Skip to content
Snippets Groups Projects

Repository graph

You can move around the graph by using the arrow keys.
Select Git revision
  • hpvm-release-epochs
  • hpvm-release-epochs0
  • main default
  • v2.0
  • v2.0rc
  • v1.0
  • v0.5
7 results
Created with Raphaël 2.2.020Mar1918171615131211987543227Feb1615983Dec19Nov1897427Oct16Sep1211109876542131Aug231815542130Jul26212016149230Jun22201814131153229May201312432129Apr10927Mar12987642127Feb26231413121110964130Jan282717Dec165430Nov2927262524211918171614119654331Oct30191612711Sep413Aug12111065432131Jul3029262315108727Jun322May24Apr17143Oct20Sep1Aug4Jul3230Jun28262513Added support for umin and umax visc intrinsicsAdding 2 files to histo. host and ptx to run visc generated host code with opencl driver generated ptxAdded a single kernel file for histogram opencl_nvidia kernels. Improved timing calculations for viscAdded single kernel file to opencl_nvidia version of histogramvisc version similar to nvidia versionHistogram improvedMerge branch 'master' of bitbucket.org:psrivas2/viscRemove llvm intrinsic declarationsFixed bug in histo. Was using get_group_id instead of visc intrinsicMerge branch 'master' of bitbucket.org:psrivas2/viscPorted opencl_base version to visc, since opencl_base kernel was faster than opencl_nvidiaMerge branch 'master' of bitbucket.org:psrivas2/viscSupport fot atomics in DFG2LLVM_SPIR (to be tested)Merge branch 'master' of bitbucket.org:psrivas2/viscAdded code to output the compiled ptx binaryMerge branch 'master' of bitbucket.org:psrivas2/viscSupport for llvm.sqrt intrinsic in DFG2LLVM_SPIR backendAdded optimization to promote certain arguements to constant memoryFixed DFG2LLVM_SPIR bugFixed DFG2LLVM_SPIR bugChanged SPIR backend to generate bitcode again. Added code to remove unsupported function attributes. Backend is still not working thoughFixed the __visc__node return type in two sgemm versions(visc_vec_opt and visc_opt)sgemm visc_sh and opencl_nvidia instrumentedDESCRIPTION files added to define input format for histo benchmark. Required by python scriptparboil python script modified to have result for 4 new benchmarks(1) Instrumented cutcp opencl_nvidia and visc version for timingsInstrumented tpacf (opencl and visc versions for timing)Instrumented bfs for timingInstrumented the histogram opencl_nvidia and visc versions with timersRemoved unnecessary printfs from histogramFixed histogramReverted back the way command line arguments are read in tpacf (visc,Histogram working correctly on visc. This commit also adds some debugging mechanisms to the opencl_nvidia and visc versionFixed tpacf opencl_nvidia and comparison toolDebugging histo. Added print statements. Would be useful laterCommited changes to sgemm, bfs, tpacf, cutcp which would be used for experiments. Required to have a checkpoint on the collected numbersImplemented translation of floor intrinsic in PTX backendMerged DFG2LLVM_NVPTXFixed a bug introduced by commit bfe38be in ptx backend. declarations were not being copied to kernel module. Fixed thoseGenerate .ll file from SPIR backend not .bc
Loading