linux-kernel - Re: [RFC PATCH v2] dm ioctl: fix erroneous EINVAL when signaled

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <CACGdZY+bY+i=pbe9MXJQFVRVXehTsHT60ew_QxS7z4xXXvyvbw@mail.gmail.com>
Date: Tue, 23 Jul 2024 10:17:28 -0700
From: Khazhy Kumykov <khazhy@...omium.org>
To: Zdenek Kabelac <zdenek.kabelac@...il.com>
Cc: Mikulas Patocka <mpatocka@...hat.com>, Alasdair Kergon <agk@...hat.com>, 
	Mike Snitzer <snitzer@...nel.org>, Zdenek Kabelac <zkabelac@...hat.com>, 
	Joe Thornber <thornber@...hat.com>, Heinz Mauelshagen <heinzm@...hat.com>, dm-devel@...ts.linux.dev, 
	linux-kernel@...r.kernel.org
Subject: Re: [RFC PATCH v2] dm ioctl: fix erroneous EINVAL when signaled

On Tue, Jul 23, 2024 at 6:11 AM Zdenek Kabelac <zdenek.kabelac@...il.com> wrote:
>
> Dne 23. 07. 24 v 14:51 Mikulas Patocka napsal(a):
> >
> >
> > On Wed, 17 Jul 2024, Khazhismel Kumykov wrote:
> >
> >> do_resume when loading a new map first calls dm_suspend, which could
> >> silently fail. When we proceeded to dm_swap_table, we would bail out
> >> with EINVAL. Instead, attempt to restore new_map and return ERESTARTSYS
> >> when signaled.
> >>
> >> Signed-off-by: Khazhismel Kumykov <khazhy@...gle.com>
> >> ---
> >>   drivers/md/dm-ioctl.c | 23 +++++++++++++++++++++--
> >>   1 file changed, 21 insertions(+), 2 deletions(-)
> >>
> >> v2: don't leak new_map if we can't assign it back to hc.
> >>
> >> diff --git a/drivers/md/dm-ioctl.c b/drivers/md/dm-ioctl.c
> >> index c2c07bfa6471..0591455ad63c 100644
> >> --- a/drivers/md/dm-ioctl.c
> >> +++ b/drivers/md/dm-ioctl.c
> >> @@ -1181,8 +1181,27 @@ static int do_resume(struct dm_ioctl *param)
> >>                      suspend_flags &= ~DM_SUSPEND_LOCKFS_FLAG;
> >>              if (param->flags & DM_NOFLUSH_FLAG)
> >>                      suspend_flags |= DM_SUSPEND_NOFLUSH_FLAG;
> >> -            if (!dm_suspended_md(md))
> >> -                    dm_suspend(md, suspend_flags);
> >> +            if (!dm_suspended_md(md)) {
> >> +                    r = dm_suspend(md, suspend_flags);
> >> +                    if (r == -EINTR)
> >> +                            r = -ERESTARTSYS;
> >
> > I'd like to ask why the "EINTR -> ERESTARTSYS" conversion is here and why
> > it isn't in dm_suspend?
I proposed ERESTARTSYS here since the act of waiting for the device to
suspend successfully seems "restartable" - I think the same reasoning
would apply to do_suspend.
> >
> > What do libdevmapper+lvm maintainers think about it? Does lvm hadle EINTR
> > by restarting the ioctl syscall? Should we return ERESTARTSYS when suspend
> > is interrupted?
>
> In general - with suspend failures - we are just stopping whole operation -
> and restoring previous state - so user can run operation again.
>
> There is no special check for exact reason of ioctl failure.
>
> Regards
>
> Zdenek
>