linux-kernel - Re: [PATCH] ARM: hw_breakpoint: Handle inexact watchpoint addresses

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]

Message-ID: <CAD=FV=Wntf0TCwdtNNvPY-CXX1VL_SZK8Y8yw1r=UfeayHfwgw@mail.gmail.com>
Date:   Mon, 2 Dec 2019 08:36:19 -0800
From:   Doug Anderson <dianders@...omium.org>
To:     Will Deacon <will@...nel.org>
Cc:     Mark Rutland <mark.rutland@....com>,
        Pavel Labath <labath@...gle.com>,
        Pratyush Anand <panand@...hat.com>,
        Matthias Kaehlcke <mka@...omium.org>,
        Kazuhiro Inaba <kinaba@...gle.com>,
        Russell King <linux@...linux.org.uk>,
        Linux ARM <linux-arm-kernel@...ts.infradead.org>,
        LKML <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] ARM: hw_breakpoint: Handle inexact watchpoint addresses

Hi,

On Wed, Nov 20, 2019 at 11:18 AM Will Deacon <will@...nel.org> wrote:
>
> On Sat, Oct 19, 2019 at 11:12:26AM -0700, Douglas Anderson wrote:
> > This is commit fdfeff0f9e3d ("arm64: hw_breakpoint: Handle inexact
> > watchpoint addresses") but ported to arm32, which has the same
> > problem.
> >
> > This problem was found by Android CTS tests, notably the
> > "watchpoint_imprecise" test [1].  I tested locally against a copycat
> > (simplified) version of the test though.
> >
> > [1] https://android.googlesource.com/platform/bionic/+/master/tests/sys_ptrace_test.cpp
> >
> > Signed-off-by: Douglas Anderson <dianders@...omium.org>
> > ---
> >
> >  arch/arm/kernel/hw_breakpoint.c | 96 ++++++++++++++++++++++++---------
> >  1 file changed, 70 insertions(+), 26 deletions(-)
>
> Sorry for taking so long to look at this. After wrapping my head around the
> logic again

Yeah.  It was a little weird and (unfortunately) arbitrarily different
in some places compared to the arm64 code.


> I think it looks fine, so please put it into the patch system
> with my Ack:
>
> Acked-by: Will Deacon <will@...nel.org>

Thanks!  Submitted as:

https://www.arm.linux.org.uk/developer/patches/viewpatch.php?id=8944/1


> One interesting difference between the implementation here and the arm64
> code is that I think if you have multiple watchpoints, all of which fire
> with a distance != 0, then arm32 will actually report them all whereas
> you'd only get one on arm64.

Are you sure about that?  The "/* No exact match found. */" code is
outside the for loop so it should only be able to trigger for exactly
one breakpoint, no?

-Doug