[TVM Discuss] [Development/RFC] [RFC][VTA]A HLS C VTA bug

2020-05-22 Thread Liangfu Chen via TVM Discuss
Thanks for locating the bug @c! Is there any way to modify this in HLS instead of modifying the generated Verilog, so that more people could avoid the pitfall? --- [Visit Topic](https://discuss.tvm.ai/t/rfc-vta-a-hls-c-vta-bug/6743/2) to respond. You are receiving this because you en

[TVM Discuss] [Development/RFC] [RFC][VTA]A HLS C VTA bug

2020-05-22 Thread Hanting Huang via TVM Discuss
What about modifying finsh instruction to write a bool into a piece of shared memory? By this way, cpu can clear it after reading. @liangfu Clear On Read in register is the definately efficient way. But if hls programming cannot easily achieve this aim. Clear On Read in shared memory might be

[TVM Discuss] [Development] How to add assertion checks in Tensor compute definition?

2020-05-22 Thread Mahesh Ambule via TVM Discuss
While adding an operator in Topi, most of the time we need to put certain checks on dynamic input data values for which values are known at runtime only. Is there a way in the form of tensor expression, to assert values in compute definition? For ex. An operator accepts indices as dynamic in

[TVM Discuss] [Development] How to add assertion checks in Tensor compute definition?

2020-05-22 Thread Junru Shao via TVM Discuss
if you are using hybrid script, I am sure that there is an assert statement that you may be interested. CC: @were --- [Visit Topic](https://discuss.tvm.ai/t/how-to-add-assertion-checks-in-tensor-compute-definition/6773/2) to respond. You are receiving this because you enabled mailing lis

[TVM Discuss] [Development/RFC] [BYOC] Use pattern language to create composite functions

2020-05-22 Thread Cody H. Yu via TVM Discuss
New `MergeComposite` pass PR filed. https://github.com/apache/incubator-tvm/pull/5656 --- [Visit Topic](https://discuss.tvm.ai/t/byoc-use-pattern-language-to-create-composite-functions/6727/29) to respond. You are receiving this because you enabled mailing list mode. To unsubscribe from

[TVM Discuss] [Development/RFC] Add bfloat16 data type

2020-05-22 Thread Menooker via TVM Discuss
Hi all, We have completed a workable draft of bfloat16 (bf16) in TVM and the bf16 related codegen in LLVM. We add bfloat16 as a new type named "bf16" in the frontend. Completed LLVM backend for generating bf16. * Use int16 as the storage type in LLVM * Add legalization to enable computation

[TVM Discuss] [Development/RFC] [RFC] Add bfloat16 data type

2020-05-22 Thread Menooker via TVM Discuss
@tqchen As you have proposed to eliminate "bfloat16" dtype after the legalization pass, I have two concerns. * The legalization pass will be more compilicated. It has to check every TIR node to replace the bf16 dtype * As the casting between bf16 and fp32 will be lowered as function calls as