linux-kernel - Re: [PATCH printk v2 3/4] printk: Skip unfinalized records in panic

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: <ZS_ahOpCQGNjDgSm@alley>
Date:   Wed, 18 Oct 2023 15:15:48 +0200
From:   Petr Mladek <pmladek@...e.com>
To:     John Ogness <john.ogness@...utronix.de>
Cc:     Sergey Senozhatsky <senozhatsky@...omium.org>,
        Steven Rostedt <rostedt@...dmis.org>,
        Thomas Gleixner <tglx@...utronix.de>,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH printk v2 3/4] printk: Skip unfinalized records in panic

On Wed 2023-10-18 10:30:46, John Ogness wrote:
> On 2023-10-17, John Ogness <john.ogness@...utronix.de> wrote:
> >>> -	if (!prb_read_valid(prb, seq, &r))
> >>> -		return false;
> >>> +	while (!prb_read_valid(prb, seq, &r)) {
> >>> +		if (this_cpu_in_panic() && seq < prb_next_seq(prb)) {
> >>> +			/*
> >>> +			 * The record @seq is not finalized and there may be
> 
> I am wondering if we should put the this_cpu_in_panic() check into
> _prb_read_valid() instead. This has the advantage that it could more
> intelligently skip non-finalized records and not need to rely on
> prb_next_seq(). Also, it would allow any other readers (kmsg_dump) to
> skip non-finalized records during panic.

Yeah, this look like a better solution. Anyway, we need to try reading
the messages behind prb_next_seq() up to the head of the ring buffer.
And @head_id is not available outside prb API.

> Looking over the code, it would be easy to extend
> desc_read_finalized_seq() to return a separate error code for exists but
> is non-finalized. This would make it trivial for _prb_read_valid() to
> skip over it when this_cpu_in_panic().

Sounds reasonable.

Best Regards,
Petr