[Prism54-devel] handling of -EINPROGRESS

[Luis R. Rodriguez]

--itqfrb9Qq3wY07cp
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable


Attention to: Jean Tourrilhes <jt@bougret.hpl.hp.com>

On Sun, Feb 22, 2004 at 03:01:41PM -0500, Luis R. Rodriguez wrote:
> On Sun, Feb 22, 2004 at 08:43:05PM +0200, Denis Vlasenko wrote:
> > I can't locate code which should "Call commit handler".
> > I tracked ioctl to wireless.c and dev.c, it seems to me it must
> > propagate error code all the way to userspace, but obviosly
> > it doesn't...
> >=20
> > What am I missing here?
>=20
> I don't remember exactly, but I think that use of -EINPROGRESS came from
> modeling our ioctl by what others do. ie, see airo.c airo_set_freq
> which sets:
>=20
> int rc =3D -EINPROGRESS;          /* Call commit handler */
>=20
> Looking through wireless.c I see call_commit_handler and I see it called
> only in ioctl_standard_call:
>=20
>         if(ret =3D=3D -EIWCOMMIT)
> 	                ret =3D call_commit_handler(dev);
>=20
> So perhaps we should be returning EIWCOMMIT instead.
>=20
> 	Luis
>=20

Jean,=20

can you tell us what we should return to force a commit? I see in
current kernel code EIWCOMMIT but some drivers are using EINPROGRESS.

	Luis

--=20
Key fingerprint =3D 113F B290 C6D2 0251 4D84  A34A 6ADD 4937 E20A 525E

--itqfrb9Qq3wY07cp
Content-Type: application/pgp-signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (GNU/Linux)

iD8DBQFAOis1at1JN+IKUl4RAgLMAJ4gAE7Uuwic+/GbZQn/gbdtVypxEgCfTvl7
BQNWfLaX5X9EegqyhI9yhpM=
=HVTG
-----END PGP SIGNATURE-----

--itqfrb9Qq3wY07cp--