linux-kernel - Re: [PATCH 11/18] staging: gasket: always allow root open for write

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:   Sun, 15 Jul 2018 21:41:50 +0200
From:   Greg Kroah-Hartman <gregkh@...uxfoundation.org>
To:     Todd Poynor <toddpoynor@...il.com>
Cc:     Dmitry Torokhov <dtor@...omium.org>, devel@...verdev.osuosl.org,
        Zhongze Hu <frankhu@...omium.org>,
        John Joseph <jnjoseph@...gle.com>,
        lkml <linux-kernel@...r.kernel.org>,
        Simon Que <sque@...omium.org>,
        Rob Springer <rspringer@...gle.com>,
        Guenter Roeck <groeck@...omium.org>,
        Todd Poynor <toddpoynor@...gle.com>
Subject: Re: [PATCH 11/18] staging: gasket: always allow root open for write

On Sun, Jul 15, 2018 at 11:15:26AM -0700, Todd Poynor wrote:
> On Sun, Jul 15, 2018 at 3:03 AM, Greg Kroah-Hartman
> <gregkh@...uxfoundation.org> wrote:
> > On Sun, Jul 15, 2018 at 12:53:09PM +0300, Dmitry Torokhov wrote:
> >> > I can't wait for people to just realize this whole "new" subsystem can
> >> > be replaced with UIO, but that's a topic for a different thread...
> >>
> >> Yes, that is true and that is why I am not sure why we are going
> >> through all this staging exercise.
> >>
> >> As far as I understand we'd still need to have quite a bit of kernel
> >> code so that we can safely program DMA controller (it does not look
> >> like uio_dmem_genirq.c is sufficient as is for gasket needs), but that
> >> should be solvable.
> >
> > I agree, it should be solvable, and much smaller and simpler than this
> > whole large chunk of "subsystem+driver" code.  But I'm not the one
> > having to do this work, and it provides a bunch of easy cleanups for
> > people looking to get into kernel development, so I don't mind :)
> >
> > But the "maintainers" should keep this in mind, as it is, this code is
> > _not_ acceptable for the main kernel tree because of the UIO framework
> > already present.
> 
> My own preference is to rewrite the apex driver entirely in-kernel and
> pull in its userspace parts here.  If I don't receive significant
> pushback on that I'll start doing that real soon.

Why would we object to that?