From: Sun Chao <[email protected]> ---
Jeff King <[email protected]> writes: > I'm undecided on this. I think reftables are still a while off, and even > once they are here, many people will still be using the older format. So > it makes sense to still apply fixes to the old code. Got it, thanks for explainning. > What I wonder, though, is whether always refreshing will cause a > noticeable performance impact (and that's why I was so slow in > responding -- I had hoped to try to come up with some numbers, but I > just hadn't gotten around to it). > > My gut says it's _probably_ not an issue, but it would be nice to have > some data to back it up. Sorry for responding after 4 days because I have been away on official business. Tody I have tryied some tools like trace logs, time, and strace, tring to figure out if there are some noticeable numbers. I tried different repositories with different ref numbers and blob numbers, I also can not recognize how much the refreshing impact the performance, perhaps I need to find a better computer for benchmark testing. --- Junio C Hamano <[email protected]> writes: > I am tempted to let correctness (and ease-of-reasoning about the > code) take precedence over potential and unknown performance issue, > at least for now. A single liner is rather simple to revert (or in > the worst case we could add "allow pack-refs to efficiently lose a > ref to a race" configuration option) anyway. Thanks a lot :) -- 2.17.2 (Apple Git-113)
