* [PATCH v3 1/6] crypto: testmgr - avoid overlap in chunked tests
2016-12-05 18:42 [PATCH v3 0/6] crypto: ARM/arm64 CRC-T10DIF/CRC32/CRC32C roundup Ard Biesheuvel
@ 2016-12-05 18:42 ` Ard Biesheuvel
2016-12-07 19:19 ` Eric Biggers
2016-12-05 18:42 ` [PATCH v3 2/6] crypto: testmgr - add/enhance test cases for CRC-T10DIF Ard Biesheuvel
` (5 subsequent siblings)
6 siblings, 1 reply; 11+ messages in thread
From: Ard Biesheuvel @ 2016-12-05 18:42 UTC (permalink / raw)
To: linux-arm-kernel
The IDXn offsets are chosen such that tap values (which may go up to
255) end up overlapping in the xbuf allocation. In particular, IDX1
and IDX3 are too close together, so update IDX3 to avoid this issue.
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
---
crypto/testmgr.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/crypto/testmgr.c b/crypto/testmgr.c
index ded50b67c757..670893bcf361 100644
--- a/crypto/testmgr.c
+++ b/crypto/testmgr.c
@@ -63,7 +63,7 @@ int alg_test(const char *driver, const char *alg, u32 type, u32 mask)
*/
#define IDX1 32
#define IDX2 32400
-#define IDX3 1
+#define IDX3 511
#define IDX4 8193
#define IDX5 22222
#define IDX6 17101
--
2.7.4
^ permalink raw reply related [flat|nested] 11+ messages in thread* [PATCH v3 1/6] crypto: testmgr - avoid overlap in chunked tests
2016-12-05 18:42 ` [PATCH v3 1/6] crypto: testmgr - avoid overlap in chunked tests Ard Biesheuvel
@ 2016-12-07 19:19 ` Eric Biggers
2016-12-07 19:53 ` Ard Biesheuvel
0 siblings, 1 reply; 11+ messages in thread
From: Eric Biggers @ 2016-12-07 19:19 UTC (permalink / raw)
To: linux-arm-kernel
On Mon, Dec 05, 2016 at 06:42:23PM +0000, Ard Biesheuvel wrote:
> The IDXn offsets are chosen such that tap values (which may go up to
> 255) end up overlapping in the xbuf allocation. In particular, IDX1
> and IDX3 are too close together, so update IDX3 to avoid this issue.
>
Hi Ard,
This patch is causing the self-tests for "xts(ecb(aes-asm))" to fail.
This is on x86. Any idea why? Here's what I see in the log:
alg: skcipher: Chunk test 1 failed on encryption at page 0 for xts(ecb(aes-asm))
00000000: 1c 3b 3a 10 2f 77 03 86 e4 83 6c 99 e3 70 cf 9b
00000010: ea 00 80 3f 5e 48 23 57 a4 ae 12 d4 14 a3 e6 3b
00000020: 5d 31 e2 76 f8 fe 4a 8d 66 b3 17 f9 ac 68 3f 44
00000030: 68 0a 86 ac 35 ad fc 33 45 be fe cb 4b b1 88 fd
00000040: 57 76 92 6c 49 a3 09 5e b1 08 fd 10 98 ba ec 70
00000050: aa a6 69 99 a7 2a 82 f2 7d 84 8b 21 d4 a7 41 b0
00000060: c5 cd 4d 5f ff 9d ac 89 ae ba 12 29 61 d0 3a 75
00000070: 71 23 e9 87 0f 8a cf 10 00 02 08 87 89 14 29 ca
00000080: 2a 3e 7a 7d 7d f7 b1 03 55 16 5c 8b 9a 6d 0a 7d
00000090: e8 b0 62 c4 50 0d c4 cd 12 0c 0f 74 18 da e3 d0
000000a0: b5 78 1c 34 80 3f a7 54 21 c7 90 df e1 de 18 34
000000b0: f2 80 d7 66 7b 32 7f 6c 8c d7 55 7e 12 ac 3a 0f
000000c0: 93 ec 05 c5 2e 04 93 ef 31 a1 2d 3d 92 60 f7 9a
000000d0: 28 9d 6a 37 9b c7 0c 50 84 14 73 d1 a8 cc 81 ec
000000e0: 58 3e 96 45 e0 7b 8d 96 70 65 5b a5 bb cf ec c6
000000f0: dc 39 66 38 0a d8 fe cb 17 b6 ba 02 46 9a 02 0a
00000100: 84 e1 8e 8f 84 25 20 70 c1 3e 9f 1f 28 9b e5 4f
00000110: bc 48 14 57 77 8f 61 60 15 e1 32 7a 02 b1 40 f1
00000120: 50 5e b3 09 32 6d 68 37 8f 83 74 59 5c 84 9d 84
00000130: f4 c3 33 ec 44 23 88 51 43 cb 47 bd 71 c5 ed ae
00000140: 9b e6 9a 2f fe ce b1 be c9 de 24 4f be 15 99 2b
00000150: 11 b7 7c 04 0f 12 bd 8f 6a 97 5a 44 a0 f9 0c 29
00000160: a9 ab c3 d4 d8 93 92 72 84 c5 87 54 cc e2 94 52
00000170: 9f 86 14 dc d2 ab a9 91 92 5f ed c4 ae 74 ff ac
00000180: 6e 33 3b 93 eb 4a ff 04 79 da 9a 41 0e 44 50 e0
00000190: dd 7a e4 c6 e2 91 09 00 57 5d a4 01 fc 07 05 9f
000001a0: 64 5e 8b 7e 9b fd ef 33 94 30 54 ff 84 01 14 93
000001b0: c2 7b 34 29 ea ed b4 ed 53 76 44 1a 77 ed 43 85
000001c0: 1a d7 7f 16 f5 41 df d2 69 d5 0d 6a 5f 14 fb 0a
000001d0: 1e 2a 8f 42 61 9e 5e c2 59 bd 96 d0 e5 cc 23 1f
000001e0: fb 84 ed 15 a8 eb 66 07 31 6b f6 ef
Eric
^ permalink raw reply [flat|nested] 11+ messages in thread
* [PATCH v3 1/6] crypto: testmgr - avoid overlap in chunked tests
2016-12-07 19:19 ` Eric Biggers
@ 2016-12-07 19:53 ` Ard Biesheuvel
2016-12-07 20:23 ` Eric Biggers
0 siblings, 1 reply; 11+ messages in thread
From: Ard Biesheuvel @ 2016-12-07 19:53 UTC (permalink / raw)
To: linux-arm-kernel
On 7 December 2016 at 19:19, Eric Biggers <ebiggers@google.com> wrote:
> On Mon, Dec 05, 2016 at 06:42:23PM +0000, Ard Biesheuvel wrote:
>> The IDXn offsets are chosen such that tap values (which may go up to
>> 255) end up overlapping in the xbuf allocation. In particular, IDX1
>> and IDX3 are too close together, so update IDX3 to avoid this issue.
>>
>
> Hi Ard,
>
> This patch is causing the self-tests for "xts(ecb(aes-asm))" to fail.
> This is on x86. Any idea why?
Does this help at all?
diff --git a/crypto/testmgr.c b/crypto/testmgr.c
index 670893bcf361..59e67f5b544b 100644
--- a/crypto/testmgr.c
+++ b/crypto/testmgr.c
@@ -63,7 +63,7 @@ int alg_test(const char *driver, const char *alg,
u32 type, u32 mask)
*/
#define IDX1 32
#define IDX2 32400
-#define IDX3 511
+#define IDX3 1511
#define IDX4 8193
#define IDX5 22222
#define IDX6 17101
> Here's what I see in the log:
>
> alg: skcipher: Chunk test 1 failed on encryption at page 0 for xts(ecb(aes-asm))
> 00000000: 1c 3b 3a 10 2f 77 03 86 e4 83 6c 99 e3 70 cf 9b
> 00000010: ea 00 80 3f 5e 48 23 57 a4 ae 12 d4 14 a3 e6 3b
> 00000020: 5d 31 e2 76 f8 fe 4a 8d 66 b3 17 f9 ac 68 3f 44
> 00000030: 68 0a 86 ac 35 ad fc 33 45 be fe cb 4b b1 88 fd
> 00000040: 57 76 92 6c 49 a3 09 5e b1 08 fd 10 98 ba ec 70
> 00000050: aa a6 69 99 a7 2a 82 f2 7d 84 8b 21 d4 a7 41 b0
> 00000060: c5 cd 4d 5f ff 9d ac 89 ae ba 12 29 61 d0 3a 75
> 00000070: 71 23 e9 87 0f 8a cf 10 00 02 08 87 89 14 29 ca
> 00000080: 2a 3e 7a 7d 7d f7 b1 03 55 16 5c 8b 9a 6d 0a 7d
> 00000090: e8 b0 62 c4 50 0d c4 cd 12 0c 0f 74 18 da e3 d0
> 000000a0: b5 78 1c 34 80 3f a7 54 21 c7 90 df e1 de 18 34
> 000000b0: f2 80 d7 66 7b 32 7f 6c 8c d7 55 7e 12 ac 3a 0f
> 000000c0: 93 ec 05 c5 2e 04 93 ef 31 a1 2d 3d 92 60 f7 9a
> 000000d0: 28 9d 6a 37 9b c7 0c 50 84 14 73 d1 a8 cc 81 ec
> 000000e0: 58 3e 96 45 e0 7b 8d 96 70 65 5b a5 bb cf ec c6
> 000000f0: dc 39 66 38 0a d8 fe cb 17 b6 ba 02 46 9a 02 0a
> 00000100: 84 e1 8e 8f 84 25 20 70 c1 3e 9f 1f 28 9b e5 4f
> 00000110: bc 48 14 57 77 8f 61 60 15 e1 32 7a 02 b1 40 f1
> 00000120: 50 5e b3 09 32 6d 68 37 8f 83 74 59 5c 84 9d 84
> 00000130: f4 c3 33 ec 44 23 88 51 43 cb 47 bd 71 c5 ed ae
> 00000140: 9b e6 9a 2f fe ce b1 be c9 de 24 4f be 15 99 2b
> 00000150: 11 b7 7c 04 0f 12 bd 8f 6a 97 5a 44 a0 f9 0c 29
> 00000160: a9 ab c3 d4 d8 93 92 72 84 c5 87 54 cc e2 94 52
> 00000170: 9f 86 14 dc d2 ab a9 91 92 5f ed c4 ae 74 ff ac
> 00000180: 6e 33 3b 93 eb 4a ff 04 79 da 9a 41 0e 44 50 e0
> 00000190: dd 7a e4 c6 e2 91 09 00 57 5d a4 01 fc 07 05 9f
> 000001a0: 64 5e 8b 7e 9b fd ef 33 94 30 54 ff 84 01 14 93
> 000001b0: c2 7b 34 29 ea ed b4 ed 53 76 44 1a 77 ed 43 85
> 000001c0: 1a d7 7f 16 f5 41 df d2 69 d5 0d 6a 5f 14 fb 0a
> 000001d0: 1e 2a 8f 42 61 9e 5e c2 59 bd 96 d0 e5 cc 23 1f
> 000001e0: fb 84 ed 15 a8 eb 66 07 31 6b f6 ef
>
> Eric
^ permalink raw reply related [flat|nested] 11+ messages in thread
* [PATCH v3 1/6] crypto: testmgr - avoid overlap in chunked tests
2016-12-07 19:53 ` Ard Biesheuvel
@ 2016-12-07 20:23 ` Eric Biggers
0 siblings, 0 replies; 11+ messages in thread
From: Eric Biggers @ 2016-12-07 20:23 UTC (permalink / raw)
To: linux-arm-kernel
On Wed, Dec 07, 2016 at 07:53:51PM +0000, Ard Biesheuvel wrote:
> Does this help at all?
>
> diff --git a/crypto/testmgr.c b/crypto/testmgr.c
> index 670893bcf361..59e67f5b544b 100644
> --- a/crypto/testmgr.c
> +++ b/crypto/testmgr.c
> @@ -63,7 +63,7 @@ int alg_test(const char *driver, const char *alg,
> u32 type, u32 mask)
> */
> #define IDX1 32
> #define IDX2 32400
> -#define IDX3 511
> +#define IDX3 1511
> #define IDX4 8193
> #define IDX5 22222
> #define IDX6 17101
>
Yes, with that change made the self-tests pass.
Eric
^ permalink raw reply [flat|nested] 11+ messages in thread
* [PATCH v3 2/6] crypto: testmgr - add/enhance test cases for CRC-T10DIF
2016-12-05 18:42 [PATCH v3 0/6] crypto: ARM/arm64 CRC-T10DIF/CRC32/CRC32C roundup Ard Biesheuvel
2016-12-05 18:42 ` [PATCH v3 1/6] crypto: testmgr - avoid overlap in chunked tests Ard Biesheuvel
@ 2016-12-05 18:42 ` Ard Biesheuvel
2016-12-05 18:42 ` [PATCH v3 3/6] crypto: arm64/crct10dif - port x86 SSE implementation to arm64 Ard Biesheuvel
` (4 subsequent siblings)
6 siblings, 0 replies; 11+ messages in thread
From: Ard Biesheuvel @ 2016-12-05 18:42 UTC (permalink / raw)
To: linux-arm-kernel
The existing test cases only exercise a small slice of the various
possible code paths through the x86 SSE/PCLMULQDQ implementation,
and the upcoming ports of it for arm64. So add one that exceeds 256
bytes in size, and convert another to a chunked test.
