>>>>> 4501 Tower Road >>>>>>>> Greensboro, NC 27410 >>>>>>>> br...@sy... >>>>>>>> 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 Wed, Jan 21, 2015 at 7:51 AM, Kapil Agarwal <ka...@ga...> >>>>>>>> wrote: >>>>>>>> >>>>>>>>> ᐧ >>>>>>>>> I am using Tesla K40c on Ubuntu 14.04. >>>>>>>>> >>>>>>>>> Best, Kapil >>>>>>>>> >>>>>>>>> On Wed, Jan 21, 2015 at 7:45 AM, Bryan Thompson <br...@sy...> >>>>>>>>> wrote: >>>>>>>>> >>>>>>>>>> What GPU are you using? What operating system? >>>>>>>>>> >>>>>>>>>> Thanks, >>>>>>>>>> Bryan >>>>>>>>>> >>>>>>>>>> ---- >>>>>>>>>> Bryan Thompson >>>>>>>>>> Chief Scientist & Founder >>>>>>>>>> SYSTAP, LLC >>>>>>>>>> 4501 Tower Road >>>>>>>>>> Greensboro, NC 27410 >>>>>>>>>> br...@sy... >>>>>>>>>> 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, Jan 20, 2015 at 7:51 PM, Kapil Agarwal <ka...@ga... >>>>>>>>>> > wrote: >>>>>>>>>> >>>>>>>>>>> Hi >>>>>>>>>>> >>>>>>>>>>> I installed MapGraph and ran it using >>>>>>>>>>> ./Algorithms/PageRank/PageRank -g webbase-1M.mtx >>>>>>>>>>> >>>>>>>>>>> I get the following errrors >>>>>>>>>>> [../../GASengine/enactor_vertex_centric.cuh, 205] >>>>>>>>>>> EnactorVertexCentric cudaBindTexture bitmask_tex_ref failed (CUDA error 11: >>>>>>>>>>> invalid argument) >>>>>>>>>>> >>>>>>>>>>> Please help me to fix it. >>>>>>>>>>> >>>>>>>>>>> Best, Kapil >>>>>>>>>>> >>>>>>>>>>> >>>>>>>>>>> ------------------------------------------------------------------------------ >>>>>>>>>>> New Year. New Location. New Benefits. New Data Center in >>>>>>>>>>> Ashburn, VA. >>>>>>>>>>> GigeNET is offering a free month of service with a new server in >>>>>>>>>>> Ashburn. >>>>>>>>>>> Choose from 2 high performing configs, both with 100TB of >>>>>>>>>>> bandwidth. >>>>>>>>>>> Higher redundancy.Lower latency.Increased capacity.Completely >>>>>>>>>>> compliant. >>>>>>>>>>> http://p.sf.net/sfu/gigenet >>>>>>>>>>> _______________________________________________ >>>>>>>>>>> mpgraph-developers mailing list >>>>>>>>>>> mpg...@li... >>>>>>>>>>> https://lists.sourceforge.net/lists/listinfo/mpgraph-developers >>>>>>>>>>> >>>>>>>>>>> >>>>>>>>>> >>>>>>>>> >>>>>>>> >>>>>>> >>>>>> >>>>> >>>> >>> >> >> >> ------------------------------------------------------------------------------ >> New Year. New Location. New Benefits. New Data Center in Ashburn, VA. >> GigeNET is offering a free month of service with a new server in Ashburn. >> Choose from 2 high performing configs, both with 100TB of bandwidth. >> Higher redundancy.Lower latency.Increased capacity.Completely compliant. >> http://p.sf.net/sfu/gigenet >> _______________________________________________ >> mpgraph-developers mailing list >> mpg...@li... >> https://lists.sourceforge.net/lists/listinfo/mpgraph-developers >> >> > |
|
From: Kapil A. <ka...@ga...> - 2015-01-25 19:30:53
|
Hi The link to the Bitcoin dataset in the paper doesn't work anymore and I require that dataset. Could you please send it to me or a link from where I can download it ? Thanks ᐧ Best, Kapil On Thu, Jan 22, 2015 at 9:00 AM, Bryan Thompson <br...@sy...> wrote: > I just checked the SVN code on SourceForge. This fix was applied to SVN > several months ago. I assume that you were using the download rather than > the SVN trunk. The SVN trunk should be Ok for CUDA 6. > > Thanks, > Bryan > > ---- > Bryan Thompson > Chief Scientist & Founder > SYSTAP, LLC > 4501 Tower Road > Greensboro, NC 27410 > br...@sy... > 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 Thu, Jan 22, 2015 at 8:43 AM, Kapil Agarwal <ka...@ga...> wrote: > >> Thanks Bryan. This worked. >> ᐧ >> >> Best, Kapil >> >> On Thu, Jan 22, 2015 at 7:00 AM, Bryan Thompson <br...@sy...> wrote: >> >>> Kapil, >>> >>> Please try the following patch for CUDA 6 support. Let me know if this >>> makes it possible for you to compile and execute correctly under CUDA 6 and >>> I will apply this patch to the SF repository. >>> >>> Thanks, >>> Bryan >>> >>> GASengine/enactor_vertex_centric.cuh >>> >>> cudaBindTexture(0, vertex_centric::contract_atomic::BitmaskTex< >>> VisitedMask>::ref,- graph_slice->d_visited_mask, bitmask_desc,+ >>> graph_slice->d_visited_mask, + //bitmask_desc, bytes), "EnactorVertexCentric >>> cudaBindTexture bitmask_tex_ref failed", __FILE__, __LINE__)) >>> >>> >>> ---- >>> Bryan Thompson >>> Chief Scientist & Founder >>> SYSTAP, LLC >>> 4501 Tower Road >>> Greensboro, NC 27410 >>> br...@sy... >>> 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 Wed, Jan 21, 2015 at 8:22 AM, Kapil Agarwal <ka...@ga...> >>> wrote: >>> >>>> CUDA 6.0 >>>> ᐧ >>>> >>>> Best, Kapil >>>> >>>> On Wed, Jan 21, 2015 at 8:18 AM, Bryan Thompson <br...@sy...> >>>> wrote: >>>> >>>>> What CUDA version? >>>>> >>>>> Bryan >>>>> >>>>> ---- >>>>> Bryan Thompson >>>>> Chief Scientist & Founder >>>>> SYSTAP, LLC >>>>> 4501 Tower Road >>>>> Greensboro, NC 27410 >>>>> br...@sy... >>>>> 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 Wed, Jan 21, 2015 at 8:17 AM, Kapil Agarwal <ka...@ga...> >>>>> wrote: >>>>> >>>>>> I added the following line to all the algorithm files >>>>>> #include <unistd.h> >>>>>> as gethostname() was not found >>>>>> >>>>>> I still get the same error >>>>>> ᐧ >>>>>> >>>>>> Best, Kapil >>>>>> >>>>>> On Wed, Jan 21, 2015 at 8:01 AM, Bryan Thompson <br...@sy...> >>>>>> wrote: >>>>>> >>>>>>> Can you try "make clean" and then recompile? Let me know what you >>>>>>> have modified (if anything) over the version that is in SVN. >>>>>>> >>>>>>> Thanks, >>>>>>> Bryan >>>>>>> >>>>>>> ---- >>>>>>> Bryan Thompson >>>>>>> Chief Scientist & Founder >>>>>>> SYSTAP, LLC >>>>>>> 4501 Tower Road >>>>>>> Greensboro, NC 27410 >>>>>>> br...@sy... >>>>>>> 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 Wed, Jan 21, 2015 at 7:51 AM, Kapil Agarwal <ka...@ga...> >>>>>>> wrote: >>>>>>> >>>>>>>> ᐧ >>>>>>>> I am using Tesla K40c on Ubuntu 14.04. >>>>>>>> >>>>>>>> Best, Kapil >>>>>>>> >>>>>>>> On Wed, Jan 21, 2015 at 7:45 AM, Bryan Thompson <br...@sy...> >>>>>>>> wrote: >>>>>>>> >>>>>>>>> What GPU are you using? What operating system? >>>>>>>>> >>>>>>>>> Thanks, >>>>>>>>> Bryan >>>>>>>>> >>>>>>>>> ---- >>>>>>>>> Bryan Thompson >>>>>>>>> Chief Scientist & Founder >>>>>>>>> SYSTAP, LLC >>>>>>>>> 4501 Tower Road >>>>>>>>> Greensboro, NC 27410 >>>>>>>>> br...@sy... >>>>>>>>> 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, Jan 20, 2015 at 7:51 PM, Kapil Agarwal <ka...@ga...