Messages in this thread | | | Date | Mon, 25 Apr 2005 22:47:53 -0600 (MDT) | From | Zwane Mwaikambo <> | Subject | Re: [PATCH]broadcast IPI race condition on CPU hotplug |
| |
On Tue, 26 Apr 2005, Li Shaohua wrote:
> On Tue, 2005-04-26 at 12:25, Zwane Mwaikambo wrote: > > > > > After a CPU is booted but before it's officially up (set online map, and > > > enable interrupt), the CPU possibly will receive a broadcast IPI. After > > > it's up, it will handle the stale interrupt soon and maybe cause oops if > > > it's a smp-call-function-interrupt. This is quite possible in CPU > > > hotplug case, but nearly can't occur at boot time. Below patch replaces > > > broadcast IPI with send_ipi_mask just like the cluster mode. > > > > Ok, but isn't it sufficient to use send_ipi_mask in smp_call_function > > instead? > I'm not sure if other routines using broadcast IPI have this bug. Fixing > the send_ipi_all API looks more generic. Is there any reason we should > use broadcast IPI?
I'd prefer only touching smp_call_function because the others are primitives, we should only be fixing the users of the primitives, otherwise we'll end up with code which won't be as versatile.
Thanks, Zwane - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
| |