Fix some cppcheck warnings

Fix some of the warnings in the base phosphor-logging code.

It doesn't fix all of them - some are false positives and some are just
suggestions, but it does bring down th

Fix some cppcheck warnings

Fix some of the warnings in the base phosphor-logging code.

It doesn't fix all of them - some are false positives and some are just
suggestions, but it does bring down the number a bit to make it easier
to spot real issues.

Signed-off-by: Matt Spinler <spinler@us.ibm.com>
Change-Id: I8a131cbc2941ccba79ef3775151007c770207bd5

show more ...

sdbusplus: use shorter type aliases

The sdbusplus headers provide shortened aliases for many types.
Switch to using them to provide better code clarity and shorter
lines. Possible replacements are

sdbusplus: use shorter type aliases

The sdbusplus headers provide shortened aliases for many types.
Switch to using them to provide better code clarity and shorter
lines. Possible replacements are for:
* bus_t
* exception_t
* manager_t
* match_t
* message_t
* object_t
* slot_t

Signed-off-by: Patrick Williams <patrick@stwcx.xyz>
Change-Id: Ibc25db433a6926f7ee43ea83312c3ac14f480c33

show more ...

PEL: Add BMC uptime to PELs in UserData section

A UserData section has been added to each PEL with additional debug
information, now there is a need to add the output of the uptime
command to UserDa

PEL: Add BMC uptime to PELs in UserData section

A UserData section has been added to each PEL with additional debug
information, now there is a need to add the output of the uptime
command to UserData and display it, but for Hostboot doesn't care
about this property, so skip adding it here it.

Tested: unit test passed
"User Data 0": {
"Section Version": "1",
"Sub-section type": "1",
"Created by": "0x2000",
...
"Uptime": "3y 332d 21h 33m 9s",
"Load": "1.47 0.94 0.61",
},

Signed-off-by: George Liu <liuxiwei@inspur.com>
Change-Id: I3d4c78bb1650da9a91804fc83de60597992ffc8a

show more ...

PEL: Update D-Bus event sev based on PEL severity

Put in checks to ensure the D-Bus event log severity agrees with the
final PEL severity for PELs created by the BMC. The D-bus property is
what is

PEL: Update D-Bus event sev based on PEL severity

Put in checks to ensure the D-Bus event log severity agrees with the
final PEL severity for PELs created by the BMC. The D-bus property is
what is used in the Redfish event log, and we want to avoid cases like
having a Critical event log for an informational PEL.

This could happen in the case where the PEL message registry entry for
the event has a hardcoded or manufacturing specific severity value that
is different than the severity the D-Bus event log is first created
with.

This commit ensures that:
* If the PEL is nonError/recovered, the D-Bus severity is one of the
non error ones.
* If the PEL isn't nonError/recovered, then the D-Bus severity also
isn't.
* If the PEL is critical, the D-Bus severity is also critical.

It doesn't try to update the D-Bus severity for every PEL severity
because there isn't a one to one mapping - e.g. Notice, Debug, and
Informational D-Bus severities all map to a PEL severity of
nonError(informational).

Signed-off-by: Matt Spinler <spinler@us.ibm.com>
Change-Id: I6b0006034090e6d8e33e9f8b136ae50cce489f6e

show more ...

PEL: Sanitize D-Bus fields that come from PELs

The Resolution D-Bus property is filled with values from the PEL callout
fields like the serial and part numbers. These fields usually come from
EEPRO

PEL: Sanitize D-Bus fields that come from PELs

The Resolution D-Bus property is filled with values from the PEL callout
fields like the serial and part numbers. These fields usually come from
EEPROMs, and may not necessarily be valid printable values if the code
that creates the PEL doesn't sanitize the values first. If any of the
characters were invalid, the daemon would crash when D-Bus broker
disconnects it from D-Bus since it doesn't allow invalid values on
D-Bus.

Prevent this crash by converting any non printable characters in the
Resolution property to spaces. This also sanitizes the EventId property
which contains the SRC ASCII string and hex words just to be safe.

This is really just a concern for non-BMC created PELs, since the BMC
code that reads EEPROMs already sanitizes the values when it reads them,
but PELs have come in from other subsystems that don't.

Signed-off-by: Matt Spinler <spinler@us.ibm.com>
Change-Id: If0e80fd9db27446f5367ed046e4bca2eb62e3fb2

show more ...

