Fix up the TODO for the deglitching

The TODO for deglitching faults should have referenced a GitHub issue.
The deglitching support will be added with openbmc/openbmc#2484.

Change-Id: Ia963ff6598855

Fix up the TODO for the deglitching

The TODO for deglitching faults should have referenced a GitHub issue.
The deglitching support will be added with openbmc/openbmc#2484.

Change-Id: Ia963ff6598855faeda5d5db39e6ce066ffcb56c0
Signed-off-by: Brandon Wyman <bjwyman@gmail.com>

show more ...

Add in a delay before setting present to true

When a power supply is inserted, allow for a time delay before starting
the fault checks. This will give the individual servicing a power supply
some ti

Add in a delay before setting present to true

When a power supply is inserted, allow for a time delay before starting
the fault checks. This will give the individual servicing a power supply
some time to insert the supply and then apply power.

Change-Id: Ic957be927cea26a6c011a0a634f84093e040b454
Signed-off-by: Brandon Wyman <bjwyman@gmail.com>

show more ...

Delay power supply on fault checks

After the input power is restored to the power supply, there will be a
delay before the 12V is within spec. This will delay the activation of
the PGOOD indicator.

Delay power supply on fault checks

After the input power is restored to the power supply, there will be a
delay before the 12V is within spec. This will delay the activation of
the PGOOD indicator. When the inputFault is cleared, set the powered on
state to false, and start up the timer that will set it to true after
the adequate time has passed.

Change-Id: I9fe6224df6e01058700f9dec4fd3666839f9f595
Signed-off-by: Brandon Wyman <bjwyman@gmail.com>

show more ...

Merge VIN_UV_FAULT check with the inputFault

Since VIN_UV_FAULT is for input voltage, and the desired behavior was to
"treat as OR condition if either asserts", this change is removing the
variable

Merge VIN_UV_FAULT check with the inputFault

Since VIN_UV_FAULT is for input voltage, and the desired behavior was to
"treat as OR condition if either asserts", this change is removing the
variable used to track VIN_UV_FAULT, and adding the condition check to
the path that sets inputFault.

Change-Id: Id4451753f6d627a62588b0d2211614174f83abb0
Signed-off-by: Brandon Wyman <bjwyman@gmail.com>

show more ...

Change to not check powerOn faults if input fault

If there was a VIN_UV_FAULT or INPUT_FAULT, do not bother to check for
the fan fault, temperature fault, output over-voltage fault, output over
curr

Change to not check powerOn faults if input fault

If there was a VIN_UV_FAULT or INPUT_FAULT, do not bother to check for
the fan fault, temperature fault, output over-voltage fault, output over
current fault, PGOOD, or UNIT_IS_OFF faults. The fans will not be
running, there will be no output voltage or current (aside from Vcs),
the PGOOD will not be active, and the unit will be off. Logging those
additional errors will just create confusion.

Change-Id: Ia9f02509a057300e4992da3a6edd90edf025e766
Signed-off-by: Brandon Wyman <bjwyman@gmail.com>

show more ...

Add power supply call out to UV and input faults

Add the inventory string to the call out for the VIN_UV_FAULT and
INPUT_FAULT conditions. This should result in the enclosure LED turning
on as an in

Add power supply call out to UV and input faults

Add the inventory string to the call out for the VIN_UV_FAULT and
INPUT_FAULT conditions. This should result in the enclosure LED turning
on as an indicator that the system requires some service attention.

Change-Id: I1d00086bf1809a424c81e77b1f22c24204e15198
Signed-off-by: Brandon Wyman <bjwyman@gmail.com>

show more ...

Update clearFaults and use for power and presence

If the power supply changes to the powered on state, or changes to
present, call the clearFaults() function instead of individually
clearing the var

Update clearFaults and use for power and presence

If the power supply changes to the powered on state, or changes to
present, call the clearFaults() function instead of individually
clearing the variables.

Ensure powerOnFault is in the clearFaults() list.

Change-Id: I02ca222c4a5ddc8041c6796f0aba23bbcf30ff8f
Signed-off-by: Brandon Wyman <bjwyman@gmail.com>

show more ...

Implement de-glitch for PGOOD and UNIT_IS_OFF

De-glitch the powerOnFault so that a fault is logged only if the
POWER_GOOD Negated or UNIT_IS_OFF bits are in the wrong state three
times in a row.

Re

Implement de-glitch for PGOOD and UNIT_IS_OFF

De-glitch the powerOnFault so that a fault is logged only if the
POWER_GOOD Negated or UNIT_IS_OFF bits are in the wrong state three
times in a row.

Resolves openbmc/openbmc#2473

Change-Id: Icbb45010f921bd3e11018f5e7bdc2d155242d464
Signed-off-by: Brandon Wyman <bjwyman@gmail.com>

show more ...

