Re: [PATCH] make psaux reconnect adjustable

Kurt Garloff (garloff@suse.de)
Tue, 14 Aug 2001 19:35:43 +0200


--GOaLjq+VdFesH+wR
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

Hi Linus,

thanks for your comments.

On Tue, Aug 14, 2001 at 09:58:55AM -0700, Linus Torvalds wrote:
> I really have two comments, but I haven't followed the whole discussion,
> so feel free to just say that it's been hashed out already:
>=20
> - sysconf entries are suspicious for stuff like this. If some code really
> requires this to work correctly, that's exactly the kind of code that
> would run automatically at bootup. A sysconf doesn't really help people
> in that case - we'd be much better off with just a bootup switch.

Maybe that's the difference of whether you define the kernel behaviour by
deciding what goes in and the one that tries to avoid a breakage without
changing the default behaviour ...
Of course the sysctl is less intrusive (from a user's point of view).

> - do we actually need the config switch AT ALL, whether at bootup or not?
> What exactly breaks if we just always pass the AA 00 values through?
> Apparently nothing ever breaks, which makes me suspect that people are
> just being unnecessarily defensive.

PS2 mouses need this ping thing to be operational after being plugged.
But, there's no reason it needs to be done in the kernel.
* It works for plain PS2 only (not: imps2, synps2, ...)
* The userspace driver (gpm, X11) can do it as well, AFAICS

I guess the thing has been introduced because it was more convenient than
fixing userspace.
Maybe the kernel does it a bit more efficient by throwing away the queue ...

> In short, I'd prefer a patch that just unconditionally removes the code,
> unless somebody KNOWS that it could break something. That failing, a
> simple kernel command line option sounds better than more files in /proc.

I'd be happy with removing it.
Patch will follow!

> Remember: the biggest mistake to do is to overdesign. The road to hell is
> paved with good intentions.

A few sysctls don't qualify as overdesign yet, I hope.

Regards,
--=20
Kurt Garloff <garloff@suse.de> Eindhoven, NL
GPG key: See mail header, key servers Linux kernel development
SuSE GmbH, Nuernberg, DE SCSI, Security

--GOaLjq+VdFesH+wR
Content-Type: application/pgp-signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (GNU/Linux)
Comment: For info see http://www.gnupg.org

iD8DBQE7eWFvxmLh6hyYd04RAid0AJkBRbQAelKjAXzaA8rRQifEPscH2ACfUUsp
BqZqeLRrmag8Bsup3yELdnY=
=dhec
-----END PGP SIGNATURE-----

--GOaLjq+VdFesH+wR--
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/