commit 8951c87e8fca6ba57f78344119ab42866c190c80
author Harshit Aghera <haghera@nvidia.com> Wed Jun 25 15:25:33 2025 +0530
committer Ed Tanous <ed@tanous.net> Tue Jul 22 01:25:38 2025 +0000
tree 6c8fe6bcb22aeebd9bb63ee1d2228e497aafcfbf
parent a499a168f10d3913fbfd01a71ca7c0b207483b6c

nvidia-gpu: add SMA Temperature Sensor

Add support for device type SMA (System Management Agent) and its
temperature sensor. It is typically an MCU device.

Tested: Build an image for gb200nvl-obmc machine with the following
patches cherry picked. This patches are needed to enable the mctp stack.

https://gerrit.openbmc.org/c/openbmc/openbmc/+/79422

```
$ curl -s -k -u 'root:0penBmc' https://10.137.203.193/redfish/v1/Chassis/NVIDIA_GB200_1/Sensors/temperature_NVIDIA_GB200_GPU_SMA_255_TEMP_0
{
  "@odata.id": "/redfish/v1/Chassis/NVIDIA_GB200_1/Sensors/temperature_NVIDIA_GB200_GPU_SMA_255_TEMP_0",
  "@odata.type": "#Sensor.v1_2_0.Sensor",
  "Id": "temperature_NVIDIA_GB200_GPU_SMA_255_TEMP_0",
  "Name": "NVIDIA GB200 GPU SMA 255 TEMP 0",
  "Reading": 34.0,
  "ReadingRangeMax": 127.0,
  "ReadingRangeMin": -128.0,
  "ReadingType": "Temperature",
  "ReadingUnits": "Cel",
  "Status": {
    "Health": "OK",
    "State": "Enabled"
  }
}%
```

Change-Id: I560864758036a5b6ea6c1745145736c7bfa0a1c5
Signed-off-by: Harshit Aghera <haghera@nvidia.com>

src/nvidia-gpu/NvidiaGpuSensorMain.cpp

diff --git a/src/nvidia-gpu/NvidiaGpuSensorMain.cpp b/src/nvidia-gpu/NvidiaGpuSensorMain.cpp
index 6ccbb05..7e404f5 100644
--- a/src/nvidia-gpu/NvidiaGpuSensorMain.cpp
+++ b/src/nvidia-gpu/NvidiaGpuSensorMain.cpp
@@ -8,6 +8,7 @@
 #include "Utils.hpp"
 
 #include <NvidiaDeviceDiscovery.hpp>
+#include <NvidiaSmaDevice.hpp>
 #include <boost/asio/error.hpp>
 #include <boost/asio/io_context.hpp>
 #include <boost/asio/post.hpp>
@@ -27,6 +28,7 @@
 #include <vector>
 
 boost::container::flat_map<std::string, std::shared_ptr<GpuDevice>> gpuDevices;
+boost::container::flat_map<std::string, std::shared_ptr<SmaDevice>> smaDevices;
 
 void configTimerExpiryCallback(
     boost::asio::io_context& io, sdbusplus::asio::object_server& objectServer,
@@ -37,7 +39,8 @@
     {
         return; // we're being canceled
     }
-    createSensors(io, objectServer, gpuDevices, dbusConnection, mctpRequester);
+    createSensors(io, objectServer, gpuDevices, smaDevices, dbusConnection,
+                  mctpRequester);
 }
 
 int main()
@@ -51,7 +54,8 @@
     mctp::MctpRequester mctpRequester(io);
 
     boost::asio::post(io, [&]() {
-        createSensors(io, objectServer, gpuDevices, systemBus, mctpRequester);
+        createSensors(io, objectServer, gpuDevices, smaDevices, systemBus,
+                      mctpRequester);
     });
 
     boost::asio::steady_timer configTimer(io);
@@ -76,7 +80,9 @@
         static_cast<sdbusplus::bus_t&>(*systemBus),
         sdbusplus::bus::match::rules::interfacesRemovedAtPath(
             std::string(inventoryPath)),
-        [](sdbusplus::message_t& msg) { interfaceRemoved(msg, gpuDevices); });
+        [](sdbusplus::message_t& msg) {
+            interfaceRemoved(msg, gpuDevices, smaDevices);
+        });
 
     io.run();
     return 0;