> >>>>>>>>> wrote: >>>>>>>>> >>>>>>>>>> Hi >>>>>>>>>> >>>>>>>>>> I installed MapGraph and ran it using >>>>>>>>>> ./Algorithms/PageRank/PageRank -g webbase-1M.mtx >>>>>>>>>> >>>>>>>>>> I get the following errrors >>>>>>>>>> [../../GASengine/enactor_vertex_centric.cuh, 205] >>>>>>>>>> EnactorVertexCentric cudaBindTexture bitmask_tex_ref failed (CUDA error 11: >>>>>>>>>> invalid argument) >>>>>>>>>> >>>>>>>>>> Please help me to fix it. >>>>>>>>>> >>>>>>>>>> Best, Kapil >>>>>>>>>> >>>>>>>>>> >>>>>>>>>> ------------------------------------------------------------------------------ >>>>>>>>>> New Year. New Location. New Benefits. New Data Center in Ashburn, >>>>>>>>>> VA. >>>>>>>>>> GigeNET is offering a free month of service with a new server in >>>>>>>>>> Ashburn. >>>>>>>>>> Choose from 2 high performing configs, both with 100TB of >>>>>>>>>> bandwidth. >>>>>>>>>> Higher redundancy.Lower latency.Increased capacity.Completely >>>>>>>>>> compliant. >>>>>>>>>> http://p.sf.net/sfu/gigenet >>>>>>>>>> _______________________________________________ >>>>>>>>>> mpgraph-developers mailing list >>>>>>>>>> mpg...@li... >>>>>>>>>> https://lists.sourceforge.net/lists/listinfo/mpgraph-developers >>>>>>>>>> >>>>>>>>>> >>>>>>>>> >>>>>>>> >>>>>>> >>>>>> >>>>> >>>> >>> >> > > > ------------------------------------------------------------------------------ > New Year. New Location. New Benefits. New Data Center in Ashburn, VA. > GigeNET is offering a free month of service with a new server in Ashburn. > Choose from 2 high performing configs, both with 100TB of bandwidth. > Higher redundancy.Lower latency.Increased capacity.Completely compliant. > http://p.sf.net/sfu/gigenet > _______________________________________________ > mpgraph-developers mailing list > mpg...@li... > https://lists.sourceforge.net/lists/listinfo/mpgraph-developers > > |
|
From: Bryan T. <br...@sy...> - 2015-01-22 14:00:44
|
I just checked the SVN code on SourceForge. This fix was applied to SVN several months ago. I assume that you were using the download rather than the SVN trunk. The SVN trunk should be Ok for CUDA 6. Thanks, Bryan ---- Bryan Thompson Chief Scientist & Founder SYSTAP, LLC 4501 Tower Road Greensboro, NC 27410 br...@sy... 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 Thu, Jan 22, 2015 at 8:43 AM, Kapil Agarwal <ka...@ga...> wrote: > Thanks Bryan. This worked. > ᐧ > > Best, Kapil > > On Thu, Jan 22, 2015 at 7:00 AM, Bryan Thompson <br...@sy...> wrote: > >> Kapil, >> >> Please try the following patch for CUDA 6 support. Let me know if this >> makes it possible for you to compile and execute correctly under CUDA 6 and >> I will apply this patch to the SF repository. >> >> Thanks, >> Bryan >> >> GASengine/enactor_vertex_centric.cuh >> >> cudaBindTexture(0, vertex_centric::contract_atomic::BitmaskTex< >> VisitedMask>::ref,- graph_slice->d_visited_mask, bitmask_desc,+ >> graph_slice->d_visited_mask, + //bitmask_desc, bytes), "EnactorVertexCentric >> cudaBindTexture bitmask_tex_ref failed", __FILE__, __LINE__)) >> >> >> ---- >> Bryan Thompson >> Chief Scientist & Founder >> SYSTAP, LLC >> 4501 Tower Road >> Greensboro, NC 27410 >> br...@sy... >> 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 Wed, Jan 21, 2015 at 8:22 AM, Kapil Agarwal <ka...@ga...> wrote: >> >>> CUDA 6.0 >>> ᐧ >>> >>> Best, Kapil >>> >>> On Wed, Jan 21, 2015 at 8:18 AM, Bryan Thompson <br...@sy...> >>> wrote: >>> >>>> What CUDA version? >>>> >>>> Bryan >>>> >>>> ---- >>>> Bryan Thompson >>>> Chief Scientist & Founder >>>> SYSTAP, LLC >>>> 4501 Tower Road >>>> Greensboro, NC 27410 >>>> br...@sy... >>>> 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 Wed, Jan 21, 2015 at 8:17 AM, Kapil Agarwal <ka...@ga...> >>>> wrote: >>>> >>>>> I added the following line to all the algorithm files >>>>> #include <unistd.h> >>>>> as gethostname() was not found >>>>> >>>>> I still get the same error >>>>> ᐧ >>>>> >>>>> Best, Kapil >>>>> >>>>> On Wed, Jan 21, 2015 at 8:01 AM, Bryan Thompson <br...@sy...> >>>>> wrote: >>>>> >>>>>> Can you try "make clean" and then recompile? Let me know what you >>>>>> have modified (if anything) over the version that is in SVN. >>>>>> >>>>>> Thanks, >>>>>> Bryan >>>>>> >>>>>> ---- >>>>>> Bryan Thompson >>>>>> Chief Scientist & Founder >>>>>> SYSTAP, LLC >>>>>> 4501 Tower Road >>>>>> Greensboro, NC 27410 >>>>>> br...@sy... >>>>>> 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 Wed, Jan 21, 2015 at 7:51 AM, Kapil Agarwal <ka...@ga...> >>>>>> wrote: >>>>>> >>>>>>> ᐧ >>>>>>> I am using Tesla K40c on Ubuntu 14.04. >>>>>>> >>>>>>> Best, Kapil >>>>>>> >>>>>>> On Wed, Jan 21, 2015 at 7:45 AM, Bryan Thompson <br...@sy...> >>>>>>> wrote: >>>>>>> >>>>>>>> What GPU are you using? What operating system? >>>>>>>> >>>>>>>> Thanks, >>>>>>>> Bryan >>>>>>>> >>>>>>>> ---- >>>>>>>> Bryan Thompson >>>>>>>> Chief Scientist & Founder >>>>>>>> SYSTAP, LLC >>>>>>>> 4501 Tower Road >>>>>>>> Greensboro, NC 27410 >>>>>>>> br...@sy... >>>>>>>> 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, Jan 20, 2015 at 7:51 PM, Kapil Agarwal <ka...@ga...> >>>>>>>> wrote: >>>>>>>> >>>>>>>>> Hi >>>>>>>>> >>>>>>>>> I installed MapGraph and ran it using >>>>>>>>> ./Algorithms/PageRank/PageRank -g webbase-1M.mtx >>>>>>>>> >>>>>>>>> I get the following errrors >>>>>>>>> [../../GASengine/enactor_vertex_centric.cuh, 205] >>>>>>>>> EnactorVertexCentric cudaBindTexture bitmask_tex_ref failed (CUDA error 11: >>>>>>>>> invalid argument) >>>>>>>>> >>>>>>>>> Please help me to fix it. >>>>>>>>> >>>>>>>>> Best, Kapil >>>>>>>>> >>>>>>>>> >>>>>>>>> ------------------------------------------------------------------------------ >>>>>>>>> New Year. New Location. New Benefits. New Data Center in Ashburn, >>>>>>>>> VA. >>>>>>>>> GigeNET is offering a free month of service with a new server in >>>>>>>>> Ashburn. >>>>>>>>> Choose from 2 high performing configs, both with 100TB of >>>>>>>>> bandwidth. >>>>>>>>> Higher redundancy.Lower latency.Increased capacity.Completely >>>>>>>>> compliant. >>>>>>>>> http://p.sf.net/sfu/gigenet >>>>>>>>> _______________________________________________ >>>>>>>>> mpgraph-developers mailing list >>>>>>>>> mpg...@li... >>>>>>>>> https://lists.sourceforge.net/lists/listinfo/mpgraph-developers >>>>>>>>> >>>>>>>>> >>>>>>>> >>>>>>> >>>>>> >>>>> >>>> >>> >> > |
|
From: Kapil A. <ka...@ga...> - 2015-01-22 13:44:15
|
