Re: [musl] Re: Tweaking the program name for <err.h> functions

[enh <enh@google.com>]

Android's libc actually does do this for everything except for
first-stage `init`, the one process that doesn't have a /dev/null
equivalent available yet:
https://android.googlesource.com/platform/bionic/+/master/libc/bionic/libc_init_common.cpp#358

On Mon, Mar 11, 2024 at 11:49 AM Skyler Ferrante (RIT Student)
<sjf5462@rit.edu> wrote:
>
> Hi Florian,
>
> > it's not running SUID (in AT_SECURE mode)
>
> I see. I didn't realize that it had different behavior for setuid/not
> setuid. That makes sense though, sorry for the confusion.
>
> Skyler
>
>
> On Mon, Mar 11, 2024 at 2:23 PM Florian Weimer <fweimer@redhat.com> wrote:
> >
> > * Skyler Ferrante:
> >
> > > Hmm, maybe I'm missing something, but it seems you can close(fd) for
> > > the standard fds and then call execve, and the new process image will
> > > have no fd 0,1,2. I've tried this on a default Ubuntu 22.04 system.
> > > This seems to affect shadow-utils and other setuid/setgid binaries.
> > >
> > > Here is a repo I built for testing,
> > > https://github.com/skyler-ferrante/fd_omission/. What is the correct
> > > glibc behavior? Am I misunderstanding something?
> >
> > If you run it under strace, it's not running SUID (in AT_SECURE mode).
> > I'm not saying we don't have bugs (although we do have some end-to-end
> > AT_SECURE tests in the testsuite, but probably not for this legacy
> > behavior), just that this approach to testing is questionable.
> >
> > Thanks,
> > Florian
> >