Re: [lttng-dev] Disabling sys_membarrier in lttng urcu

[Mathieu Desnoyers via lttng-dev <lttng-dev@lists.lttng.org>]

----- On Jun 17, 2021, at 6:13 PM, lttng-dev lttng-dev@lists.lttng.org wrote:

> Hello,
> 
> Some old topic, see
> https://lists.lttng.org/pipermail/lttng-dev/2019-November/029411.html.
> (I actually thought this was solved meanwhile).
> 
> With lttng 2.13 liburcu is replicated in lttng-ust so my old custom
> hack aint helping.
> 
> Aside from another crude hack, I thought about doing this:
> 
> extern int lttng_ust_urcu_has_sys_membarrier;
> int setup() {
>  lttng_ust_urcu_has_sys_membarrier = 0;
> }
> 
> this is obvious possible in my own program, but I don't know if some
> lttng daemon would need to update RCU structures in shared memory that
> should sync to other processes (and wont do that with sys_membarrier
> in case of Xenomai threads)?
> 
> Seems safer to me to hack it out once more...

We currently don't have any RCU synchronization across shared memory,
and the implementation we have in lttng-ust only tracks threads within
the same process, so it would be safe to tweak the lttng-ust-urcu behavior
to use membarrier or not based on an environment variable.

So I think we could expose a new "--disable-sys-membarrier" configure option
to lttng-ust, and maybe add a LTTNG_UST_MEMBARRIER environment variable.
This would likely be 2.14 material though.

I also notice that lttng-ust 2.13's implementation of RCU still refers to
CONFIG_RCU_FORCE_SYS_MEMBARRIER which belongs to liburcu. This is
probably something we'll want to fix right away before we release 2.13
final.

Thanks,

Mathieu

-- 
Mathieu Desnoyers
EfficiOS Inc.
http://www.efficios.com
_______________________________________________
lttng-dev mailing list
lttng-dev@lists.lttng.org
https://lists.lttng.org/cgi-bin/mailman/listinfo/lttng-dev