On 11/7/2023 5:04 PM, Patrick Robb wrote: > Sorry, now I understand what is causing confusion. When a retest is > requested, that retest is done on the original dpdk artifact we created > for the patchseries. We don't pull again, re-apply, and then rerun the > testing. I think that is the behavior you were expecting, Ferruh. >
Hi Patrick, Thanks for clarification, yes that was my expectation. > We have discussed whether this is an important option to support in > future iterations of the retest-request feature. One idea is to add > another option in the retest request schema which allows for the > requester to indicate they want the patch re-applied onto the newer > branch, and testing rerun with that new dpdk. It's good to see that this > is a use case which would see application. Will be aiming to add this > support in the future. > I am also not sure if this is the common usecase, lets record the need, we can decide what to support based on it. And as a generic approach, what about following syntax: Recheck-request,attribute=value: ... And for above case, attribute can be "pull=true" ...
