linux-kernel - Re: [PATCH 6/9] x86/mm: Introduce ptep_set_wrprotect

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

Date:   Thu, 7 Jun 2018 11:24:45 -0700
From:   Andy Lutomirski <luto@...nel.org>
To:     Dave Hansen <dave.hansen@...ux.intel.com>
Cc:     Yu-cheng Yu <yu-cheng.yu@...el.com>,
        LKML <linux-kernel@...r.kernel.org>, linux-doc@...r.kernel.org,
        Linux-MM <linux-mm@...ck.org>,
        linux-arch <linux-arch@...r.kernel.org>, X86 ML <x86@...nel.org>,
        "H. Peter Anvin" <hpa@...or.com>,
        Thomas Gleixner <tglx@...utronix.de>,
        Ingo Molnar <mingo@...hat.com>,
        "H. J. Lu" <hjl.tools@...il.com>,
        "Shanbhogue, Vedvyas" <vedvyas.shanbhogue@...el.com>,
        "Ravi V. Shankar" <ravi.v.shankar@...el.com>,
        Jonathan Corbet <corbet@....net>,
        Oleg Nesterov <oleg@...hat.com>, Arnd Bergmann <arnd@...db.de>,
        mike.kravetz@...cle.com
Subject: Re: [PATCH 6/9] x86/mm: Introduce ptep_set_wrprotect_flush and
 related functions

On Thu, Jun 7, 2018 at 11:23 AM Dave Hansen <dave.hansen@...ux.intel.com> wrote:
>
> On 06/07/2018 09:24 AM, Andy Lutomirski wrote:
> >> +static inline void ptep_set_wrprotect_flush(struct vm_area_struct *vma,
> >> +                                           unsigned long addr, pte_t *ptep)
> >> +{
> >> +       bool rw;
> >> +
> >> +       rw = test_and_clear_bit(_PAGE_BIT_RW, (unsigned long *)&ptep->pte);
> >> +       if (IS_ENABLED(CONFIG_X86_INTEL_SHADOW_STACK_USER)) {
> >> +               struct mm_struct *mm = vma->vm_mm;
> >> +               pte_t pte;
> >> +
> >> +               if (rw && (atomic_read(&mm->mm_users) > 1))
> >> +                       pte = ptep_clear_flush(vma, addr, ptep);
> > Why are you clearing the pte?
>
> I think I insisted on this being in there.
>
> First of all, we need to flush the TLB eventually because we need the
> shadowstack PTE permissions to be in effect.
>
> But, generally, we can't clear a dirty bit in a "live" PTE without
> flushing.  The processor can keep writing until we flush, and even keep
> setting it whenever _it_ allows a write, which it can do based on stale
> TLB contents.  Practically, I think a walk to set the dirty bit is
> mostly the same as a TLB miss, but that's certainly not guaranteed forever.
>
> That's even ignoring all the fun errata we have.

Maybe make it a separate patch, then?  It seems like this patch is
doing two almost entirely separate things: adding some flushes and
adding the magic hwdirty -> swdirty transition.  As it stands, it's
quite difficult to read the patch and figure out what it does.