This patch has been tested on the mcf54415-based stmark2
board. The eSDHC driver works reliably using DMA mode.

Signed-off-by: Angelo Dureghello <angelo@sysam.it>

This patch is to make get_cd work well when DM_MMC enabled

Signed-off-by: Yinbo Zhu <yinbo.zhu@nxp.com>
Reviewed-by: York Sun <york.sun@nxp.com>

The following hang is observed on a Hummingboard 2 MicroSOM
i2eX iMX6D - rev 1.3 with no eMMC populated on board:

U-Boot SPL 2018.11+gf6206f8587 (Nov 16 2018 - 00:56:34 +0000)
Trying to boot from MMC1

U-Boot 2018.11+gf6206f8587 (Nov 16 2018 - 00:56:34 +0000)

CPU:   Freescale i.MX6D rev1.5 996 MHz (running at 792 MHz)
CPU:   Extended Commercial temperature grade (-20C to 105C) at 33C
Reset cause: POR
Board: MX6 HummingBoard2
DRAM:  1 GiB
MMC:   FSL_SDHC: 0, FSL_SDHC: 1
Loading Environment from MMC... *** Warning - bad CRC, using default environment

No panel detected: default to HDMI
Display: HDMI (1024x768)
In:    serial
Out:   serial
Err:   serial
---> hangs

which is caused by the following infinite loop inside esdhc_send_cmd_common()

	while (!(esdhc_read32(&regs->irqstat) & flags))
		;

Instead of looping forever, provide an exit path so that a timeout
error can be propagated in the case irqstat does not report
any interrupts, which may happen when no eMMC is populated on
board.

Reported-by: Ricardo Salveti <rsalveti@rsalveti.net>
Signed-off-by: Fabio Estevam <festevam@gmail.com>
Tested-by: Peter Robinson <pbrobinson@gmail.com>
Tested-by: Ricardo Salveti <rsalveti@rsalveti.net>

When CONIFG_CLK is enabled, use uclass clk api to handle
the clock.

Signed-off-by: Peng Fan <peng.fan@nxp.com>
Reviewed-by: Anatolij Gustschin <agust@denx.de>
Cc: Jaehoon Chung <jh80.chung@samsung.com>
Cc: Stefano Babic <sbabic@denx.de>
Signed-off-by: Troy Kisky <troy.kisky@boundarydevices.com>

The strobe dll code is ported from Linux Kernel:
drivers/mmc/host/sdhci-esdhc-imx.c
The comments are from the above file,
"For HS400 eMMC, there is a data_strobe line. This signal is generated
by the device and used for data output and CRC status response output
in HS400 mode. The frequency of this signal follows the frequency of
CLK generated by host. The host receives the data which is aligned to the
edge of data_strobe line. Due to the time delay between CLK line and
data_strobe line, if the delay time is larger than one clock cycle,
then CLK and data_strobe line will be misaligned, read error shows up.
So when the CLK is higher than 100MHz, each clock cycle is short enough,
host should configure the delay target. "

Signed-off-by: Peng Fan <peng.fan@nxp.com>
Cc: Jaehoon Chung <jh80.chung@samsung.com>
Cc: Stefano Babic <sbabic@denx.de>

In SPL, all the available mmc devices gets initialized during boot.
This might not work in cases where clocks are not available for
certain mmc devices(other than boot device) and the support for
enabling device might not be ready.

Texas Instruments' K3 J721E device having a central system controller
(dmsc) is one such example falling in this category. Below is the
sequence for the failing scenario:
- ROM comes up in SD mode and loads SPL by just initialing SD card.
- SPL loads dmsc firmware from SD Card.
Since ROM has enabled SD, SPL need not enable the SD, just need
to re initialize the card. But SPL is trying to initialize other MMC
instances which are in disabled state. Since dmsc firmware is not yet
available, devices cannot be enabled. So in SPL, initialize only the
mmc device that is needed.

