Merge origin/master into xcvrd-refactor-9

- Integrated fast-reboot caching feature from master
- Updated is_fast_reboot_enabled() calls to use new cached method
- Preserved refactored CMIS state handler functions

Author: bobby-nexthop

.github/copilot-instructions.md

@@ -0,0 +1,157 @@
+# Copilot Instructions for sonic-platform-daemons
+
+## Project Overview
+
+sonic-platform-daemons contains the platform monitoring daemons for SONiC. These daemons run as services on the switch, continuously monitoring hardware components (fans, PSUs, thermals, transceivers, LEDs, PCIe, storage) and publishing their state to the SONiC Redis databases. They consume the platform APIs defined in sonic-platform-common.
+
+## Architecture
+
+```
+sonic-platform-daemons/
+├── sonic-xcvrd/          # Transceiver daemon (SFP/QSFP monitoring)
+│   ├── xcvrd/
+│   │   ├── xcvrd.py      # Main transceiver monitoring daemon
+│   │   └── ...
+│   ├── tests/            # xcvrd tests
+│   └── setup.py
+├── sonic-psud/           # PSU daemon (power supply monitoring)
+│   ├── scripts/psud
+│   ├── tests/
+│   └── setup.py
+├── sonic-thermalctld/    # Thermal control daemon
+│   ├── scripts/thermalctld
+│   ├── tests/
+│   └── setup.py
+├── sonic-ledd/           # LED daemon
+│   ├── scripts/ledd
+│   ├── tests/
+│   └── setup.py
+├── sonic-pcied/          # PCIe monitoring daemon
+│   ├── scripts/pcied
+│   ├── tests/
+│   └── setup.py
+├── sonic-syseepromd/     # System EEPROM daemon
+│   ├── scripts/syseepromd
+│   ├── tests/
+│   └── setup.py
+├── sonic-chassisd/       # Chassis daemon (modular chassis)
+│   ├── scripts/chassisd
+│   ├── tests/
+│   └── setup.py
+├── sonic-ycabled/        # Y-cable daemon (dual-ToR)
+│   ├── ycable/
+│   ├── tests/
+│   └── setup.py
+├── sonic-sensormond/     # Sensor monitoring daemon
+├── sonic-stormond/       # Storage monitoring daemon
+└── .github/              # GitHub configuration
+```
+
+### Key Concepts
+- **Each daemon is a standalone Python package** with its own `setup.py` and tests
+- **Platform API consumer**: Daemons call `sonic-platform-common` base class methods
+- **DB publishers**: Daemons write hardware state to STATE_DB, update COUNTERS_DB
+- **Event-driven monitoring**: Daemons poll hardware at intervals, detect state changes
+
+## Language & Style
+
+- **Primary language**: Python 3
+- **Indentation**: 4 spaces
+- **Naming conventions**:
+  - Daemon scripts: lowercase (e.g., `psud`, `thermalctld`, `xcvrd`)
+  - Functions: `snake_case`
+  - Classes: `PascalCase`
+  - Constants: `UPPER_CASE`
+- **Logging**: Use `sonic_py_common.daemon_base.Logger`
+- **Docstrings**: Required for public methods
+
+## Build Instructions
+
+```bash
+# Each daemon is a separate Python package
+cd sonic-xcvrd
+python3 setup.py bdist_wheel
+
+cd sonic-psud
+python3 setup.py bdist_wheel
+# etc.
+```
+
+## Testing
+
+```bash
+# Run tests for a specific daemon
+cd sonic-xcvrd
+pytest tests/ -v
+
+cd sonic-psud
+pytest tests/ -v
+
+# Run with coverage
+pytest tests/ --cov --cov-report=term-missing
+```
+
+- Each daemon has its own `tests/` directory
+- Tests use **pytest** with mock objects
+- Platform APIs are mocked (no real hardware needed)
+- Tests verify state machine logic, DB updates, and error handling
+
+## PR Guidelines
+
+- **Commit format**: `[daemon]: Description` (e.g., `[xcvrd]: Add DOM monitoring support`)
+- **Signed-off-by**: REQUIRED (`git commit -s`)
+- **CLA**: Sign Linux Foundation EasyCLA
+- **Testing**: All changes must include or update unit tests
+- **Platform compatibility**: Changes must not break any vendor platform
+- **DB schema**: Document any STATE_DB table changes
+
+## Common Patterns
+
+### Daemon Structure
+```python
+from sonic_py_common.daemon_base import DaemonBase
+
+class MyDaemon(DaemonBase):
+    def __init__(self):
+        super().__init__('mydaemon')
+        self.platform_chassis = load_platform_chassis()
+    
+    def run(self):
+        while True:
+            # Monitor hardware
+            status = self.platform_chassis.get_fan(0).get_status()
+            # Update DB
+            self.state_db.set('FAN_INFO|FAN0', {'status': str(status)})
+            time.sleep(self.polling_interval)
+```
+
+### DB Update Pattern
+```python
+# Daemons typically write to STATE_DB
+# Table format: TABLE_NAME|KEY
+# Fields: key-value pairs representing hardware state
+fvs = swsscommon.FieldValuePairs([
+    ('status', 'OK'),
+    ('speed', '12000'),
+    ('temperature', '35.0')
+])
+tbl.set(key, fvs)
+```
+
+## Dependencies
+
+- **sonic-platform-common**: Platform base classes (sonic_platform_base)
+- **sonic-py-common**: Common Python utilities, DaemonBase
+- **python-swsscommon**: Redis database bindings
+- **sonic-buildimage**: Packages are built within the buildimage system
+
+## Gotchas
+
+- **Platform plugin loading**: Daemons dynamically load vendor platform plugins — handle `ImportError` gracefully
+- **Polling intervals**: Too-frequent polling wastes CPU; too-infrequent misses events
+- **Error resilience**: Daemons must not crash on platform API errors — catch, log, continue
+- **Signal handling**: Daemons must handle SIGTERM/SIGINT for clean shutdown
+- **Multi-ASIC**: xcvrd and other daemons must be namespace-aware for multi-ASIC platforms
+- **DB consistency**: Always update all related fields atomically
+- **Warm restart**: Consider state preservation during warm restart scenarios
+- **Resource leaks**: Ensure file handles and DB connections are properly closed

