pldm: Use std::expected for instance ID allocation

Refactor InstanceIdDb::next() to return
std::expected<uint8_t, InstanceIdError> instead of throwing exceptions.
With this change, errors are no lon

pldm: Use std::expected for instance ID allocation

Refactor InstanceIdDb::next() to return
std::expected<uint8_t, InstanceIdError> instead of throwing exceptions.
With this change, errors are no longer thrown by default.
Instead, the caller can inspect allocation errors as values, extract
the error message or code as needed, or explicitly throw the error
further up the stack if appropriate.
This approach allows for more flexible and explicit error handling, and
reduces reliance on exception-based control flow.

Motivation:
This change aims to improve robustness by enabling explicit error
handling and reducing the risk of uncaught exceptions that could lead
to a pldmd core dump if an instance ID allocation fails for a single
EID. Such a core dump could disrupt ongoing operations such as sensor
polling or firmware updates for other EIDs

Note:
Currently, some call sites are still void-returning functions and
cannot propagate errors via return values.
As a result, errors from InstanceIdDb::next() are thrown as
InstanceIdError at these locations to avoid silent failures.
Ideally, it would be preferable for these functions to propagate errors
by value (for example, using std::expected), so that errors can always
be handled explicitly without relying on exceptions.

Change-Id: I7189e97f206bae1eed5c8db44ebbbcd4393c8f8d
Signed-off-by: Eric Yang <eric.yang.wiwynn@gmail.com>

show more ...

fix ci: move to stdexec::inline_scheduler

Move from `exec::inline_scheduler` to `stdexec::inline_scheduler`.

Looks like the definition has been deprecated and is pointing to
`stdexec::inline_schedu

fix ci: move to stdexec::inline_scheduler

Move from `exec::inline_scheduler` to `stdexec::inline_scheduler`.

Looks like the definition has been deprecated and is pointing to
`stdexec::inline_scheduler` anyways.

the relevant header
```

namespace exec {
// A simple scheduler that executes its continuation inline, on the
// thread of the caller of start().
using inline_scheduler
[[deprecated("Use stdexec::inline_scheduler instead")]] = stdexec::inline_scheduler;
} // namespace exec
```

Tested: Inspection only.

Fixes: 10d0b4b7d1498cfd5c3d37edea271a54d1984e41
Change-Id: I1d23cb624e8902c541eeb31bd31155a3207a2da9
Signed-off-by: Alexander Hansen <alexander.hansen@9elements.com>

show more ...

Revert "pldm: use std::expected for instance ID allocation"

This reverts commit 70262ed7bf854b25d4b65628bc3c892ddfe9380f.

Change-Id: I1f54d6b60bf3cfa9f00764140eec6167e1e182d6
Signed-off-by: Manojki

Revert "pldm: use std::expected for instance ID allocation"

This reverts commit 70262ed7bf854b25d4b65628bc3c892ddfe9380f.

Change-Id: I1f54d6b60bf3cfa9f00764140eec6167e1e182d6
Signed-off-by: Manojkiran Eda <manojkiran.eda@gmail.com>

show more ...

pldm: use std::expected for instance ID allocation

Refactor InstanceIdDb::next() to return
std::expected<uint8_t, InstanceIdError> instead of throwing exceptions.
This change enables callers to expl

pldm: use std::expected for instance ID allocation

Refactor InstanceIdDb::next() to return
std::expected<uint8_t, InstanceIdError> instead of throwing exceptions.
This change enables callers to explicitly handle allocation errors via
value inspection, rather than relying on exception handling.

This approach prevents core dumps from uncaught exceptions and
eliminates the need for pervasive try-catch blocks.
Callers can now access the error code and message directly, improving
clarity and control of error propagation.

Note:
Errors from InstanceIdDb::next() are currently handled via early
return, which may silently discard failures.
This is a temporary solution; APIs and callers will need to be updated
in the future to propagate and handle errors explicitly.

Change-Id: Ibf2e0034b0ee725cb59adfd93b74e48db8c42cba
Signed-off-by: Eric Yang <eric.yang.wiwynn@gmail.com>

show more ...

