Re: [PATCH 8/7] net/netfilter/nf_conntrack_core: Remove another memory barrier

2016-09-06 Thread Will Deacon
On Mon, Sep 05, 2016 at 08:57:19PM +0200, Manfred Spraul wrote: > On 09/02/2016 09:22 PM, Peter Zijlstra wrote: > Anyone around with a ppc or arm? How slow is the loop of the > spin_unlock_wait() calls? > Single CPU is sufficient. > > Question 1: How large is the difference between: >

Re: [PATCH 8/7] net/netfilter/nf_conntrack_core: Remove another memory barrier

2016-09-06 Thread Will Deacon
On Mon, Sep 05, 2016 at 08:57:19PM +0200, Manfred Spraul wrote: > On 09/02/2016 09:22 PM, Peter Zijlstra wrote: > Anyone around with a ppc or arm? How slow is the loop of the > spin_unlock_wait() calls? > Single CPU is sufficient. > > Question 1: How large is the difference between: >

Re: [PATCH 8/7] net/netfilter/nf_conntrack_core: Remove another memory barrier

2016-09-05 Thread Manfred Spraul
Hi Peter, On 09/02/2016 09:22 PM, Peter Zijlstra wrote: On Fri, Sep 02, 2016 at 08:35:55AM +0200, Manfred Spraul wrote: On 09/01/2016 06:41 PM, Peter Zijlstra wrote: On Thu, Sep 01, 2016 at 04:30:39PM +0100, Will Deacon wrote: On Thu, Sep 01, 2016 at 05:27:52PM +0200, Manfred Spraul wrote:

Re: [PATCH 8/7] net/netfilter/nf_conntrack_core: Remove another memory barrier

2016-09-05 Thread Manfred Spraul
Hi Peter, On 09/02/2016 09:22 PM, Peter Zijlstra wrote: On Fri, Sep 02, 2016 at 08:35:55AM +0200, Manfred Spraul wrote: On 09/01/2016 06:41 PM, Peter Zijlstra wrote: On Thu, Sep 01, 2016 at 04:30:39PM +0100, Will Deacon wrote: On Thu, Sep 01, 2016 at 05:27:52PM +0200, Manfred Spraul wrote:

Re: [PATCH 8/7] net/netfilter/nf_conntrack_core: Remove another memory barrier

2016-09-02 Thread Manfred Spraul
On 09/02/2016 09:22 PM, Peter Zijlstra wrote: On Fri, Sep 02, 2016 at 08:35:55AM +0200, Manfred Spraul wrote: On 09/01/2016 06:41 PM, Peter Zijlstra wrote: On Thu, Sep 01, 2016 at 04:30:39PM +0100, Will Deacon wrote: On Thu, Sep 01, 2016 at 05:27:52PM +0200, Manfred Spraul wrote: Since

Re: [PATCH 8/7] net/netfilter/nf_conntrack_core: Remove another memory barrier

2016-09-02 Thread Manfred Spraul
On 09/02/2016 09:22 PM, Peter Zijlstra wrote: On Fri, Sep 02, 2016 at 08:35:55AM +0200, Manfred Spraul wrote: On 09/01/2016 06:41 PM, Peter Zijlstra wrote: On Thu, Sep 01, 2016 at 04:30:39PM +0100, Will Deacon wrote: On Thu, Sep 01, 2016 at 05:27:52PM +0200, Manfred Spraul wrote: Since

Re: [PATCH 8/7] net/netfilter/nf_conntrack_core: Remove another memory barrier

2016-09-02 Thread Peter Zijlstra
On Fri, Sep 02, 2016 at 08:35:55AM +0200, Manfred Spraul wrote: > On 09/01/2016 06:41 PM, Peter Zijlstra wrote: > >On Thu, Sep 01, 2016 at 04:30:39PM +0100, Will Deacon wrote: > >>On Thu, Sep 01, 2016 at 05:27:52PM +0200, Manfred Spraul wrote: > >>>Since spin_unlock_wait() is defined as equivalent

Re: [PATCH 8/7] net/netfilter/nf_conntrack_core: Remove another memory barrier

2016-09-02 Thread Peter Zijlstra
On Fri, Sep 02, 2016 at 08:35:55AM +0200, Manfred Spraul wrote: > On 09/01/2016 06:41 PM, Peter Zijlstra wrote: > >On Thu, Sep 01, 2016 at 04:30:39PM +0100, Will Deacon wrote: > >>On Thu, Sep 01, 2016 at 05:27:52PM +0200, Manfred Spraul wrote: > >>>Since spin_unlock_wait() is defined as equivalent

