Hi Dave, I added "-arch=sm_21" to CUDACodeGenerator.java:1067. Observe the small change when "f" is appended to the String[].
This is certainly not the correct fix. Looks like the ideal thing would be generating PTX files instead. Anyway, I'm able to continue my tests on my new GTX 460. :) More details at http://developer.download.nvidia.com/compute/cuda/3_2_prod/toolkit/docs/Fermi_Compatibility_Guide.pdf String[] nvccCmd = { "nvcc", "--cubin", "-arch=sm_21", "-Xptxas", "-v", "-I" + CXXCommandBuilder.X10_DIST + "/include", null }; for (String f : compilationUnits) { if (f.endsWith(".cu")) { nvccCmd[nvccCmd.length-1] = f; Cheers Richard Gomes M: +44(77)9955-6813 http://tinyurl.com/frgomes twitter: frgomes JQuantLib is a library for Quantitative Finance written in Java. http://www.jquantlib.org/ twitter: jquantlib On 12/12/10 20:36, Richard Gomes wrote: > Hi Dave, > > From the programming guide: > > > > 3.1.2 Binary Compatibility > > Binary code is architecture-specific. A cubin object is generated using > the compiler option –code that specifies the targeted architecture: For > example, compiling with –code=sm_13 produces binary code for devices of > compute capability 1.3. Binary compatibility is guaranteed from one > minor revision to the next one, but not from one minor revision to the > previous one or across major revisions. In other words, a cubin object > generated for compute capability X.y is only guaranteed to execute on > devices of compute capability X.z where z≥y. > > > > This makes me believe that code intended to GTX 460 should be compiled > with compute capability 2.1 set, i.e. employing -code=sm_21. > > Does it make sense? > How I could force this option when I call x10c++ ? > > Thanks > > Richard Gomes > M: +44(77)9955-6813 > http://tinyurl.com/frgomes > twitter: frgomes > > JQuantLib is a library for Quantitative Finance written in Java. > http://www.jquantlib.org/ > twitter: jquantlib > > On 11/12/10 00:33, Dave Cunningham wrote: >> This is what I got with our tesla 2070 too. I assumed it was something to >> do with the system but if you get it too then it's more likely to be >> something we're doing wrong. Maybe we're invoking nvcc improperly. >> >> On Fri, Dec 10, 2010 at 5:36 PM, Richard Gomes<rgomes1...@yahoo.co.uk>wrote: >> >>> Hi guys, >>> >>> Today I started tests with a new toy: GeForce GTX 460 :) >>> >>> In a nutshell, I removed the old GeForce 8300 and installed the new one. >>> >>> Now all X10 programs are failing, except CUDATopology which works fine. >>> All programs are behaving like this: >>> >>> $ runx10 CUDAKernelTest >>> X10_NPLACES not set. Assuming 1 place, running locally >>> CUDA_ERROR_INVALID_SOURCE (At common/x10rt_cuda.cc:383) >>> Aborted >>> >>> I've recompiled from trunk, as usual (and should always do before >>> reporting issues!). >>> >>> All Nvidia apps, part of the dev kit are working as expected. >>> >>> >>> This is what deviceQuery says: >>> >>> >>> $ ./deviceQuery >>> ./deviceQuery Starting... >>> >>> CUDA Device Query (Runtime API) version (CUDART static linking) >>> >>> There is 1 device supporting CUDA >>> >>> Device 0: "GeForce GTX 460" >>> CUDA Driver Version: 3.20 >>> CUDA Runtime Version: 3.20 >>> CUDA Capability Major/Minor version number: 2.1 >>> Total amount of global memory: 2146631680 bytes >>> Multiprocessors x Cores/MP = Cores: 7 (MP) x 48 (Cores/MP) >>> = 336 (Cores) >>> Total amount of constant memory: 65536 bytes >>> Total amount of shared memory per block: 49152 bytes >>> Total number of registers available per block: 32768 >>> Warp size: 32 >>> Maximum number of threads per block: 1024 >>> Maximum sizes of each dimension of a block: 1024 x 1024 x 64 >>> Maximum sizes of each dimension of a grid: 65535 x 65535 x 1 >>> Maximum memory pitch: 2147483647 bytes >>> Texture alignment: 512 bytes >>> Clock rate: 1.40 GHz >>> Concurrent copy and execution: Yes >>> Run time limit on kernels: Yes >>> Integrated: No >>> Support host page-locked memory mapping: Yes >>> Compute mode: Default (multiple host >>> threads can use this device simultaneously) >>> Concurrent kernel execution: Yes >>> Device has ECC support enabled: No >>> Device is using TCC driver mode: No >>> >>> deviceQuery, CUDA Driver = CUDART, CUDA Driver Version = 3.20, CUDA >>> Runtime Version = 3.20, NumDevs = 1, Device = GeForce GTX 460 >>> >>> >>> PASSED >>> >>> Press<Enter> to Quit... >>> ----------------------------------------------------------- >>> >>> >>> Any idea? >>> >>> Thanks a lot :) >>> >>> -- >>> Richard Gomes >>> M: +44(77)9955-6813 >>> http://tinyurl.com/frgomes >>> twitter: frgomes >>> >>> JQuantLib is a library for Quantitative Finance written in Java. >>> http://www.jquantlib.org/ >>> twitter: jquantlib >>> >>> >>> ------------------------------------------------------------------------------ >>> Oracle to DB2 Conversion Guide: Learn learn about native support for >>> PL/SQL, >>> new data types, scalar functions, improved concurrency, built-in packages, >>> OCI, SQL*Plus, data movement tools, best practices and more. >>> http://p.sf.net/sfu/oracle-sfdev2dev >>> _______________________________________________ >>> X10-users mailing list >>> X10-users@lists.sourceforge.net >>> https://lists.sourceforge.net/lists/listinfo/x10-users >>> >> ------------------------------------------------------------------------------ >> Oracle to DB2 Conversion Guide: Learn learn about native support for PL/SQL, >> new data types, scalar functions, improved concurrency, built-in packages, >> OCI, SQL*Plus, data movement tools, best practices and more. >> http://p.sf.net/sfu/oracle-sfdev2dev >> _______________________________________________ >> X10-users mailing list >> X10-users@lists.sourceforge.net >> https://lists.sourceforge.net/lists/listinfo/x10-users >> > > ------------------------------------------------------------------------------ > Oracle to DB2 Conversion Guide: Learn learn about native support for PL/SQL, > new data types, scalar functions, improved concurrency, built-in packages, > OCI, SQL*Plus, data movement tools, best practices and more. > http://p.sf.net/sfu/oracle-sfdev2dev > _______________________________________________ > X10-users mailing list > X10-users@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/x10-users ------------------------------------------------------------------------------ Oracle to DB2 Conversion Guide: Learn learn about native support for PL/SQL, new data types, scalar functions, improved concurrency, built-in packages, OCI, SQL*Plus, data movement tools, best practices and more. http://p.sf.net/sfu/oracle-sfdev2dev _______________________________________________ X10-users mailing list X10-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/x10-users
