Re: [PATCH] pci: Avoid unnecessary calls to work_on_cpu

2013-07-07 Thread Andi Kleen
> > But Rusty added this (in 873392ca51), and he knows way more about this > > stuff than I do. > > Actually, I just stopped the code from playing cpumask games, which is > what it used to do. You're right the numa_node_id() check ptimization is not 100% safe on preempt kernels and should be pro

Re: [PATCH] pci: Avoid unnecessary calls to work_on_cpu

2013-07-07 Thread Rusty Russell
Bjorn Helgaas writes: > [+cc Rusty] > > On Mon, Jun 24, 2013 at 2:05 PM, Alexander Duyck > wrote: >> This patch is meant to address the fact that we are making unnecessary calls >> to work_on_cpu. To resolve this I have added a check to see if the current >> node is the correct node for the devi

Re: [PATCH] pci: Avoid unnecessary calls to work_on_cpu

2013-07-05 Thread Benjamin Herrenschmidt
On Fri, 2013-07-05 at 17:36 -0600, Bjorn Helgaas wrote: > It seems a little strange to me that this "run the driver probe method > on the correct node" code is in PCI. I would think this behavior > would be desirable for *all* bus types, not just PCI, so maybe it > would make sense to do this up i

Re: [PATCH] pci: Avoid unnecessary calls to work_on_cpu

2013-07-05 Thread Bjorn Helgaas
[+cc Rusty] On Mon, Jun 24, 2013 at 2:05 PM, Alexander Duyck wrote: > This patch is meant to address the fact that we are making unnecessary calls > to work_on_cpu. To resolve this I have added a check to see if the current > node is the correct node for the device before we decide to assign the

[PATCH] pci: Avoid unnecessary calls to work_on_cpu

2013-06-24 Thread Alexander Duyck
This patch is meant to address the fact that we are making unnecessary calls to work_on_cpu. To resolve this I have added a check to see if the current node is the correct node for the device before we decide to assign the probe task to another CPU. The advantages to this approach is that we can