asyncio.unix_events._compute_returncode() can be replaced with os._wait_status_to_returncode() , but I prefer to do that in a separated PR since Lib/test/test_asyncio/test_unix_events.py has extensive tests on calls to os.W*() functions using waitpid_mocks() helper.

Well, I prefer to keep this PR short to ease its review.

I updated my PR to rename the function os.status_to_exitcode() and make it public.

TODO: Add unit tests by spawning a child process (maybe using fork?) and use waitpid() to get a status.

I modified my PR to add Windows support. On Windows, os.waitpid() status also requires an operation (shif right by 8 bits) to get an exitcode from the waitpid status. So IMO it's worth it to add it to Windows as well, which makes the function even more useful ;-)

But the Windows CI should be checked.

Fixed. I also added an unit test for signals.

I didn't know that os.spawnv() uses a shell. Apparently, Windows shell doesn't like spaces. I changed the test to use a filename instead.

But if the API is provided by the public API, it would help to handling status code through standard library suggest. :)

I'm not sure that I understood this point. Do you mean that the stdlib should use the new function? My PR modify 3 modules to use use: os, multiprocessing and subprocess. And also many tests.

Hum, maybe run_cgi() http.server log should be modified in a separated PR to use the function:

pid, sts = os.waitpid(pid, 0)
...
if sts:
    self.log_error("CGI script exit status %#x", sts)

IMO logging an exit code rather than an exit status is more convenient.

Tools/scripts/which.py may use this function as well to display an exit code rather than a raw waitpid exit status:

sts = os.system('ls ' + longlist + ' ' + filename)
if sts: msg('"ls -l" exit status: ' + repr(sts))

But it changes the behavior, so I prefer to do it in a separated PR.

mailcap.test() may also uses it for:

print("Executing:", command)
sts = os.system(command)
if sts:
    print("Exit status:", sts)

I renamed the function to os.waitstatus_to_exitcode(). IMHO this name avoids any risk of confusion and it's short enough.

I moved the function documentation at the right place: close to waitpid() documentation.

I also rebased my PR and squashed commits to be able to modify the commit message.

(Oops, "make regen-all" failed: it's now fixed.)

@pablogsal, @methane, @gpshead, @pitrou: Hi. Do you think that it would be an useful addition to the stdlib?

IMO it is worth it: I was able to use the new function in 3 modules of the stdlib (os, multiprocessing, subprocess), but also _bootsubprocess, and even more tests.

Oh, another possible usage of waitstatus_to_exitcode(): os.popen().close() result. Example from test_popen.py:

    def test_return_code(self):
        self.assertEqual(os.popen("exit 0").close(), None)
        if os.name == 'nt':
            self.assertEqual(os.popen("exit 42").close(), 42)
        else:
            self.assertEqual(os.popen("exit 42").close(), 42 << 8)

I pushed as much changes as possible in https://bugs.python.org/issue40094 (non-controversial changes) to make this PR as small as possible. So remaining changes should really be focused on the value of the addition of the os.waitstatus_to_exitcode() function.

I rebased my PR on top of this changes and I modify popen().close() documentation and test to use waitstatus_to_exitcode().