[PATCH 2/5] introduce-resume-dependency.patch

[Andy Green]

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

Somebody in the thread at some point said:
| On Monday 02 June 2008 14:27:48 Andy Green wrote:
|> Defines a way for drivers to defer execution of resume callbacks
|> until one or more other driver they are dependent on has itself
|> resumed.
|
| okay, this is something that is unlikely to end in mainline? Besides
that what
| is the policy on capitalisation of the macros?
|
| Maybe move the fixes from 4/5 to this patch? Also called_flag seems to be
| unused, it needed some time until I understood how this is supposed to
work? I

No called_flag has a nice use, it's explained in the comment at the
bottom of the include file.  We use it later in jbt6k74 to make sure we
are resuming only after BOTH glamo and pcf50633 resumed.  It's better
than using a counter because we could stage the bits of resumption if we
like based on which dependent devices resumed so far.

| like the list_del as it is avoiding calling any _remove on module
| removal.

Yeah doing them at suspend time means no trouble on that account.

- -Andy
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (GNU/Linux)
Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org

iEYEARECAAYFAkhFJh0ACgkQOjLpvpq7dMpqlACghs6S0qUhtD4jRaDcQ5tGd1St
Dl4AoJBLR2xPuL2V3tZFl8TpecJ3VIxE
=yobd
-----END PGP SIGNATURE-----