OPTIM: http: improve parsing performance of long URIs

Searching the trailing space in long URIs takes some time. This can
happen especially on static files and some blogs. By skipping valid
character ranges by 32-bit blocks, it's possible to increase the
HTTP performance from 212k to 216k req/s on requests features a
100-character URI, which is an increase of 2%. This is done for
architectures supporting unaligned accesses (x86_64, x86, armv7a).
There's only a 32-bit version because URIs are rarely long and very
often short, so it's more efficient to limit the systematic overhead
than to try to optimize for the rarest requests.

diff --git a/src/proto_http.c b/src/proto_http.c
index c9f38b1b12790f535e266576d4dd920fae4e3244..36428b164024aaf929b081fd32bcd7bf5b15403e 100644 (file)
--- a/src/proto_http.c
+++ b/src/proto_http.c
@@ -1519,8 +1519,25 @@ const char *http_parse_reqline(struct http_msg *msg,
 
        case HTTP_MSG_RQURI:
        http_msg_rquri:
+#if defined(__x86_64__) ||                                             \
+    defined(__i386__) || defined(__i486__) || defined(__i586__) || defined(__i686__) || \
+    defined(__ARM_ARCH_7A__)
+               /* speedup: skip bytes not between 0x21 and 0x7e inclusive */
+               while (ptr <= end - sizeof(int)) {
+                       int x = *(int *)ptr - 0x21212121;
+                       if (x & 0x80808080)
+                               break;
+
+                       x -= 0x5e5e5e5e;
+                       if (!(x & 0x80808080))
+                               break;
+
+                       ptr += sizeof(int);
+               }
+#endif
+       http_msg_rquri2:
                if (likely((unsigned char)(*ptr - 33) <= 93)) /* 33 to 126 included */
-                       EAT_AND_JUMP_OR_RETURN(http_msg_rquri, HTTP_MSG_RQURI);
+                       EAT_AND_JUMP_OR_RETURN(http_msg_rquri2, HTTP_MSG_RQURI);
 
                if (likely(HTTP_IS_SPHT(*ptr))) {
                        msg->sl.rq.u_l = ptr - msg_start - msg->sl.rq.u;