Re: parallel autotest [2/3]: Implement 'testsuite --jobs'.

[Eric Blake]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

According to Ralf Wildenhues on 9/29/2008 3:35 PM:
>>> address@hidden address@hidden@address@hidden
>>> address@hidden address@hidden
>>> address@hidden -j
>> This second line is redundant, since you used @ovar on the line above.
> 
> The first line addresses --jobs=N, the second line addresses -jN.
> What is redundant there?  (see below at --help output, for more info)

I guess I didn't phrase that well.  I meant that the second -j line:

- -j[N]
- -j

looks redundant; I think it could be better rendered as:

- --jobs[=N]
- -jN
- -j

in the manual (@var, not @ovar), and

- -j, --jobs[=N]

in the --help.  But I can also see why you are using -j[N] in the --help,
so I could still be swayed to go your way.

>> Looks sane, on a first reading.
> 
> Just to be sure, you did notice that commented bit:
> 
>>> +      . "$at_test_source" # AT_JOB_FIFO_FD<&-
> 
> that would be nice to have without the '#' but doesn't work this way?

I did wonder about that comment.  Closing the fd would mean one less
leaked fd into the test; what went wrong?

> 
> Hmm.  I'm starting to think this whole test is a bit too fragile; if the
> load is too high (for example because you run ./testsuite -jN), then it
> can fail sometimes.

Is there some other sort of test, even at -j2, where we can kick off two
tests that do some handshaking interleaved by sleeps?  Perhaps not; tests
have to be independent because the testsuite must gracefully fall back on
a serialized run on deficient platforms.

> Making it robust enough will probably cause the test to take an unduly
> long time,

I'm certainly not coming up with any bright ideas, which means this
feature will not get as much test coverage on the platforms where it is
most needed.  On the other hand, using a parallel test is so enticing that
I still think it is worth checking in, even if we can't figure out a
robust test for it.

- --
Don't work too hard, make some time for fun as well!

Eric Blake             address@hidden
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (Cygwin)
Comment: Public key at home.comcast.net/~ericblake/eblake.gpg
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iEYEARECAAYFAkjhjAEACgkQ84KuGfSFAYBjaACgjwllZ3Y+JJaTxylVvI9ZY9yu
5gMAoIcEz0ku8vN/i5Jsn7piAlqd6NfM
=xPKA
-----END PGP SIGNATURE-----