Signed-off-by: Lokesh Vutla <lokeshvutla@ti.com>
Reviewed-by: Peng Fan <peng.fan@nxp.com>

It's quite hard to figure out time units for various function that have
timeout parameters. This leads to possible errors when one forgets to
convert ms to us, for example. Let's rename those parameters
correspondingly to 'timeout_us' and 'timeout_ms' to prevent such issues
further.

While at it, add time units info as comments to struct mmc fields.

This commit doesn't change the behavior, only renames parameters names.
Buildman should report no changes at all.

Signed-off-by: Sam Protsenko <semen.protsenko@linaro.org>
Reviewed-by: Peng Fan <peng.fan@nxp.com>
Reviewed-by: Igor Opaniuk <igor.opaniuk@gmail.com>

mmc_wait_dat0() expects timeout argument to be in usec units. But some
overlying functions operate on timeout in msec units. Convert timeout
from msec to usec when passing it to mmc_wait_dat0().

This fixes 'avb' commands on BeagleBoard X15, because next chain was
failing:

    get_partition() -> mmc_switch_part() -> __mmc_switch() ->
    mmc_wait_dat0()

when passing incorrect timeout from __mmc_switch() to mmc_wait_dat0().

Fixes: bb98b8c5c06a ("mmc: During a switch, poll on dat0 if available and check the final status")
Signed-off-by: Sam Protsenko <semen.protsenko@linaro.org>
Reviewed-by: Eugeniu Rosca <rosca.eugeniu@gmail.com>
Tested-by: Eugeniu Rosca <rosca.eugeniu@gmail.com>
Reviewed-by: Peng Fan <peng.fan@nxp.com>
Tested-by: Igor Opaniuk <igor.opaniuk@gmail.com>
Reviewed-by: Igor Opaniuk <igor.opaniuk@gmail.com>

The comment about init op being NULL used to be next to the NULL check
code. Commit 8ca51e51 ("dm: mmc: Add a way to use driver model for
MMC operations") separated the comment from the code. Put them back
together.

Fixes: 8ca51e51 ("dm: mmc: Add a way to use driver model for MMC operations")
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Reviewed-by: Peng Fan <peng.fan@nxp.com>

eMMC 5.1+ supports HS400 Enhances Strobe mode without the need for
tuning procedure.
The flow is as following:
 - set HS_TIMIMG (Highspeed)
 - Host change freq to <= 52Mhz
 - set the bus width to Enhanced strobe and DDR8Bit(CMD6),
   EXT_CSD[183] = 0x86 instead of 0x80
 - set HS_TIMING to 0x3 (HS400)
 - Host change freq to <= 200Mhz
 - Host select HS400 enhanced strobe complete

Signed-off-by: Peng Fan <peng.fan@nxp.com>

The MMC mode and width are fixed for MMC SPI host hence we skip
sd_select_mode_and_width() and mmc_select_mode_and_width() for
MMC SPI host.

Signed-off-by: Anup Patel <anup.patel@wdc.com>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Tested-by: Bin Meng <bmeng.cn@gmail.com>

This operation may fail. Retry it a few times before giving up and report
a failure.

Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>

Accessing the boot partition had been error prone with HS200 and HS400 and
was disabled. The driver first switched to a lesser mode and then switched
the partition access. It was mostly due to a bad handling of the switch and
has been fixed, so let's remove this 'feature'

Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>

The switch operation can sometimes make the bus unreliable, in that case
the send_status parameter should be false to indicate not to poll using
CMD13. If polling on dat0 is possible, we should use it to detect the end
of the operation.
At the end of the operation it is safe to use CMD13 to get the status of
the card. It is important to do so because the operation may have failed.

Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>

The e-MMC spec allows the e-MMC to specify a timeout for the partition
switch command. It can take up to 2550 ms. There is no lower limit to this
value in the spec, but do as the the linux driver does and force it to be
at least 300ms.

Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>

Starting with rev 4.5, the eMMC can define a generic timeout for the
SWITCH command.

Following Linux Kernel code, the timeout also changed from 1000 -> 500

Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>

Using the DAT0 line as a rdy/busy line is an alternative to reading the
status register of the card. It especially useful in situation where the
bus is not in a good shape, like when modes are switched.
This is also how the linux driver behaves.

