noinline version of is_valid_utf8

boostorg · Mar 6, 2024 · d79a586 · d79a586
1 parent a284d0d
commit d79a586
Show file tree

Hide file tree

Showing 2 changed files with 44 additions and 88 deletions.
diff --git a/include/boost/json/detail/sse2.hpp b/include/boost/json/detail/sse2.hpp
@@ -136,7 +136,7 @@ count_valid<false>(
         uint8_t len = first & 0xFF;
         if(BOOST_JSON_UNLIKELY(end - p < len))
             break;
-        if(BOOST_JSON_UNLIKELY(! is_valid_utf8(p, first)))
+        if(BOOST_JSON_UNLIKELY(! is_valid_utf8_no_inline(p, first)))
             break;
         p += len;
     }
@@ -185,7 +185,7 @@ count_valid<false>(
         uint8_t len = first & 0xFF;
         if(BOOST_JSON_UNLIKELY(end - p < len))
             break;
-        if(BOOST_JSON_UNLIKELY(! is_valid_utf8(p, first)))
+        if(BOOST_JSON_UNLIKELY(! is_valid_utf8_no_inline(p, first)))
             break;
         p += len;
     }

diff --git a/include/boost/json/detail/utf8.hpp b/include/boost/json/detail/utf8.hpp
@@ -21,58 +21,12 @@ namespace boost {
 namespace json {
 namespace detail {
 
-template<endian::order = endian::order::little>
-constexpr
-std::uint32_t
-make_u32_impl(std::uint8_t b4, std::uint8_t b3, std::uint8_t b2, std::uint8_t b1)
-{
-    return (b4 << 24) | (b3 << 16) | (b2 << 8) | b1;
-}
-
-template<>
-constexpr
-std::uint32_t
-make_u32_impl<endian::order::big>(
-    std::uint8_t b4, std::uint8_t b3, std::uint8_t b2, std::uint8_t b1)
-{
-    return (b1 << 24) | (b2 << 16) | (b3 << 8) | b4;
-}
-
-constexpr
-std::uint32_t
-make_u32(std::uint8_t b4, std::uint8_t b3, std::uint8_t b2, std::uint8_t b1)
-{
-    return make_u32_impl<endian::order::native>(b4, b3, b2, b1);
-}
-
-template<endian::order = endian::order::little>
-constexpr
-bool
-utf8_case5(std::uint32_t v)
-{
-    return ( ( ( v & make_u32(0xC0,0xC0,0xF0,0x00) )
-        + make_u32(0x7F,0x7F,0x70,0x00) ) | make_u32(0x00,0x00,0x30,0x00) )
-        == make_u32(0x00,0x00,0x30,0x00);
-}
-
-template<>
-constexpr
-bool
-utf8_case5<endian::order::big>(std::uint32_t v)
-{
-    return ( ( ( v & make_u32(0xC0,0xC0,0xF0,0x00) )
-        + make_u32(0x00,0x00,0x70,0xFF) ) | make_u32(0x00,0x00,0x30,0x00) )
-        == make_u32(0x80,0x80,0x30,0x00);
-}
-
 template<int N>
 std::uint32_t
 load_little_endian(void const* p)
 {
-    std::uint32_t v = 0;
-    std::memcpy(&v, p, N);
-    endian::little_to_native_inplace(v);
-    return v;
+    auto const up = reinterpret_cast<unsigned char const*>(p);
+    return endian::endian_load<std::uint32_t, N, endian::order::little>(up);
 }
 
 inline
@@ -122,47 +76,49 @@ is_valid_utf8(const char* p, uint16_t first)
 
     // 2 bytes, second byte [80, BF]
     case 1:
-        std::memcpy(&v, p, 2);
-        return ( v & make_u32(0x00,0x00,0xC0,0x00) )
-            == make_u32(0x00,0x00,0x80,0x00);
-
-    // 3 bytes, second byte [A0, BF]
-    case 2:
-        std::memcpy(&v, p, 3);
-        return ( v & make_u32(0x00,0xC0,0xE0,0x00) )
-            == make_u32(0x00,0x80,0xA0,0x00);
-
-    // 3 bytes, second byte [80, BF]
-    case 3:
-        std::memcpy(&v, p, 3);
-        return ( v & make_u32(0x00,0xC0,0xC0,0x00) )
-            == make_u32(0x00,0x80,0x80,0x00);
-
-    // 3 bytes, second byte [80, 9F]
-    case 4:
-        std::memcpy(&v, p, 3);
-        return ( v & make_u32(0x00,0xC0,0xE0,0x00) )
-            == make_u32(0x00,0x80,0x80,0x00);
-
-    // 4 bytes, second byte [90, BF]
-    case 5:
-        std::memcpy(&v, p, 4);
-        return utf8_case5<endian::order::native>(v);
-
-    // 4 bytes, second byte [80, BF]
-    case 6:
-        std::memcpy(&v, p, 4);
-        return ( v & make_u32(0xC0,0xC0,0xC0,0x00) )
-            == make_u32(0x80,0x80,0x80,0x00);
-
-    // 4 bytes, second byte [80, 8F]
-    case 7:
-        std::memcpy(&v, p, 4);
-        return ( v & make_u32(0xC0,0xC0,0xF0,0x00) )
-            == make_u32(0x80,0x80,0x80,0x00);
+        v = load_little_endian<2>(p);
+        return (v & 0xC000) == 0x8000;
+
+     // 3 bytes, second byte [A0, BF]
+     case 2:
+        v = load_little_endian<3>(p);
+        return (v & 0xC0E000) == 0x80A000;
+
+     // 3 bytes, second byte [80, BF]
+     case 3:
+        v = load_little_endian<3>(p);
+        return (v & 0xC0C000) == 0x808000;
+
+     // 3 bytes, second byte [80, 9F]
+     case 4:
+        v = load_little_endian<3>(p);
+        return (v & 0xC0E000) == 0x808000;
+
+     // 4 bytes, second byte [90, BF]
+     case 5:
+        v = load_little_endian<4>(p);
+        return (v & 0xC0C0FF00) + 0x7F7F7000 <= 0x2F00;
+
+     // 4 bytes, second byte [80, BF]
+     case 6:
+        v = load_little_endian<4>(p);
+        return (v & 0xC0C0C000) == 0x80808000;
+
+     // 4 bytes, second byte [80, 8F]
+     case 7:
+        v = load_little_endian<4>(p);
+        return (v & 0xC0C0F000) == 0x80808000;
     }
 }
 
+BOOST_NOINLINE
+inline
+bool
+is_valid_utf8_no_inline(const char* p, uint16_t first)
+{
+    return is_valid_utf8(p, first);
+}
+
 class utf8_sequence
 {
     char seq_[4];