search for: therm

As you changed the return value of `temp_get()` to solely be the error code, or absence of an error, I would change all those tests that checked whether the returned value was strictly less, or greater than, 0 to now only compare against 0 (no error). For example, if (therm && therm->attr_get && nvkm_therm_temp_get(therm, &val) < 0) if (therm->func->temp_get(therm, &val) >= 0) could become if (therm && therm->attr_get && nvkm_therm_temp_get(therm, &val)) if (!therm->func->temp_get(therm, &...

The temp_get() function currently returns negative error numbers or a temperature. However, the thermal sensors can (in theory) measure negative temperatures. Some implementations of temp_get() correctly clamp negative temperature readings to 0 so that users do not mistake them for errors, but some, like gp100_temp_get(), do not. Rather than relying on implementations remembering to clamp values,...

...don't even know on how many bits 0x20400 uses... If you are indeed right, I would like to see g84.c's temp_get(). > > Signed-off-by: Karol Herbst <karolherbst at gmail.com> > --- > drm/nouveau/include/nvkm/subdev/clk.h | 4 ++-- > drm/nouveau/include/nvkm/subdev/therm.h | 2 +- > drm/nouveau/nouveau_hwmon.c | 15 +++++++++------ > drm/nouveau/nvkm/subdev/clk/base.c | 2 +- > drm/nouveau/nvkm/subdev/therm/base.c | 19 ++++++++++++++----- > drm/nouveau/nvkm/subdev/therm/g84.c | 13 ++++++++----- > drm/nouveau/nvkm/subdev/th...

On Wed, Sep 16, 2020 at 10:01 PM Karol Herbst <kherbst at redhat.com> wrote: > > On Wed, Sep 16, 2020 at 9:47 PM Jeremy Cline <jcline at redhat.com> wrote: > > > > The temp_get() function currently returns negative error numbers or a > > temperature. However, the thermal sensors can (in theory) measure > > negative temperatures. Some implementations of temp_get() correctly > > clamp negative temperature readings to 0 so that users do not mistake > > them for errors, but some, like gp100_temp_get(), do not. > > > > Rather than relying...

...egative values. Since Pascal (and maybe earlier) we have sensors with improved precision. Adjust the nvkm_get_temp method to be able to deal with those changes and let hwmon return an error properly. Signed-off-by: Karol Herbst <karolherbst at gmail.com> --- drm/nouveau/include/nvkm/subdev/therm.h | 2 +- drm/nouveau/nouveau_hwmon.c | 15 +++++++++------ drm/nouveau/nvkm/subdev/therm/base.c | 19 ++++++++++++++----- drm/nouveau/nvkm/subdev/therm/g84.c | 11 ++++++----- drm/nouveau/nvkm/subdev/therm/gp100.c | 9 +++++---- drm/nouveau/nvkm/subdev/therm/nv40.c | 9 +...

...w many bits >> 0x20400 uses... >> >> If you are indeed right, I would like to see g84.c's temp_get(). >> > > Nvidia uses 0x020460 on Pascal and I thought we had that in envytools already? Nothing in nvbios. > > [0] 228.412618 MMIO32 R 0x020460 0x20002f20 PTHERM.I2C_SLAVE+0x60 => 0x20002f20 > [0] 229.413039 MMIO32 R 0x020460 0x20002f10 PTHERM.I2C_SLAVE+0x60 => 0x20002f10 > [0] 230.416976 MMIO32 R 0x020460 0x20002f68 PTHERM.I2C_SLAVE+0x60 => 0x20002f68 > [0] 231.417407 MMIO32 R 0x020460 0x20002fe8 PTHERM.I2C_SLAVE+0x60 => 0x20002fe8 &gt...

Commit d32656373857 ("drm/nouveau/therm/gp100: initial implementation of new gp1xx temperature sensor") added support for reading finer-grain temperatures, but continued to report temperatures in 1 degree Celsius increments via nvkm_therm_temp_get(). Rather than altering nvkm_therm_temp_get() to report finer-grain temperatures, whi...

