Re: [PATCH v2] refcount: Create unchecked atomic_t implementation

2017-06-21 Thread Ingo Molnar
* Kees Cook wrote: > Many subsystems will not use refcount_t unless there is a way to build the > kernel so that there is no regression in speed compared to atomic_t. This > adds CONFIG_REFCOUNT_FULL to enable the full refcount_t implementation > which has the validation

Re: [PATCH v2] refcount: Create unchecked atomic_t implementation

2017-06-21 Thread Ingo Molnar
* Kees Cook wrote: > Many subsystems will not use refcount_t unless there is a way to build the > kernel so that there is no regression in speed compared to atomic_t. This > adds CONFIG_REFCOUNT_FULL to enable the full refcount_t implementation > which has the validation but is slightly slower.

Re: [PATCH v2] refcount: Create unchecked atomic_t implementation

2017-06-19 Thread Kees Cook
On Wed, Jun 7, 2017 at 10:56 PM, Greg KH wrote: > On Wed, Jun 07, 2017 at 07:58:31PM -0700, Kees Cook wrote: >> Many subsystems will not use refcount_t unless there is a way to build the >> kernel so that there is no regression in speed compared to atomic_t. This >>

Re: [PATCH v2] refcount: Create unchecked atomic_t implementation

2017-06-19 Thread Kees Cook
On Wed, Jun 7, 2017 at 10:56 PM, Greg KH wrote: > On Wed, Jun 07, 2017 at 07:58:31PM -0700, Kees Cook wrote: >> Many subsystems will not use refcount_t unless there is a way to build the >> kernel so that there is no regression in speed compared to atomic_t. This >> adds CONFIG_REFCOUNT_FULL to

Re: [PATCH v2] refcount: Create unchecked atomic_t implementation

2017-06-09 Thread Peter Zijlstra
On Fri, Jun 09, 2017 at 06:24:04AM +0200, Manfred Spraul wrote: > Hi Davidlohr, > > On 06/08/2017 10:09 PM, Davidlohr Bueso wrote: > > > >Yes, this would be a prerequisite for ipc; which I initially thought > >didn't > >take a performance hit. > > > Did you see a regression for ipc? I'd be most

Re: [PATCH v2] refcount: Create unchecked atomic_t implementation

2017-06-09 Thread Peter Zijlstra
On Fri, Jun 09, 2017 at 06:24:04AM +0200, Manfred Spraul wrote: > Hi Davidlohr, > > On 06/08/2017 10:09 PM, Davidlohr Bueso wrote: > > > >Yes, this would be a prerequisite for ipc; which I initially thought > >didn't > >take a performance hit. > > > Did you see a regression for ipc? I'd be most

Re: [PATCH v2] refcount: Create unchecked atomic_t implementation

2017-06-08 Thread Manfred Spraul
Hi Davidlohr, On 06/08/2017 10:09 PM, Davidlohr Bueso wrote: Yes, this would be a prerequisite for ipc; which I initially thought didn't take a performance hit. Did you see a regression for ipc? The fast paths don't use the refcount, it is only used for rare situations: - GETALL, SETALL

Re: [PATCH v2] refcount: Create unchecked atomic_t implementation

2017-06-08 Thread Manfred Spraul
Hi Davidlohr, On 06/08/2017 10:09 PM, Davidlohr Bueso wrote: Yes, this would be a prerequisite for ipc; which I initially thought didn't take a performance hit. Did you see a regression for ipc? The fast paths don't use the refcount, it is only used for rare situations: - GETALL, SETALL

Re: [PATCH v2] refcount: Create unchecked atomic_t implementation

2017-06-08 Thread Davidlohr Bueso
On Thu, 08 Jun 2017, Reshetova, Elena wrote: On Wed, Jun 07, 2017 at 07:58:31PM -0700, Kees Cook wrote: > Many subsystems will not use refcount_t unless there is a way to build the > kernel so that there is no regression in speed compared to atomic_t. This > adds CONFIG_REFCOUNT_FULL to enable