PEL: Allow pretty symFRU/proc name in callout JSON

When callouts are passed in via JSON FFDC files, allow maintenance
procedures and symbolic FRUs to be the pretty names used in the
message registry

PEL: Allow pretty symFRU/proc name in callout JSON

When callouts are passed in via JSON FFDC files, allow maintenance
procedures and symbolic FRUs to be the pretty names used in the
message registry as well as the raw 7 character names.

Do this by attempting to look up the name as a key in the pretty name ->
raw name maps, and if found then use the raw name.

Signed-off-by: Matt Spinler <spinler@us.ibm.com>
Change-Id: I576e56932dd101e32caf495c65de6fed65de43e7

show more ...

rsyslog-config: fix disabled config

When no remote syslog is used action ~ (drop) isn't appropriate as it
inhibits processing of the current message for any additional configs
that might come after

rsyslog-config: fix disabled config

When no remote syslog is used action ~ (drop) isn't appropriate as it
inhibits processing of the current message for any additional configs
that might come after server.conf. This file shouldn't be empty either
as there're no other output actions defined in default configuration.

Make it a no-op by sending to /dev/null instead. Add a linefeed at the
end so that "cat server.conf" would look nicer.

Signed-off-by: Paul Fertser <fercerpav@gmail.com>
Change-Id: I6636c76cf0304b3c36f579225846ab19e992c333

show more ...

PEL: Remove PowerFault from message entries, registries, and code

The PowerFault flag is deprecated and should be removed in both the
registry and code references. Its functionality is superseded by

PEL: Remove PowerFault from message entries, registries, and code

The PowerFault flag is deprecated and should be removed in both the
registry and code references. Its functionality is superseded by the
critical_system severity level.

Signed-off-by: Mike Capps <mikepcapps@gmail.com>
Change-Id: I1bc1898d9589e0678b5a37b87289c6588859f2ee

show more ...

PEL: Support 'PossibleSubsystems' in msg registry

Add a 'PossibleSubsystems' field to the message registry schema where
someone can list which subsystems they might pass into the PEL_SUBSYSTEM
Addit

PEL: Support 'PossibleSubsystems' in msg registry

Add a 'PossibleSubsystems' field to the message registry schema where
someone can list which subsystems they might pass into the PEL_SUBSYSTEM
AdditionalData property field when creating a PEL.

The PossibleSubsystems value will be used by a off-BMC script that
generates field documentation for SRCs, so it knows what all SRC ASCII
string values are possible for that entry since the subsystem is part of
them. For example, if that field was ['processor', 'memory'], then that
means the possible SRCs are BD10AAAA and BD20AAAA for reason code
'AAAA'.

The Subsystem and PossibleSubsystems fields are mutually exclusive in
the message registry, so code changes had to be made to support the
subsystem field now being optional.

There is now a chance that even though the subsystem is supposed to be
passed in using PEL_SUBSYSTEM, it isn't. In that case, the 'Other'
subsystem will be used, since it seems fairly generic.

Signed-off-by: Matt Spinler <spinler@us.ibm.com>
Change-Id: Ic8087fc46b2a192459b7287186f8972610e84730

show more ...

PEL: Skip purging PEL if it has associated hw isolation entry

When PELs reaches to certain limits, either in total size or
total number of them, the code starts removing older pels
based on certain

PEL: Skip purging PEL if it has associated hw isolation entry

When PELs reaches to certain limits, either in total size or
total number of them, the code starts removing older pels
based on certain criteria. During this process make sure
no pel is removed that has an associated hardware isolation
entry record or error_log event record.

Signed-off-by: Sumit Kumar <sumit_kumar@in.ibm.com>
Change-Id: Ia7c9af6911bcf2ac8f8ec046d4c7a7a12d9b2bbd

show more ...

PEL: Wait a bit after host up before sending PELs

Wait a minute after the host state switches to running before sending
any PELs to the host. This applies both on a normal boot and also when
the BM

PEL: Wait a bit after host up before sending PELs

Wait a minute after the host state switches to running before sending
any PELs to the host. This applies both on a normal boot and also when
the BMC is rebooted when the host is already up.

This is being done for 2 reasons:
1) Give some relief to the PLDM traffic and CPU load that occurs after a
BMC reboot when the host is already up, because the host will also be
exchanging PLDM PDRs with the BMC then and pretty much every
application is also starting up at the time as well.
2) A lot of times on a normal boot the host notifier code ends up having
to do a few retries when it first starts sending the PELs because the
first few commands fail, possibly because the code on the receiving
end isn't quite ready yet.

I arrived at that amount of time by deciding 30 seconds was reasonable,
and then doubling it just to be sure. During testing of the reboots the
minute timeout gave plenty of time for the BMC to get back to BMC ready.

