ROCm/composable_kernel
1
0
Fork 0
mirror of https://github.com/ROCm/composable_kernel.git synced 2026-05-15 10:37:44 +00:00
Code Issues Packages Projects Releases Wiki Activity

Merge commit 'd7497d26948ca90d0224920472712e0f657fb744' into develop

Browse Source
This commit is contained in:
assistant-librarian[bot]
2026-01-07 08:16:44 +00:00
parent 9ec8eac079
commit bb614ee8b2
1 changed files with 15 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files

19
include/ck_tile/core/arch/amd_buffer_addressing_builtins.hpp
View File

@@ -2376,12 +2376,23 @@ amd_buffer_load_invalid_element_return_zero(const T* p_src_wave,
return amd_buffer_load_impl<T, N, coherence>(
src_wave_buffer_resource, src_addr_shift + src_thread_addr_offset, 0);
#else
thread_buffer<T, N> tmp =
amd_buffer_load_impl<T, N, coherence>(src_wave_buffer_resource, src_thread_addr_offset, 0);
if constexpr(oob_conditional_check)
return src_thread_element_valid ? tmp : thread_buffer<T, N>{numeric<T>::zero()};
{
if(src_thread_element_valid)
{
return amd_buffer_load_impl<T, N, coherence>(
src_wave_buffer_resource, src_thread_addr_offset, 0);
}
else
{
return thread_buffer<T, N>{numeric<T>::zero()};
}
}
else
return tmp;
{
return amd_buffer_load_impl<T, N, coherence>(
src_wave_buffer_resource, src_thread_addr_offset, 0);
}
#endif
}