Re: [Xen-devel] [PATCH 00/12] cpumask handling scalability improvements

["Jan Beulich" <JBeulich@xxxxxxxx>]

>>> On 20.10.11 at 17:49, Andrew Cooper <andrew.cooper3@xxxxxxxxxx> wrote:
> The saving of not allocating masks for unused irq_desc's (and irq_cfg's)
> will be significant in the general case.  (3 * NR_UNUSED_IRQs *
> sizeof(mask)) where the average system is wasting most of 224 IRQs per CPU.
> 
> However, I am against moving the masks out of irq_desc (perhaps this is
> the C++ coder inside me).

I didn't propose moving them out of there.

> Would an acceptable alternative be to change irq_desc to use
> cpumask_var_t's and allocate them on first use? (I have not spent long
> thinking about this, so it is possible that the extra checks for Null
> pointers on the irq path might be counter productive?)

That's what it does (or really it allocates them for all GSIs at boot, and
then in create_irq() for the dynamically allocated ones). There's no
reason to worry about NULL checks - any access to a never initialized
irq_desc is a bug anyway.

Jan


_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-devel