Hello Eugene, Ping on the below... Thanks, Michael On 10/14/18 4:25 PM, Michael Kerrisk (man-pages) wrote: > Hi Eugene, > > On 09/19/2018 07:12 PM, Eugene Syromyatnikov wrote: >> Specifically, on s390/s390x. >> >> Signed-off-by: Eugene Syromyatnikov <evgsyr@xxxxxxxxx> >> --- >> man2/execve.2 | 13 +++++++++++++ >> 1 file changed, 13 insertions(+) >> >> diff --git a/man2/execve.2 b/man2/execve.2 >> index 788e62d..3103c0e 100644 >> --- a/man2/execve.2 >> +++ b/man2/execve.2 >> @@ -607,6 +607,19 @@ but the new image could not be completely built. >> In such cases, the kernel kills the process with a >> .BR SIGKILL >> signal. >> +.\" https://lists.strace.io/pipermail/strace-devel/2017-August/007236.html >> +.PP >> +Technically, a successful call to >> +.BR execve () >> +does return, in the new program image, with return value available >> +for inspection (in an architecture-specific register as described in >> +.BR syscall (2)). >> +It makes little sense for the program itself to check it, but tracers >> +(such as >> +.BR strace (1)) >> +may show it; that said, some architectures (such as s390/s390x) do not bother >> +setting proper return value during exec, and that's can be exposed, >> +if not handled by the tracer specifically. > > The last two lines above are a bit garbled. I can't make out what you mean > by "and that's can be exposed, if not handled by the tracer specifically". > Could you say something more about this please? > > Thanks, > > Michael > >> .\" >> .SS Interpreter scripts >> A maximum line length of 127 characters is allowed for the first line in >> > > -- Michael Kerrisk Linux man-pages maintainer; http://www.kernel.org/doc/man-pages/ Linux/UNIX System Programming Training: http://man7.org/training/
