On Thu, Nov 10, 2016 at 01:11:18PM +0530, Parav Pandit wrote: > Hi Leon, Tejun, Christoph, Liran, Doug, Matan, > > So are you ok with below proposal? I'm fine with it and it looks like very clean approach to solve our multi-object future. > > 1. Define two resources by rdmacg. > (a) hca_handles (covers doorbell pages) > (b) hca_resources (mr, pd, qp, srq, vendor defined, all consolidated count) > Both cannot be combined as explained in [1]. > > 2. User configures absolute count for above two resources (similar to > today's file descriptors, pid cgroup controller max limit) > > Leon, > Let us know if you have any further discussions during LPC on > questions of [2] in using percentage based scheme or otherwise. No, we didn't have. > > Parav > > [1] https://www.spinics.net/lists/linux-rdma/msg42771.html > [2] https://www.spinics.net/lists/linux-rdma/msg42768.html > > > > On Tue, Nov 8, 2016 at 1:42 PM, Liran Liss <liranl@xxxxxxxxxxxx> wrote: > >> From: Parav Pandit [mailto:pandit.parav@xxxxxxxxx] > > > >> > > >> > Hmm.. > >> > I guess that you are right. > >> > > >> > So we can add another count for "HCA handles", > >> I prefer this. This keeps it vendor agnostic and clean if we don't go percentage > >> route. > > > > OK; let's do it. > > > >> Would indirection table also fall in this category? > >> > > > > No. It's just another HCA resource... > > > >> > or alternatively, each provider will restrict the number of handles > >> > per device to a reasonable small number (which won't be treated as one of the > >> "HCA resources"). > >> This would require vendor drivers to get the understanding of cgroup object > >> and pid and that breaks the modular approach. I like to avoid this. > >> > >> > Typically, a process shouldn't need to open more than a single handle... > >> Right. well behaved application won't do multiple handles.
Attachment:
signature.asc
Description: PGP signature