Re: [PATCH 8/7] net/netfilter/nf_conntrack_core: Remove another memory barrier

2016-09-02 Thread Manfred Spraul
On 09/01/2016 06:41 PM, Peter Zijlstra wrote: On Thu, Sep 01, 2016 at 04:30:39PM +0100, Will Deacon wrote: On Thu, Sep 01, 2016 at 05:27:52PM +0200, Manfred Spraul wrote: Since spin_unlock_wait() is defined as equivalent to spin_lock(); spin_unlock(), the memory barrier before

Re: [PATCH 8/7] net/netfilter/nf_conntrack_core: Remove another memory barrier

2016-09-02 Thread Manfred Spraul
On 09/01/2016 06:41 PM, Peter Zijlstra wrote: On Thu, Sep 01, 2016 at 04:30:39PM +0100, Will Deacon wrote: On Thu, Sep 01, 2016 at 05:27:52PM +0200, Manfred Spraul wrote: Since spin_unlock_wait() is defined as equivalent to spin_lock(); spin_unlock(), the memory barrier before

Re: [PATCH 8/7] net/netfilter/nf_conntrack_core: Remove another memory barrier

2016-09-02 Thread Boqun Feng
Hi Manfred, On Thu, Sep 01, 2016 at 06:41:26PM +0200, Peter Zijlstra wrote: > On Thu, Sep 01, 2016 at 04:30:39PM +0100, Will Deacon wrote: > > On Thu, Sep 01, 2016 at 05:27:52PM +0200, Manfred Spraul wrote: > > > Since spin_unlock_wait() is defined as equivalent to spin_lock(); > > >

Re: [PATCH 8/7] net/netfilter/nf_conntrack_core: Remove another memory barrier

2016-09-02 Thread Boqun Feng
Hi Manfred, On Thu, Sep 01, 2016 at 06:41:26PM +0200, Peter Zijlstra wrote: > On Thu, Sep 01, 2016 at 04:30:39PM +0100, Will Deacon wrote: > > On Thu, Sep 01, 2016 at 05:27:52PM +0200, Manfred Spraul wrote: > > > Since spin_unlock_wait() is defined as equivalent to spin_lock(); > > >

Re: [PATCH 8/7] net/netfilter/nf_conntrack_core: Remove another memory barrier

2016-09-01 Thread Peter Zijlstra
On Thu, Sep 01, 2016 at 04:30:39PM +0100, Will Deacon wrote: > On Thu, Sep 01, 2016 at 05:27:52PM +0200, Manfred Spraul wrote: > > Since spin_unlock_wait() is defined as equivalent to spin_lock(); > > spin_unlock(), the memory barrier before spin_unlock_wait() is > > also not required. Note that

Re: [PATCH 8/7] net/netfilter/nf_conntrack_core: Remove another memory barrier

2016-09-01 Thread Peter Zijlstra
On Thu, Sep 01, 2016 at 04:30:39PM +0100, Will Deacon wrote: > On Thu, Sep 01, 2016 at 05:27:52PM +0200, Manfred Spraul wrote: > > Since spin_unlock_wait() is defined as equivalent to spin_lock(); > > spin_unlock(), the memory barrier before spin_unlock_wait() is > > also not required. Note that

Re: [PATCH 8/7] net/netfilter/nf_conntrack_core: Remove another memory barrier

2016-09-01 Thread Will Deacon
On Thu, Sep 01, 2016 at 05:27:52PM +0200, Manfred Spraul wrote: > Since spin_unlock_wait() is defined as equivalent to spin_lock(); > spin_unlock(), the memory barrier before spin_unlock_wait() is > also not required. > > Not for stable! > > Signed-off-by: Manfred Spraul

Re: [PATCH 8/7] net/netfilter/nf_conntrack_core: Remove another memory barrier

2016-09-01 Thread Will Deacon
On Thu, Sep 01, 2016 at 05:27:52PM +0200, Manfred Spraul wrote: > Since spin_unlock_wait() is defined as equivalent to spin_lock(); > spin_unlock(), the memory barrier before spin_unlock_wait() is > also not required. > > Not for stable! > > Signed-off-by: Manfred Spraul > Cc: Pablo Neira Ayuso

[PATCH 8/7] net/netfilter/nf_conntrack_core: Remove another memory barrier

2016-09-01 Thread Manfred Spraul
Since spin_unlock_wait() is defined as equivalent to spin_lock(); spin_unlock(), the memory barrier before spin_unlock_wait() is also not required. Not for stable! Signed-off-by: Manfred Spraul Cc: Pablo Neira Ayuso Cc:

[PATCH 8/7] net/netfilter/nf_conntrack_core: Remove another memory barrier

2016-09-01 Thread Manfred Spraul
Since spin_unlock_wait() is defined as equivalent to spin_lock(); spin_unlock(), the memory barrier before spin_unlock_wait() is also not required. Not for stable! Signed-off-by: Manfred Spraul Cc: Pablo Neira Ayuso Cc: netfilter-de...@vger.kernel.org --- net/netfilter/nf_conntrack_core.c | 8