On 2022/7/22 14:17, Christoph Hellwig wrote: > On Thu, Jul 21, 2022 at 10:48:07PM -0700, Christoph Hellwig wrote: >> I think __ublk_destroy_dev just needs to go away in that form. >> Also I'd much rather do the copy_to_user before the ublk_add_chdev >> as that means we never remove a devic already marked life due to a >> failure. Something like the patch below, which will need testing first >> before I'd dare to submit it: > > Improved and tested version: > > --- > From 49ba6d0c5788ea9d3a6ef88d910b702152f5d75a Mon Sep 17 00:00:00 2001 > From: Christoph Hellwig <hch@xxxxxx> > Date: Fri, 22 Jul 2022 07:38:59 +0200 > Subject: ublk_drv: fix error handling of ublk_add_dev > > __ublk_destroy_dev() is called for handling error in ublk_add_dev(), > but either tagset isn't allocated or mutex isn't initialized. > > So fix the issue by letting replacing ublk_add_dev with a > ublk_add_tag_set function that is much more limited in scope and > instead unwind every single step directly in ublk_ctrl_add_dev. > To allow for this refactor the device freeing so that there is > a helper for freeing the device number instead of coupling that > with freeing the mutex and the memory. > > Note that this now copies the dev_info to userspace before adding > the character device. This not only simplifies the erro handling > in ublk_ctrl_add_dev, but also means that the character device > can only be seen by userspace if the device addition succeeded. > > Based on a patch from Ming Lei. > > Signed-off-by: Christoph Hellwig <hch@xxxxxx> > --- Reviewed-by: ZiyangZhang <ZiyangZhang@xxxxxxxxxxxxxxxxx>
