[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <CACVXFVOr+nFvFge-VOR+Jsci2hp263r7F-NZ8N2d5wsfDDOpWg@mail.gmail.com>
Date: Wed, 6 Jul 2016 23:39:29 +0800
From: Ming Lei <tom.leiming@...il.com>
To: Tejun Heo <tj@...nel.org>
Cc: aaronlee0817 <aaronlee0817@....com>, Jens Axboe <axboe@...nel.dk>,
Ming Lin <mlin@...nel.org>, Shaohua Li <shli@...nel.org>,
Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
cgroup@...r.kernel.org, jiale0817.li@...sung.com,
yanzi.zhang@...sung.com, zhen1.zhang@...sung.com,
linux-block <linux-block@...r.kernel.org>
Subject: Re: Re: cgroup: Fix split bio been throttled more than once
On Wed, Jul 6, 2016 at 11:06 PM, Tejun Heo <tj@...nel.org> wrote:
> Hello,
>
> On Wed, Jul 06, 2016 at 10:58:55PM +0800, aaronlee0817 wrote:
>> >But that said, can't we just copy BIO_THROTLED while splitting?
>>
>> In my opinion, copying BIO_THROTLED while splitting doesn't work
>> with this, because, the bio within the limit will not set BIO_THROTLED
>> but will be split, and back to call generic_make_request(), and will be
>> count twice.
>
> I see. It's just a bit weird that we're marking this from clone site
> when it's an issue which is specific to throttling. The thing is
> depending on how bio clone is used, it could misbehave in the other
The cloned bio often need to be submitted via generic_make_request().
> direction (bio not accounted at all). What we want to know is whether
> a bio has been previously accounted or not, rather than whether a bio
> is a result of splitting. I think it'd be better to track that
> instead.
Looks the flag of BIO_THROTLED doesn't work well, and it should have
been set if the bio is accounted, and seems not related with splitting.
Thanks,
Ming
>
> Thanks.
>
> --
> tejun
> --
> To unsubscribe from this list: send the line "unsubscribe linux-block" in
> the body of a message to majordomo@...r.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
--
Ming Lei
Powered by blists - more mailing lists