azure-pipelines.yml

@@ -87,6 +87,22 @@ jobs:
         artifact: 'sonic-buildimage.vs'
         runVersion: 'latestFromBranch'
         runBranch: "$(sourceBranch)"
+        path: $(Build.ArtifactStagingDirectory)/download
+        patterns: |
+          target/debs/bookworm/libnl-3-200_*.deb
+          target/debs/bookworm/libnl-genl-3-200_*.deb
+          target/debs/bookworm/libnl-route-3-200_*.deb
+          target/debs/bookworm/libnl-nf-3-200_*.deb
+          target/debs/bookworm/libyang_1.0.73_amd64.deb
+          target/debs/bookworm/libswsscommon_1.0.0_amd64.deb
+          target/debs/bookworm/python3-swsscommon_1.0.0_amd64.deb
+
+          target/python-wheels/bookworm/swsssdk-2.0.1-py3-none-any.whl
+          target/python-wheels/bookworm/sonic_py_common-1.0-py3-none-any.whl
+          target/python-wheels/bookworm/sonic_yang_mgmt-1.0-py3-none-any.whl
+          target/python-wheels/bookworm/sonic_yang_models-1.0-py3-none-any.whl
+          target/python-wheels/bookworm/sonic_config_engine-1.0-py3-none-any.whl
+          target/python-wheels/bookworm/sonic_platform_common-1.0-py3-none-any.whl
       displayName: "Download artifacts from latest sonic-buildimage build"
 
     - script: |
@@ -99,7 +115,7 @@ jobs:
         sudo dpkg -i libyang_1.0.73_amd64.deb
         sudo dpkg -i libswsscommon_1.0.0_amd64.deb
         sudo dpkg -i python3-swsscommon_1.0.0_amd64.deb
-      workingDirectory: $(Pipeline.Workspace)/target/debs/bookworm/
+      workingDirectory: $(Build.ArtifactStagingDirectory)/download/target/debs/bookworm/
       displayName: 'Install Debian dependencies'
 
