The spin lock is not necessary in set_ios. Anything that is racing with
changes to the I/O state is already broken. The mmc core already provides
synchronization via "claiming" the host. So remove spin_lock and friends
from sdhci_set_ios and related callbacks.
Signed-off-by: NAdrian Hunter <adrian.hunter@intel.com>
Signed-off-by: NUlf Hansson <ulf.hansson@linaro.org>
Tested-by: NLudovic Desroches <ludovic.desroches@microchip.com>

Drivers can use the host operation directly, so remove this now unused
callback.
Signed-off-by: NAdrian Hunter <adrian.hunter@intel.com>
Signed-off-by: NUlf Hansson <ulf.hansson@linaro.org>
Tested-by: NLudovic Desroches <ludovic.desroches@microchip.com>

Make use  of an Intel ACPI _DSM that provides eMMC driver strength.
Signed-off-by: NAdrian Hunter <adrian.hunter@intel.com>
Signed-off-by: NUlf Hansson <ulf.hansson@linaro.org>
Tested-by: NLudovic Desroches <ludovic.desroches@microchip.com>

Make use  of an Intel ACPI _DSM that indicates if re-tuning is needed after
D3.
Signed-off-by: NAdrian Hunter <adrian.hunter@intel.com>
Signed-off-by: NUlf Hansson <ulf.hansson@linaro.org>
Tested-by: NLudovic Desroches <ludovic.desroches@microchip.com>

Let devices define their own private data to facilitate device-specific
operations. The size of the private structure is specified in the
sdhci_pci_fixes structure, then sdhci_pci_probe_slot() will allocate extra
space for it, and sdhci_pci_priv() can be used to get a reference to it.
Signed-off-by: NAdrian Hunter <adrian.hunter@intel.com>
Signed-off-by: NUlf Hansson <ulf.hansson@linaro.org>
Tested-by: NLudovic Desroches <ludovic.desroches@microchip.com>

Devices might save and restore tuning values so that re-tuning might not be
needed after a pm transition.  Let drivers decide by pushing the
mmc_retune_needed() logic down to them.
Signed-off-by: NAdrian Hunter <adrian.hunter@intel.com>
Signed-off-by: NUlf Hansson <ulf.hansson@linaro.org>
Tested-by: NLudovic Desroches <ludovic.desroches@microchip.com>

The delay loops for reset and clock enable always take at least 1 ms
because they use mdelay(1). However they can take a lot less time e.g. less
than 50us. Use ktime and reduce the delay to 10 microseconds per loop.
Signed-off-by: NAdrian Hunter <adrian.hunter@intel.com>
Signed-off-by: NUlf Hansson <ulf.hansson@linaro.org>
Tested-by: NLudovic Desroches <ludovic.desroches@microchip.com>

The 2835 has two SD controllers: The Arasan sdhci controller (supported
by the iproc driver) and a custom sdhost controller.  This patch adds a
driver for the latter.

The sdhci controller supports both sdcard and sdio.  The sdhost
controller supports the sdcard only, but has better performance.  Also
note that the rpi3 has sdio wifi, so driving the sdcard with the sdhost
controller allows to use the sdhci controller for wifi support.

The configuration is done by devicetree via pin muxing.  Both SD
controller are available on the same pins (2 pin groups = pin 22 to 27 +
pin 48 to 53).  So it's possible to use both SD controllers at the same
time with different pin groups.

The code was originally written by Phil Elwell in the downstream
Rasbperry Pi tree.   In preparation for the upstream merge it was
cleaned up and the code base was moderized by Eric Anholt, Stefan
Wahren and Gerd Hoffmann.
Signed-off-by: NEric Anholt <eric@anholt.net>
Signed-off-by: NStefan Wahren <stefan.wahren@i2se.com>
Signed-off-by: NGerd Hoffmann <kraxel@redhat.com>
Signed-off-by: NUlf Hansson <ulf.hansson@linaro.org>

This is the other SD controller on the platform, which can be swapped
to the role of SD card host using pin muxing.
Signed-off-by: NEric Anholt <eric@anholt.net>
Signed-off-by: NGerd Hoffmann <kraxel@redhat.com>
Acked-by: NRob Herring <robh@kernel.org>
Signed-off-by: NUlf Hansson <ulf.hansson@linaro.org>

File contains multiple functions doing variations on the same thing,
sdio_readb(), sdio_writeb()f, sdio_readw(), sdio_writew()
etc. Although the functions have very similar logic the code is laid
out in a variety of ways. This makes it overly complicated to
read. There is a already a nice clean chunk of code, if we use this
format for all instances then we will have cleaned up the code,
reduced the line count and lessened the cognitive load required while
reading. Less lines equals less bugs.

