On Tue, 1 Aug 2006 21:31:22 -0700
"Jesse Brandeburg" <jesse.brandeburg@xxxxxxxxx> wrote:
> On 7/31/06, Alan Stern <stern@xxxxxxxxxxxxxxxxxxx> wrote:
> > On Mon, 31 Jul 2006, Andrew Morton wrote:
> > > core_initcall() would suit. That's actually a bit late for this sort of
> > > thing, but we can always add a new section later if it becomes a problem.
> > > I'd suggest that we ensure that srcu_notifier_chain_register() performs a
> > > reliable BUG() if it gets called too early.
> > Here's a patch to test. I can't try it out on my machine because
> > 2.6.18-rc2-mm1 (even without the patch) crashes partway through a
> > suspend-to-disk, in a way that's extremely hard to debug. Some sort of
> > spinlock-related bug occurs within ioapic_write_entry.
> can't test because I also can't suspend or hibernate with rc2-mm1
> (resume causes hard hang with the backlight and screen off) The issue
> i reported was against linus' 2.6.18-rc3 kernel.
This might help?
author Jiri Slaby <ku@xxxxxxxxxxxxxxxxxxx> Tue, 01 Aug 2006 01:16:13 +0159
@@ -2360,6 +2360,7 @@ static int ioapic_resume(struct sys_devi
reg_00.bits.ID = mp_ioapics[dev->id].mpc_apicid;
io_apic_write(dev->id, 0, reg_00.raw);
+ spin_unlock_irqrestore(&ioapic_lock, flags);
for (i = 0; i < nr_ioapic_registers[dev->id]; i ++)
ioapic_write_entry(dev->id, i, entry[i]);