Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1394764 > unrolled thread

Re: [PATCH v4 10/10] x86/xsaves: Re-enable XSAVES

Back to article view | Back to linux.kernel

This discussion starts older than the indexed window; earlier articles aren't shown. The article labeled Started by below is the oldest one visible, not the original post.

Contents

  Re: [PATCH v4 10/10] x86/xsaves: Re-enable XSAVES Dave Hansen <dave.hansen@linux.intel.com> - 2016-05-05 00:50 +0200
    Re: [PATCH v4 10/10] x86/xsaves: Re-enable XSAVES Yu-cheng Yu <yu-cheng.yu@intel.com> - 2016-05-05 01:00 +0200

#1394764 — Re: [PATCH v4 10/10] x86/xsaves: Re-enable XSAVES

It's my fault, but you also need to go update
	
	fpu__xfeature_set_state()
and
	__raw_xsave_addr()

The theoretical problem is that you might ask for a __raw_xsave_addr()
of a component which has been compacted out of an XSAVES buffer and thus
has no address.  We could work around this by doing a memmove() and
moving the components "up" after the one we are trying to set in order
to make space.

But, since we *always* call XSAVES with an instruction mask of -1 and
end up with a requested feature bitmap (RFBM) equal to XCR0, I think we
can do a shortcut because we'll practically *always* have an
xcomp_bv==RFBM==XCR0, which means that all (present) components will
always have an address.

So, the alternative to doing the memmove() is to add some WARN_ON_FPU()
checks to enforce xcomp_bv==RFBM==XCR0 in places where we call
XSAVES/XRSTORS and __raw_xsave_addr(), maybe more.

[toc] | [next] | [standalone]

#1394768

On Wed, May 04, 2016 at 03:41:49PM -0700, Dave Hansen wrote:
> It's my fault, but you also need to go update
> 	
> 	fpu__xfeature_set_state()
> and
> 	__raw_xsave_addr()
> 
> The theoretical problem is that you might ask for a __raw_xsave_addr()
> of a component which has been compacted out of an XSAVES buffer and thus
> has no address.  We could work around this by doing a memmove() and
> moving the components "up" after the one we are trying to set in order
> to make space.
> 
> But, since we *always* call XSAVES with an instruction mask of -1 and
> end up with a requested feature bitmap (RFBM) equal to XCR0, I think we
> can do a shortcut because we'll practically *always* have an
> xcomp_bv==RFBM==XCR0, which means that all (present) components will
> always have an address.
> 
> So, the alternative to doing the memmove() is to add some WARN_ON_FPU()
> checks to enforce xcomp_bv==RFBM==XCR0 in places where we call
> XSAVES/XRSTORS and __raw_xsave_addr(), maybe more.

In the coming version 5 patches, we are going to have one additional 
patch for fixing __fpu_restore_sig() for the compacted format.  I changed
my existing patch a little and run into some problems. Fixing it now.

Our ptrace tests went OK before, but are failing now. It might be relating
to what you are saying? I will check it.
 

[toc] | [prev] | [standalone]

Back to top | Article view | linux.kernel

csiph-web