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
---
crypto/testmgr.h | 70 ++++++++++++--------
1 file changed, 42 insertions(+), 28 deletions(-)
diff --git a/crypto/testmgr.h b/crypto/testmgr.h
index e64a4ef9d8ca..9b656be7f52f 100644
--- a/crypto/testmgr.h
+++ b/crypto/testmgr.h
@@ -1334,36 +1334,50 @@ static struct hash_testvec rmd320_tv_template[] = {
}
};
-#define CRCT10DIF_TEST_VECTORS 3
+#define CRCT10DIF_TEST_VECTORS ARRAY_SIZE(crct10dif_tv_template)
static struct hash_testvec crct10dif_tv_template[] = {
{
- .plaintext = "abc",
- .psize = 3,
-#ifdef __LITTLE_ENDIAN
- .digest = "\x3b\x44",
-#else
- .digest = "\x44\x3b",
-#endif
- }, {
- .plaintext = "1234567890123456789012345678901234567890"
- "123456789012345678901234567890123456789",
- .psize = 79,
-#ifdef __LITTLE_ENDIAN
- .digest = "\x70\x4b",
-#else
- .digest = "\x4b\x70",
-#endif
- }, {
- .plaintext =
- "abcddddddddddddddddddddddddddddddddddddddddddddddddddddd",
- .psize = 56,
-#ifdef __LITTLE_ENDIAN
- .digest = "\xe3\x9c",
-#else
- .digest = "\x9c\xe3",
-#endif
- .np = 2,
- .tap = { 28, 28 }
+ .plaintext = "abc",
+ .psize = 3,
+ .digest = (u8 *)(u16 []){ 0x443b },
+ }, {
+ .plaintext = "1234567890123456789012345678901234567890"
+ "123456789012345678901234567890123456789",
+ .psize = 79,
+ .digest = (u8 *)(u16 []){ 0x4b70 },
+ .np = 2,
+ .tap = { 63, 16 },
+ }, {
+ .plaintext = "abcdddddddddddddddddddddddddddddddddddddddd"
+ "ddddddddddddd",
+ .psize = 56,
+ .digest = (u8 *)(u16 []){ 0x9ce3 },
+ .np = 8,
+ .tap = { 1, 2, 28, 7, 6, 5, 4, 3 },
+ }, {
+ .plaintext = "1234567890123456789012345678901234567890"
+ "1234567890123456789012345678901234567890"
+ "1234567890123456789012345678901234567890"
+ "1234567890123456789012345678901234567890"
+ "1234567890123456789012345678901234567890"
+ "1234567890123456789012345678901234567890"
+ "1234567890123456789012345678901234567890"
+ "123456789012345678901234567890123456789",
+ .psize = 319,
+ .digest = (u8 *)(u16 []){ 0x44c6 },
+ }, {
+ .plaintext = "1234567890123456789012345678901234567890"
+ "1234567890123456789012345678901234567890"
+ "1234567890123456789012345678901234567890"
+ "1234567890123456789012345678901234567890"
+ "1234567890123456789012345678901234567890"
+ "1234567890123456789012345678901234567890"
+ "1234567890123456789012345678901234567890"
+ "123456789012345678901234567890123456789",
+ .psize = 319,
+ .digest = (u8 *)(u16 []){ 0x44c6 },
+ .np = 4,
+ .tap = { 1, 255, 57, 6 },
}
};
--
2.7.4
^ permalink raw reply related [flat|nested] 11+ messages in thread* [PATCH v3 3/6] crypto: arm64/crct10dif - port x86 SSE implementation to arm64
2016-12-05 18:42 [PATCH v3 0/6] crypto: ARM/arm64 CRC-T10DIF/CRC32/CRC32C roundup Ard Biesheuvel
2016-12-05 18:42 ` [PATCH v3 1/6] crypto: testmgr - avoid overlap in chunked tests Ard Biesheuvel
2016-12-05 18:42 ` [PATCH v3 2/6] crypto: testmgr - add/enhance test cases for CRC-T10DIF Ard Biesheuvel
@ 2016-12-05 18:42 ` Ard Biesheuvel
2016-12-05 18:42 ` [PATCH v3 4/6] crypto: arm/crct10dif - port x86 SSE implementation to ARM Ard Biesheuvel
` (3 subsequent siblings)
6 siblings, 0 replies; 11+ messages in thread
From: Ard Biesheuvel @ 2016-12-05 18:42 UTC (permalink / raw)
To: linux-arm-kernel
This is a transliteration of the Intel algorithm implemented
using SSE and PCLMULQDQ instructions that resides in the file
arch/x86/crypto/crct10dif-pcl-asm_64.S, but simplified to only
operate on buffers that are 16 byte aligned (but of any size)
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
---
arch/arm64/crypto/Kconfig | 5 +
arch/arm64/crypto/Makefile | 3 +
arch/arm64/crypto/crct10dif-ce-core.S | 392 ++++++++++++++++++++
arch/arm64/crypto/crct10dif-ce-glue.c | 95 +++++
4 files changed, 495 insertions(+)
diff --git a/arch/arm64/crypto/Kconfig b/arch/arm64/crypto/Kconfig
index 2cf32e9887e1..d773c0659202 100644
--- a/arch/arm64/crypto/Kconfig
+++ b/arch/arm64/crypto/Kconfig
@@ -23,6 +23,11 @@ config CRYPTO_GHASH_ARM64_CE
depends on ARM64 && KERNEL_MODE_NEON
select CRYPTO_HASH
+config CRYPTO_CRCT10DIF_ARM64_CE
+ tristate "CRCT10DIF digest algorithm using PMULL instructions"
+ depends on KERNEL_MODE_NEON && CRC_T10DIF
+ select CRYPTO_HASH
+
config CRYPTO_AES_ARM64_CE
tristate "AES core cipher using ARMv8 Crypto Extensions"
depends on ARM64 && KERNEL_MODE_NEON
diff --git a/arch/arm64/crypto/Makefile b/arch/arm64/crypto/Makefile
index abb79b3cfcfe..36fd3eb4201b 100644
--- a/arch/arm64/crypto/Makefile
+++ b/arch/arm64/crypto/Makefile
@@ -17,6 +17,9 @@ sha2-ce-y := sha2-ce-glue.o sha2-ce-core.o
obj-$(CONFIG_CRYPTO_GHASH_ARM64_CE) += ghash-ce.o
ghash-ce-y := ghash-ce-glue.o ghash-ce-core.o
+obj-$(CONFIG_CRYPTO_CRCT10DIF_ARM64_CE) += crct10dif-ce.o
+crct10dif-ce-y := crct10dif-ce-core.o crct10dif-ce-glue.o
+
obj-$(CONFIG_CRYPTO_AES_ARM64_CE) += aes-ce-cipher.o
CFLAGS_aes-ce-cipher.o += -march=armv8-a+crypto
diff --git a/arch/arm64/crypto/crct10dif-ce-core.S b/arch/arm64/crypto/crct10dif-ce-core.S
new file mode 100644
index 000000000000..d5b5a8c038c8
--- /dev/null
+++ b/arch/arm64/crypto/crct10dif-ce-core.S
@@ -0,0 +1,392 @@
+//
+// Accelerated CRC-T10DIF using arm64 NEON and Crypto Extensions instructions
+//
+// Copyright (C) 2016 Linaro Ltd <ard.biesheuvel@linaro.org>
+//
+// This program is free software; you can redistribute it and/or modify
+// it under the terms of the GNU General Public License version 2 as
+// published by the Free Software Foundation.
+//
+
+//
+// Implement fast CRC-T10DIF computation with SSE and PCLMULQDQ instructions
+//
+// Copyright (c) 2013, Intel Corporation
+//
+// Authors:
+// Erdinc Ozturk <erdinc.ozturk@intel.com>
+// Vinodh Gopal <vinodh.gopal@intel.com>
+// James Guilford <james.guilford@intel.com>
+// Tim Chen <tim.c.chen@linux.intel.com>
+//
+// This software is available to you under a choice of one of two
+// licenses. You may choose to be licensed under the terms of the GNU
+// General Public License (GPL) Version 2, available from the file
+// COPYING in the main directory of this source tree, or the
+// OpenIB.org BSD license below:
+//
+// Redistribution and use in source and binary forms, with or without
+// modification, are permitted provided that the following conditions are
+// met:
+//
+// * Redistributions of source code must retain the above copyright
+// notice, this list of conditions and the following disclaimer.
+//
+// * Redistributions in binary form must reproduce the above copyright
+// notice, this list of conditions and the following disclaimer in the
+// documentation and/or other materials provided with the
+// distribution.
+//
+// * Neither the name of the Intel Corporation nor the names of its
+// contributors may be used to endorse or promote products derived from
+// this software without specific prior written permission.
+//
+//
+// THIS SOFTWARE IS PROVIDED BY INTEL CORPORATION ""AS IS"" AND ANY
+// EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+// IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
+// PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL INTEL CORPORATION OR
+// CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
+// EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
+// PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
+// PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
+// LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
+// NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
+// SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+//
+// Function API:
+// UINT16 crc_t10dif_pcl(
+// UINT16 init_crc, //initial CRC value, 16 bits
+// const unsigned char *buf, //buffer pointer to calculate CRC on
+// UINT64 len //buffer length in bytes (64-bit data)
+// );
+//
+// Reference paper titled "Fast CRC Computation for Generic
+// Polynomials Using PCLMULQDQ Instruction"
+// URL: http://www.intel.com/content/dam/www/public/us/en/documents
+// /white-papers/fast-crc-computation-generic-polynomials-pclmulqdq-paper.pdf
+//
+//
+
+#include <linux/linkage.h>
+#include <asm/assembler.h>
+
+ .text
+ .cpu generic+crypto
+
+ arg1_low32 .req w0
+ arg2 .req x1
+ arg3 .req x2
+
+ vzr .req v13
+
+ENTRY(crc_t10dif_pmull)
+ movi vzr.16b, #0 // init zero register
+
+ // adjust the 16-bit initial_crc value, scale it to 32 bits
+ lsl arg1_low32, arg1_low32, #16
+
+ // check if smaller than 256
+ cmp arg3, #256
+
+ // for sizes less than 128, we can't fold 64B at a time...
+ b.lt _less_than_128
+
+ // load the initial crc value
+ // crc value does not need to be byte-reflected, but it needs
+ // to be moved to the high part of the register.
+ // because data will be byte-reflected and will align with
+ // initial crc at correct place.
+ movi v10.16b, #0
+ mov v10.s[3], arg1_low32 // initial crc
+
+ // receive the initial 64B data, xor the initial crc value
+ ldp q0, q1, [arg2]
+ ldp q2, q3, [arg2, #0x20]
+ ldp q4, q5, [arg2, #0x40]
+ ldp q6, q7, [arg2, #0x60]
+ add arg2, arg2, #0x80
+
+CPU_LE( rev64 v0.16b, v0.16b )
+CPU_LE( rev64 v1.16b, v1.16b )
+CPU_LE( rev64 v2.16b, v2.16b )
+CPU_LE( rev64 v3.16b, v3.16b )
+CPU_LE( rev64 v4.16b, v4.16b )
+CPU_LE( rev64 v5.16b, v5.16b )
+CPU_LE( rev64 v6.16b, v6.16b )
+CPU_LE( rev64 v7.16b, v7.16b )
+
+CPU_LE( ext v0.16b, v0.16b, v0.16b, #8 )
+CPU_LE( ext v1.16b, v1.16b, v1.16b, #8 )
+CPU_LE( ext v2.16b, v2.16b, v2.16b, #8 )
+CPU_LE( ext v3.16b, v3.16b, v3.16b, #8 )
+CPU_LE( ext v4.16b, v4.16b, v4.16b, #8 )
+CPU_LE( ext v5.16b, v5.16b, v5.16b, #8 )
+CPU_LE( ext v6.16b, v6.16b, v6.16b, #8 )
+CPU_LE( ext v7.16b, v7.16b, v7.16b, #8 )
+
+ // XOR the initial_crc value
+ eor v0.16b, v0.16b, v10.16b
+
+ ldr q10, rk3 // xmm10 has rk3 and rk4
+ // type of pmull instruction
+ // will determine which constant to use
+
+ //
+ // we subtract 256 instead of 128 to save one instruction from the loop
+ //
+ sub arg3, arg3, #256
+
+ // at this section of the code, there is 64*x+y (0<=y<64) bytes of
+ // buffer. The _fold_64_B_loop will fold 64B at a time
+ // until we have 64+y Bytes of buffer
+
+
+ // fold 64B at a time. This section of the code folds 4 vector
+ // registers in parallel
+_fold_64_B_loop:
+
+ .macro fold64, reg1, reg2
+ ldp q11, q12, [arg2], #0x20
+
+ pmull2 v8.1q, \reg1\().2d, v10.2d
+ pmull \reg1\().1q, \reg1\().1d, v10.1d
+
+CPU_LE( rev64 v11.16b, v11.16b )
+CPU_LE( rev64 v12.16b, v12.16b )
+
+ pmull2 v9.1q, \reg2\().2d, v10.2d
+ pmull \reg2\().1q, \reg2\().1d, v10.1d
+
+CPU_LE( ext v11.16b, v11.16b, v11.16b, #8 )
+CPU_LE( ext v12.16b, v12.16b, v12.16b, #8 )
+
+ eor \reg1\().16b, \reg1\().16b, v8.16b
+ eor \reg2\().16b, \reg2\().16b, v9.16b
+ eor \reg1\().16b, \reg1\().16b, v11.16b
+ eor \reg2\().16b, \reg2\().16b, v12.16b
+ .endm
+
+ fold64 v0, v1
+ fold64 v2, v3
+ fold64 v4, v5
+ fold64 v6, v7
+
+ subs arg3, arg3, #128
+
+ // check if there is another 64B in the buffer to be able to fold
+ b.ge _fold_64_B_loop
+
+ // at this point, the buffer pointer is pointing at the last y Bytes
+ // of the buffer the 64B of folded data is in 4 of the vector
+ // registers: v0, v1, v2, v3
+
+ // fold the 8 vector registers to 1 vector register with different
+ // constants
+
+ ldr q10, rk9
+
+ .macro fold16, reg, rk
+ pmull v8.1q, \reg\().1d, v10.1d
+ pmull2 \reg\().1q, \reg\().2d, v10.2d
+ .ifnb \rk
+ ldr q10, \rk
+ .endif
+ eor v7.16b, v7.16b, v8.16b
+ eor v7.16b, v7.16b, \reg\().16b
+ .endm
+
+ fold16 v0, rk11
+ fold16 v1, rk13
+ fold16 v2, rk15
+ fold16 v3, rk17
+ fold16 v4, rk19
+ fold16 v5, rk1
+ fold16 v6
+
+ // instead of 64, we add 48 to the loop counter to save 1 instruction
+ // from the loop instead of a cmp instruction, we use the negative
+ // flag with the jl instruction
+ adds arg3, arg3, #(128-16)
+ b.lt _final_reduction_for_128
+
+ // now we have 16+y bytes left to reduce. 16 Bytes is in register v7
+ // and the rest is in memory. We can fold 16 bytes@a time if y>=16
+ // continue folding 16B at a time
+
+_16B_reduction_loop:
+ pmull v8.1q, v7.1d, v10.1d
+ pmull2 v7.1q, v7.2d, v10.2d
+ eor v7.16b, v7.16b, v8.16b
+
+ ldr q0, [arg2], #16
+CPU_LE( rev64 v0.16b, v0.16b )
+CPU_LE( ext v0.16b, v0.16b, v0.16b, #8 )
+ eor v7.16b, v7.16b, v0.16b
+ subs arg3, arg3, #16
+
+ // instead of a cmp instruction, we utilize the flags with the
+ // jge instruction equivalent of: cmp arg3, 16-16
+ // check if there is any more 16B in the buffer to be able to fold
+ b.ge _16B_reduction_loop
+
+ // now we have 16+z bytes left to reduce, where 0<= z < 16.