Fix build issues after migrating to Clang 20

With the recent migration to Clang 20 in openbmc-build-scripts, the
current PLDM codebase fails to compile due to stricter type checking
and unused field

Fix build issues after migrating to Clang 20

With the recent migration to Clang 20 in openbmc-build-scripts, the
current PLDM codebase fails to compile due to stricter type checking
and unused field warnings introduced by the updated compiler.

This commit addresses the following issues to restore build stability:

1. Callback compatibility: Replaced std::bind_front with lambda
expressions that explicitly capture this, resolving type mismatch
errors when assigning member function callbacks to std::function.
2. Unused field warning: Removed the unused `bus` member variable from
the `SoftPowerOff` class and updated its constructor accordingly to
eliminate `-Wunused-private-field` warnings.

These changes are required to allow successful CI builds under the
Clang 20 toolchain.

Change-Id: I8eb909c2b1263bf7b2dbf53cdfc0424d425d352b
Signed-off-by: Eric Yang <eric.yang.wiwynn@gmail.com>

show more ...

requester: associate Entity Manager configs with MCTP endpoints

Currently, pldmd listens for new MCTP endpoint exposed by mctpd, but
they only shows their EID, Network Id, and SupportedMessageTypes,

requester: associate Entity Manager configs with MCTP endpoints

Currently, pldmd listens for new MCTP endpoint exposed by mctpd, but
they only shows their EID, Network Id, and SupportedMessageTypes, which
cannot fulfill some requirements, e.g., get the device's name or which
board it belongs to.

In openbmc, the additional information are exposed by Entity
Manager[1][2], so add this ability to `MctpDiscovery, it queries the
association between MCTP endpoints and Entity Manager configurations
from MCTP Reactor, when a new MCTP endpoint has been register by mctpd.

Added unit test for this commit to validate the association does work,
passed.

[1]: https://github.com/openbmc/entity-manager/blob/master/schemas/mctp.json
[2]: https://gerrit.openbmc.org/c/openbmc/dbus-sensors/+/69111

Change-Id: Ibf1717f1840e527f21bd8397e747ae121e2dcd25
Signed-off-by: Unive Tien <unive.tien.wiwynn@gmail.com>

show more ...

tests: Migrate to placement_new from reinterpret casting

reinterpret_cast is prohibited by the C++ core guidelines because
it takes the behavior outside the language definition and gives
problems wi

tests: Migrate to placement_new from reinterpret casting

reinterpret_cast is prohibited by the C++ core guidelines because
it takes the behavior outside the language definition and gives
problems with type safety. Placement-new on the other-hand allows
to control the object storage while still properly instantiating
an object,keeping the behavior inside the C++ language
specification.

Change-Id: Ifab9ea58b932db11d7af0b9def119bed1bfdc44d
Signed-off-by: Pavithra Barithaya <pavithrabarithaya07@gmail.com>

show more ...

Support `terminus_name` option in `dbus_to_terminus_effecter`

Support `terminus_name` configuration option in
`dbus_to_terminus_effecter` to allow setting the destination terminus
beside `mctp_eid`.

Support `terminus_name` option in `dbus_to_terminus_effecter`

Support `terminus_name` configuration option in
`dbus_to_terminus_effecter` to allow setting the destination terminus
beside `mctp_eid`. This is helpful when the mctp endpoint Eid is not
static.

Change-Id: I8b1ed15741807086254146017c99c13ae667dac1
Signed-off-by: Thu Nguyen <thu@os.amperecomputing.com>

show more ...

platform-mc: Handle `Connectivity` propertiesChanged signal

From Mctp codeConstruct version 2.0 [1], mctpd supports `.Connectivity`
property under `au.com.CodeConstruct.MCTP.Endpoint` interface of t

platform-mc: Handle `Connectivity` propertiesChanged signal

From Mctp codeConstruct version 2.0 [1], mctpd supports `.Connectivity`
property under `au.com.CodeConstruct.MCTP.Endpoint` interface of the
endpoint. This commit handles the propertiesChanged signal from this
interface, and updates the Availability of the MCTP Endpoint accordingly
in the source to enable or disable message sending/receiving via that
endpoint of the terminus.

[1] https://github.com/CodeConstruct/mctp/blob/v2.0/docs/endpoint-recovery.md#proposed-design

When the discovery process first starts, it will only handle the
endpoints that have `Available` `.Connectivity`. It lets the
propertiesChanged signal trigger the adding of the endpoints when they
are back to` Available`.