Pick the most simple and clear code flow and change all functions to
be the same.
Signed-off-by: NTobin C. Harding <me@tobin.cc>
Signed-off-by: NUlf Hansson <ulf.hansson@linaro.org>

Various functions take as parameter an optional pointer. Pointer
should be guarded with non-NULL check before dereferencing.

Add non-NULL check before dereference of pointer.
Signed-off-by: NTobin C. Harding <me@tobin.cc>
Signed-off-by: NUlf Hansson <ulf.hansson@linaro.org>

This patch fixes the case where meson_mmc_probe function fails before
core_clk is enabled.
Originally, that would result in a try to disable the core clock, and if
it was not already enabled, it would result in a kernel warning.
This issue is similar to the one with div_clk.
Fix it by introducing another error path used only between successfully
enabling the core clock, and successfully enabling the div clock.
That would ensure that core clock is disabled only if it was enabled before.
Signed-off-by: NMichał Zegan <webczat@webczatnet.pl>
Signed-off-by: NUlf Hansson <ulf.hansson@linaro.org>

The SDHCI controller found on Tegra186 in very similar to the controller
found on earlier generations of Tegra.
Signed-off-by: NThierry Reding <treding@nvidia.com>
Signed-off-by: NUlf Hansson <ulf.hansson@linaro.org>

The list of compatible strings is somewhat difficult to read and extend.
Reformat it into a list to make it more easily extensible.

While at it, also remove the "plus one of the above" clause because it
isn't actually valid.
Signed-off-by: NThierry Reding <treding@nvidia.com>
Signed-off-by: NUlf Hansson <ulf.hansson@linaro.org>

sdhci_pltfm_init() has already set the platform drvdata for us, remove
the unnecessary platform_set_drvdata() call.
Signed-off-by: NJisheng Zhang <jszhang@marvell.com>
Acked-by: NPatrice Chotard <patrice.chotard@st.com>
Signed-off-by: NUlf Hansson <ulf.hansson@linaro.org>

sdhci_pltfm_init() has already set the platform drvdata for us, remove
the unnecessary platform_set_drvdata() call.
Signed-off-by: NJisheng Zhang <jszhang@marvell.com>
Signed-off-by: NUlf Hansson <ulf.hansson@linaro.org>

sdhci_pltfm_init() has already set the platform drvdata for us, remove
the unnecessary platform_set_drvdata() call.
Signed-off-by: NJisheng Zhang <jszhang@marvell.com>
Signed-off-by: NUlf Hansson <ulf.hansson@linaro.org>

Add support for HS400ES mode to Cadence SDHCI driver.
Signed-off-by: NPiotr Sroka <piotrs@cadence.com>
Reviewed-by: NMasahiro Yamada <yamada.masahiro@socionext.com>
Signed-off-by: NUlf Hansson <ulf.hansson@linaro.org>

The SDHCI controller found on NVIDIA Tegra186 SoCs is very similar to
the one on prior generations of Tegra and can be supported by the same
driver.
Acked-by: NAdrian Hunter <adrian.hunter@intel.com>
Signed-off-by: NThierry Reding <treding@nvidia.com>
Signed-off-by: NUlf Hansson <ulf.hansson@linaro.org>

The device tree binding for the SDHCI controller found on Tegra SoCs
specifies that a reset control can be provided by the device tree. No
code was ever added to support the module reset, which can cause the
driver to try and access registers from a module that's in reset. On
most Tegra SoC generations doing so would cause a hang.

Note that it's unlikely to see this happen because on most platforms
these resets will have been deasserted by the bootloader. However the
portability can be improved by making sure the driver deasserts the
reset before accessing any registers.

