OpenBMC as a project has moved away from using yocto x86 sdks for unit
testing. Instead now the preferred way to build and test code is by two
ways:

1. Leveraging the meson subproject support
2. Using openbmc docker container

Since x86 sdks needed some weird hacks in the code base like figuring
out the linker arguments ourselves & running custom bash commands from
the meson build files. Also with changing yocto code base, its a large
maintainer burden to make sure this support works.

Tested:
Code builds & unit tests are passed with both the approaches mentioned
above.

Change-Id: Iaf55e2c003ffd9ee1a295de5bdfd14d81222e94b
Signed-off-by: Manojkiran Eda <manojkiran.eda@gmail.com>

It was reported that, there is an inaccurate check in host_pdr_handler
code that could lead to an unbound access of an array resulting in pldm
crash. This commit aims at fixing it.

Resolves openbmc/pldm/#36

Change-Id: I1bba25f5b1cad55a30ce956f83b8c15b93149219
Signed-off-by: Manojkiran Eda <manojkiran.eda@gmail.com>

OpenBMC CI has migrated from using `pycodestyle` to `flake8` as part of
this[1] commit. Unlike `pycodestyle` , `flake8` does not rely on the
presence of setup.cfg file in their root path as a trigger, but it runs
on all repositories by default. Hence there is no need of having
setup.cfg file, so removing it from the repository.

[1]: https://github.com/openbmc/openbmc-build-scripts/commit/c5ad7ff440cfd94fc025efbd45a3859475b18820



Change-Id: I8e648f3230d1f2a83b620978b36bda88aba7ebc0
Signed-off-by: Manojkiran Eda <manojkiran.eda@gmail.com>

This patchset is improved version of the existing code for several
reasons:

1.The code is more concise and easier to read. It uses modern C++
features such as `std::ranges::for_each` and `std::format`, which
express the intent of the code more clearly compared to the nested
if-else statements and manual formatting.

2.The code now has fewer lines of code and avoids the need for creating
and managing an `std::ostringstream` object manually, which simplifies
the function and reduces the risk of errors.

3.This patch uses type-safe range-based for loop and `uint8_t` type
directly, whereas the existing code uses `int` for iterating over the
buffer elements, which may lead to unintentional type conversions or
loss of precision.

4.While the performance impact may not be significant for this patch,
using `std::ranges::for_each` with lambdas might allow for more
optimization opportunities by the compiler compared to the traditional
loop used in existing code.

Overall, this patchset leverages modern C++ features to achieve the same
functionality in a more concise, readable, and type-safe manner, making
it preferable.

Change-Id: I7be547ade053638cb4ca459ee795195f6f0883bf
Signed-off-by: Manojkiran Eda <manojkiran.eda@gmail.com>

Initialize "totalNumComponentUpdates" and "compUpdateCompletedCount" to
0 to avoid potential errors caused by uninitialized variables.

Change-Id: I27f57330e77f98109a09b3270a3dd6fd50e8fb52
Signed-off-by: DelphineCCChiu <delphine_cc_chiu@wiwynn.com>

cppcheck flags style warnings to indicate when we declare a variable and
never use it. Its effective dead code, this patch aims at fixing all the
[unusedVariable] warnings reported by the openbmc repo-ci.

Change-Id: Ie51176c6b2ced7251b467b2d456e9f3088637121
Signed-off-by: Manojkiran Eda <manojkiran.eda@gmail.com>

This commit fixes odd looking inconsistent spacing
while tracing the BIOS attribute updates.

Change-Id: Ib9dba3ca28e86ac64d67fe2b7ef5e3f66ce971b5
Signed-off-by: Manojkiran Eda <manojkiran.eda@gmail.com>

In the current state, pldm does create the sensor/effecter even in the
absence of the underlying dbus resource. There is a bug in the responder
code due to which we emplace an empty dbus value map into the handler
with a sensor ID. And if the remote PLDM endpoint queries for the state
of the sensor, pldm would crash since it tries to read content from the
empty dbus value map. This commit would fix the crash by adding
necessary bound checks.