+ // first, we reduce the data in the xmm7 register
+
+_final_reduction_for_128:
+ // check if any more data to fold. If not, compute the CRC of
+ // the final 128 bits
+ adds arg3, arg3, #16
+ b.eq _128_done
+
+ // here we are getting data that is less than 16 bytes.
+ // since we know that there was data before the pointer, we can
+ // offset the input pointer before the actual point, to receive
+ // exactly 16 bytes. after that the registers need to be adjusted.
+_get_last_two_regs:
+ add arg2, arg2, arg3
+ ldr q1, [arg2, #-16]
+CPU_LE( rev64 v1.16b, v1.16b )
+CPU_LE( ext v1.16b, v1.16b, v1.16b, #8 )
+
+ // get rid of the extra data that was loaded before
+ // load the shift constant
+ adr x4, tbl_shf_table + 16
+ sub x4, x4, arg3
+ ld1 {v0.16b}, [x4]
+
+ // shift v2 to the left by arg3 bytes
+ tbl v2.16b, {v7.16b}, v0.16b
+
+ // shift v7 to the right by 16-arg3 bytes
+ movi v9.16b, #0x80
+ eor v0.16b, v0.16b, v9.16b
+ tbl v7.16b, {v7.16b}, v0.16b
+
+ // blend
+ sshr v0.16b, v0.16b, #7 // convert to 8-bit mask
+ bsl v0.16b, v2.16b, v1.16b
+
+ // fold 16 Bytes
+ pmull v8.1q, v7.1d, v10.1d
+ pmull2 v7.1q, v7.2d, v10.2d
+ eor v7.16b, v7.16b, v8.16b
+ eor v7.16b, v7.16b, v0.16b
+
+_128_done:
+ // compute crc of a 128-bit value
+ ldr q10, rk5 // rk5 and rk6 in xmm10
+
+ // 64b fold
+ ext v0.16b, vzr.16b, v7.16b, #8
+ mov v7.d[0], v7.d[1]
+ pmull v7.1q, v7.1d, v10.1d
+ eor v7.16b, v7.16b, v0.16b
+
+ // 32b fold
+ ext v0.16b, v7.16b, vzr.16b, #4
+ mov v7.s[3], vzr.s[0]
+ pmull2 v0.1q, v0.2d, v10.2d
+ eor v7.16b, v7.16b, v0.16b
+
+ // barrett reduction
+_barrett:
+ ldr q10, rk7
+ mov v0.d[0], v7.d[1]
+
+ pmull v0.1q, v0.1d, v10.1d
+ ext v0.16b, vzr.16b, v0.16b, #12
+ pmull2 v0.1q, v0.2d, v10.2d
+ ext v0.16b, vzr.16b, v0.16b, #12
+ eor v7.16b, v7.16b, v0.16b
+ mov w0, v7.s[1]
+
+_cleanup:
+ // scale the result back to 16 bits
+ lsr x0, x0, #16
+ ret
+
+_less_than_128:
+ cbz arg3, _cleanup
+
+ movi v0.16b, #0
+ mov v0.s[3], arg1_low32 // get the initial crc value
+
+ ldr q7, [arg2], #0x10
+CPU_LE( rev64 v7.16b, v7.16b )
+CPU_LE( ext v7.16b, v7.16b, v7.16b, #8 )
+ eor v7.16b, v7.16b, v0.16b // xor the initial crc value
+
+ cmp arg3, #16
+ b.eq _128_done // exactly 16 left
+ b.lt _less_than_16_left
+
+ ldr q10, rk1 // rk1 and rk2 in xmm10
+
+ // update the counter. subtract 32 instead of 16 to save one
+ // instruction from the loop
+ subs arg3, arg3, #32
+ b.ge _16B_reduction_loop
+
+ add arg3, arg3, #16
+ b _get_last_two_regs
+
+_less_than_16_left:
+ // shl r9, 4
+ adr x0, tbl_shf_table + 16
+ sub x0, x0, arg3
+ ld1 {v0.16b}, [x0]
+ movi v9.16b, #0x80
+ eor v0.16b, v0.16b, v9.16b
+ tbl v7.16b, {v7.16b}, v0.16b
+ b _128_done
+ENDPROC(crc_t10dif_pmull)
+
+// precomputed constants
+// these constants are precomputed from the poly:
+// 0x8bb70000 (0x8bb7 scaled to 32 bits)
+ .align 4
+// Q = 0x18BB70000
+// rk1 = 2^(32*3) mod Q << 32
+// rk2 = 2^(32*5) mod Q << 32
+// rk3 = 2^(32*15) mod Q << 32
+// rk4 = 2^(32*17) mod Q << 32
+// rk5 = 2^(32*3) mod Q << 32
+// rk6 = 2^(32*2) mod Q << 32
+// rk7 = floor(2^64/Q)
+// rk8 = Q
+
+rk1: .octa 0x06df0000000000002d56000000000000
+rk3: .octa 0x7cf50000000000009d9d000000000000
+rk5: .octa 0x13680000000000002d56000000000000
+rk7: .octa 0x000000018bb7000000000001f65a57f8
+rk9: .octa 0xbfd6000000000000ceae000000000000
+rk11: .octa 0x713c0000000000001e16000000000000
+rk13: .octa 0x80a6000000000000f7f9000000000000
+rk15: .octa 0xe658000000000000044c000000000000
+rk17: .octa 0xa497000000000000ad18000000000000
+rk19: .octa 0xe7b50000000000006ee3000000000000
+
+tbl_shf_table:
+// use these values for shift constants for the tbl/tbx instruction
+// different alignments result in values as shown:
+// DDQ 0x008f8e8d8c8b8a898887868584838281 # shl 15 (16-1) / shr1
+// DDQ 0x01008f8e8d8c8b8a8988878685848382 # shl 14 (16-3) / shr2
+// DDQ 0x0201008f8e8d8c8b8a89888786858483 # shl 13 (16-4) / shr3
+// DDQ 0x030201008f8e8d8c8b8a898887868584 # shl 12 (16-4) / shr4
+// DDQ 0x04030201008f8e8d8c8b8a8988878685 # shl 11 (16-5) / shr5
+// DDQ 0x0504030201008f8e8d8c8b8a89888786 # shl 10 (16-6) / shr6
+// DDQ 0x060504030201008f8e8d8c8b8a898887 # shl 9 (16-7) / shr7
+// DDQ 0x07060504030201008f8e8d8c8b8a8988 # shl 8 (16-8) / shr8
+// DDQ 0x0807060504030201008f8e8d8c8b8a89 # shl 7 (16-9) / shr9
+// DDQ 0x090807060504030201008f8e8d8c8b8a # shl 6 (16-10) / shr10
+// DDQ 0x0a090807060504030201008f8e8d8c8b # shl 5 (16-11) / shr11
+// DDQ 0x0b0a090807060504030201008f8e8d8c # shl 4 (16-12) / shr12
+// DDQ 0x0c0b0a090807060504030201008f8e8d # shl 3 (16-13) / shr13
+// DDQ 0x0d0c0b0a090807060504030201008f8e # shl 2 (16-14) / shr14
+// DDQ 0x0e0d0c0b0a090807060504030201008f # shl 1 (16-15) / shr15
+
+ .byte 0x0, 0x81, 0x82, 0x83, 0x84, 0x85, 0x86, 0x87
+ .byte 0x88, 0x89, 0x8a, 0x8b, 0x8c, 0x8d, 0x8e, 0x8f
+ .byte 0x0, 0x1, 0x2, 0x3, 0x4, 0x5, 0x6, 0x7
+ .byte 0x8, 0x9, 0xa, 0xb, 0xc, 0xd, 0xe , 0x0
diff --git a/arch/arm64/crypto/crct10dif-ce-glue.c b/arch/arm64/crypto/crct10dif-ce-glue.c
new file mode 100644
index 000000000000..60cb590c2590
--- /dev/null
+++ b/arch/arm64/crypto/crct10dif-ce-glue.c
@@ -0,0 +1,95 @@
+/*
+ * Accelerated CRC-T10DIF using arm64 NEON and Crypto Extensions instructions
+ *
+ * Copyright (C) 2016 Linaro Ltd <ard.biesheuvel@linaro.org>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/cpufeature.h>
+#include <linux/crc-t10dif.h>
+#include <linux/init.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/string.h>
+
+#include <crypto/internal/hash.h>
+
+#include <asm/neon.h>
+
+#define CRC_T10DIF_PMULL_CHUNK_SIZE 16U
+
+asmlinkage u16 crc_t10dif_pmull(u16 init_crc, const u8 buf[], u64 len);
+
+static int crct10dif_init(struct shash_desc *desc)
+{
+ u16 *crc = shash_desc_ctx(desc);
+
+ *crc = 0;
+ return 0;
+}
+
+static int crct10dif_update(struct shash_desc *desc, const u8 *data,
+ unsigned int length)
+{
+ u16 *crc = shash_desc_ctx(desc);
+ unsigned int l;
+
+ if (unlikely((u64)data % CRC_T10DIF_PMULL_CHUNK_SIZE)) {
+ l = min_t(u32, length, CRC_T10DIF_PMULL_CHUNK_SIZE -
+ ((u64)data % CRC_T10DIF_PMULL_CHUNK_SIZE));
+
+ *crc = crc_t10dif_generic(*crc, data, l);
+
+ length -= l;
+ data += l;
+ }
+
+ if (length > 0) {
+ kernel_neon_begin_partial(14);
+ *crc = crc_t10dif_pmull(*crc, data, length);
+ kernel_neon_end();
+ }
+
+ return 0;
+}
+
+static int crct10dif_final(struct shash_desc *desc, u8 *out)
+{
+ u16 *crc = shash_desc_ctx(desc);
+
+ *(u16 *)out = *crc;
+ return 0;
+}
+
+static struct shash_alg crc_t10dif_alg = {
+ .digestsize = CRC_T10DIF_DIGEST_SIZE,
+ .init = crct10dif_init,
+ .update = crct10dif_update,
+ .final = crct10dif_final,
+ .descsize = CRC_T10DIF_DIGEST_SIZE,
+
+ .base.cra_name = "crct10dif",
+ .base.cra_driver_name = "crct10dif-arm64-ce",
+ .base.cra_priority = 200,
+ .base.cra_blocksize = CRC_T10DIF_BLOCK_SIZE,
+ .base.cra_module = THIS_MODULE,
+};
+
+static int __init crc_t10dif_mod_init(void)
+{
+ return crypto_register_shash(&crc_t10dif_alg);
+}
+
+static void __exit crc_t10dif_mod_exit(void)
+{
+ crypto_unregister_shash(&crc_t10dif_alg);
+}
+
+module_cpu_feature_match(PMULL, crc_t10dif_mod_init);
+module_exit(crc_t10dif_mod_exit);
+
+MODULE_AUTHOR("Ard Biesheuvel <ard.biesheuvel@linaro.org>");
+MODULE_LICENSE("GPL v2");
--
2.7.4
^ permalink raw reply related [flat|nested] 11+ messages in thread* [PATCH v3 4/6] crypto: arm/crct10dif - port x86 SSE implementation to ARM
2016-12-05 18:42 [PATCH v3 0/6] crypto: ARM/arm64 CRC-T10DIF/CRC32/CRC32C roundup Ard Biesheuvel
` (2 preceding siblings ...)
