Messages in this thread | | | Date | Sat, 9 Aug 2008 00:30:06 -0700 | From | "Yinghai Lu" <> | Subject | Re: [PATCH 08/42] introduce nr_irqs |
| |
On Fri, Aug 8, 2008 at 6:59 PM, H. Peter Anvin <hpa@zytor.com> wrote: > Yinghai Lu wrote: >> >> On Fri, Aug 8, 2008 at 6:00 PM, Eric W. Biederman <ebiederm@xmission.com> >> wrote: >>> >>> Yinghai Lu <yhlu.kernel@gmail.com> writes: >>> >>>> and at this point it is equal NR_IRQS >>> >>> This patch is ok, but this approach is fundamentally broken. >>> Only 19 architectures use GENERIC_HARDIRQS >>> >>> The only required interface is interrupt.h not irq.h >>> >>> So you may not replace NR_IRQS with nr_irqs in anything other >>> then genirq code and some architecture code. >>> >>> Doing so will break the build on several architectures. >> >> some platforms will not use include/linux/irq.h and kernel/irq/handle.c ? >> > > Also, what's the point, if it's just a renaming?
that is the start point. nr_irqs is variable, and will be probed later. and use that number to init dyn_alloc.
YH
| |