Testing:
1. Unit tests passed.
2. Functionally verified the fix by giving a dummy dbus object path and
making sure that the sensor/effecter is not created & also ensured that
pldm does not crash.

Change-Id: I5fb06677f6ae1bd74f9ec741b311f59737caf79d
Signed-off-by: Manojkiran Eda <manojkiran.eda@gmail.com>

Below are the, System specific bios attribute infrastructure commits
Commit - https://gerrit.openbmc.org/c/openbmc/pldm/+/55997
       - https://gerrit.openbmc.org/c/openbmc/pldm/+/68451



Adding BIOS Attribute files for system types mentioned below
    1. Rainier-1s4u
    2. Rainier-2u
    3. Rainier-4u
    4. Everest
    5. Bonnell

Change-Id: I209dcff347cf83322857a2ac23353269f23217f3
Signed-off-by: Archana Kakani <archana.kakani@ibm.com>

This reverts commit d28f08cf.

Reason for revert : Pel severity is completely openpower-specific , we should not polluting common code with IBM specific things, hence reverting this change for now.

Change-Id: If11721b1522097460ba07d51335b7ebbe821311a
Signed-off-by: Manojkiran Eda <manojkiran.eda@gmail.com>

Declaring the const-ness of a parameter is just another form of
type safety. The benefit of const correctness is that it prevents
you from inadvertently modifying something you didn’t expect
would be modified. The commit fixes places in the code where a
parameter could be a constant.

Change-Id: Ie6f1d4b37373979a3efef8cb617294d41dec888c
Signed-off-by: Pavithra Barithaya <pavithrabarithaya07@gmail.com>

PLDM users who want to use bios attributes based on the system type
can enable this meson option.
With disabled option default bios json files are picked to build
bios attribute tables.

Added meson option "system-specific-bios-json" to add the support
for system specific bios attributes.
Below is the pre-requisite to support system specific bios
attributes:
1. Entity Manager service is active
2. Entity Manager should have Decorator.Compatible interface
   and system type in Names property under this interface
3. BIOS Attribute json files are added into the folder(Named on
   the System Type property value) and should be installed
   under /usr/share/pldm/bios/

With disabled "system-specific-bios-json" option default bios
attributes are populated.

Tested:
Poweron system-specific-bios-json enabled/disabled

Change-Id: I95a953cdb12c344d22f487b83040356a1b5fa937
Signed-off-by: Archana Kakani <archana.kakani@ibm.com>

This is the infrastructure commit to support PCIe Topology

PCIe topology refers to the physical arrangement of PCIe devices within
the system. Topology data contains links connecting PCIe device to the
PCIe Controller.  This commit adds support for PCIe Topology and Cable
information filetype. Topology information is provided by host using
FILIO mechanism.
This commit adds support for handling
1. Topology file(PLDM_FILE_TYPE_PCIE_TOPOLOGY)
2. Cable information file(PLDM_FILE_TYPE_CABLE_INFO)

Tested: Updated Unit test and tested power on.

Change-Id: Ic77f4f9921dd855c61c3929b5a458d80b1691b48
Signed-off-by: Archana Kakani <archana.kakani@ibm.com>

Entity manager service hosts the compatible interface and populates the
Names property with the chassis type. It is seen that the Names property
is updated pretty late on the way to BMC ready state. And by that time
pldmd is already creating the bios table assuming that the system type
is not detected.

To fix this behavior, rather than creating the bios table early, we need
to wait until we get the system type from the entity manager daemon.

System Type is fetched from Entity Manager Decorator.Compatible
interface [1].

