GitLab

* porting fmoe_sorting from moe_sorting

* pass default example test

* remod

* comp v4 setup

* add a file

* Finished the coding of the feature, Compiler not in the way we supposed to have

* Update some of the code to better format

* get tback the restrict variable name, need to switch out to solve the transpose issue

* Solve the compiler issue on SHMEM conflict

* roll back to compute pipeline

* Add the changes from include/ck_tile

* Address the comments

* Merge from internal (#1857)

* enable batched_gemm_softmax_gemm_perm_wmma for gfx12

* disable instances with blocksize=256 in attention examples

* debuggging

* debug

* fixed lds_enabled

* debugging

* Fix and add limit to skiplds feature

* Enable skipLds feature and fix compilation bugs

* add ck_tile definitions for gfx12

* fix clang format and test/wmma_op

* updage instances cmake for gfx12

* disable the test_wmma_op on gfx12

* fix the builds for gfx950

* add gfx12 and gfx950 to default target list

* ...