netdev - Re: [v4 PATCH 10/13] ubifs: Use crypto

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

Message-ID: <1b6176d7-31a8-7211-b648-a79bd25af6dc@huawei.com>
Date: Sat, 15 Mar 2025 17:27:22 +0800
From: Zhihao Cheng <chengzhihao1@...wei.com>
To: Herbert Xu <herbert@...dor.apana.org.au>
CC: Linux Crypto Mailing List <linux-crypto@...r.kernel.org>, Richard
 Weinberger <richard@....at>, <linux-mtd@...ts.infradead.org>, "Rafael J.
 Wysocki" <rafael@...nel.org>, Pavel Machek <pavel@....cz>,
	<linux-pm@...r.kernel.org>, Steffen Klassert <steffen.klassert@...unet.com>,
	<netdev@...r.kernel.org>
Subject: Re: [v4 PATCH 10/13] ubifs: Use crypto_acomp interface

在 2025/3/15 17:12, Herbert Xu 写道:
> On Sat, Mar 15, 2025 at 05:08:47PM +0800, Zhihao Cheng wrote:
>>
>> According to the warning message, current compressor is zstd. The output
>> buffer size is limited only for LZO compressor by [1].
> 
> Any algorithm can and will produce output longer than the input,
> if you give it enough output buffer.
> 
> Previously an output buffer length of 2x the input length was given
> to all algorithms, meaning that they would all succeed no matter
> whether the input can be compressed or not.
> 
> This has now been changed so that incompressible data is not
> needlessly compressed all the way to the end.  In fact we should
> reduce it further to eliminate the UBIFS_MIN_COMPRESS_DIFF check.

Ah, I get it. Thanks for reminding, and I verify that the root cause is 
the output buffer becomes smaller.
> 
> I will remove the warning on compression since failures are
> expected and reduce the output buffer length further to remove
> the post-compression length check.
> 

I think we should keep the warning, it would be better to distinguish 
the different error types.