2016-12-05 18:42 ` [PATCH v3 3/6] crypto: arm64/crct10dif - port x86 SSE implementation to arm64 Ard Biesheuvel
@ 2016-12-05 18:42 ` Ard Biesheuvel
2016-12-05 18:42 ` [PATCH v3 5/6] crypto: arm64/crc32 - accelerated support based on x86 SSE implementation Ard Biesheuvel
` (2 subsequent siblings)
6 siblings, 0 replies; 11+ messages in thread
From: Ard Biesheuvel @ 2016-12-05 18:42 UTC (permalink / raw)
To: linux-arm-kernel
This is a transliteration of the Intel algorithm implemented
using SSE and PCLMULQDQ instructions that resides in the file
arch/x86/crypto/crct10dif-pcl-asm_64.S, but simplified to only
operate on buffers that are 16 byte aligned (but of any size)
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
---
arch/arm/crypto/Kconfig | 5 +
arch/arm/crypto/Makefile | 2 +
arch/arm/crypto/crct10dif-ce-core.S | 427 ++++++++++++++++++++
arch/arm/crypto/crct10dif-ce-glue.c | 101 +++++
4 files changed, 535 insertions(+)
diff --git a/arch/arm/crypto/Kconfig b/arch/arm/crypto/Kconfig
index 27ed1b1cd1d7..fce801fa52a1 100644
--- a/arch/arm/crypto/Kconfig
+++ b/arch/arm/crypto/Kconfig
@@ -120,4 +120,9 @@ config CRYPTO_GHASH_ARM_CE
that uses the 64x64 to 128 bit polynomial multiplication (vmull.p64)
that is part of the ARMv8 Crypto Extensions
+config CRYPTO_CRCT10DIF_ARM_CE
+ tristate "CRCT10DIF digest algorithm using PMULL instructions"
+ depends on KERNEL_MODE_NEON && CRC_T10DIF
+ select CRYPTO_HASH
+
endif
diff --git a/arch/arm/crypto/Makefile b/arch/arm/crypto/Makefile
index fc5150702b64..fc77265014b7 100644
--- a/arch/arm/crypto/Makefile
+++ b/arch/arm/crypto/Makefile
@@ -13,6 +13,7 @@ ce-obj-$(CONFIG_CRYPTO_AES_ARM_CE) += aes-arm-ce.o
ce-obj-$(CONFIG_CRYPTO_SHA1_ARM_CE) += sha1-arm-ce.o
ce-obj-$(CONFIG_CRYPTO_SHA2_ARM_CE) += sha2-arm-ce.o
ce-obj-$(CONFIG_CRYPTO_GHASH_ARM_CE) += ghash-arm-ce.o
+ce-obj-$(CONFIG_CRYPTO_CRCT10DIF_ARM_CE) += crct10dif-arm-ce.o
ifneq ($(ce-obj-y)$(ce-obj-m),)
ifeq ($(call as-instr,.fpu crypto-neon-fp-armv8,y,n),y)
@@ -36,6 +37,7 @@ sha1-arm-ce-y := sha1-ce-core.o sha1-ce-glue.o
sha2-arm-ce-y := sha2-ce-core.o sha2-ce-glue.o
aes-arm-ce-y := aes-ce-core.o aes-ce-glue.o
ghash-arm-ce-y := ghash-ce-core.o ghash-ce-glue.o
+crct10dif-arm-ce-y := crct10dif-ce-core.o crct10dif-ce-glue.o
quiet_cmd_perl = PERL $@
cmd_perl = $(PERL) $(<) > $(@)
diff --git a/arch/arm/crypto/crct10dif-ce-core.S b/arch/arm/crypto/crct10dif-ce-core.S
new file mode 100644
index 000000000000..ce45ba0c0687
--- /dev/null
+++ b/arch/arm/crypto/crct10dif-ce-core.S
@@ -0,0 +1,427 @@
+//
+// Accelerated CRC-T10DIF using ARM NEON and Crypto Extensions instructions
+//
+// Copyright (C) 2016 Linaro Ltd <ard.biesheuvel@linaro.org>
+//
+// This program is free software; you can redistribute it and/or modify
+// it under the terms of the GNU General Public License version 2 as
+// published by the Free Software Foundation.
+//
+
+//
+// Implement fast CRC-T10DIF computation with SSE and PCLMULQDQ instructions
+//
+// Copyright (c) 2013, Intel Corporation
+//
+// Authors:
+// Erdinc Ozturk <erdinc.ozturk@intel.com>
+// Vinodh Gopal <vinodh.gopal@intel.com>
+// James Guilford <james.guilford@intel.com>
+// Tim Chen <tim.c.chen@linux.intel.com>
+//
+// This software is available to you under a choice of one of two
+// licenses. You may choose to be licensed under the terms of the GNU
+// General Public License (GPL) Version 2, available from the file
+// COPYING in the main directory of this source tree, or the
+// OpenIB.org BSD license below:
+//
+// Redistribution and use in source and binary forms, with or without
+// modification, are permitted provided that the following conditions are
+// met:
+//
+// * Redistributions of source code must retain the above copyright
+// notice, this list of conditions and the following disclaimer.
+//
+// * Redistributions in binary form must reproduce the above copyright
+// notice, this list of conditions and the following disclaimer in the
+// documentation and/or other materials provided with the
+// distribution.
+//
+// * Neither the name of the Intel Corporation nor the names of its
+// contributors may be used to endorse or promote products derived from
+// this software without specific prior written permission.
+//
+//
+// THIS SOFTWARE IS PROVIDED BY INTEL CORPORATION ""AS IS"" AND ANY
+// EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+// IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
+// PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL INTEL CORPORATION OR
+// CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
+// EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
+// PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
+// PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
+// LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
+// NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
+// SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+//
+// Function API:
+// UINT16 crc_t10dif_pcl(
+// UINT16 init_crc, //initial CRC value, 16 bits
+// const unsigned char *buf, //buffer pointer to calculate CRC on
+// UINT64 len //buffer length in bytes (64-bit data)
+// );
+//
+// Reference paper titled "Fast CRC Computation for Generic
+// Polynomials Using PCLMULQDQ Instruction"
+// URL: http://www.intel.com/content/dam/www/public/us/en/documents
+// /white-papers/fast-crc-computation-generic-polynomials-pclmulqdq-paper.pdf
+//
+//
+
+#include <linux/linkage.h>
+#include <asm/assembler.h>
+
+#ifdef CONFIG_CPU_ENDIAN_BE8
+#define CPU_LE(code...)
+#else
+#define CPU_LE(code...) code
+#endif
+
+ .text
+ .fpu crypto-neon-fp-armv8
+
+ arg1_low32 .req r0
+ arg2 .req r1
+ arg3 .req r2
+
+ qzr .req q13
+
+ q0l .req d0
+ q0h .req d1
+ q1l .req d2
+ q1h .req d3
+ q2l .req d4
+ q2h .req d5
+ q3l .req d6
+ q3h .req d7
+ q4l .req d8
+ q4h .req d9
+ q5l .req d10
+ q5h .req d11
+ q6l .req d12
+ q6h .req d13
+ q7l .req d14
+ q7h .req d15
+
+ENTRY(crc_t10dif_pmull)
+ vmov.i8 qzr, #0 // init zero register
+
+ // adjust the 16-bit initial_crc value, scale it to 32 bits
+ lsl arg1_low32, arg1_low32, #16
+
+ // check if smaller than 256
+ cmp arg3, #256
+
+ // for sizes less than 128, we can't fold 64B at a time...
+ blt _less_than_128
+
+ // load the initial crc value
+ // crc value does not need to be byte-reflected, but it needs
+ // to be moved to the high part of the register.
+ // because data will be byte-reflected and will align with
+ // initial crc at correct place.
+ vmov s0, arg1_low32 // initial crc
+ vext.8 q10, qzr, q0, #4
+
+ // receive the initial 64B data, xor the initial crc value
+ vld1.64 {q0-q1}, [arg2, :128]!
+ vld1.64 {q2-q3}, [arg2, :128]!
+ vld1.64 {q4-q5}, [arg2, :128]!
+ vld1.64 {q6-q7}, [arg2, :128]!
+CPU_LE( vrev64.8 q0, q0 )
+CPU_LE( vrev64.8 q1, q1 )
+CPU_LE( vrev64.8 q2, q2 )
+CPU_LE( vrev64.8 q3, q3 )
+CPU_LE( vrev64.8 q4, q4 )
+CPU_LE( vrev64.8 q5, q5 )
+CPU_LE( vrev64.8 q6, q6 )
+CPU_LE( vrev64.8 q7, q7 )
+
+ vswp d0, d1
+ vswp d2, d3
+ vswp d4, d5
+ vswp d6, d7
+ vswp d8, d9
+ vswp d10, d11
+ vswp d12, d13
+ vswp d14, d15
+
+ // XOR the initial_crc value
+ veor.8 q0, q0, q10
+
+ adr ip, rk3
+ vld1.64 {q10}, [ip, :128] // xmm10 has rk3 and rk4
+
+ //
+ // we subtract 256 instead of 128 to save one instruction from the loop
+ //
+ sub arg3, arg3, #256
+
+ // at this section of the code, there is 64*x+y (0<=y<64) bytes of
+ // buffer. The _fold_64_B_loop will fold 64B at a time
+ // until we have 64+y Bytes of buffer
+
+
+ // fold 64B at a time. This section of the code folds 4 vector
+ // registers in parallel
+_fold_64_B_loop:
+
+ .macro fold64, reg1, reg2
+ vld1.64 {q11-q12}, [arg2, :128]!
+
+ vmull.p64 q8, \reg1\()h, d21
+ vmull.p64 \reg1, \reg1\()l, d20
+ vmull.p64 q9, \reg2\()h, d21
+ vmull.p64 \reg2, \reg2\()l, d20
+
+CPU_LE( vrev64.8 q11, q11 )
+CPU_LE( vrev64.8 q12, q12 )
+ vswp d22, d23
+ vswp d24, d25
+
+ veor.8 \reg1, \reg1, q8
+ veor.8 \reg2, \reg2, q9
+ veor.8 \reg1, \reg1, q11
+ veor.8 \reg2, \reg2, q12
+ .endm
+
+ fold64 q0, q1
+ fold64 q2, q3
+ fold64 q4, q5
+ fold64 q6, q7
+
+ subs arg3, arg3, #128
+
+ // check if there is another 64B in the buffer to be able to fold
+ bge _fold_64_B_loop
+
+ // at this point, the buffer pointer is pointing at the last y Bytes
+ // of the buffer the 64B of folded data is in 4 of the vector
+ // registers: v0, v1, v2, v3
+
+ // fold the 8 vector registers to 1 vector register with different
+ // constants
+
+ adr ip, rk9
+ vld1.64 {q10}, [ip, :128]!
+
+ .macro fold16, reg, rk
+ vmull.p64 q8, \reg\()l, d20
+ vmull.p64 \reg, \reg\()h, d21
+ .ifnb \rk
+ vld1.64 {q10}, [ip, :128]!
+ .endif
+ veor.8 q7, q7, q8
+ veor.8 q7, q7, \reg
+ .endm
+
+ fold16 q0, rk11
+ fold16 q1, rk13
+ fold16 q2, rk15
+ fold16 q3, rk17
+ fold16 q4, rk19
+ fold16 q5, rk1
+ fold16 q6
+
+ // instead of 64, we add 48 to the loop counter to save 1 instruction
+ // from the loop instead of a cmp instruction, we use the negative
+ // flag with the jl instruction
+ adds arg3, arg3, #(128-16)
+ blt _final_reduction_for_128
+
+ // now we have 16+y bytes left to reduce. 16 Bytes is in register v7
+ // and the rest is in memory. We can fold 16 bytes@a time if y>=16
+ // continue folding 16B at a time
+
+_16B_reduction_loop:
+ vmull.p64 q8, d14, d20
+ vmull.p64 q7, d15, d21
+ veor.8 q7, q7, q8
+
+ vld1.64 {q0}, [arg2, :128]!
+CPU_LE( vrev64.8 q0, q0 )
+ vswp d0, d1
+ veor.8 q7, q7, q0
+ subs arg3, arg3, #16
+
+ // instead of a cmp instruction, we utilize the flags with the
+ // jge instruction equivalent of: cmp arg3, 16-16
+ // check if there is any more 16B in the buffer to be able to fold
+ bge _16B_reduction_loop
+
+ // now we have 16+z bytes left to reduce, where 0<= z < 16.
+ // first, we reduce the data in the xmm7 register
+
+_final_reduction_for_128:
+ // check if any more data to fold. If not, compute the CRC of
+ // the final 128 bits
+ adds arg3, arg3, #16
+ beq _128_done
+
+ // here we are getting data that is less than 16 bytes.
+ // since we know that there was data before the pointer, we can
+ // offset the input pointer before the actual point, to receive
+ // exactly 16 bytes. after that the registers need to be adjusted.
+_get_last_two_regs:
+ add arg2, arg2, arg3
+ sub arg2, arg2, #16
+ vld1.64 {q1}, [arg2]
+CPU_LE( vrev64.8 q1, q1 )
+ vswp d2, d3
+
+ // get rid of the extra data that was loaded before
+ // load the shift constant
+ adr ip, tbl_shf_table + 16
+ sub ip, ip, arg3
+ vld1.8 {q0}, [ip]
+
+ // shift v2 to the left by arg3 bytes
+ vtbl.8 d4, {d14-d15}, d0
+ vtbl.8 d5, {d14-d15}, d1
+
+ // shift v7 to the right by 16-arg3 bytes
+ vmov.i8 q9, #0x80
+ veor.8 q0, q0, q9
+ vtbl.8 d18, {d14-d15}, d0
+ vtbl.8 d19, {d14-d15}, d1
+
+ // blend
+ vshr.s8 q0, q0, #7 // convert to 8-bit mask
+ vbsl.8 q0, q2, q1
+
+ // fold 16 Bytes
+ vmull.p64 q8, d18, d20
+ vmull.p64 q7, d19, d21
+ veor.8 q7, q7, q8
+ veor.8 q7, q7, q0
+
+_128_done:
+ // compute crc of a 128-bit value
+ vldr d20, rk5
+ vldr d21, rk6 // rk5 and rk6 in xmm10
+
+ // 64b fold
+ vext.8 q0, qzr, q7, #8
+ vmull.p64 q7, d15, d20
+ veor.8 q7, q7, q0
+
+ // 32b fold
+ vext.8 q0, q7, qzr, #12
+ vmov s31, s3
+ vmull.p64 q0, d0, d21
+ veor.8 q7, q0, q7
+
+ // barrett reduction
+_barrett:
+ vldr d20, rk7
+ vldr d21, rk8
+
+ vmull.p64 q0, d15, d20
+ vext.8 q0, qzr, q0, #12
+ vmull.p64 q0, d1, d21
+ vext.8 q0, qzr, q0, #12
+ veor.8 q7, q7, q0
+ vmov r0, s29
+
+_cleanup:
+ // scale the result back to 16 bits
+ lsr r0, r0, #16
+ bx lr
+
+_less_than_128:
+ teq arg3, #0
+ beq _cleanup
+
+ vmov.i8 q0, #0
+ vmov s3, arg1_low32 // get the initial crc value
+
+ vld1.64 {q7}, [arg2, :128]!
