From: Shuxin Yang <shuxinyang2006@gmail.com>
Date: Sun, 20 Apr 2014 22:50:33 +0000 (-0700)
Subject: Minor enhancement to put_short() macro. This change saw marginal speedup
X-Git-Tag: 1.9.9-b1~933
X-Git-Url: http://git.ipfire.org/cgi-bin/gitweb.cgi?a=commitdiff_plain;h=666581bbc17eaea62acb47cf47ab35a037e9b9d0;p=thirdparty%2Fzlib-ng.git

Minor enhancement to put_short() macro. This change saw marginal speedup
(about 0% to 3% depending on the compression level and input). I guess
the speedup likely arises from following facts:

 1) "s->pending" now is loaded once, and stored once. In the original
    implementation, it needs to be loaded and stored twice as the
    compiler isn't able to disambiguate "s->pending" and
    "s->pending_buf[]"

 2) better code generations:
   2.1) no instruction are needed for extracting two bytes from a short.
   2.2) need less registers
   2.3) stores to adjacent bytes are merged into a single store, albeit
        at the cost of penalty of potentially unaligned access.

Conflicts:
	trees.c
---

diff --git a/deflate.h b/deflate.h
index 5cf2ce126..28140d2ec 100644
--- a/deflate.h
+++ b/deflate.h
@@ -291,6 +291,30 @@ typedef enum {
  */
 #define put_byte(s, c) {s->pending_buf[s->pending++] = (c);}
 
+/* ===========================================================================
+ * Output a short LSB first on the stream.
+ * IN assertion: there is enough room in pendingBuf.
+ */
+#if defined(__x86_64) || defined(__i386_)
+/* Compared to the else-clause's implementation, there are few advantages:
+ *  - s->pending is loaded only once (else-clause's implementation needs to
+ *    load s->pending twice due to the alias between s->pending and
+ *    s->pending_buf[].
+ *  - no instructions for extracting bytes from short.
+ *  - needs less registers
+ *  - stores to adjacent bytes are merged into a single store, albeit at the
+ *    cost of penalty of potentially unaligned access. 
+ */
+#define put_short(s, w) { \
+    s->pending += 2; \
+    *(ush*)(&s->pending_buf[s->pending - 2]) = (w) ; \
+}
+#else
+#define put_short(s, w) { \
+    put_byte(s, (uch)((w) & 0xff)); \
+    put_byte(s, (uch)((ush)(w) >> 8)); \
+}
+#endif
 
 #define MIN_LOOKAHEAD (MAX_MATCH+MIN_MATCH+1)
 /* Minimum amount of lookahead, except at the end of the input file.