[PATCH 0/4] crypto: add sha256() function

[PATCH 0/4] crypto: add sha256() function

[Eric Biggers]

This series adds a function sha256() to the sha256 library so that users
who want to compute a hash in one step can just call sha256() instead of
sha256_init() + sha256_update() + sha256_final().

Patches 2-4 then convert some users to use it.

Eric Biggers (4):
  crypto: lib/sha256 - add sha256() function
  efi: use sha256() instead of open coding
  mptcp: use sha256() instead of open coding
  ASoC: cros_ec_codec: use sha256() instead of open coding

 drivers/firmware/efi/embedded-firmware.c |  9 +++-----
 include/crypto/sha.h                     |  1 +
 lib/crypto/sha256.c                      | 10 +++++++++
 net/mptcp/crypto.c                       | 15 +++----------
 sound/soc/codecs/cros_ec_codec.c         | 27 ++----------------------
 5 files changed, 19 insertions(+), 43 deletions(-)


base-commit: 57c8aa43b9f272c382c253573c82be5cb68fe22d
-- 
2.27.0

[PATCH 1/4] crypto: lib/sha256 - add sha256() function

[Eric Biggers]

From: Eric Biggers <ebiggers@google.com>

Add a function sha256() which computes a SHA-256 digest in one step,
combining sha256_init() + sha256_update() + sha256_final().

This is similar to how we also have blake2s().

Signed-off-by: Eric Biggers <ebiggers@google.com>
---
 include/crypto/sha.h |  1 +
 lib/crypto/sha256.c  | 10 ++++++++++
 2 files changed, 11 insertions(+)

diff --git a/include/crypto/sha.h b/include/crypto/sha.h
index 10753ff71d46..4ff3da816630 100644
--- a/include/crypto/sha.h
+++ b/include/crypto/sha.h
@@ -147,6 +147,7 @@ static inline void sha256_init(struct sha256_state *sctx)
 }
 void sha256_update(struct sha256_state *sctx, const u8 *data, unsigned int len);
 void sha256_final(struct sha256_state *sctx, u8 *out);
+void sha256(const u8 *data, unsigned int len, u8 *out);
 
 static inline void sha224_init(struct sha256_state *sctx)
 {
diff --git a/lib/crypto/sha256.c b/lib/crypto/sha256.c
index 2e621697c5c3..2321f6cb322f 100644
--- a/lib/crypto/sha256.c
+++ b/lib/crypto/sha256.c
@@ -280,4 +280,14 @@ void sha224_final(struct sha256_state *sctx, u8 *out)
 }
 EXPORT_SYMBOL(sha224_final);
 
+void sha256(const u8 *data, unsigned int len, u8 *out)
+{
+	struct sha256_state sctx;
+
+	sha256_init(&sctx);
+	sha256_update(&sctx, data, len);
+	sha256_final(&sctx, out);
+}
+EXPORT_SYMBOL(sha256);
+
 MODULE_LICENSE("GPL");
-- 
2.27.0

[PATCH 2/4] efi: use sha256() instead of open coding

[Eric Biggers]

From: Eric Biggers <ebiggers@google.com>

Now that there's a function that calculates the SHA-256 digest of a
buffer in one step, use it instead of sha256_init() + sha256_update() +
sha256_final().

Cc: linux-efi@vger.kernel.org
Cc: Ard Biesheuvel <ardb@kernel.org>
Cc: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Eric Biggers <ebiggers@google.com>
---
 drivers/firmware/efi/embedded-firmware.c | 9 +++------
 1 file changed, 3 insertions(+), 6 deletions(-)

diff --git a/drivers/firmware/efi/embedded-firmware.c b/drivers/firmware/efi/embedded-firmware.c
index a1b199de9006..e97a9c9d010c 100644
--- a/drivers/firmware/efi/embedded-firmware.c
+++ b/drivers/firmware/efi/embedded-firmware.c
@@ -37,9 +37,8 @@ static const struct dmi_system_id * const embedded_fw_table[] = {
 static int __init efi_check_md_for_embedded_firmware(
 	efi_memory_desc_t *md, const struct efi_embedded_fw_desc *desc)
 {
-	struct sha256_state sctx;
 	struct efi_embedded_fw *fw;
-	u8 sha256[32];
+	u8 hash[32];
 	u64 i, size;
 	u8 *map;
 
@@ -54,10 +53,8 @@ static int __init efi_check_md_for_embedded_firmware(
 		if (memcmp(map + i, desc->prefix, EFI_EMBEDDED_FW_PREFIX_LEN))
 			continue;
 
-		sha256_init(&sctx);
-		sha256_update(&sctx, map + i, desc->length);
-		sha256_final(&sctx, sha256);
-		if (memcmp(sha256, desc->sha256, 32) == 0)
+		sha256(map + i, desc->length, hash);
+		if (memcmp(hash, desc->sha256, 32) == 0)
 			break;
 	}
 	if ((i + desc->length) > size) {
-- 
2.27.0

Re: [PATCH 0/4] crypto: add sha256() function

[Ard Biesheuvel]

On Tue, 7 Jul 2020 at 21:59, Eric Biggers <ebiggers@kernel.org> wrote:
>
> This series adds a function sha256() to the sha256 library so that users
> who want to compute a hash in one step can just call sha256() instead of
> sha256_init() + sha256_update() + sha256_final().
>
> Patches 2-4 then convert some users to use it.
>
> Eric Biggers (4):
>   crypto: lib/sha256 - add sha256() function
>   efi: use sha256() instead of open coding
>   mptcp: use sha256() instead of open coding
>   ASoC: cros_ec_codec: use sha256() instead of open coding
>

For the series,

Reviewed-by: Ard Biesheuvel <ardb@kernel.org>

Feel free to take the EFI patch through the crypto tree.


>  drivers/firmware/efi/embedded-firmware.c |  9 +++-----
>  include/crypto/sha.h                     |  1 +
>  lib/crypto/sha256.c                      | 10 +++++++++
>  net/mptcp/crypto.c                       | 15 +++----------
>  sound/soc/codecs/cros_ec_codec.c         | 27 ++----------------------
>  5 files changed, 19 insertions(+), 43 deletions(-)
>
>
> base-commit: 57c8aa43b9f272c382c253573c82be5cb68fe22d
> --
> 2.27.0
>

Re: [PATCH 0/4] crypto: add sha256() function

[Hans de Goede]

Hi,

On 7/7/20 8:58 PM, Eric Biggers wrote:
> This series adds a function sha256() to the sha256 library so that users
> who want to compute a hash in one step can just call sha256() instead of
> sha256_init() + sha256_update() + sha256_final().
> 
> Patches 2-4 then convert some users to use it.
> 
> Eric Biggers (4):
>    crypto: lib/sha256 - add sha256() function
>    efi: use sha256() instead of open coding
>    mptcp: use sha256() instead of open coding
>    ASoC: cros_ec_codec: use sha256() instead of open coding
> 
>   drivers/firmware/efi/embedded-firmware.c |  9 +++-----
>   include/crypto/sha.h                     |  1 +
>   lib/crypto/sha256.c                      | 10 +++++++++
>   net/mptcp/crypto.c                       | 15 +++----------
>   sound/soc/codecs/cros_ec_codec.c         | 27 ++----------------------
>   5 files changed, 19 insertions(+), 43 deletions(-)
> 
> 
> base-commit: 57c8aa43b9f272c382c253573c82be5cb68fe22d

I've done some quick tests on this series to make sure that
the efi embedded-firmware support did not regress.
That still works fine, so this series is;

Tested-by: Hans de Goede <hdegoede@redhat.com>

Regards,

Hans