On interfaceAdded signal, it assumes that mctpd only publishes available
endpoints to D-Bus, so it adds the endpoints to the terminus anyway.

Tested:
1. Enable `unsafe-writable-connectivity` option in PACKAGECONFIG of mctp
recipe.
2. After PLDM discovers all the endpoints, write `Degraded` to
`.Connectivity` of one of the endpoint.
3. Write it back to `Available` to see how message is stopped from being
sent/received via the endpoint.

Signed-off-by: Chau Ly <chaul@amperecomputing.com>
Signed-off-by: Thu Nguyen <thu@os.amperecomputing.com>
Change-Id: I5b7a38ae72e655b60d71396a1118f2809aaa3838

show more ...

clang-format: update latest spec and reformat

Copy the latest format file from the docs repository and apply.

Change-Id: I95f756bab7f403af49a94011bbb1fe4e51f985ad
Signed-off-by: Patrick Williams <p

clang-format: update latest spec and reformat

Copy the latest format file from the docs repository and apply.

Change-Id: I95f756bab7f403af49a94011bbb1fe4e51f985ad
Signed-off-by: Patrick Williams <patrick@stwcx.xyz>

show more ...

Format meson files with meson.format

Meson 1.5.0 introduced a new feature to format all the meson
files. Formatting all the meson files is now as simple as running
`meson format -i -r` command in th

Format meson files with meson.format

Meson 1.5.0 introduced a new feature to format all the meson
files. Formatting all the meson files is now as simple as running
`meson format -i -r` command in the repository root folder.

more details : https://mesonbuild.com/Commands.html#format

Change-Id: I9c5468cc502ae78b7a055e2de2a10296930cb9ec
Signed-off-by: Manojkiran Eda <manojkiran.eda@gmail.com>

show more ...

clang-format: re-format for clang-18

clang-format-18 isn't compatible with the clang-format-17 output, so we
need to reformat the code with the latest version. The way clang-18
handles lambda forma

clang-format: re-format for clang-18

clang-format-18 isn't compatible with the clang-format-17 output, so we
need to reformat the code with the latest version. The way clang-18
handles lambda formatting also changed, so we have made changes to the
organization default style format to better handle lambda formatting.

See I5e08687e696dd240402a2780158664b7113def0e for updated style.
See Iea0776aaa7edd483fa395e23de25ebf5a6288f71 for clang-18 enablement.

Change-Id: I8c84201cb2343a8c8a5507a49de0721a1bee7063
Signed-off-by: Patrick Williams <patrick@stwcx.xyz>

show more ...

stdexec: pass scheduler to default_task_context

The latest version of stdexec[1] does not allow a default constructed
`default_task_context` but requires a scheduler to be passed in. The
previous d

stdexec: pass scheduler to default_task_context

The latest version of stdexec[1] does not allow a default constructed
`default_task_context` but requires a scheduler to be passed in. The
previous default was to use an `inline_scheduler` so explicitly pass
that in.

[1]: https://github.com/openbmc/sdbusplus/commit/5cee91570368554a7fcbbd9418f65efda449fa70

Signed-off-by: Patrick Williams <patrick@stwcx.xyz>
Change-Id: Ib210b42ba5272e0176cd4423b156ca6372115a55

show more ...

pldm: Drop pessimizing moves

Fixes issues such as:

```
../host-bmc/host_pdr_handler.cpp:684:9: error: moving a temporary object prevents copy elision [-Werror,-Wpessimizing-move]
684 | st

pldm: Drop pessimizing moves

Fixes issues such as:

```
../host-bmc/host_pdr_handler.cpp:684:9: error: moving a temporary object prevents copy elision [-Werror,-Wpessimizing-move]
684 | std::move(std::vector<uint8_t>(1, PLDM_PDR_ENTITY_ASSOCIATION)),
| ^
../host-bmc/host_pdr_handler.cpp:684:9: note: remove std::move call here
684 | std::move(std::vector<uint8_t>(1, PLDM_PDR_ENTITY_ASSOCIATION)),
| ^~~~~~~~~~ ~
```

Change-Id: Id26f0d6fc21837e1eb76ae3c294c222782a4e69f
Signed-off-by: Andrew Jeffery <andrew@codeconstruct.com.au>

show more ...

requester: Use return code instead of throwing exception

The coroutine API should only forward error code and response data
to the caller when sends/receives PLDM message instead check the
response

requester: Use return code instead of throwing exception

The coroutine API should only forward error code and response data
to the caller when sends/receives PLDM message instead check the
response and thrown exception.

The `sendRecvMsg` API will response tuple of error code, response and
response length as below:

+ [error_code, _, _] if registerRequest fails with `error_code`.
+ [PLDM_ERROR_NOT_READY, nullptr, 0] if the request is timed out.
+ [PLDM_SUCCESS, resp, len] if succeeded.

Signed-off-by: Khang Nguyen Duy <khangng@amperecomputing.com>
Signed-off-by: Thu Nguyen <thu@os.amperecomputing.com>
Change-Id: Id8262d147e9fed50af8f55f1c611a3a3b153b6e3

show more ...

requester: Add coroutine API

Added coroutine API to Handler using stdexec to help send and receive
messages in one function call.

For example, to get all PDRs from terminus needs multiple getPDR
co

requester: Add coroutine API

Added coroutine API to Handler using stdexec to help send and receive
messages in one function call.

For example, to get all PDRs from terminus needs multiple getPDR
commands. Except first getPDR commands, the following getPDR command
request data(e.g. dataTransferHandle) needs the previous getPDR command
response data(e.g. nextDataTransferHandle). By using C++ co_await, the
code to get all PDRs can straightforward be implemented by a while loop.

Pseudo code example:
do
{
auto rc = co_await getPDR(dataTransferHndl, ...);

// update dataTransferHndl for next getPDR command
dataTransferHndl = nextDataTransferHndl;

} while(...);

Signed-off-by: Gilbert Chen <gilbert.chen@arm.com>
Signed-off-by: Khang Nguyen <khangng@amperecomputing.com>
Signed-off-by: Thu Nguyen <thu@os.amperecomputing.com>
Change-Id: I7b47d15ac15f6ae661ec94dca6a281844b939a44

show more ...

requester: Modified MctpDiscovery class

Modified MctpDiscovery class to take list of managers instead of single
fwManager. The change is for adding platform-mc manager.
Added loadStaticEndpoints API

requester: Modified MctpDiscovery class

Modified MctpDiscovery class to take list of managers instead of single
fwManager. The change is for adding platform-mc manager.
Added loadStaticEndpoints API for MCTP layer which doesn't implement
/xyz/openbmc_project/MCTP/Endpoint.Interface

The patch is part of implementation of design document below.
https://gerrit.openbmc-project.xyz/c/openbmc/docs/+/47252

Signed-off-by: Gilbert Chen <gilbert.chen@arm.com>
Signed-off-by: Thu Nguyen <thu@os.amperecomputing.com>
Change-Id: I1e1673504583a87f2a9bc3adf76fb49c2dc30254

show more ...

Replace sizeof() with .size()

This change replaces the usage of sizeof() with .size() to
calculate size of arrays and expressions of STL container types.
Using .size() with containers like std::vect

Replace sizeof() with .size()

This change replaces the usage of sizeof() with .size() to
calculate size of arrays and expressions of STL container types.
Using .size() with containers like std::vector, std::array, and
std::string improves code readability and adheres to modern C++
standards. This also helps in clang-tidy checks.

Change-Id: I1e9e6e453a3b9218682098887763a4908f186815
Signed-off-by: Pavithra Barithaya <pavithra.b@ibm.com>

show more ...

Deprecate the support for x86 sdk

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

Deprecate the support for x86 sdk

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>

show more ...

meson: adjust nlohmann-json dependency

Simplify the meson dependency for nlohmann-json, use the consistent
dependency name, and align the wrap file name with other repositories.

Signed-off-by: Patr

meson: adjust nlohmann-json dependency