+CPU_LE( vrev64.8 q7, q7 )
+ vswp d14, d15
+ veor.8 q7, q7, q0
+
+ cmp arg3, #16
+ beq _128_done // exactly 16 left
+ blt _less_than_16_left
+
+ // now if there is, load the constants
+ vldr d20, rk1
+ vldr d21, rk2 // rk1 and rk2 in xmm10
+
+ // check if there is enough buffer to be able to fold 16B at a time
+ subs arg3, arg3, #32
+ addlt arg3, arg3, #16
+ blt _get_last_two_regs
+ b _16B_reduction_loop
+
+_less_than_16_left:
+ // shl r9, 4
+ adr ip, tbl_shf_table + 16
+ sub ip, ip, arg3
+ vld1.8 {q0}, [ip]
+ vmov.i8 q9, #0x80
+ veor.8 q0, q0, q9
+ vtbl.8 d18, {d14-d15}, d0
+ vtbl.8 d15, {d14-d15}, d1
+ vmov d14, d18
+ b _128_done
+ENDPROC(crc_t10dif_pmull)
+
+// precomputed constants
+// these constants are precomputed from the poly:
+// 0x8bb70000 (0x8bb7 scaled to 32 bits)
+ .align 4
+// Q = 0x18BB70000
+// rk1 = 2^(32*3) mod Q << 32
+// rk2 = 2^(32*5) mod Q << 32
+// rk3 = 2^(32*15) mod Q << 32
+// rk4 = 2^(32*17) mod Q << 32
+// rk5 = 2^(32*3) mod Q << 32
+// rk6 = 2^(32*2) mod Q << 32
+// rk7 = floor(2^64/Q)
+// rk8 = Q
+
+rk3: .quad 0x9d9d000000000000
+rk4: .quad 0x7cf5000000000000
+rk5: .quad 0x2d56000000000000
+rk6: .quad 0x1368000000000000
+rk7: .quad 0x00000001f65a57f8
+rk8: .quad 0x000000018bb70000
+rk9: .quad 0xceae000000000000
+rk10: .quad 0xbfd6000000000000
+rk11: .quad 0x1e16000000000000
+rk12: .quad 0x713c000000000000
+rk13: .quad 0xf7f9000000000000
+rk14: .quad 0x80a6000000000000
+rk15: .quad 0x044c000000000000
+rk16: .quad 0xe658000000000000
+rk17: .quad 0xad18000000000000
+rk18: .quad 0xa497000000000000
+rk19: .quad 0x6ee3000000000000
+rk20: .quad 0xe7b5000000000000
+rk1: .quad 0x2d56000000000000
+rk2: .quad 0x06df000000000000
+
+tbl_shf_table:
+// use these values for shift constants for the tbl/tbx instruction
+// different alignments result in values as shown:
+// DDQ 0x008f8e8d8c8b8a898887868584838281 # shl 15 (16-1) / shr1
+// DDQ 0x01008f8e8d8c8b8a8988878685848382 # shl 14 (16-3) / shr2
+// DDQ 0x0201008f8e8d8c8b8a89888786858483 # shl 13 (16-4) / shr3
+// DDQ 0x030201008f8e8d8c8b8a898887868584 # shl 12 (16-4) / shr4
+// DDQ 0x04030201008f8e8d8c8b8a8988878685 # shl 11 (16-5) / shr5
+// DDQ 0x0504030201008f8e8d8c8b8a89888786 # shl 10 (16-6) / shr6
+// DDQ 0x060504030201008f8e8d8c8b8a898887 # shl 9 (16-7) / shr7
+// DDQ 0x07060504030201008f8e8d8c8b8a8988 # shl 8 (16-8) / shr8
+// DDQ 0x0807060504030201008f8e8d8c8b8a89 # shl 7 (16-9) / shr9
+// DDQ 0x090807060504030201008f8e8d8c8b8a # shl 6 (16-10) / shr10
+// DDQ 0x0a090807060504030201008f8e8d8c8b # shl 5 (16-11) / shr11
+// DDQ 0x0b0a090807060504030201008f8e8d8c # shl 4 (16-12) / shr12
+// DDQ 0x0c0b0a090807060504030201008f8e8d # shl 3 (16-13) / shr13
+// DDQ 0x0d0c0b0a090807060504030201008f8e # shl 2 (16-14) / shr14
+// DDQ 0x0e0d0c0b0a090807060504030201008f # shl 1 (16-15) / shr15
+
+ .byte 0x0, 0x81, 0x82, 0x83, 0x84, 0x85, 0x86, 0x87
+ .byte 0x88, 0x89, 0x8a, 0x8b, 0x8c, 0x8d, 0x8e, 0x8f
+ .byte 0x0, 0x1, 0x2, 0x3, 0x4, 0x5, 0x6, 0x7
+ .byte 0x8, 0x9, 0xa, 0xb, 0xc, 0xd, 0xe , 0x0
diff --git a/arch/arm/crypto/crct10dif-ce-glue.c b/arch/arm/crypto/crct10dif-ce-glue.c
new file mode 100644
index 000000000000..d428355cf38d
--- /dev/null
+++ b/arch/arm/crypto/crct10dif-ce-glue.c
@@ -0,0 +1,101 @@
+/*
+ * Accelerated CRC-T10DIF using ARM NEON and Crypto Extensions instructions
+ *
+ * Copyright (C) 2016 Linaro Ltd <ard.biesheuvel@linaro.org>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/crc-t10dif.h>
+#include <linux/init.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/string.h>
+
+#include <crypto/internal/hash.h>
+
+#include <asm/neon.h>
+#include <asm/simd.h>
+
+#define CRC_T10DIF_PMULL_CHUNK_SIZE 16U
+
+asmlinkage u16 crc_t10dif_pmull(u16 init_crc, const u8 buf[], u32 len);
+
+static int crct10dif_init(struct shash_desc *desc)
+{
+ u16 *crc = shash_desc_ctx(desc);
+
+ *crc = 0;
+ return 0;
+}
+
+static int crct10dif_update(struct shash_desc *desc, const u8 *data,
+ unsigned int length)
+{
+ u16 *crc = shash_desc_ctx(desc);
+ unsigned int l;
+
+ if (!may_use_simd()) {
+ *crc = crc_t10dif_generic(*crc, data, length);
+ } else {
+ if (unlikely((u32)data % CRC_T10DIF_PMULL_CHUNK_SIZE)) {
+ l = min_t(u32, length, CRC_T10DIF_PMULL_CHUNK_SIZE -
+ ((u32)data % CRC_T10DIF_PMULL_CHUNK_SIZE));
+
+ *crc = crc_t10dif_generic(*crc, data, l);
+
+ length -= l;
+ data += l;
+ }
+ if (length > 0) {
+ kernel_neon_begin();
+ *crc = crc_t10dif_pmull(*crc, data, length);
+ kernel_neon_end();
+ }
+ }
+ return 0;
+}
+
+static int crct10dif_final(struct shash_desc *desc, u8 *out)
+{
+ u16 *crc = shash_desc_ctx(desc);
+
+ *(u16 *)out = *crc;
+ return 0;
+}
+
+static struct shash_alg crc_t10dif_alg = {
+ .digestsize = CRC_T10DIF_DIGEST_SIZE,
+ .init = crct10dif_init,
+ .update = crct10dif_update,
+ .final = crct10dif_final,
+ .descsize = CRC_T10DIF_DIGEST_SIZE,
+
+ .base.cra_name = "crct10dif",
+ .base.cra_driver_name = "crct10dif-arm-ce",
+ .base.cra_priority = 200,
+ .base.cra_blocksize = CRC_T10DIF_BLOCK_SIZE,
+ .base.cra_module = THIS_MODULE,
+};
+
+static int __init crc_t10dif_mod_init(void)
+{
+ if (!(elf_hwcap2 & HWCAP2_PMULL))
+ return -ENODEV;
+
+ return crypto_register_shash(&crc_t10dif_alg);
+}
+
+static void __exit crc_t10dif_mod_exit(void)
+{
+ crypto_unregister_shash(&crc_t10dif_alg);
+}
+
+module_init(crc_t10dif_mod_init);
+module_exit(crc_t10dif_mod_exit);
+
+MODULE_AUTHOR("Ard Biesheuvel <ard.biesheuvel@linaro.org>");
+MODULE_LICENSE("GPL v2");
+MODULE_ALIAS_CRYPTO("crct10dif");
--
2.7.4
^ permalink raw reply related [flat|nested] 11+ messages in thread* [PATCH v3 5/6] crypto: arm64/crc32 - accelerated support based on x86 SSE implementation
2016-12-05 18:42 [PATCH v3 0/6] crypto: ARM/arm64 CRC-T10DIF/CRC32/CRC32C roundup Ard Biesheuvel
` (3 preceding siblings ...)
2016-12-05 18:42 ` [PATCH v3 4/6] crypto: arm/crct10dif - port x86 SSE implementation to ARM Ard Biesheuvel
@ 2016-12-05 18:42 ` Ard Biesheuvel
2016-12-05 18:42 ` [PATCH v3 6/6] crypto: arm/crc32 " Ard Biesheuvel
2016-12-07 12:09 ` [PATCH v3 0/6] crypto: ARM/arm64 CRC-T10DIF/CRC32/CRC32C roundup Herbert Xu
6 siblings, 0 replies; 11+ messages in thread
From: Ard Biesheuvel @ 2016-12-05 18:42 UTC (permalink / raw)
To: linux-arm-kernel
This is a combination of the the Intel algorithm implemented using SSE
and PCLMULQDQ instructions from arch/x86/crypto/crc32-pclmul_asm.S, and
the new CRC32 extensions introduced for both 32-bit and 64-bit ARM in
version 8 of the architecture. Two versions of the above combo are
provided, one for CRC32 and one for CRC32C.
The PMULL/NEON algorithm is faster, but operates on blocks of at least
64 bytes, and on multiples of 16 bytes only. For the remaining input,
or for all input on systems that lack the PMULL 64x64->128 instructions,
the CRC32 instructions will be used.
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
---
arch/arm64/crypto/Kconfig | 6 +
arch/arm64/crypto/Makefile | 3 +
arch/arm64/crypto/crc32-ce-core.S | 266 ++++++++++++++++++++
arch/arm64/crypto/crc32-ce-glue.c | 212 ++++++++++++++++
4 files changed, 487 insertions(+)
diff --git a/arch/arm64/crypto/Kconfig b/arch/arm64/crypto/Kconfig
index d773c0659202..21835deb1ab9 100644
--- a/arch/arm64/crypto/Kconfig
+++ b/arch/arm64/crypto/Kconfig
@@ -28,6 +28,11 @@ config CRYPTO_CRCT10DIF_ARM64_CE
depends on KERNEL_MODE_NEON && CRC_T10DIF
select CRYPTO_HASH
+config CRYPTO_CRC32_ARM64_CE
+ tristate "CRC32 and CRC32C digest algorithms using PMULL instructions"
+ depends on KERNEL_MODE_NEON && CRC32
+ select CRYPTO_HASH
+
config CRYPTO_AES_ARM64_CE
tristate "AES core cipher using ARMv8 Crypto Extensions"
depends on ARM64 && KERNEL_MODE_NEON
@@ -58,4 +63,5 @@ config CRYPTO_CRC32_ARM64
tristate "CRC32 and CRC32C using optional ARMv8 instructions"
depends on ARM64
select CRYPTO_HASH
+
endif
diff --git a/arch/arm64/crypto/Makefile b/arch/arm64/crypto/Makefile
index 36fd3eb4201b..144387805a46 100644
--- a/arch/arm64/crypto/Makefile
+++ b/arch/arm64/crypto/Makefile
@@ -20,6 +20,9 @@ ghash-ce-y := ghash-ce-glue.o ghash-ce-core.o
obj-$(CONFIG_CRYPTO_CRCT10DIF_ARM64_CE) += crct10dif-ce.o
crct10dif-ce-y := crct10dif-ce-core.o crct10dif-ce-glue.o
+obj-$(CONFIG_CRYPTO_CRC32_ARM64_CE) += crc32-ce.o
+crc32-ce-y:= crc32-ce-core.o crc32-ce-glue.o
+
obj-$(CONFIG_CRYPTO_AES_ARM64_CE) += aes-ce-cipher.o
CFLAGS_aes-ce-cipher.o += -march=armv8-a+crypto
diff --git a/arch/arm64/crypto/crc32-ce-core.S b/arch/arm64/crypto/crc32-ce-core.S
new file mode 100644
index 000000000000..18f5a8442276
--- /dev/null
+++ b/arch/arm64/crypto/crc32-ce-core.S
@@ -0,0 +1,266 @@
+/*
+ * Accelerated CRC32(C) using arm64 CRC, NEON and Crypto Extensions instructions
+ *
+ * Copyright (C) 2016 Linaro Ltd <ard.biesheuvel@linaro.org>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+/* GPL HEADER START
+ *
+ * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 only,
+ * as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * General Public License version 2 for more details (a copy is included
+ * in the LICENSE file that accompanied this code).