     - script: |
@@ -110,7 +126,7 @@ jobs:
         sudo pip3 install sonic_yang_models-1.0-py3-none-any.whl
         sudo pip3 install sonic_config_engine-1.0-py3-none-any.whl
         sudo pip3 install sonic_platform_common-1.0-py3-none-any.whl
-      workingDirectory: $(Pipeline.Workspace)/target/python-wheels/bookworm/
+      workingDirectory: $(Build.ArtifactStagingDirectory)/download/target/python-wheels/bookworm/
       displayName: 'Install Python dependencies'
 
     - script: |

sonic-chassisd/scripts/chassis_db_init

@@ -41,12 +41,12 @@ NOT_AVAILABLE = 'N/A'
 # Helper functions =============================================================
 #
 
-# try get information from platform API and return a default value if caught NotImplementedError
+# try get information from platform API and return a default value if caught NotImplementedError or TimeoutError
 
 
 def try_get(callback, *args, **kwargs):
     """
-    Handy function to invoke the callback and catch NotImplementedError
+    Handy function to invoke the callback and catch NotImplementedError or TimeoutError
     :param callback: Callback to be invoked
     :param args: Arguments to be passed to callback
     :param kwargs: Default return value if exception occur
@@ -57,7 +57,7 @@ def try_get(callback, *args, **kwargs):
         ret = callback(*args)
         if ret is None:
             ret = default
-    except NotImplementedError:
+    except (NotImplementedError, TimeoutError):
         ret = default
 
     return ret

sonic-chassisd/scripts/chassisd

@@ -102,7 +102,6 @@ INVALID_IP = '0.0.0.0'
 CHASSIS_MODULE_ADMIN_STATUS = 'admin_status'
 MODULE_ADMIN_DOWN = 0
 MODULE_ADMIN_UP = 1
-MODULE_PRE_SHUTDOWN = 2
 MODULE_REBOOT_CAUSE_DIR = "/host/reboot-cause/module/"
 MAX_HISTORY_FILES = 10
 
@@ -1359,12 +1358,6 @@ class ChassisdDaemon(daemon_base.DaemonBase):
 
     def submit_dpu_callback(self, module_index, admin_state):
         module = self.module_updater.chassis.get_module(module_index)
-
-        # This is only valid on platforms which have pci_detach and sensord changes required. If it is not implemented,
-        # there are no actions taken during this function execution.
-        if admin_state == MODULE_PRE_SHUTDOWN:
-            try_get(module.module_pre_shutdown, default=False)
-        # Set admin_state change in progress using the centralized method
         if admin_state == MODULE_ADMIN_DOWN:
             try_get(module.set_admin_state_gracefully, admin_state, default=False)
 
@@ -1387,10 +1380,7 @@ class ChassisdDaemon(daemon_base.DaemonBase):
                 # Get admin state of DPU
                 admin_state = self.module_updater.get_module_admin_status(module_name)
                 if admin_state == ModuleBase.MODULE_STATUS_EMPTY:
-                    op = MODULE_PRE_SHUTDOWN
-                    if operational_state != ModuleBase.MODULE_STATUS_OFFLINE:
-                        # shutdown DPU
-                        op = MODULE_ADMIN_DOWN
+                    op = MODULE_ADMIN_DOWN
 
                 # Initialize DPU_STATE DB table on bootup
                 dpu_state_key = "DPU_STATE|" + module_name
@@ -1439,6 +1429,7 @@ class ChassisdDaemon(daemon_base.DaemonBase):
         try:
             # Start configuration manager task
             if self.smartswitch:
+                self.set_initial_dpu_admin_state()
                 self.config_manager = SmartSwitchConfigManagerTask()
                 self.config_manager.task_run()
             elif self.module_updater.supervisor_slot == self.module_updater.my_slot:
@@ -1450,10 +1441,6 @@ class ChassisdDaemon(daemon_base.DaemonBase):
             # Start main loop
             self.log_info("Start daemon main loop")
 
