Messages in this thread | | | Date | Wed, 26 Jul 2006 13:22:24 -0700 (PDT) | From | Linus Torvalds <> | Subject | Re: [patch] Reorganize the cpufreq cpu hotplug locking to not be totally bizare |
| |
On Wed, 26 Jul 2006, Arjan van de Ven wrote: > > As a quick hack I made non-lock_cpu_hotplug()'ing versions of the 3 key > workqueue functions (patch below). It works, it's correct, it's just so > ugly that I'm almost too ashamed to post it. I haven't found a better > solution yet though... time to take a step back I suppose.
That really is _way_ too ugly for words.
For 2.6.18, we may just have to leave the recursive locking in place, and just remove the warning. With the recursive lock, if/when somebody needs to take that lock early, the code can just do so, and then the inner lock-taker ends up being a no-op.
Of course, that's why people want recursive locks in the first place, and it's also why we've (largely successfully) have avoided them - it allows for people being way too lazy about locking, and allows for really broken schenarios like this.
I wonder if we could just make the workqueue code just run with preemption disabled - that should also automatically protect against any CPU hotplug events on the local CPU (and I think "local CPU" is all that the wq code cares about, no?)
Linus - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
| |