...have sensors with improved precision. Adjust the nvkm_get_temp method to be able to deal with those changes and let hwmon return an error properly. Signed-off-by: Karol Herbst <karolherbst at gmail.com> --- drm/nouveau/include/nvkm/subdev/clk.h | 4 ++-- drm/nouveau/include/nvkm/subdev/therm.h | 2 +- drm/nouveau/nouveau_hwmon.c | 15 +++++++++------ drm/nouveau/nvkm/subdev/clk/base.c | 2 +- drm/nouveau/nvkm/subdev/therm/base.c | 19 ++++++++++++++----- drm/nouveau/nvkm/subdev/therm/g84.c | 13 ++++++++----- drm/nouveau/nvkm/subdev/therm/gp100.c | 9 +++++-...

...y screw up the value because we don't even know on how many bits > 0x20400 uses... > > If you are indeed right, I would like to see g84.c's temp_get(). > Nvidia uses 0x020460 on Pascal and I thought we had that in envytools already? [0] 228.412618 MMIO32 R 0x020460 0x20002f20 PTHERM.I2C_SLAVE+0x60 => 0x20002f20 [0] 229.413039 MMIO32 R 0x020460 0x20002f10 PTHERM.I2C_SLAVE+0x60 => 0x20002f10 [0] 230.416976 MMIO32 R 0x020460 0x20002f68 PTHERM.I2C_SLAVE+0x60 => 0x20002f68 [0] 231.417407 MMIO32 R 0x020460 0x20002fe8 PTHERM.I2C_SLAVE+0x60 => 0x20002fe8 [0] 232.417742 MMI...

...pci-0100 Adapter: PCI adapter GPU core: N/A (min = +0.60 V, max = +1.20 V) temp1: N/A (high = +95.0°C, hyst = +3.0°C) (crit = +105.0°C, hyst = +5.0°C) (emerg = +135.0°C, hyst = +5.0°C) power1: N/A Karol Herbst (3): therm: split return code and value in nvkm_get_temp hwmon: properly check for errors subdev/volt/gk104: return error when read fails drm/nouveau/include/nvkm/subdev/therm.h | 2 +- drm/nouveau/nouveau_hwmon.c | 48 ++++++++++++++++++++++++--------- drm/nouveau/nvkm/subdev/therm/base.c...

On Wed, Sep 16, 2020 at 9:47 PM Jeremy Cline <jcline at redhat.com> wrote: > > The temp_get() function currently returns negative error numbers or a > temperature. However, the thermal sensors can (in theory) measure > negative temperatures. Some implementations of temp_get() correctly > clamp negative temperature readings to 0 so that users do not mistake > them for errors, but some, like gp100_temp_get(), do not. > > Rather than relying on implementations remem...

...0:01 PM Karol Herbst <kherbst at redhat.com> wrote: > > > > On Wed, Sep 16, 2020 at 9:47 PM Jeremy Cline <jcline at redhat.com> wrote: > > > > > > The temp_get() function currently returns negative error numbers or a > > > temperature. However, the thermal sensors can (in theory) measure > > > negative temperatures. Some implementations of temp_get() correctly > > > clamp negative temperature readings to 0 so that users do not mistake > > > them for errors, but some, like gp100_temp_get(), do not. > > > > >...

This seems to be absolutely necessary for a lot of NV40. Reported-by: gsgf on IRC/freenode Signed-off-by: Martin Peres <martin.peres at free.fr> --- drm/nouveau/nvkm/subdev/therm/base.c | 22 +++++++++++++++++++--- 1 file changed, 19 insertions(+), 3 deletions(-) diff --git a/drm/nouveau/nvkm/subdev/therm/base.c b/drm/nouveau/nvkm/subdev/therm/base.c index 8894fee..df949fa 100644 --- a/drm/nouveau/nvkm/subdev/therm/base.c +++ b/drm/nouveau/nvkm/subdev/therm/base.c @@ -64,1...

