From: Hans Kristian Rosbach <hk-git@circlestorm.org>
Date: Thu, 23 Feb 2017 08:04:48 +0000 (+0100)
Subject: Let all platforms defining UNALIGNED_OK use the optimized put_short
X-Git-Tag: 1.9.9-b1~675
X-Git-Url: http://git.ipfire.org/cgi-bin/gitweb.cgi?a=commitdiff_plain;h=47c616ce343c47b406e20ce6ae7cf644c3e8f6cf;p=thirdparty%2Fzlib-ng.git

Let all platforms defining UNALIGNED_OK use the optimized put_short
implementation. Also change from pre-increment to post-increment to
prevent a double-store on non-x86 platforms.
---

diff --git a/deflate.h b/deflate.h
index acd342d10..37e559680 100644
--- a/deflate.h
+++ b/deflate.h
@@ -307,7 +307,7 @@ typedef enum {
  * Output a short LSB first on the stream.
  * IN assertion: there is enough room in pendingBuf.
  */
-#if defined(__x86_64) || defined(__i386_)
+#ifdef UNALIGNED_OK
 /* 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
@@ -318,8 +318,8 @@ typedef enum {
  *    cost of penalty of potentially unaligned access. 
  */
 #define put_short(s, w) { \
+    *(uint16_t*)(&s->pending_buf[s->pending]) = (w) ; \
     s->pending += 2; \
-    *(uint16_t*)(&s->pending_buf[s->pending - 2]) = (w) ; \
 }
 #else
 #define put_short(s, w) { \