Re: [PATCH v3 0/3] epoll: introduce round robin wakeup mode

2015-03-07 Thread Jason Baron
On 03/05/2015 04:15 AM, Ingo Molnar wrote: > * Jason Baron wrote: > >> 2) We are using the wakeup in this case to 'assign' work more >> permanently to the thread. That is, in the case of a listen socket >> we then add the connected socket to the woken up threads local set >> of epoll events.

Re: [PATCH v3 0/3] epoll: introduce round robin wakeup mode

2015-03-07 Thread Jason Baron
On 03/05/2015 04:15 AM, Ingo Molnar wrote: * Jason Baron jba...@akamai.com wrote: 2) We are using the wakeup in this case to 'assign' work more permanently to the thread. That is, in the case of a listen socket we then add the connected socket to the woken up threads local set of epoll

Re: [PATCH v3 0/3] epoll: introduce round robin wakeup mode

2015-03-05 Thread Jason Baron
On 03/05/2015 04:15 AM, Ingo Molnar wrote: > * Jason Baron wrote: > >> 2) We are using the wakeup in this case to 'assign' work more >> permanently to the thread. That is, in the case of a listen socket >> we then add the connected socket to the woken up threads local set >> of epoll events.

Re: [PATCH v3 0/3] epoll: introduce round robin wakeup mode

2015-03-05 Thread Ingo Molnar
* Jason Baron wrote: > 2) We are using the wakeup in this case to 'assign' work more > permanently to the thread. That is, in the case of a listen socket > we then add the connected socket to the woken up threads local set > of epoll events. So the load persists past the wake up. And in this

Re: [PATCH v3 0/3] epoll: introduce round robin wakeup mode

2015-03-05 Thread Jason Baron
On 03/05/2015 04:15 AM, Ingo Molnar wrote: * Jason Baron jba...@akamai.com wrote: 2) We are using the wakeup in this case to 'assign' work more permanently to the thread. That is, in the case of a listen socket we then add the connected socket to the woken up threads local set of epoll

Re: [PATCH v3 0/3] epoll: introduce round robin wakeup mode

2015-03-05 Thread Ingo Molnar
* Jason Baron jba...@akamai.com wrote: 2) We are using the wakeup in this case to 'assign' work more permanently to the thread. That is, in the case of a listen socket we then add the connected socket to the woken up threads local set of epoll events. So the load persists past the wake

Re: [PATCH v3 0/3] epoll: introduce round robin wakeup mode

2015-03-04 Thread Jason Baron
On 03/04/2015 07:02 PM, Ingo Molnar wrote: > * Andrew Morton wrote: > >> On Fri, 27 Feb 2015 17:01:32 -0500 Jason Baron wrote: >> I don't really understand the need for rotation/round-robin. We can solve the thundering herd via exclusive wakeups, but what is the point in choosing

Re: [PATCH v3 0/3] epoll: introduce round robin wakeup mode

2015-03-04 Thread Ingo Molnar
* Andrew Morton wrote: > On Fri, 27 Feb 2015 17:01:32 -0500 Jason Baron wrote: > > > > > > > > > I don't really understand the need for rotation/round-robin. We can > > > solve the thundering herd via exclusive wakeups, but what is the point > > > in choosing to wake the task which has been

Re: [PATCH v3 0/3] epoll: introduce round robin wakeup mode

2015-03-04 Thread Ingo Molnar
* Andrew Morton a...@linux-foundation.org wrote: On Fri, 27 Feb 2015 17:01:32 -0500 Jason Baron jba...@akamai.com wrote: I don't really understand the need for rotation/round-robin. We can solve the thundering herd via exclusive wakeups, but what is the point in choosing to

Re: [PATCH v3 0/3] epoll: introduce round robin wakeup mode

2015-03-04 Thread Jason Baron
On 03/04/2015 07:02 PM, Ingo Molnar wrote: * Andrew Morton a...@linux-foundation.org wrote: On Fri, 27 Feb 2015 17:01:32 -0500 Jason Baron jba...@akamai.com wrote: I don't really understand the need for rotation/round-robin. We can solve the thundering herd via exclusive wakeups, but what

Re: [PATCH v3 0/3] epoll: introduce round robin wakeup mode

2015-03-01 Thread Jason Baron
On 02/27/2015 04:31 PM, Jonathan Corbet wrote: > On Fri, 27 Feb 2015 13:10:34 -0800 > Andrew Morton wrote: > >> I don't really understand the need for rotation/round-robin. We can >> solve the thundering herd via exclusive wakeups, but what is the point >> in choosing to wake the task which has

Re: [PATCH v3 0/3] epoll: introduce round robin wakeup mode

2015-03-01 Thread Jason Baron
On 02/27/2015 04:31 PM, Jonathan Corbet wrote: On Fri, 27 Feb 2015 13:10:34 -0800 Andrew Morton a...@linux-foundation.org wrote: I don't really understand the need for rotation/round-robin. We can solve the thundering herd via exclusive wakeups, but what is the point in choosing to wake the

