Skip to contentSkip to conversation

oeqa/runtime: fix race-condition in minidebuginfo test

PATCH Superseded [email protected] 2 participants 3 messages Jun 10, 2024 – Jun 11, 2024

v3 of 4 Superseded Latest is v4 →

lore
Share & tools
lore
#0 lore raw

PATCH v3 1/1 oeqa/runtime: fix race-condition in minidebuginfo test

From: Etienne Cordonnier <[email protected]>

Fix this error where 'coredumpctl info' warns that the coredump is still being processed:

```
AssertionError: 1 != 0 : MiniDebugInfo Test failed: No match found.
-- Notice: 1 [email protected] unit is running, output may be incomplete.
```
---
 meta/lib/oeqa/runtime/cases/systemd.py | 7 +++++++
 1 file changed, 7 insertions(+)
meta/lib/oeqa/runtime/cases/systemd.py +7 -0
index 80fdae240a..aed230708e 100644
@@ -155,6 +155,13 @@ class SystemdServiceTests(SystemdTest):
         self.target.run('kill -SEGV %s' % output)
         self.assertEqual(status, 0, msg = 'Not able to find process that runs sleep, output : %s' % output)
 
+        # Give some time to [email protected] to process the coredump
+        for x in range(20):
+            status, output = self.target.run('coredumpctl --quiet --no-pager --no-legend | wc -l')
+            if output == "1":
+                break
+            time.sleep(1)
+
         (status, output) = self.target.run('coredumpctl info')
         self.assertEqual(status, 0, msg='MiniDebugInfo Test failed: %s' % output)
         self.assertEqual('sleep_for_duration (busybox.nosuid' in output or 'xnanosleep (sleep.coreutils' in output,
Signature · 1 line
2.36.1.vfs.0.0
#1 lore raw
Quoted 22 lines · Etienne Cordonnier via lists.openembedded.org

From: Etienne Cordonnier <[email protected]>

Fix this error where 'coredumpctl info' warns that the coredump is still being processed:

```
AssertionError: 1 != 0 : MiniDebugInfo Test failed: No match found.
-- Notice: 1 [email protected] unit is running, output may be incomplete.
```

---
meta/lib/oeqa/runtime/cases/systemd.py | 7 +++++++

1 file changed, 7 insertions(+)
meta/lib/oeqa/runtime/cases/systemd.py +0 -0
index 80fdae240a..aed230708e 100644
@@ -155,6 +155,13 @@ class SystemdServiceTests(SystemdTest):
        self.target.run('kill -SEGV %s' % output)

meta/lib/oeqa/runtime/cases/systemd.py class SystemdServiceTests(SystemdTest): · line 165

Quoted 10 lines
meta/lib/oeqa/runtime/cases/systemd.py +7 -0
        self.assertEqual(status, 0, msg = 'Not able to find process that runs sleep, output : %s' % output)
 
+        # Give some time to [email protected] to process the coredump
+        for x in range(20):
+            status, output = self.target.run('coredumpctl --quiet --no-pager --no-legend | wc -l')
+            if output == "1":
+                break
+            time.sleep(1)
+
        (status, output) = self.target.run('coredumpctl info')

As someone who spent too long staring at mysterious failures in the debuginfod tests, I recommend you add an ‘else’ case to that loop so that you can explicitly fail if the timeout expires without breaking out.

Also it feels like the test could be made more reliably by using matches instead of just asking it to list all crashes and expecting it to be one (what if something else crashed?), and assuming the latest crash is the one we’re after. If you use matches then the exit code is set to non-zero if the match isn’t found, so you could combine the retry loop and the actual test into one.

Ross

#2 lore raw

Hi Ross,
OK, I addressed your feedback. On a side note, coredumpctl matches based on process names are really confusing when used with symlinks. I used a match based on pid to avoid the confusion:
"

# coredumpctl list
TIME                         PID UID GID SIG     COREFILE EXE
   SIZE
Tue 2024-05-21 05:36:50 GMT 3442   0   0 SIGSEGV present
 /usr/bin/coreutils 109.0K
# coredumpctl list coreutils
No coredumps found.
# coredumpctl list sleep
TIME                         PID UID GID SIG     COREFILE EXE
   SIZE
Tue 2024-05-21 05:36:50 GMT 3442   0   0 SIGSEGV present
 /usr/bin/coreutils 109.0K
"
Quoted 53 lines · Ross Burton
Quoted 4 lines · Etienne Cordonnier via lists.openembedded.org

From: Etienne Cordonnier <[email protected]>

Fix this error where 'coredumpctl info' warns that the coredump is still

being

Quoted 5 lines

processed:

```
AssertionError: 1 != 0 : MiniDebugInfo Test failed: No match found. -- Notice: 1 [email protected] unit is running, output may be

incomplete.

Review

Quoted 8 lines

```

---
meta/lib/oeqa/runtime/cases/systemd.py | 7 +++++++

1 file changed, 7 insertions(+)
(quoted diff) +0 -0

b/meta/lib/oeqa/runtime/cases/systemd.py

meta/lib/oeqa/runtime/cases/systemd.py class SystemdServiceTests(SystemdTest): · line 156

Quoted 6 lines

index 80fdae240a..aed230708e 100644

meta/lib/oeqa/runtime/cases/systemd.py +0 -0
@@ -155,6 +155,13 @@ class SystemdServiceTests(SystemdTest):
        self.target.run('kill -SEGV %s' % output)
        self.assertEqual(status, 0, msg = 'Not able to find process that

runs sleep, output : %s' % output)

meta/lib/oeqa/runtime/cases/systemd.py class SystemdServiceTests(SystemdTest): · line 157

Quoted 2 lines
meta/lib/oeqa/runtime/cases/systemd.py +1 -0
+        # Give some time to [email protected] to process the

coredump

meta/lib/oeqa/runtime/cases/systemd.py class SystemdServiceTests(SystemdTest): · line 159

Quoted 2 lines
meta/lib/oeqa/runtime/cases/systemd.py +2 -0
+        for x in range(20):
+            status, output = self.target.run('coredumpctl --quiet

--no-pager --no-legend | wc -l')

meta/lib/oeqa/runtime/cases/systemd.py class SystemdServiceTests(SystemdTest): · line 165

Quoted 5 lines
meta/lib/oeqa/runtime/cases/systemd.py +4 -0
+            if output == "1":
+                break
+            time.sleep(1)
+
        (status, output) = self.target.run('coredumpctl info')

As someone who spent too long staring at mysterious failures in the debuginfod tests, I recommend you add an ‘else’ case to that loop so that you can explicitly fail if the timeout expires without breaking out.

Also it feels like the test could be made more reliably by using matches instead of just asking it to list all crashes and expecting it to be one (what if something else crashed?), and assuming the latest crash is the one
we’re after. If you use matches then the exit code is set to non-zero if
the match isn’t found, so you could combine the retry loop and the actual test into one.

Ross