linux.git/arch/x86/kernel/fpu, branch v5.6

x86/fpu: Deactivate FPU state after failure during state load

2020-01-07T12:44:42+00:00

In __fpu__restore_sig(), fpu_fpregs_owner_ctx needs to be reset if the
FPU state was not fully restored. Otherwise the following may happen (on
the same CPU):

  Task A                     Task B               fpu_fpregs_owner_ctx
  *active*                                        A.fpu
  __fpu__restore_sig()
                             ctx switch           load B.fpu
                             *active*             B.fpu
  fpregs_lock()
  copy_user_to_fpregs_zeroing()
    copy_kernel_to_xregs() *modify*
    copy_user_to_xregs() *fails*
  fpregs_unlock()
                            ctx switch            skip loading B.fpu,
                            *active*              B.fpu

In the success case, fpu_fpregs_owner_ctx is set to the current task.

In the failure case, the FPU state might have been modified by loading
the init state.

In this case, fpu_fpregs_owner_ctx needs to be reset in order to ensure
that the FPU state of the following task is loaded from saved state (and
not skipped because it was the previous state).

Reset fpu_fpregs_owner_ctx after a failure during restore occurred, to
ensure that the FPU state for the next task is always loaded.

The problem was debugged-by Yu-cheng Yu .

 [ bp: Massage commit message. ]

Fixes: 5f409e20b7945 ("x86/fpu: Defer FPU state load until return to userspace")
Reported-by: Yu-cheng Yu 
Signed-off-by: Sebastian Andrzej Siewior 
Signed-off-by: Borislav Petkov 
Cc: Andy Lutomirski 
Cc: Dave Hansen 
Cc: Fenghua Yu 
Cc: "H. Peter Anvin" 
Cc: Ingo Molnar 
Cc: Jann Horn 
Cc: Peter Zijlstra 
Cc: "Ravi V. Shankar" 
Cc: Rik van Riel 
Cc: Thomas Gleixner 
Cc: Tony Luck 
Cc: x86-ml 
Link: https://lkml.kernel.org/r/20191220195906.plk6kpmsrikvbcfn@linutronix.de

x86/fpu/xstate: Make xfeature_is_supervisor()/xfeature_is_user() return bool

2020-01-06T18:08:40+00:00

Have both xfeature_is_supervisor()/xfeature_is_user() return bool
because they are used only in boolean context.

Suggested-by: Borislav Petkov 
Signed-off-by: Yu-cheng Yu 
Signed-off-by: Borislav Petkov 
Acked-by: Sebastian Andrzej Siewior 
Cc: Andy Lutomirski 
Cc: Dave Hansen 
Cc: Fenghua Yu 
Cc: "H. Peter Anvin" 
Cc: Ingo Molnar 
Cc: Peter Zijlstra 
Cc: "Ravi V. Shankar" 
Cc: Rik van Riel 
Cc: Thomas Gleixner 
Cc: Tony Luck 
Cc: x86-ml 
Link: https://lkml.kernel.org/r/20191212210855.19260-3-yu-cheng.yu@intel.com

x86/fpu/xstate: Fix small issues

2020-01-06T16:31:11+00:00

In response to earlier comments, fix small issues before introducing
XSAVES supervisor states:

- Fix comments of xfeature_is_supervisor().
- Replace ((u64)1 << 63) with XCOMP_BV_COMPACTED_FORMAT.

No functional changes.

Signed-off-by: Yu-cheng Yu 
Signed-off-by: Borislav Petkov 
Reviewed-by: Dave Hansen 
Reviewed-by: Tony Luck 
Acked-by: Sebastian Andrzej Siewior 
Cc: Andy Lutomirski 
Cc: Dave Hansen 
Cc: Fenghua Yu 
Cc: "H. Peter Anvin" 
Cc: Ingo Molnar 
Cc: Peter Zijlstra 
Cc: "Ravi V. Shankar" 
Cc: Rik van Riel 
Cc: Thomas Gleixner 
Cc: x86-ml 
Link: https://lkml.kernel.org/r/20191212210855.19260-2-yu-cheng.yu@intel.com

treewide: Use sizeof_field() macro

2019-12-09T18:36:44+00:00

Replace all the occurrences of FIELD_SIZEOF() with sizeof_field() except
at places where these are defined. Later patches will remove the unused
definition of FIELD_SIZEOF().

This patch is generated using following script:

EXCLUDE_FILES="include/linux/stddef.h|include/linux/kernel.h"

git grep -l -e "\bFIELD_SIZEOF\b" | while read file;
do

	if [[ "$file" =~ $EXCLUDE_FILES ]]; then
		continue
	fi
	sed -i  -e 's/\bFIELD_SIZEOF\b/sizeof_field/g' $file;
done

Signed-off-by: Pankaj Bharadiya 
Link: https://lore.kernel.org/r/20190924105839.110713-3-pankaj.laxminarayan.bharadiya@intel.com
Co-developed-by: Kees Cook 
Signed-off-by: Kees Cook 
Acked-by: David Miller  # for net

x86/fpu: Use XFEATURE_FP/SSE enum values instead of hardcoded numbers

2019-11-04T21:10:07+00:00

When setting up sizes and offsets for legacy header entries the code uses
hardcoded 0/1 instead of the corresponding enum values XFEATURE_FP and
XFEATURE_SSE.

