linux-kernel - Re: [PATCH 1/3] nvme-tcp: rate limit error message in send path

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: <20250131072947.GB16012@lst.de>
Date: Fri, 31 Jan 2025 08:29:47 +0100
From: Christoph Hellwig <hch@....de>
To: Daniel Wagner <dwagner@...e.de>
Cc: Christoph Hellwig <hch@....de>, Daniel Wagner <wagi@...nel.org>,
	Keith Busch <kbusch@...nel.org>, Jens Axboe <axboe@...nel.dk>,
	Sagi Grimberg <sagi@...mberg.me>, Ming Lei <ming.lei@...hat.com>,
	James Smart <james.smart@...adcom.com>,
	Hannes Reinecke <hare@...e.de>, linux-nvme@...ts.infradead.org,
	linux-kernel@...r.kernel.org, linux-block@...r.kernel.org
Subject: Re: [PATCH 1/3] nvme-tcp: rate limit error message in send path

On Thu, Jan 30, 2025 at 04:25:35PM +0100, Daniel Wagner wrote:
> On Wed, Jan 29, 2025 at 07:05:34AM +0100, Christoph Hellwig wrote:
> > On Tue, Jan 28, 2025 at 05:34:46PM +0100, Daniel Wagner wrote:
> > > If a lot of request are in the queue, this message is spamming the logs,
> > > thus rate limit it.
> > 
> > Are in the queue when what happens?  Not that I'm against this,
> > but if we have a known condition where this error is printed a lot
> > we should probably skip it entirely for that?
> 
> The condition is that all the elements in the queue->send_list could fail as a
> batch. I had a bug in my patches which re-queued all the failed command
> immediately and semd them out again, thus spamming the log.
> 
> This behavior doesn't exist in upstream. I just thought it might make
> sense to rate limit as precaution. I don't know if it is worth the code
> churn.

I'm fine with the rate limiting.  I was just wondering if there is
a case where we'd easily hit it and could do even better.