From mboxrd@z Thu Jan 1 00:00:00 1970 From: Darren Hart Subject: Re: [PATCH] pwm_lpss: Add support for PCI devices Date: Mon, 12 May 2014 16:18:16 -0700 Message-ID: <20140512231816.GA65896@dvhart-mac01.jf.intel.com> References: <1397311131-13371-1-git-send-email-chiau.ee.chew@intel.com> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: QUOTED-PRINTABLE Return-path: Received: from mga02.intel.com ([134.134.136.20]:51894 "EHLO mga02.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751288AbaELXR6 (ORCPT ); Mon, 12 May 2014 19:17:58 -0400 Content-Disposition: inline In-Reply-To: <1397311131-13371-1-git-send-email-chiau.ee.chew@intel.com> Sender: linux-pwm-owner@vger.kernel.org List-Id: linux-pwm@vger.kernel.org To: Chew Chiau Ee Cc: Thierry Reding , Alan Cox , Mika Westerberg , linux-pwm@vger.kernel.org, linux-kernel@vger.kernel.org On Sat, Apr 12, 2014 at 09:58:51PM +0800, Chew Chiau Ee wrote: > From: Alan Cox >=20 > Not all systems enumerate the PWM devices via ACPI. They can also be = exposed > via the PCI interface. >=20 > Signed-off-by: Alan Cox > Signed-off-by: Chew, Chiau Ee > --- > drivers/pwm/pwm-lpss.c | 160 ++++++++++++++++++++++++++++++++++++++= --------- > 1 files changed, 129 insertions(+), 31 deletions(-) >=20 > diff --git a/drivers/pwm/pwm-lpss.c b/drivers/pwm/pwm-lpss.c > index 449e372..6f79bf8 100644 > --- a/drivers/pwm/pwm-lpss.c > +++ b/drivers/pwm/pwm-lpss.c > @@ -6,6 +6,7 @@ > * Author: Chew Kean Ho > * Author: Chang Rebecca Swee Fun > * Author: Chew Chiau Ee > + * Author: Alan Cox > * > * This program is free software; you can redistribute it and/or mod= ify > * it under the terms of the GNU General Public License version 2 as > @@ -19,6 +20,9 @@ > #include > #include > #include > +#include > + > +static int pci_drv, plat_drv; /* So we know which drivers registered= */ > =20 > #define PWM 0x00000000 > #define PWM_ENABLE BIT(31) > @@ -34,6 +38,15 @@ struct pwm_lpss_chip { > struct pwm_chip chip; > void __iomem *regs; > struct clk *clk; > + unsigned long clk_rate; > +}; > + > +struct pwm_lpss_boardinfo { > + unsigned long clk_rate; > +}; > + > +static const struct pwm_lpss_boardinfo byt_info =3D { > + 25000000 > }; > =20 > static inline struct pwm_lpss_chip *to_lpwm(struct pwm_chip *chip) > @@ -55,7 +68,7 @@ static int pwm_lpss_config(struct pwm_chip *chip, s= truct pwm_device *pwm, > /* The equation is: base_unit =3D ((freq / c) * 65536) + correction= */ > base_unit =3D freq * 65536; > =20 > - c =3D clk_get_rate(lpwm->clk); > + c =3D lpwm->clk_rate; > if (!c) > return -EINVAL; > =20 > @@ -113,52 +126,47 @@ static const struct pwm_ops pwm_lpss_ops =3D { > .owner =3D THIS_MODULE, > }; > =20 > -static const struct acpi_device_id pwm_lpss_acpi_match[] =3D { > - { "80860F09", 0 }, > - { }, > -}; > -MODULE_DEVICE_TABLE(acpi, pwm_lpss_acpi_match); > - > -static int pwm_lpss_probe(struct platform_device *pdev) > +static struct pwm_lpss_chip *pwm_lpss_probe(struct device *dev, > + struct resource *r, struct pwm_lpss_boardinfo *info) Has this landed anywhere? I didn't see it in mainline or next, am I loo= king in the wrong place? If it's just still pending, I ran into one issue integ= rating it with 3.14.2: CC [M] drivers/pwm/pwm-lpss.o drivers/pwm/pwm-lpss.c: In function =E2=80=98pwm_lpss_probe_pci=E2=80= =99: drivers/pwm/pwm-lpss.c:192:2: warning: passing argument 3 of =E2=80=98= pwm_lpss_probe=E2=80=99 discards =E2=80=98const=E2=80=99 qualifier from pointer target type [= enabled by default] drivers/pwm/pwm-lpss.c:130:30: note: expected =E2=80=98struct pwm_lps= s_boardinfo *=E2=80=99 but argument is of type =E2=80=98const struct pwm_lpss_boardinfo *=E2= =80=99 Can we make the third argument to pwm_lpss_probe a const? The following= is working for me: static struct pwm_lpss_chip *pwm_lpss_probe(struct device *dev, struct resource *r, const struct pwm_lpss_boar= dinfo *info) Thanks, -- Darren Hart Open Source Technology Center darren.hart@intel.com Intel Corporation