Replace the hardcoded numbers which enhances readability of the code and
also makes this code the first user of those enum values..

Signed-off-by: Cyrill Gorcunov 
Signed-off-by: Thomas Gleixner 
Link: https://lkml.kernel.org/r/20191101130153.GG1615@uranus.lan

x86/fpu: Shrink space allocated for xstate_comp_offsets

2019-11-04T21:04:19+00:00

commit 8ff925e10f2c ("x86/xsaves: Clean up code in xstate offsets
computation in xsave area") introduced an allocation of 64 entries for
xstate_comp_offsets while the code only handles up to XFEATURE_MAX entries.

For this reason xstate_offsets and xstate_sizes are already defined with
the explicit XFEATURE_MAX limit.

Do the same for compressed format for consistency sake.

As the changelog of that commit is not giving any information it's assumed
that the main idea was to cover all possible bits in xfeatures_mask, but
this doesn't explain why other variables such as the non-compacted offsets
and sizes are explicitely limited to XFEATURE_MAX.

For consistency it's better to use the XFEATURE_MAX limit everywhere and
extend it on demand when new features get implemented at the hardware
level and subsequently supported by the kernel.

Signed-off-by: Cyrill Gorcunov 
Signed-off-by: Thomas Gleixner 
Link: https://lkml.kernel.org/r/20191101124228.GF1615@uranus.lan

x86/fpu: Update stale variable name in comment

2019-11-04T21:04:19+00:00

When the fpu code was reworked pcntxt_mask was renamed to
xfeatures_mask. Reflect it in the comment as well.

Signed-off-by: Cyrill Gorcunov 
Signed-off-by: Thomas Gleixner 
Link: https://lkml.kernel.org/r/20191101123850.GE1615@uranus.lan

Merge branch 'x86-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip

2019-07-09T00:28:57+00:00

Pull x86 AVX512 status update from Ingo Molnar:
 "This adds a new ABI that the main scheduler probably doesn't want to
  deal with but HPC job schedulers might want to use: the
  AVX512_elapsed_ms field in the new /proc//arch_status task status
  file, which allows the user-space job scheduler to cluster such tasks,
  to avoid turbo frequency drops"

* 'x86-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  Documentation/filesystems/proc.txt: Add arch_status file
  x86/process: Add AVX-512 usage elapsed time to /proc/pid/arch_status
  proc: Add /proc//arch_status

x86/fpu: Inline fpu__xstate_clear_all_cpu_caps()

2019-07-07T10:01:47+00:00

All fpu__xstate_clear_all_cpu_caps() does is to invoke one simple
function since commit

  73e3a7d2a7c3b ("x86/fpu: Remove the explicit clearing of XSAVE dependent features")

so invoke that function directly and remove the wrapper.

Signed-off-by: Sebastian Andrzej Siewior 
Signed-off-by: Thomas Gleixner 
Link: https://lkml.kernel.org/r/20190704060743.rvew4yrjd6n33uzx@linutronix.de

x86/fpu: Make 'no387' and 'nofxsr' command line options useful

2019-07-07T10:01:46+00:00

The command line option `no387' is designed to disable the FPU
entirely. This only 'works' with CONFIG_MATH_EMULATION enabled.

But on 64bit this cannot work because user space expects SSE to work which
required basic FPU support. MATH_EMULATION does not help because SSE is not
emulated.

The command line option `nofxsr' should also be limited to 32bit because
FXSR is part of the required flags on 64bit so turning it off is not
possible.

Clearing X86_FEATURE_FPU without emulation enabled will not work anyway and
hang in fpu__init_system_early_generic() before the console is enabled.

Setting additioal dependencies, ensures that the CPU still boots on a
modern CPU. Otherwise, dropping FPU will leave FXSR enabled causing the
kernel to crash early in fpu__init_system_mxcsr().

With XSAVE support it will crash in fpu__init_cpu_xstate(). The problem is
that xsetbv() with XMM set and SSE cleared is not allowed.  That means
XSAVE has to be disabled. The XSAVE support is disabled in
fpu__init_system_xstate_size_legacy() but it is too late. It can be
removed, it has been added in commit

  1f999ab5a1360 ("x86, xsave: Disable xsave in i387 emulation mode")

to use `no387' on a CPU with XSAVE support.

All this happens before console output.

After hat, the next possible crash is in RAID6 detect code because MMX
remained enabled. With a 3DNOW enabled config it will explode in memcpy()
for instance due to kernel_fpu_begin() but this is unconditionally enabled.

This is enough to boot a Debian Wheezy on a 32bit qemu "host" CPU which
supports everything up to XSAVES, AVX2 without 3DNOW. Later, Debian
increased the minimum requirements to i686 which means it does not boot
userland atleast due to CMOV.

After masking the additional features it still keeps SSE4A and 3DNOW*
enabled (if present on the host) but those are unused in the kernel.

Restrict `no387' and `nofxsr' otions to 32bit only. Add dependencies for
FPU, FXSR to additionaly mask CMOV, MMX, XSAVE if FXSR or FPU is cleared.

Reported-by: Vegard Nossum 
Signed-off-by: Sebastian Andrzej Siewior 
Signed-off-by: Thomas Gleixner 
Link: https://lkml.kernel.org/r/20190703083247.57kjrmlxkai3vpw3@linutronix.de