Signed-off-by: Matt Spinler <spinler@us.ibm.com>
Change-Id: I0929a4acdda0a9fe8fbdcd624c5821a642ad83a6

show more ...

Pull file path into elog Entry constructor

The path property on the FilePath interfaces was previously being set
after the entry object was created, sending a properties changed signal
after the int

Pull file path into elog Entry constructor

The path property on the FilePath interfaces was previously being set
after the entry object was created, sending a properties changed signal
after the interfaces added signal.

Since the path can be known at the time the entry is constructed, pass
it into the constructor instead so no extra signal will be sent.

Change-Id: I1150236d23cecb1df78e3fb4ae641c75b488af37

show more ...

openpower-pels: only fail on hw callouts

After further discussion with the IBM service and manufacturing team, it
was determined that firmware should only be automatically going to
Quiesce and preve

openpower-pels: only fail on hw callouts

After further discussion with the IBM service and manufacturing team, it
was determined that firmware should only be automatically going to
Quiesce and preventing the system boot in QuiesceOnError mode when the
PEL has a hardware callout present (vs. any type of callout).

Tested:
- New unit test case
- Verified that an unrecoverable PEL with only a symbolic FRU and a
maintenance procedure callout did not trigger the Quiesce logic
- Verified that a BMC log with a HW callout still triggers the Quiesce
logic
- Verified a boot in simulation with Quiesce enabled

Signed-off-by: Andrew Geissler <geissonator@yahoo.com>
Change-Id: I8912d59fd69b699c7da5a44da887cce4e987f6d2

show more ...

PEL: Pass in subsystem to use with additional data

There is a requirement to pass in the subsystem to use with
the additional data field. In the process update the user
header code and SRC with the

PEL: Pass in subsystem to use with additional data

There is a requirement to pass in the subsystem to use with
the additional data field. In the process update the user
header code and SRC with the passed-in subsystem value
instead from the message registry.

Signed-off-by: Sumit Kumar <sumit_kumar@in.ibm.com>
Change-Id: I08c8c3bab100fec581df9043f4bb27b28cb556e1

show more ...

PEL: Fix for error - unable to read PEL ID file

This is a case when PEL daemon can't read pelID file to get the next
pel Id to use. It could be that the file got corrupted during code
update or fact

PEL: Fix for error - unable to read PEL ID file

This is a case when PEL daemon can't read pelID file to get the next
pel Id to use. It could be that the file got corrupted during code
update or factory reset leading to zeros data in file. The fix is to
delete the file as there is already code later to recreate file if
not available.

Signed-off-by: Sumit Kumar <sumit_kumar@in.ibm.com>
Change-Id: I42372cc1d79ece5914e7fdb51f3fa2a2d9cb07e7

show more ...

PEL: Write terminating SRC to the progress SRC

Check for incoming pels for severity type 0x51 - Critical error,
System termination. If found, fetch the SRC and add this SRC to
progress SRC interface

PEL: Write terminating SRC to the progress SRC

Check for incoming pels for severity type 0x51 - Critical error,
System termination. If found, fetch the SRC and add this SRC to
progress SRC interface on dbus. In addition set the terminate bit
in BMC created pels only.

Signed-off-by: Sumit Kumar <sumit_kumar@in.ibm.com>
Change-Id: I26194a26743263183dcb61e097c745c4705fa006

show more ...

PEL: pad hex words displayed in error message

In the message registry under "Documentation" and "Message", users can
include the hex words as part of the error message. This change pads
the displaye

PEL: pad hex words displayed in error message

In the message registry under "Documentation" and "Message", users can
include the hex words as part of the error message. This change pads
the displayed hex words so that the full eight digit hex number is
displayed.

Signed-off-by: Zane Shelley <zshelle@us.ibm.com>
Change-Id: I32c4204a1643b585dc1309f67976f4abb44fb04c

show more ...

PEL: Add guard record entry create function

Added D-Bus wrapper function to create guard record
entry D-Bus object.

hardware isolation API's used no no-reply mode to avoid
cyclic dependency with lo

PEL: Add guard record entry create function

Added D-Bus wrapper function to create guard record
entry D-Bus object.

hardware isolation API's used no no-reply mode to avoid
cyclic dependency with log manager. PEL log already have
the details of guard records properties. For any failure
during record creation is hardware isolation application
is ignored by log manager. User will get details on this
failure by looking hardware isolation log information.

Tested: verified D-Bus create method.