Thanks Bryan. This worked. ᐧ Best, Kapil On Thu, Jan 22, 2015 at 7:00 AM, Bryan Thompson <br...@sy...> wrote: > Kapil, > > Please try the following patch for CUDA 6 support. Let me know if this > makes it possible for you to compile and execute correctly under CUDA 6 and > I will apply this patch to the SF repository. > > Thanks, > Bryan > > GASengine/enactor_vertex_centric.cuh > > cudaBindTexture(0, vertex_centric::contract_atomic::BitmaskTex< > VisitedMask>::ref,- graph_slice->d_visited_mask, bitmask_desc,+ > graph_slice->d_visited_mask, + //bitmask_desc, bytes), "EnactorVertexCentric > cudaBindTexture bitmask_tex_ref failed", __FILE__, __LINE__)) > > > ---- > Bryan Thompson > Chief Scientist & Founder > SYSTAP, LLC > 4501 Tower Road > Greensboro, NC 27410 > br...@sy... > 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 Wed, Jan 21, 2015 at 8:22 AM, Kapil Agarwal <ka...@ga...> wrote: > >> CUDA 6.0 >> ᐧ >> >> Best, Kapil >> >> On Wed, Jan 21, 2015 at 8:18 AM, Bryan Thompson <br...@sy...> wrote: >> >>> What CUDA version? >>> >>> Bryan >>> >>> ---- >>> Bryan Thompson >>> Chief Scientist & Founder >>> SYSTAP, LLC >>> 4501 Tower Road >>> Greensboro, NC 27410 >>> br...@sy... >>> 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 Wed, Jan 21, 2015 at 8:17 AM, Kapil Agarwal <ka...@ga...> >>> wrote: >>> >>>> I added the following line to all the algorithm files >>>> #include <unistd.h> >>>> as gethostname() was not found >>>> >>>> I still get the same error >>>> ᐧ >>>> >>>> Best, Kapil >>>> >>>> On Wed, Jan 21, 2015 at 8:01 AM, Bryan Thompson <br...@sy...> >>>> wrote: >>>> >>>>> Can you try "make clean" and then recompile? Let me know what you >>>>> have modified (if anything) over the version that is in SVN. >>>>> >>>>> Thanks, >>>>> Bryan >>>>> >>>>> ---- >>>>> Bryan Thompson >>>>> Chief Scientist & Founder >>>>> SYSTAP, LLC >>>>> 4501 Tower Road >>>>> Greensboro, NC 27410 >>>>> br...@sy... >>>>> 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 Wed, Jan 21, 2015 at 7:51 AM, Kapil Agarwal <ka...@ga...> >>>>> wrote: >>>>> >>>>>> ᐧ >>>>>> I am using Tesla K40c on Ubuntu 14.04. >>>>>> >>>>>> Best, Kapil >>>>>> >>>>>> On Wed, Jan 21, 2015 at 7:45 AM, Bryan Thompson <br...@sy...> >>>>>> wrote: >>>>>> >>>>>>> What GPU are you using? What operating system? >>>>>>> >>>>>>> Thanks, >>>>>>> Bryan >>>>>>> >>>>>>> ---- >>>>>>> Bryan Thompson >>>>>>> Chief Scientist & Founder >>>>>>> SYSTAP, LLC >>>>>>> 4501 Tower Road >>>>>>> Greensboro, NC 27410 >>>>>>> br...@sy... >>>>>>> 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, Jan 20, 2015 at 7:51 PM, Kapil Agarwal <ka...@ga...> >>>>>>> wrote: >>>>>>> >>>>>>>> Hi >>>>>>>> >>>>>>>> I installed MapGraph and ran it using >>>>>>>> ./Algorithms/PageRank/PageRank -g webbase-1M.mtx >>>>>>>> >>>>>>>> I get the following errrors >>>>>>>> [../../GASengine/enactor_vertex_centric.cuh, 205] >>>>>>>> EnactorVertexCentric cudaBindTexture bitmask_tex_ref failed (CUDA error 11: >>>>>>>> invalid argument) >>>>>>>> >>>>>>>> Please help me to fix it. >>>>>>>> >>>>>>>> Best, Kapil >>>>>>>> >>>>>>>> >>>>>>>> ------------------------------------------------------------------------------ >>>>>>>> New Year. New Location. New Benefits. New Data Center in Ashburn, >>>>>>>> VA. >>>>>>>> GigeNET is offering a free month of service with a new server in >>>>>>>> Ashburn. >>>>>>>> Choose from 2 high performing configs, both with 100TB of bandwidth. >>>>>>>> Higher redundancy.Lower latency.Increased capacity.Completely >>>>>>>> compliant. >>>>>>>> http://p.sf.net/sfu/gigenet >>>>>>>> _______________________________________________ >>>>>>>> mpgraph-developers mailing list >>>>>>>> mpg...@li... >>>>>>>> https://lists.sourceforge.net/lists/listinfo/mpgraph-developers >>>>>>>> >>>>>>>> >>>>>>> >>>>>> >>>>> >>>> >>> >> > |
|
From: Bryan T. <br...@sy...> - 2015-01-22 12:00:29
|
Kapil, Please try the following patch for CUDA 6 support. Let me know if this makes it possible for you to compile and execute correctly under CUDA 6 and I will apply this patch to the SF repository. Thanks, Bryan GASengine/enactor_vertex_centric.cuh cudaBindTexture(0, vertex_centric::contract_atomic::BitmaskTex< VisitedMask>::ref,- graph_slice->d_visited_mask, bitmask_desc,+ graph_slice->d_visited_mask, + //bitmask_desc, bytes), "EnactorVertexCentric cudaBindTexture bitmask_tex_ref failed", __FILE__, __LINE__)) ---- Bryan Thompson Chief Scientist & Founder SYSTAP, LLC 4501 Tower Road Greensboro, NC 27410 br...@sy... 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 Wed, Jan 21, 2015 at 8:22 AM, Kapil Agarwal <ka...@ga...> wrote: > CUDA 6.0 > ᐧ > > Best, Kapil > > On Wed, Jan 21, 2015 at 8:18 AM, Bryan Thompson <br...@sy...> wrote: > >> What CUDA version? >> >> Bryan >> >> ---- >> Bryan Thompson >> Chief Scientist & Founder >> SYSTAP, LLC >> 4501 Tower Road >> Greensboro, NC 27410 >> br...@sy... >> 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 Wed, Jan 21, 2015 at 8:17 AM, Kapil Agarwal <ka...@ga...> wrote: >> >>> I added the following line to all the algorithm files >>> #include <unistd.h> >>> as gethostname() was not found >>> >>> I still get the same error >>> ᐧ >>> >>> Best, Kapil >>> >>> On Wed, Jan 21, 2015 at 8:01 AM, Bryan Thompson <br...@sy...> >>> wrote: >>> >>>> Can you try "make clean" and then recompile? Let me know what you have >>>> modified (if anything) over the version that is in SVN. >>>> >>>> Thanks, >>>> Bryan >>>> >>>> ---- >>>> Bryan Thompson >>>> Chief Scientist & Founder >>>> SYSTAP, LLC >>>> 4501 Tower Road >>>> Greensboro, NC 27410 >>>> br...@sy... >>>> 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 Wed, Jan 21, 2015 at 7:51 AM, Kapil Agarwal <ka...@ga...> >>>> wrote: >>>> >>>>> ᐧ >>>>> I am using Tesla K40c on Ubuntu 14.04. >>>>> >>>>> Best, Kapil >>>>> >>>>> On Wed, Jan 21, 2015 at 7:45 AM, Bryan Thompson <br...@sy...> >>>>> wrote: >>>>> >>>>>> What GPU are you using? What operating system? >>>>>> >>>>>> Thanks, >>>>>> Bryan >>>>>> >>>>>> ---- >>>>>> Bryan Thompson >>>>>> Chief Scientist & Founder >>>>>> SYSTAP, LLC >>>>>> 4501 Tower Road >>>>>> Greensboro, NC 27410 >>>>>> br...@sy... >>>>>> 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, Jan 20, 2015 at 7:51 PM, Kapil Agarwal <ka...@ga...> >>>>>> wrote: >>>>>> >>>>>>> Hi >>>>>>> >>>>>>> I installed MapGraph and ran it using >>>>>>> ./Algorithms/PageRank/PageRank -g webbase-1M.mtx >>>>>>> >>>>>>> I get the following errrors >>>>>>> [../../GASengine/enactor_vertex_centric.cuh, 205] >>>>>>> EnactorVertexCentric cudaBindTexture bitmask_tex_ref failed (CUDA error 11: >>>>>>> invalid argument) >>>>>>> >>>>>>> Please help me to fix it. >>>>>>> >>>>>>> Best, Kapil >>>>>>> >>>>>>> >>>>>>> ------------------------------------------------------------------------------ >>>>>>> New Year. New Location. New Benefits. New Data Center in Ashburn, VA. >>>>>>> GigeNET is offering a free month of service with a new server in >>>>>>> Ashburn. >>>>>>> Choose from 2 high performing configs, both with 100TB of bandwidth. >>>>>>> Higher redundancy.Lower latency.Increased capacity.Completely >>>>>>> compliant. >>>>>>> http://p.sf.net/sfu/gigenet >>>>>>> _______________________________________________ >>>>>>> mpgraph-developers mailing list >>>>>>> mpg...@li... >>>>>>> https://lists.sourceforge.net/lists/listinfo/mpgraph-developers >>>>>>> >>>>>>> >>>>>> >>>>> >>>> >>> >> > |
