Re: Rough notes from sys_membarrier() lightning BoF

From: Mathieu Desnoyers
Date: Thu Sep 21 2017 - 14:02:24 EST


----- On Sep 21, 2017, at 9:15 AM, Peter Zijlstra peterz@xxxxxxxxxxxxx wrote:

> On Wed, Sep 20, 2017 at 06:13:50PM +0000, Mathieu Desnoyers wrote:
>> My proposed RFC for private expedited membarrier enforces that all
>> architectures perform the registration step. Using the "PRIVATE_EXPEDITED"
>> command without prior process registration returns an error on all
>> architectures. The goal here is to make all architectures behave in the
>> same way, and it allows us to rely on process registration to deal
>> with future arch-specific optimizations.
>>
>> Adding the "core_sync" behavior could then be done for the next kernel
>> merge window. I'm currently foreseeing two possible ABI approaches to
>> expose it:
>>
>> Approach 1:
>>
>> Add MEMBARRIER_CMD_PRIVATE_EXPEDITED_SYNC_CORE and
>> MEMBARRIER_CMD_REGISTER_PRIVATE_EXPEDITED_SYNC_CORE commands. This
>> allows us to return their availability through MEMBARRIER_CMD_QUERY.
>>
>> Approach 2:
>>
>> Add a "MEMBARRIER_FLAG_SYNC_CORE" as flag parameter. It could be set
>> when issuing both MEMBARRIER_CMD_REGISTER_PRIVATE_EXPEDITED and
>> MEMBARRIER_CMD_PRIVATE_EXPEDITED, thus ensuring core serializing
>> behavior. Querying whether core serialization is supported could
>> be done by issuing the MEMBARRIER_CMD_QUERY command with the
>> MEMBARRIER_FLAG_SYNC_CORE flag set.
>>
>> Any other ideas ? Any approach seems better ?
>
> So we really need another FLAG for that? AFAICT the current
> PRIVATE_EXPEDITED is already sufficient for the cross modifying code,
> since the IPI triggers an exception return on all currently running CPUs
> and the future running CPUs will have the return to userspace doing the
> exception return.
>
> The only issue is Andy fudging our x86 ret-to-userspace to not use IRET,
> which we can fix by forcing it into the slowpath (that needs to exist
> anyway) using that new TIF flag.

I agree that x86, as it stands today, would provide core serialization
with the private expedited membarrier command. And we can deal with
future optimization of ret-to-userspace using the TIF flag set on
registration.

I'm wondering whether all architectures guarantee core serialization
on return from interrupt triggered by the IPI, and on ret-to-userspace ?

Thanks,

Mathieu

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