+ *
+ * You should have received a copy of the GNU General Public License
+ * version 2 along with this program; If not, see http://www.gnu.org/licenses
+ *
+ * Please visit http://www.xyratex.com/contact if you need additional
+ * information or have any questions.
+ *
+ * GPL HEADER END
+ */
+
+/*
+ * Copyright 2012 Xyratex Technology Limited
+ *
+ * Using hardware provided PCLMULQDQ instruction to accelerate the CRC32
+ * calculation.
+ * CRC32 polynomial:0x04c11db7(BE)/0xEDB88320(LE)
+ * PCLMULQDQ is a new instruction in Intel SSE4.2, the reference can be found
+ * at:
+ * http://www.intel.com/products/processor/manuals/
+ * Intel(R) 64 and IA-32 Architectures Software Developer's Manual
+ * Volume 2B: Instruction Set Reference, N-Z
+ *
+ * Authors: Gregory Prestas <Gregory_Prestas@us.xyratex.com>
+ * Alexander Boyko <Alexander_Boyko@xyratex.com>
+ */
+
+#include <linux/linkage.h>
+#include <asm/assembler.h>
+
+ .text
+ .align 6
+ .cpu generic+crypto+crc
+
+.Lcrc32_constants:
+ /*
+ * [x4*128+32 mod P(x) << 32)]' << 1 = 0x154442bd4
+ * #define CONSTANT_R1 0x154442bd4LL
+ *
+ * [(x4*128-32 mod P(x) << 32)]' << 1 = 0x1c6e41596
+ * #define CONSTANT_R2 0x1c6e41596LL
+ */
+ .octa 0x00000001c6e415960000000154442bd4
+
+ /*
+ * [(x128+32 mod P(x) << 32)]' << 1 = 0x1751997d0
+ * #define CONSTANT_R3 0x1751997d0LL
+ *
+ * [(x128-32 mod P(x) << 32)]' << 1 = 0x0ccaa009e
+ * #define CONSTANT_R4 0x0ccaa009eLL
+ */
+ .octa 0x00000000ccaa009e00000001751997d0
+
+ /*
+ * [(x64 mod P(x) << 32)]' << 1 = 0x163cd6124
+ * #define CONSTANT_R5 0x163cd6124LL
+ */
+ .quad 0x0000000163cd6124
+ .quad 0x00000000FFFFFFFF
+
+ /*
+ * #define CRCPOLY_TRUE_LE_FULL 0x1DB710641LL
+ *
+ * Barrett Reduction constant (u64`) = u` = (x**64 / P(x))`
+ * = 0x1F7011641LL
+ * #define CONSTANT_RU 0x1F7011641LL
+ */
+ .octa 0x00000001F701164100000001DB710641
+
+.Lcrc32c_constants:
+ .octa 0x000000009e4addf800000000740eef02
+ .octa 0x000000014cd00bd600000000f20c0dfe
+ .quad 0x00000000dd45aab8
+ .quad 0x00000000FFFFFFFF
+ .octa 0x00000000dea713f10000000105ec76f0
+
+ vCONSTANT .req v0
+ dCONSTANT .req d0
+ qCONSTANT .req q0
+
+ BUF .req x0
+ LEN .req x1
+ CRC .req x2
+
+ vzr .req v9
+
+ /**
+ * Calculate crc32
+ * BUF - buffer
+ * LEN - sizeof buffer (multiple of 16 bytes), LEN should be > 63
+ * CRC - initial crc32
+ * return %eax crc32
+ * uint crc32_pmull_le(unsigned char const *buffer,
+ * size_t len, uint crc32)
+ */
+ENTRY(crc32_pmull_le)
+ adr x3, .Lcrc32_constants
+ b 0f
+
+ENTRY(crc32c_pmull_le)
+ adr x3, .Lcrc32c_constants
+
+0: bic LEN, LEN, #15
+ ld1 {v1.16b-v4.16b}, [BUF], #0x40
+ movi vzr.16b, #0
+ fmov dCONSTANT, CRC
+ eor v1.16b, v1.16b, vCONSTANT.16b
+ sub LEN, LEN, #0x40
+ cmp LEN, #0x40
+ b.lt less_64
+
+ ldr qCONSTANT, [x3]
+
+loop_64: /* 64 bytes Full cache line folding */
+ sub LEN, LEN, #0x40
+
+ pmull2 v5.1q, v1.2d, vCONSTANT.2d
+ pmull2 v6.1q, v2.2d, vCONSTANT.2d
+ pmull2 v7.1q, v3.2d, vCONSTANT.2d
+ pmull2 v8.1q, v4.2d, vCONSTANT.2d
+
+ pmull v1.1q, v1.1d, vCONSTANT.1d
+ pmull v2.1q, v2.1d, vCONSTANT.1d
+ pmull v3.1q, v3.1d, vCONSTANT.1d
+ pmull v4.1q, v4.1d, vCONSTANT.1d
+
+ eor v1.16b, v1.16b, v5.16b
+ ld1 {v5.16b}, [BUF], #0x10
+ eor v2.16b, v2.16b, v6.16b
+ ld1 {v6.16b}, [BUF], #0x10
+ eor v3.16b, v3.16b, v7.16b
+ ld1 {v7.16b}, [BUF], #0x10
+ eor v4.16b, v4.16b, v8.16b
+ ld1 {v8.16b}, [BUF], #0x10
+
+ eor v1.16b, v1.16b, v5.16b
+ eor v2.16b, v2.16b, v6.16b
+ eor v3.16b, v3.16b, v7.16b
+ eor v4.16b, v4.16b, v8.16b
+
+ cmp LEN, #0x40
+ b.ge loop_64
+
+less_64: /* Folding cache line into 128bit */
+ ldr qCONSTANT, [x3, #16]
+
+ pmull2 v5.1q, v1.2d, vCONSTANT.2d
+ pmull v1.1q, v1.1d, vCONSTANT.1d
+ eor v1.16b, v1.16b, v5.16b
+ eor v1.16b, v1.16b, v2.16b
+
+ pmull2 v5.1q, v1.2d, vCONSTANT.2d
+ pmull v1.1q, v1.1d, vCONSTANT.1d
+ eor v1.16b, v1.16b, v5.16b
+ eor v1.16b, v1.16b, v3.16b
+
+ pmull2 v5.1q, v1.2d, vCONSTANT.2d
+ pmull v1.1q, v1.1d, vCONSTANT.1d
+ eor v1.16b, v1.16b, v5.16b
+ eor v1.16b, v1.16b, v4.16b
+
+ cbz LEN, fold_64
+
+loop_16: /* Folding rest buffer into 128bit */
+ subs LEN, LEN, #0x10
+
+ ld1 {v2.16b}, [BUF], #0x10
+ pmull2 v5.1q, v1.2d, vCONSTANT.2d
+ pmull v1.1q, v1.1d, vCONSTANT.1d
+ eor v1.16b, v1.16b, v5.16b
+ eor v1.16b, v1.16b, v2.16b
+
+ b.ne loop_16
+
+fold_64:
+ /* perform the last 64 bit fold, also adds 32 zeroes
+ * to the input stream */
+ ext v2.16b, v1.16b, v1.16b, #8
+ pmull2 v2.1q, v2.2d, vCONSTANT.2d
+ ext v1.16b, v1.16b, vzr.16b, #8
+ eor v1.16b, v1.16b, v2.16b
+
+ /* final 32-bit fold */
+ ldr dCONSTANT, [x3, #32]
+ ldr d3, [x3, #40]
+
+ ext v2.16b, v1.16b, vzr.16b, #4
+ and v1.16b, v1.16b, v3.16b
+ pmull v1.1q, v1.1d, vCONSTANT.1d
+ eor v1.16b, v1.16b, v2.16b
+
+ /* Finish up with the bit-reversed barrett reduction 64 ==> 32 bits */
+ ldr qCONSTANT, [x3, #48]
+
+ and v2.16b, v1.16b, v3.16b
+ ext v2.16b, vzr.16b, v2.16b, #8
+ pmull2 v2.1q, v2.2d, vCONSTANT.2d
+ and v2.16b, v2.16b, v3.16b
+ pmull v2.1q, v2.1d, vCONSTANT.1d
+ eor v1.16b, v1.16b, v2.16b
+ mov w0, v1.s[1]
+
+ ret
+ENDPROC(crc32_pmull_le)
+ENDPROC(crc32c_pmull_le)
+
+ .macro __crc32, c
+0: subs x2, x2, #16
+ b.mi 8f
+ ldp x3, x4, [x1], #16
+CPU_BE( rev x3, x3 )
+CPU_BE( rev x4, x4 )
+ crc32\c\()x w0, w0, x3
+ crc32\c\()x w0, w0, x4
+ b.ne 0b
+ ret
+
+8: tbz x2, #3, 4f
+ ldr x3, [x1], #8
+CPU_BE( rev x3, x3 )
+ crc32\c\()x w0, w0, x3
+4: tbz x2, #2, 2f
+ ldr w3, [x1], #4
+CPU_BE( rev w3, w3 )
+ crc32\c\()w w0, w0, w3
+2: tbz x2, #1, 1f
+ ldrh w3, [x1], #2
+CPU_BE( rev16 w3, w3 )
+ crc32\c\()h w0, w0, w3
+1: tbz x2, #0, 0f
+ ldrb w3, [x1]
+ crc32\c\()b w0, w0, w3
+0: ret
+ .endm
+
+ .align 5
+ENTRY(crc32_armv8_le)
+ __crc32
+ENDPROC(crc32_armv8_le)
+
+ .align 5
+ENTRY(crc32c_armv8_le)
+ __crc32 c
+ENDPROC(crc32c_armv8_le)
diff --git a/arch/arm64/crypto/crc32-ce-glue.c b/arch/arm64/crypto/crc32-ce-glue.c
new file mode 100644
index 000000000000..8594127d5e01
--- /dev/null
+++ b/arch/arm64/crypto/crc32-ce-glue.c
@@ -0,0 +1,212 @@
+/*
+ * Accelerated CRC32(C) using arm64 NEON and Crypto Extensions instructions
+ *
+ * Copyright (C) 2016 Linaro Ltd <ard.biesheuvel@linaro.org>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/cpufeature.h>
+#include <linux/crc32.h>
+#include <linux/init.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/string.h>
+
+#include <crypto/internal/hash.h>
+
+#include <asm/hwcap.h>
+#include <asm/neon.h>
+#include <asm/unaligned.h>
+
+#define PMULL_MIN_LEN 64L /* minimum size of buffer
+ * for crc32_pmull_le_16 */
+#define SCALE_F 16L /* size of NEON register */
+
+asmlinkage u32 crc32_pmull_le(const u8 buf[], u64 len, u32 init_crc);
+asmlinkage u32 crc32_armv8_le(u32 init_crc, const u8 buf[], size_t len);
+
+asmlinkage u32 crc32c_pmull_le(const u8 buf[], u64 len, u32 init_crc);
+asmlinkage u32 crc32c_armv8_le(u32 init_crc, const u8 buf[], size_t len);
+
+static u32 (*fallback_crc32)(u32 init_crc, const u8 buf[], size_t len);
+static u32 (*fallback_crc32c)(u32 init_crc, const u8 buf[], size_t len);
+
+static int crc32_pmull_cra_init(struct crypto_tfm *tfm)
+{
+ u32 *key = crypto_tfm_ctx(tfm);
+
+ *key = 0;
+ return 0;
+}
+
+static int crc32c_pmull_cra_init(struct crypto_tfm *tfm)
+{
+ u32 *key = crypto_tfm_ctx(tfm);
+
+ *key = ~0;
+ return 0;
+}
+
+static int crc32_pmull_setkey(struct crypto_shash *hash, const u8 *key,
+ unsigned int keylen)
+{
+ u32 *mctx = crypto_shash_ctx(hash);
+
+ if (keylen != sizeof(u32)) {
+ crypto_shash_set_flags(hash, CRYPTO_TFM_RES_BAD_KEY_LEN);
+ return -EINVAL;
+ }
+ *mctx = le32_to_cpup((__le32 *)key);
+ return 0;
+}
+
+static int crc32_pmull_init(struct shash_desc *desc)
+{
+ u32 *mctx = crypto_shash_ctx(desc->tfm);
+ u32 *crc = shash_desc_ctx(desc);
+
+ *crc = *mctx;
+ return 0;
+}
+
+static int crc32_pmull_update(struct shash_desc *desc, const u8 *data,
+ unsigned int length)
+{
+ u32 *crc = shash_desc_ctx(desc);
+ unsigned int l;
+
+ if ((u64)data % SCALE_F) {
+ l = min_t(u32, length, SCALE_F - ((u64)data % SCALE_F));
+
+ *crc = fallback_crc32(*crc, data, l);
+
+ data += l;
+ length -= l;
+ }
+
+ if (length >= PMULL_MIN_LEN) {
+ l = round_down(length, SCALE_F);
+
+ kernel_neon_begin_partial(10);
+ *crc = crc32_pmull_le(data, l, *crc);
+ kernel_neon_end();
+
+ data += l;
+ length -= l;
+ }
+
+ if (length > 0)
+ *crc = fallback_crc32(*crc, data, length);
+
+ return 0;
+}
+
+static int crc32c_pmull_update(struct shash_desc *desc, const u8 *data,
+ unsigned int length)
+{
+ u32 *crc = shash_desc_ctx(desc);
+ unsigned int l;
+
+ if ((u64)data % SCALE_F) {
+ l = min_t(u32, length, SCALE_F - ((u64)data % SCALE_F));
+
+ *crc = fallback_crc32c(*crc, data, l);
+
+ data += l;
+ length -= l;
+ }
+
+ if (length >= PMULL_MIN_LEN) {
+ l = round_down(length, SCALE_F);
+
+ kernel_neon_begin_partial(10);
+ *crc = crc32c_pmull_le(data, l, *crc);
+ kernel_neon_end();
+
+ data += l;
+ length -= l;
+ }
+
+ if (length > 0) {
+ *crc = fallback_crc32c(*crc, data, length);
+ }
+
+ return 0;
+}
+
+static int crc32_pmull_final(struct shash_desc *desc, u8 *out)
+{
+ u32 *crc = shash_desc_ctx(desc);
+
+ put_unaligned_le32(*crc, out);
+ return 0;
+}
+
+static int crc32c_pmull_final(struct shash_desc *desc, u8 *out)
+{
+ u32 *crc = shash_desc_ctx(desc);
+
+ put_unaligned_le32(~*crc, out);
+ return 0;
+}
+
+static struct shash_alg crc32_pmull_algs[] = { {
+ .setkey = crc32_pmull_setkey,
+ .init = crc32_pmull_init,
+ .update = crc32_pmull_update,
+ .final = crc32_pmull_final,
+ .descsize = sizeof(u32),
+ .digestsize = sizeof(u32),
+
+ .base.cra_ctxsize = sizeof(u32),
+ .base.cra_init = crc32_pmull_cra_init,
+ .base.cra_name = "crc32",
+ .base.cra_driver_name = "crc32-arm64-ce",
+ .base.cra_priority = 200,
+ .base.cra_blocksize = 1,
+ .base.cra_module = THIS_MODULE,
+}, {
+ .setkey = crc32_pmull_setkey,
+ .init = crc32_pmull_init,
+ .update = crc32c_pmull_update,
+ .final = crc32c_pmull_final,
+ .descsize = sizeof(u32),
+ .digestsize = sizeof(u32),
+
+ .base.cra_ctxsize = sizeof(u32),
+ .base.cra_init = crc32c_pmull_cra_init,
+ .base.cra_name = "crc32c",
+ .base.cra_driver_name = "crc32c-arm64-ce",
+ .base.cra_priority = 200,
+ .base.cra_blocksize = 1,
+ .base.cra_module = THIS_MODULE,
+} };
+
+static int __init crc32_pmull_mod_init(void)
+{
+ if (elf_hwcap & HWCAP_CRC32) {
+ fallback_crc32 = crc32_armv8_le;
+ fallback_crc32c = crc32c_armv8_le;
+ } else {
+ fallback_crc32 = crc32_le;
+ fallback_crc32c = __crc32c_le;
+ }
+
+ return crypto_register_shashes(crc32_pmull_algs,
+ ARRAY_SIZE(crc32_pmull_algs));
+}
+
+static void __exit crc32_pmull_mod_exit(void)
+{
+ crypto_unregister_shashes(crc32_pmull_algs,
+ ARRAY_SIZE(crc32_pmull_algs));
+}
+
+module_cpu_feature_match(PMULL, crc32_pmull_mod_init);
+module_exit(crc32_pmull_mod_exit);
+
+MODULE_AUTHOR("Ard Biesheuvel <ard.biesheuvel@linaro.org>");
+MODULE_LICENSE("GPL v2");
--
2.7.4
^ permalink raw reply related [flat|nested] 11+ messages in thread* [PATCH v3 6/6] crypto: arm/crc32 - accelerated support based on x86 SSE implementation
2016-12-05 18:42 [PATCH v3 0/6] crypto: ARM/arm64 CRC-T10DIF/CRC32/CRC32C roundup Ard Biesheuvel
` (4 preceding siblings ...)
