Wed, Feb 26, 2025 at 04:06:19PM +0100, przemyslaw.kits...@intel.com wrote: >On 2/26/25 15:48, Jiri Pirko wrote: >> Tue, Feb 25, 2025 at 04:40:49PM +0100, przemyslaw.kits...@intel.com wrote: >> > On 2/25/25 15:35, Jiri Pirko wrote: >> > > Tue, Feb 25, 2025 at 12:30:49PM +0100, przemyslaw.kits...@intel.com >> > > wrote: >> >> [...] >> >> > > > output, for all PFs and VFs on given device: >> > > > >> > > > pci/0000:af:00: >> > > > name rss size 8 unit entry size_min 0 size_max 24 size_gran 1 >> > > > resources: >> > > > name lut_512 size 0 unit entry size_min 0 size_max 16 size_gran >> > > > 1 >> > > > name lut_2048 size 8 unit entry size_min 0 size_max 8 size_gran >> > > > 1 >> > > > >> > > > What is contributing to the hardness, this is not just one for all ice >> > > > PFs, but one per device, which we distinguish via pci BDF. >> > > >> > > How? >> > >> > code is in ice_adapter_index() >> >> If you pass 2 pfs of the same device to a VM with random BDF, you get 2 >> ice_adapters, correct? > >Right now, yes
That is a bug. > >> >> [...] > >What I want is to keep two ice_adapters for two actual devices (SDNs)