Since resets are synchronous on Tegra SoCs, the platform driver needs
to implement a custom ->remove() callback now to make sure the clock
is disabled after the reset is asserted.
Acked-by: NAdrian Hunter <adrian.hunter@intel.com>
Signed-off-by: NThierry Reding <treding@nvidia.com>
Signed-off-by: NUlf Hansson <ulf.hansson@linaro.org>

The mmc host was added in meson_mmc_probe, but never removed in
meson_mmc_remove.
Fix that by removing the host before deallocating other resources.
Signed-off-by: NMichał Zegan <webczat@webczatnet.pl>
Tested-by: NMichał Zegan <webczat@webczatnet.pl>
Acked-by: NKevin Hilman <khilman@baylibre.com>
Signed-off-by: NUlf Hansson <ulf.hansson@linaro.org>

After we received the dataend interrupt, R1 response register carries
the value from the automatically generated stop command. Report that
info back to the MMC block layer, so we will be notified in case of e.g.
ECC errors which happened during the last transfer.
Reviewed-by: NSimon Horman <horms+renesas@verge.net.au>
Tested-by: NYoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
Signed-off-by: NWolfram Sang <wsa+renesas@sang-engineering.com>
Signed-off-by: NUlf Hansson <ulf.hansson@linaro.org>

Halting the kernel on an unsupported stop command seems overkill, report
the error and say what we already did (due to autocmd12) instead.
Reviewed-by: NSimon Horman <horms+renesas@verge.net.au>
Signed-off-by: NWolfram Sang <wsa+renesas@sang-engineering.com>
Signed-off-by: NUlf Hansson <ulf.hansson@linaro.org>

Making sure we match the actual register names.
Reviewed-by: NSimon Horman <horms+renesas@verge.net.au>
Signed-off-by: NWolfram Sang <wsa+renesas@sang-engineering.com>
Signed-off-by: NUlf Hansson <ulf.hansson@linaro.org>

Reviewed-by: NSimon Horman <horms+renesas@verge.net.au>
Signed-off-by: NWolfram Sang <wsa+renesas@sang-engineering.com>
Signed-off-by: NUlf Hansson <ulf.hansson@linaro.org>

The current code assumes that DMA is finished before SD access end is
flagged. Thus, it schedules the 'dma_complete' tasklet in the SD card
interrupt routine when DATAEND is set. The assumption is not safe,
though. Even by mounting an SD card, it can be seen that sometimes DMA
complete is first, sometimes DATAEND. It seems they are usually close
enough timewise to not cause problems. However, a customer reported that
with CMD53 sometimes things really break apart. As a result, the BSP has
a patch which introduces flags for both events and makes sure both flags
are set before scheduling the tasklet. The customer accepted the patch,
yet it doesn't seem a proper upstream solution to me.

This patch refactors the code to replace the tasklet with already
existing and more lightweight mechanisms. First of all, we set the
callback in a DMA descriptor to automatically get notified when DMA is
done. In the callback, we then use a completion to make sure the SD
access has already ended. Then, we proceed as before.
Signed-off-by: NWolfram Sang <wsa+renesas@sang-engineering.com>
Signed-off-by: NUlf Hansson <ulf.hansson@linaro.org>

There is no check if clk_prepare_enable() succeed in sdhci_pxav2_probe().

Found by Linux Driver Verification project (linuxtesting.org).
Signed-off-by: NAlexey Khoroshilov <khoroshilov@ispras.ru>
Acked-by: NAdrian Hunter <adrian.hunter@intel.com>
Signed-off-by: NUlf Hansson <ulf.hansson@linaro.org>

Replace timeout magic numbers with proper constants.
Signed-off-by: NHeiner Kallweit <hkallweit1@gmail.com>
Acked-by: NKevin Hilman <khilman@baylibre.com>
Tested-by: NKevin Hilman <khilman@baylibre.com>
Signed-off-by: NUlf Hansson <ulf.hansson@linaro.org>

Struct mmc_command includes a reference to the related mmc_request.
Therefore we don't have to store mrq separately in struct meson_host.
And we can remove some now unneeded WARN_ON's.
Signed-off-by: NHeiner Kallweit <hkallweit1@gmail.com>
Acked-by: NKevin Hilman <khilman@baylibre.com>
Tested-by: NKevin Hilman <khilman@baylibre.com>
Signed-off-by: NUlf Hansson <ulf.hansson@linaro.org>

Config values which are not changed during runtime we can set
in the probe function already.
The block size setting is overwritten later in
meson_mmc_start_cmd anyway if needed, so it doesn't harm if we
remove this setting in meson_mmc_set_ios.

In addition write config register only if configuration changed.