...te: > This adds support for enabling automatic clockgating on nvidia GPUs for > Kepler1. While this is not technically a clockgating level, it does > enable clockgating using the clockgating values initially set by the > vbios (which should be safe to use). > > This introduces two therm helpers for controlling basic clockgating: > nvkm_therm_clkgate_enable() - enables clockgating through > CG_CTRL, done after initializing the GPU fully > nvkm_therm_clkgate_fini() - prepares clockgating for suspend or > driver unload > > As well, we...

From: Thierry Reding <treding at nvidia.com> The recently introduced clock gate support breaks on Tegra chips because no thermal support is enabled for those devices. Conditionalize the code on the existence of thermal support to fix this. Fixes: b138eca661cc ("drm/nouveau: Add support for basic clockgating on Kepler1") Cc: Martin Peres <martin.peres at free.fr> Cc: Lyude Paul <lyude at redhat.com> S...

...my kepler's idle power consumption from ~23.6W to ~21.65W. Similar to how the nvidia driver seems to handle this, we enable clockgating for each engine that supports it after it's initialization. Signed-off-by: Lyude <lyude at redhat.com> --- .../gpu/drm/nouveau/include/nvkm/subdev/therm.h | 4 ++ drivers/gpu/drm/nouveau/nvkm/core/engine.c | 20 +++++- drivers/gpu/drm/nouveau/nvkm/engine/device/base.c | 14 ++-- drivers/gpu/drm/nouveau/nvkm/subdev/therm/Kbuild | 2 + drivers/gpu/drm/nouveau/nvkm/subdev/therm/base.c | 2 + .../gpu/drm/nouveau/nvkm/subdev/therm/clk...

Signed-off-by: Rhys Kidd <rhyskidd at gmail.com> --- .../gpu/drm/nouveau/include/nvkm/subdev/therm.h | 1 + drivers/gpu/drm/nouveau/nvkm/engine/device/base.c | 6 +++ drivers/gpu/drm/nouveau/nvkm/subdev/therm/Kbuild | 1 + drivers/gpu/drm/nouveau/nvkm/subdev/therm/base.c | 3 +- drivers/gpu/drm/nouveau/nvkm/subdev/therm/gp100.c | 57 ++++++++++++++++++++++ 5 files changed, 67 insert...

...gt; Similar to how the nvidia driver seems to handle this, we enable > > clockgating for each engine that supports it after it's > > initialization. > > > > Signed-off-by: Lyude <lyude at redhat.com> > > --- > >  .../gpu/drm/nouveau/include/nvkm/subdev/therm.h    |  4 ++ > >  drivers/gpu/drm/nouveau/nvkm/core/engine.c         | 20 +++++- > >  drivers/gpu/drm/nouveau/nvkm/engine/device/base.c  | 14 ++-- > >  drivers/gpu/drm/nouveau/nvkm/subdev/therm/Kbuild   |  2 + > >  drivers/gpu/drm/nouveau/nvkm/subdev/therm/base.c   |  2 + &g...

This adds support for enabling automatic clockgating on nvidia GPUs for Kepler1. While this is not technically a clockgating level, it does enable clockgating using the clockgating values initially set by the vbios (which should be safe to use). This introduces two therm helpers for controlling basic clockgating: nvkm_therm_clkgate_enable() - enables clockgating through CG_CTRL, done after initializing the GPU fully nvkm_therm_clkgate_fini() - prepares clockgating for suspend or driver unload As well, we add the nouveau kernel config parameter NvPmEnableGating...

This adds support for enabling automatic clockgating on nvidia GPUs for Kepler1. While this is not technically a clockgating level, it does enable clockgating using the clockgating values initially set by the vbios (which should be safe to use). This introduces two therm helpers for controlling basic clockgating: nvkm_therm_clkgate_enable() - enables clockgating through CG_CTRL, done after initializing the GPU fully nvkm_therm_clkgate_fini() - prepares clockgating for suspend or driver unload As well, we add the nouveau kernel config parameter NvPmEnableGating...