Mike Gilbert

This fails in the same way with 2020.0.0-alpha1.

Portage sets an environment variable `T`, and that seems to trigger the failure here. You should be able to reproduce this like so: ``` meson build cd build ninja T=/tmp...

With this patch applied, I get a different assert failure: ``` ================================== 479/1064 ================================== test: systemd / test-mountpoint-util start time: 20:08:20 duration: 0.09s result: killed by signal 6 SIGABRT command:...

That fixed it, thanks.

We need to mount a new instance of /proc to get pids from the new namespace to show up; otherwise, /proc remains tied to the parent namespace. Is there some...

I still get the same failure on Gentoo with Portage's pid-sandbox feature enabled. The change in af918c481864116fdf92eceb3376a014053eaf5f does not help because we do not run the tests as root. ```...

> AFAICS, the failures have nothing to do with this PR, am I wrong? I'm not sure what gives you that impression; it looks like this PR does cause the...

From the log: ``` 2022-09-09T19:01:27.9583005Z ====================================================================== 2022-09-09T19:01:27.9585459Z FAIL: testSimple (portage.tests.emerge.test_simple.SimpleEmergeTestCase) 2022-09-09T19:01:27.9586076Z ---------------------------------------------------------------------- 2022-09-09T19:01:27.9589072Z Traceback (most recent call last): 2022-09-09T19:01:27.9601848Z File "/home/runner/work/portage/portage/build/lib/portage/tests/__init__.py", line 273, in run 2022-09-09T19:01:27.9602233Z testMethod() 2022-09-09T19:01:27.9602618Z File "/home/runner/work/portage/portage/build/lib/portage/tests/emerge/test_simple.py",...

I plan to take a look at this tomorrow.

Sorry, but I have to ask you to clean up the commit history a bit. - Rebase your changes on current master (no merge commits!). - Some of the commits...