-            # Set the initial DPU admin state for SmartSwitch
-            if self.smartswitch:
-                self.set_initial_dpu_admin_state()
-
             while not self.stop.wait(self.loop_interval):
                 self.module_updater.module_db_update()
                 self.module_updater.check_midplane_reachability()

sonic-chassisd/tests/test_chassis_db_init.py

@@ -38,3 +38,19 @@ def test_provision_db():
     assert serial == fvs[CHASSIS_INFO_SERIAL_FIELD]
     assert model == fvs[CHASSIS_INFO_MODEL_FIELD]
     assert revision == fvs[CHASSIS_INFO_REV_FIELD]
+
+def test_try_get_timeout_error():
+    def raise_timeout():
+        raise TimeoutError("timeout")
+
+    result = try_get(raise_timeout)
+
+    assert result == NOT_AVAILABLE
+
+def test_try_get_not_implemented_error():
+    def raise_not_implemented():
+        raise NotImplementedError("not implemented")
+
+    result = try_get(raise_not_implemented)
+
+    assert result == NOT_AVAILABLE

sonic-chassisd/tests/test_chassisd.py

@@ -1549,8 +1549,8 @@ def test_set_initial_dpu_admin_state_empty_offline():
         # Verify DPU state was updated with 'down' since operational state is OFFLINE
         mock_update_dpu_state.assert_called_once_with("DPU_STATE|DPU0", 'down')
 
-        # Verify callback was submitted with MODULE_PRE_SHUTDOWN since admin state is EMPTY and oper state is OFFLINE
-        mock_submit_callback.assert_called_once_with(0, MODULE_PRE_SHUTDOWN)
+        # Verify callback was submitted with MODULE_ADMIN_DOWN when admin state is EMPTY
+        mock_submit_callback.assert_called_once_with(0, MODULE_ADMIN_DOWN)
 
 
 def test_set_initial_dpu_admin_state_empty_not_offline():
@@ -1600,7 +1600,7 @@ def test_set_initial_dpu_admin_state_empty_not_offline():
         # Verify DPU state was updated with 'down' since operational state is not ONLINE
         mock_update_dpu_state.assert_called_once_with("DPU_STATE|DPU0", 'down')
 
-        # Verify callback was submitted with MODULE_ADMIN_DOWN since admin state is EMPTY and oper state is not OFFLINE
+        # Verify callback was submitted with MODULE_ADMIN_DOWN when admin state is EMPTY
         mock_submit_callback.assert_called_once_with(0, MODULE_ADMIN_DOWN)
 
 
@@ -2041,26 +2041,11 @@ def test_submit_dpu_callback():
     daemon_chassisd.module_updater = module_updater
     module_updater.module_table.get = MagicMock(return_value=(True, []))
 
-    # Test MODULE_ADMIN_DOWN scenario
-    with patch.object(module, 'module_pre_shutdown') as mock_pre_shutdown, \
-         patch.object(module, 'set_admin_state_gracefully') as mock_set_admin_state_gracefully:
+    # Test MODULE_ADMIN_DOWN scenario - set_admin_state_gracefully is called
+    with patch.object(module, 'set_admin_state_gracefully') as mock_set_admin_state_gracefully:
         daemon_chassisd.submit_dpu_callback(index, MODULE_ADMIN_DOWN)
-        # Verify pre_shutdown is not called for admin down
-        mock_pre_shutdown.assert_not_called()
-        # Verify set_admin_state_gracefully is called with MODULE_ADMIN_DOWN
         mock_set_admin_state_gracefully.assert_called_once_with(MODULE_ADMIN_DOWN)
 
-    # Test MODULE_PRE_SHUTDOWN scenario
-    with patch.object(module, 'module_pre_shutdown') as mock_pre_shutdown, \
-         patch.object(module, 'set_admin_state_gracefully') as mock_set_admin_state_gracefully:
-
-        module_updater.module_table.get = MagicMock(return_value=(True, []))
-        daemon_chassisd.submit_dpu_callback(index, MODULE_PRE_SHUTDOWN)
-
-        # Verify only pre_shutdown is called for pre-shutdown state
-        mock_pre_shutdown.assert_called_once()
-        mock_set_admin_state_gracefully.assert_not_called()
-
 def test_chassis_daemon_assertion():
     chassis = MockChassis()