On Tue, 22 Nov 2016 21:15:56 -0000 Thomas Gleixner <[email protected]> wrote:
> The driver fails to: > > - initialize packages which are not available at driver init time, > though the value of that initialization is completely unclear as > nothing ever uses these values. I fixed it up nevertheless and leave > it to the maintainers to decide whether it should completely go away. > > - to propagate error codes in the hotplug online path, where a > registration fails and the package data is freed, but return code > is 0. > > The initialization/removal code of that driver is a maze of > duplicated code which is more or less the same as the cpu hotplug > code. After switching over the driver to the hotplug statemachine, > the whole init/removal machinery can be replaced by > installing/removing the hotplug state. > > The total damage is: > > intel_rapl.c | 363 > ++++++++++++++++------------------------------------------- 1 file > changed, 104 insertions(+), 259 deletions(-) > > and the binary size shrinks as well: > > text data bss dec hex > 7996 625 32 8653 > 21cd Before 7216 593 32 > 7841 1ea1 After > I have successfully tested this patchset with various cpu online/offline scenarios on both single and dual socket systems. Looks good to me. The cpu topology management is much more streamlined. Thanks. I also sent out this patch below on top of yours. [PATCH] powercap/intel_rapl: fix and tidy up error handling Thanks, Jacob > Thanks, > > tglx > > > > [Jacob Pan]
