> On Mar 11, 2021, at 9:41 AM, Tom Ellis > <tom-lists-haskell-cafe-2...@jaguarpaw.co.uk> wrote: > > I'm not really following the details, but is this useful to you? > > % cat g.c && cc g.c -o g && ./g > #include <sys/types.h> > #include <grp.h> > #include <errno.h> > #include <stdio.h> > > int main(int argc, char **argv) > { > char buf[1024]; > struct group g, *p; > int rc; > > errno = 0; > rc = getgrnam_r(argc > 1 ? argv[1] : "nosuchgrouphere", > &g, buf, sizeof(buf), &p); > printf("%s(%p) %m(%d)\n", p ? g.gr_name : NULL, p, errno); > return (rc == 0 && p == NULL); > } > (null)((nil)) No such process(3)
Yes, it means that the reported error is not an artefact of the Haskell "unix" package, but rather originates directly from normal use of the getpwnam_r(3) glibc API on these systems. It would now be useful to also post: - The output of "./g root" or some other group known to exist. - The output of "./g xyzzy" or some other short group name known to not exist - The output of "grep group /etc/nsswitch.conf" - Attach an strace output file (g.trace.txt) from: strace -o g.trace.txt ./g -- Viktor. _______________________________________________ ghc-devs mailing list ghc-devs@haskell.org http://mail.haskell.org/cgi-bin/mailman/listinfo/ghc-devs