systemctl start obmc-host-start at 0.target fail

Deng Tyler tyler.sabdon at gmail.com
Tue Jan 7 00:20:59 AEDT 2020 

Hi all:
    Thanks for your sharing and that's really helpful. I still have a
question for chassis power on/off.  I add service to run poweron shell
while obmc-host-start at t0.target triggered and run poweroff shell while
obms-host-shutdown at 0.target triggered. After issuing chassis power on
command, my platform can power on system but power off after few seconds. I
check journalctl as below:
Jan 06 11:05:39   systemd[1]: Starting Start Watchdog0...
Jan 06 11:05:39   systemd[1]: Starting Assert bmc_booted LED...
Jan 06 11:05:39   phosphor-watchdog[352]: watchdog: enabled and started
Jan 06 11:05:39   phosphor-ledmanager[222]: [[0;1;31m[[0;1;39m[[0;1;31mNo
service providers for physical LED[[0m
Jan 06 11:05:39   systemd[1]: Started Start Watchdog0.
Jan 06 11:05:39   systemd[1]: Started Assert bmc_booted LED.
Jan 06 11:05:39   systemd[1]: Reached target Start Host0 Minimum.
Jan 06 11:05:39   systemd[1]: Reached target Start Host0.
Jan 06 11:05:39   systemd[1]: Startup finished in 4.527s (kernel) + 1min
32.169s (userspace) = 1min 36.697s.
Jan 06 11:05:39   phosphor-host-state-manager[315]: Received signal that
host is running
Jan 06 11:05:39   phosphor-host-state-manager[315]: Change to Host State
Jan 06 11:05:39   phosphor-timemanager[320]: Changing time settings is
*deferred* now
Jan 06 11:06:09   phosphor-watchdog[352]: watchdog: Timed out
Jan 06 11:06:09   systemd[1]: unit_file_find_fragment:
obmc-host-timeout at .target+0
→ obmc-host-timeout at 0.target
Jan 06 11:06:09   phosphor-watchdog[352]: watchdog: disabled
Jan 06 11:06:09   systemd[1]: Reached target Quiesce Target.
Jan 06 11:06:09   systemd[1]: Reached target Host instance 0 timed out
refreshing Watchdog.
Jan 06 11:06:09   systemd[1]: Stopping Phosphor poweron watchdog...
Jan 06 11:06:09   systemd[1]: phosphor-watchdog at poweron.service: Succeeded.
Jan 06 11:06:09   systemd[1]: Stopped Phosphor poweron watchdog.
Jan 06 11:06:09   phosphor-host-state-manager[315]: Auto reboot enabled,
rebooting
Jan 06 11:06:09   phosphor-host-state-manager[315]: Beginning reboot...
Jan 06 11:06:09   phosphor-host-state-manager[315]: Host State transaction
request
Jan 06 11:06:09   systemd[1]: unit_file_find_fragment:
obmc-host-reboot at .target+0
→ obmc-host-reboot at 0.target
Jan 06 11:06:09   systemd[1]: unit_file_find_fragment:
obmc-host-shutdown at .target+0
→ obmc-host-shutdown at 0.target
Jan 06 11:06:09   systemd[1]: unit_file_find_fragment:
obmc-host-stopping at .target+0
→ obmc-host-stopping at 0.target
Jan 06 11:06:09   systemd[1]: unit_file_find_fragment:
obmc-host-stopped at .target+0
→ obmc-host-stopped at 0.target
Jan 06 11:06:09   systemd[1]: unit_file_find_fragment:
obmc-power-stop-pre at .target+0
→ obmc-power-stop-pre at 0.target
Jan 06 11:06:09   systemd[1]: unit_file_find_fragment:
obmc-host-stop-pre at .target+0
→ obmc-host-stop-pre at 0.target
Jan 06 11:06:09   systemd[1]: Created slice
system-phosphor\x2dreboot\x2dhost.slice.
Jan 06 11:06:09   systemd[1]: Reached target Chassis0 (Power Off).
Jan 06 11:06:09   systemd[1]: Stopped target Start Host0.
Jan 06 11:06:09   systemd[1]: Stopped target Start Host0 Minimum.
Jan 06 11:06:09   systemd[1]: Reached target Stop Host0 (Pre).
Jan 06 11:06:09   systemd[1]: Stopped target Host instance 0 timed out
refreshing Watchdog.
Jan 06 11:06:09   systemd[1]: Stopped target Quiesce Target.
Jan 06 11:06:09   systemd[1]: Starting Configure GPIOs for Host Power
Control...
Jan 06 11:06:09   systemd[1]: Starting Wait for /org/openbmc/HostIpmi/1...
Jan 06 11:06:09   systemd[1]: Started Reboot host0.
Jan 06 11:06:09   init_once.sh[404]: init_once.sh
Jan 06 11:06:09   phosphor-chassis-state-manager[290]: Received signal that
power OFF is complete
Jan 06 11:06:09   phosphor-chassis-state-manager[290]: Change to Chassis
Power State
Jan 06 11:06:09   systemd[1]: Started Configure GPIOs for Host Power
Control.
Jan 06 11:06:09   systemd[1]: Starting Stop Host...
Jan 06 11:06:09   poweroff.sh[408]: out -> poweroff

The power on process suspended by phosphor-watchdog timeout. I have no idea
to clear watchdog timer. Could you provide any suggestion for this
situation? Thanks.

