Re: [PATCH 0/7] devcg: device cgroup extension for rdma resource

[Haggai Eran <haggaie-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>]

On 07/09/2015 23:38, Parav Pandit wrote:
> Currently user space applications can easily take away all the rdma
> device specific resources such as AH, CQ, QP, MR etc. Due to which other
> applications in other cgroup or kernel space ULPs may not even get chance
> to allocate any rdma resources.
> 
> This patch-set allows limiting rdma resources to set of processes.
> It extend device cgroup controller for limiting rdma device limits.
I don't think extending the device cgroup is the right place for these
limits. It is currently a very generic controller and adding various
RDMA resources to it looks out of place. Why not create a new controller
for rdma?

Another thing I noticed is that all limits in this cgroup are global,
while the resources they control are hardware device specific.
I think it would be better if the cgroup controlled the limits of each
device separately.

> With this patch, user verbs module queries rdma device cgroup controller
> to query process's limit to consume such resource. It uncharge resource 
> counter after resource is being freed.
This is another reason why per-device limits would be better. Since
limits are reflected to user-space when querying a specific device, it
will show the same maximum limit on every device opened. If the user
opens 3 devices they might expect to be able to open 3 times the number
of the resources they actually can.

Regards,
Haggai
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html