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

Groups > linux.kernel > #1565182

Re: [tip:x86/urgent] x86/fpu: Set the xcomp_bv when we fake up a XSAVES area

Show all headers | View raw

On 01/23/2017 08:55 AM, Yu-cheng Yu wrote:
> On Mon, Jan 23, 2017 at 07:36:20AM -0800, Dave Hansen wrote:
>> The kernel xsave buffer should *ALWAYS* have the
>> XCOMP_BV_COMPACTED_FORMAT bit set.  It should have been set before the
>> copyin and it should be set when it's finished.
>>
>> The best fix here would be not to paper over the issue in the copy
>> function but find where it got clobbered, or where some initialization
>> code failed to set it.
> 
> Someone else reported different issues from the same bug and a different
> patch was just tested OK this morning.  I think that adding xfeatures bits
> to xcomp_bv should have been done in fpstate_init().

Right.  So where did it get cleared out?

> Also, in copy_init_fpstate_to_fpregs(), we do:
> 
> 	copy_kernel_to_xregs(&init_fpstate.xsave, -1).
> 
> That (-1) could mean (0) because the parameters are declared as:
> 
> 	copy_kernel_to_xregs(struct xregs_state *, u64)

I'm not sure what you're saying.  -1 just means "all 1's" when cast to
an unsigned type.  This shouldn't case any problems.

Back to linux.kernel | Previous | Next — Previous in thread | Next in thread | Find similar | Unroll thread

Thread

[PATCH] x86/fpu: set the xcomp_bv when we fake up a XSAVES area Kevin Hao <haokexin@gmail.com> - 2017-01-22 10:00 +0100
  [tip:x86/urgent] x86/fpu: Set the xcomp_bv when we fake up a XSAVES  area tip-bot for Kevin Hao <tipbot@zytor.com> - 2017-01-23 09:30 +0100
    Re: [tip:x86/urgent] x86/fpu: Set the xcomp_bv when we fake up a  XSAVES area Dave Hansen <dave.hansen@linux.intel.com> - 2017-01-23 16:40 +0100
      Re: [tip:x86/urgent] x86/fpu: Set the xcomp_bv when we fake up a  XSAVES area Yu-cheng Yu <yu-cheng.yu@intel.com> - 2017-01-23 18:00 +0100
        Re: [tip:x86/urgent] x86/fpu: Set the xcomp_bv when we fake up a  XSAVES area Dave Hansen <dave.hansen@linux.intel.com> - 2017-01-23 18:30 +0100
          Re: [tip:x86/urgent] x86/fpu: Set the xcomp_bv when we fake up a  XSAVES area Yu-cheng Yu <yu-cheng.yu@intel.com> - 2017-01-23 22:10 +0100
            Re: [tip:x86/urgent] x86/fpu: Set the xcomp_bv when we fake up a  XSAVES area Dave Hansen <dave.hansen@linux.intel.com> - 2017-01-23 22:20 +0100
              Re: [tip:x86/urgent] x86/fpu: Set the xcomp_bv when we fake up a  XSAVES area Yu-cheng Yu <yu-cheng.yu@intel.com> - 2017-01-23 22:30 +0100
                Re: [tip:x86/urgent] x86/fpu: Set the xcomp_bv when we fake up a  XSAVES area Dave Hansen <dave.hansen@linux.intel.com> - 2017-01-23 22:30 +0100
                Re: [tip:x86/urgent] x86/fpu: Set the xcomp_bv when we fake up a  XSAVES area Kevin Hao <haokexin@gmail.com> - 2017-01-24 01:20 +0100
                Re: [tip:x86/urgent] x86/fpu: Set the xcomp_bv when we fake up a  XSAVES area Dave Hansen <dave.hansen@linux.intel.com> - 2017-01-24 02:00 +0100
                Re: [tip:x86/urgent] x86/fpu: Set the xcomp_bv when we fake up a  XSAVES area Kevin Hao <haokexin@gmail.com> - 2017-01-24 03:00 +0100
                Re: [tip:x86/urgent] x86/fpu: Set the xcomp_bv when we fake up a  XSAVES area Dave Hansen <dave.hansen@linux.intel.com> - 2017-01-24 03:10 +0100
                Re: [tip:x86/urgent] x86/fpu: Set the xcomp_bv when we fake up a  XSAVES area Kevin Hao <haokexin@gmail.com> - 2017-01-24 03:20 +0100
                Re: [tip:x86/urgent] x86/fpu: Set the xcomp_bv when we fake up a  XSAVES area Dave Hansen <dave.hansen@linux.intel.com> - 2017-01-24 03:40 +0100
                Re: [tip:x86/urgent] x86/fpu: Set the xcomp_bv when we fake up a  XSAVES area Kevin Hao <haokexin@gmail.com> - 2017-01-24 06:20 +0100
                Re: [tip:x86/urgent] x86/fpu: Set the xcomp_bv when we fake up a  XSAVES area Ingo Molnar <mingo@kernel.org> - 2017-01-24 09:10 +0100
  [tip:x86/urgent] x86/fpu: Set the xcomp_bv when we fake up a XSAVES  area tip-bot for Kevin Hao <tipbot@zytor.com> - 2017-01-23 10:50 +0100

csiph-web