> >>
> >> ---------------------------------------------------------------------
> >> -----------
> > After the call
> >
> > Ira
> >
>
> Ira,
> On the call we discussed that today there is no 'incarnation' value exposed for
> the uverbs%.
> You suggested an optimization by using the cdev's file creation timestamp as
> option for libibverbs to understand that the driver did or did-not restart since
> mapping of that ibv_device.
> Main pitfalls with this is if system time is being modified (admin or ntp/ptp). In
> the rare case that we have a hoplug event and time is rewinded we can
> accidentally reach a very hard debugging exercise.
> What do you think?
It was actually Jason who suggested that.
Anyway I think the scenario you explain is probably a very rare case. So I would not worry about it.
Going forward we could put an instance number in sysfs if we really are concerned about making this 100% robust.
But I'm not sure that is required.
Ira
��.n��������+%������w��{.n�����{���fk��ܨ}���Ơz�j:+v�����w����ޙ��&�)ߡ�a����z�ޗ���ݢj��w�f
