diff options
author | Jeffrey Walton <noloader@gmail.com> | 2018-11-17 00:41:49 -0500 |
---|---|---|
committer | Jeffrey Walton <noloader@gmail.com> | 2018-11-17 00:41:49 -0500 |
commit | bbc5c63d332e317ca7ceb7b71e59a517a16b71d8 (patch) | |
tree | f93d3c627340776b2119f4d4408005e73a814127 /gcm_simd.cpp | |
parent | 0ac7b252213cac5a9be61f45319d90f9cf207150 (diff) | |
download | cryptopp-git-bbc5c63d332e317ca7ceb7b71e59a517a16b71d8.tar.gz |
Drop GCM to POWER7 on PowerPC
GCM can do some bulk XOR's using the SIMD unit. However, we still need loads and stores to be fast. Fast loads and stores of unaligned data requires the VSX unit
Diffstat (limited to 'gcm_simd.cpp')
-rw-r--r-- | gcm_simd.cpp | 4 |
1 files changed, 2 insertions, 2 deletions
diff --git a/gcm_simd.cpp b/gcm_simd.cpp index 3e42f16b..b701f398 100644 --- a/gcm_simd.cpp +++ b/gcm_simd.cpp @@ -740,8 +740,8 @@ void GCM_ReverseHashBufferIfNeeded_CLMUL(byte *hashBuffer) // ***************************** POWER8 ***************************** //
-#if CRYPTOPP_ALTIVEC_AVAILABLE
-void GCM_Xor16_ALTIVEC(byte *a, const byte *b, const byte *c)
+#if CRYPTOPP_POWER7_AVAILABLE
+void GCM_Xor16_POWER7(byte *a, const byte *b, const byte *c)
{
VecStore(VecXor(VecLoad(b), VecLoad(c)), a);
}
|