2016-12-05 18:42 ` [PATCH v3 5/6] crypto: arm64/crc32 - accelerated support based on x86 SSE implementation Ard Biesheuvel
@ 2016-12-05 18:42 ` Ard Biesheuvel
2016-12-07 12:09 ` [PATCH v3 0/6] crypto: ARM/arm64 CRC-T10DIF/CRC32/CRC32C roundup Herbert Xu
6 siblings, 0 replies; 11+ messages in thread
From: Ard Biesheuvel @ 2016-12-05 18:42 UTC (permalink / raw)
To: linux-arm-kernel
This is a combination of the the Intel algorithm implemented using SSE
and PCLMULQDQ instructions from arch/x86/crypto/crc32-pclmul_asm.S, and
the new CRC32 extensions introduced for both 32-bit and 64-bit ARM in
version 8 of the architecture. Two versions of the above combo are
provided, one for CRC32 and one for CRC32C.
The PMULL/NEON algorithm is faster, but operates on blocks of at least
64 bytes, and on multiples of 16 bytes only. For the remaining input,
or for all input on systems that lack the PMULL 64x64->128 instructions,
the CRC32 instructions will be used.
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
---
arch/arm/crypto/Kconfig | 5 +
arch/arm/crypto/Makefile | 2 +
arch/arm/crypto/crc32-ce-core.S | 306 ++++++++++++++++++++
arch/arm/crypto/crc32-ce-glue.c | 242 ++++++++++++++++
4 files changed, 555 insertions(+)
diff --git a/arch/arm/crypto/Kconfig b/arch/arm/crypto/Kconfig
index fce801fa52a1..de7bb20815bf 100644
--- a/arch/arm/crypto/Kconfig
+++ b/arch/arm/crypto/Kconfig
@@ -125,4 +125,9 @@ config CRYPTO_CRCT10DIF_ARM_CE
depends on KERNEL_MODE_NEON && CRC_T10DIF
select CRYPTO_HASH
+config CRYPTO_CRC32_ARM_CE
+ tristate "CRC32(C) digest algorithm using CRC and/or PMULL instructions"
+ depends on KERNEL_MODE_NEON && CRC32
+ select CRYPTO_HASH
+
endif
diff --git a/arch/arm/crypto/Makefile b/arch/arm/crypto/Makefile
index fc77265014b7..b578a1820ab1 100644
--- a/arch/arm/crypto/Makefile
+++ b/arch/arm/crypto/Makefile
@@ -14,6 +14,7 @@ ce-obj-$(CONFIG_CRYPTO_SHA1_ARM_CE) += sha1-arm-ce.o
ce-obj-$(CONFIG_CRYPTO_SHA2_ARM_CE) += sha2-arm-ce.o
ce-obj-$(CONFIG_CRYPTO_GHASH_ARM_CE) += ghash-arm-ce.o
ce-obj-$(CONFIG_CRYPTO_CRCT10DIF_ARM_CE) += crct10dif-arm-ce.o
+ce-obj-$(CONFIG_CRYPTO_CRC32_ARM_CE) += crc32-arm-ce.o
ifneq ($(ce-obj-y)$(ce-obj-m),)
ifeq ($(call as-instr,.fpu crypto-neon-fp-armv8,y,n),y)
@@ -38,6 +39,7 @@ sha2-arm-ce-y := sha2-ce-core.o sha2-ce-glue.o
aes-arm-ce-y := aes-ce-core.o aes-ce-glue.o
ghash-arm-ce-y := ghash-ce-core.o ghash-ce-glue.o
crct10dif-arm-ce-y := crct10dif-ce-core.o crct10dif-ce-glue.o
+crc32-arm-ce-y:= crc32-ce-core.o crc32-ce-glue.o
quiet_cmd_perl = PERL $@
cmd_perl = $(PERL) $(<) > $(@)
diff --git a/arch/arm/crypto/crc32-ce-core.S b/arch/arm/crypto/crc32-ce-core.S
new file mode 100644
index 000000000000..e63d400dc5c1
--- /dev/null
+++ b/arch/arm/crypto/crc32-ce-core.S
@@ -0,0 +1,306 @@
+/*
+ * Accelerated CRC32(C) using ARM CRC, NEON and Crypto Extensions instructions
+ *
+ * Copyright (C) 2016 Linaro Ltd <ard.biesheuvel@linaro.org>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+/* GPL HEADER START
+ *
+ * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 only,
+ * as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * General Public License version 2 for more details (a copy is included
+ * in the LICENSE file that accompanied this code).
+ *
+ * You should have received a copy of the GNU General Public License
+ * version 2 along with this program; If not, see http://www.gnu.org/licenses
+ *
+ * Please visit http://www.xyratex.com/contact if you need additional
+ * information or have any questions.
+ *
+ * GPL HEADER END
+ */
+
+/*
+ * Copyright 2012 Xyratex Technology Limited
+ *
+ * Using hardware provided PCLMULQDQ instruction to accelerate the CRC32
+ * calculation.
+ * CRC32 polynomial:0x04c11db7(BE)/0xEDB88320(LE)
+ * PCLMULQDQ is a new instruction in Intel SSE4.2, the reference can be found
+ * at:
+ * http://www.intel.com/products/processor/manuals/
+ * Intel(R) 64 and IA-32 Architectures Software Developer's Manual
+ * Volume 2B: Instruction Set Reference, N-Z
+ *
+ * Authors: Gregory Prestas <Gregory_Prestas@us.xyratex.com>
+ * Alexander Boyko <Alexander_Boyko@xyratex.com>
+ */
+
+#include <linux/linkage.h>
+#include <asm/assembler.h>
+
+ .text
+ .align 6
+ .arch armv8-a
+ .arch_extension crc
+ .fpu crypto-neon-fp-armv8
+
+.Lcrc32_constants:
+ /*
+ * [x4*128+32 mod P(x) << 32)]' << 1 = 0x154442bd4
+ * #define CONSTANT_R1 0x154442bd4LL
+ *
+ * [(x4*128-32 mod P(x) << 32)]' << 1 = 0x1c6e41596
+ * #define CONSTANT_R2 0x1c6e41596LL
+ */
+ .quad 0x0000000154442bd4
+ .quad 0x00000001c6e41596
+
+ /*
+ * [(x128+32 mod P(x) << 32)]' << 1 = 0x1751997d0
+ * #define CONSTANT_R3 0x1751997d0LL
+ *
+ * [(x128-32 mod P(x) << 32)]' << 1 = 0x0ccaa009e
+ * #define CONSTANT_R4 0x0ccaa009eLL
+ */
+ .quad 0x00000001751997d0
+ .quad 0x00000000ccaa009e
+
+ /*
+ * [(x64 mod P(x) << 32)]' << 1 = 0x163cd6124
+ * #define CONSTANT_R5 0x163cd6124LL
+ */
+ .quad 0x0000000163cd6124
+ .quad 0x00000000FFFFFFFF
+
+ /*
+ * #define CRCPOLY_TRUE_LE_FULL 0x1DB710641LL
+ *
+ * Barrett Reduction constant (u64`) = u` = (x**64 / P(x))`
+ * = 0x1F7011641LL
+ * #define CONSTANT_RU 0x1F7011641LL
+ */
+ .quad 0x00000001DB710641
+ .quad 0x00000001F7011641
+
+.Lcrc32c_constants:
+ .quad 0x00000000740eef02
+ .quad 0x000000009e4addf8
+ .quad 0x00000000f20c0dfe
+ .quad 0x000000014cd00bd6
+ .quad 0x00000000dd45aab8
+ .quad 0x00000000FFFFFFFF
+ .quad 0x0000000105ec76f0
+ .quad 0x00000000dea713f1
+
+ dCONSTANTl .req d0
+ dCONSTANTh .req d1
+ qCONSTANT .req q0
+
+ BUF .req r0
+ LEN .req r1
+ CRC .req r2
+
+ qzr .req q9
+
+ /**
+ * Calculate crc32
+ * BUF - buffer
+ * LEN - sizeof buffer (multiple of 16 bytes), LEN should be > 63
+ * CRC - initial crc32
+ * return %eax crc32
+ * uint crc32_pmull_le(unsigned char const *buffer,
+ * size_t len, uint crc32)
+ */
+ENTRY(crc32_pmull_le)
+ adr r3, .Lcrc32_constants
+ b 0f
+
+ENTRY(crc32c_pmull_le)
+ adr r3, .Lcrc32c_constants
+
+0: bic LEN, LEN, #15
+ vld1.8 {q1-q2}, [BUF, :128]!
+ vld1.8 {q3-q4}, [BUF, :128]!
+ vmov.i8 qzr, #0
+ vmov.i8 qCONSTANT, #0
+ vmov dCONSTANTl[0], CRC
+ veor.8 d2, d2, dCONSTANTl
+ sub LEN, LEN, #0x40
+ cmp LEN, #0x40
+ blt less_64
+
+ vld1.64 {qCONSTANT}, [r3]
+
+loop_64: /* 64 bytes Full cache line folding */
+ sub LEN, LEN, #0x40
+
+ vmull.p64 q5, d3, dCONSTANTh
+ vmull.p64 q6, d5, dCONSTANTh
+ vmull.p64 q7, d7, dCONSTANTh
+ vmull.p64 q8, d9, dCONSTANTh
+
+ vmull.p64 q1, d2, dCONSTANTl
+ vmull.p64 q2, d4, dCONSTANTl
+ vmull.p64 q3, d6, dCONSTANTl
+ vmull.p64 q4, d8, dCONSTANTl
+
+ veor.8 q1, q1, q5
+ vld1.8 {q5}, [BUF, :128]!
+ veor.8 q2, q2, q6
+ vld1.8 {q6}, [BUF, :128]!
+ veor.8 q3, q3, q7
+ vld1.8 {q7}, [BUF, :128]!
+ veor.8 q4, q4, q8
+ vld1.8 {q8}, [BUF, :128]!
