|
|
|
|
|
|
|
|
|
|
xen-devel
Re: [Xen-devel] [PATCH] x86: fix the potential of encountering panic "IO
On 11/12/2008 10:39, "Jan Beulich" <jbeulich@xxxxxxxxxx> wrote:
> I have to admit that in addition I find it quite disturbing that there
> is a dependency on the side effect of smp_prepare_cpus()
> (check_nmi_watchdog() or until now setup_IO_APIC()) or init_xen_time()
> (init_platform_timer() -> plt_overflow()) enabling interrupts: c/s 5604
> removed the explicit local_irq_enable() from start_of_day() (which is
> __start_xen() now), putting it into init_xen_time(). There it got
> removed by c/s 18698, without an explicit replacement. I'd think there
> should be an explicit local_irq_enable() in __start_xen() in any case,
> and as early as reasonable (i.e. the latest before smp_prepare_cpus()).
Sounds reasonable. Send a patch.
-- Keir
_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-devel
|
|
|
|
|