Jun Liu

@seungmanhan we have a HIP tidy issue: ``` [2024-05-23T13:46:38.046Z] /home/jenkins/workspace/MLLIBS_MIOpen_impl_getitem_bwd/src/include/miopen/miopen_internal.h:119:1: warning: function 'miopenConvolutionCKBackwardWeightsGetWorkSpaceSize' has a definition with different parameter names [readability-inconsistent-declaration-parameter-name] [2024-05-23T13:46:38.046Z] 119 | miopenConvolutionCKBackwardWeightsGetWorkSpaceSize(const miopenAlphaBetaCase_t ab_case, [2024-05-23T13:46:38.046Z] | ^...

@JehandadKhan @randyspauldingamd @CAHEK7 last ping before last round of CI and then plan to merge

@seungmanhan we still get a few tests persistently failing at `Fp32 Hip All gfx90a` stage, could you help to check them out?

@seungmanhan could you help to resolve the conflict? We are having some CI issues and let me know if you need help. Thanks!

@seungmanhan @apwojcik The Windows build is failing while the Linux side is passing: > lld-link: error: undefined symbol: enum miopenStatus_t __cdecl miopen::GetitemBackward(struct miopen::Handle &, void *, unsigned __int64, struct miopen::TensorDescriptor...

@Zakhrov it might be a Boost version issue, are you following the build guide closely to build MIOpen dependencies? ``` export LC_ALL=C.UTF-8 && \ export LANG=C.UTF-8 && \ apt-get update...

Just curious, should this topic belong or eventually belong to Wiki or Contribution Guide page? It looks like a guideline which we should follow. **The intention for the issue is...

@jamaliki are you using Pytorch wheel? If so the KDB needs to be embedded and cannot be copied over. If you are building MIOpen from code, the KDB file is...

Ping @seungmanhan and also from Slack

@zjchen77 thanks! Our CK kernels https://github.com/ROCmSoftwarePlatform/Composable_kernel already supports and in the process of refining them, Indeed exposing these kernels via MIOpen might be useful in many cases.