Re: [Xen-devel] Check for irqs_disabled in evtchn_fifo_unmask

[Juergen Gross <jgross@xxxxxxxx>]

On 01/06/17 19:01, Anoob Soman wrote:
> Hi,
> 
> Can someone explain, why evtchn_fifo_unmask() requires irqs_disabled().
> What happens, if irqs are not disabled ?

evtchn_fifo_unmask() seems to be not critical, but evtchn_2l_unmask()
might be racy if another event is coming in in the middle. The
BUG_ON() in evtchn_fifo_unmask() just makes sure adding a new caller
not obeying the restriction of evtchn_2l_unmask() won't be unnoticed.


Juergen

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
https://lists.xen.org/xen-devel