[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <fb2e6c9d-028c-3ed8-1068-0e654bdb63a4@kernel.dk>
Date: Mon, 24 Jul 2023 12:22:07 -0600
From: Jens Axboe <axboe@...nel.dk>
To: Phil Elwell <phil@...pberrypi.com>
Cc: Greg KH <gregkh@...uxfoundation.org>, andres@...razel.de,
asml.silence@...il.com, david@...morbit.com, hch@....de,
io-uring@...r.kernel.org, LKML <linux-kernel@...r.kernel.org>,
linux-xfs@...r.kernel.org, stable <stable@...r.kernel.org>
Subject: Re: [PATCH] io_uring: Use io_schedule* in cqring wait
On 7/24/23 10:48 AM, Phil Elwell wrote:
> Hi Jens,
>
> On Mon, 24 Jul 2023 at 17:08, Jens Axboe <axboe@...nel.dk> wrote:
>>
>> On 7/24/23 10:07?AM, Phil Elwell wrote:
>>>> Even though I don't think this is an actual problem, it is a bit
>>>> confusing that you get 100% iowait while waiting without having IO
>>>> pending. So I do think the suggested patch is probably worthwhile
>>>> pursuing. I'll post it and hopefully have Andres test it too, if he's
>>>> available.
>>>
>>> If you CC me I'll happily test it for you.
>>
>> Here it is.
>
> < snip >
>
> Thanks, that works for me on top of 6.5-rc3. Going to 6.1 is a
> non-trivial (for me) back-port - the switch from "ret = 0" in 6.5 to
> "ret = 1" in 6.1 is surprising.
Great, thanks for testing. I'll take care of the stable backports
once the patch lands in upstream -git later this week.
--
Jens Axboe
Powered by blists - more mailing lists