PLDM registers the service name only after building the BIOS tables, to
avoids failures to bios updates from other applications

Tested: Power off/On successfully in Simulator

[1] https://github.com/openbmc/entity-manager/commit/9bac6409d4bd684f058517f41de33ba1d17e5666



Change-Id: I4d431061eaaf8842f6382c2e83807f725653e19b
Signed-off-by: Archana Kakani <archana.kakani@ibm.com>

This commit updates the pldm-softpoweroff application so that it can
properly locate the PLDM effecter and sensor PDRs for the graceful
poweroff request. It also updates the host terminus ID to the correct
value.

Tested by:

Running the application when Hostboot was IPLing, and also when PHYP
was at standby.

  root@ever39bmc:~# /tmp/pldm-softpoweroff
  Getting the response. PLDM RC = 0
  Timer started waiting for host soft off, TIMEOUT_IN_SEC = 0x1c20
  root@ever39bmc:~# echo $?
  0

Signed-off-by: Zach Clark <zach@ibm.com>
Change-Id: I7408611881abecfb5169816af032e2fb9d3a8a09

This commit introduces a config file for the soft power off
app. The config file specifies the order of remote PDRs to
be checked during the soft power off process.

Previously, the entity type was hardcoded in the codebase,
limiting scalability. With this change, the softoff app gains
flexibility for future expansions.

TESTED: By running 'obmcutil poweroff' with various remote
entities during boot-up.

Signed-off-by: Sagar Srinivas <sagar.srinivas@ibm.com>
Change-Id: Ib8d2aba6a05fc4e9da4b042fd67890e6282ed784

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

Change-Id: Id9161c24e5ed91f7194d6d536a29ecda02e8dc84
Signed-off-by: Patrick Williams <patrick@stwcx.xyz>

This commit implements functionality to retrieve the managed object of a
specific DBUS service on a particular path.  Additionally implements a
function template for inventory objects which efficiently enables
retrieval of managed object for Inventory Manager.

Tested: Added unit test cases for checking the return value.

Change-Id: Ide652f843db1623bdacebf3e269e03895bbb7f1a
Signed-off-by: Riya Dixit <riyadixitagra@gmail.com>

`pldmTransport.recvMsg` in `pldmd` will call
`pldm_transport_af_mctp_recv` in libpldm. This function allocates memory
of `requestMsg` and copy the MCTP request message data to `requestMsg`.
`pldmd` should free memory of requestMsg after using.

Tested:
1. Check the memory usage of `pldmd`. There is no memory leak.

Signed-off-by: Thu Nguyen <thu@os.amperecomputing.com>
Change-Id: Ie9270509e60790618c1a399fa3c9bf34f6f54907

This clang error is generated when the value stored to a variable
is never read in the code.

Change-Id: Ib68741b3aaf8856010d0073675abc2fdb3012712
Signed-off-by: Pavithra Barithaya <pavithra.b@ibm.com>

It seems that we are trying to convert file objects to strings which is
considered as a misfeature since it does not work reliably.

This is the warning that is seen while building pldm:

``` WARNING: Broken features used:
 * 1.3.0: {'str.format: Value other than strings, integers, bools, options,
   dictionaries and lists thereof.'}
```

Tested:

Rebuild pldm repository with the fix and it builds pldm without the
mentioned warning above.

Change-Id: If8fedf488d33de6dccc2c1b45c2bfe124fbc3384
Signed-off-by: Manojkiran Eda <manojkiran.eda@gmail.com>

Severity is part of the PEL User Header section, and is used
to specify the PEL severity. This commit adds an option to pass
severity parameter while logging PEL. This will help to classify PELs
based on their severity. The commit also maintains uniformity in
the PEL Error logged.

Tested: By creating the error logs(pel) in error scenarios.

Signed-off-by: Pavithra Barithaya <pavithra.b@ibm.com>
Change-Id: Ic084d49b8ffe9aaea2c36c9fefa95a10a9c1c3ec

This commit replaces occurrences of 'find' with
'contains' to access elements in a std::map
according to the latest C++ standard (since C++20).

