diff mbox series

[v2,5/6] crypto: lib/sha256 - Unroll LOAD and BLEND loops

Message ID 20201020203957.3512851-6-nivedita@alum.mit.edu
State Accepted
Commit 18d05ca4486fe38991c3166b1f4df26b8a029665
Headers show
Series crypto: lib/sha256 - cleanup/optimization | expand

Commit Message

Arvind Sankar Oct. 20, 2020, 8:39 p.m. UTC
Unrolling the LOAD and BLEND loops improves performance by ~8% on x86_64
(tested on Broadwell Xeon) while not increasing code size too much.

Signed-off-by: Arvind Sankar <nivedita@alum.mit.edu>
---
 lib/crypto/sha256.c | 24 ++++++++++++++++++++----
 1 file changed, 20 insertions(+), 4 deletions(-)

Comments

Eric Biggers Oct. 22, 2020, 5:02 a.m. UTC | #1
On Tue, Oct 20, 2020 at 04:39:56PM -0400, Arvind Sankar wrote:
> Unrolling the LOAD and BLEND loops improves performance by ~8% on x86_64

> (tested on Broadwell Xeon) while not increasing code size too much.

> 

> Signed-off-by: Arvind Sankar <nivedita@alum.mit.edu>

> ---


Looks good,

Reviewed-by: Eric Biggers <ebiggers@google.com>
diff mbox series

Patch

diff --git a/lib/crypto/sha256.c b/lib/crypto/sha256.c
index 5efd390706c6..3a8802d5f747 100644
--- a/lib/crypto/sha256.c
+++ b/lib/crypto/sha256.c
@@ -68,12 +68,28 @@  static void sha256_transform(u32 *state, const u8 *input, u32 *W)
 	int i;
 
 	/* load the input */
-	for (i = 0; i < 16; i++)
-		LOAD_OP(i, W, input);
+	for (i = 0; i < 16; i += 8) {
+		LOAD_OP(i + 0, W, input);
+		LOAD_OP(i + 1, W, input);
+		LOAD_OP(i + 2, W, input);
+		LOAD_OP(i + 3, W, input);
+		LOAD_OP(i + 4, W, input);
+		LOAD_OP(i + 5, W, input);
+		LOAD_OP(i + 6, W, input);
+		LOAD_OP(i + 7, W, input);
+	}
 
 	/* now blend */
-	for (i = 16; i < 64; i++)
-		BLEND_OP(i, W);
+	for (i = 16; i < 64; i += 8) {
+		BLEND_OP(i + 0, W);
+		BLEND_OP(i + 1, W);
+		BLEND_OP(i + 2, W);
+		BLEND_OP(i + 3, W);
+		BLEND_OP(i + 4, W);
+		BLEND_OP(i + 5, W);
+		BLEND_OP(i + 6, W);
+		BLEND_OP(i + 7, W);
+	}
 
 	/* load the state into our registers */
 	a = state[0];  b = state[1];  c = state[2];  d = state[3];