#
a370b023 |
| 05-Sep-2024 |
Bright Cheng <bright_cheng@wiwynn.com> |
meta-facebook: yosemite4: set powerstate for slot hot-plug
Add function in slot-plug-in and slot-unplug services to make slot power state meet with actual status. 1. RequestedPowerTransition and Req
meta-facebook: yosemite4: set powerstate for slot hot-plug
Add function in slot-plug-in and slot-unplug services to make slot power state meet with actual status. 1. RequestedPowerTransition and RequestedHostTransition to Off while slot being plug-out. 2. RequestedPowerTransition to On while slot being plug-in.
Get slot presence and keep with files for other process that needs to judge with slot presence.
Tested: 1. Plug out slot then the corresponding power state should be off - pass. 2. Plug in slot then the corresponding power staet should be on - pass. 3. The value of slot presence should be changed with plugging-out and plugging-in - pass.
Change-Id: I501fe40019ef1f2ff6b7de193645514bc3b648ef Signed-off-by: Bright Cheng <bright.cheng.wiwynn@gmail.com>
show more ...
|
#
6a1c7b17 |
| 06-Nov-2024 |
Eric Yang <eric.yang.wiwynn@gmail.com> |
meta-facebook: yosemite4: Relocated I3C HUB Enable
Relocated I3C Hub Enable to standby power good pin rising before rescan FRU to ensure SupportedMessageTypes initialization during EID setup. This c
meta-facebook: yosemite4: Relocated I3C HUB Enable
Relocated I3C Hub Enable to standby power good pin rising before rescan FRU to ensure SupportedMessageTypes initialization during EID setup. This change ensures successful initialization of PLDM devices.
Tested: Completed 10 rounds: 1. Successfully verified PLDM devices appear after performing 12V power cycles sequentially on all 8 slots - PASS 2. Build code - PASS
Change-Id: I24e9c09a499434877e517a6733cb508e9f76e909 Signed-off-by: Eric Yang <eric.yang.wiwynn@gmail.com>
show more ...
|
#
248da882 |
| 30-Sep-2024 |
Rush Chen <rush.chen.wiwynn@gmail.com> |
meta-facebook: yosemite4: NIC[0~3] power fault event log
Summary: The system shall register an event when NIC[0~3] power fault. BMC monitors GPIO "ALT_SMB_BMC_CPLD2_N" and then read IOE pin.
Test P
meta-facebook: yosemite4: NIC[0~3] power fault event log
Summary: The system shall register an event when NIC[0~3] power fault. BMC monitors GPIO "ALT_SMB_BMC_CPLD2_N" and then read IOE pin.
Test Plan: With a testing CPLD firmware, set IOE pin high to trigger NIC power fault.
Testing: 1. trigger NIC1 power fault by setting IOE reg 0x0a, bit1 = 1: root@bmc:~# i2ctransfer -f -y 28 w2@0x25 0x0a 0x02 r1 0x00
2. read IOE reg 0x01 (bit1 = 1): root@bmc:~# i2ctransfer -f -y 28 w1@0x25 0x01 r1 0x02
3. Check event log root@bmc:~# mfg-tool log-display "2": { "additional_data": [ "NIC1 Power Fault=1" ], "event_id": "", "message": "NIC1 Power Fault", "resolution": "", "resolved": false, "severity": "xyz.openbmc_project.Logging.Entry.Level.Error", "timestamp": "2024-08-19T07:50:01.311000000Z", "updated_timestamp": "2024-08-19T07:50:01.311000000Z" },
Change-Id: I12b297b14b8be4c97c1f7d0949b0b6c16b91fa40 Signed-off-by: Rush Chen <rush.chen.wiwynn@gmail.com>
show more ...
|
#
9d7cf2fb |
| 28-Aug-2024 |
Lora Lin <lora.lin.wiwynn@gmail.com> |
meta-facebook: yosemite4: Monitor host power status
Description: - Monitor host power status. - Request the host power transition when status change.
Note: - Compatible table - SD CPLD firmware
meta-facebook: yosemite4: Monitor host power status
Description: - Monitor host power status. - Request the host power transition when status change.
