Eric Dumazet wrote:
On 5/3/19 2:07 AM, Steffen Klassert wrote:
> > On Wed, Apr 24, 2019 at 12:40:23PM +0200, Florian Westphal wrote:
> >> I'm not sure this is a good idea to begin with, refcount
> >> is right next to state spinlock which is taken for both tx and rx ops,
> >> plus this complicates
On 5/3/19 2:07 AM, Steffen Klassert wrote:
> On Wed, Apr 24, 2019 at 12:40:23PM +0200, Florian Westphal wrote:
>> I'm not sure this is a good idea to begin with, refcount
>> is right next to state spinlock which is taken for both tx and rx ops,
>> plus this complicates debugging quite a bit.
>
> -Original Message-
> From: Steffen Klassert
> Sent: Friday, May 3, 2019 12:16 PM
> To: Vakul Garg
> Cc: Florian Westphal ; netdev@vger.kernel.org
> Subject: Re: [RFC HACK] xfrm: make state refcounting percpu
>
> On Fri, May 03, 2019 at 06:34:29AM
On Fri, May 03, 2019 at 06:34:29AM +, Vakul Garg wrote:
> > -Original Message-
> > From: Steffen Klassert
> >
> > Also, is this a new problem or was it always like that?
>
> It is always like this. On 4-core, 8-core platforms as well, these atomics
> consume significant cpu
> (8 co
> -Original Message-
> From: Steffen Klassert
> Sent: Friday, May 3, 2019 11:52 AM
> To: Vakul Garg
> Cc: Florian Westphal ; netdev@vger.kernel.org
> Subject: Re: [RFC HACK] xfrm: make state refcounting percpu
>
> On Fri, May 03, 2019 at 06:13:22AM
On Fri, May 03, 2019 at 06:13:22AM +, Vakul Garg wrote:
>
>
> > -Original Message-
> > From: Steffen Klassert
> > Sent: Friday, May 3, 2019 11:38 AM
> > To: Florian Westphal
> > Cc: Vakul Garg ; netdev@vger.kernel.org
> > Subject: Re: [RFC
> -Original Message-
> From: Steffen Klassert
> Sent: Friday, May 3, 2019 11:38 AM
> To: Florian Westphal
> Cc: Vakul Garg ; netdev@vger.kernel.org
> Subject: Re: [RFC HACK] xfrm: make state refcounting percpu
>
> On Wed, Apr 24, 2019 at 12:40:23PM +0200
On Wed, Apr 24, 2019 at 12:40:23PM +0200, Florian Westphal wrote:
> I'm not sure this is a good idea to begin with, refcount
> is right next to state spinlock which is taken for both tx and rx ops,
> plus this complicates debugging quite a bit.
Hm, what would be the usecase where this could help?
On 04/24/2019 03:40 AM, Florian Westphal wrote:
> I'm not sure this is a good idea to begin with, refcount
> is right next to state spinlock which is taken for both tx and rx ops,
> plus this complicates debugging quite a bit.
>
Why not simply using include/linux/percpu-refcount.h ?
> Also e