Note of warning: As per the specification, while polling on DAT0 the CLK
must not turned off: "[...] Without a clock edge the Device (unless
previously disconnected by a deselect command (CMD7)) will force the DAT0
line down, forever. [...]"

Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>

mmc_send_status() is currently used to poll the card until it is ready, not
actually returning the status of the card.
Make it return the status and add another function to poll the card.

Also remove the 'extern' declaration in the mmc-private.h header to comply
with the coding standard.

Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>

This reverts commit 318a7a57.

The last and only user of this callback had been the omap_hsmmc driver.
It is not used anymore. Removing the callback.

Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>

U-Boot code currently only applies this restriction to HS200 mode,
extend this to HS400 mode as well.

Currently U-Boot code not support accessing boot partition in HS200/400
mode. This needs more check.

Signed-off-by: Marek Vasut <marek.vasut+renesas@gmail.com>
Cc: Jean-Jacques Hiblot <jjhiblot@ti.com>
Cc: Nobuhiro Iwamatsu <iwamatsu@nigauri.org>
Cc: Peng Fan <peng.fan@nxp.com>
Reviewed-by: Peng Fan <peng.fan@nxp.com>

The tiny printf implementation only supports %x format specifier,
it does not support %X . Since it makes little difference whether
the debug output prints hex numbers in capitals or not, change it
to %x and make the MMC_TRACE output work with tiny printf too.

Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Jaehoon Chung <jh80.chung@samsung.com>
Cc: Simon Glass <sjg@chromium.org>

Commit d0851c893706 ("blk: Call part_init() in the post_probe() method")
removed the call to part_init() in mmc.c, as this is done by the DM_MMC
framework.
However Allwinner is (still) relying on a non-DM MMC driver, so we are
now missing the implicit partition init, leading to failing MMC accesses
due to the missing partition information.

Bring the call back just for non-DM MMC driver to fix this regression.

Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Tested-by: Soeren Moch <smoch@web.de>

part_init() is currently called in every DM BLK driver, either
in its bind() or probe() method. However we can use the BLK
uclass driver's post_probe() method to do it automatically.

Update all DM BLK drivers to adopt this change.

Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>

Add HS400 support.
Selecting HS400 needs first select HS200 according to spec, so use
a dedicated function for HS400.
Add HS400 related macros.
Remove the restriction of only using the low 6 bits of
EXT_CSD_CARD_TYPE, using all the 8 bits.

Signed-off-by: Peng Fan <peng.fan@nxp.com>
Cc: Jaehoon Chung <jh80.chung@samsung.com>
Cc: Jean-Jacques Hiblot <jjhiblot@ti.com>
Cc: Stefano Babic <sbabic@denx.de>
Cc: Kishon Vijay Abraham I <kishon@ti.com>
Cc: Faiz Abbas <faiz_abbas@ti.com>
Cc: Marek Vasut <marex@denx.de>
Signed-off-by: Peng Fan <peng.fan@nxp.com>
Signed-off-by: Troy Kisky <troy.kisky@boundarydevices.com>

This code is useful for testing the existance of devices that
do not have card detect capabilities.  This breaks out the core
functionality and leaves the actual init logic and error reporting
in mmc_start_init().

Signed-off-by: Jon Nettleton <jon@solid-run.com>
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Reviewed-by: Stefano Babic <sbabic@denx.de>

After the commit b9a2a0e2e9c0 ("mmc: Add support for downgrading
HS200/HS400 to HS mode"), it add a parameter in mmc_set_card_speed()
which indicates that the HS200/HS400 to HS downgrade is happening.

During the HS400 initialization, first select to HS200, and config
the related clock rate, then downgrade to HS mode. So here also need
to config the downgrade value to be true for two reasons. First,
make sure in the function mmc_set_card_speed(), after switch to HS
mode, first config the clock rate, then read the EXT_CSD, avoid
receiving data of EXT_CSD in HS mode at 200MHz. Second, after issue
the MMC_CMD_SWITCH command, it need to wait a bit then switch bus
properties.

