Message-ID: <7fdf700c-6747-fa58-ee7a-2cca6397fa05@linux.alibaba.com>
Date:   Wed, 1 Feb 2023 17:43:00 +0800
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.15; rv:102.0)
 Gecko/20100101 Thunderbird/102.6.1
Subject: Re: [PATCH] crypto: arm64/aes-ccm - Rewrite skcipher walker loop
Content-Language: en-US
To:     Herbert Xu <herbert@gondor.apana.org.au>
Cc:     Linux Crypto Mailing List <linux-crypto@vger.kernel.org>,
        Ard Biesheuvel <ardb@kernel.org>
References: <Y9eGyzZ+JAqRQvtm@gondor.apana.org.au>
 <b83ca139-1e8c-60f3-939f-15f727710c36@linux.alibaba.com>
 <Y9opEyTFKXAVjk/D@gondor.apana.org.au>
 <477c73bd-b56e-3c63-1ad6-6a2a08af42af@linux.alibaba.com>
 <Y9ovNyYP03rUBPlq@gondor.apana.org.au>
From:   Tianjia Zhang <tianjia.zhang@linux.alibaba.com>
In-Reply-To: <Y9ovNyYP03rUBPlq@gondor.apana.org.au>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
Precedence: bulk

Hi Herbert,

On 2/1/23 5:21 PM, Herbert Xu wrote:
> On Wed, Feb 01, 2023 at 05:15:23PM +0800, Tianjia Zhang wrote:
>>
>> According to your previous reply, walker will ensure that the nbytes of
>> each iteration is at least the size of the chunk. If walk.nbytes == 0,
> 
> walk.nbytes == 0 is used to indicate error.  Of course you could
> check for an error return in addition to checking walk.nbytes but
> that's how this bug got created in the first place.
> 
> So always check for walk.nbytes == 0.
> 
> Cheers,

It seems that only need to fix the loop condition, so if change the
loop condition of the code just now to
while (walk.nbytes && walk.nbytes != walk.total), in this way, the
last chunk cryption is separated out of the loop, which will be clearer
logically. What is your opinion?

Thanks,
Tianjia