Messages in this thread | | | Date | Wed, 16 Mar 2022 11:55:15 -0700 | From | Jakub Kicinski <> | Subject | Re: [PATCH net-next 1/3] net: gre_demux: add skb drop reasons to gre_rcv() |
| |
On Wed, 16 Mar 2022 10:12:00 +0000 David Laight wrote: > Is it worth considering splitting the 'reason' into two parts? > Eg x << 16 | y > One part being the overall reason - and probably a define. > The other qualifying the actual failing test and probably just > being a number. > > Then you get an overall view of the fails (which might even > be counted) while still being able to locate the actual > failing test.
That popped to my mind, but other than the fact that it "seems fine" I can't really convince myself that (a) 2 levels are enough, why not 3; (b) I personally don't often look at the drops, so IDK what'd fit the needs of the consumer of this API. TCP bad csum drops are the only ones I have experience with, and we have those already covered.
| |