However, contains() is not suitable in all situations
as in the case where we want to use the element contained
in the map using contains(), then it would do 2 searches
instead of 1 in the case where we use find() which fetches
the element from the map also.

Here are some relevant links to C++ documentation and
resources:

[1]: https://en.cppreference.com/w/cpp/container/map/find
[2]: https://en.cppreference.com/w/cpp/container/map/contains
[3]: https://wg21.link/N4861



Signed-off-by: Sagar Srinivas <sagar.srinivas@ibm.com>
Change-Id: If589e52026417c1c14a6a946e48826b749034e9a

This commit sets the PCIE config space data sent from
host with a prefix '0x' and the hex value. This change
was needed as the Redfish validator [1] expects the fields
to be in a particular Regex form otherwise it fails to
validate.

Tested: Ran a Redfish Validator against the changes.
Result:
Before this change:
ERROR - VendorId: String '5348' does not match
pattern ''^0[xX](([a-fA-F]|[0-9]){2}){2}$''
ERROR - DeviceId: String '5719' does not match
pattern ''^0[xX](([a-fA-F]|[0-9]){2}){2}$''
.....

After this change : Successfully passed

[1] https://www.dmtf.org/sites/default/files/standards/documents/


DSP0268_2022.2.pdf

Change-Id: I2b1d34532e94b84b34577cb2d1d7f5b955633bf6
Signed-off-by: Pavithra Barithaya <pavithrabarithaya07@gmail.com>

Added GetSensorReading command to get the current sensor
value and its state.

Tested results:
```
root@bmc:~# pldmtool platform getSensorReading -h
get the numeric sensor reading
Usage: pldmtool platform GetSensorReading [OPTIONS]

Options:
  -h,--help                   Print this help message and exit
  -m,--mctp_eid UINT          MCTP endpoint ID
  -v,--verbose
  -n,--retry-count UINT       Number of retry when PLDM request
message is failed
  -i,--sensor_id UINT REQUIRED
                              Sensor ID that is used to identify
and access the sensor
  -r,--rearm UINT REQUIRED    Manually re-arm EventState after
responding to this request
root@bmc:~# pldmtool platform getSensorReading -m9 -r0 -i300
{
    "sensorDataSize": "uint8",
    "sensorOperationalState": "Sensor Enabled",
    "sensorEventMessageEnable": "Sensor Events Enabled",
    "presentState": "Sensor Normal",
    "previousState": "Sensor Unknown",
    "eventState": "Sensor Normal",
    "presentReading": 38
}
```

Change-Id: I83a111eef73ee20a06c699405a9cdf1f848dd66a
Signed-off-by: Ruslan Magomedov <velochimc@mail.ru>

Previously, the sensor events generated by the PLDM stack only
included the current state of the sensor without any reference
to the previous state.

This commit introduces a new feature to address this limitation
by caching the previous states of all sensors.

The enhancement ensures that platform event messages(sensor events)
now contain the correct previous state information alongside the
current state.

Behavior Model:
1. Upon the initial occurrence of a sensor event, both the current
   state and the previous state are identical, adhering to the
   specification.
2. Subsequently, when the PLDM stack sends out sensor events, it
   retrieves the previous state from its cache and populates it
   accordingly.

Testing:
1. Example: Change the value of a sensor, such as dimm8 identify, to
   false using busctl.
   Tx: 81 02 0a 01 00 00 4d 00 01 00 [ 01 01 ]
- Note: Initially, both the event state and the previous state are 01.

2. Change the value of dimm8 identify to true.
   Tx: 81 02 0a 01 00 00 4d 00 01 00 [ 02 01 ]
- Result: Event state: 02, Previous state: 01.

3. Change the value of dimm8 identify back to false.
   Tx: 81 02 0a 01 00 00 4d 00 01 00 [ 01 02 ]
- Result: Event state: 01, Previous state: 02.

