chacha20: add ARMv7/NEON implementation

* cipher/Makefile.am: Add 'chacha20-armv7-neon.S'.
* cipher/chacha20-armv7-neon.S: New.
* cipher/chacha20.c (USE_NEON): New.
[USE_NEON] (_gcry_chacha20_armv7_neon_blocks): New.
(chacha20_do_setkey) [USE_NEON]: Use Neon implementation if
HWF_ARM_NEON flag set.
(selftest): Self-test encrypting buffer byte by byte.
* configure.ac [neonsupport=yes]: Add 'chacha20-armv7-neon.lo'.
--

Add Andrew Moon's public domain ARMv7/NEON implementation of ChaCha20. Original
source is available at: https://github.com/floodyberry/chacha-opt

Benchmark on Cortex-A8 (--cpu-mhz 1008):

Old:
 CHACHA20       |  nanosecs/byte   mebibytes/sec   cycles/byte
     STREAM enc |     13.45 ns/B     70.92 MiB/s     13.56 c/B
     STREAM dec |     13.45 ns/B     70.90 MiB/s     13.56 c/B

New:
 CHACHA20       |  nanosecs/byte   mebibytes/sec   cycles/byte
     STREAM enc |      6.20 ns/B     153.9 MiB/s      6.25 c/B
     STREAM dec |      6.20 ns/B     153.9 MiB/s      6.25 c/B

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

1 files changed, 34 insertions, 0 deletions

diff --git a/cipher/chacha20.c b/cipher/chacha20.c
index ebba2fcb..c1847aa5 100644
--- a/cipher/chacha20.c
+++ b/cipher/chacha20.c
@@ -67,6 +67,16 @@
 # define USE_AVX2 1
 #endif
 
+/* USE_NEON indicates whether to enable ARM NEON assembly code. */
+#undef USE_NEON
+#ifdef ENABLE_NEON_SUPPORT
+# if defined(HAVE_ARM_ARCH_V6) && defined(__ARMEL__) \
+     && defined(HAVE_COMPATIBLE_GCC_ARM_PLATFORM_AS) \
+     && defined(HAVE_GCC_INLINE_ASM_NEON)
+#  define USE_NEON 1
+# endif
+#endif /*ENABLE_NEON_SUPPORT*/
+
 
 struct CHACHA20_context_s;
 
@@ -104,6 +114,13 @@ unsigned int _gcry_chacha20_amd64_avx2_blocks(u32 *state, const byte *in,
 
 #endif /* USE_AVX2 */
 
+#ifdef USE_NEON
+
+unsigned int _gcry_chacha20_armv7_neon_blocks(u32 *state, const byte *in,
+                                              byte *out, size_t bytes);
+
+#endif /* USE_NEON */
+
 
 static void chacha20_setiv (void *context, const byte * iv, size_t ivlen);
 static const char *selftest (void);
@@ -353,6 +370,10 @@ chacha20_do_setkey (CHACHA20_context_t * ctx,
   if (features & HWF_INTEL_AVX2)
     ctx->blocks = _gcry_chacha20_amd64_avx2_blocks;
 #endif
+#ifdef USE_NEON
+  if (features & HWF_ARM_NEON)
+    ctx->blocks = _gcry_chacha20_armv7_neon_blocks;
+#endif
 
   (void)features;
 
@@ -541,6 +562,19 @@ selftest (void)
     if (buf[i] != (byte) i)
       return "ChaCha20 encryption test 2 failed.";
 
+  chacha20_setkey (&ctx, key_1, sizeof key_1);
+  chacha20_setiv (&ctx, nonce_1, sizeof nonce_1);
+  /* encrypt */
+  for (i = 0; i < sizeof buf; i++)
+    chacha20_encrypt_stream (&ctx, &buf[i], &buf[i], 1);
+  /* decrypt */
+  chacha20_setkey (&ctx, key_1, sizeof key_1);
+  chacha20_setiv (&ctx, nonce_1, sizeof nonce_1);
+  chacha20_encrypt_stream (&ctx, buf, buf, sizeof buf);
+  for (i = 0; i < sizeof buf; i++)
+    if (buf[i] != (byte) i)
+      return "ChaCha20 encryption test 3 failed.";
+
   return NULL;
 }