Skip to content

Conversation

@coldav
Copy link

@coldav coldav commented Oct 3, 2024

Native cpu can make calls to mux builtins such as shuffle which are ABI compliant but are not what is expected by ock passes. This fixes them up by remove the vector versions from libnativecpu.cpp and using a pass to convert from parameters which relate to the ABI to calling the mux functions with the set interface unaffected by the ABI.

This currently only handle a small number of cases for shuffle such as when a vector i2 is replaced with double or byval is used. It will be expanded over time as needed.

…o mux

Native cpu can make calls to mux builtins such as shuffle which are
ABI compliant but are not what is expected by ock passes. This fixes
them up by remove the vector versions from libnativecpu.cpp and
using a pass to convert from parameters which relate to the ABI to
calling the mux functions with the set interface unaffected by the ABI.

This currently only handle a small number of cases for shuffle such
as when a vector i2 is replaced with double or byval is used. It will
be expanded over time as needed.
@coldav coldav requested a review from a team as a code owner October 3, 2024 15:43
@coldav
Copy link
Author

coldav commented Oct 4, 2024

@intel/llvm-gatekeepers please merge.

@sommerlukas sommerlukas merged commit ddd23ad into intel:sycl Oct 4, 2024
13 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants