chacha20: add SSSE3 assembly implementation

* cipher/Makefile.am: Add 'chacha20-ssse3-amd64.S'.
* cipher/chacha20-ssse3-amd64.S: New.
* cipher/chacha20.c (USE_SSSE3): New macro.
[USE_SSSE3] (_gcry_chacha20_amd64_ssse3_blocks): New.
(chacha20_do_setkey): Select SSSE3 implementation if there is HW
support.
* configure.ac [host=x86-64]: Add 'chacha20-ssse3-amd64.lo'.
--

Add SSSE3 optimized implementation for ChaCha20. Based on implementation
by Andrew Moon.

Before (Intel Haswell):

 CHACHA20       |  nanosecs/byte   mebibytes/sec   cycles/byte
     STREAM enc |      1.97 ns/B     483.6 MiB/s      6.31 c/B
     STREAM dec |      1.97 ns/B     484.0 MiB/s      6.31 c/B

After:

 CHACHA20       |  nanosecs/byte   mebibytes/sec   cycles/byte
     STREAM enc |     0.742 ns/B    1284.8 MiB/s      2.38 c/B
     STREAM dec |     0.741 ns/B    1286.5 MiB/s      2.37 c/B

Signed-off-by: Jussi Kivilinna <jussi.kivilinna@iki.fi>

1 files changed, 22 insertions, 0 deletions

diff --git a/cipher/chacha20.c b/cipher/chacha20.c
index ff0366d3..de8982be 100644
--- a/cipher/chacha20.c
+++ b/cipher/chacha20.c
@@ -47,6 +47,13 @@
 #define CHACHA20_MAX_IV_SIZE  12        /* Bytes.  */
 #define CHACHA20_INPUT_LENGTH (CHACHA20_BLOCK_SIZE / 4)
 
+/* USE_SSSE3 indicates whether to compile with Intel SSSE3 code. */
+#undef USE_SSSE3
+#if defined(__x86_64__) && defined(HAVE_COMPATIBLE_GCC_AMD64_PLATFORM_AS) && \
+    defined(HAVE_GCC_INLINE_ASM_SSSE3)
+# define USE_SSSE3 1
+#endif
+
 
 struct CHACHA20_context_s;
 
@@ -63,6 +70,14 @@ typedef struct CHACHA20_context_s
 } CHACHA20_context_t;
 
 
+#ifdef USE_SSSE3
+
+unsigned int _gcry_chacha20_amd64_ssse3_blocks(u32 *state, const byte *in,
+                                               byte *out, size_t bytes);
+
+#endif /* USE_SSSE3 */
+
+
 static void chacha20_setiv (void *context, const byte * iv, size_t ivlen);
 static const char *selftest (void);
 
@@ -279,6 +294,7 @@ chacha20_do_setkey (CHACHA20_context_t * ctx,
 {
   static int initialized;
   static const char *selftest_failed;
+  unsigned int features = _gcry_get_hw_features ();
 
   if (!initialized)
     {
@@ -294,6 +310,12 @@ chacha20_do_setkey (CHACHA20_context_t * ctx,
     return GPG_ERR_INV_KEYLEN;
 
   ctx->blocks = chacha20_blocks;
+#ifdef USE_SSSE3
+  if (features & HWF_INTEL_SSSE3)
+    ctx->blocks = _gcry_chacha20_amd64_ssse3_blocks;
+#endif
+
+  (void)features;
 
   chacha20_keysetup (ctx, key, keylen);