linux-kernel - Re: [syzbot] [integrity] [overlayfs] possible deadlock in mnt_want

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 for Android: free password hash cracker in your pocket

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

Date:   Thu, 5 Oct 2023 16:22:20 +0300
From:   Amir Goldstein <amir73il@...il.com>
To:     Mimi Zohar <zohar@...ux.ibm.com>
Cc:     syzbot <syzbot+b42fe626038981fb7bfa@...kaller.appspotmail.com>,
        hdanton@...a.com, linux-fsdevel@...r.kernel.org,
        linux-integrity@...r.kernel.org, linux-kernel@...r.kernel.org,
        linux-mm@...ck.org, linux-unionfs@...r.kernel.org,
        miklos@...redi.hu, mszeredi@...hat.com,
        syzbot@...kalhler.appspotmail.com, syzkaller-bugs@...glegroups.com,
        viro@...iv.linux.org.uk
Subject: Re: [syzbot] [integrity] [overlayfs] possible deadlock in
 mnt_want_write (2)

On Thu, Oct 5, 2023 at 4:14 PM Mimi Zohar <zohar@...ux.ibm.com> wrote:
>
> On Thu, 2023-10-05 at 13:26 +0300, Amir Goldstein wrote:
> > On Thu, Oct 5, 2023 at 12:59 PM syzbot
> > <syzbot+b42fe626038981fb7bfa@...kaller.appspotmail.com> wrote:
> > >
> > > Hello,
> > >
> > > syzbot tried to test the proposed patch but the build/boot failed:
> >
> > My mistake. Please try again:
> >
> > #syz test: https://github.com/amir73il/linux ima-ovl-fix
>
> Thanks, Amir.   "mutext_init(&iint->mutex); moved, but the status
> initialization lines 161-166 were dropped.   They're needed by IMA-
> appraisal for signature verification.
>
>         iint->ima_file_status = INTEGRITY_UNKNOWN;
>         iint->ima_mmap_status = INTEGRITY_UNKNOWN;
>         iint->ima_bprm_status = INTEGRITY_UNKNOWN;
>         iint->ima_read_status = INTEGRITY_UNKNOWN;
>         iint->ima_creds_status = INTEGRITY_UNKNOWN;
>         iint->evm_status = INTEGRITY_UNKNOWN;
>

They are dropped from iint_init_once()
They are not needed there because there are now set
in every iint allocation in iint_init_always()
instead of being set in iint_free()

This is the standard practice for slab objects.
See inode_init_once()/inode_init_always().

Thanks,
Amir.