Greg Lueck
Greg Lueck
Add specifications for two proposed extensions: * "sycl_ext_oneapi_device_if": Allows device code to conditionally use "optional kernel features" based on the device's aspects. * "sycl_ext_intel_device_architecture": Allows device code to conditionally use...
Add a proposed extension specification and design for `if_device`, an alternative to `__SYCL_DEVICE_ONLY__`, which is implementable in a 1-pass compiler. This commit also renames the proposed extension "sycl_ext_oneapi_device_if" to "sycl_ext_oneapi_if_device_has"...
Revise the design for handling optional kernel features for AOT compilations. There is no need to create a separate tool "aspect-filter" because the filtering can be done in the "sycl-post-link"...
Update the proposed specification for the kernel compiler to allow kernels to be written in the SYCL language, using the new proposed "free function kernel" format. This PR also adds...
Update the sycl_ext_oneapi_dot_accumulate extension spec to: * Use the latest specification template. * Document the "packed" APIs. These were previously shown in the "sample header" section, but there was no...
Add a proposed extension specification for `work_group_memory`, a lighter weight API to allocate device local memory for an nd-range kernel. Also related, add a list of restrictions that, when followed,...
Add the "phase 2" design for `if_device_has` and `if_architecture_is` extensions. Whereas the phase 1 design worked only in some limited cases for AOT compilation, the phase 2 design fully supports...
Fix several problems with the example in this extension: * The `use_root_sync` property is a kernel property, not a kernel launch property. Therefore, it should decorate the kernel definition, not...