OpenVZ Forum: Devel » BC: resource beancounters (v6) (with userpages reclamation + configfs)

Home » Mailing lists » Devel » BC: resource beancounters (v6) (with userpages reclamation + configfs)

Show: Today's Messages :: Show Polls :: Message Navigator
E-mail to friend

Re: [ckrm-tech] [PATCH 4/13] BC: context handling [message #8534 is a reply to message #8532]

Thu, 23 November 2006 10:45

Pavel Emelianov
Messages: 1149
Registered: September 2006

Senior Member

Paul Menage wrote:
> On 11/23/06, Pavel Emelianov <xemul@openvz.org> wrote:
>> Paul Menage wrote:
>> > On 11/23/06, Pavel Emelianov <xemul@openvz.org> wrote:
>> >> You mean moving is like this:
>> >>
>> >> old_bc = task->real_bc;
>> >> task->real_bc = new_bc;
>> >> cmpxchg(&tsk->exec_bc, old_bc, new_bc);
>> >>
>> >> ? Then this won't work:
>> >>
>> >> Initialisation:
>> >> current->exec_bc = init_bc;
>> >> current->real_bc = init_bc;
>> >> ...
>> >> IRQ:
>> >> current->exec_bc = init_bc;
>> >> ...
>> >> old_bc = tsk->real_bc; /* init_bc */
>> >> tsk->real_bc = bc1;
>> >> cx(tsk->exec_bc, init_bc, bc1); /* ok */
>> >> ...
>> >> Here at the middle of an interrupt
>> >> we have bc1 set as exec_bc on task
>> >> which IS wrong!
>> >
>> > You could get round that by having a separate "irq_bc" that's never
>> > valid for a task not in an interrupt.
>>
>> No no no. This is not what is needed. You see, we do have to
>> set exec_bc as temporary (and atomic) context. Having temporary
>> context is 1. flexible 2. needed by beancounters' network accountig.
>
> I don't see why having an irq_bc wouldn't solve this. At the start of
> the interrupt handler, set current->exec_bc to &irq_bc; at the end set
> it to current->real_bc; use the cmpxchg() that I suggested to ensure
> that you never update task->exec_bc from another task if it's not
> equal to task->real_bc; use RCU to ensure that a beancounter is never
> freed while someone might be accessing it.

Oh, I see. I just didn't get your idea. This will work, but
1. we separate interrupt accounting from all the others'
2. for interrupts only. In case we want to set init_bc as
temporary context all will be broken...

We need some generic solution independent from what
exactly is set as temporary exec_bc.

>>
>> Maybe we can make smth similar to wait_task_inactive and change
>> it's beancounter before unlocking the runqueue?
>
> That could work too.

Could work, but whether everyone will like such intrusion...
I agree that stop_machine isn't nicer. This is a temporary
solution that works for sure. Better one will follow...

Report message to a moderator

[Message index]

		BC: resource beancounters (v6) (with userpages reclamation + configfs) By: dev on Thu, 09 November 2006 16:42
		[PATCH 1/13] BC: atomic_dec_and_lock_irqsave() helper By: dev on Thu, 09 November 2006 16:47
		Re: [PATCH 1/13] BC: atomic_dec_and_lock_irqsave() helper By: Cedric Le Goater on Fri, 10 November 2006 15:19
		Re: [PATCH 1/13] BC: atomic_dec_and_lock_irqsave() helper By: dev on Fri, 10 November 2006 16:40
		[PATCH 2/13] BC: Kconfig and Makefile By: dev on Thu, 09 November 2006 16:47
		[PATCH 3/13] BC: beancounters core and API By: dev on Thu, 09 November 2006 16:49
		[PATCH 4/13] BC: context handling By: dev on Thu, 09 November 2006 16:51
		Re: [ckrm-tech] [PATCH 4/13] BC: context handling By: Paul Menage on Thu, 23 November 2006 07:48
		Re: [ckrm-tech] [PATCH 4/13] BC: context handling By: Pavel Emelianov on Thu, 23 November 2006 08:35
		Re: [ckrm-tech] [PATCH 4/13] BC: context handling By: Paul Menage on Thu, 23 November 2006 08:53
		Re: [ckrm-tech] [PATCH 4/13] BC: context handling By: Pavel Emelianov on Thu, 23 November 2006 09:20
		Re: [ckrm-tech] [PATCH 4/13] BC: context handling By: Paul Menage on Thu, 23 November 2006 09:31
		Re: [ckrm-tech] [PATCH 4/13] BC: context handling By: Pavel Emelianov on Thu, 23 November 2006 09:56
		Re: [ckrm-tech] [PATCH 4/13] BC: context handling By: Paul Menage on Thu, 23 November 2006 10:18
		Re: [ckrm-tech] [PATCH 4/13] BC: context handling By: Pavel Emelianov on Thu, 23 November 2006 10:45
		Re: [ckrm-tech] [PATCH 4/13] BC: context handling By: Pavel Emelianov on Fri, 24 November 2006 10:10
		Re: [ckrm-tech] [PATCH 4/13] BC: context handling By: Paul Menage on Sat, 25 November 2006 00:09
		Re: [ckrm-tech] [PATCH 4/13] BC: context handling By: Pavel Emelianov on Mon, 27 November 2006 08:27
		[PATCH 5/13] BC: configfs interface By: dev on Thu, 09 November 2006 16:52
		[PATCH 6/13] BC: kmemsize accounting (core) By: dev on Thu, 09 November 2006 16:53
		Re: [PATCH 6/13] BC: kmemsize accounting (core) By: Paul Jackson on Thu, 09 November 2006 19:05
		Re: [PATCH 6/13] BC: kmemsize accounting (core) By: Pekka Enberg on Fri, 10 November 2006 22:50
		Re: [PATCH 6/13] BC: kmemsize accounting (core) By: Pekka Enberg on Fri, 10 November 2006 22:46
		Re: [PATCH 6/13] BC: kmemsize accounting (core) By: Pavel Emelianov on Mon, 13 November 2006 12:13
		Re: [PATCH 6/13] BC: kmemsize accounting (core) By: Pavel Machek on Sat, 11 November 2006 05:50
		[PATCH 7/13] BC: kmemsize accounting (hooks) By: dev on Thu, 09 November 2006 16:54
		[PATCH 8/13] BC: privvmpages accounting (core) By: dev on Thu, 09 November 2006 16:56
		[PATCH 9/13] BC: privvmpages accounting (hooks) By: dev on Thu, 09 November 2006 16:57
		[PATCH 10/13] BC: physpages accounting (core) By: dev on Thu, 09 November 2006 16:59
		[PATCH 11/13] BC: physpages accounting (hooks) By: dev on Thu, 09 November 2006 17:01
		[PATCH 12/13] BC: numtasks accounting By: dev on Thu, 09 November 2006 17:02
		[PATCH 13/13] BC: numfiles accounting By: dev on Thu, 09 November 2006 17:03
		Re: BC: resource beancounters (v6) (with userpages reclamation + configfs) By: Herbert Poetzl on Sun, 19 November 2006 19:41

Previous Topic:	Re: Re: [RFC] [PATCH 0/4] uid_ns: introduction
Next Topic:	[PATCH] ext3: small fix for previous retries patch in ext3_prepare_write()

Goto Forum:

-=] Back to Top [=-

[ Syndicate this forum (XML) ] [

] [

]

Current Time: Sun Sep 07 22:16:21 GMT 2025

Total time taken to generate the page: 0.13359 seconds