[PATCH v3 0/7] Rework device probing
David Marchand
david.marchand at redhat.com
Sat Apr 11 12:23:41 CEST 2026
On Sat, 11 Apr 2026 at 12:10, David Marchand <david.marchand at redhat.com> wrote:
>
> On Fri, 10 Apr 2026 at 16:35, Morten Brørup <mb at smartsharesystems.com> wrote:
> >
> > Is there a workaround until this patch is added, if I don't want rte_eal_init() to probe any devices, but I want my application to probe devices later by calling rte_bus_scan() followed by rte_bus_probe()?
> >
> > It's not something we really need, so if you don't know it off the top of your head, don't spend time trying to figure it out.
Some (untested) other idea... write some "plugin" loaded last (via
-d), that would list all buses, then unregister all of them (or a part
of the list, for buses known to be a problem..).
IOVA discovery would end up as DC, so probably fine (and we have an
option to force it if needed).
Then after call to rte_eal_init, the application would get a hold of
this list, re-register buses... ?
I think it could work, but clearly hacky...
--
David Marchand
More information about the dev
mailing list