Tyler.

zhang_cy1989 <zhang_cy1989 at 163.com> 於 2019年12月26日 週四 上午11:11寫道：

> Dear Tim
>   Thanks a milion.
>   The infromation is very useful for me.
>
>   Thank you again for your prompt reply.
>
>   Best Regards
>
> Felix
> 在 2019-12-26 10:32:54，"CS20 CHLi30" <CHLI30 at nuvoton.com> 写道：
>
> Hi Felixzhang,
>
>
>
> Yes, you can using op-power-control on X86 motherboard for sure.
>
>
>
> Example about using op-power-control:
>
> systemctl list-dependencies obmc-host-start at 0.target
>
>
>
> obmc-host-start at 0.target (Power ON)
>
> -> mapper-wait at -org-openbmc-control-chassis0.service
>
> -> phosphor-reset-host-reboot-attempts at 0.service
>
> -> obmc-host-startmin at 0.target
>
>      -> mapper-wait at -org-openbmc-control-chassis0.service
>
>      -> obmc-enable-host-watchdog at 0.service
>
>      -> phosphor-watchdog at poweron.service
>
>      -> obmc-chassis-poweron at 0.target
>
>           -> op-power-start at 0.service
>
>           -> mapper-wait at -xyz-openbmc_project-state-chassis0.service
>
>           -> phosphor-fan-control-init at 0.service
>
>           -> phosphor-fan-monitor-init at 0.service
>
>           -> phosphor-fan-presence-tach at 0.service
>
>
>
> BTW, here is the list about the sequence of obmc-host-start at 0.target for
> you refer it.
>
> When you presss Power ON button from WebUI the state
> RequestedHostTransition change to ON, what happens behind that is done by
> systemd units, roughly described as below:
>
> 1.      Host state manager starts *obmc-host-start at 0.target*
>
> 2.      The above target requires *obmc-host-startmin at 0.target*
>
> 3.      The above target requires *obmc-chassis-poweron at 0.target* and
> *start_host at 0.service*
>
> 4.      *obmc-chassis-poweron at 0.target* requires
> *op-power-start at 0.service* and *op-wait-power-on at 0.service*
>
> 5.      *op-power-start at 0.service* is to call the PowerControl service to
> DC on. And *op-wait-power-on at 0.service* is to wait for PGood after DC on.
>
> 6.      After that (DC is on and PGood is OK), *start_host at 0.service* is
> to start the host CPU.
>
>
>
> Best regards,
>
> Tim
>
> *From:* zhang_cy1989 [mailto:zhang_cy1989 at 163.com]
> *Sent:* Thursday, December 26, 2019 10:22 AM
> *To:* CS20 CHLi30 <CHLI30 at nuvoton.com>
> *Cc:* openbmc at lists.ozlabs.org
> *Subject:* Re:RE: systemctl start obmc-host-start at 0.target fail
>
>
>
> Dear,Tim
>
>    Thank you for guidance.
>
>    I will study the dbus-monitor and the dependencies of
> obmc-host-start at 0.target.
>
>  @CS20 CHLi30
>
>   I need ask you another advice.
>
>
>
>   I use the romulus recipes to build BMC firmware. And use this fw to the
> X86 motherboard + AST2500.
>
>   The romulus is base on openpower. And my platform is X86 motherboard.
>
>   *Is it feasible that using romulus BMC firmware to X86 motherboard,
> specially the power control part?*
>
>
>
>   Would you have better suggestion?
>
>
>
>   Looking forward to your replay.
>
>
>
>  Best Regards
>
>  Felixzhang
>
>
>
> At 2019-12-26 09:34:09, "CS20 CHLi30" <CHLI30 at nuvoton.com> wrote:
>
> Hi Felixzhang,
>
>
>
> About this kind of systemd target start fail, you can use *dbus-monitor*
> to check target dependency more detail.
>
> BTW, you can also use *systemctl list-dependecies
> obmc-host-start at 0.target <obmc-host-start at 0.target>* to get all related
> services.
>
>
>
> For example for using x86-power-control:
>
> *systemctl list-dependencies obmc-host-start at 0.target
> <obmc-host-start at 0.target>*
>
>
>
> obmc-host-start at 0.target (Power ON)
>
> -> mapper-wait at -xyz-openbmc_project-Chassis-Control-Power0.service
>
> -> phosphor-reset-host-reboot-attempts at 0.service
>
> -> obmc-host-startmin at 0.target
