From mboxrd@z Thu Jan 1 00:00:00 1970 From: Maximilian Heyne Date: Fri, 6 May 2022 06:44:40 +0000 Subject: [Intel-wired-lan] [PATCH net-next v2] drivers, ixgbe: show VF statistics In-Reply-To: <20220504201632.2a41a3b9@kernel.org> Message-ID: <20220506064440.57940-1-mheyne@amazon.de> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: intel-wired-lan@osuosl.org List-ID: On 2022-05-04T20:16:32-07:00 Jakub Kicinski wrote: > On Tue, 3 May 2022 15:00:17 +0000 Maximilian Heyne wrote: > > + for (i = 0; i < adapter->num_vfs; i++) { > > + ethtool_sprintf(&p, "VF %u Rx Packets", i); > > + ethtool_sprintf(&p, "VF %u Rx Bytes", i); > > + ethtool_sprintf(&p, "VF %u Tx Packets", i); > > + ethtool_sprintf(&p, "VF %u Tx Bytes", i); > > + ethtool_sprintf(&p, "VF %u MC Packets", i); > > + } > > Please drop the ethtool stats. We've been trying to avoid duplicating > the same stats in ethtool and iproute2 for a while now. > I can see the point that standard metrics should only be reported via the iproute2 interface. However, in this special case this patch was intended to converge the out-of-tree driver with the in-tree version. These missing stats were breaking our userspace. If we now switch solely to iproute2 based statistics both driver versions would diverge even more. So depending on where a user gets the ixgbe driver from, they have to work-around. I can change the patch as requested, but it will contradict the inital intention. At least Intel should then port this change to the external driver as well. Let's get a statement from them. Amazon Development Center Germany GmbH Krausenstr. 38 10117 Berlin Geschaeftsfuehrung: Christian Schlaeger, Jonathan Weiss Eingetragen am Amtsgericht Charlottenburg unter HRB 149173 B Sitz: Berlin Ust-ID: DE 289 237 879