- Dec 21, 2020
-
-
Akash Kothari authored
-
Akash Kothari authored
-
- Dec 20, 2020
-
-
Akash Kothari authored
-
- Dec 08, 2020
-
-
Akash Kothari authored
-
- Nov 21, 2020
-
-
Hashim Sharif authored
-
- Feb 16, 2020
-
-
Yifan Zhao authored
-
- Dec 26, 2019
-
-
Hashim Sharif authored
-
- Nov 04, 2019
-
-
Yifan Zhao authored
-
- Oct 20, 2019
-
-
Yifan Zhao authored
-
- Jul 14, 2019
-
-
Hashim Sharif authored
-
- Jul 12, 2019
-
-
Hashim Sharif authored
-
- Dec 14, 2018
-
-
Hashim Sharif authored
-
- Dec 13, 2018
-
-
Hashim Sharif authored
-
- Dec 06, 2018
- Oct 29, 2018
-
-
Prakalp Srivastava authored
-
- Oct 22, 2018
-
-
Prakalp Srivastava authored
-
- Aug 09, 2017
-
-
kotsifa2 authored
(benchmark not updated)
-
- Aug 07, 2017
-
-
kotsifa2 authored
(CPU/GPU/SPIR) based on policy. Support for scheduling per iteration included. Further testing required.
-
- Aug 03, 2017
-
-
Prakalp Srivastava authored
Fixed spmv. LLVM 4.0 allows argument type promotion which was breaking the type check between __visc__node call and kernel function
-
- Aug 02, 2017
-
-
Prakalp Srivastava authored
Fixed visc_sh sgemm benchmark. The bug was that mtc bind when fixing function type was not handled correctly
-
- Jul 26, 2017
-
-
Prakalp Srivastava authored
-
- Jun 14, 2017
-
-
Prakalp Srivastava authored
-
- Jun 13, 2017
-
-
kotsifa2 authored
-
- Jun 07, 2017
-
-
Prakalp Srivastava authored
Added parsing support for visc in,out,inout attributes. Commit ef703216 in previous llvm 3.4 based repo
-
- Jun 05, 2017
-
-
Prakalp Srivastava authored
-
- Jun 02, 2017
-
-
kotsifa2 authored
-
Prakalp Srivastava authored
Added the function to clone any function into a different typesHandiling two cases: (1) The caller is attempting to add data size argument after every pointer(2) The caller wants to add some extra arguments at the end
-
- Jun 01, 2017
-
-
Prakalp Srivastava authored
-
- May 25, 2017
-
-
Prakalp Srivastava authored
-
- May 20, 2017
-
-
kotsifa2 authored
-
- Apr 06, 2017
-
-
kotsifa2 authored
The SymVT stuff has not been moved yet, I removed it for now. The test/VISC tests also have not been moved yet, so they may not work.
-
- Dec 01, 2016
-
-
Prakalp Srivastava authored
-
- Sep 07, 2016
-
-
Prakalp Srivastava authored
Added argument to __visc__edge to allow user to specify 1-to-1 or all-to-all edge. all-to-all = true, one-to-one = false
-
- May 01, 2016
-
-
Prakalp Srivastava authored
-
- Apr 26, 2016
-
-
Prakalp Srivastava authored
Changed GenVISC pass to automatically infer return type of internal nodes from bind output intrinsics
-
- Apr 23, 2016
-
-
Prakalp Srivastava authored
-
- Mar 20, 2016
-
-
Prakalp Srivastava authored
-
- Mar 15, 2016
-
-
Prakalp Srivastava authored
So no need to generate backend specific intrinsics for PTX backend. Now only using nvvm specifc intrinsic for rsqrt
-
Prakalp Srivastava authored
get_global_size (2) Code gen for nvvm intrinsics in GenVISC pass (3) Commented out code added to GenVISC and DFG2LLVM_X86 to allow wait instruction in a different BB than launch (with phi nodes as uses of launch)
-