Note: - Compatible table - SD CPLD firmware version: 00010302 BMC INT_SMB_BMC_SLOT1_4_N or INT_SMB_BMC_SLOT5_8_N will be triggered when the CPU power status changes or a power fault occurs in the slot.
Test Plan: - Check host power state after power cycle/on/off: 200 rounds Pass - Check host power state after sled-cycle: 20 rounds Pass
Change-Id: I2b7cfd748cf3e56fb0b4527deede7bcecd9397d0 Signed-off-by: Lora Lin <lora.lin.wiwynn@gmail.com>
show more ...
|
#
1a69cc1e |
| 05-Aug-2024 |
PeterHo-wiwynn <peter.ho.wiwynn@gmail.com> |
meta-facebook: yosemite4: Revise HSC OCP setting
Revise HSC OCP setting between T1/T2 configuration.
Tested: Tested and verified on yosemite4. Change-Id: I36f832dc6c3bc91b3f3339f638872735be9c5a2e S
meta-facebook: yosemite4: Revise HSC OCP setting
Revise HSC OCP setting between T1/T2 configuration.
Tested: Tested and verified on yosemite4. Change-Id: I36f832dc6c3bc91b3f3339f638872735be9c5a2e Signed-off-by: PeterHo-wiwynn <peter.ho.wiwynn@gmail.com>
show more ...
|
#
78c198b9 |
| 26-Aug-2024 |
Patrick Williams <patrick@stwcx.xyz> |
meta-facebook: remove .sh suffix for scripts
Scripts and executables should not have a language-based extension in order to simplify user operations and to avoid excessive refactoring when rewrites
meta-facebook: remove .sh suffix for scripts
Scripts and executables should not have a language-based extension in order to simplify user operations and to avoid excessive refactoring when rewrites are done. Remove all `.sh` extensions on scripts.
Signed-off-by: Patrick Williams <patrick@stwcx.xyz> Change-Id: I408fa0f31e6064c0ce3c1559a4182cd625e91f21
show more ...
|
#
fade53eb |
| 15-Aug-2024 |
Eric Yang <eric.yang.wiwynn@gmail.com> |
meta-facebook: yosemite4: Prevent BMC reboot by disabling/enabling i3c hub
During hot service, disable and enable the i3c hub to prevent potential BMC reboots caused by noise from slot plug-in.
Thi
meta-facebook: yosemite4: Prevent BMC reboot by disabling/enabling i3c hub
During hot service, disable and enable the i3c hub to prevent potential BMC reboots caused by noise from slot plug-in.
This change addresses the issue where enabling the i3c port during slot plug-in can generate electrical noise due to leakage.
These noises, when sent to the BMC by the i3c hub, cause interrupt storming, leading to the BMC CPU getting stuck and unexpected reboots.
Tested: - Verified that the i3c hub is correctly disabled and enabled during hot service -pass. - Performed 60 plug-in and unplug operations, ensuring only one slot was operated at a time. No BMC reboots were observed -pass.
Change-Id: I030daf8b4b68a43204847eeb2153757ccb18292f Signed-off-by: Eric Yang <eric.yang.wiwynn@gmail.com>
show more ...
|
#
aae4247f |
| 20-Aug-2024 |
Yikai Tsai <yikai.tsai.wiwynn@gmail.com> |
meta-facebook: yosemite4: Fan board eFuse fault event
Summary: The system shall register an event in response to fan board eFuse fault. BMC is triggered by INT_FANBOARD0_IOEXP_N or INT_FANBOARD1_IOE
meta-facebook: yosemite4: Fan board eFuse fault event
Summary: The system shall register an event in response to fan board eFuse fault. BMC is triggered by INT_FANBOARD0_IOEXP_N or INT_FANBOARD1_IOEXP_N to identify which fan board has eFuse fault. Then, BMC reads relevant fan board IOE to check affected eFuse.
Test Plan: Pull FLT_EFUSE0_N low to make INT_FANBOARD0_IOEXP_N be low.
Testing: Take fan board 0 and eFuse 0 for example.
