Did that mixture of shifting by 8 and 16 generate better code
on some architecture? This is easier to read.
---
crypto/des_generic.c | 22 +++++++++++-----------
1 files changed, 11 insertions(+), 11 deletions(-)
diff --git a/crypto/des_generic.c b/crypto/des_generic.c
index afde5b4..1b74adf 100644
--- a/crypto/des_generic.c
+++ b/crypto/des_generic.c
@@ -553,17 +553,17 @@ static const u32 S8[64] = {
ROR(R, 4);
#define ROUND(L, R, A, B, K, d) \
- B = K[0]; A = K[1]; K += d; \
- B ^= R; A ^= R; \
- B &= 0x3f3f3f3f; ROR(A, 4); \
- L ^= S8[0xff & B]; A &= 0x3f3f3f3f; \
- L ^= S6[0xff & (B >> 8)]; B >>= 16; \
- L ^= S7[0xff & A]; \
- L ^= S5[0xff & (A >> 8)]; A >>= 16; \
- L ^= S4[0xff & B]; \
- L ^= S2[0xff & (B >> 8)]; \
- L ^= S3[0xff & A]; \
- L ^= S1[0xff & (A >> 8)];
+ B = K[0]; A = K[1]; K += d; \
+ B ^= R; A ^= R; \
+ B &= 0x3f3f3f3f; ROR(A, 4); \
+ A &= 0x3f3f3f3f; L ^= S8[0xff & B]; B >>= 8;\
+ L ^= S7[0xff & A]; A >>= 8; \
+ L ^= S6[0xff & B]; B >>= 8; \
+ L ^= S5[0xff & A]; A >>= 8; \
+ L ^= S4[0xff & B]; B >>= 8; \
+ L ^= S3[0xff & A]; A >>= 8; \
+ L ^= S2[0xff & B]; \
+ L ^= S1[0xff & A];
/*
* PC2 lookup tables are organized as 2 consecutive sets of 4 interleaved
--
1.6.0.6