Re: [PATCH v2] refcount: Create unchecked atomic_t implementation

2017-06-08 Thread Davidlohr Bueso
On Thu, 08 Jun 2017, Reshetova, Elena wrote: On Wed, Jun 07, 2017 at 07:58:31PM -0700, Kees Cook wrote: > Many subsystems will not use refcount_t unless there is a way to build the > kernel so that there is no regression in speed compared to atomic_t. This > adds CONFIG_REFCOUNT_FULL to enable

RE: [PATCH v2] refcount: Create unchecked atomic_t implementation

2017-06-08 Thread Reshetova, Elena
> On Wed, Jun 07, 2017 at 07:58:31PM -0700, Kees Cook wrote: > > Many subsystems will not use refcount_t unless there is a way to build the > > kernel so that there is no regression in speed compared to atomic_t. This > > adds CONFIG_REFCOUNT_FULL to enable the full refcount_t implementation > >

RE: [PATCH v2] refcount: Create unchecked atomic_t implementation

2017-06-08 Thread Reshetova, Elena
> On Wed, Jun 07, 2017 at 07:58:31PM -0700, Kees Cook wrote: > > Many subsystems will not use refcount_t unless there is a way to build the > > kernel so that there is no regression in speed compared to atomic_t. This > > adds CONFIG_REFCOUNT_FULL to enable the full refcount_t implementation > >

Re: [PATCH v2] refcount: Create unchecked atomic_t implementation

2017-06-08 Thread Christoph Hellwig
On Wed, Jun 07, 2017 at 07:58:31PM -0700, Kees Cook wrote: > Many subsystems will not use refcount_t unless there is a way to build the > kernel so that there is no regression in speed compared to atomic_t. This > adds CONFIG_REFCOUNT_FULL to enable the full refcount_t implementation > which has

Re: [PATCH v2] refcount: Create unchecked atomic_t implementation

2017-06-08 Thread Christoph Hellwig
On Wed, Jun 07, 2017 at 07:58:31PM -0700, Kees Cook wrote: > Many subsystems will not use refcount_t unless there is a way to build the > kernel so that there is no regression in speed compared to atomic_t. This > adds CONFIG_REFCOUNT_FULL to enable the full refcount_t implementation > which has

Re: [PATCH v2] refcount: Create unchecked atomic_t implementation

2017-06-07 Thread Greg KH
On Wed, Jun 07, 2017 at 07:58:31PM -0700, Kees Cook wrote: > Many subsystems will not use refcount_t unless there is a way to build the > kernel so that there is no regression in speed compared to atomic_t. This > adds CONFIG_REFCOUNT_FULL to enable the full refcount_t implementation > which has

Re: [PATCH v2] refcount: Create unchecked atomic_t implementation

2017-06-07 Thread Greg KH
On Wed, Jun 07, 2017 at 07:58:31PM -0700, Kees Cook wrote: > Many subsystems will not use refcount_t unless there is a way to build the > kernel so that there is no regression in speed compared to atomic_t. This > adds CONFIG_REFCOUNT_FULL to enable the full refcount_t implementation > which has

[PATCH v2] refcount: Create unchecked atomic_t implementation

2017-06-07 Thread Kees Cook
Many subsystems will not use refcount_t unless there is a way to build the kernel so that there is no regression in speed compared to atomic_t. This adds CONFIG_REFCOUNT_FULL to enable the full refcount_t implementation which has the validation but is slightly slower. When not enabled, refcount_t

[PATCH v2] refcount: Create unchecked atomic_t implementation

2017-06-07 Thread Kees Cook
Many subsystems will not use refcount_t unless there is a way to build the kernel so that there is no regression in speed compared to atomic_t. This adds CONFIG_REFCOUNT_FULL to enable the full refcount_t implementation which has the validation but is slightly slower. When not enabled, refcount_t