1. Check fan board 0 IOE init status root@bmc:~# i2ctransfer -f -y 30 w1@0x22 0 r1 0xff
2. Pull FLT_EFUSE0_N low and check fan board 0 IOE init status root@bmc:~# i2ctransfer -f -y 30 w1@0x22 0 r1 0xfb
3. Check event log root@bmc:~# mfg-tool log-display ... "2": { "additional_data": [ "FAULT=eFuse0" ], "event_id": "", "message": "Fan board 0 eFuse fault", "resolution": "", "resolved": false, "severity": "xyz.openbmc_project.Logging.Entry.Level.Error", "timestamp": "2024-02-27T17:31:55.420000000Z", "updated_timestamp": "2024-02-27T17:31:55.420000000Z" }
Change-Id: Idb1a6a767b1e412931d229eb9cfaac7624fdec0a Signed-off-by: Yikai Tsai <yikai.tsai.wiwynn@gmail.com>
show more ...
|
#
817a2d7f |
| 23-Jul-2024 |
Yikai Tsai <yikai.tsai.wiwynn@gmail.com> |
meta-facebook: yosemite4: Power fault event
Summary: The system shall register an event in the case of BIC and CPU power fault. BMC is triggered by INT_SMB_BMC_SLOT1_4_N or INT_SMB_BMC_SLOT5_8_N, ch
meta-facebook: yosemite4: Power fault event
Summary: The system shall register an event in the case of BIC and CPU power fault. BMC is triggered by INT_SMB_BMC_SLOT1_4_N or INT_SMB_BMC_SLOT5_8_N, checking which SLOT got power fault via reading the MUX (i2c-bus 8, 0x70 for SLOT1-4, and i2c-bus 9, 0x71 for SLOT1-4). Then, BMC reads CPLD IOE for each failed SLOT to check affected power rails.
For BIC power fault, check address 0x24, port 1:
bit [1] PWRGD_P1V2_STBY_FAULT [0] PWRGD_P1V8_STBY_FAULT
For CPU power fault, check address 0x24, port 2:
bit [4] PWRGD_PVDD11_S3_R_FAULT [3] PWRGD_PVDDIO_FAULT [2] PWRGD_PVDDCR_CPU1_FAULT [1] PWRGD_PVDDCR_SOC_FAULT [0] PWRGD_PVDDCR_CPU0_FAULT
Test Plan: Trigger power fault by setting CPLD (0x0F) on server board to pull INT_SMB_BMC_SLOT1_4_N or INT_SMB_BMC_SLOT5_8_N low.
Testing: Take SLOT2 for example.
1. Check gpio status root@bmc:~# cat /sys/kernel/debug/gpio | grep -i 1_4 gpio-612 (INT_SMB_BMC_SLOT1_4_|gpio_monitor ) in hi IRQ
2. Check (1 to 4)MUX status root@bmc:~# i2ctransfer -f -y 8 w1@0x70 0 r1 0x00
3. Check SLOT2 CPLD IOE status root@bmc:~# i2ctransfer -y -f 1 w1@0x24 0x01 r1 0x00 root@bmc:~# i2ctransfer -y -f 1 w1@0x24 0x02 r1 0x00
4. Trigger power fault by setting CPLD (0x0F) on server board root@bmc:~# i2ctransfer -y -f 1 w2@0xf 0x3f 0x1
5. Check event log root@bmc:~# mfg-tool log-display ... }, "2": { "additional_data": [ "FAULT=PWRGD_P1V8_STBY_FAULT" ], "event_id": "", "message": "SLOT2 AC Power Fault", "resolution": "", "resolved": false, "severity": "xyz.openbmc_project.Logging.Entry.Level.Error", "timestamp": "2024-07-18T08:37:13.179000000Z", "updated_timestamp": "2024-07-18T08:37:13.179000000Z" }, "3": { "additional_data": [ "FAULT=PWRGD_PVDDCR_CPU0_FAULT" ], "event_id": "", "message": "SLOT2 DC Power Fault", "resolution": "", "resolved": false, "severity": "xyz.openbmc_project.Logging.Entry.Level.Error", "timestamp": "2024-07-18T08:37:13.231000000Z", "updated_timestamp": "2024-07-18T08:37:13.231000000Z" }
6. Check gpio status root@bmc:~# cat /sys/kernel/debug/gpio | grep -i 1_4 gpio-612 (INT_SMB_BMC_SLOT1_4_|gpio_monitor ) in lo IRQ
7. Check (1 to 4)MUX status root@bmc:~# i2ctransfer -f -y 8 w1@0x70 0 r1 0x20
8. Check SLOT2 CPLD IOE status root@bmc:~# i2ctransfer -y -f 1 w1@0x24 0x01 r1 0x01 root@bmc:~# i2ctransfer -y -f 1 w1@0x24 0x02 r1 0x01
Change-Id: I67b93e9a55c949af54aee4ca90dc998448bdfdcc Signed-off-by: Yikai Tsai <yikai.tsai.wiwynn@gmail.com>
show more ...
