On 11/3/19 11:31 PM, enh wrote: > i hit this bug in practice, but it did remind me of the reported id(1) > bug https://github.com/landley/toybox/issues/117. re-reading that,
I need to go through those again, but I have _years_ of todo items queued up in case I find time for them... Meanwhile, I'm off doing http://lists.j-core.org/pipermail/j-core/2019-October/000867.html and so on through at least the 21st... > it's actually kind of the opposite bug --- "what does id(1) do if it > has to _output_ a group that it only has a number for?". i did look > into fixing that bug too while i was in the area (after knocking up a > quick hack to setgroups(2) a bogus group and exec toybox id), but That's what mkroot is intended for, I need to get that actually running the tests... > found that i couldn't get glibc to reliably indicate getgrgid(3) > failure. none of the claims in the man page appear to be true, at The FSF! > least for glibc 2.28. it seems like i get back an irrelevant group > entry and errno left at 0. so i'm just ignoring that for now... If it works in musl and bionic, glibc can go hang. Ship static binaries linked against bionic for the AOSP prebuilts and call it good. Maybe they'll fix it someday, but there isn't enough popcorn in the world to make https://lwn.net/Articles/801482/ something I'd want to get near. >> Rob Rob _______________________________________________ Toybox mailing list [email protected] http://lists.landley.net/listinfo.cgi/toybox-landley.net
