summaryrefslogtreecommitdiff
path: root/open_issues/increasing_bogus_port_at_boot.mdwn
blob: 483b2844948dcbfebcfab2ce77fb1f41c9a0cf0d (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
[[!meta copyright="Copyright © 2016 Free Software
Foundation, Inc."]]

[[!meta license="""[[!toggle id="license" text="GFDL 1.2+"]][[!toggleable
id="license" text="Permission is granted to copy, distribute and/or modify this
document under the terms of the GNU Free Documentation License, Version 1.2 or
any later version published by the Free Software Foundation; with no Invariant
Sections, no Front-Cover Texts, and no Back-Cover Texts.  A copy of the license
is included in the section entitled [[GNU Free Documentation
License|/fdl]]."]]"""]]

[[!tag open_issue_porting]]

When the ntpdate package is installed, one gets at boot something like:

    Debian GNU/Hurd stretch/sid debian console
    task ext2fs increasing a bogus port 947 by 1, most probably a bug.
    task ext2fs increasing a bogus port 947 by 1, most probably a bug.
    task ext2fs deallocating a bogus port 947, most probably a bug.
    task ext2fs deallocating a bogus port 947, most probably a bug.
    login:

This is coming from the execution of the shell script
/etc/network/if-up.d/ntpdate, whose stdout/stderr is on the Mach console, but
part of which gets executed after getty starts on it. It happens that getty uses
revoke() to revoke access to it from other programs, and thus the ntpdate shell
scripts gets its stdout/stderr in a bogus state, which libc doesn't really cope
with correctly.

Commenting `c:23:respawn:/sbin/getty 38400 console` from `/etc/inittab` works
around the issue (but removes the getty from the Mach console)