[llvm-dev] retpoline mitigation and 6.0
Chandler Carruth via llvm-dev
llvm-dev at lists.llvm.org
Fri Feb 2 18:04:31 PST 2018
On Fri, Feb 2, 2018 at 5:56 PM Guenter Roeck <linux at roeck-us.net> wrote:
> On 02/02/2018 04:27 PM, Chandler Carruth wrote:
> > On Fri, Feb 2, 2018 at 4:23 PM Chandler Carruth <chandlerc at google.com
> <mailto:chandlerc at google.com>> wrote:
> > On Fri, Feb 2, 2018 at 4:03 PM David Woodhouse <dwmw2 at infradead.org
> <mailto:dwmw2 at infradead.org>> wrote:
> > On Thu, 2018-02-01 at 10:10 +0100, Hans Wennborg via llvm-dev
> > >
> > > I saw the retpoline mitigation landed in r323155. Are we
> ready to
> > > merge this to 6.0, or are there any open issues that we're
> > > for? Also, were there any followups I should know about? Also,
> > > release notes please :-)
> > Eep, please can we keep the command line option for clang and
> the thunk
> > ABI matching what GCC, the Linux kernel and Xen are all doing?
> > To say that I am not stunningly keen on
> > https://lkml.org/lkml/2018/2/2/975 would be a bit of an
> > understatement...
> > A minor note on this specific patch:
> > You don't need '-mretpoline -mretpoline-external-thunk'. The second flag
> implies the first. (Or should, if not, its a bug.) Our goal was that you
> needed exactly one flag to enable this in whatever form.
> The llvm commit log says:
> "... They can write this custom thunk and use `-mretpoline-external-thunk`
> *in addition* to `-mretpoline`. In this case, on x86-64 thu thunk names
> must be: ... "
> which I understand to mean that _both_ are needed. I'll be happy to stand
> corrected if that is guaranteed not to be the case.
Gah, sorry. I thought I had updated that before landing, but I must have
The intent is that you *do not* need both. That changed during review, but
that was the intent. The test seems to confirm that this should work. If
you see anything that makes you think it isn't working, please give a shout
(or file a bug) and we'll fix.
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the llvm-dev