Lennart Poettering
Lennart Poettering
So I think there's a misunderstanding here: resolved stays with the same DNS server as long as it responds, and it will ask on all links in parallel if there's...
lgtm, some minor things.
This sounds like premature optimization. I have serious doubts that this is really a good thing to do. Can you provide us with any kind of profiling/performance data that shows...
> I might be missing some historical context, but bypass is also enabled for DO packets: yeah, if clients want DNSSEC RRs we better step out of the way and...
any reason you marked this as draft btw? (i usually don't look at draft stuff with the same priority as on other stuff)
so, hmm, taking a step back on this i *do* wonder what the right approach here is. the idea of the bypass stuff originally was that we bypass all our...
> so, hmm, taking a step back on this i _do_ wonder what the right approach here is. the idea of the bypass stuff originally was that we bypass all...
looks generally good to me, but i really hate macros that make use of local variables of the functions they are called in, in basically a "hidden" way. can you...
Sorry, but i am strictly against this. subuids/subgids are a terrible concept. They should go away, not be hooked into more and more stuff.
> > Sorry, but i am strictly against this. subuids/subgids are a terrible concept. They should go away, not be hooked into more and more stuff. > > The reality...