Don't change the location of clock initialization as in an
earlier version of the patch, this change causes a hang.
This issue was reported and fix suggested by:
Helmut Klein <hgkr.klein@gmail.com>
Signed-off-by: NHeiner Kallweit <hkallweit1@gmail.com>
Acked-by: NKevin Hilman <khilman@baylibre.com>
Tested-by: NKevin Hilman <khilman@baylibre.com>
Signed-off-by: NUlf Hansson <ulf.hansson@linaro.org>

CLK core does a deep copy of init.name, therefore it's fully ok to
provide a local variable.
Signed-off-by: NHeiner Kallweit <hkallweit1@gmail.com>
Acked-by: NKevin Hilman <khilman@baylibre.com>
Tested-by: NKevin Hilman <khilman@baylibre.com>
Signed-off-by: NUlf Hansson <ulf.hansson@linaro.org>

The condition should be "if (ret)" as the disable/unprepare is
supposed to be executed if the previous command fails.
In addition adjust the error path in probe to properly deal
with the case that cfg_div_clk can be registered successfully
but enable/prepare fails.
In this case we shouldn't call clk_disable_unprepare.
Reported-by: NMichał Zegan <webczat_200@poczta.onet.pl>
Signed-off-by: NHeiner Kallweit <hkallweit1@gmail.com>
Acked-by: NKevin Hilman <khilman@baylibre.com>
Tested-by: NKevin Hilman <khilman@baylibre.com>
Signed-off-by: NUlf Hansson <ulf.hansson@linaro.org>

Member mux_parent isn't used outside meson_mmc_clk_init. So remove it
and replace it with a local variable in meson_mmc_clk_init.
Signed-off-by: NHeiner Kallweit <hkallweit1@gmail.com>
Acked-by: NKevin Hilman <khilman@baylibre.com>
Tested-by: NKevin Hilman <khilman@baylibre.com>
Signed-off-by: NUlf Hansson <ulf.hansson@linaro.org>

Because the DT requires a fixed number of mux parent clocks,
variable mux_parent_count can be replaced with constant
MUX_CLK_NUM_PARENTS, so remove it.
Signed-off-by: NHeiner Kallweit <hkallweit1@gmail.com>
Acked-by: NKevin Hilman <khilman@baylibre.com>
Tested-by: NKevin Hilman <khilman@baylibre.com>
Signed-off-by: NUlf Hansson <ulf.hansson@linaro.org>

Member ocr_mask is never used and member irq we can replace with a
local variable in meson_mmc_probe. So let's remove both members.
Signed-off-by: NHeiner Kallweit <hkallweit1@gmail.com>
Acked-by: NKevin Hilman <khilman@baylibre.com>
Tested-by: NKevin Hilman <khilman@baylibre.com>
Signed-off-by: NUlf Hansson <ulf.hansson@linaro.org>

The return value of meson_mmc_request_done and meson_mmc_read_resp
isn't used, so make both functions return void.
Signed-off-by: NHeiner Kallweit <hkallweit1@gmail.com>
Acked-by: NKevin Hilman <khilman@baylibre.com>
Tested-by: NKevin Hilman <khilman@baylibre.com>
Signed-off-by: NUlf Hansson <ulf.hansson@linaro.org>

Core ensures that there are no commands with cmd->data being set and
nothing to transfer. And we don't have to reset bit CMD_CFG_DATA_NUM
because cmd_cfg was zero-initialized and this bit isn't set.
Signed-off-by: NHeiner Kallweit <hkallweit1@gmail.com>
Acked-by: NKevin Hilman <khilman@baylibre.com>
Tested-by: NKevin Hilman <khilman@baylibre.com>
Signed-off-by: NUlf Hansson <ulf.hansson@linaro.org>

Too much condition iteration makes the code
less readable. Slightly improve it.
Signed-off-by: NShawn Lin <shawn.lin@rock-chips.com>
Signed-off-by: NJaehoon Chung <jh80.chung@samsung.com>
Signed-off-by: NUlf Hansson <ulf.hansson@linaro.org>

No functional change intended.
Signed-off-by: NShawn Lin <shawn.lin@rock-chips.com>
Signed-off-by: NJaehoon Chung <jh80.chung@samsung.com>
Signed-off-by: NUlf Hansson <ulf.hansson@linaro.org>

No need to declar it there, remove it.
Signed-off-by: NShawn Lin <shawn.lin@rock-chips.com>
Signed-off-by: NJaehoon Chung <jh80.chung@samsung.com>
Signed-off-by: NUlf Hansson <ulf.hansson@linaro.org>