|
From: Bryan T. <br...@sy...> - 2015-01-21 13:27:31
|
Ok. I think that this issue emerged with CUDA 6.0. We have a fix in a private branch. I will try to track down the code change for this and apply a patch to the SVN trunk. In the meanwhile, you might try CUDA 5.0. It should work out of the box. Thanks, Bryan ---- Bryan Thompson Chief Scientist & Founder SYSTAP, LLC 4501 Tower Road Greensboro, NC 27410 br...@sy... 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 Wed, Jan 21, 2015 at 8:22 AM, Kapil Agarwal <ka...@ga...> wrote: > CUDA 6.0 > ᐧ > > Best, Kapil > > On Wed, Jan 21, 2015 at 8:18 AM, Bryan Thompson <br...@sy...> wrote: > >> What CUDA version? >> >> Bryan >> >> ---- >> Bryan Thompson >> Chief Scientist & Founder >> SYSTAP, LLC >> 4501 Tower Road >> Greensboro, NC 27410 >> br...@sy... >> 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 Wed, Jan 21, 2015 at 8:17 AM, Kapil Agarwal <ka...@ga...> wrote: >> >>> I added the following line to all the algorithm files >>> #include <unistd.h> >>> as gethostname() was not found >>> >>> I still get the same error >>> ᐧ >>> >>> Best, Kapil >>> >>> On Wed, Jan 21, 2015 at 8:01 AM, Bryan Thompson <br...@sy...> >>> wrote: >>> >>>> Can you try "make clean" and then recompile? Let me know what you have >>>> modified (if anything) over the version that is in SVN. >>>> >>>> Thanks, >>>> Bryan >>>> >>>> ---- >>>> Bryan Thompson >>>> Chief Scientist & Founder >>>> SYSTAP, LLC >>>> 4501 Tower Road >>>> Greensboro, NC 27410 >>>> br...@sy... >>>> 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 Wed, Jan 21, 2015 at 7:51 AM, Kapil Agarwal <ka...@ga...> >>>> wrote: >>>> >>>>> ᐧ >>>>> I am using Tesla K40c on Ubuntu 14.04. >>>>> >>>>> Best, Kapil >>>>> >>>>> On Wed, Jan 21, 2015 at 7:45 AM, Bryan Thompson <br...@sy...> >>>>> wrote: >>>>> >>>>>> What GPU are you using? What operating system? >>>>>> >>>>>> Thanks, >>>>>> Bryan >>>>>> >>>>>> ---- >>>>>> Bryan Thompson >>>>>> Chief Scientist & Founder >>>>>> SYSTAP, LLC >>>>>> 4501 Tower Road >>>>>> Greensboro, NC 27410 >>>>>> br...@sy... >>>>>> 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, Jan 20, 2015 at 7:51 PM, Kapil Agarwal <ka...@ga...> >>>>>> wrote: >>>>>> >>>>>>> Hi >>>>>>> >>>>>>> I installed MapGraph and ran it using >>>>>>> ./Algorithms/PageRank/PageRank -g webbase-1M.mtx >>>>>>> >>>>>>> I get the following errrors >>>>>>> [../../GASengine/enactor_vertex_centric.cuh, 205] >>>>>>> EnactorVertexCentric cudaBindTexture bitmask_tex_ref failed (CUDA error 11: >>>>>>> invalid argument) >>>>>>> >>>>>>> Please help me to fix it. >>>>>>> >>>>>>> Best, Kapil >>>>>>> >>>>>>> >>>>>>> ------------------------------------------------------------------------------ >>>>>>> New Year. New Location. New Benefits. New Data Center in Ashburn, VA. >>>>>>> GigeNET is offering a free month of service with a new server in >>>>>>> Ashburn. >>>>>>> Choose from 2 high performing configs, both with 100TB of bandwidth. >>>>>>> Higher redundancy.Lower latency.Increased capacity.Completely >>>>>>> compliant. >>>>>>> http://p.sf.net/sfu/gigenet >>>>>>> _______________________________________________ >>>>>>> mpgraph-developers mailing list >>>>>>> mpg...@li... >>>>>>> https://lists.sourceforge.net/lists/listinfo/mpgraph-developers >>>>>>> >>>>>>> >>>>>> >>>>> >>>> >>> >> > |
|
From: Kapil A. <ka...@ga...> - 2015-01-21 13:23:12
|
CUDA 6.0 ᐧ Best, Kapil On Wed, Jan 21, 2015 at 8:18 AM, Bryan Thompson <br...@sy...> wrote: > What CUDA version? > > Bryan > > ---- > Bryan Thompson > Chief Scientist & Founder > SYSTAP, LLC > 4501 Tower Road > Greensboro, NC 27410 > br...@sy... > 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 Wed, Jan 21, 2015 at 8:17 AM, Kapil Agarwal <ka...@ga...> wrote: > >> I added the following line to all the algorithm files >> #include <unistd.h> >> as gethostname() was not found >> >> I still get the same error >> ᐧ >> >> Best, Kapil >> >> On Wed, Jan 21, 2015 at 8:01 AM, Bryan Thompson <br...@sy...> wrote: >> >>> Can you try "make clean" and then recompile? Let me know what you have >>> modified (if anything) over the version that is in SVN. >>> >>> Thanks, >>> Bryan >>> >>> ---- >>> Bryan Thompson >>> Chief Scientist & Founder >>> SYSTAP, LLC >>> 4501 Tower Road >>> Greensboro, NC 27410 >>> br...@sy... >>> 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 Wed, Jan 21, 2015 at 7:51 AM, Kapil Agarwal <ka...@ga...> >>> wrote: >>> >>>> ᐧ >>>> I am using Tesla K40c on Ubuntu 14.04. >>>> >>>> Best, Kapil >>>> >>>> On Wed, Jan 21, 2015 at 7:45 AM, Bryan Thompson <br...@sy...> >>>> wrote: >>>> >>>>> What GPU are you using? What operating system? >>>>> >>>>> Thanks, >>>>> Bryan >>>>> >>>>> ---- >>>>> Bryan Thompson >>>>> Chief Scientist & Founder >>>>> SYSTAP, LLC >>>>> 4501 Tower Road >>>>> Greensboro, NC 27410 >>>>> br...@sy... >>>>> 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, Jan 20, 2015 at 7:51 PM, Kapil Agarwal <ka...@ga...> >>>>> wrote: >>>>> >>>>>> Hi >>>>>> >>>>>> I installed MapGraph and ran it using >>>>>> ./Algorithms/PageRank/PageRank -g webbase-1M.mtx >>>>>> >>>>>> I get the following errrors >>>>>> [../../GASengine/enactor_vertex_centric.cuh, 205] >>>>>> EnactorVertexCentric cudaBindTexture bitmask_tex_ref failed (CUDA error 11: >>>>>> invalid argument) >>>>>> >>>>>> Please help me to fix it. >>>>>> >>>>>> Best, Kapil >>>>>> >>>>>> >>>>>> ------------------------------------------------------------------------------ >>>>>> New Year. New Location. New Benefits. New Data Center in Ashburn, VA. >>>>>> GigeNET is offering a free month of service with a new server in >>>>>> Ashburn. >>>>>> Choose from 2 high performing configs, both with 100TB of bandwidth. >>>>>> Higher redundancy.Lower latency.Increased capacity.Completely >>>>>> compliant. >>>>>> http://p.sf.net/sfu/gigenet >>>>>> _______________________________________________ >>>>>> mpgraph-developers mailing list >>>>>> mpg...@li... >>>>>> https://lists.sourceforge.net/lists/listinfo/mpgraph-developers >>>>>> >>>>>> >>>>> >>>> >>> >> > |
|
From: Bryan T. <br...@sy...> - 2015-01-21 13:19:03
|
