[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20180108210355.GA11256@1wt.eu>
Date: Mon, 8 Jan 2018 22:03:55 +0100
From: Willy Tarreau <w@....eu>
To: Thomas Gleixner <tglx@...utronix.de>
Cc: Linus Torvalds <torvalds@...ux-foundation.org>,
Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
the arch/x86 maintainers <x86@...nel.org>,
One Thousand Gnomes <gnomes@...rguk.ukuu.org.uk>,
Andy Lutomirski <luto@...nel.org>
Subject: Re: [PATCH RFC 2/4] x86/arch_prctl: add ARCH_GET_NOPTI and
ARCH_SET_NOPTI to enable/disable PTI
On Mon, Jan 08, 2018 at 09:49:23PM +0100, Thomas Gleixner wrote:
> On Mon, 8 Jan 2018, Willy Tarreau wrote:
> > On Mon, Jan 08, 2018 at 09:54:05AM -0800, Linus Torvalds wrote:
> >
> > > So make the thing per-mm, and then at task switch time as you switch
> > > mms, you set the bit in a percpu variable for testing at kernel entry.
> >
> > I'll see how to do that, this is not yet 100% clear to me, I'm still
> > discovering (this code has immensely changed since last time I *really*
> > dug into it). So I suspect I'll have to set this variable in
> > __switch_to() based on this other MM flag.
>
> The right thing is probably the cpu entry area because that's what you can
> access before switching CR3 if PTI is enabled for the task
In fact I didn't need any such thing, because when coming from user
space I can simply check CR3.12 : if 0, we had PTI disabled for the
task, otherwise it's enabled.
I'll repost the cleaned up series ASAP, unfortunately I need to go now.
Thanks,
Willy
Powered by blists - more mailing lists