Re: [RFC PATCH 1/2] perf_events: add support for per-cpu per-cgroupmonitoring (v2)

From: Stephane Eranian
Date: Wed Sep 08 2010 - 09:59:18 EST


On Wed, Sep 8, 2010 at 3:57 PM, Peter Zijlstra <peterz@xxxxxxxxxxxxx> wrote:
> On Wed, 2010-09-08 at 15:56 +0200, stephane eranian wrote:
>> On Wed, Sep 8, 2010 at 3:52 PM, Peter Zijlstra <peterz@xxxxxxxxxxxxx> wrote:
>> > On Wed, 2010-09-08 at 15:30 +0200, Stephane Eranian wrote:
>> >> + Â Â Â } times[NR_CPUS] ____cacheline_aligned_in_smp;
>> >
>> > That's fail! NR_CPUS can be like 4k for distro configs.
>> >
>> How do you do this on struct that are dynamically allocated?
>> The number of css is dynamic.
>
> alloc_percpu()?
>
Ok, I will look into this. Thanks.
Clearly we can kmalloc() the array. But it would be nicer to have one
struct allocated per-cpu (avoid the cacheline_aligned_in_smp stuff).
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/