What CUDA version? Bryan ---- Bryan Thompson Chief Scientist & Founder SYSTAP, LLC 4501 Tower Road Greensboro, NC 27410 br...@sy... 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 Wed, Jan 21, 2015 at 8:17 AM, Kapil Agarwal <ka...@ga...> wrote: > I added the following line to all the algorithm files > #include <unistd.h> > as gethostname() was not found > > I still get the same error > ᐧ > > Best, Kapil > > On Wed, Jan 21, 2015 at 8:01 AM, Bryan Thompson <br...@sy...> wrote: > >> Can you try "make clean" and then recompile? Let me know what you have >> modified (if anything) over the version that is in SVN. >> >> Thanks, >> Bryan >> >> ---- >> Bryan Thompson >> Chief Scientist & Founder >> SYSTAP, LLC >> 4501 Tower Road >> Greensboro, NC 27410 >> br...@sy... >> 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 Wed, Jan 21, 2015 at 7:51 AM, Kapil Agarwal <ka...@ga...> wrote: >> >>> ᐧ >>> I am using Tesla K40c on Ubuntu 14.04. >>> >>> Best, Kapil >>> >>> On Wed, Jan 21, 2015 at 7:45 AM, Bryan Thompson <br...@sy...> >>> wrote: >>> >>>> What GPU are you using? What operating system? >>>> >>>> Thanks, >>>> Bryan >>>> >>>> ---- >>>> Bryan Thompson >>>> Chief Scientist & Founder >>>> SYSTAP, LLC >>>> 4501 Tower Road >>>> Greensboro, NC 27410 >>>> br...@sy... >>>> 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, Jan 20, 2015 at 7:51 PM, Kapil Agarwal <ka...@ga...> >>>> wrote: >>>> >>>>> Hi >>>>> >>>>> I installed MapGraph and ran it using >>>>> ./Algorithms/PageRank/PageRank -g webbase-1M.mtx >>>>> >>>>> I get the following errrors >>>>> [../../GASengine/enactor_vertex_centric.cuh, 205] EnactorVertexCentric >>>>> cudaBindTexture bitmask_tex_ref failed (CUDA error 11: invalid argument) >>>>> >>>>> Please help me to fix it. >>>>> >>>>> Best, Kapil >>>>> >>>>> >>>>> ------------------------------------------------------------------------------ >>>>> New Year. New Location. New Benefits. New Data Center in Ashburn, VA. >>>>> GigeNET is offering a free month of service with a new server in >>>>> Ashburn. >>>>> Choose from 2 high performing configs, both with 100TB of bandwidth. >>>>> Higher redundancy.Lower latency.Increased capacity.Completely >>>>> compliant. >>>>> http://p.sf.net/sfu/gigenet >>>>> _______________________________________________ >>>>> mpgraph-developers mailing list >>>>> mpg...@li... >>>>> https://lists.sourceforge.net/lists/listinfo/mpgraph-developers >>>>> >>>>> >>>> >>> >> > |
|
From: Kapil A. <ka...@ga...> - 2015-01-21 13:17:29
|
I added the following line to all the algorithm files #include <unistd.h> as gethostname() was not found I still get the same error ᐧ Best, Kapil On Wed, Jan 21, 2015 at 8:01 AM, Bryan Thompson <br...@sy...> wrote: > Can you try "make clean" and then recompile? Let me know what you have > modified (if anything) over the version that is in SVN. > > Thanks, > Bryan > > ---- > Bryan Thompson > Chief Scientist & Founder > SYSTAP, LLC > 4501 Tower Road > Greensboro, NC 27410 > br...@sy... > 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 Wed, Jan 21, 2015 at 7:51 AM, Kapil Agarwal <ka...@ga...> wrote: > >> ᐧ >> I am using Tesla K40c on Ubuntu 14.04. >> >> Best, Kapil >> >> On Wed, Jan 21, 2015 at 7:45 AM, Bryan Thompson <br...@sy...> wrote: >> >>> What GPU are you using? What operating system? >>> >>> Thanks, >>> Bryan >>> >>> ---- >>> Bryan Thompson >>> Chief Scientist & Founder >>> SYSTAP, LLC >>> 4501 Tower Road >>> Greensboro, NC 27410 >>> br...@sy... >>> 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, Jan 20, 2015 at 7:51 PM, Kapil Agarwal <ka...@ga...> >>> wrote: >>> >>>> Hi >>>> >>>> I installed MapGraph and ran it using >>>> ./Algorithms/PageRank/PageRank -g webbase-1M.mtx >>>> >>>> I get the following errrors >>>> [../../GASengine/enactor_vertex_centric.cuh, 205] EnactorVertexCentric >>>> cudaBindTexture bitmask_tex_ref failed (CUDA error 11: invalid argument) >>>> >>>> Please help me to fix it. >>>> >>>> Best, Kapil >>>> >>>> >>>> ------------------------------------------------------------------------------ >>>> New Year. New Location. New Benefits. New Data Center in Ashburn, VA. >>>> GigeNET is offering a free month of service with a new server in >>>> Ashburn. >>>> Choose from 2 high performing configs, both with 100TB of bandwidth. >>>> Higher redundancy.Lower latency.Increased capacity.Completely compliant. >>>> http://p.sf.net/sfu/gigenet >>>> _______________________________________________ >>>> mpgraph-developers mailing list >>>> mpg...@li... >>>> https://lists.sourceforge.net/lists/listinfo/mpgraph-developers >>>> >>>> >>> >> > |
|
From: Bryan T. <br...@sy...> - 2015-01-21 13:13:30
|
What GPU are you using? What operating system? Thanks, Bryan ---- Bryan Thompson Chief Scientist & Founder SYSTAP, LLC 4501 Tower Road Greensboro, NC 27410 br...@sy... 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, Jan 20, 2015 at 7:51 PM, Kapil Agarwal <ka...@ga...> wrote: > Hi > > I installed MapGraph and ran it using > ./Algorithms/PageRank/PageRank -g webbase-1M.mtx > > I get the following errrors > [../../GASengine/enactor_vertex_centric.cuh, 205] EnactorVertexCentric > cudaBindTexture bitmask_tex_ref failed (CUDA error 11: invalid argument) > > Please help me to fix it. > > Best, Kapil > ᐧ > > > ------------------------------------------------------------------------------ > New Year. New Location. New Benefits. New Data Center in Ashburn, VA. > GigeNET is offering a free month of service with a new server in Ashburn. > Choose from 2 high performing configs, both with 100TB of bandwidth. > Higher redundancy.Lower latency.Increased capacity.Completely compliant. > http://p.sf.net/sfu/gigenet > _______________________________________________ > mpgraph-developers mailing list > mpg...@li... > https://lists.sourceforge.net/lists/listinfo/mpgraph-developers > > |
|
From: Bryan T. <br...@sy...> - 2015-01-21 13:01:43
|