+
+ veor.8 q1, q1, q5
+ veor.8 q2, q2, q6
+ veor.8 q3, q3, q7
+ veor.8 q4, q4, q8
+
+ cmp LEN, #0x40
+ bge loop_64
+
+less_64: /* Folding cache line into 128bit */
+ vldr dCONSTANTl, [r3, #16]
+ vldr dCONSTANTh, [r3, #24]
+
+ vmull.p64 q5, d3, dCONSTANTh
+ vmull.p64 q1, d2, dCONSTANTl
+ veor.8 q1, q1, q5
+ veor.8 q1, q1, q2
+
+ vmull.p64 q5, d3, dCONSTANTh
+ vmull.p64 q1, d2, dCONSTANTl
+ veor.8 q1, q1, q5
+ veor.8 q1, q1, q3
+
+ vmull.p64 q5, d3, dCONSTANTh
+ vmull.p64 q1, d2, dCONSTANTl
+ veor.8 q1, q1, q5
+ veor.8 q1, q1, q4
+
+ teq LEN, #0
+ beq fold_64
+
+loop_16: /* Folding rest buffer into 128bit */
+ subs LEN, LEN, #0x10
+
+ vld1.8 {q2}, [BUF, :128]!
+ vmull.p64 q5, d3, dCONSTANTh
+ vmull.p64 q1, d2, dCONSTANTl
+ veor.8 q1, q1, q5
+ veor.8 q1, q1, q2
+
+ bne loop_16
+
+fold_64:
+ /* perform the last 64 bit fold, also adds 32 zeroes
+ * to the input stream */
+ vmull.p64 q2, d2, dCONSTANTh
+ vext.8 q1, q1, qzr, #8
+ veor.8 q1, q1, q2
+
+ /* final 32-bit fold */
+ vldr dCONSTANTl, [r3, #32]
+ vldr d6, [r3, #40]
+ vmov.i8 d7, #0
+
+ vext.8 q2, q1, qzr, #4
+ vand.8 d2, d2, d6
+ vmull.p64 q1, d2, dCONSTANTl
+ veor.8 q1, q1, q2
+
+ /* Finish up with the bit-reversed barrett reduction 64 ==> 32 bits */
+ vldr dCONSTANTl, [r3, #48]
+ vldr dCONSTANTh, [r3, #56]
+
+ vand.8 q2, q1, q3
+ vext.8 q2, qzr, q2, #8
+ vmull.p64 q2, d5, dCONSTANTh
+ vand.8 q2, q2, q3
+ vmull.p64 q2, d4, dCONSTANTl
+ veor.8 q1, q1, q2
+ vmov r0, s5
+
+ bx lr
+ENDPROC(crc32_pmull_le)
+ENDPROC(crc32c_pmull_le)
+
+ .macro __crc32, c
+ subs ip, r2, #8
+ bmi .Ltail\c
+
+ tst r1, #3
+ bne .Lunaligned\c
+
+ teq ip, #0
+.Laligned8\c:
+ ldrd r2, r3, [r1], #8
+ARM_BE8(rev r2, r2 )
+ARM_BE8(rev r3, r3 )
+ crc32\c\()w r0, r0, r2
+ crc32\c\()w r0, r0, r3
+ bxeq lr
+ subs ip, ip, #8
+ bpl .Laligned8\c
+
+.Ltail\c:
+ tst ip, #4
+ beq 2f
+ ldr r3, [r1], #4
+ARM_BE8(rev r3, r3 )
+ crc32\c\()w r0, r0, r3
+
+2: tst ip, #2
+ beq 1f
+ ldrh r3, [r1], #2
+ARM_BE8(rev16 r3, r3 )
+ crc32\c\()h r0, r0, r3
+
+1: tst ip, #1
+ bxeq lr
+ ldrb r3, [r1]
+ crc32\c\()b r0, r0, r3
+ bx lr
+
+.Lunaligned\c:
+ tst r1, #1
+ beq 2f
+ ldrb r3, [r1], #1
+ subs r2, r2, #1
+ crc32\c\()b r0, r0, r3
+
+ tst r1, #2
+ beq 0f
+2: ldrh r3, [r1], #2
+ subs r2, r2, #2
+ARM_BE8(rev16 r3, r3 )
+ crc32\c\()h r0, r0, r3
+
+0: subs ip, r2, #8
+ bpl .Laligned8\c
+ b .Ltail\c
+ .endm
+
+ .align 5
+ENTRY(crc32_armv8_le)
+ __crc32
+ENDPROC(crc32_armv8_le)
+
+ .align 5
+ENTRY(crc32c_armv8_le)
+ __crc32 c
+ENDPROC(crc32c_armv8_le)
diff --git a/arch/arm/crypto/crc32-ce-glue.c b/arch/arm/crypto/crc32-ce-glue.c
new file mode 100644
index 000000000000..e1566bec1016
--- /dev/null
+++ b/arch/arm/crypto/crc32-ce-glue.c
@@ -0,0 +1,242 @@
+/*
+ * Accelerated CRC32(C) using ARM CRC, NEON and Crypto Extensions instructions
+ *
+ * Copyright (C) 2016 Linaro Ltd <ard.biesheuvel@linaro.org>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/crc32.h>
+#include <linux/init.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/string.h>
+
+#include <crypto/internal/hash.h>
+
+#include <asm/hwcap.h>
+#include <asm/neon.h>
+#include <asm/simd.h>
+#include <asm/unaligned.h>
+
+#define PMULL_MIN_LEN 64L /* minimum size of buffer
+ * for crc32_pmull_le_16 */
+#define SCALE_F 16L /* size of NEON register */
+
+asmlinkage u32 crc32_pmull_le(const u8 buf[], u32 len, u32 init_crc);
+asmlinkage u32 crc32_armv8_le(u32 init_crc, const u8 buf[], u32 len);
+
+asmlinkage u32 crc32c_pmull_le(const u8 buf[], u32 len, u32 init_crc);
+asmlinkage u32 crc32c_armv8_le(u32 init_crc, const u8 buf[], u32 len);
+
+static u32 (*fallback_crc32)(u32 init_crc, const u8 buf[], u32 len);
+static u32 (*fallback_crc32c)(u32 init_crc, const u8 buf[], u32 len);
+
+static int crc32_cra_init(struct crypto_tfm *tfm)
+{
+ u32 *key = crypto_tfm_ctx(tfm);
+
+ *key = 0;
+ return 0;
+}
+
+static int crc32c_cra_init(struct crypto_tfm *tfm)
+{
+ u32 *key = crypto_tfm_ctx(tfm);
+
+ *key = ~0;
+ return 0;
+}
+
+static int crc32_setkey(struct crypto_shash *hash, const u8 *key,
+ unsigned int keylen)
+{
+ u32 *mctx = crypto_shash_ctx(hash);
+
+ if (keylen != sizeof(u32)) {
+ crypto_shash_set_flags(hash, CRYPTO_TFM_RES_BAD_KEY_LEN);
+ return -EINVAL;
+ }
+ *mctx = le32_to_cpup((__le32 *)key);
+ return 0;
+}
+
+static int crc32_init(struct shash_desc *desc)
+{
+ u32 *mctx = crypto_shash_ctx(desc->tfm);
+ u32 *crc = shash_desc_ctx(desc);
+
+ *crc = *mctx;
+ return 0;
+}
+
+static int crc32_update(struct shash_desc *desc, const u8 *data,
+ unsigned int length)
+{
+ u32 *crc = shash_desc_ctx(desc);
+
+ *crc = crc32_armv8_le(*crc, data, length);
+ return 0;
+}
+
+static int crc32c_update(struct shash_desc *desc, const u8 *data,
+ unsigned int length)
+{
+ u32 *crc = shash_desc_ctx(desc);
+
+ *crc = crc32c_armv8_le(*crc, data, length);
+ return 0;
+}
+
+static int crc32_final(struct shash_desc *desc, u8 *out)
+{
+ u32 *crc = shash_desc_ctx(desc);
+
+ put_unaligned_le32(*crc, out);
+ return 0;
+}
+
+static int crc32c_final(struct shash_desc *desc, u8 *out)
+{
+ u32 *crc = shash_desc_ctx(desc);
+
+ put_unaligned_le32(~*crc, out);
+ return 0;
+}
+
+static int crc32_pmull_update(struct shash_desc *desc, const u8 *data,
+ unsigned int length)
+{
+ u32 *crc = shash_desc_ctx(desc);
+ unsigned int l;
+
+ if (may_use_simd()) {
+ if ((u32)data % SCALE_F) {
+ l = min_t(u32, length, SCALE_F - ((u32)data % SCALE_F));
+
+ *crc = fallback_crc32(*crc, data, l);
+
+ data += l;
+ length -= l;
+ }
+
+ if (length >= PMULL_MIN_LEN) {
+ l = round_down(length, SCALE_F);
+
+ kernel_neon_begin();
+ *crc = crc32_pmull_le(data, l, *crc);
+ kernel_neon_end();
+
+ data += l;
+ length -= l;
+ }
+ }
+
+ if (length > 0)
+ *crc = fallback_crc32(*crc, data, length);
+
+ return 0;
+}
+
+static int crc32c_pmull_update(struct shash_desc *desc, const u8 *data,
+ unsigned int length)
+{
+ u32 *crc = shash_desc_ctx(desc);
+ unsigned int l;
+
+ if (may_use_simd()) {
+ if ((u32)data % SCALE_F) {
+ l = min_t(u32, length, SCALE_F - ((u32)data % SCALE_F));
+
+ *crc = fallback_crc32c(*crc, data, l);
+
+ data += l;
+ length -= l;
+ }
+
+ if (length >= PMULL_MIN_LEN) {
+ l = round_down(length, SCALE_F);
+
+ kernel_neon_begin();
+ *crc = crc32c_pmull_le(data, l, *crc);
+ kernel_neon_end();
+
+ data += l;
+ length -= l;
+ }
+ }
+
+ if (length > 0)
+ *crc = fallback_crc32c(*crc, data, length);
+
+ return 0;
+}
+
+static struct shash_alg crc32_pmull_algs[] = { {
+ .setkey = crc32_setkey,
+ .init = crc32_init,
+ .update = crc32_update,
+ .final = crc32_final,
+ .descsize = sizeof(u32),
+ .digestsize = sizeof(u32),
+
+ .base.cra_ctxsize = sizeof(u32),
+ .base.cra_init = crc32_cra_init,
+ .base.cra_name = "crc32",
+ .base.cra_driver_name = "crc32-arm-ce",
+ .base.cra_priority = 200,
+ .base.cra_blocksize = 1,
+ .base.cra_module = THIS_MODULE,
+}, {
+ .setkey = crc32_setkey,
+ .init = crc32_init,
+ .update = crc32c_update,
+ .final = crc32c_final,
+ .descsize = sizeof(u32),
+ .digestsize = sizeof(u32),
+
+ .base.cra_ctxsize = sizeof(u32),
+ .base.cra_init = crc32c_cra_init,
+ .base.cra_name = "crc32c",
+ .base.cra_driver_name = "crc32c-arm-ce",
+ .base.cra_priority = 200,
+ .base.cra_blocksize = 1,
+ .base.cra_module = THIS_MODULE,
+} };
+
+static int __init crc32_pmull_mod_init(void)
+{
+ if (elf_hwcap2 & HWCAP2_PMULL) {
+ crc32_pmull_algs[0].update = crc32_pmull_update;
+ crc32_pmull_algs[1].update = crc32c_pmull_update;
+
+ if (elf_hwcap2 & HWCAP2_CRC32) {
+ fallback_crc32 = crc32_armv8_le;
+ fallback_crc32c = crc32c_armv8_le;
+ } else {
+ fallback_crc32 = crc32_le;
+ fallback_crc32c = __crc32c_le;
+ }
+ } else if (!(elf_hwcap2 & HWCAP2_CRC32)) {
+ return -ENODEV;
+ }
+
+ return crypto_register_shashes(crc32_pmull_algs,
+ ARRAY_SIZE(crc32_pmull_algs));
+}
+
+static void __exit crc32_pmull_mod_exit(void)
+{
+ crypto_unregister_shashes(crc32_pmull_algs,
+ ARRAY_SIZE(crc32_pmull_algs));
+}
+
+module_init(crc32_pmull_mod_init);
+module_exit(crc32_pmull_mod_exit);
+
+MODULE_AUTHOR("Ard Biesheuvel <ard.biesheuvel@linaro.org>");
+MODULE_LICENSE("GPL v2");
+MODULE_ALIAS_CRYPTO("crc32");
+MODULE_ALIAS_CRYPTO("crc32c");
--
2.7.4
^ permalink raw reply related [flat|nested] 11+ messages in thread* [PATCH v3 0/6] crypto: ARM/arm64 CRC-T10DIF/CRC32/CRC32C roundup
2016-12-05 18:42 [PATCH v3 0/6] crypto: ARM/arm64 CRC-T10DIF/CRC32/CRC32C roundup Ard Biesheuvel
` (5 preceding siblings ...)
2016-12-05 18:42 ` [PATCH v3 6/6] crypto: arm/crc32 " Ard Biesheuvel
@ 2016-12-07 12:09 ` Herbert Xu
6 siblings, 0 replies; 11+ messages in thread
From: Herbert Xu @ 2016-12-07 12:09 UTC (permalink / raw)
To: linux-arm-kernel
On Mon, Dec 05, 2016 at 06:42:22PM +0000, Ard Biesheuvel wrote:
> This v3 combines the CRC-T10DIF and CRC32 implementations for both ARM and
> arm64 that I sent out a couple of weeks ago, and adds support to the latter
> for CRC32C.
All applied. Thanks.
--
Email: Herbert Xu <herbert@gondor.apana.org.au>
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt
^ permalink raw reply [flat|nested] 11+ messages in thread