- dev - lists.frrouting.org

FRR 10.2 Bug report - RIP/RIPNG distribution list fails to filter routes
by Russell Yount 12 Jan '25

12 Jan '25

I attempted to upgrade some FreeBSD 14.2 systems from frr8 ( 8.5.6) to frr10 (10.2) both from the FreeBSD ports collection. I skipped frr9 due to a bug that did not initialize routes from the kernel on startup. The systems are VMs running under Proxmox virtual environment. The distribution-list functionality no longer works for both rip and ripng. The following configuration can reproduce the problem. I would be happy to test any proposed fixes for this. -- router rip network vtnet8 network vtnet9 distribute-list rip-dist-out out exit router ripng network vtnet8 network vtnet9 ipv6 distribute-list ripng-dist-out out exit access-list rip-dist-out seq 5 deny any ipv6 access-list ripng-dist-out seq 5 deny any -- Also I notice in frr10 the routes reflecting the interface addresses going to loopback interface display in "show ip route" and "show ipv6 route" route as a new local(L) route type. I assume this is intentional but thought I would mention it. -Russ Russell J. Yount <Russell.Yount(a)gmail.com>

1 0

New Defects reported by Coverity Scan for freerangerouting/frr
by scan-admin＠coverity.com 09 Jan '25

09 Jan '25

Hi, Please find the latest report on new defect(s) introduced to freerangerouting/frr found with Coverity Scan. 2 new defect(s) introduced to freerangerouting/frr found with Coverity Scan. New defect(s) Reported-by: Coverity Scan Showing 2 of 2 defect(s) ** CID 1613932: Null pointer dereferences (FORWARD_NULL) ________________________________________________________________________________________________________ *** CID 1613932: Null pointer dereferences (FORWARD_NULL) /bgpd/bgp_route.c: 15206 in peer_adj_routes() 15200 } 15201 15202 output_count += output_count_per_rd; 15203 filtered_count += filtered_count_per_rd; 15204 } 15205 } else { >>> CID 1613932: Null pointer dereferences (FORWARD_NULL) >>> Passing null pointer "table" to "show_adj_route", which dereferences it. 15206 show_adj_route(vty, peer, table, afi, safi, type, rmap_name, 15207 json, json_ar, show_flags, &header1, &header2, 15208 rd_str, match, &output_count, &filtered_count); 15209 15210 if (use_json) { 15211 if (type == bgp_show_adj_route_advertised || ** CID 1613931: API usage errors (PRINTF_ARGS) ________________________________________________________________________________________________________ *** CID 1613931: API usage errors (PRINTF_ARGS) /mgmtd/mgmt_fe_adapter.c: 1240 in fe_adapter_handle_session_req() 1234 void *__msg, size_t msg_len) 1235 { 1236 struct mgmt_msg_session_req *msg = __msg; 1237 struct mgmt_fe_session_ctx *session; 1238 uint64_t client_id; 1239 >>> CID 1613931: API usage errors (PRINTF_ARGS) >>> Length modifier "L" not applicable to conversion specifier in "%Lu". 1240 __dbg("Got session-req is create %u req-id %Lu for refer-id %Lu from '%s'", 1241 msg->refer_id == 0, msg->req_id, msg->refer_id, adapter->name); 1242 1243 if (msg->refer_id) { 1244 uint64_t session_id = msg->refer_id; 1245 ________________________________________________________________________________________________________ To view the defects in Coverity Scan visit, http://url2497.blackduck.com/ls/click?upn=u001.Ji18sHaXCxZb7Rfw8sC51j9Suwl8…

1 0

Stop testing and updates for CentOS 7 / RedHat 7
by Martin Winter 09 Jan '25

09 Jan '25

Hi there, I'm planning to stop releasing any new CentOS 7 / RedHat 7 packages and stop the testing for it. - No more CI tests on CentOS 7 - No new packages for CentOS 7 / RedHat 7 (not even minor updates or security fixes) CentOS 7 is EoL for some time. So far, we still used it for build Redhat 7 / CentOS 7 packages, but there is now the time where repositories get removed or moved. As such, I think the work to maintain this in the CI Infrastructure and Release Build is too much and I'm considering abandoning it. This does NOT impact RedHat 8 or RedHat 9. Please reply if this causes any issues to you or have any other reasons why this should be reconsidered. Regards, Martin Winter

