linux-kernel - Re: [PATCH 2/7] crypto: stm32

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <a584c152-329e-9c79-ec62-795485302a55@foss.st.com>
Date:   Wed, 12 Jul 2023 09:58:36 +0200
From:   Thomas BOURGOIN <thomas.bourgoin@...s.st.com>
To:     Linus Walleij <linus.walleij@...aro.org>
CC:     Herbert Xu <herbert@...dor.apana.org.au>,
        "David S . Miller" <davem@...emloft.net>,
        Rob Herring <robh+dt@...nel.org>,
        Krzysztof Kozlowski <krzysztof.kozlowski+dt@...aro.org>,
        Conor Dooley <conor+dt@...nel.org>,
        Maxime Coquelin <mcoquelin.stm32@...il.com>,
        Alexandre Torgue <alexandre.torgue@...s.st.com>,
        Lionel Debieve <lionel.debieve@...s.st.com>,
        <linux-crypto@...r.kernel.org>, <devicetree@...r.kernel.org>,
        <linux-stm32@...md-mailman.stormreply.com>,
        <linux-arm-kernel@...ts.infradead.org>,
        <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH 2/7] crypto: stm32 - add new algorithms support

Hi Linus,

Did you run your test only with the patch adding the support for 
STM32MP13 or did you try the whole patch set ?

The error is on the test vector number 4, which is an HASH of 64 bytes 
which is exactly the size of a blcok for SHA1.

Did you try to run the test for SHA256 ? (I guess you will see the same 
error on test vector 4)

I have two hypothesis :
	- it could be related to the padding of the data. (I don't think, the 
patch does not modify these lines)
	- it could come from the way the first block of data is handle. To 
start the first pre computation, we have to write 1 block + 1 word. The 
test only write one block so maybe the issue is here.

On 7/6/23 09:37, Thomas BOURGOIN wrote:
 > +		/*
 > +		 * After first block + 1 words are fill up,
 > +		 * we only need to fill 1 block to start partial computation
 > +		 */
 > +		rctx->state.blocklen -= sizeof(u32);
 > +

I found a typo in the number of CSR to save/restore for the SHA1 and 
SHA256 algorithm. It should be 38 instead of 22.
Tell me if it fixes the regression.

It could be possible to divide the patch in 2 (one patch rework 
preparing MP13 and one patch with the new algorithm) but for the 
upstream I do not know if it is relevant to have 2 patches instead of one.

Best regards,

Thomas

---
  drivers/crypto/stm32/stm32-hash.c | 2 +-
  1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/crypto/stm32/stm32-hash.c 
b/drivers/crypto/stm32/stm32-hash.c
index a50d73c18d5c..88a186c3dd78 100644
--- a/drivers/crypto/stm32/stm32-hash.c
+++ b/drivers/crypto/stm32/stm32-hash.c
@@ -74,7 +74,7 @@

  /* Context swap register */
  #define HASH_CSR_NB_SHA256_HMAC		54
-#define HASH_CSR_NB_SHA256		22
+#define HASH_CSR_NB_SHA256		38
  #define HASH_CSR_NB_SHA512_HMAC		103
  #define HASH_CSR_NB_SHA512		91
  #define HASH_CSR_NB_SHA3_HMAC		88
-- 
2.25.1