Thread (4 messages) 4 messages, 2 authors, 2013-12-16

Re: [RFC PATCH] time: Support in tick broadcast framework for archs without an external wakeup source

From: Preeti U Murthy <hidden>
Date: 2013-12-13 05:24:53
Also in: linux-pm, lkml

Hi Ben,

On 12/13/2013 10:47 AM, Benjamin Herrenschmidt wrote:
On Fri, 2013-12-13 at 09:49 +0530, Preeti U Murthy wrote:
quoted
On some architectures, in certain CPU deep idle states the local timers stop.
An external clock device is used to wakeup these CPUs. The kernel support for the
wakeup of these CPUs is provided by the tick broadcast framework by using the
external clock device as the wakeup source.
quoted
However on architectures like PowerPC there is no external clock device. 
Minor nit ...

I wouldn't make this an architectural statement. Some PowerPC's do have
external clock devices (for example the old MPIC interrupt controller
had timers). In fact, if we really need it, I'm sure we *could* find
something somewhere in P8 that could act as a timer, probably hijacking
a bit of the OCC or similar but at this stage, that's not on the radar.

So make it an implementation statement. "However, not all
implementations, such as some PowerPC ones, provide such an external
timer ...".
Thanks for this information. I will update this going ahead.

Regards
Preeti U Murthy
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help