|
#
ab989dc6 |
| 23-Jul-2024 |
Yikai Tsai <yikai.tsai.wiwynn@gmail.com> |
meta-facebook: yosemite4: HSC fault event
Summary: The system shall register an event in response to an HSC fault. BMC monitors Medusa HSC fault pin (FLT_HSC_SERVER_SLOTx_N) on spider board CPLD IO
meta-facebook: yosemite4: HSC fault event
Summary: The system shall register an event in response to an HSC fault. BMC monitors Medusa HSC fault pin (FLT_HSC_SERVER_SLOTx_N) on spider board CPLD IO expander. Check power fault reason via I2C to HSC STATUS_WORD (0x79).
Test Plan: Trigger HSC OTP fault by setting lower OT_FAULT_LIMIT (0x4F) to pull FLT_HSC_SERVER_SLOTx_N low.
Testing: Take FLT_HSC_SERVER_SLOT7_N for example.
1. Check init status root@bmc:~# cat /sys/kernel/debug/gpio | grep -i FLT_HSC_SERVER_SLOT7 gpio-698 (FLT_HSC_SERVER_SLOT7|gpio_monitor ) in hi IRQ root@bmc:~# i2ctransfer -f -y 6 w1@0x40 0x79 r2 0x00 0x00 root@bmc:~# i2ctransfer -f -y 6 w1@0x40 0x4f r2 0xff 0x0f
2. Trigger HSC OTP fault root@bmc:~# i2ctransfer -f -y 6 w3@0x40 0x4f 0x02 0x1 root@bmc:~# i2ctransfer -f -y 6 w1@0x40 0x4f r2 0x02 0x01 root@bmc:~# i2ctransfer -f -y 6 w1@0x40 0x79 r2 0x45 0x18 root@bmc:~# cat /sys/kernel/debug/gpio | grep -i FLT_HSC_SERVER_SLOT7 gpio-698 (FLT_HSC_SERVER_SLOT7|gpio_monitor ) in lo IRQ
3. Check event log root@bmc:~# mfg-tool log-display ... }, "97": { "additional_data": [ "STATUS_WORD=0x1845" ], "event_id": "", "message": "SLOT7 HSC Fault", "resolution": "", "resolved": false, "severity": "xyz.openbmc_project.Logging.Entry.Level.Error", "timestamp": "2024-07-17T02:21:30.707000000Z", "updated_timestamp": "2024-07-17T02:21:30.707000000Z" },
Change-Id: I88bac394a174165d91b471db25c0e43ccc744eec Signed-off-by: Yikai Tsai <yikai.tsai.wiwynn@gmail.com>
show more ...
|
Revision tags: 2.17.0-dev |
|
#
c678673d |
| 10-May-2024 |
Allen.Wang <Allen_Wang@quantatw.com> |
meta-facebook: yosemite4:Reconfig network interface for NIC hot service
While NIC hot remove/plug or replace another NIC, it must down/up to NCSI re-initialization and reconfigure net interface to r
meta-facebook: yosemite4:Reconfig network interface for NIC hot service
While NIC hot remove/plug or replace another NIC, it must down/up to NCSI re-initialization and reconfigure net interface to recovery network link.
