Checking DMesg logs for Panics/Errors post test case execution (#3674)

* changes to check kernel errors using function call

* assign default false to assert_kernel_error

* addressing node reference issue

* Adding kernel panic check as well in the after_case

* Fixing Nox Null Item access error

* Variable and Function Renaming with minor fixes

* Fixing flake8 long line error

Author: umfranci

lisa/node.py

@@ -24,7 +24,7 @@
 from lisa.nic import Nics, NicsBSD
 from lisa.operating_system import OperatingSystem
 from lisa.secret import add_secret
-from lisa.tools import Chmod, Df, Echo, Lsblk, Mkfs, Mount, Reboot, Uname, Wsl
+from lisa.tools import Chmod, Df, Dmesg, Echo, Lsblk, Mkfs, Mount, Reboot, Uname, Wsl
 from lisa.tools.mkfs import FileSystem
 from lisa.util import (
     ContextMixin,
@@ -110,6 +110,7 @@ def __init__(
         self._support_sudo: Optional[bool] = None
         self._is_dirty: bool = False
         self.capture_boot_time: bool = False
+        self.assert_kernel_error_after_test: bool = False
         self.capture_azure_information: bool = False
         self.capture_kernel_config: bool = False
         self.has_checked_bash_prompt: bool = False
@@ -501,6 +502,12 @@ def get_information(self) -> Dict[str, str]:
 
         return final_information
 
+    def check_kernel_error(self) -> None:
+        # Check if the kernel is in a healthy state without errors or panics.
+        dmesg = self.tools[Dmesg]
+        dmesg.check_kernel_errors(force_run=True, throw_error=True)
+        self.check_kernel_panic()
+
     def expand_env_path(self, raw_path: str) -> str:
         echo = self.tools[Echo]
         result = echo.run(raw_path, shell=True)

lisa/platform_.py

@@ -195,6 +195,9 @@ def deploy_environment(self, environment: Environment) -> None:
                 node.features = Features(node, self)
             node.capture_azure_information = platform_runbook.capture_azure_information
             node.capture_boot_time = platform_runbook.capture_boot_time
+            node.assert_kernel_error_after_test = (
+                platform_runbook.assert_kernel_error_after_test
+            )
             node.capture_kernel_config = (
                 platform_runbook.capture_kernel_config_information
             )

lisa/schema.py

@@ -1444,6 +1444,8 @@ class Platform(TypedSchema, ExtendableSchemaMixin):
     capture_azure_information: bool = False
     # capture boot time info or not
     capture_boot_time: bool = False
+    # to check if dmesg logs need to be analyzed after each test case
+    assert_kernel_error_after_test: bool = False
     # capture kernel config info or not
     capture_kernel_config_information: bool = False
     capture_vm_information: bool = True

lisa/testsuite.py

@@ -856,6 +856,11 @@ def __run_case(
                 timeout=timeout,
                 test_kwargs=test_kwargs,
             )
+            if case_result.environment is not None:
+                nodes = case_result.environment.nodes
+                for node in nodes.list():
+                    if node.assert_kernel_error_after_test:
+                        node.check_kernel_error()
             case_result.set_status(TestStatus.PASSED, "")
         except Exception as identifier:
             case_result.handle_exception(exception=identifier, log=log)