Jesper Dangaard Brouer <jbrouer@xxxxxxxxxx> writes: > On 22/08/2023 18.56, Toke Høiland-Jørgensen wrote: >> Jesper Dangaard Brouer <hawk@xxxxxxxxxx> writes: >> >>> On 22/08/2023 16.22, Toke Høiland-Jørgensen wrote: >>>> The functionality of these utilities have been incorporated into the >>>> xdp-bench utility in xdp-tools. Remove the unmaintained versions in >>>> samples. >>>> >>> >>> I think it will be worth our time if we give some examples of how the >>> removed utility translates to some given xdp-bench commands. There is >>> not a 1-1 mapping. >>> >>> XDP driver changes need to be verified on physical NIC hardware, so >>> these utilities are still being run by QA. I know Red Hat, Intel and >>> Linaro QA people are using these utilities. It will save us time if we >>> can reference a commit message instead of repeatable describing this. >>> E.g. for Intel is it often contingent workers that adds a tested-by >>> (that all need to update their knowledge). >> >> I did think about putting that in the commit message for these, but I >> figured it was too obscure a place to put it, compared to (for instance) >> putting it into the xdp-bench man page. >> >> If you prefer to have it in the commit message as well, I can respin >> adding it - WDYT? >> > > It is super nice that xdp-bench already have a man page, but I was > actually looking at this and it was a bit overwhelming (520 lines) > explaining every possible option. Haha, I think that's the first time I've had anyone complain that I write too *much* documentation ;) > I really think its worth giving examples in the commit, to ease the > transition to this new tool. OK, I'll respin tomorrow with some examples in the commit messages... -Toke