BMC/tmp/openbmc.git/meta-facebook/conf/machine/include, branch master

meta-facebook: darwin: add machine

2025-07-23T14:52:36+00:00

Initial commit for darwin. Change-Id: I65ff2a4fc671b63bdbd982b950c29896ef696d44 Signed-off-by: Robert Ercean Dima

meta-facebook: add liquid-cooled machine feature

2025-06-04T21:03:48+00:00

Automatically enable features such as leak-detection based on the liquid-cooled machine feature. Signed-off-by: Patrick Williams Change-Id: I034694dfa2f624f7f13f2101582860052bee35fe

slp: remove from default images

2025-05-15T09:51:21+00:00

slpd-lite[1] is a home grown application within OpenBMC that does not appear to get much use. There are other more full function options like Avahi or LLDP now. The slpd-lite daemon provides an unauthenticated network interface which was the result of a fairly severe PSIRT last year[2]. Given the above, this application should not be included by default when building OpenBMC. Tested: - Confirmed slpd application and service no longer in a Romulus image [1]: https://github.com/openbmc/slpd-lite [2]: https://github.com/openbmc/slpd-lite/security/advisories/GHSA-wmgv-jffg-v3xr Change-Id: Ifd67651cb00a411bf12f0330c4b4fab626953361 Signed-off-by: Andrew Geissler

meta-facebook: Enable cgroup support for all meta BMCs

2025-05-02T16:43:45+00:00

Tested: Built a few of these and saw that the kernel configurations for cgroups were as we wanted Change-Id: Iccaa81bf6071c3184d69543c84e10d8d38722d72 Signed-off-by: James Athappilly

meta-facebook: Introduce reconfig-interface-duid-ll service support

2025-04-09T19:46:18+00:00

Introduce the reconfig-interface-duid-ll service for platforms that use configuration files generated by phosphor-networkd. This service helps adding the DUID-LL (DHCP Unique Identifier – Link Layer) setting to the generated network interface configuration files, ensuring proper DHCP configuration for network identification. Tested on bletchley: ``` root@bmc:~# networkctl status eth0 * 2: eth0 Link File: /usr/lib/systemd/network/99-default.link Network File: /etc/systemd/network/00-bmc-eth0.network State: routable (configured) Online state: online Type: ether Path: platform-1e670000.ethernet Driver: ftgmac100 Hardware Address: c0:18:50:f1:d7:96 MTU: 1500 (min: 68, max: 1500) QDisc: pfifo_fast IPv6 Address Generation Mode: eui64 Number of Queues (Tx/Rx): 1/1 Auto negotiation: yes Speed: 1Gbps Duplex: full Port: mii Address: 10.10.14.25 (DHCPv4 via 10.10.10.106) 2018:5::aeba 2019:7::c218:50ff:fef1:d796 fe80::c218:50ff:fef1:d796 Gateway: 10.10.10.204 fe80::a00:27ff:fee1:5408 DNS: 8.8.8.8 168.95.1.1 Search Domains: dhcpserver.bu9bmc.local mmb.ipv6.com Activation Policy: up Required For Online: yes DHCPv4 Client ID: c0:18:50:f1:d7:96 DHCPv6 Client IAID: 0x0 DHCPv6 Client DUID: DUID-LL:0001c01850f1d796 ``` Change-Id: I4276708dec4e53e354fc8a92cbf5fc9d7a92a61c Signed-off-by: Potin Lai Signed-off-by: Patrick Williams

meta-phosphor: add distro feature production-level tweaks

2025-04-08T12:51:22+00:00

There are some settings that are desired to be set for production-level systems, but may be less desired in development. Such as settings that turn hard-errors into soft-errors. Add a distro feature to allow these settings to be enabled. Add as a production-level tweak, a config for the kernel default setting of `hung_task_panic`. This prevents unnecessary panics due to a busy system, usually due to poor driver implementations. Enable the production-level tweaks on meta-facebook. Tested: ``` root@yosemite4:~# cat /proc/sys/kernel/hung_task_panic 0 ``` Change-Id: I0542547c77ddacae44d98e9ab6edc3743ff0016c Signed-off-by: Cosmo Chou Signed-off-by: Patrick Williams

meta-facebook: explicitly disable avahi, ldap, slp

2024-08-19T18:47:23+00:00

Meta doesn't use these protocols on our systems, so disable them. Signed-off-by: Patrick Williams Change-Id: I30b6723b061beabb7d8b6dc6a6cf024e6ce5bc77

meta-facebook: fix ipv6-ll address

2024-01-18T20:18:51+00:00

The correct MAC address of NCSI managerd ethernet interface is obtained during NCSI config process. It will need run extra workaround to regenerated IPv6-LL address based on new MAC address from NCSI. Add service for checking the MAC address is correctly get from NIC card via NCSI, then regenerate the IPv6-LL address. Tested on Harma: ``` root@harma:~# ifconfig eth0 eth0 Link encap:Ethernet HWaddr 10:70:FD:E4:3F:63 inet addr:10.10.12.110 Bcast:10.10.255.255 Mask:255.255.0.0 inet6 addr: fe80::1270:fdff:fee4:3f63/64 Scope:Link inet6 addr: 2018:5::e52e/128 Scope:Global inet6 addr: 2019:7::1270:fdff:fee4:3f63/64 Scope:Global UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:97642 errors:0 dropped:539 overruns:0 frame:0 TX packets:1625 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:1000 RX bytes:15804387 (15.0 MiB) TX bytes:140035 (136.7 KiB) Interrupt:42 root@harma:~# systemctl status network-wait-ipv6-ll@eth0.service * network-wait-ipv6-ll@eth0.service - Wait IPv6 link-local initialization Loaded: loaded (/usr/lib/systemd/system/network-wait-ipv6-ll@.service; static) Active: active (exited) since Tue 2024-01-09 03:33:43 PST; 38s ago Process: 271 ExecStart=/usr/libexec/check-ipv6-ll eth0 (code=exited, status=0/SUCCESS) Main PID: 271 (code=exited, status=0/SUCCESS) CPU: 148ms Jan 09 03:33:41 harma systemd[1]: Starting Wait IPv6 link-local initialization... Jan 09 03:33:41 harma check-ipv6-ll[271]: non-global mac address, start fix eth0 ipv6-ll Jan 09 03:33:43 harma check-ipv6-ll[271]: eth0 mac: 10:70:fd:e4:3f:63 Jan 09 03:33:43 harma check-ipv6-ll[271]: expected ipv6-ll: fe80::1270:fdff:fee4:3f63 Jan 09 03:33:43 harma systemd[1]: Finished Wait IPv6 link-local initialization. ``` Signed-off-by: Potin Lai Change-Id: Id3c40beda9f290d36565f0e8e146be772c9a9983

meta-facebook: opt out of webui

2023-10-11T05:29:09+00:00

Signed-off-by: Patrick Williams Change-Id: I80149293de370f5f979d657ba43efa4b7d556190

meta-facebook: remove IPMI hw from multihost systems

2023-08-18T00:19:50+00:00

All meta-facebook systems use ipmb to interact with the OCP debug card. The multi-host machines also use ipmb to talk with the BIC, but the 'ipmi-ipmb' package is not considered an 'ipmi-hw' package (because it is often installed as an addition). The single host machine(s) use 'ipmi-kcs'. Rearrange the inc file specifications so that the multi-host systems do not get the 'ipmi-kcs' package. Signed-off-by: Patrick Williams Change-Id: I756984d32d424dd0b08a44b434318ccee6672889