2019-05-31 15:14:51

by Marco Elver

Subject: [PATCH v3 2/3] x86: Use static_cpu_has in uaccess region to avoid instrumentation

This patch is a pre-requisite for enabling KASAN bitops instrumentation;
using static_cpu_has instead of boot_cpu_has avoids instrumentation of
test_bit inside the uaccess region. With instrumentation, the KASAN
check would otherwise be flagged by objtool.

For consistency, kernel/signal.c was changed to mirror this change,
however, is never instrumented with KASAN (currently unsupported under
x86 32bit).

Signed-off-by: Marco Elver <[email protected]>
Suggested-by: H. Peter Anvin <[email protected]>
---
Changes in v3:
* Use static_cpu_has instead of moving boot_cpu_has outside uaccess
region.

Changes in v2:
* Replaces patch: 'tools/objtool: add kasan_check_* to uaccess
whitelist'
---
arch/x86/ia32/ia32_signal.c | 2 +-
arch/x86/kernel/signal.c | 2 +-
2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/arch/x86/ia32/ia32_signal.c b/arch/x86/ia32/ia32_signal.c
index 629d1ee05599..1cee10091b9f 100644
--- a/arch/x86/ia32/ia32_signal.c
+++ b/arch/x86/ia32/ia32_signal.c
@@ -358,7 +358,7 @@ int ia32_setup_rt_frame(int sig, struct ksignal *ksig,
put_user_ex(ptr_to_compat(&frame->uc), &frame->puc);

/* Create the ucontext. */
- if (boot_cpu_has(X86_FEATURE_XSAVE))
+ if (static_cpu_has(X86_FEATURE_XSAVE))
put_user_ex(UC_FP_XSTATE, &frame->uc.uc_flags);
else
put_user_ex(0, &frame->uc.uc_flags);
diff --git a/arch/x86/kernel/signal.c b/arch/x86/kernel/signal.c
index 364813cea647..52eb1d551aed 100644
--- a/arch/x86/kernel/signal.c
+++ b/arch/x86/kernel/signal.c
@@ -391,7 +391,7 @@ static int __setup_rt_frame(int sig, struct ksignal *ksig,
put_user_ex(&frame->uc, &frame->puc);

/* Create the ucontext. */
- if (boot_cpu_has(X86_FEATURE_XSAVE))
+ if (static_cpu_has(X86_FEATURE_XSAVE))
put_user_ex(UC_FP_XSTATE, &frame->uc.uc_flags);
else
put_user_ex(0, &frame->uc.uc_flags);
--
2.22.0.rc1.257.g3120a18244-goog

2019-06-07 10:01:16

by Marco Elver

Subject: Re: [PATCH v3 2/3] x86: Use static_cpu_has in uaccess region to avoid instrumentation

Gentle ping. I would appreciate quick feedback if this approach is reasonable.

Peter: since you suggested that we should not change objtool, did you
have a particular approach in mind that is maybe different from v2 and
v3? Or is this what you were thinking of?

Many thanks!

