Re: [PATCH V6 for-6.11/block] loop: Fix a race between loop detach and loop open

[Christoph Hellwig <hch@lst.de>]

On Wed, Jun 19, 2024 at 08:21:35AM +0000, Gulam Mohamed wrote:
> > To: Gulam Mohamed <gulam.mohamed@oracle.com>
> > Cc: linux-block@vger.kernel.org; linux-kernel@vger.kernel.org;
> > yukuai1@huaweicloud.com; hch@lst.de; axboe@kernel.dk
> > Subject: Re: [PATCH V6 for-6.11/block] loop: Fix a race between loop detach
> > and loop open
> > 
> > Do we need the re-addition of the open method to fix the ltp test case?  I
> > kinda hate it, but if that is what it takes:
> > 
> I don't think its needed but I kept it because your following comment in the suggested change says " switch the state to roundown here to prevent new openers from coming in":

Let's keep it.  I meant to say new I/O coming in, but letting a new
opener come in and then fail I/O isn't really nice behavior.