The old socket code did something like
int fd = socket(…);
if (fd == -1) {
error handling
}
if (fd < 100) {
int newfd = fcntl(fd, F_DUPFD, 100);
if (newfd == -1) {
error handling
}
close(fd);
fd = newfd;
}
to leave file descriptor space for stdio to work in where the descriptors where
stored in a (char).
> On 14 Sep 2022, at 09:06, Casey Deccio <ca...@deccio.net> wrote:
>
>
>> On Sep 13, 2022, at 3:35 PM, Graham Clinch <g.cli...@lancaster.ac.uk> wrote:
>>
>> I suspect nrpe-ng is closing stdin before launching nslookup.
>>
>>
>> With mac homebrew's build of bind 9.18.6 and a bit of shell redirection to
>> close stdin, I get:
>>
>> ---
>> $ /opt/homebrew/bin/nslookup -version
>> nslookup 9.18.6
>>
>> $ /opt/homebrew/bin/nslookup example.net 0<&-
>> [... usual output snipped ...]
>>
>> Assertion failed: (fd > STDERR_FILENO), function uv__close, file core.c,
>> line 602.
>> Abort trap: 6
>>
>> $ echo $?
>> 134
>> ---
>>
>
> Nice find!!
>
>> This might count as a regression of sorts from the migration to libuv, as
>> older nslookup doesn't fail in the same way:
>
> Could be. At the very least it's a behavioral difference, which affects some
> packages that depend on it. I'll leave it to the BIND devs to determine
> if/how to handle it.
>
>> (dig & delv are also affected in the same way)
>>
>> The cmake group came across the same situation with libuv and open missing
>> standard fds against /dev/null to compensate:
>> https://gitlab.kitware.com/cmake/cmake/-/merge_requests/3282
>
> I've patched the code for nrpe_ng, and it works!
>
> $ diff -u /usr/lib/python3/dist-packages/nrpe_ng/commands.py{.old,}
> --- /usr/lib/python3/dist-packages/nrpe_ng/commands.py.old 2017-08-08
> 13:05:02.000000000 -0600
> +++ /usr/lib/python3/dist-packages/nrpe_ng/commands.py 2022-09-13
> 17:00:36.767239885 -0600
> @@ -85,6 +85,7 @@
>
> proc = tornado.process.Subprocess(
> run_args,
> + stdin=subprocess.DEVNULL,
> stdout=tornado.process.Subprocess.STREAM,
> close_fds=True,
> env=env)
>
> I'll request that something get pushed upstream. Many thanks for help
> tracking that down!
>
> Casey
> --
> Visit https://lists.isc.org/mailman/listinfo/bind-users to unsubscribe from
> this list
>
> ISC funds the development of this software with paid support subscriptions.
> Contact us at https://www.isc.org/contact/ for more information.
>
>
> bind-users mailing list
> bind-users@lists.isc.org
> https://lists.isc.org/mailman/listinfo/bind-users
--
Mark Andrews, ISC
1 Seymour St., Dundas Valley, NSW 2117, Australia
PHONE: +61 2 9871 4742 INTERNET: ma...@isc.org
--
Visit https://lists.isc.org/mailman/listinfo/bind-users to unsubscribe from
this list
ISC funds the development of this software with paid support subscriptions.
Contact us at https://www.isc.org/contact/ for more information.
bind-users mailing list
bind-users@lists.isc.org
https://lists.isc.org/mailman/listinfo/bind-users