On Fri, 31 May 2019 at 17:11, Marco Elver <[email protected]> wrote:
>
> This patch is a pre-requisite for enabling KASAN bitops instrumentation;
> using static_cpu_has instead of boot_cpu_has avoids instrumentation of
> test_bit inside the uaccess region. With instrumentation, the KASAN
> check would otherwise be flagged by objtool.
>
> For consistency, kernel/signal.c was changed to mirror this change,
> however, is never instrumented with KASAN (currently unsupported under
> x86 32bit).
>
> Signed-off-by: Marco Elver <[email protected]>
> Suggested-by: H. Peter Anvin <[email protected]>
> ---
> Changes in v3:
> * Use static_cpu_has instead of moving boot_cpu_has outside uaccess
> region.
>
> Changes in v2:
> * Replaces patch: 'tools/objtool: add kasan_check_* to uaccess
> whitelist'
> ---
> arch/x86/ia32/ia32_signal.c | 2 +-
> arch/x86/kernel/signal.c | 2 +-
> 2 files changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/arch/x86/ia32/ia32_signal.c b/arch/x86/ia32/ia32_signal.c
> index 629d1ee05599..1cee10091b9f 100644
> --- a/arch/x86/ia32/ia32_signal.c
> +++ b/arch/x86/ia32/ia32_signal.c
> @@ -358,7 +358,7 @@ int ia32_setup_rt_frame(int sig, struct ksignal *ksig,
> put_user_ex(ptr_to_compat(&frame->uc), &frame->puc);
>
> /* Create the ucontext. */
> - if (boot_cpu_has(X86_FEATURE_XSAVE))
> + if (static_cpu_has(X86_FEATURE_XSAVE))
> put_user_ex(UC_FP_XSTATE, &frame->uc.uc_flags);
> else
> put_user_ex(0, &frame->uc.uc_flags);
> diff --git a/arch/x86/kernel/signal.c b/arch/x86/kernel/signal.c
> index 364813cea647..52eb1d551aed 100644
> --- a/arch/x86/kernel/signal.c
> +++ b/arch/x86/kernel/signal.c
> @@ -391,7 +391,7 @@ static int __setup_rt_frame(int sig, struct ksignal *ksig,
> put_user_ex(&frame->uc, &frame->puc);
>
> /* Create the ucontext. */
> - if (boot_cpu_has(X86_FEATURE_XSAVE))
> + if (static_cpu_has(X86_FEATURE_XSAVE))
> put_user_ex(UC_FP_XSTATE, &frame->uc.uc_flags);
> else
> put_user_ex(0, &frame->uc.uc_flags);
> --
> 2.22.0.rc1.257.g3120a18244-goog
>

2019-06-12 18:00:22

by Dmitry Vyukov

Subject: Re: [PATCH v3 2/3] x86: Use static_cpu_has in uaccess region to avoid instrumentation

On Fri, Jun 7, 2019 at 11:44 AM Marco Elver <[email protected]> wrote:
>
> Gentle ping. I would appreciate quick feedback if this approach is reasonable.
>
> Peter: since you suggested that we should not change objtool, did you
> have a particular approach in mind that is maybe different from v2 and
> v3? Or is this what you were thinking of?
>
> Many thanks!
>
> On Fri, 31 May 2019 at 17:11, Marco Elver <[email protected]> wrote:
> >
> > This patch is a pre-requisite for enabling KASAN bitops instrumentation;
> > using static_cpu_has instead of boot_cpu_has avoids instrumentation of
> > test_bit inside the uaccess region. With instrumentation, the KASAN
> > check would otherwise be flagged by objtool.
> >
> > For consistency, kernel/signal.c was changed to mirror this change,
> > however, is never instrumented with KASAN (currently unsupported under
> > x86 32bit).
> >
> > Signed-off-by: Marco Elver <[email protected]>
> > Suggested-by: H. Peter Anvin <[email protected]>
> > ---
> > Changes in v3:
> > * Use static_cpu_has instead of moving boot_cpu_has outside uaccess
> > region.
> >
> > Changes in v2:
> > * Replaces patch: 'tools/objtool: add kasan_check_* to uaccess
> > whitelist'
> > ---
> > arch/x86/ia32/ia32_signal.c | 2 +-
> > arch/x86/kernel/signal.c | 2 +-
> > 2 files changed, 2 insertions(+), 2 deletions(-)
> >
> > diff --git a/arch/x86/ia32/ia32_signal.c b/arch/x86/ia32/ia32_signal.c
> > index 629d1ee05599..1cee10091b9f 100644
> > --- a/arch/x86/ia32/ia32_signal.c
> > +++ b/arch/x86/ia32/ia32_signal.c
> > @@ -358,7 +358,7 @@ int ia32_setup_rt_frame(int sig, struct ksignal *ksig,
> > put_user_ex(ptr_to_compat(&frame->uc), &frame->puc);
> >
> > /* Create the ucontext. */
> > - if (boot_cpu_has(X86_FEATURE_XSAVE))
> > + if (static_cpu_has(X86_FEATURE_XSAVE))


Peter Z or A, does it look good to you? Could you please Ack this?