Simplify the meson dependency for nlohmann-json, use the consistent
dependency name, and align the wrap file name with other repositories.

Signed-off-by: Patrick Williams <patrick@stwcx.xyz>
Change-Id: I3a682a3e07e38d0a1bfe35858e24f1bb74b4ec34

show more ...

sdbusplus: avoid deprecated phosphor::Timer

sdbusplus had an older type named `phosphor::Timer` which was
recently renamed to `sdbusplus::Timer`. Update the code to use
the new type alias.

Change-

sdbusplus: avoid deprecated phosphor::Timer

sdbusplus had an older type named `phosphor::Timer` which was
recently renamed to `sdbusplus::Timer`. Update the code to use
the new type alias.

Change-Id: I68d3968bcb359bee5ea0c9e695572b68fcf34136
Signed-off-by: Patrick Williams <patrick@stwcx.xyz>

show more ...

build: use allowed over enabled

Meson feature options are typically in a tri-state of enabled, disabled,
or auto. The enabled and disabled functions on an option (from
`get_option`) no longer retur

build: use allowed over enabled

Meson feature options are typically in a tri-state of enabled, disabled,
or auto. The enabled and disabled functions on an option (from
`get_option`) no longer return true for auto features. Instead, the
expectation is to use `allowed()` which is true for both enabled and auto.

Switch all uses of `enabled` to `allowed`.

Change-Id: I794c658531483266bfce6f9b28127a166ed01490
Signed-off-by: Patrick Williams <patrick@stwcx.xyz>

show more ...

requestHandler: Queue the request messages to one endpoint

Section "Requirements for requesters" in DSP0240 details "A PLDM
terminus that issues PLDM requests to another PLDM terminus shall wait
unt

requestHandler: Queue the request messages to one endpoint

Section "Requirements for requesters" in DSP0240 details "A PLDM
terminus that issues PLDM requests to another PLDM terminus shall wait
until one of the following occurs before issuing a new PLDM request:
it gets the response to a particular request, it times out waiting for
the response, or it receives an indication that transmission of the
particular request failed." So the registered request messages to one
endpoint have to be queued before sending to meet the requester
requirement.
When a sensor manager is added to pldmd daemon to manage the sensors
of one terminus, the sensor manager will support the sensor polling
timers. The timers will poll the PLDM sensors with different intervals
so there can be many `getSensorReading` requests to one endpoint to
get the sensor values. Moreover, BMC also sends
`PollForPlatformEventMessage` requests to get the event data when it
receives the `pldmMessagePollEvent` event from the terminus. So those
TX request messages have to be queued.

Signed-off-by: Thu Nguyen <thu@os.amperecomputing.com>
Change-Id: I9756e1029b41a297a8f73e2ab7be47bc143ef710

show more ...

pldm: Convert to using libpldm transport APIs

A significant amount of logic can be removed by exploiting the new
transport APIs provided by libpldm. Switch the pldm repository over to
use these by i

pldm: Convert to using libpldm transport APIs

A significant amount of logic can be removed by exploiting the new
transport APIs provided by libpldm. Switch the pldm repository over to
use these by introducing an RAII wrapper for the APIs. The current
stance is to continue using the legacy mctp-demux transport
implementation, but we also provide a build option to switch to the
AF_MCTP transport.

We don't currently have the infrastructure in place to get the correct
TIDs, so to keep everything working as before use the EID as the TID in
the EID-to-TID mapping.

Change-Id: I366f079082b102cfc0e90db0f62208581eb8693e
Signed-off-by: Rashmica Gupta <rashmica@linux.ibm.com>
Signed-off-by: Delphine CC Chiu <Delphine_CC_Chiu@wiwynn.com>
Signed-off-by: Thu Nguyen <thu@os.amperecomputing.com>
Signed-off-by: Andrew Jeffery <andrew@aj.id.au>
Signed-off-by: Konstantin Aladyshev <aladyshev22@gmail.com>

show more ...

InstanceIdDb: Move header from pldmd/ to common/

The header is used across all executables produced by the project, which
makes it "common".

Change-Id: I022b179fad1de8dba2ef39fd33bc240ec73a2d33
Sig

InstanceIdDb: Move header from pldmd/ to common/

The header is used across all executables produced by the project, which
makes it "common".

Change-Id: I022b179fad1de8dba2ef39fd33bc240ec73a2d33
Signed-off-by: Andrew Jeffery <andrew@codeconstruct.com.au>

show more ...