Can you try "make clean" and then recompile? Let me know what you have modified (if anything) over the version that is in SVN. Thanks, Bryan ---- Bryan Thompson Chief Scientist & Founder SYSTAP, LLC 4501 Tower Road Greensboro, NC 27410 br...@sy... 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 Wed, Jan 21, 2015 at 7:51 AM, Kapil Agarwal <ka...@ga...> wrote: > ᐧ > I am using Tesla K40c on Ubuntu 14.04. > > Best, Kapil > > On Wed, Jan 21, 2015 at 7:45 AM, Bryan Thompson <br...@sy...> wrote: > >> What GPU are you using? What operating system? >> >> Thanks, >> Bryan >> >> ---- >> Bryan Thompson >> Chief Scientist & Founder >> SYSTAP, LLC >> 4501 Tower Road >> Greensboro, NC 27410 >> br...@sy... >> 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, Jan 20, 2015 at 7:51 PM, Kapil Agarwal <ka...@ga...> wrote: >> >>> Hi >>> >>> I installed MapGraph and ran it using >>> ./Algorithms/PageRank/PageRank -g webbase-1M.mtx >>> >>> I get the following errrors >>> [../../GASengine/enactor_vertex_centric.cuh, 205] EnactorVertexCentric >>> cudaBindTexture bitmask_tex_ref failed (CUDA error 11: invalid argument) >>> >>> Please help me to fix it. >>> >>> Best, Kapil >>> >>> >>> ------------------------------------------------------------------------------ >>> New Year. New Location. New Benefits. New Data Center in Ashburn, VA. >>> GigeNET is offering a free month of service with a new server in Ashburn. >>> Choose from 2 high performing configs, both with 100TB of bandwidth. >>> Higher redundancy.Lower latency.Increased capacity.Completely compliant. >>> http://p.sf.net/sfu/gigenet >>> _______________________________________________ >>> mpgraph-developers mailing list >>> mpg...@li... >>> https://lists.sourceforge.net/lists/listinfo/mpgraph-developers >>> >>> >> > |
|
From: Kapil A. <ka...@ga...> - 2015-01-21 12:52:09
|
ᐧ I am using Tesla K40c on Ubuntu 14.04. Best, Kapil On Wed, Jan 21, 2015 at 7:45 AM, Bryan Thompson <br...@sy...> wrote: > What GPU are you using? What operating system? > > Thanks, > Bryan > > ---- > Bryan Thompson > Chief Scientist & Founder > SYSTAP, LLC > 4501 Tower Road > Greensboro, NC 27410 > br...@sy... > 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, Jan 20, 2015 at 7:51 PM, Kapil Agarwal <ka...@ga...> wrote: > >> Hi >> >> I installed MapGraph and ran it using >> ./Algorithms/PageRank/PageRank -g webbase-1M.mtx >> >> I get the following errrors >> [../../GASengine/enactor_vertex_centric.cuh, 205] EnactorVertexCentric >> cudaBindTexture bitmask_tex_ref failed (CUDA error 11: invalid argument) >> >> Please help me to fix it. >> >> Best, Kapil >> >> >> ------------------------------------------------------------------------------ >> New Year. New Location. New Benefits. New Data Center in Ashburn, VA. >> GigeNET is offering a free month of service with a new server in Ashburn. >> Choose from 2 high performing configs, both with 100TB of bandwidth. >> Higher redundancy.Lower latency.Increased capacity.Completely compliant. >> http://p.sf.net/sfu/gigenet >> _______________________________________________ >> mpgraph-developers mailing list >> mpg...@li... >> https://lists.sourceforge.net/lists/listinfo/mpgraph-developers >> >> > |
|
From: Kapil A. <ka...@ga...> - 2015-01-21 00:51:57
|
Hi I installed MapGraph and ran it using ./Algorithms/PageRank/PageRank -g webbase-1M.mtx I get the following errrors [../../GASengine/enactor_vertex_centric.cuh, 205] EnactorVertexCentric cudaBindTexture bitmask_tex_ref failed (CUDA error 11: invalid argument) Please help me to fix it. Best, Kapil ᐧ |
|
From: Zhisong Fu <fuz...@sy...> - 2014-11-29 20:44:19
|
*Hi, Levs,* *I will be looking at this and let you know. * *Zhisong* On Fri, Nov 28, 2014 at 7:44 PM, Лев Долгов <lev...@gm...> wrote: > *Hi,* > > *When running PageRank to test, I encountered* > [MapGraphV3]$ Algorithms/PageRank/PageRank -graph > smallRegressionGraphs/small.mtx > Running on host: > Detected 1 CUDA Capable device(s) > Running on this device: > Device 0: "GeForce GTX TITAN Black" > CUDA Driver Version / Runtime Version 6.5 / 6.5 > CUDA Capability Major/Minor version number: 3.5 > Total amount of global memory: 6144 MBytes (6442254336 > bytes) > Reading from smallRegressionGraphs/small.mtx: > Parsing MARKET COO format (7 nodes, 11 directed edges)... Done parsing > (0s). > Converting 7 vertices, 11 directed edges (unordered tuples) to CSR > format... Done converting (0s). > GPU 0 column_indices: 11 elements (176 bytes) > GPU 0 row_offsets: 8 elements (32 bytes) > CPU to GPU memory transfer time: 0.046699 ms > [../../GASengine/enactor_vertex_centric.cuh, 208] EnactorVertexCentric > cudaBindTexture bitmask_tex_ref failed (CUDA error 11: invalid argument) > > *and, when trying to correct a seeming syntax error (referencing C API > cudaBindTexture) > <http://developer.download.nvidia.com/compute/cuda/4_1/rel/toolkit/docs/online/group__CUDART__TEXTURE_g3ea5f64f8e1255f2adfa393e0b4a7622.html#g3ea5f64f8e1255f2adfa393e0b4a7622>, > I got (same as above, but different error):* > [../../GASengine/enactor_vertex_centric.cuh, 209] EnactorVertexCentric > cudaBindTexture bitmask_tex_ref failed (CUDA error 18: invalid texture > reference) > *Below are the corrections I tried to make (in * > GASengine/enactor_vertex_centric.cuh > *). What am I doing wrong? (makefile attached, compiling with CUDA Tooklit > 6.5)* // Bind bitmask texture > int bytes = (graph_slice->nodes + 8 - 1) / 8; > cudaChannelFormatDesc bitmask_desc = cudaCreateChannelDesc<char>(); > if(cudaGetLastError()) > printf("ERROR!\n"); > size_t offset = (size_t) 0; > struct textureReference texref = > vertex_centric::contract_atomic::BitmaskTex< > VisitedMask>::ref; > if (retval = > util::B40CPerror( > cudaBindTexture(&offset, > &texref, > graph_slice->d_visited_mask, &bitmask_desc, > (size_t) bytes), > "EnactorVertexCentric cudaBindTexture bitmask_tex_ref > failed", > __FILE__, __LINE__)) > break; > } > while (0); > > > > *Thanks a lot,Levs* > > > > ------------------------------------------------------------------------------ > Download BIRT iHub F-Type - The Free Enterprise-Grade BIRT Server > from Actuate! Instantly Supercharge Your Business Reports and Dashboards > with Interactivity, Sharing, Native Excel Exports, App Integration & more > Get technology previously reserved for billion-dollar corporations, FREE > > http://pubads.g.doubleclick.net/gampad/clk?id=157005751&iu=/4140/ostg.clktrk > _______________________________________________ > mpgraph-developers mailing list > mpg...@li... > https://lists.sourceforge.net/lists/listinfo/mpgraph-developers > > |
|
From: Лев Д. <lev...@gm...> - 2014-11-29 02:44:51
|
