Yeah, the wrapper_cmp version looks rather good, I didn’t realize it would go down to one instruction. Or that an old patch would have been stalled for five months ;-) OG Le lun. 9 août 2021 à 01:05, Rich Felker a écrit : > On Sun, Aug 08, 2021 at 10:21:06PM +0100, Jon Chesterfield wrote: > > On Sun, Aug 8, 2021 at 9:39 PM Érico Nogueira > wrote: > > > > > This is the favored approach, from what I understood of the discussions > > > surrounding it. It's implemented with musl's namespacing rules and such > > > in [1]. > > > > > > > Unsurprisingly I like your patch. > > > > > > > > It is badly optimized for some archs, unfortunately, as explained in > the > > > thread from [2]. I believe that's what's holding it up. > > > > > > [1] > > > > https://inbox.vuxu.org/musl/20210309210213.29539-1-ericonr@disroot.org/ > > > [2] > > > > https://inbox.vuxu.org/musl/20210309150320.GU32655@brightrain.aerifal.cx/ > > > > > > And that's what I get for not reading the list carefully enough. I missed > > the thread > > from March entirely. > > No problem. Indeed, this was stalled because of concerns about bad > compiler behavior on some archs, but I really don't want to allow that > to dictate how we solve the problem. The compiler is doing the wrong > thing (failing to tail call) and should just be fixed. I'll probably > merge the tail-call patch as-is, or with any minimal fixes needed, > right after this release so as not to make major API changes at last > minute. > > Rich >