llvm-project.git/libc/src/string/memory_utils/arm/inline_memset.h, branch main Unnamed repository; edit this file 'description' to name the repository. [libc] Improve Cortex `memset` and `memcpy` functions (#149044) 2025-07-17T08:15:43+00:00 Guillaume Chatelet gchatelet@google.com 2025-07-17T08:15:43+00:00 e0cce5cdcb8a7829389d910a9204447646e69407 The code for `memcpy` is the same as in #148204 but it fixes the build bot error by using `static_assert(cpp::always_false<decltype(access)>)` instead of `static_assert(false)` (older compilers fails on `static_assert(false)` in `constexpr` `else` bodies). The code for `memset` is new and vastly improves performance over the current byte per byte implementation. Both `memset` and `memcpy` implementations use prefetching for sizes >= 64. This lowers a bit the performance for sizes between 64 and 256 but improves throughput for greater sizes. 
The code for `memcpy` is the same as in #148204 but it fixes the build
bot error by using `static_assert(cpp::always_false<decltype(access)>)`
instead of `static_assert(false)` (older compilers fails on
`static_assert(false)` in `constexpr` `else` bodies).

The code for `memset` is new and vastly improves performance over the
current byte per byte implementation.

Both `memset` and `memcpy` implementations use prefetching for sizes >=
64. This lowers a bit the performance for sizes between 64 and 256 but
improves throughput for greater sizes.