[RTEMS Project] #4373: qemu: qemu-system-arm hangs during cron'ed test runs with rtems-test (was: rtems-test: qemu-system-arm hangs during cron'ed test runs)

RTEMS trac trac at rtems.org
Thu Apr 8 01:36:29 UTC 2021

#4373: qemu: qemu-system-arm hangs during cron'ed test runs with rtems-test
 Reporter:  Joel Sherrill  |       Owner:  (none)
     Type:  defect         |      Status:  new
 Priority:  normal         |   Milestone:  6.1
Component:  tool           |     Version:  6
 Severity:  normal         |  Resolution:
 Keywords:                 |  Blocked By:
 Blocking:                 |

Comment (by Chris Johns):

 I am sorry but you are going to have to debug this to resolve it because
 it is not been seen else where and/or we do not have a suitable set up to
 duplicate it.

 At the core of the bug is qemu stopping and it should not unless our test
 code or the BSP has a problem you have exposed. I agree there is a
 secondary bug where `rtems-test` is not killing the process but I am
 reluctant to look at that until the reason qemu hangs is found. A fixed
 `rtems-test` only results in more tests being (maybe) incorrectly tag as
 `indeterminate` and I prefer we look to get better qemu test results.

 A few suggestions...

 1. Is console output still being generated?

 2. See if you can find out where in an RTEMS test executable things are
 hanging. This means adding the gdb server flag to each qemu however I am
 not sure if this breaks when a number of qemu sessions are run in
 parallel, ie port binding.

 3. Can you connect a python debugger to an active process based on pid? If
 you could debug what is happening that would help the `rtems-test` of
 things but as stated IMO that is secondary to the real issue.

 4. Hack `rtems-test` to add some trace around the job handling. Again only
 `rtems-test` related.

Ticket URL: <http://devel.rtems.org/ticket/4373#comment:1>
RTEMS Project <http://www.rtems.org/>
RTEMS Project

More information about the bugs mailing list