Some integration tests (TestClientIPCRender) still touch the graphics hardware, despite default option --tests-use-real-graphics == FALSE
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
Mir |
Won't Fix
|
Medium
|
Unassigned |
Bug Description
Some integration tests (TestClientIPCR
So then, if the build machine doesn't have its graphics drivers in order, we notice tests failing when they actually shouldn't be running at all:
[0;32m[ RUN ] [mTestClientIP
libEGL warning: DRI2: xcb_connect failed
libEGL warning: DRI2: xcb_connect failed
/tmp/buildd/
Value of: red_pattern.
Actual: false
Expected: true
[0;31m[ FAILED ] [mTestClientIP
[0;32m[ RUN ] [mTestClientIP
libEGL warning: DRI2: xcb_connect failed
libEGL warning: DRI2: xcb_connect failed
/tmp/buildd/
Value of: red_pattern.
Actual: false
Expected: true
/tmp/buildd/
Value of: green_pattern.
Actual: false
Expected: true
[0;31m[ FAILED ] [mTestClientIP
See also: bug 1221373
Related branches
- Kevin DuBois (community): Approve
- PS Jenkins bot (community): Approve (continuous-integration)
- Alexandros Frantzis (community): Approve
- Andreas Pokorny (community): Approve
-
Diff: 64 lines (+6/-30)2 files modifiedtests/integration-tests/graphics/mesa/test_buffer_integration.cpp (+1/-21)
tests/mir_test_framework/stubbed_server_configuration.cpp (+5/-9)
Changed in mir: | |
assignee: | nobody → Daniel van Vugt (vanvugt) |
Changed in mir: | |
assignee: | Daniel van Vugt (vanvugt) → nobody |
I am OK with making the tests respect the --tests- use-real- graphics flag, but I would argue that we should then make the Android acceptance and integration tests run with --tests- use-real- graphics= true (since they do run on a capable device), so that we keep exercising the Mir/driver boundary.