On Sat, Sep 04, 2021 at 11:49:06AM +0900, Tetsuo Handa wrote:
> On 2021/09/04 10:39, Luis Chamberlain wrote:
> > diff --git a/fs/block_dev.c b/fs/block_dev.c
> > index 45df6cbccf12..81a4738910a8 100644
> > --- a/fs/block_dev.c
> > +++ b/fs/block_dev.c
> > @@ -1144,10 +1144,13 @@ struct block_device *blkdev_get_no_open(dev_t dev)
> > {
> > struct block_device *bdev;
> > struct inode *inode;
> > + int ret;
> >
> > inode = ilookup(blockdev_superblock, dev);
> > if (!inode) {
> > - blk_request_module(dev);
> > + ret = blk_request_module(dev);
> > + if (ret)
> > + return NULL;
>
> Since e.g. loop_add() from loop_probe() returns -EEXIST when /dev/loop$num already
> exists (e.g. raced with ioctl(LOOP_CTL_ADD)), isn't unconditionally failing an over-failing?
It's not clear to me how. What do we loose by capturing the failure on
blk_request_module()?
Luis
