From c2a9915a451a3c6ad618f55a2871a43403333b7f Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Bj=C3=B6rn=20Gustavsson?= <bjorn@erlang.org>
Date: Wed, 17 Jan 2018 16:01:28 +0100
Subject: Optimize matching of an 'utf8' segment in the binary syntax

Matching out an 8-bit integer is faster than matching out
an utf8-encoded code point, even if the value of the code
point is less than 128. The reason is that matching out
an 8-bit integer is specially optimized to avoid a function
call. Do a similar optimization for matching out an utf8
segment.
---
 erts/emulator/beam/bs_instrs.tab | 16 +++++++++++++++-
 1 file changed, 15 insertions(+), 1 deletion(-)

(limited to 'erts')

diff --git a/erts/emulator/beam/bs_instrs.tab b/erts/emulator/beam/bs_instrs.tab
index 9f03b19731..b11903a47b 100644
--- a/erts/emulator/beam/bs_instrs.tab
+++ b/erts/emulator/beam/bs_instrs.tab
@@ -919,9 +919,23 @@ i_bs_get_integer(Fail, Live, FlagsAndUnit, Ms, Sz, Dst) {
 }
 
 i_bs_get_utf8(Ctx, Fail, Dst) {
+    Eterm result;
     ErlBinMatchBuffer* mb = ms_matchbuffer($Ctx);
-    Eterm result = erts_bs_get_utf8(mb);
 
+    if (mb->size - mb->offset < 8) {
+        $FAIL($Fail);
+    }
+    if (BIT_OFFSET(mb->offset) != 0) {
+        result = erts_bs_get_utf8(mb);
+    } else {
+        byte b = mb->base[BYTE_OFFSET(mb->offset)];
+        if (b < 128) {
+            result = make_small(b);
+            mb->offset += 8;
+        } else {
+            result = erts_bs_get_utf8(mb);
+        }
+    }
     if (is_non_value(result)) {
         $FAIL($Fail);
     }
-- 
cgit v1.2.3