[libc-commits] [libc] [libc] Add support for string/memory_utils functions for AArch64 without HW FP/SIMD (PR #137592)

[Guillaume Chatelet via libc-commits]

================
@@ -19,13 +19,35 @@
 
 namespace LIBC_NAMESPACE_DECL {
 
-[[maybe_unused]] LIBC_INLINE BcmpReturnType inline_bcmp_aarch64(CPtr p1,
-                                                                CPtr p2,
-                                                                size_t count) {
-  if (LIBC_LIKELY(count <= 32)) {
-    if (LIBC_UNLIKELY(count >= 16)) {
-      return aarch64::Bcmp<16>::head_tail(p1, p2, count);
-    }
+[[maybe_unused]] LIBC_INLINE BcmpReturnType
+inline_bcmp_aarch64_no_fp(CPtr p1, CPtr p2, size_t count) {
+  return generic::Bcmp<uint64_t>::loop_and_tail_align_above(256, p1, p2, count);
+}
+
+#ifdef __ARM_NEON
+[[maybe_unused]] LIBC_INLINE BcmpReturnType
+inline_bcmp_aarch64_with_fp(CPtr p1, CPtr p2, size_t count) {
+  if (count <= 32) {
+    return aarch64::Bcmp<16>::head_tail(p1, p2, count);
+  }
+
+  if (count <= 64) {
+    return aarch64::Bcmp<32>::head_tail(p1, p2, count);
+  }
+
+  if (LIBC_UNLIKELY(count > 256)) {
+    if (auto value = aarch64::Bcmp<32>::block(p1, p2))
+      return value;
+    align_to_next_boundary<16, Arg::P1>(p1, p2, count);
+  }
+
+  return aarch64::Bcmp<32>::loop_and_tail(p1, p2, count);
+}
+#endif
+
+[[gnu::flatten]] LIBC_INLINE BcmpReturnType
+inline_bcmp_aarch64_dispatch(CPtr p1, CPtr p2, size_t count) {
+  if (LIBC_LIKELY(count <= 16)) {
----------------
gchatelet wrote:

Can you completely duplicate the code in `inline_bcmp_aarch64_no_fp` and `inline_bcmp_aarch64_with_fp` so we can make sure the code is exactly the same as before for the latter. Currently the `if (LIBC_UNLIKELY(count >= 16))` is gone.

https://github.com/llvm/llvm-project/pull/137592