Tested: build and test on yosemite4 system
NIC1 (eth0) connected CX7 NIC
root@bmc:~# ifconfig eth0 eth0 Link encap:Ethernet HWaddr E8:EB:D3:46:D8:71 inet addr:10.10.14.236 Bcast:10.10.255.255 Mask:255.255.0.0 inet6 addr: 2018:5::4ca9/128 Scope:Global inet6 addr: fe80::eaeb:d3ff:fe46:d871/64 Scope:Link inet6 addr: 2019:7::eaeb:d3ff:fe46:d871/64 Scope:Global UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:23692 errors:0 dropped:11 overruns:0 frame:0 TX packets:263 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:1000 RX bytes:2510627 (2.3 MiB) TX bytes:21546 (21.0 KiB) Interrupt:42
Test 1: remove and plug same NIC
bmc still pingable
ping 10.10.14.236
Pinging 10.10.14.236 with 32 bytes of data: Reply from 10.10.14.236: bytes=32 time=1ms TTL=64 Reply from 10.10.14.236: bytes=32 time=1ms TTL=64 Reply from 10.10.14.236: bytes=32 time=1ms TTL=64 Reply from 10.10.14.236: bytes=32 time<1ms TTL=64
Test2: Remove CX7 NIC and plug another BRCM NIC
eth0 will reconfigured by correct MAC address and its pingable
ifconfig eth0 eth0 Link encap:Ethernet HWaddr D4:04:E6:36:58:19 inet addr:10.10.12.226 Bcast:10.10.255.255 Mask:255.255.0.0 inet6 addr: fe80::d604:e6ff:fe36:5819/64 Scope:Link UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:25736 errors:0 dropped:12 overruns:0 frame:0 TX packets:404 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:1000 RX bytes:2749245 (2.6 MiB) TX bytes:32907 (32.1 KiB) Interrupt:42
ping 10.10.12.226
Pinging 10.10.12.226 with 32 bytes of data: Reply from 10.10.12.226: bytes=32 time=1ms TTL=64 Reply from 10.10.12.226: bytes=32 time=1ms TTL=64 Reply from 10.10.12.226: bytes=32 time=7ms TTL=64 Reply from 10.10.12.226: bytes=32 time=1ms TTL=64
Change-Id: I8130eb4ba317ebd66378d646ca95e03dd7481fd7 Signed-off-by: Allen.Wang <Allen_Wang@quantatw.com>
show more ...
|
#
5fd67bdf |
| 10-Apr-2024 |
Delphine CC Chiu <Delphine_CC_Chiu@wiwynn.com> |
meta-facebook: yosemite4: Rescan WF BIC and CXL EID
The WF BIC and CXL might not ready when BMC setting up MCTP network. Add rescan mechanism to make sure it get the EIDs.
Change-Id: I7d57f549a4b7c
meta-facebook: yosemite4: Rescan WF BIC and CXL EID
The WF BIC and CXL might not ready when BMC setting up MCTP network. Add rescan mechanism to make sure it get the EIDs.
Change-Id: I7d57f549a4b7c9441eadab08d4df0d1e4b46e4d1 Signed-off-by: Delphine CC Chiu <Delphine_CC_Chiu@wiwynn.com>
show more ...
|
#
0598b574 |
| 12-Mar-2024 |
Chris Wang <Chris_Wang@wiwynn.com> |
meta-facebook: yosemite4: Support NIC endpoint setup
Setup the MCTP Endpoint for the NIC when inserted. Remove the MCTP Endpoint for the NIC when unplugged.
Change-Id: I0c36cecda6509a64e4f26c70bf61
meta-facebook: yosemite4: Support NIC endpoint setup
Setup the MCTP Endpoint for the NIC when inserted. Remove the MCTP Endpoint for the NIC when unplugged.
Change-Id: I0c36cecda6509a64e4f26c70bf61590467571a7b Signed-off-by: Delphine CC Chiu <Delphine_CC_Chiu@wiwynn.com>
show more ...
|
#
a2fb20c3 |
| 05-Feb-2024 |
Patrick Williams <patrick@stwcx.xyz> |
meta-facebook: remove excess override syntaxes
When a bbappend file is already in a meta-machine layer, there is no reason for extra ":machine" override syntax usage. Remove them all.
Signed-off-b
meta-facebook: remove excess override syntaxes
When a bbappend file is already in a meta-machine layer, there is no reason for extra ":machine" override syntax usage. Remove them all.