1 0

New Defects reported by Coverity Scan for freerangerouting/frr
by scan-admin＠coverity.com 01 Jan '25

01 Jan '25

Hi, Please find the latest report on new defect(s) introduced to freerangerouting/frr found with Coverity Scan. 1 new defect(s) introduced to freerangerouting/frr found with Coverity Scan. New defect(s) Reported-by: Coverity Scan Showing 1 of 1 defect(s) ** CID 1602708: Memory - illegal accesses (USE_AFTER_FREE) /bgpd/bgp_route.c: 15789 in bgp_address_family_distance_delete() ________________________________________________________________________________________________________ *** CID 1602708: Memory - illegal accesses (USE_AFTER_FREE) /bgpd/bgp_route.c: 15789 in bgp_address_family_distance_delete() 15783 safi_t safi = SAFI_UNSPEC; 15784 struct bgp_dest *dest = NULL; 15785 struct bgp_distance *bdistance = NULL; 15786 15787 FOREACH_AFI_SAFI (afi, safi) { 15788 for (dest = bgp_table_top(bgp_distance_table[afi][safi]); dest; >>> CID 1602708: Memory - illegal accesses (USE_AFTER_FREE) >>> Passing freed pointer "dest" as an argument to "bgp_route_next". 15789 dest = bgp_route_next(dest)) { 15790 if (!bgp_dest_has_bgp_path_info_data(dest)) 15791 continue; 15792 bdistance = bgp_dest_get_bgp_distance_info(dest); 15793 XFREE(MTYPE_AS_LIST, bdistance->access_list); 15794 bgp_distance_free(bdistance); ________________________________________________________________________________________________________ To view the defects in Coverity Scan visit, https://u15810271.ct.sendgrid.net/ls/click?upn=u001.AxU2LYlgjL6eX23u9ErQy-2…

1 0

New Defects reported by Coverity Scan for freerangerouting/frr
by scan-admin＠coverity.com 16 Dec '24

16 Dec '24

Hi, Please find the latest report on new defect(s) introduced to freerangerouting/frr found with Coverity Scan. 2 new defect(s) introduced to freerangerouting/frr found with Coverity Scan. New defect(s) Reported-by: Coverity Scan Showing 2 of 2 defect(s) ** CID 1602521: Error handling issues (CHECKED_RETURN) /pimd/pim_rp.c: 1110 in pim_rp_g() ________________________________________________________________________________________________________ *** CID 1602521: Error handling issues (CHECKED_RETURN) /pimd/pim_rp.c: 1110 in pim_rp_g() 1104 if (PIM_DEBUG_PIM_NHT_RP) 1105 zlog_debug( 1106 "%s: NHT Register RP addr %pPA grp %pFX with Zebra", 1107 __func__, &nht_p, &rp_info->group); 1108 pim_nht_find_or_track(pim, nht_p, NULL, rp_info, NULL); 1109 pim_rpf_set_refresh_time(pim); >>> CID 1602521: Error handling issues (CHECKED_RETURN) >>> Calling "pim_nht_lookup_ecmp" without checking return value (as is done elsewhere 9 out of 11 times). 1110 pim_nht_lookup_ecmp(pim, &rp_info->rp.source_nexthop, nht_p, &rp_info->group, true); 1111 return (&rp_info->rp); 1112 } 1113 1114 // About to Go Down 1115 return NULL; ** CID 1602520: Error handling issues (CHECKED_RETURN) /pimd/pim_tib.c: 38 in tib_sg_oil_setup() ________________________________________________________________________________________________________ *** CID 1602520: Error handling issues (CHECKED_RETURN) /pimd/pim_tib.c: 38 in tib_sg_oil_setup() 32 33 pim_addr_to_prefix(&grp, sg.grp); 34 35 up = pim_upstream_find(pim, &sg); 36 if (up) { 37 memcpy(&nexthop, &up->rpf.source_nexthop, sizeof(struct pim_nexthop)); >>> CID 1602520: Error handling issues (CHECKED_RETURN) >>> Calling "pim_nht_lookup_ecmp" without checking return value (as is done elsewhere 9 out of 11 times). 38 pim_nht_lookup_ecmp(pim, &nexthop, vif_source, &grp, false); 39 if (nexthop.interface) 40 input_iface_vif_index = pim_if_find_vifindex_by_ifindex( 41 pim, nexthop.interface->ifindex); 42 } else 43 input_iface_vif_index = pim_nht_lookup_ecmp_if_vif_index(pim, vif_source, &grp); ________________________________________________________________________________________________________ To view the defects in Coverity Scan visit, https://u15810271.ct.sendgrid.net/ls/click?upn=u001.AxU2LYlgjL6eX23u9ErQy-2…