Signed-off-by: Jayanth Othayoth <ojayanth@in.ibm.com>
Change-Id: I8c7703dc28aaf7b39793e3bd1dfe4fac374d9ad5

show more ...

catch exceptions as const

Signed-off-by: Patrick Williams <patrick@stwcx.xyz>
Change-Id: Ic8e6ade739bd5ea6e79cac6b9bb2b182748e10c8

PEL: Update pel_values to support bmcweb

Updates done to support display of Message property in bmcweb.
With the redfish command the message property showed up as
--old-
"Message": "xyz.openbmc_proj

PEL: Update pel_values to support bmcweb

Updates done to support display of Message property in bmcweb.
With the redfish command the message property showed up as
--old-
"Message": "xyz.openbmc_project.Power.PowerSupply.Error.InputFault"
--New--
"Message": "110015F0 event in subsystem: Power Supply",

Note that the "Power Supply" is a $subsystem property and had ':' in the
description, to enhance readability the $subsystem for CEC Hardware is
update with '-' in place of ':'.

Change-Id: Ic6af0937f4e560753814462692b95185a6d84d07
Signed-off-by: Vijay Lobo <vijaylobo@gmail.com>

show more ...

PEL: Store system boot state in user data section

For every incoming pel data interface watches the dbus property
that includes the latest boot progress/state value. This value would be
captured and

PEL: Store system boot state in user data section

For every incoming pel data interface watches the dbus property
that includes the latest boot progress/state value. This value would be
captured and stored in user data section of pel.

Signed-off-by: Sumit Kumar <sumit_kumar@in.ibm.com>
Change-Id: Ie48e7ee9fdd965af672f4bd563f8c9fc3bfacc43

show more ...

PEL: Update bmc & platform dump status in SRC section

With every PEL creation, the status of bmc and platform dumps -
Hardware and Hypervisor would be checked and status bits in SRC
section updated

PEL: Update bmc & platform dump status in SRC section

With every PEL creation, the status of bmc and platform dumps -
Hardware and Hypervisor would be checked and status bits in SRC
section updated accordingly.

Change-Id: I3ec7626611cf330f2ce235a97ee3046c0d32b6ab
Signed-off-by: Sumit Kumar <sumit_kumar@in.ibm.com>

show more ...

PEL: Change BMCSP01 maint procedure to BMC0001

This name change allows 4 characters of uniqueness instead of just 2.

Signed-off-by: Matt Spinler <spinler@us.ibm.com>
Change-Id: I12cb883619a3a3c2f84

PEL: Change BMCSP01 maint procedure to BMC0001

This name change allows 4 characters of uniqueness instead of just 2.

Signed-off-by: Matt Spinler <spinler@us.ibm.com>
Change-Id: I12cb883619a3a3c2f843dd7ccc3db6524dea3459

show more ...

PEL: Remove 'no_vpd_for_fru' maintenance procedure

This procedure was being used in a callout when the location code could
not be found on an inventory item. The code was changed to just not add
a

PEL: Remove 'no_vpd_for_fru' maintenance procedure

This procedure was being used in a callout when the location code could
not be found on an inventory item. The code was changed to just not add
a callout in this case.

This was done for the following reasons:
1) There's no expected reason that the inventory would be missing a
location code for a valid inventory path.
2) There wasn't a way to give a hint to the end user about what the
callout should be in that case. Usually maintenance procedures should
have steps one can take to still do the appropriate replacement.
3) Because of 1), this case indicates a bad inventory path passed in by
the caller. Theoretically there is a way to have a 'bmc_code'
callout, but that wouldn't be appropriate to add in this log since
the intent of this one is for another problem.
3a) There is no way to create a new error log from inside a PEL section,
and I don't think this one case warrants coming up with a way to do
so.
4) The inventory path in question is still being added into a UserData
section so that development can debug it.

Signed-off-by: Matt Spinler <spinler@us.ibm.com>
Change-Id: I0757080f04942807cc34029d8667387db6b959fe

show more ...

PEL: Guard against hostboot sending down duplicate PEL Ids

This commit is to guard against hostboot sending down PEL Id
that we already have in our repository. This caused PEL to
get orphaned in the

PEL: Guard against hostboot sending down duplicate PEL Ids

This commit is to guard against hostboot sending down PEL Id
that we already have in our repository. This caused PEL to
get orphaned in the filesystem without a corresponding valid
openbmc event log Id.
The action is to move such PELs to archive folder.

Signed-off-by: Sumit Kumar <sumit_kumar@in.ibm.com>
Change-Id: I46bb865b4b87ec0b59f362e3f79a1b5a6710a45c

show more ...