libgrapheme

commit e917e53a05de7cab2591d6b2fa3f2ce5ece89bcf
parent 10a95f711451bc653ba5d1f14c0f514e4d854c44
Author: Laslo Hunhold <dev@frign.de>
Date:   Tue,  4 Jan 2022 18:56:38 +0100

Mark likely branches

The likely() and unlikely() macros (known from the Linux kernel) in
src/util.h are defined in a portable way.

This addition brings a performance improvment of around 6%, which also
shows how well-optimized grapheme_is_character_break() already is. One
break check for two codepoints now takes only around 10µs on average on
my machine, which is insanely fast when you consider the complexity of
the algorithm behind it.

Signed-off-by: Laslo Hunhold <dev@frign.de>

Diffstat:
M
src/character.c
 | 
16
++++++++--------
M
src/util.h
 | 
15
+++++++++++++++

2 files changed, 23 insertions(+), 8 deletions(-)
diff --git a/src/character.c b/src/character.c
@@ -105,10 +105,10 @@ static const uint_least16_t dont_break_gb12_13[2 * NUM_BREAK_PROPS] = {
 static enum break_property
 get_break_prop(uint_least32_t cp)
 {
-	if (cp > 0x10FFFF) {
-		return BREAK_PROP_OTHER;
-	} else {
+	if (likely(cp <= 0x10FFFF)) {
 		return prop[minor[major[cp >> 8] + (cp & 0xff)]].break_property;
+	} else {
+		return BREAK_PROP_OTHER;
 	}
 }
 
@@ -118,11 +118,11 @@ grapheme_is_character_break(uint_least32_t cp0, uint_least32_t cp1, GRAPHEME_STA
 	enum break_property cp0_prop, cp1_prop;
 	bool notbreak = false;
 
-	if (state) {
-		if (!state->prop_set) {
-			cp0_prop = get_break_prop(cp0);
-		} else {
+	if (likely(state)) {
+		if (likely(state->prop_set)) {
 			cp0_prop = state->prop;
+		} else {
+			cp0_prop = get_break_prop(cp0);
 		}
 		cp1_prop = get_break_prop(cp1);
 
@@ -153,7 +153,7 @@ grapheme_is_character_break(uint_least32_t cp0, uint_least32_t cp1, GRAPHEME_STA
 		            UINT16_C(1 << cp1_prop));
 
 		/* update or reset flags (when we have a break) */
-		if (!notbreak) {
+		if (likely(!notbreak)) {
 			state->gb11_flag = state->gb12_13_flag = false;
 		}
 	} else {
diff --git a/src/util.h b/src/util.h
@@ -10,4 +10,19 @@
 
 #define LEN(x) (sizeof(x) / sizeof(*(x)))
 
+#undef likely
+#undef unlikely
+#ifdef __has_builtin
+	#if __has_builtin(__builtin_expect)
+		#define likely(expr) __builtin_expect(!!(expr), 1)
+		#define unlikely(expr) __builtin_expect(!!(expr), 0)
+	#else
+		#define likely(expr) (expr)
+		#define unlikely(expr) (expr)
+	#endif
+#else
+	#define likely(expr) (expr)
+	#define unlikely(expr) (expr)
+#endif
+
 #endif /* UTIL_H */