*Hi,* *When running PageRank to test, I encountered* [MapGraphV3]$ Algorithms/PageRank/PageRank -graph smallRegressionGraphs/small.mtx Running on host: Detected 1 CUDA Capable device(s) Running on this device: Device 0: "GeForce GTX TITAN Black" CUDA Driver Version / Runtime Version 6.5 / 6.5 CUDA Capability Major/Minor version number: 3.5 Total amount of global memory: 6144 MBytes (6442254336 bytes) Reading from smallRegressionGraphs/small.mtx: Parsing MARKET COO format (7 nodes, 11 directed edges)... Done parsing (0s). Converting 7 vertices, 11 directed edges (unordered tuples) to CSR format... Done converting (0s). GPU 0 column_indices: 11 elements (176 bytes) GPU 0 row_offsets: 8 elements (32 bytes) CPU to GPU memory transfer time: 0.046699 ms [../../GASengine/enactor_vertex_centric.cuh, 208] EnactorVertexCentric cudaBindTexture bitmask_tex_ref failed (CUDA error 11: invalid argument) *and, when trying to correct a seeming syntax error (referencing C API cudaBindTexture) <http://developer.download.nvidia.com/compute/cuda/4_1/rel/toolkit/docs/online/group__CUDART__TEXTURE_g3ea5f64f8e1255f2adfa393e0b4a7622.html#g3ea5f64f8e1255f2adfa393e0b4a7622>, I got (same as above, but different error):* [../../GASengine/enactor_vertex_centric.cuh, 209] EnactorVertexCentric cudaBindTexture bitmask_tex_ref failed (CUDA error 18: invalid texture reference) *Below are the corrections I tried to make (in * GASengine/enactor_vertex_centric.cuh *). What am I doing wrong? (makefile attached, compiling with CUDA Tooklit 6.5)* // Bind bitmask texture int bytes = (graph_slice->nodes + 8 - 1) / 8; cudaChannelFormatDesc bitmask_desc = cudaCreateChannelDesc<char>(); if(cudaGetLastError()) printf("ERROR!\n"); size_t offset = (size_t) 0; struct textureReference texref = vertex_centric::contract_atomic::BitmaskTex< VisitedMask>::ref; if (retval = util::B40CPerror( cudaBindTexture(&offset, &texref, graph_slice->d_visited_mask, &bitmask_desc, (size_t) bytes), "EnactorVertexCentric cudaBindTexture bitmask_tex_ref failed", __FILE__, __LINE__)) break; } while (0); *Thanks a lot,Levs* |
|
From: Лев Д. <lev...@gm...> - 2014-11-29 02:29:24
|
*Hi,* *When running PageRank to test, I encountered* [MapGraphV3]$ Algorithms/PageRank/PageRank -graph smallRegressionGraphs/small.mtx Running on host: Detected 1 CUDA Capable device(s) Running on this device: Device 0: "GeForce GTX TITAN Black" CUDA Driver Version / Runtime Version 6.5 / 6.5 CUDA Capability Major/Minor version number: 3.5 Total amount of global memory: 6144 MBytes (6442254336 bytes) Reading from smallRegressionGraphs/small.mtx: Parsing MARKET COO format (7 nodes, 11 directed edges)... Done parsing (0s). Converting 7 vertices, 11 directed edges (unordered tuples) to CSR format... Done converting (0s). GPU 0 column_indices: 11 elements (176 bytes) GPU 0 row_offsets: 8 elements (32 bytes) CPU to GPU memory transfer time: 0.046699 ms [../../GASengine/enactor_vertex_centric.cuh, 208] EnactorVertexCentric cudaBindTexture bitmask_tex_ref failed (CUDA error 11: invalid argument) *and, when trying to correct a seeming syntax error (referencing C API cudaBindTexture) <http://developer.download.nvidia.com/compute/cuda/4_1/rel/toolkit/docs/online/group__CUDART__TEXTURE_g3ea5f64f8e1255f2adfa393e0b4a7622.html#g3ea5f64f8e1255f2adfa393e0b4a7622>, I got (same as above, but different error):* [../../GASengine/enactor_vertex_centric.cuh, 209] EnactorVertexCentric cudaBindTexture bitmask_tex_ref failed (CUDA error 18: invalid texture reference) *Below are the corrections I tried to make (in * GASengine/enactor_vertex_centric.cuh *). What am I doing wrong? (makefile attached, compiling with CUDA Tooklit 6.5)* // Bind bitmask texture int bytes = (graph_slice->nodes + 8 - 1) / 8; cudaChannelFormatDesc bitmask_desc = cudaCreateChannelDesc<char>(); if(cudaGetLastError()) printf("ERROR!\n"); size_t offset = (size_t) 0; struct textureReference texref = vertex_centric::contract_atomic::BitmaskTex< VisitedMask>::ref; if (retval = util::B40CPerror( cudaBindTexture(&offset, &texref, graph_slice->d_visited_mask, &bitmask_desc, (size_t) bytes), "EnactorVertexCentric cudaBindTexture bitmask_tex_ref failed", __FILE__, __LINE__)) break; } while (0); *Thanks a lot,Levs* |
|
From: Bryan T. <br...@sy...> - 2014-11-03 12:31:57
|
I believe that this question was answered through email. Please follow up if you are still having problems. Thanks, Bryan ---- Bryan Thompson Chief Scientist & Founder SYSTAP, LLC 4501 Tower Road Greensboro, NC 27410 br...@sy... 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 Fri, Oct 17, 2014 at 5:38 AM, Abhilasha R <abh...@gm...> wrote: > Hey, > > We originally tried to run mapgraph on CUDA version 5.0, and were unable > to install as we got "Calling a __host__ function from a __device__ > function is not allowed." > > We then ran it on Cuda 6.0 and got an error that gethostname function in > the.cu files could not be identified. On commenting out of these lines in > all the .cu files within /Algorithms, the compilation was successful. > > We would like to know how to test if mapgraph was installed successfully > and how to run a few algorithms. Any help would be greatly appreciated. > > Thanks and regards, > Abhilasha > > > > ------------------------------------------------------------------------------ > > _______________________________________________ > mpgraph-developers mailing list > mpg...@li... > https://lists.sourceforge.net/lists/listinfo/mpgraph-developers > > |
|
From: Bryan T. <br...@sy...> - 2014-11-03 12:31:29
|
No. Allocation needs to be done outside of the device kernels. Use Initialize(). Bryan ---- Bryan Thompson Chief Scientist & Founder SYSTAP, LLC 4501 Tower Road Greensboro, NC 27410 br...@sy... 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 Sun, Oct 19, 2014 at 2:53 AM, Лев Долгов <lev...@gm...> wrote: > Hi! > > Is there a way to allocate objects dynamically in gather_sum? Say, I want > to choose a max of N candidates, and N is determined at runtime? > > Thank you! > Leo > > > ------------------------------------------------------------------------------ > > _______________________________________________ > mpgraph-developers mailing list > mpg...@li... > https://lists.sourceforge.net/lists/listinfo/mpgraph-developers > > |
|
From: Лев Д. <lev...@gm...> - 2014-10-19 06:54:35
|
Hi! Is there a way to allocate objects dynamically in gather_sum? Say, I want to choose a max of N candidates, and N is determined at runtime? Thank you! Leo |
|
From: Abhilasha R <abh...@gm...> - 2014-10-17 09:38:32
|
Hey, We originally tried to run mapgraph on CUDA version 5.0, and were unable to install as we got "Calling a __host__ function from a __device__ function is not allowed." We then ran it on Cuda 6.0 and got an error that gethostname function in the.cu files could not be identified. On commenting out of these lines in all the .cu files within /Algorithms, the compilation was successful. We would like to know how to test if mapgraph was installed successfully and how to run a few algorithms. Any help would be greatly appreciated. Thanks and regards, Abhilasha |
|
From: Bryan T. <br...@sy...> - 2013-08-04 21:44:21
|