Re: [PATCH v3 0/3] epoll: introduce round robin wakeup mode

2015-02-27 Thread Andrew Morton
On Fri, 27 Feb 2015 17:01:32 -0500 Jason Baron wrote: > > > > > I don't really understand the need for rotation/round-robin. We can > > solve the thundering herd via exclusive wakeups, but what is the point > > in choosing to wake the task which has been sleeping for the longest > > time? Why

Re: [PATCH v3 0/3] epoll: introduce round robin wakeup mode

2015-02-27 Thread Jason Baron
On 02/27/2015 04:10 PM, Andrew Morton wrote: > On Wed, 25 Feb 2015 11:27:04 -0500 Jason Baron wrote: > >>> Libenzi inactive eventpoll appears to be without a >>> dedicated maintainer since 2011 or so. Is there anyone who >>> knows the code and its usages in detail and does final ABI >>>

Re: [PATCH v3 0/3] epoll: introduce round robin wakeup mode

2015-02-27 Thread Jonathan Corbet
On Fri, 27 Feb 2015 13:10:34 -0800 Andrew Morton wrote: > I don't really understand the need for rotation/round-robin. We can > solve the thundering herd via exclusive wakeups, but what is the point > in choosing to wake the task which has been sleeping for the longest > time? Why is that

Re: [PATCH v3 0/3] epoll: introduce round robin wakeup mode

2015-02-27 Thread Andrew Morton
On Wed, 25 Feb 2015 11:27:04 -0500 Jason Baron wrote: > > Libenzi inactive eventpoll appears to be without a > > dedicated maintainer since 2011 or so. Is there anyone who > > knows the code and its usages in detail and does final ABI > > decisions on eventpoll - Andrew, Al or Linus? > > >

Re: [PATCH v3 0/3] epoll: introduce round robin wakeup mode

2015-02-27 Thread Jonathan Corbet
On Fri, 27 Feb 2015 13:10:34 -0800 Andrew Morton a...@linux-foundation.org wrote: I don't really understand the need for rotation/round-robin. We can solve the thundering herd via exclusive wakeups, but what is the point in choosing to wake the task which has been sleeping for the longest

Re: [PATCH v3 0/3] epoll: introduce round robin wakeup mode

2015-02-27 Thread Jason Baron
On 02/27/2015 04:10 PM, Andrew Morton wrote: On Wed, 25 Feb 2015 11:27:04 -0500 Jason Baron jba...@akamai.com wrote: Libenzi inactive eventpoll appears to be without a dedicated maintainer since 2011 or so. Is there anyone who knows the code and its usages in detail and does final ABI

Re: [PATCH v3 0/3] epoll: introduce round robin wakeup mode

2015-02-27 Thread Andrew Morton
On Wed, 25 Feb 2015 11:27:04 -0500 Jason Baron jba...@akamai.com wrote: Libenzi inactive eventpoll appears to be without a dedicated maintainer since 2011 or so. Is there anyone who knows the code and its usages in detail and does final ABI decisions on eventpoll - Andrew, Al or Linus?

Re: [PATCH v3 0/3] epoll: introduce round robin wakeup mode

2015-02-27 Thread Andrew Morton
On Fri, 27 Feb 2015 17:01:32 -0500 Jason Baron jba...@akamai.com wrote: I don't really understand the need for rotation/round-robin. We can solve the thundering herd via exclusive wakeups, but what is the point in choosing to wake the task which has been sleeping for the longest

Re: [PATCH v3 0/3] epoll: introduce round robin wakeup mode

2015-02-25 Thread Jason Baron
On 02/25/2015 02:38 AM, Ingo Molnar wrote: > * Jason Baron wrote: > >> Hi, >> >> When we are sharing a wakeup source among multiple epoll >> fds, we end up with thundering herd wakeups, since there >> is currently no way to add to the wakeup source >> exclusively. This series introduces a new

Re: [PATCH v3 0/3] epoll: introduce round robin wakeup mode

2015-02-25 Thread Jason Baron
On 02/25/2015 02:38 AM, Ingo Molnar wrote: * Jason Baron jba...@akamai.com wrote: Hi, When we are sharing a wakeup source among multiple epoll fds, we end up with thundering herd wakeups, since there is currently no way to add to the wakeup source exclusively. This series introduces a

Re: [PATCH v3 0/3] epoll: introduce round robin wakeup mode

2015-02-24 Thread Ingo Molnar
* Jason Baron wrote: > Hi, > > When we are sharing a wakeup source among multiple epoll > fds, we end up with thundering herd wakeups, since there > is currently no way to add to the wakeup source > exclusively. This series introduces a new EPOLL_ROTATE > flag to allow for round robin

Re: [PATCH v3 0/3] epoll: introduce round robin wakeup mode

2015-02-24 Thread Ingo Molnar
* Jason Baron jba...@akamai.com wrote: Hi, When we are sharing a wakeup source among multiple epoll fds, we end up with thundering herd wakeups, since there is currently no way to add to the wakeup source exclusively. This series introduces a new EPOLL_ROTATE flag to allow for round