4. Similar results can be obtained using the getStateSensorReadings
   command for the same sensor at respective times.

Change-Id: Ic93a55e61fd5128cecc698b3555a6639876882ed
Signed-off-by: Manojkiran Eda <manojkiran.eda@gmail.com>

I have been actively contributing to OpenBMC/PLDM for quite
some time now. Attached is a summary of my contributions to
the repository [1].

I would like to learn and help out on pldm reviews.

[1]: https://github.com/openbmc/pldm/commits?author=Pavithrab7



Change-Id: I374c55f7e01a7648010c2964b3cd60d5deb0ef57
Signed-off-by: Pavithra Barithaya <pavithrabarithaya07@gmail.com>

This commit adds a common function to set the present property
to true for all the present FRUs.

Signed-off-by: Pavithra Barithaya <pavithra.b@ibm.com>
Change-Id: Ia218f0de7883ff399e9dec60f03d88c27afed3e9

An “Update Access Key” (UAK) is checked when applying system
firmware updates. Entitled updates are applied if the embedded
expiration date of the UAK is later than the firmware release date.

Host provides us the Firmware User Access Key (FW UAK) Date to BMC in
the genesis boot as it will have license information and only Host
has the knowledge of it. Also the BMC can use this value even when
host is not running.

The commit implements the support for access key
update in PLDM. The host sends a setFruRecordTable command with
the value to be updated.
PLDM uses the WriteKeyword method [1] since that will update both the
D-Bus cache and the VPD backplane.

[1] https://github.com/openbmc/phosphor-dbus-interfaces/blob/master/yaml/com/ibm/VPD/Manager.interface.yaml



Signed-off-by: Pavithra Barithaya <pavithra.b@ibm.com>
Change-Id: Ib089787db3a3c42eedb833852dc48b8fa87be7e4

The major difference between the standard FRUs
and the IBM specific FRUs is the format in which
the VPD data is written and its model type.
If the model property is filled it is a IBM specific FRU,
if not then that can be an industry standard FRU.

Change-Id: I9c01b99e37107bc3caeb88336626bc9e514e78f0
Signed-off-by: Sampa Misra <sampmisr@in.ibm.com>

In its present condition, the instance database gets initialized
within the hostSoftOff function. Consequently, once the function
execution is complete, the database goes out of scope, triggering
the invocation of the instanceDb destructor. This, in turn,
prematurely destroys the instance database in the middle of the
softpoweroff flow, resulting in a double free corruption.

Tested:
After implementing the fix, softoff was tested to ensure it does
not crash.

Change-Id: I251201b06864a8a5273bfaa23468a06ad678763a
Signed-off-by: Manojkiran Eda <manojkiran.eda@gmail.com>

Infrastructure commit-
https://gerrit.openbmc.org/c/openbmc/pldm/+/69548



This commit would add similar change for oem sensor json too.

TESTED: with pldmtool raw commands sending sensor events to the BMC.

Signed-off-by: Sagar Srinivas <sagar.srinivas@ibm.com>
Change-Id: I82bc849195fa3b65d6afb353c852aada163bba46

I appreciate the contributions made by @Sampa to the OpenBMC
project. However, I haven't seen any active involvement from her
since 2021. While it's understandable that priorities may shift,
I want to express gratitude for her past contributions.

Thank you Sampa, for your past efforts.

Change-Id: I91c6ba7553753dfbc0c31b42b1553f3942bc3ffa
Signed-off-by: Manojkiran Eda <manojkiran.eda@gmail.com>

This update resolves a crash occurring in pldm when attempting
to retrieve the system type from the entity manager service. The
issue is resolved by implementing exception handling for potential
errors during the dbus call.

Testing:
Tested in simulator and pldm is working fine without crash.

Change-Id: I6dae2c6b5a5001b801f02d359a48acb497ad7323
Signed-off-by: Kamalkumar Patel <kamalkumar.patel@ibm.com>

With the adoption of LG2 we do not need iostream
prints to the journal. The library was referred but
never used in these cases.

