Re: [Xen-devel] [PATCH] x86/amd: don't set X86_BUG_SYSRET_SS_ATTRS if forced to zero

From	Andrew Cooper <andrew.cooper3@citrix.com>
Newsgroups	linux.kernel
Subject	Re: [Xen-devel] [PATCH] x86/amd: don't set X86_BUG_SYSRET_SS_ATTRS if forced to zero
Date	2017-04-25 22:20 +0200
Message-ID	<tAf6i-Vi-9@gated-at.bofh.it> (permalink)
References	<tAd4t-84X-3@gated-at.bofh.it> <tAdnQ-8d5-39@gated-at.bofh.it> <tAdxw-8iW-47@gated-at.bofh.it> <tAead-kQ-13@gated-at.bofh.it>
Organization	linux.* mail to news gateway

Show all headers | View raw

On 25/04/17 20:18, Borislav Petkov wrote:
> On Tue, Apr 25, 2017 at 08:34:34PM +0200, Juergen Gross wrote:
>> And what happens when there is a scheduling event right here?
>> __switch_to() will see X86_BUG_SYSRET_SS_ATTRS set and take a wrong
>> path.
> So the whole thing we're doing right now is wrong: set bit and then
> clear bit.
>
> We should not set the bit at all and there won't be any window to get it
> wrong.
>
> So can we do something like this instead:
>
> 	if (!cpu_has(c, X86_FEATURE_XENPV))
> 		set_cpu_bug(c, X86_BUG_SYSRET_SS_ATTRS);
>
> or is XENPV the wrong thing to test?
>

X86_BUG_SYSRET_SS_ATTRS only actually causes a problem if you enter the
kernel via an IDT vector (forcing %ss to NULL), then exiting the kernel
via the optimistic sysret path, which on AMD loads the %ss selector, but
apparently doesn't update the segment cache (and %ss.dpl in particular).

The problem (for all ring-deprivileged virtuailsation; not just Xen PV),
is that

savesegment(ss, ss_sel);
if (ss_sel != __KERNEL_DS)
    loadsegment(ss, __KERNEL_DS);

tries to load %ss with an RPL of 0, and things blow up.

~Andrew

Thread

[PATCH] x86/amd: don't set X86_BUG_SYSRET_SS_ATTRS if forced to zero Juergen Gross <jgross@suse.com> - 2017-04-25 20:10 +0200
  Re: [PATCH] x86/amd: don't set X86_BUG_SYSRET_SS_ATTRS if forced to  zero Borislav Petkov <bp@alien8.de> - 2017-04-25 20:30 +0200
    Re: [PATCH] x86/amd: don't set X86_BUG_SYSRET_SS_ATTRS if forced to  zero Juergen Gross <jgross@suse.com> - 2017-04-25 20:40 +0200
      Re: [PATCH] x86/amd: don't set X86_BUG_SYSRET_SS_ATTRS if forced to  zero Borislav Petkov <bp@alien8.de> - 2017-04-25 21:20 +0200
        Re: [Xen-devel] [PATCH] x86/amd: don't set X86_BUG_SYSRET_SS_ATTRS if  forced to zero Andrew Cooper <andrew.cooper3@citrix.com> - 2017-04-25 22:20 +0200
          Re: [Xen-devel] [PATCH] x86/amd: don't set X86_BUG_SYSRET_SS_ATTRS  if forced to zero Borislav Petkov <bp@alien8.de> - 2017-04-25 22:30 +0200
        Re: [PATCH] x86/amd: don't set X86_BUG_SYSRET_SS_ATTRS if forced to  zero Juergen Gross <jgross@suse.com> - 2017-04-26 06:50 +0200
          Re: [PATCH] x86/amd: don't set X86_BUG_SYSRET_SS_ATTRS if forced to  zero Borislav Petkov <bp@alien8.de> - 2017-04-26 08:40 +0200
            Re: [PATCH] x86/amd: don't set X86_BUG_SYSRET_SS_ATTRS if forced to  zero Juergen Gross <jgross@suse.com> - 2017-04-26 20:30 +0200
              Re: [PATCH] x86/amd: don't set X86_BUG_SYSRET_SS_ATTRS if forced to  zero Borislav Petkov <bp@alien8.de> - 2017-04-27 00:10 +0200
                Re: [PATCH] x86/amd: don't set X86_BUG_SYSRET_SS_ATTRS if forced to  zero Juergen Gross <jgross@suse.com> - 2017-04-27 06:50 +0200

csiph-web