On Sun, Nov 04, 2018 at 01:02:18PM +0100, Thomas Gleixner wrote: > Jens, > > On Sat, 3 Nov 2018, Jens Axboe wrote: > > > On 11/2/18 8:59 AM, Ming Lei wrote: > > > Hi Jens, > > > > > > As I mentioned, there are at least two issues in the patch of ' > > > irq: add support for allocating (and affinitizing) sets of IRQs': > > > > > > 1) it is wrong to pass 'mask + usedvec' to irq_build_affinity_masks() > > > > > > 2) we should spread all possible CPUs in 2-stage way on each set of IRQs > > > > > > The fix isn't trivial, and I introduce two extra patches as preparation, > > > then the implementation can be more clean. > > > > > > The patchset is against mq-maps branch of block tree, feel free to > > > integrate into the whole patchset of multiple queue maps. > > > > Thanks Ming, I ran this through my testing, and I end up with the > > same maps and affinities for all the cases I cared about. I'm going > > to drop my initial version, and add the three. > > So I assume, that I can pick up Mings series instead. > > There is another patch pending affecting the irq affinity spreading. Can > you folks please have a look at it? > > https://lkml.kernel.org/r/20181102180248.13583-1-longli@xxxxxxxxxxxxxxxxx This patch looks fine. It ensures that all CPUs are covered in irq's affinity when required vector number is <= nr_numa_nodes. Reviewed-by: Ming Lei <ming.lei@xxxxxxxxxx> Thanks, Ming
