All of lore.kernel.org
 help / color / mirror / Atom feed
* How to invalidate all cache
@ 2014-10-13  3:03 Zheng Liu
  2014-10-13  3:16 ` Slava Pestov
  0 siblings, 1 reply; 6+ messages in thread
From: Zheng Liu @ 2014-10-13  3:03 UTC (permalink / raw)
  To: linux-bcache; +Cc: Zheng Liu

Hi all,

I am trying to do some benchmarks against bcache, espeically in cache
miss scenario, but I don't find a user interface to invalidate all
cache.  Anyone can tell me how to do that.

Thanks in advance,
                                                - Zheng

^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: How to invalidate all cache
  2014-10-13  3:03 How to invalidate all cache Zheng Liu
@ 2014-10-13  3:16 ` Slava Pestov
  2014-10-13  3:45   ` Zheng Liu
  0 siblings, 1 reply; 6+ messages in thread
From: Slava Pestov @ 2014-10-13  3:16 UTC (permalink / raw)
  To: Zheng Liu; +Cc: linux-bcache, Zheng Liu

You can try to detach the backing device, which will cause all dirty
data to be written back, and attach it again.

On Sun, Oct 12, 2014 at 8:03 PM, Zheng Liu <gnehzuil.liu@gmail.com> wrote:
> Hi all,
>
> I am trying to do some benchmarks against bcache, espeically in cache
> miss scenario, but I don't find a user interface to invalidate all
> cache.  Anyone can tell me how to do that.
>
> Thanks in advance,
>                                                 - Zheng
> --
> To unsubscribe from this list: send the line "unsubscribe linux-bcache" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re[2]: How to invalidate all cache
  2014-10-13  3:45   ` Zheng Liu
@ 2014-10-13  3:43     ` Pavel Goran
  2014-10-13 13:30       ` Zheng Liu
  0 siblings, 1 reply; 6+ messages in thread
From: Pavel Goran @ 2014-10-13  3:43 UTC (permalink / raw)
  To: linux-bcache

Hello Zheng,

Monday, October 13, 2014, 10:45:37 AM, you wrote:

> Yes, detaching a backing device will make all dirty data to be written
> back to the backing device.  But if I understand correctly it couldn't
> invalidate all data in cache device.  That means that after attaching
> backing device the data in cache device is still valid.

It  would  be  a  bug if cached data were considered valid after detaching and
re-attaching. Because while a cache device an a backing device were separated,
data  could be written to a backing device, this would possibly make data in a
"disconnected" cache obsolete.

Pavel Goran
  

^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: How to invalidate all cache
  2014-10-13  3:16 ` Slava Pestov
@ 2014-10-13  3:45   ` Zheng Liu
  2014-10-13  3:43     ` Re[2]: " Pavel Goran
  0 siblings, 1 reply; 6+ messages in thread
From: Zheng Liu @ 2014-10-13  3:45 UTC (permalink / raw)
  To: Slava Pestov; +Cc: linux-bcache, Zheng Liu

Hi Slava,

Thanks for your reply.

On Sun, Oct 12, 2014 at 08:16:31PM -0700, Slava Pestov wrote:
> You can try to detach the backing device, which will cause all dirty
> data to be written back, and attach it again.

Yes, detaching a backing device will make all dirty data to be written
back to the backing device.  But if I understand correctly it couldn't
invalidate all data in cache device.  That means that after attaching
backing device the data in cache device is still valid.  All I want to
do is to remove all data in cache device.  Any idea?

Regards,
                                                - Zheng

> 
> On Sun, Oct 12, 2014 at 8:03 PM, Zheng Liu <gnehzuil.liu@gmail.com> wrote:
> > Hi all,
> >
> > I am trying to do some benchmarks against bcache, espeically in cache
> > miss scenario, but I don't find a user interface to invalidate all
> > cache.  Anyone can tell me how to do that.
> >
> > Thanks in advance,
> >                                                 - Zheng
> > --
> > To unsubscribe from this list: send the line "unsubscribe linux-bcache" in
> > the body of a message to majordomo@vger.kernel.org
> > More majordomo info at  http://vger.kernel.org/majordomo-info.html

^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: How to invalidate all cache
  2014-10-13  3:43     ` Re[2]: " Pavel Goran
@ 2014-10-13 13:30       ` Zheng Liu
  2014-10-13 18:13         ` (unknown) Eric Wheeler
  0 siblings, 1 reply; 6+ messages in thread
From: Zheng Liu @ 2014-10-13 13:30 UTC (permalink / raw)
  To: Pavel Goran; +Cc: linux-bcache

On Mon, Oct 13, 2014 at 10:43:14AM +0700, Pavel Goran wrote:
> Hello Zheng,
> 
> Monday, October 13, 2014, 10:45:37 AM, you wrote:
> 
> > Yes, detaching a backing device will make all dirty data to be written
> > back to the backing device.  But if I understand correctly it couldn't
> > invalidate all data in cache device.  That means that after attaching
> > backing device the data in cache device is still valid.
> 
> It  would  be  a  bug if cached data were considered valid after detaching and
> re-attaching. Because while a cache device an a backing device were separated,
> data  could be written to a backing device, this would possibly make data in a
> "disconnected" cache obsolete.

Ah, Thanks for pointing it out.  I got it.

BTW, there is a bug when I tried to detach/attach a backing device.
A patch has sent out to the mailing list.  That would be great if some
one can review it.

Regards,
                                                - Zheng

^ permalink raw reply	[flat|nested] 6+ messages in thread

* (unknown)
  2014-10-13 13:30       ` Zheng Liu
@ 2014-10-13 18:13         ` Eric Wheeler
  0 siblings, 0 replies; 6+ messages in thread
From: Eric Wheeler @ 2014-10-13 18:13 UTC (permalink / raw)
  To: linux-bcache

unsubscribe linux-bcache

^ permalink raw reply	[flat|nested] 6+ messages in thread

end of thread, other threads:[~2014-10-13 18:13 UTC | newest]

Thread overview: 6+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2014-10-13  3:03 How to invalidate all cache Zheng Liu
2014-10-13  3:16 ` Slava Pestov
2014-10-13  3:45   ` Zheng Liu
2014-10-13  3:43     ` Re[2]: " Pavel Goran
2014-10-13 13:30       ` Zheng Liu
2014-10-13 18:13         ` (unknown) Eric Wheeler

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.