netdev - RE: [PATCH net-next 1/2] netdev: Add queue stats for TX stop and wake

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]

Date: Thu, 9 May 2024 21:19:52 +0000
From: Dan Jurgens <danielj@...dia.com>
To: Andrew Lunn <andrew@...n.ch>
CC: "netdev@...r.kernel.org" <netdev@...r.kernel.org>, "mst@...hat.com"
	<mst@...hat.com>, "jasowang@...hat.com" <jasowang@...hat.com>,
	"xuanzhuo@...ux.alibaba.com" <xuanzhuo@...ux.alibaba.com>,
	"virtualization@...ts.linux.dev" <virtualization@...ts.linux.dev>,
	"davem@...emloft.net" <davem@...emloft.net>, "edumazet@...gle.com"
	<edumazet@...gle.com>, "kuba@...nel.org" <kuba@...nel.org>,
	"pabeni@...hat.com" <pabeni@...hat.com>, Jiri Pirko <jiri@...dia.com>
Subject: RE: [PATCH net-next 1/2] netdev: Add queue stats for TX stop and wake

> From: Andrew Lunn <andrew@...n.ch>
> Sent: Thursday, May 9, 2024 3:47 PM
> To: Dan Jurgens <danielj@...dia.com>
> Subject: Re: [PATCH net-next 1/2] netdev: Add queue stats for TX stop and
> wake
> 
> On Thu, May 09, 2024 at 11:32:15AM -0500, Daniel Jurgens wrote:
> > TX queue stop and wake are counted by some drivers.
> > Support reporting these via netdev-genl queue stats.
> >
> > +        name: tx-wake
> > +        doc: |
> > +          Number of times the tx queue was restarted.
> > +        type: uint
> 
> I'm curious where these names came from. The opposite of stop would be
> start. The opposite of wake would be sleep. Are these meant to be
> opposites of each other? If they are opposites, why would they differ by
> more than 1? And if they can only differ by 1, why do we need both?

The names come from the API. netif_tx_stop_queue, netif_tx_wake_queue. It's true that they can only ever differ by 1, but when they do that's interesting.  Though eventually a TX timeout will occur if it's due to something like a lost interrupt.

The most useful thing is knowing if queues are being stopped frequently, so if there's objection to the wake side it can be dropped.

> 
> 	Andrew