linux-kernel - Re: [PATCH v1] printk_ringbuffer: don't needlessly wrap data blocks around

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: <aLmicQkB5RRJaxCE@pathway>
Date: Thu, 4 Sep 2025 16:30:09 +0200
From: Petr Mladek <pmladek@...e.com>
To: Daniil Tatianin <d-tatianin@...dex-team.ru>
Cc: linux-kernel@...r.kernel.org, Steven Rostedt <rostedt@...dmis.org>,
	John Ogness <john.ogness@...utronix.de>,
	Sergey Senozhatsky <senozhatsky@...omium.org>
Subject: Re: [PATCH v1] printk_ringbuffer: don't needlessly wrap data blocks
 around

On Wed 2025-09-03 03:10:08, Daniil Tatianin wrote:
> Previously, data blocks that perfectly fit the data ring buffer would
> get wrapped around to the beginning for no reason since the calculated
> offset of the next data block would belong to the next wrap. Since this
> offset is not actually part of the data block, but rather the offset of
> where the next data block is going to start, there is no reason to
> include it when deciding whether the current block fits the buffer.

I am afraid to touch this code. I am curious how you found this ;-)

> diff --git a/kernel/printk/printk_ringbuffer.c b/kernel/printk/printk_ringbuffer.c
> index d9fb053cff67..f885ba8be5e6 100644
> --- a/kernel/printk/printk_ringbuffer.c
> +++ b/kernel/printk/printk_ringbuffer.c
> @@ -1002,6 +1002,18 @@ static bool desc_reserve(struct printk_ringbuffer *rb, unsigned long *id_out)
>  	return true;
>  }
>  
> +static bool same_lpos_wraps(struct prb_data_ring *data_ring,
> +			     unsigned long begin_lpos, unsigned long next_lpos)

The name might be slightly confusing because it might return true
even when the two given values do not belong into the same wrap.

I would make more clear that it is about the data block starting
on begin_lpos. Either I would call it

	same_wrap_data_block()

or I would invert the logic and call it:

	is_wrapped_data_block()

The 2nd variant looks more self-explanatory to me.


Otherwise the patch looks good to me. I was not able to find
any scenario when it would blow up just by reading the code.

Well, I would still like to do some tests.

Best Regards,
Petr