> > put_user_ex(UC_FP_XSTATE, &frame->uc.uc_flags);
> > else
> > put_user_ex(0, &frame->uc.uc_flags);
> > diff --git a/arch/x86/kernel/signal.c b/arch/x86/kernel/signal.c
> > index 364813cea647..52eb1d551aed 100644
> > --- a/arch/x86/kernel/signal.c
> > +++ b/arch/x86/kernel/signal.c
> > @@ -391,7 +391,7 @@ static int __setup_rt_frame(int sig, struct ksignal *ksig,
> > put_user_ex(&frame->uc, &frame->puc);
> >
> > /* Create the ucontext. */
> > - if (boot_cpu_has(X86_FEATURE_XSAVE))
> > + if (static_cpu_has(X86_FEATURE_XSAVE))
> > put_user_ex(UC_FP_XSTATE, &frame->uc.uc_flags);
> > else
> > put_user_ex(0, &frame->uc.uc_flags);
> > --
> > 2.22.0.rc1.257.g3120a18244-goog
> >

2019-06-13 15:34:41

by Andrey Ryabinin

Subject: Re: [PATCH v3 2/3] x86: Use static_cpu_has in uaccess region to avoid instrumentation

On 5/31/19 6:08 PM, Marco Elver wrote:
> This patch is a pre-requisite for enabling KASAN bitops instrumentation;
> using static_cpu_has instead of boot_cpu_has avoids instrumentation of
> test_bit inside the uaccess region. With instrumentation, the KASAN
> check would otherwise be flagged by objtool.
>
> For consistency, kernel/signal.c was changed to mirror this change,
> however, is never instrumented with KASAN (currently unsupported under
> x86 32bit).
>
> Signed-off-by: Marco Elver <[email protected]>
> Suggested-by: H. Peter Anvin <[email protected]>

Reviewed-by: Andrey Ryabinin <[email protected]>

2019-06-13 15:50:19

by Peter Zijlstra

Subject: Re: [PATCH v3 2/3] x86: Use static_cpu_has in uaccess region to avoid instrumentation

On Fri, May 31, 2019 at 05:08:30PM +0200, Marco Elver wrote:
> This patch is a pre-requisite for enabling KASAN bitops instrumentation;
> using static_cpu_has instead of boot_cpu_has avoids instrumentation of
> test_bit inside the uaccess region. With instrumentation, the KASAN
> check would otherwise be flagged by objtool.
>
> For consistency, kernel/signal.c was changed to mirror this change,
> however, is never instrumented with KASAN (currently unsupported under
> x86 32bit).

Acked-by: Peter Zijlstra (Intel) <[email protected]>

Thanks!

>
> Signed-off-by: Marco Elver <[email protected]>
> Suggested-by: H. Peter Anvin <[email protected]>
> ---
> Changes in v3:
> * Use static_cpu_has instead of moving boot_cpu_has outside uaccess
> region.
>
> Changes in v2:
> * Replaces patch: 'tools/objtool: add kasan_check_* to uaccess
> whitelist'
> ---
> arch/x86/ia32/ia32_signal.c | 2 +-
> arch/x86/kernel/signal.c | 2 +-
> 2 files changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/arch/x86/ia32/ia32_signal.c b/arch/x86/ia32/ia32_signal.c
> index 629d1ee05599..1cee10091b9f 100644
> --- a/arch/x86/ia32/ia32_signal.c
> +++ b/arch/x86/ia32/ia32_signal.c
> @@ -358,7 +358,7 @@ int ia32_setup_rt_frame(int sig, struct ksignal *ksig,
> put_user_ex(ptr_to_compat(&frame->uc), &frame->puc);
>
> /* Create the ucontext. */
> - if (boot_cpu_has(X86_FEATURE_XSAVE))
> + if (static_cpu_has(X86_FEATURE_XSAVE))
> put_user_ex(UC_FP_XSTATE, &frame->uc.uc_flags);
> else
> put_user_ex(0, &frame->uc.uc_flags);
> diff --git a/arch/x86/kernel/signal.c b/arch/x86/kernel/signal.c
> index 364813cea647..52eb1d551aed 100644
> --- a/arch/x86/kernel/signal.c
> +++ b/arch/x86/kernel/signal.c
> @@ -391,7 +391,7 @@ static int __setup_rt_frame(int sig, struct ksignal *ksig,
> put_user_ex(&frame->uc, &frame->puc);
>
> /* Create the ucontext. */
> - if (boot_cpu_has(X86_FEATURE_XSAVE))
> + if (static_cpu_has(X86_FEATURE_XSAVE))
> put_user_ex(UC_FP_XSTATE, &frame->uc.uc_flags);
> else
> put_user_ex(0, &frame->uc.uc_flags);
> --
> 2.22.0.rc1.257.g3120a18244-goog
>