Remove try/catch defaulting PSU presence to false

During an earlier commit for openbmc/openbmc#1731, a change was made to
default the power supply presence to false if the inventory manager did
not

Remove try/catch defaulting PSU presence to false

During an earlier commit for openbmc/openbmc#1731, a change was made to
default the power supply presence to false if the inventory manager did
not yet have a present property available for a power supply instance.
This should have been, and eventually was, better handled via the
service files used for starting the application.

https://gerrit.openbmc-project.xyz/#/c/6018/15..25/

Change-Id: Idcde1d9c405465606dbffd23372122e893caf3d2
Signed-off-by: Brandon Wyman <bjwyman@gmail.com>

show more ...

Use Common.Device.Error for device failures

The ReadFailure and WriteFailure errors are now
available in Common.Device, and code will now use
those instead of the ones in Sensor.Device and
Control.D

Use Common.Device.Error for device failures

The ReadFailure and WriteFailure errors are now
available in Common.Device, and code will now use
those instead of the ones in Sensor.Device and
Control.Device, which were meant for sensor errors.

Change-Id: Id15b56ab56f3ee4acee024941840a17b9f5bf9c6
Signed-off-by: Matt Spinler <spinler@us.ibm.com>

show more ...

Move power errors to org.open_power.witherspoon

Move xyz.openbmc_project.Power to org.open_power.Witherspoon.
This repo is Witherspoon specific, so moving to specific tree to have
its own namespace.

Move power errors to org.open_power.witherspoon

Move xyz.openbmc_project.Power to org.open_power.Witherspoon.
This repo is Witherspoon specific, so moving to specific tree to have
its own namespace. Update generated file, and corresponding namespace
statements in the source files.

Change-Id: I10bd1eeeac5b5f391ca54bf361b249b7ad531852
Signed-off-by: Brandon Wyman <bjwyman@gmail.com>

show more ...

Update the delay from state=1 to powerOn=true

Increasing the power on delay timer from 1 second to 2 seconds, rounding
up from the 1250ms that should be the max from BMC_POWER_UP to DC_GOOD.

Resolv

Update the delay from state=1 to powerOn=true

Increasing the power on delay timer from 1 second to 2 seconds, rounding
up from the 1250ms that should be the max from BMC_POWER_UP to DC_GOOD.

Resolves: openbmc/openbmc#2439

Change-Id: I90f363d31514c38efec8b3594ba705aea7ccb2bc
Signed-off-by: Brandon Wyman <bjwyman@gmail.com>

show more ...

Change order of powered on fault checks

It seems to make more sense to see the checks logging errors in the
order that they lead to the power supply shutting off.

A fan fault could lead to a therma

Change order of powered on fault checks

It seems to make more sense to see the checks logging errors in the
order that they lead to the power supply shutting off.

A fan fault could lead to a thermal fault, which could lead to the power
supply shutting down when it should be on (if it did not have the fan or
thermal fault).

I am re-ordering the powered on checks to be in an order that seems to
make more sense.

Resolves: openbmc/openbmc#1732

Change-Id: I5a36678d03722a2de70d0e8ec5d9d4c04e07f943
Signed-off-by: Brandon Wyman <bjwyman@gmail.com>

show more ...

Add a helper function to capture command metadata

Due to openbmc/openbmc#2364, certain command response files may be
missing. Add a helper function to call the PMBus::exists() function to
skip tryin

Add a helper function to capture command metadata

Due to openbmc/openbmc#2364, certain command response files may be
missing. Add a helper function to call the PMBus::exists() function to
skip trying to read in metadata for missing files.

Change-Id: I101c5ccebe76dd29fb07369fd082072ce94f021b
Signed-off-by: Brandon Wyman <bjwyman@gmail.com>

show more ...

Add in support for power supply over-temperature

If a power supply encounters an over-temperature condition, the FAN
FAULT OR WARNING bit of the STATUS_WORD (low byte) command response
should be tur

Add in support for power supply over-temperature

If a power supply encounters an over-temperature condition, the FAN
FAULT OR WARNING bit of the STATUS_WORD (low byte) command response
should be turned on. Ideally, when this is encountered, both power
supplies would be called out, as one condition that could lead to this
is due to the other supply putting out less current. Since each
monitoring application of the power supplies only knows its own
information, just the power supply indicating the condition will be
called out along with metadata containing potentially relevant command
response data.

Change-Id: I8f96828f85161050f73bb080392e1e8fef4a179b
Signed-off-by: Brandon Wyman <bjwyman@gmail.com>

show more ...

Add in support for power supply fan fault

If the FAN FAULT OR WARN bit in the STATUS_WORD turns on, log an error
calling out the power supply, include the contents from STATUS_WORD,
MFR_SPECIFIC, ST

Add in support for power supply fan fault

If the FAN FAULT OR WARN bit in the STATUS_WORD turns on, log an error
calling out the power supply, include the contents from STATUS_WORD,
MFR_SPECIFIC, STATUS_TEMPERATURE, and STATUS_FANS_1_2 in the metadata.

