Re: [PATCH] fs: Add additional checks for block devices during mount

[Christian Brauner]

On Mon, Jul 21, 2025 at 08:47:12AM +0200, Christoph Hellwig wrote:
> On Mon, Jul 21, 2025 at 09:20:27AM +0800, Zizhi Wo wrote:
> > Sorry, disk_live() is only declared but not defined when CONFIG_BLOCK is
> > not set...
> 
> You can just add a if (IS_ENABLED(CONFIG_BLOCK)) check around it.
> 
> 
> But the layering here feels wrong.  sget_dev and it's helper operate
> purely on the dev_t.  Anything actually dealing with a block device /
> gendisk should be in the helpers that otherwise use it.

Either we add a lookup_bdev_alive() variant that checks whether the
inode is still hashed when looking up the dev_t and we accept that this
deals with the obvious cases and accept that this is racy...

Or we add lookup_bdev_no_open() that returns the block device with the
reference bumped, paired with lookup_bdev_put_no_open(). Afaiu, that
would prevent deletion. We could even put this behind a
guard(bdev_no_open)(fc->source). The reference count bump shouldn't
matter there. Christoph?