On Fri, Aug 10, 2018 at 09:08:02PM +0300, Aapo Vienamo wrote:

Hi all,

This series implements support for faster signaling modes on Tegra
SDHCI controllers. This series consist of several parts: changes
requried for 1.8 V signaling and pad control, pad calibration, and
tuning. Following earlies patch sets have been merged into this
larger set: "Tegra PMC pinctrl pad configuration", "Tegra SDHCI enable
1.8 V signaling on Tegar210 and Tegra186", "Tegra SDHCI update the
padautocal procedure". Also the patches for enabling SDHCI tuning
are added.

Changelog:
v2:
	- Fix grammar in PMC device tree bindings docs
	- Remove a stray line from tegra sdhci bindings
	- Cosmetic changes to PMC pinctrl driver
	- Fix a typo in "soc/tegra: pmc: Implement
	  tegra_io_pad_is_powered()" commit message
	- Declare mask and value on the same line in
	  tegra_io_pad_is_powered()
	- Move the call to tegra_sdhci_is_pad_and_regulator_valid() to
	  inside the if condition in tegra_sdhci_reset()
	- Use usleep_range() in tegra_sdhci_configure_cal_pad()
	- Move sdhci_writel() out of the enable if-else body in
	  tegra_sdhci_configure_cal_pad()
	- Add a delay before starting polling in
	  tegra_sdhci_pad_autocalib()
	- Use usleep_range() in tegra_sdhci_set_tap()
	- Rename orig_enabled to status in
	  tegra_sdhci_configure_card_clk()
	- Fix if condition wrapping alignment in tegra_sdhci_set_tap()

v1:
	- Probe the regulator voltage capabilities to determine whether pinctrl
	  is needed in tegra_sdhci_r eset
	- Don't remove tegra_sdhci_voltage_switch()
	- Use dev_warn() in tegra_sdhci_init_pinctrl_info()
	- Don't change start_signal_voltage_switch callback if pinctrl info
	  invalid
	- Only call udelay(1) on enable in tegra_sdhci_configure_cal_pad()
	- Add nvidia, prefix to pad autocal offset dt props in the example

See the original patch sets for earlier changelogs.

Aapo Vienamo (40):
  dt-bindings: Add Tegra PMC pad configuration bindings
  dt-bindings: mmc: tegra: Add pad voltage control properties
  dt-bindings: Add Tegra SDHCI pad pdpu offset bindings
  dt-bindings: mmc: Add Tegra SDHCI sampling trimmer values
  soc/tegra: pmc: Fix pad voltage configuration for Tegra186
  soc/tegra: pmc: Factor out DPD register bit calculation
  soc/tegra: pmc: Implement tegra_io_pad_is_powered()
  soc/tegra: pmc: Use X macro to generate IO pad tables
  soc/tegra: pmc: Remove public pad voltage APIs
  soc/tegra: pmc: Implement pad configuration via pinctrl
  mmc: sdhci: Add a quirk to skip clearing the transfer mode register on
    tuning
  mmc: tegra: Reconfigure pad voltages during voltage switching
  mmc: tegra: Poll for calibration completion
  mmc: tegra: Set calibration pad voltage reference
  mmc: tegra: Power on the calibration pad
  mmc: tegra: Disable card clock during pad calibration
  mmc: tegra: Program pad autocal offsets from dt
  mmc: tegra: Perform pad calibration after voltage switch
  mmc: tegra: Enable pad calibration on Tegra210 and Tegra186
  mmc: tegra: Add a workaround for tap value change glitch
  mmc: tegra: Parse default trim and tap from dt
  mmc: tegra: Configure default tap values
  mmc: tegra: Configure default trim value on reset
  mmc: tegra: Use standard SDHCI tuning on Tegra210 and Tegra186
  mmc: sdhci: Add a quirk to disable card clock during tuning
  mmc: tegra: Enable workaround for tuning transfer mode bug
  mmc: tegra: Set SDHCI_QUIRK2_TUNE_DIS_CARD_CLK on Tegra210
  mmc: tegra: Enable UHS and HS200 modes for Tegra210
  mmc: tegra: Enable UHS and HS200 modes for Tegra186
  arm64: dts: Add Tegra210 sdmmc pinctrl voltage states
  arm64: dts: Add Tegra186 sdmmc pinctrl voltage states
  arm64: dts: tegra210-p2180: Allow ldo2 to go down to 1.8 V
  arm64: dts: tegra210-p2180: Correct sdmmc4 vqmmc-supply
  arm64: dts: tegra210-p2597: Remove no-1-8-v from sdmmc1
  arm64: dts: tegra186: Add sdmmc pad auto calibration offsets
  arm64: dts: tegra210: Add sdmmc pad auto calibration offsets
  arm64: dts: tegra210: Add SDHCI tap and trim values
  arm64: dts: tegra186: Add SDHCI tap and trim values
  arm64: dts: tegra186: Assign clocks for sdmmc1 and sdmmc4
  arm64: dts: tegra210: Assign clocks for sdmmc1 and sdmmc4

 .../bindings/arm/tegra/nvidia,tegra186-pmc.txt     |  93 ++++
 .../bindings/arm/tegra/nvidia,tegra20-pmc.txt      | 103 ++++
 .../bindings/mmc/nvidia,tegra20-sdhci.txt          |  68 +++
 arch/arm64/boot/dts/nvidia/tegra186.dtsi           |  74 +++
 arch/arm64/boot/dts/nvidia/tegra210-p2180.dtsi     |  12 +-
 arch/arm64/boot/dts/nvidia/tegra210-p2597.dtsi     |   1 -
 arch/arm64/boot/dts/nvidia/tegra210.dtsi           |  55 ++
 drivers/mmc/host/sdhci-tegra.c                     | 556 +++++++++++++++++++--
 drivers/mmc/host/sdhci.c                           |  21 +
 drivers/mmc/host/sdhci.h                           |   4 +
 drivers/soc/tegra/pmc.c                            | 511 ++++++++++++++-----
 include/dt-bindings/pinctrl/pinctrl-tegra-io-pad.h |  18 +
 include/soc/tegra/pmc.h                            |  20 +-
 13 files changed, 1324 insertions(+), 212 deletions(-)
 create mode 100644 include/dt-bindings/pinctrl/pinctrl-tegra-io-pad.h

Hi Ulf, Adrian,

I've been running some tests on this on various Tegra boards and this
all seems to work fine. I'm also pretty happy with how this turned out,
so the whole series is:

Acked-by: Thierry Reding <redacted>

In order to get this merged, I think it'd be best for you guys to pick
up the MMC patches, since they are all independent of the rest. There is
a runtime dependency on the PMC bits for the faster modes, but things
will fallback to the status quo if those bits are not enabled. The only
dependency here is between the PMC and DTS changes, which I usually pick
up into the Tegra tree anyway, so I can sort that out there.

Since I've already gone through the process of splitting up the series,
I can offer to rebase the MMC patches on top of v4.19-rc1 when it's out
and send a pull request for the MMC patches. Alternatively I can further
split that branch up into two parts, one with the two patches that touch
the core and another with the remainder, if that's what you prefer. Or
you can of course feel free to apply all of the MMC patches yourselves
with my above Acked-by if you want to.

Any thought on the above?

Thanks,
Thierry