Re: [Xen-devel] [PATCH] x86/MSI: drop local cpumask_t variable from msi_compose_msg()

[Keir Fraser <keir.xen@xxxxxxxxx>]

On 14/10/2011 13:42, "Jan Beulich" <JBeulich@xxxxxxxx> wrote:

>> I wonder whether the cpus_empty() check should be a BUG_ON. Or an ASSERT
>> pushed into cpu_mask_to_apicid.
> 
> An ASSERT may be reasonable, but simply dropping the check here
> may be too - no other code path invoking cpu_mask_to_apicid() has
> a similar check.

It is certainly not valid to call cpu_mask_to_apicid with an empty mask.
Hence an ASSERT would check that precondition, and for all callers.

 -- Keir



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