Re: [BUG] RCU hang with io_uring nvme polling

[Keith Busch]

On Mon, Jun 29, 2026 at 04:47:00PM -0400, Ben Carey wrote:
> On Fri, Jun 26, 2026 at 12:33 PM Keith Busch <kbusch@xxxxxxxxxx> wrote:
> > The test has 1 polling queue with 2 jobs dispatching. One of the job's
> > polled the completions for both. The other job is polling for no reason
> > at all with nothing outstanding. The only thing that can break us out of
> > that loop now is need_resched(), but that appears to never return true.
> 
> Inspired by this I tried to find a place where one thread polls on a job that's
> already finished. I found that a race to io_check_iopoll causes one thread to
> enter the polling loop when another has already finished on it. Putting
> io_check_iopoll behind a spinlock seems to fix it, though I imagine a more
> elegant fix is out there (reusing a different lock, not using expensive locks,
> a smarter place to check for racing, etc.)

I can see why that resolves your observation, but I don't think we can
do this. We're ultimately polling for a hardware event, and this layer
is too high a level for serializing these things. This could be a
multi-device or multi-queue backing storage where the completion
pollings should occur concurrently.