linux-kernel - Re: [BUG] crypto: tcrypt - data corruption in ahash tests with CRYPTO_AHASH_ALG_BLOCK

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: Windows password security audit tool. GUI, reports in PDF.

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

Message-ID: <f9890350-cbe2-4cff-b5f2-1c76cd42ab4b@ti.com>
Date: Fri, 16 Jan 2026 14:03:19 +0530
From: T Pratham <t-pratham@...com>
To: Herbert Xu <herbert@...dor.apana.org.au>
CC: "David S. Miller" <davem@...emloft.net>, <linux-crypto@...r.kernel.org>,
	<linux-kernel@...r.kernel.org>, Manorit Chawdhry <m-chawdhry@...com>, "Shiva
 Tripathi" <s-tripathi1@...com>, Kamlesh Gurudasani <kamlesh@...com>
Subject: Re: [BUG] crypto: tcrypt - data corruption in ahash tests with
 CRYPTO_AHASH_ALG_BLOCK_ONLY

On 16/01/26 08:00, Herbert Xu wrote:
> On Thu, Jan 15, 2026 at 03:38:50PM +0530, T Pratham wrote:
>>
>> However, in tcrypt, the wait struct is accessed as below:
>>
>> crypto/tcrypt.c:
>> static inline int do_one_ahash_op(struct ahash_request *req, int ret)
>> {
>> 	struct crypto_wait *wait = req->base.data;
> 
> This is just broken.  Even before the partial hash block handling
> there are code paths where req->base.data will be modified prior
> to final return.  That's why the completion function takes a void *
> instead of just passing the request back.

Yeah, I realized this a couple of hours after hitting the send button.
Only that partial block handling is the first place where I hit this.

> 
> So we should fix tcrypt to not do this crazy thing.

Yup, looking forward for this.

> 
> Thanks,

-- 
Regards
T Pratham <t-pratham@...com>