From mboxrd@z Thu Jan 1 00:00:00 1970 From: Thierry Reding Subject: Re: [PATCH] backlight: add PWM dependencies Date: Wed, 26 Feb 2014 14:25:30 +0100 Message-ID: <20140226132529.GC29779@ulmo.nvidia.com> References: <1391518634-6472-1-git-send-email-linus.walleij@linaro.org> <20140210104032.GB20143@ulmo.nvidia.com> Mime-Version: 1.0 Content-Type: multipart/signed; micalg=pgp-sha1; protocol="application/pgp-signature"; boundary="VywGB/WGlW4DM4P8" Return-path: Received: from mail-ee0-f42.google.com ([74.125.83.42]:33938 "EHLO mail-ee0-f42.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751947AbaBZNZe (ORCPT ); Wed, 26 Feb 2014 08:25:34 -0500 Content-Disposition: inline In-Reply-To: Sender: linux-pwm-owner@vger.kernel.org List-Id: linux-pwm@vger.kernel.org To: Linus Walleij Cc: Jingoo Han , Arnd Bergmann , "linux-kernel@vger.kernel.org" , "linux-arm-kernel@lists.infradead.org" , "linux-pwm@vger.kernel.org" --VywGB/WGlW4DM4P8 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Content-Transfer-Encoding: quoted-printable On Mon, Feb 10, 2014 at 12:09:29PM +0100, Linus Walleij wrote: > On Mon, Feb 10, 2014 at 11:40 AM, Thierry Reding > wrote: >=20 > > it seems like at least BACKLIGHT_LP8788 is missing a corresponding > > dependency as well. > > > > I have applied Sascha's patch to remove the obsolete HAVE_PWM symbol, > > and this will fix at least the build issues. However it will also cause > > the driver to fail at runtime because the pwm_*() functions won't work. >=20 > So it definately needs that API, not just stubs. >=20 > But isn't it proper for Kconfig to allow you to break things > like that by configuring out stuff and have stubs come in? >=20 > I'm a bit torn here. After thinking about this some more, I've come to the conclusion that the drivers should have the dependency. Kconfig should make sure that the resulting drivers work. If somebody wanted to knowingly build this driver with a configuration that won't work at runtime, then they should be using COMPILE_TEST instead. So I'm leaning towards applying this patch if there are no objections. Thierry --VywGB/WGlW4DM4P8 Content-Type: application/pgp-signature -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.22 (GNU/Linux) iQIcBAEBAgAGBQJTDetJAAoJEN0jrNd/PrOh4coP/Als6SaArd/CT4bpdFqdKj3Z QjwO4TnR8rKVpdgscGHGJmvdIJMvGP1euoYIFtgpOw4BccZfPJtYVNVLWWS+ovby jCmcj66wa++tsDsZWwVVVXOOxyBDTKzPuIBnAK3SQar4cntzufasiBqIjAchSx4l uj089YlQ99zq7PD0lTs4tzmkKqDd1UqkwqUZrjOmHdfQDSikjBv4bjjrmYCfbFcl 2LLNcpRb7DnlYW5oKWwV9XC45LZ0YdXooVsoYrYDqqGXsHMcnxx3d6TTx8eMCshU UiyzNrNcvPaJx78xNhVagoQ+P9GididRqtnoahofgLmPYZmtZK9rOWwmoFLFLZeA +q3UTCW+iZou5YcXy0Zep6qrLRd9PAf5BcxEYf+gTVpZINwKXGY8W5GkVYZWnMKG /AAfF/PeUUm3wtfgN5zhfR1WwWdaHTyV61waI82c0hR2q2TiHxsQM/w6HF50HpD/ QBpjgZrtVMbrvE2MfcSZKQnBmuJKOC5PZtcc48pGlZ2bWn2cbog0oo1k2zwN5qrd h3MnIwAbZRBxJJ4/aTxhPerBhwHGB6nkLnKK1THiBUUvXyLoqk8uph8XRFOAfY9J DTcLSXNUeuX3TPBdPx158VdiYGS1OZKxZ2GUGWcSNCNRgpdg5vlnP79jaZaMatsJ 5O7fgf54k4kc5YN7zwc8 =F3kA -----END PGP SIGNATURE----- --VywGB/WGlW4DM4P8-- From mboxrd@z Thu Jan 1 00:00:00 1970 From: thierry.reding@gmail.com (Thierry Reding) Date: Wed, 26 Feb 2014 14:25:30 +0100 Subject: [PATCH] backlight: add PWM dependencies In-Reply-To: References: <1391518634-6472-1-git-send-email-linus.walleij@linaro.org> <20140210104032.GB20143@ulmo.nvidia.com> Message-ID: <20140226132529.GC29779@ulmo.nvidia.com> To: linux-arm-kernel@lists.infradead.org List-Id: linux-arm-kernel.lists.infradead.org On Mon, Feb 10, 2014 at 12:09:29PM +0100, Linus Walleij wrote: > On Mon, Feb 10, 2014 at 11:40 AM, Thierry Reding > wrote: > > > it seems like at least BACKLIGHT_LP8788 is missing a corresponding > > dependency as well. > > > > I have applied Sascha's patch to remove the obsolete HAVE_PWM symbol, > > and this will fix at least the build issues. However it will also cause > > the driver to fail at runtime because the pwm_*() functions won't work. > > So it definately needs that API, not just stubs. > > But isn't it proper for Kconfig to allow you to break things > like that by configuring out stuff and have stubs come in? > > I'm a bit torn here. After thinking about this some more, I've come to the conclusion that the drivers should have the dependency. Kconfig should make sure that the resulting drivers work. If somebody wanted to knowingly build this driver with a configuration that won't work at runtime, then they should be using COMPILE_TEST instead. So I'm leaning towards applying this patch if there are no objections. Thierry -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 836 bytes Desc: not available URL: