Re: [RFC PATCH for 4.15 05/10] x86: Introduce sync_core_before_usermode

[Mathieu Desnoyers]

----- On Nov 10, 2017, at 5:32 PM, Mathieu Desnoyers mathieu.desnoyers@xxxxxxxxxxxx wrote:

> ----- On Nov 10, 2017, at 5:20 PM, Mathieu Desnoyers
> mathieu.desnoyers@xxxxxxxxxxxx wrote:
> 
>> ----- On Nov 10, 2017, at 5:02 PM, Andy Lutomirski luto@xxxxxxxxxx wrote:
>> 
>>> On Fri, Nov 10, 2017 at 1:37 PM, Mathieu Desnoyers
>>> <mathieu.desnoyers@xxxxxxxxxxxx> wrote:
>>>> Introduce an architecture function that ensures the current CPU
>>>> issues a core serializing instruction before returning to usermode.
>>>>
>>>> This is needed to fix an existing core serialization bug on
>>>> thread migration, and also needed by the membarrier "sync_core" command.
>>>>
>>>> Architectures defining the sync_core_before_usermode() static inline
>>>> need to define ARCH_HAS_SYNC_CORE_BEFORE_USERMODE.
>>>>
>>>> Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@xxxxxxxxxxxx>
>>>> CC: Peter Zijlstra <peterz@xxxxxxxxxxxxx>
>>>> CC: Andy Lutomirski <luto@xxxxxxxxxx>
>>>> CC: Paul E. McKenney <paulmck@xxxxxxxxxxxxxxxxxx>
>>>> CC: Boqun Feng <boqun.feng@xxxxxxxxx>
>>>> CC: Andrew Hunter <ahh@xxxxxxxxxx>
>>>> CC: Maged Michael <maged.michael@xxxxxxxxx>
>>>> CC: Avi Kivity <avi@xxxxxxxxxxxx>
>>>> CC: Benjamin Herrenschmidt <benh@xxxxxxxxxxxxxxxxxxx>
>>>> CC: Paul Mackerras <paulus@xxxxxxxxx>
>>>> CC: Michael Ellerman <mpe@xxxxxxxxxxxxxx>
>>>> CC: Dave Watson <davejwatson@xxxxxx>
>>>> CC: Thomas Gleixner <tglx@xxxxxxxxxxxxx>
>>>> CC: Ingo Molnar <mingo@xxxxxxxxxx>
>>>> CC: "H. Peter Anvin" <hpa@xxxxxxxxx>
>>>> CC: Andrea Parri <parri.andrea@xxxxxxxxx>
>>>> CC: Russell King <linux@xxxxxxxxxxxxxxx>
>>>> CC: Greg Hackmann <ghackmann@xxxxxxxxxx>
>>>> CC: Will Deacon <will.deacon@xxxxxxx>
>>>> CC: David Sehr <sehr@xxxxxxxxxx>
>>>> CC: Linus Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx>
>>>> CC: x86@xxxxxxxxxx
>>>> CC: linux-arch@xxxxxxxxxxxxxxx
>>>> ---
>>>>  arch/x86/Kconfig                 |  1 +
>>>>  arch/x86/include/asm/processor.h | 10 ++++++++++
>>>>  include/linux/processor.h        |  6 ++++++
>>>>  3 files changed, 17 insertions(+)
>>>>
>>>> diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig
>>>> index 01f78c1d40b5..54fbb8960d94 100644
>>>> --- a/arch/x86/Kconfig
>>>> +++ b/arch/x86/Kconfig
>>>> @@ -62,6 +62,7 @@ config X86
>>>>         select ARCH_HAS_SG_CHAIN
>>>>         select ARCH_HAS_STRICT_KERNEL_RWX
>>>>         select ARCH_HAS_STRICT_MODULE_RWX
>>>> +       select ARCH_HAS_SYNC_CORE_BEFORE_USERMODE
>>>>         select ARCH_HAS_UBSAN_SANITIZE_ALL
>>>>         select ARCH_HAS_ZONE_DEVICE             if X86_64
>>>>         select ARCH_HAVE_NMI_SAFE_CMPXCHG
>>>> diff --git a/arch/x86/include/asm/processor.h b/arch/x86/include/asm/processor.h
>>>> index bdac19ab2488..6ce996a7c730 100644
>>>> --- a/arch/x86/include/asm/processor.h
>>>> +++ b/arch/x86/include/asm/processor.h
>>>> @@ -706,6 +706,16 @@ static inline void sync_core(void)
>>>>  #endif
>>>>  }
>>>>
>>>> +/*
>>>> + * Ensure that a core serializing instruction is issued before returning
>>>> + * to user-mode. x86 implements return to user-space through sysexit and
>>>> + * sysretq, which are not core serializing.
>>>> + */
>>>> +static inline void sync_core_before_usermode(void)
>>>> +{
>>>> +       sync_core();
>>>> +}
>>> 
>>> Make this if (!in_interrupt()) sync_core(); please.  We can optimize
>>> it better later on.
>> 
>> Sure, done. It will be part of the next version of that patch.
> 
> Of course, using in_interrupt() from linux/interrupt.h from
> asm/processor.h is not such a good idea (circular dependency).
> 
> Any recommendation on where to move that static inline on x86 ?

Actually, for the needs of the fix, I plan on not putting the
if (!in_interrupt()) check, given that the only caller will
be the scheduler, not in an interrupt context.

We can always revisit this optimization-wise if we end up
using in interrupt handlers in the future.

Is that ok with you ?

Thanks,

Mathieu


> 
> Thanks,
> 
> Mathieu
> 
> 
>> 
>> Thanks!
>> 
>> Mathieu
>> 
>> 
>> --
>> Mathieu Desnoyers
>> EfficiOS Inc.
>> http://www.efficios.com
> 
> --
> Mathieu Desnoyers
> EfficiOS Inc.
> http://www.efficios.com

-- 
Mathieu Desnoyers
EfficiOS Inc.
http://www.efficios.com