Signed-off-by: Sagar Srinivas <sagar.srinivas@ibm.com>
Change-Id: I11d3f00284e4f16b63cddb8da2fed14d59672ba1

In the current state, PLDM does not support having a state set ID
field in the event JSONs, this commit would add that support in
PLDM there by allowing hosts to have multiple sensors with the
same entity instance number, type and container id, but with different
state sets.

TESTED: with pldmtool raw commands sending sensor events to the BMC.

Signed-off-by: Sagar Srinivas <sagar.srinivas@ibm.com>
Change-Id: If9bb9bde27d6b35f5904ac199801dbdffde0f57e

In the current state, we attempt to create an effecter PDR even if
the D-Bus object that is being modeled as an effecter is absent at
the time of creation.

However, this behavior may not be beneficial, as assuming that the
D-Bus backend will magically appear later does not necessarily mean
that we will be able to process it, we only process the JSON files
during the first getPDR request. Therefore, it would be more efficient
to skip creating the effecter in the absence of the D-Bus object.

This commit also addresses an error trace by including the dbus object
path when it is not present on the system, rather than the opaque
effecter id number which does not provide contextual information and
may make it harder to diagnose issues.

Change-Id: I84239c75684f611b7985df027a93ba4531f783d0
Signed-off-by: Manojkiran Eda <manojkiran.eda@gmail.com>

In the current state, we attempt to create a sensor PDR even if the
D-Bus object that is being modeled as a sensor is absent at the time
of creation.

However, this behavior may not be beneficial, as assuming that the
D-Bus backend will magically appear later does not necessarily mean
that we will be able to process it and send sensor events, since we
only process the JSON files only during the first getPDR request.
Therefore, it would be more efficient to skip creating a sensor during
the absence of the D-Bus object.

This commit also addresses an error trace by including the dbus object
path when it is not present on the system, rather than the opaque sensor
id number which does not provide contextual information and may make it
harder to diagnose issues.

Change-Id: I5747319f15c212ab6654b9d7accfce28782c9565
Signed-off-by: Manojkiran Eda <manojkiran.eda@gmail.com>

The commit implements the setFruRecordTable command
and also updates the DBus property for the IBM cable cards
for which the host sends a setFruRecordTable command.

Tested: using pldmtool

Signed-off-by: Pavithra Barithaya <pavithra.b@ibm.com>
Change-Id: I70e4f85f627577d8ca1bc90447b10e9e2e8e7ccd

- There are some of the listed paths below which don't have "Present"
property. Due to this, the FRU present property check was throwing
"Resource not found" error which misguides the user.

- The commit adds support to fix this misguiding error trace.

Paths:
/xyz/openbmc_project/inventory/system/chassis/motherboard/fan0/fan0_0
/xyz/openbmc_project/inventory/system/chassis/motherboard/fan0/fan0_1
/xyz/openbmc_project/inventory/system/chassis/motherboard/fan1/fan1_0
/xyz/openbmc_project/inventory/system/chassis/motherboard/fan1/fan1_1
/xyz/openbmc_project/inventory/system/chassis/motherboard/fan2/fan2_0
/xyz/openbmc_project/inventory/system/chassis/motherboard/fan2/fan2_1
/xyz/openbmc_project/inventory/system/chassis/motherboard/fan3/fan3_0
/xyz/openbmc_project/inventory/system/chassis/motherboard/fan3/fan3_1
/xyz/openbmc_project/inventory/system/chassis/motherboard/fan4/fan4_0
/xyz/openbmc_project/inventory/system/chassis/motherboard/fan4/fan4_1
/xyz/openbmc_project/inventory/system/chassis/motherboard/fan5/fan5_0
/xyz/openbmc_project/inventory/system/chassis/motherboard/fan5/fan5_1

Change-Id: I5a8fae3a5dbae428f14ccc90aeba8319a48c11b2
Signed-off-by: Kamalkumar Patel <kamalkumar.patel@ibm.com>