Russell, > On Tue, Nov 30, 2010 at 11:03:31AM +0000, Russell King - ARM Linux wrote: > > I see no point to what is being proposed in this thread. It's _soo_ > > little code that the platforms have to implement that it really is > > not worth the effort.
[...] > Note also that most implementations of this get it wrong. Realview > assumes ARMv6 MPcore layout of the auxcontrol register, which is > cut'n'pasted into s5pv310 (ARMv7) and Tegra (ARMv7) without thinking. > The only implementation which has put some thought into it is the > UX500 one, which is extremely simple. > > So, out of the four existing implementations, at least two need fixing > to be different, and I'll bet that they do have some way to power down > the offlined CPUs for greater power savings. I have patches that fix this for RealView and Versatile Express, I posted them to the list in the past (note that these are now out of date): http://lists.infradead.org/pipermail/linux-arm-kernel/2010-September/026160.html I have a newer version here: git://linux-arm.org/linux-2.6-wd.git cpu-hotplug but it relies on the Versatile Express multi-tile patches that I'm posting at the moment. If the hotplug APIs aren't about to change fundamentally, then I'll repost these patches once the multi-tile stuff is accepted. Will _______________________________________________ linaro-dev mailing list linaro-dev@lists.linaro.org http://lists.linaro.org/mailman/listinfo/linaro-dev
