Messages in this thread | | | Subject | Re: [PATCH, resend] eliminate spurious pointless WARN_ON()s | From | Peter Zijlstra <> | Date | Fri, 13 Mar 2009 09:52:35 +0100 |
| |
On Thu, 2009-03-12 at 14:48 +0100, Andi Kleen wrote: > On Thu, Mar 12, 2009 at 01:21:50PM +0000, Jan Beulich wrote: > > Namely during early boot, the panic() or BUG() paths may end up in > > smp_call_function_*() with just a single online CPU. In that situation > > the warnings generated are not only meaningless, but also result in > > relevant output being cluttered. > > I actually have patches that just fix panic/shutdown to never call > smp_call_function(), but use an own vector. It does all kinds of other things > too that are not appropiate in panic, like allocating memory.
Oh? The only allocation that it does it at cpu-hotplug, where it allocates a cpu-mask thingy, and that's only a real allocation on NR_CPUS>64.
| |