>
>      -> mapper-wait at -xyz-openbmc_project-Chassis-Control-Power0.service
>
>      -> obmc-enable-host-watchdog at 0.service
>
>      -> phosphor-watchdog at poweron.service
>
>      -> obmc-chassis-poweron at 0.target
>
>           -> intel-power-start at 0.service
>
>           ->
> mapper-wait at -xyz-openbmc_project-Chassis-Control-Power0.service
>
>           -> phosphor-fan-control-init at 0.service
>
>           -> phosphor-fan-monitor-init at 0.service
>
>           -> phosphor-fan-presence-tach at 0.service
>
>
>
> Then you should found that might some services doesn’t start normally and
> cause your obmc-host-tart at 0.target start fail.
>
>
>
> Best regards,
>
> Tim
>
> *From:* openbmc [
> mailto:openbmc-bounces+chli30=nuvoton.com at lists.ozlabs.org
> <openbmc-bounces+chli30=nuvoton.com at lists.ozlabs.org>] *On Behalf Of *
> zhang_cy1989
> *Sent:* Wednesday, December 25, 2019 7:19 PM
> *To:* openbmc at lists.ozlabs.org
> *Subject:* systemctl start obmc-host-start at 0.target fail
>
>
>
>
>
> Dear All
>
>    On ast2500+x86 platform, use the romulus bmc firmware, the detail
> information:
>
> *ID="openbmc-phosphor"*
> *NAME="Phosphor OpenBMC (Phosphor OpenBMC Project Reference Distro)"*
> *VERSION="2.7.0-0"*
> *VERSION_ID="2.7.0-0-gcb91a7773-dirty"*
> *PRETTY_NAME="Phosphor OpenBMC (Phosphor OpenBMC Project Reference Distro)
> 2.7.0-0"*
> *BUILD_ID="2.7.0"*
> *OPENBMC_TARGET_MACHINE="romulus"*
>
>
>
>   The BMC ip: 192.168.0.1
>
>   Remote Computer ip: 192.168.0.2
>
>
>
>   I use the ipmitool to communicate with BMC:
>
> ipmitool -I lanplus -H 192.168.0.1 -U root -P 0penBmc power on
>
>
>
>  And from the journalctl results, I found those infromations:
>
> ......
>
> -- The unit op-wait-power-on at 0.service has entered the 'failed' state
> with result 'timeout'.
> Nov 28 07:55:47 romulus systemd[1]: Failed to start Wait for Power0 to
> turn on.
> -- Subject: A start job for unit op-wait-power-on at 0.service has failed
> -- Defined-By: systemd
> -- Support: https://lists.freedesktop.org/mailman/listinfo/systemd-devel
> <https://urldefense.proofpoint.com/v2/url?u=https-3A__lists.freedesktop.org_mailman_listinfo_systemd-2Ddevel&d=DwMGbg&c=ue8mO8zgC4VZ4q_aNVKt8G9MC01UFDmisvMR1k-EoDM&r=_vvS1tH9U_15yp1--_mdcWxVamV74LY4Xz5Zy4-7UIo&m=EhLiDsjPMBS98eeZEQG2UAHMPrvbvGb_OublwmQy2uc&s=UO9vddeGuzrgEFSgZkdgqCUEsjWHcXBofqYYlHxCNkM&e=>
> --
> -- A start job for unit op-wait-power-on at 0.service has finished with a
> failure.
> --
> -- The job identifier is 906 and the job result is failed.
> Nov 28 07:55:47 romulus systemd[1]: Dependency failed for Chassis0 (Power
> On).
> -- Subject: A start job for unit obmc-chassis-poweron at 0.target has failed
> -- Defined-By: systemd
> -- Support: https://lists.freedesktop.org/mailman/listinfo/systemd-devel
> <https://urldefense.proofpoint.com/v2/url?u=https-3A__lists.freedesktop.org_mailman_listinfo_systemd-2Ddevel&d=DwMGbg&c=ue8mO8zgC4VZ4q_aNVKt8G9MC01UFDmisvMR1k-EoDM&r=_vvS1tH9U_15yp1--_mdcWxVamV74LY4Xz5Zy4-7UIo&m=EhLiDsjPMBS98eeZEQG2UAHMPrvbvGb_OublwmQy2uc&s=UO9vddeGuzrgEFSgZkdgqCUEsjWHcXBofqYYlHxCNkM&e=>
> --
> -- A start job for unit obmc-chassis-poweron at 0.target has finished with a
> failure.
> --
> -- The job identifier is 886 and the job result is dependency.
> Nov 28 07:55:47 romulus systemd[1]: Dependency failed for Start Host0
> Minimum.
> -- Subject: A start job for unit obmc-host-startmin at 0.target has failed
> -- Defined-By: systemd
> -- Support: https://lists.freedesktop.org/mailman/listinfo/systemd-devel
> <https://urldefense.proofpoint.com/v2/url?u=https-3A__lists.freedesktop.org_mailman_listinfo_systemd-2Ddevel&d=DwMGbg&c=ue8mO8zgC4VZ4q_aNVKt8G9MC01UFDmisvMR1k-EoDM&r=_vvS1tH9U_15yp1--_mdcWxVamV74LY4Xz5Zy4-7UIo&m=EhLiDsjPMBS98eeZEQG2UAHMPrvbvGb_OublwmQy2uc&s=UO9vddeGuzrgEFSgZkdgqCUEsjWHcXBofqYYlHxCNkM&e=>