1 0

Re: [dev] Supporting Graceful Restart for more SAFIs (e.g. EVPN)
by Donald Sharp 13 Dec '24

13 Dec '24

As I recall from the discussion during the time GR and EVPN was added was that there was uncertainty about behaviors and what we had time for in relation to testing. So it was not added. From my perspective I would love to have someone work on this bit of code. I would like though the addition of a topotest to show that it actually works. Donald > On Dec 12, 2024, at 11:05 AM, Lukas Stockner via dev <dev(a)lists.frrouting.org> wrote: > > <mime-attachment> > _______________________________________________ > dev mailing list > dev(a)lists.frrouting.org > https://lists.frrouting.org/listinfo/dev

1 0

New Defects reported by Coverity Scan for freerangerouting/frr
by scan-admin＠coverity.com 12 Dec '24

12 Dec '24

Hi, Please find the latest report on new defect(s) introduced to freerangerouting/frr found with Coverity Scan. 1 new defect(s) introduced to freerangerouting/frr found with Coverity Scan. New defect(s) Reported-by: Coverity Scan Showing 1 of 1 defect(s) ** CID 1602463: Error handling issues (CHECKED_RETURN) /pimd/pim_msdp_packet.c: 422 in pim_msdp_pkt_sa_gen() ________________________________________________________________________________________________________ *** CID 1602463: Error handling issues (CHECKED_RETURN) /pimd/pim_msdp_packet.c: 422 in pim_msdp_pkt_sa_gen() 416 417 sa_count = 0; 418 if (PIM_DEBUG_MSDP_INTERNAL) { 419 zlog_debug(" sa gen %d", local_cnt); 420 } 421 >>> CID 1602463: Error handling issues (CHECKED_RETURN) >>> Calling "pim_get_all_mcast_group" without checking return value (as is done elsewhere 9 out of 10 times). 422 pim_get_all_mcast_group(&group_all); 423 pim_msdp_originator_id(pim, &group_all, &rp); 424 local_cnt = pim_msdp_pkt_sa_fill_hdr(pim, local_cnt, rp); 425 426 for (ALL_LIST_ELEMENTS_RO(pim->msdp.sa_list, sanode, sa)) { 427 if (!(sa->flags & PIM_MSDP_SAF_LOCAL)) { ________________________________________________________________________________________________________ To view the defects in Coverity Scan visit, https://u15810271.ct.sendgrid.net/ls/click?upn=u001.AxU2LYlgjL6eX23u9ErQy-2…

1 0

Supporting Graceful Restart for more SAFIs (e.g. EVPN)
by Lukas Stockner 12 Dec '24

12 Dec '24

Hi all, I've been looking into using FRR as a RR for EVPN with Graceful Restart enabled and found that this currently does not work since FOREACH_AFI_SAFI_NSF only goes through the unicast, multicast and MPLS-VPN SAFIs, so EVPN routes aren't preserved and peers which only negotiate EVPN are deleted on disconnect. However, I can't see any reason why such a limitation would be necessary, and simply removing it and using the general FOREACH_AFI_SAFI in the GH code appears to work just fine. >From what I can tell through git blame, it seems like the GH code predates these SAFIs, its loop boundaries weren't updated when e.g. EVPN was added, and when the FOREACH #define was introduced to clean the code up, it was assumed that the difference is intentional. I'd be happy to open a PR to change this, just wanted to check if anyone is aware of a reason to keep this as-is. Cheers, Lukas

