Messages in this thread | | | Date | Fri, 20 Jul 2018 13:05:59 -0700 | From | Davidlohr Bueso <> | Subject | Re: [PATCH -next 0/2] fs/epoll: loosen irq safety when possible |
| |
On Fri, 20 Jul 2018, Andrew Morton wrote:
>On Fri, 20 Jul 2018 10:29:54 -0700 Davidlohr Bueso <dave@stgolabs.net> wrote: > >> Hi, >> >> Both patches replace saving+restoring interrupts when taking the >> ep->lock (now the waitqueue lock), with just disabling local irqs. >> This shows immediate performance benefits in patch 1 for an epoll >> workload running on Xen. > >I'm surprised. Is spin_lock_irqsave() significantly more expensive >than spin_lock_irq()? Relative to all the other stuff those functions >are doing? If so, how come? Some architectural thing makes >local_irq_save() much more costly than local_irq_disable()?
For example, if you compare x86 native_restore_fl() to xen_restore_fl(), the cost of Xen is much higher.
And at least considering ep_scan_ready_list(), the lock is taken/released twice, to deal with the ovflist when the ep->wq.lock is not held. To the point that it yields measurable results (see patch 1) across incremental thread counts.
> >> The main concern we need to have with this >> sort of changes in epoll is the ep_poll_callback() which is passed >> to the wait queue wakeup and is done very often under irq context, >> this patch does not touch this call. > >Yeah, these changes are scary. For the code as it stands now, and for >the code as it evolves.
Yes which is why I've been throwing lots of epoll workloads at it.
> >I'd have more confidence if we had some warning mechanism if we run >spin_lock_irq() when IRQs are disabled, which is probably-a-bug. But >afaict we don't have that. Probably for good reasons - I wonder what >they are? > >> Patches have been tested pretty heavily with the customer workload, >> microbenchmarks, ltp testcases and two high level workloads that >> use epoll under the hood: nginx and libevent benchmarks. >> >> Details are in the individual patches. >> >> Applies on top of mmotd. > >Please convince me about the performance benefits?
As for number I only have patch 1.
Thanks, Davidlohr
| |