> --
> -- A start job for unit obmc-host-startmin at 0.target has finished with a
> failure.
> --
> -- The job identifier is 752 and the job result is dependency.
> Nov 28 07:55:47 romulus systemd[1]: Dependency failed for Start Host0.
> -- Subject: A start job for unit obmc-host-start at 0.target has failed
> -- Defined-By: systemd
> -- Support: https://lists.freedesktop.org/mailman/listinfo/systemd-devel
> <https://urldefense.proofpoint.com/v2/url?u=https-3A__lists.freedesktop.org_mailman_listinfo_systemd-2Ddevel&d=DwMGbg&c=ue8mO8zgC4VZ4q_aNVKt8G9MC01UFDmisvMR1k-EoDM&r=_vvS1tH9U_15yp1--_mdcWxVamV74LY4Xz5Zy4-7UIo&m=EhLiDsjPMBS98eeZEQG2UAHMPrvbvGb_OublwmQy2uc&s=UO9vddeGuzrgEFSgZkdgqCUEsjWHcXBofqYYlHxCNkM&e=>
> --
> -- A start job for unit obmc-host-start at 0.target has finished with a
> failure.
>
> Nov 28 08:24:09 romulus systemd[1]: obmc-host-start at 0.target: Job
> obmc-host-start at 0.target/start failed with result 'dependency'.
> Nov 28 08:24:09 romulus systemd[1]: obmc-host-start at 0.target: Triggering
> OnFailure= dependencies.
> Nov 28 08:24:10 romulus systemd[1]: obmc-host-startmin at 0.target: Job
> obmc-host-startmin at 0.target/start failed with result 'dependency'.
> Nov 28 08:24:10 romulus systemd[1]: obmc-host-startmin at 0.target:
> Triggering OnFailure= dependencies.
> Nov 28 08:24:10 romulus systemd[1]: obmc-chassis-poweron at 0.target: Job
> obmc-chassis-poweron at 0.target/start failed with result 'dependency'.
> Nov 28 08:24:10 romulus systemd[1]: obmc-chassis-poweron at 0.target:
> Triggering OnFailure= dependencies.
>
> ......
>
>
>
> I alse do the following action:
>
> *systemctl start obmc-host-start at 0.target <obmc-host-start at 0.target> *
>
> And alse get the similar fail result.
>
>
>
> I don't how to debug this isuue.Could you give me some suggestions?
>
>
>
> Best Regards
>
> Felixzhang
>
>
>
>
>
>
>
>
>
>
>
>
> ------------------------------
>
> The privileged confidential information contained in this email is
> intended for use only by the addressees as indicated by the original sender
> of this email. If you are not the addressee indicated in this email or are
> not responsible for delivery of the email to such a person, please kindly
> reply to the sender indicating this fact and delete all copies of it from
> your computer and network server immediately. Your cooperation is highly
> appreciated. It is advised that any unauthorized use of confidential
> information of Nuvoton is strictly prohibited; and any information in this
> email irrelevant to the official business of Nuvoton shall be deemed as
> neither given nor endorsed by Nuvoton.
>
>
>
>
> ------------------------------
> The privileged confidential information contained in this email is
> intended for use only by the addressees as indicated by the original sender
> of this email. If you are not the addressee indicated in this email or are
> not responsible for delivery of the email to such a person, please kindly
> reply to the sender indicating this fact and delete all copies of it from
> your computer and network server immediately. Your cooperation is highly
> appreciated. It is advised that any unauthorized use of confidential
> information of Nuvoton is strictly prohibited; and any information in this
> email irrelevant to the official business of Nuvoton shall be deemed as
> neither given nor endorsed by Nuvoton.
>
>
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.ozlabs.org/pipermail/openbmc/attachments/20200106/e64b9aab/attachment-0001.htm>

More information about the openbmc mailing list