See https://sourceforge.net/p/mpgraph/tickets/8/. The OpenMP timings are at 3Ghz (I fixed the problem where the CPU was not clocked at its nominal frequency). The K20 and GTX570 results are also there. These data were collected using the ./regressions/Makefile. Bryan |
|
From: Bryan T. <br...@sy...> - 2013-08-02 11:25:30
|
All, The SVN repository has been re-rooted. We are now working in the trunk The SVN repository URL is unchanged: svn://svn.code.sf.net/p/mpgraph/code The URL of the checkout for the trunk is: svn://svn.code.sf.net/p/mpgraph/code/trunk You MUST obtain a clean checkout of the code from SVN. If you have existing local changes: You need to apply those changes to a clean checkout of the project. To do this, simple checkout the project into another directory. Then copy each modified source file from your old working directory (but NOT the entire directory structure since the .svn files MUST NOT be copied) into the newly checked out version of the project. Please contact me if you have ANY questions about this procedure. The SVN repository is unfrozen. Development may now continue. Thanks, Bryan From: Bryan Thompson <tho...@us...<mailto:tho...@us...>> Reply-To: "[mpgraph:tickets]" <5...@ti...<mailto:5...@ti...>> Date: Friday, August 2, 2013 7:14 AM To: "[mpgraph:tickets]" <5...@ti...<mailto:5...@ti...>> Subject: [mpgraph:tickets] #5 Re-root SVN I have completed the procedure as described at [1]. It appears that this procedure has converted all of the existing SVN revisions by +1. This is because the initial revision is now the create of the trunk, branches, and tags directories. Thus the current revision in the repository has changed from r30 to r31. I do not believe that this will influence future revisions in any manner. If you have changes against the checked out code, then you probably need to apply those changes to a clean checkout of the project. To do this, simple checkout the project into another directory. Then copy each modified source file from your old working directory (but NOT the entire directory structure since the .svn files MUST NOT be copied) into the newly checked out version of the project. Please contact me if you have ANY questions about this procedure. [1] https://sourceforge.net/p/forge/site-support/4915/ ________________________________ [tickets:#5]<http://sourceforge.net/p/mpgraph/tickets/5/> Re-root SVN Status: open Created: Thu Aug 01, 2013 10:58 AM UTC by Bryan Thompson Last Updated: Thu Aug 01, 2013 12:29 PM UTC Owner: Bryan Thompson The initial import of the code should have been into svn://svn.code.sf.net/p/mpgraph/code/trunk rather than svn://svn.code.sf.net/p/mpgraph/code The current SVN setup does not allow us to have trunk, branches, and tags. SVN needs to be re-rooted. This will probably require an SVN export, an administrative request to SF to clear the SVN repo, and an SVN import into trunk. At that point we could tag the initial release. ________________________________ Sent from sourceforge.net because you indicated interest in https://sourceforge.net/p/mpgraph/tickets/5/ To unsubscribe from further messages, please visit https://sourceforge.net/auth/subscriptions/ |
|
From: Bryan T. <br...@sy...> - 2013-08-02 10:50:04
|
All, I am going to re-root the SVN repository this morning. Please freeze commits until I have completed this procedure. I have already made a backup of the SVN repo that is current as of r30 (the current committed revision). Thanks, Bryan ----- I'm not sure what steps you took at the beginning, so I can't really advise what happened at the beginning, but here is my suggestion on how to recover from this. There are certainly a number of ways to tackle this, such as just doing a series of svn mv and svn mkdir operations to move stuff to where you want it, but personally, this is what I would do (I recommend reading through the instructions and making sure there are no surprises before starting): 1. make a backup<http://sourceforge.net/p/forge/documentation/rsync%20Backups/>, don't expect any issues, but, just in case. 2. Login to the shell service: ssh -t thompsonbry,mp...@sh...<mailto:mp...@sh...> create 3. from the shell service get an svnadmin dump: svnadmin dump /home/svn/p/mpgraph/code > /home/project-web/mpgraph/mpgraph-svndump shutdown ; exit 4. via Admin -> Tools in project web interface, delete, then recreate the svn repository 5. follow the instructions on the code page (which displays when it's an empty repository) to add the branch/tags/trunk directory structure. 6. log back into the shell, and once there, run an svnadmin load with "trunk" as the parent dir. svnadmin load --parent-dir trunk /home/svn/p/mpgraph/code/ < /home/project-web/mpgraph/mpgraph-svndump 7. Once that's done, Shutdown the shell (since you won't need it anymore), then go to Admin -> Tools again and select "Refresh repository" so the web interface scans in the new changes. Note: I shutdown the shell in between deleting/recreating the repo because the repository is mounted at shell creation time, so we need to do that for it to recognize the new repo Note 2: this will result in all the revisions being off-by-one (since rev 1 will now be adding the branch/tags/trunk structure) Note 3: I tested this concept earlier before writing these steps and you can see the repo on my test project here:https://sourceforge.net/p/strawhat/mpgraph/ Let me know how this goes! Regards, Chris Tsai, SourceForge.net Support ________________________________ [site-support:#4915] Re-root SVN (import did not respect trunk/braches/tags) Status: pending Created: Thu Aug 01, 2013 12:28 PM UTC by Bryan Thompson Last Updated: Thu Aug 01, 2013 12:28 PM UTC Owner: Chris Tsai Hello, the initial import of the mpgraph project did not respect the use of the trunk, branches, and tags directory structure. The files are at the top-level. See https://sourceforge.net/p/mpgraph/code/30/tree/ I would like to fix this so we can begin tagging releases, managing branches, etc. Please advise on the best approach. E.g., should I delete the SVN repository and re-import a snapshot? Or is there a better way of accomplishing this goal? There are very few revisions in SVN to date for this project, but we are at our first release point. It would be nice to retain the commit history up to now, but if it is not possible to do this then we can destroy the exiting SVN and reimport. Thanks, Bryan ________________________________ Sent from sourceforge.net because you indicated interest in https://sourceforge.net/p/forge/site-support/4915/ To unsubscribe from further messages, please visit https://sourceforge.net/auth/subscriptions/ |
