From mboxrd@z Thu Jan 1 00:00:00 1970 From: Wei Ni Subject: Re: [PATCH V7 03/12] thermal: tegra: get rid of PDIV/HOTSPOT hack Date: Wed, 16 Mar 2016 11:48:48 +0800 Message-ID: <56E8D7A0.9070704@nvidia.com> References: <1457665763-29781-1-git-send-email-wni@nvidia.com> <1457665763-29781-4-git-send-email-wni@nvidia.com> <20160314200539.GH1872@localhost.localdomain> <56E7AA01.4000305@nvidia.com> <20160315195624.GE26619@localhost.localdomain> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: QUOTED-PRINTABLE Return-path: In-Reply-To: <20160315195624.GE26619@localhost.localdomain> Sender: linux-kernel-owner@vger.kernel.org To: Eduardo Valentin Cc: rui.zhang@intel.com, thierry.reding@gmail.com, MLongnecker@nvidia.com, swarren@wwwdotorg.org, mikko.perttunen@kapsi.fi, linux-tegra@vger.kernel.org, linux-pm@vger.kernel.org, linux-kernel@vger.kernel.org List-Id: linux-tegra@vger.kernel.org On 2016=E5=B9=B403=E6=9C=8816=E6=97=A5 03:56, Eduardo Valentin wrote: > * PGP Signed by an unknown key >=20 > On Tue, Mar 15, 2016 at 02:21:53PM +0800, Wei Ni wrote: >> >> >> On 2016=E5=B9=B403=E6=9C=8815=E6=97=A5 04:05, Eduardo Valentin wrote= : >>>> Old Signed by an unknown key >>> >>> On Fri, Mar 11, 2016 at 11:09:14AM +0800, Wei Ni wrote: >>>> Get rid of T124-specific PDIV/HOTSPOT hack. >>>> tegra-soctherm.c contained a hack to set the SENSOR_PDIV and >>>> SENSOR_HOTSPOT_OFFSET registers - it just did two writes of >>>> T124-specific opaque values. Convert these into a form that can b= e >>>> substituted on a per-chip basis, and into structure fields that ha= ve >>>> at least some independent meaning. >>>> >>>> Signed-off-by: Wei Ni >>>> --- >>>> drivers/thermal/tegra/tegra-soctherm.c | 18 ++++++++++++++---- >>>> 1 file changed, 14 insertions(+), 4 deletions(-) >>>> >>>> diff --git a/drivers/thermal/tegra/tegra-soctherm.c b/drivers/ther= mal/tegra/tegra-soctherm.c >>>> index b3ec0faa2bee..b4b791ebfbb6 100644 >>>> --- a/drivers/thermal/tegra/tegra-soctherm.c >>>> +++ b/drivers/thermal/tegra/tegra-soctherm.c >>>> @@ -48,14 +48,12 @@ >>>> #define SENSOR_CONFIG2_THERMB_SHIFT 0 >>>> =20 >>>> #define SENSOR_PDIV 0x1c0 >>>> -#define SENSOR_PDIV_T124 0x8888 >>>> #define SENSOR_PDIV_CPU_MASK (0xf << 12) >>>> #define SENSOR_PDIV_GPU_MASK (0xf << 8) >>>> #define SENSOR_PDIV_MEM_MASK (0xf << 4) >>>> #define SENSOR_PDIV_PLLX_MASK (0xf << 0) >>>> =20 >>>> #define SENSOR_HOTSPOT_OFF 0x1c4 >>>> -#define SENSOR_HOTSPOT_OFF_T124 0x00060600 >>>> #define SENSOR_HOTSPOT_CPU_MASK (0xff << 16) >>>> #define SENSOR_HOTSPOT_GPU_MASK (0xff << 8) >>>> #define SENSOR_HOTSPOT_MEM_MASK (0xff << 0) >>>> @@ -436,6 +434,7 @@ static int tegra_soctherm_probe(struct platfor= m_device *pdev) >>>> struct resource *res; >>>> unsigned int i; >>>> int err; >>>> + u32 pdiv, hotspot; >>>> =20 >>>> const struct tegra_tsensor *tsensors =3D t124_tsensors; >>>> const struct tegra_tsensor_group **ttgs =3D tegra124_tsensor_gro= ups; >>>> @@ -493,8 +492,19 @@ static int tegra_soctherm_probe(struct platfo= rm_device *pdev) >>>> goto disable_clocks; >>>> } >>>> =20 >>>> - writel(SENSOR_PDIV_T124, tegra->regs + SENSOR_PDIV); >>>> - writel(SENSOR_HOTSPOT_OFF_T124, tegra->regs + SENSOR_HOTSPOT_OFF= ); >>>> + /* Program pdiv and hotspot offsets per THERM */ >>>> + pdiv =3D readl(tegra->regs + SENSOR_PDIV); >>>> + hotspot =3D readl(tegra->regs + SENSOR_HOTSPOT_OFF); >>>> + for (i =3D 0; i < TEGRA124_SOCTHERM_SENSOR_NUM; ++i) { >>>> + pdiv =3D REG_SET_MASK(pdiv, ttgs[i]->pdiv_mask, >>>> + ttgs[i]->pdiv); >>>> + if (ttgs[i]->id !=3D TEGRA124_SOCTHERM_SENSOR_PLLX) >>>> + hotspot =3D REG_SET_MASK(hotspot, >>>> + ttgs[i]->pllx_hotspot_mask, >>>> + ttgs[i]->pllx_hotspot_diff); >>>> + } >>>> + writel(pdiv, tegra->regs + SENSOR_PDIV); >>>> + writel(hotspot, tegra->regs + SENSOR_HOTSPOT_OFF); >>> >>> Is the above logic the same for all supported chips? e.g. do we alw= ays >>> skip pllx for hotspot configuration? >> >> Yes, this logic support Tegra124, Tegra210, and Tegra132 which I wil= l send out >> in next series. >=20 >=20 > Ok. Could you please add a comment then explaining why pllx is not > needed for the hotspot configuration? This is the hotspot offset from PLLX, so we doesn't need to configure f= or PLLX, the Tegra DRM introduced it. I will add comment for it. >=20 >> >>> >>> >>>> =20 >>>> /* Initialize thermctl sensors */ >>>> =20 >>>> --=20 >>>> 1.9.1 >>>> >>> >>> * Unknown Key >>> * 0x7DA4E256 >>> >=20 > * Unknown Key > * 0x7DA4E256 >=20