XPost: linux.debian.bugs.dist   
   From: carnil@debian.org   
      
   Control: tags -1 + moreinfo   
      
   Hi,   
      
   On Sun, Oct 12, 2025 at 11:56:09PM +0200, Garri Djavadyan wrote:   
   > Package: linux-image-amd64   
   > Version: 6.12.48-1   
   >   
   > Dear Debian Linux Kernel Maintainers,   
   >   
   > I noticed that the ipv6_route flags RTF_ADDRCONF and RTF_PREFIX_RT are   
   > not cleared when static on-link routes are added during IPv6 address   
   > configuration, and it leads to situations when the kernel updates the   
   > static on-link routes with expiration time.   
   >   
   > To replicate the problem I used the latest Debian 13.1 distribution   
   > with the kernel 6.12.48-1 and radvd (on the second machine) with the   
   > following configuration:   
   >   
   > root@localhost:~# cat /etc/debian_version   
   > 13.1   
   >   
   > root@localhost:~# uname -a   
   > Linux localhost 6.12.48+deb13-amd64 #1 SMP PREEMPT_DYNAMIC Debian   
   > 6.12.48-1 (2025-09-20) x86_64 GNU/Linux   
   >   
   > radvd.conf (on a directly connected machine):   
   > interface veth1   
   > {   
   >         AdvSendAdvert on;   
   >         MinRtrAdvInterval 45;   
   >         MaxRtrAdvInterval 60;   
   >         AdvDefaultLifetime 0;   
   >         AdvDefaultPreference low;   
   >         AdvHomeAgentFlag off;   
   >         prefix fd00::/64   
   >         {   
   >                 AdvOnLink on;   
   >                 AdvAutonomous on;   
   >                 AdvRouterAddr off;   
   >                 AdvPreferredLifetime 60;   
   >                 AdvValidLifetime 120;   
   >         };   
   > };   
   >   
   >   
   > When I first add a manual IPv6 address to the interface receiving   
   > ICMPv6 RA packets and then receive an ICMPv6 RA with the same on-link   
   > prefix, the packet is silently ignored and no ipv6_route flags,   
   > including RTF_EXPIRES, get set on the static route. Everything works as   
   > expected.   
   >   
   > However, if an ICMPv6 RA is received before a manual IPv6 address is   
   > set on the interface, and the RA route is installed in the IPv6 route   
   > table, along with the ipv6_route flags RTF_ADDRCONF, RTF_PREFIX_RT, and   
   > RTF_EXPIRES, only the flag RTF_EXPIRES gets cleared when a manual IPv6   
   > address is configured on the interface later. As a result, after   
   > configuring the IPv6 address, it does not have any associated   
   > expiration time, but the kernel still treats it as an RA-learned route,   
   > so the next received RA packet sets the expiration time again.   
   >   
   > Below are the steps leading to the described issue:   
   >   
   >   
   > # RAs are accepted by the inteface ens4   
   > root@localhost:~# cat /proc/sys/net/ipv6/conf/ens4/accept_ra   
   > 1   
   >   
   >   
   > # Nothing is assigned to ens4 yet   
   > root@localhost:~# ip -6 addr show dev ens4   
   > root@localhost:~# ip -6 ro show dev ens4   
   > root@localhost:~#   
   >   
   >   
   > # No fd00::/64 routes are present in the IPv6 route table   
   > root@localhost:~# egrep '^fd0+\ ' /proc/net/ipv6_route   
   > root@localhost:~#   
   >   
   >   
   > # Received on-link prefix fd00::/64 from the router   
   > root@localhost:~# tcpdump -vi ens4 'icmp6[0] = 134'   
   > tcpdump: listening on ens4, link-type EN10MB (Ethernet), snapshot   
   > length 262144 bytes   
   >   
   > 20:41:57.733151 IP6 (flowlabel 0x15aa0, hlim 255, next-header ICMPv6   