Signed-off-by: Patrick Williams <patrick@stwcx.xyz> Change-Id: I610e72633adb5bc36b63fe2b727e394c61070185
show more ...
|
#
d3135247 |
| 19-Jan-2024 |
Patrick Williams <patrick@stwcx.xyz> |
meta-facebook: reduce permissions on scripts
Scripts should be installed with 0755 permissions and not 0777, otherwise non-root users can potentially modify or delete them.
Signed-off-by: Patrick W
meta-facebook: reduce permissions on scripts
Scripts should be installed with 0755 permissions and not 0777, otherwise non-root users can potentially modify or delete them.
Signed-off-by: Patrick Williams <patrick@stwcx.xyz> Change-Id: I41270a2bb3fb940d8ca49ed6230545d98efb2fea
show more ...
|
#
2f1f94c4 |
| 12-Jan-2024 |
Delphine CC Chiu <Delphine_CC_Chiu@wiwynn.com> |
meta-facebook: yosemite4: Support power policy after doing slot hot-plug
Add systemd service slot-hot-plug to recover host status by power policy
Tested: plug out then in slot, check host status an
meta-facebook: yosemite4: Support power policy after doing slot hot-plug
Add systemd service slot-hot-plug to recover host status by power policy
Tested: plug out then in slot, check host status and journal
Change-Id: I8168b32a40c34d787fbfdd6c6a0c944ace919514 Signed-off-by: Delphine CC Chiu <Delphine_CC_Chiu@wiwynn.com>
show more ...
|
Revision tags: 2.16.0-dev |
|
#
085d6197 |
| 24-Nov-2023 |
Delphine CC Chiu <Delphine_CC_Chiu@wiwynn.com> |
meta-facebook: yosemite4: rescan eeproms after chassis on/off
Add service for rescanning the eeproms after chassis on/off event detected by gpio-monitor.
Tested: power on/off chassis and check dbus
meta-facebook: yosemite4: rescan eeproms after chassis on/off
Add service for rescanning the eeproms after chassis on/off event detected by gpio-monitor.
Tested: power on/off chassis and check dbus path on FruDevice. Change-Id: Ie85f1526bf14cd87f92f93e37839bc80cf9044a4 Signed-off-by: Delphine CC Chiu <Delphine_CC_Chiu@wiwynn.com>
show more ...
|
#
6e311d3c |
| 15-Nov-2023 |
Delphine CC Chiu <Delphine_CC_Chiu@wiwynn.com> |
yosemite4: Support gpio monitor for server board device probing
If 12v stby power state was off when bmc boot up, server board device driver will not be probed. so add gpio monitor support for serve
yosemite4: Support gpio monitor for server board device probing
If 12v stby power state was off when bmc boot up, server board device driver will not be probed. so add gpio monitor support for server board device probing
Tested: devices on server board were probed after 12v on or hot-pluged
Change-Id: I9e2a6cee966dd1c64cbff53b368fa6686a8c3924 Signed-off-by: Delphine CC Chiu <Delphine_CC_Chiu@wiwynn.com>
show more ...
|
#
9d5d2484 |
| 18-Oct-2023 |
Delphine CC Chiu <Delphine_CC_Chiu@wiwynn.com> |
yosemite4: phosphor-gpio-monitor: Support sled cycle from power button
Summary: Support sled cycle from power button through gpio monitor, which detecting interrupt for bmc gpiop1
Tested: Do sled c
yosemite4: phosphor-gpio-monitor: Support sled cycle from power button
Summary: Support sled cycle from power button through gpio monitor, which detecting interrupt for bmc gpiop1
Tested: Do sled cycle after pressing power button: pass root@bmc:~# /usr/bin/phosphor-multi-gpio-monitor --config /usr/share/phosphor-gpio-monitor/phosphor-multi-gpio-monitor.json ... <6> GPIO Line 121 monitoring started ... [press power button over 4 second] <6> GPIO Line 121 Asserted
Change-Id: I9197bfabff8b7e1f7a7370c0ff37219440df45b7 Signed-off-by: Delphine CC Chiu <Delphine_CC_Chiu@wiwynn.com>
show more ...
|