1 0

New Defects reported by Coverity Scan for freerangerouting/frr
by scan-admin＠coverity.com 26 Nov '24

26 Nov '24

Hi, Please find the latest report on new defect(s) introduced to freerangerouting/frr found with Coverity Scan. 1 new defect(s) introduced to freerangerouting/frr found with Coverity Scan. New defect(s) Reported-by: Coverity Scan Showing 1 of 1 defect(s) ** CID 1602058: Incorrect expression (COPY_PASTE_ERROR) /bgpd/bgpd.c: 5706 in peer_update_source_addr_set() ________________________________________________________________________________________________________ *** CID 1602058: Incorrect expression (COPY_PASTE_ERROR) /bgpd/bgpd.c: 5706 in peer_update_source_addr_set() 5700 SET_FLAG(member->flags, PEER_FLAG_UPDATE_SOURCE); 5701 member->update_source = sockunion_dup(su); 5702 XFREE(MTYPE_PEER_UPDATE_SOURCE, member->update_if); 5703 member->last_reset = PEER_DOWN_UPDATE_SOURCE_CHANGE; 5704 5705 /* Send notification or reset peer depending on state. */ >>> CID 1602058: Incorrect expression (COPY_PASTE_ERROR) >>> "peer" in "peer->connection" looks like a copy-paste error. 5706 if (!peer_notify_config_change(peer->connection)) 5707 bgp_session_reset(member); 5708 5709 /* Apply new source configuration to BFD session. */ 5710 if (member->bfd_config) 5711 bgp_peer_bfd_update_source(member); ________________________________________________________________________________________________________ To view the defects in Coverity Scan visit, https://u15810271.ct.sendgrid.net/ls/click?upn=u001.AxU2LYlgjL6eX23u9ErQy-2…

1 0

New Defects reported by Coverity Scan for freerangerouting/frr
by scan-admin＠coverity.com 17 Nov '24

17 Nov '24

Hi, Please find the latest report on new defect(s) introduced to freerangerouting/frr found with Coverity Scan. 1 new defect(s) introduced to freerangerouting/frr found with Coverity Scan. New defect(s) Reported-by: Coverity Scan Showing 1 of 1 defect(s) ** CID 1601875: Error handling issues (CHECKED_RETURN) /pimd/pim_rp.c: 1544 in pim_embedded_rp_new() ________________________________________________________________________________________________________ *** CID 1601875: Error handling issues (CHECKED_RETURN) /pimd/pim_rp.c: 1544 in pim_embedded_rp_new() 1538 pim_rp_refresh_group_to_rp_mapping(pim); 1539 if (PIM_DEBUG_PIM_NHT_RP) 1540 zlog_debug("%s: NHT Register RP addr %pPA grp %pFX with Zebra", __func__, 1541 &rp_info->rp.rpf_addr, &rp_info->group); 1542 1543 pim_find_or_track_nexthop(pim, rp_info->rp.rpf_addr, NULL, rp_info, NULL); >>> CID 1601875: Error handling issues (CHECKED_RETURN) >>> Calling "pim_ecmp_nexthop_lookup" without checking return value (as is done elsewhere 8 out of 9 times). 1544 pim_ecmp_nexthop_lookup(pim, &rp_info->rp.source_nexthop, rp_info->rp.rpf_addr, 1545 &rp_info->group, 1); 1546 } 1547 1548 void pim_embedded_rp_delete(struct pim_instance *pim, const pim_addr *group) 1549 { ________________________________________________________________________________________________________ To view the defects in Coverity Scan visit, https://u15810271.ct.sendgrid.net/ls/click?upn=u001.AxU2LYlgjL6eX23u9ErQy-2…

1 0