On Mon, Jun 27, 2016 at 04:51:49PM +0200, Quentin Monnet wrote:
> 2016-06-23 (15:59 -0700) ~ Ben Pfaff
> > On Thu, Jun 09, 2016 at 11:52:49AM +0200, Quentin Monnet wrote:
> >> Signed-off-by: David Marchand
> >> Signed-off-by: Liu Xiaofeng
2016-06-23 (15:59 -0700) ~ Ben Pfaff
> On Thu, Jun 09, 2016 at 11:52:49AM +0200, Quentin Monnet wrote:
>> Signed-off-by: David Marchand
>> Signed-off-by: Liu Xiaofeng
>> Signed-off-by: Quentin Monnet
>
>
On Thu, Jun 09, 2016 at 11:52:49AM +0200, Quentin Monnet wrote:
> From: David Marchand
>
> Relying on /proc/cpuinfo to count the number of available cores is not
> the best option:
>
> - The code is x86-specific.
> - If the process is started with a different CPU
Quentin Monnet wrote on 06/16/2016 09:55:35 AM:
> From: Quentin Monnet
> To: Ryan Moats/Omaha/IBM@IBMUS
> Cc: Ben Pfaff , David Marchand
> , ovs-dev
> Date: 06/16/2016 09:56 AM
>
Hi Ryan,
> > Other UNIX-like systems only use _SC_NPROCESSORS_ONLN.
>
> After double checking, I'm not *entirely* sure that's true, but it can
> be updated in a follow-on patch
>
I meant “Other UNIX-like systems only use _SC_NPROCESSORS_ONLN *with
this commit*”, I did not mean that there was no
Quentin Monnet wrote on 06/09/2016 04:52:49 AM:
> From: Quentin Monnet
> To: Ben Pfaff
> Cc: David Marchand , Ryan Moats/Omaha/
> IBM@IBMUS, ovs-dev
> Date: 06/09/2016 04:53 AM
>
From: David Marchand
Relying on /proc/cpuinfo to count the number of available cores is not
the best option:
- The code is x86-specific.
- If the process is started with a different CPU affinity, then it will
wrongly try to start too many threads (for an example,