   > (58) payload length: 56) fe80::f83a:f0ff:fe69:27d2 > ip6-allnodes:   
   > [icmp6 sum ok] ICMP6, router advertisement, length 56   
   >         hop limit 64, Flags [none], pref low, router lifetime 0s,   
   > reachable time 0ms, retrans timer 0ms   
   >           prefix info option (3), length 32 (4): fd00::/64, Flags   
   > [onlink, auto], valid time 120s, pref. time 60s   
   >           source link-address option (1), length 8 (1):   
   > fa:3a:f0:69:27:d2   
   >   
   >   
   > # The RA route is installed in the ipv6_route structure with the flags   
   > 0x004c0001   
   > # 0x4c: RTF_ADDRCONF, RTF_PREFIX_RT, and RTF_EXPIRES   
   > root@localhost:~# egrep '^fd0+\ ' /proc/net/ipv6_route   
   > fd000000000000000000000000000000 40 00000000000000000000000000000000 00   
   > 00000000000000000000000000000000 00000100 00000002 00000000 004c0001   
   > ens4   
   >   
   >   
   > # The route has an expiration time assigned as expected   
   > root@localhost:~# ip -6 ro show dev ens4   
   > fd00::/64 proto kernel metric 256 expires 88sec pref medium   
   >   
   >   
   > # Now, the manual IPv6 address from the subnet fd00::/64 is configured   
   > on the interface   
   > root@localhost:~# ip addr add fd00::2/64 dev ens4   
   > root@localhost:~# ip -6 addr show dev ens4   
   > 3: ens4:  mtu 1500 qdisc fq_codel   
   > state UP group default qlen 1000   
   >     altname enp0s4   
   >     altname enx525400123457   
   >     inet6 fd00::2/64 scope global   
   >        valid_lft forever preferred_lft forever   
   >     inet6 fd00::5054:ff:fe12:3457/64 scope global dynamic mngtmpaddr   
   > proto kernel_ra   
   >        valid_lft 88sec preferred_lft 28sec   
   >   
   >   
   > # The ipv6_route entry for fd00::/64 gets updated: the flags are   
   > changed to 0x000c0001.   
   > # 0x0c: RTF_ADDRCONF, RTF_PREFIX_RT   
   > # The flag RTF_EXPIRES is removed but the RA-specific flags   
   > RTF_ADDRCONF, RTF_PREFIX_RT are still set.   
   > root@localhost:~# egrep '^fd0+\ ' /proc/net/ipv6_route   
   > fd000000000000000000000000000000 40 00000000000000000000000000000000 00   
   > 00000000000000000000000000000000 00000100 00000001 00000000 000c0001   
   > ens4   
   >   
   >   
   > # From user's perspective the on-link route looks permanent, no   
   > expiration time is present   
   > root@localhost:~# ip -6 ro show dev ens4   
   > fd00::/64 proto kernel metric 256 pref medium   
   >   
   >   
   > # Next RA packet has come at this point   
   >   
   >   
   > # And the permanent route turned into a temporary one again   
   > (0x004c0001)   
   > root@localhost:~# egrep '^fd0+\ ' /proc/net/ipv6_route   
   > fd000000000000000000000000000000 40 00000000000000000000000000000000 00   
   > 00000000000000000000000000000000 00000100 00000002 00000000 004c0001   
   > ens4   
   >   
   > root@localhost:~# ip -6 ro show dev ens4   
   > fd00::/64 proto kernel metric 256 expires 77sec pref medium   
   >   
   >   
   > # The kernel is instructed not to accept RAs anymore   
   > root@localhost:~# echo 0 > /proc/sys/net/ipv6/conf/ens4/accept_ra   
   > root@localhost:~# cat /proc/sys/net/ipv6/conf/ens4/accept_ra   
   > 0   
   >   
   >   
   > # After 2 minutes, the on-link route gets vanished   
   > # while the manual IPv6 address is still installed on the interface   
   > root@localhost:~# ping -Oi 10 fd00::1   
      
   [continued in next message]   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)