Bug#1136459: openssh-server: [regression] stopped accepting connections over the WLAN

Dixi quod=E2=80=A6

>/usr/sbin/sshd -ddddde -o IPQoS=3D'lowdelay throughput'
>also works (isn=E2=80=99t that the pre-1:10.0p1-7+deb13u3 setting?).

ssh=E2=80=99ing out also fails unless -o IPQoS=3D'lowdelay throughput'
so it also affects the client.

I then ran a test which TC octets pass through, by cloning
https://github.com/mirabilos/ECN-Bits (an earlier project),
'cd c && make', then:

$ ./server/server 11111

and:

$ i=3D-1; while (( ++i < 256 )); do ./client/client 192.168.1.5 11111 $i; d=
one

The result is that the Fritzboxen do not generally drop packets
with any TC octet (my first 0x49 didn=E2=80=99t pass, but it=E2=80=99s UDP,=
 and
on retry, it worked).

However, if I start the server as=E2=80=A6

$ ./server/server +0xB8 11111

(meaning it will (try to) respond to each incoming packet with
TC 0xB8=E2=80=A50xBB), then no replies are registred at the client. I=E2=80=
=99m
not sure whether this is something done by Linux itself or a
device after it.

Reversing the construct from above, i.e. the trixie system sending=E2=80=A6
0x00=E2=80=A50x9F, 0xC0=E2=80=A50xFF pass, i.e. 0xA0=E2=80=A50xBF are dropp=
ed, which is
bit pattern 001xxx.xx (ToS.ECN), which corresponds to precedence 1
in the old model, or in DSCP, decimal 8=E2=80=A515, which Wikipedia=E2=80=
=99s list
shows as CS1, ?, AF11, ?, AF12, ?, AF13, ?.

Unfortunately, I=E2=80=99m unable to get either of the laptops to switch
APs (sudo iwconfig wlan0 ap =E2=80=A6), but I did get an iwlwifi crash on
attempting that=E2=80=A6 so I cannot test changing Fritzboxen atm.

Gru=C3=9F
//Thorsten
--=20
Thorsten Glaser
Linux / Unix Developer
Tel.: +49 160 91168501
E-Mail: [email protected]

B1 Systems GmbH
Osterfeldstra=C3=9Fe 7 / 85088 Vohburg / https://www.b1-systems.de/
GF: Ralph Dehner / Unternehmenssitz: Vohburg / AG: Ingolstadt, HRB 3537