Leaks in death tests can cause subsequent tests in the same process to fail
| Affects | Status | Importance | Assigned to | Milestone | |
|---|---|---|---|---|---|
| Mir |
Fix Released
|
Medium
|
Alexandros Frantzis | ||
| mir (Ubuntu) |
Fix Released
|
Medium
|
Unassigned | ||
Bug Description
To reproduce:
valgrind --error-exitcode=1 --trace-
The ClientLibraryThread tests should pass and not report any definite leaks. It's expected to get leaks from the DeathTests.
Note that for me (and Daniel, see [1]), this is still a problem even with Christopher's fix ([2]). So, this is a real issue, just not the one that was blocking CI.
I don't know why on some setups (e.g. Alan's) this fix seems to have no effect. Perhaps it was a combination of this leak and the one fixed by Chris that complicated the situation, and now that the other leak is fixed this will work properly on all setups.
[1] https:/
[2] https:/
Related branches
- Alan Griffiths: Abstain
- Daniel van Vugt: Abstain
- Chris Halse Rogers: Approve
- PS Jenkins bot (community): Needs Fixing (continuous-integration)
- Robert Carr (community): Approve
- Cemil Azizoglu (community): Approve
-
Diff: 28 lines (+6/-0)1 file modifiedtests/acceptance-tests/test_client_library_errors.cpp (+6/-0)
| summary: |
- Leaks in death tests can cause subsequent tests ran in the same process - to fail + Leaks in death tests can cause subsequent tests in the same process to + fail |
| Changed in mir: | |
| milestone: | none → 0.10.0 |
| assignee: | nobody → Alexandros Frantzis (afrantzis) |
| status: | New → In Progress |
| status: | In Progress → Fix Committed |
| Changed in mir: | |
| status: | Fix Committed → Fix Released |
| Daniel van Vugt (vanvugt) wrote | #1 |
| Changed in mir (Ubuntu): | |
| importance: | Undecided → Medium |
| status: | New → Fix Released |
mir (0.10.0+