[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20240905105552.GE15400@noisy.programming.kicks-ass.net>
Date: Thu, 5 Sep 2024 12:55:52 +0200
From: Peter Zijlstra <peterz@...radead.org>
To: Jens Axboe <axboe@...nel.dk>
Cc: linux-kernel@...r.kernel.org, tglx@...utronix.de
Subject: Re: [PATCH 4/4] sched/core: split iowait state into two states
On Mon, Aug 19, 2024 at 09:39:49AM -0600, Jens Axboe wrote:
> iowait is a bogus metric, but it's helpful in the sense that it allows
> short waits to not enter sleep states that have a higher exit latency
> than would've otherwise have been picked for iowait'ing tasks. However,
> it's harmless in that lots of applications and monitoring assumes that
> iowait is busy time, or otherwise use it as a health metric.
> Particularly for async IO it's entirely nonsensical.
Holdup... I should've remembered this. You're using ->in_iowait for
io_uring. And that is what started all this.
Now, having in_iowait set when you're waiting on a futex is utterly
insane, but looking at commit 7b72d661f1f2 ("io_uring: gate iowait
schedule on having pending requests") you're now only actually setting
in_iowait when you have pending IO.
And I don't think that is nonsensical as you write above.
You are after all still actually waiting while there is pending IO, no?
Powered by blists - more mailing lists