> I think this part of your patch is a bit overkill, to be honest, and am > inclined to leave it out. We don't normally do this kind of > double-checking of configure results at build time. The autopkgtest > makes sense though, and seems to be good enough to catch regressions in > this kind of thing in future. > > Would you mind if I revised your patch to omit this part? (No need to > resend; I can make the adjustments myself.)
Thanks Colin. No problem. I wrote this test _before_ the more complicated, (but arguably more valuable autopkgtest). Feel free to amend as you see appropriate. I also pushed the patch to a Gitlab fork to confirm the tests worked, (outside the container I did my local testing in). See https://salsa.debian.org/alexjfisher/openssh/-/jobs/9976488#L3041 for the autopkgtest job log.
