On Fri, 3 Apr 2015 08:55:57 +0200 Ingo Molnar <mi...@kernel.org> wrote: > So the original commit also has the problem that it unnecessary > drops/retakes the descriptor lock: > > > irq_put_desc_unlock(desc, flags); > > - /* set the initial affinity to prevent every interrupt being on CPU0 */ > > - if (m) > > - __irq_set_affinity(irq, m, false); > > > i.e. why not just call into irq_set_affinity_locked() while we still > have the descriptor locked?
I had tried that but it didn't help much. I also tried kzalloc a new descriptor like the proc functionality does, and that changes the behavior a little, but doesn't fix it AFAICS. > Now this is just a small annoyance that should not really matter - it > would be nice to figure out the real reason for why the irqs move back > to CPU#0. > > In theory the same could happen to 'irqbalanced' as well, if it calls > shortly after an irq was registered - so this is not a bug we want to > ignore. Let me know if I can do something to help, the IRQ code is a bit of a steep learning curve, so the chances of me fixing it are small. > Also, worst case we are back to where v3.19 was, right? So could we > try to analyze this a bit more? Yes, 3.19 shipped with this issue. Again, let me know if I can help. Thanks, Jesse -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/