Change-Id: Ic9261cf08517344f594a4616a91dbec47bb07d7e
Signed-off-by: Brandon Wyman <bjwyman@gmail.com>

show more ...

Add in detection of output overvoltage fault

If bit 5 of the lower byte to STATUS_WORD (STATUS_BYTE) is on when the
power is on, the power supply is indicating that it has detected an
output overvol

Add in detection of output overvoltage fault

If bit 5 of the lower byte to STATUS_WORD (STATUS_BYTE) is on when the
power is on, the power supply is indicating that it has detected an
output overvoltage fault condition. Log an error pointing to the power
supply that is indicating this overvoltage condition.

Change-Id: Ie369a30e6b82be805ae3aafd67812700ce2421f7
Signed-off-by: Brandon Wyman <bjwyman@gmail.com>

show more ...

Refactor the analyze() function to be less verbose

Split up the various fault checking sections of code to be there own
separate functions for checking the various fault bits/conditions and
logging

Refactor the analyze() function to be less verbose

Split up the various fault checking sections of code to be there own
separate functions for checking the various fault bits/conditions and
logging of appropriate error with necessary metadata for debug.

Change-Id: I65e54f53b0089d82852bee398ebb7d2303f29b6c
Signed-off-by: Brandon Wyman <bjwyman@gmail.com>

show more ...

Add in support for detecting an output overcurrent fault

When analyzing the STATUS_WORD bits, check if the IOUT_OC_FAULT bit is
on. That bit indicates that an output overcurrent fault condition has

Add in support for detecting an output overcurrent fault

When analyzing the STATUS_WORD bits, check if the IOUT_OC_FAULT bit is
on. That bit indicates that an output overcurrent fault condition has
occurred. Call out the power supply reporting that.

Change-Id: I18c7cd981ffc1b73ab52c90d32ba34195013e65e
Signed-off-by: Brandon Wyman <bjwyman@gmail.com>

show more ...

Add support for power supply off when it should be on

If the power good bit indicates false, or the UNIT_IS_OFF bit is on,
create an error log and attach STATUS_WORD, STATUS_INPUT, STATUS_VOUT,
STAT

Add support for power supply off when it should be on

If the power good bit indicates false, or the UNIT_IS_OFF bit is on,
create an error log and attach STATUS_WORD, STATUS_INPUT, STATUS_VOUT,
STATUS_IOUT, and MFR_SPECIFIC values to the metadata. The combination of
those PMBus command results should give an indication as to why the
power supply has turned off.

Change-Id: I692a8fdeac3fe208a5eb70964db7b5094cfb587c
Signed-off-by: Brandon Wyman <bjwyman@gmail.com>

show more ...

Add support for tracking system power state

Change-Id: Ie0f27f696082ff3ee60c955992f3c2e55f4b5e57
Signed-off-by: Brandon Wyman <bjwyman@gmail.com>

Update to allow missing inventory property

On startup, there is a chance that the inventory property may not be
populated.

Change-Id: Ie364a3dc5b7f19cbc587450fc54f231d1ec280f7
Signed-off-by: Brando

Update to allow missing inventory property

On startup, there is a chance that the inventory property may not be
populated.

Change-Id: Ie364a3dc5b7f19cbc587450fc54f231d1ec280f7
Signed-off-by: Brandon Wyman <bjwyman@gmail.com>

show more ...

Add in code to handle INPUT fault or warning

If the INPUT fault or warning bit in the STATUS_WORD turns on, report
a fault that includes STATUS_WORD and STATUS_INPUT values in the
metadata.

Change-

Add in code to handle INPUT fault or warning

If the INPUT fault or warning bit in the STATUS_WORD turns on, report
a fault that includes STATUS_WORD and STATUS_INPUT values in the
metadata.

Change-Id: I2c0bc187357088a667dba1248be9a1c52f9dc073
Signed-off-by: Brandon Wyman <bjwyman@gmail.com>

show more ...

Add STATUS_WORD to metadata for VIN_UV_FAULT

Change-Id: Iaa6001f7c5d0c558ad3bc01e209dc316236fea93
Signed-off-by: Brandon Wyman <bjwyman@gmail.com>

Update analyze to check VIN_UV_FAULT

The function is a pure virtual function in DeviceMonitor, add in the
implementation of that for PowerSupply. Read the file that represents
that bit from the STAT

Update analyze to check VIN_UV_FAULT

The function is a pure virtual function in DeviceMonitor, add in the
implementation of that for PowerSupply. Read the file that represents
that bit from the STATUS_WORD. If fault is on, report a fault.

Change-Id: I05a4bff997bb0c8b8b71db444e9db0e506765689
Signed-off-by: Brandon Wyman <bjwyman@gmail.com>

show more ...