Test on i.MX8QM MEK board, some Micron eMMC will stuck in transfer
mode in this case, and USDHC will never get data transfer complete
status, cause the uboot hang.

Signed-off-by: Haibo Chen <haibo.chen@nxp.com>
Acked-by: Marek Vasut <marek.vasut@gmail.com>

Older kernel versions or systems which do not connect eMMC reset line
properly may not be able to handle situations where either the eMMC
is left in HS200/HS400 mode or SD card in UHS modes by the bootloader
and may misbehave. Downgrade the eMMC to HS/HS52 mode and/or SD card
to non-UHS mode before booting the kernel to allow such older kernels
to work with modern U-Boot.

Signed-off-by: Marek Vasut <marek.vasut+renesas@gmail.com>
Cc: Tom Rini <trini@konsulko.com>

When using CMD6 to switch eMMC card timing from HS200/HS400 to HS/legacy,
do not poll for the completion status using CMD13, but rather wait 50mS.

Once the card receives the CMD6 and starts executing it, the bus is in
undefined state until both the card finishes executing the command and
until the controller switches the bus to matching timing configuration.
During this time, it is not possible to transport any commands or data
across the bus, which includes the CMD13.

Signed-off-by: Marek Vasut <marek.vasut+renesas@gmail.com>
Cc: Jaehoon Chung <jh80.chung@samsung.com>

The mmc_select_mode_and_width() function can be called while the card
is in HS200/HS400 mode and can be used to downgrade the card to lower
mode, e.g. HS. This is used for example by mmc_boot_part_access_chk()
which cannot access the card in HS200/HS400 mode and which is in turn
called by saveenv if env is in the MMC.

In such case, forcing the card clock to legacy frequency cannot work.
Instead, the card must be switched to HS mode first, from which it can
then be reprogrammed as needed.

However, this procedure needs additional code changes, since the current
implementation checks whether the card correctly switched to HS mode in
mmc_set_card_speed(). The check only expects that the card will be going
to HS mode from lower modes, not from higher modes, hence add a parameter
which indicates that the HS200/HS400 to HS downgrade is happening. This
makes the code send the switch command first, reconfigure the controller
next and finally perform the EXT_CSD readback check. The last two steps
cannot be done in reverse order as the card is already in HS mode when
the clock are being switched on the controller side.

Signed-off-by: Marek Vasut <marek.vasut+renesas@gmail.com>
Cc: Jaehoon Chung <jh80.chung@samsung.com>
Signed-off-by: Troy Kisky <troy.kisky@boundarydevices.com>

According to SD Specifications Part 1 Physical Layer Simplified Specification
Version 6.00 August 29, 2018, section 4.3.10 (Switch Function Command) and
section 5.6 (SCR register), SD cards version 1.00 and 1.01 do not support the
SD CMD 6.

Currently, U-Boot will issue CMD 6 unconditionally in sd_set_card_speed()
while configuring the bus for selected frequency. This will make SD cards
version 1.00 and 1.01 time out and thus fail detection altogether.

Fix this by not sending CMD 6 on such cards. Tested on Matsushita Electric
Industrial Co., Ltd. Japan RP-SD008B / Victor 8MB SD card, CU-SD008, which
is correctly detected with this patch as:

Device: sd@ee160000
Manufacturer ID: 1
OEM: 5041
Name: S008B
Bus Speed: 25000000
Mode : SD Legacy
Rd Block Len: 512
SD version 1.0
High Capacity: No
Capacity: 6.5 MiB
Bus Width: 4-bit
Erase Group Size: 512 Bytes

Signed-off-by: Marek Vasut <marek.vasut+renesas@gmail.com>
Cc: Masahiro Yamada <yamada.masahiro@socionext.com>

Signed-off-by: Troy Kisky <troy.kisky@boundarydevices.com>

Signed-off-by: Troy Kisky <troy.kisky@boundarydevices.com>

Signed-off-by: Troy Kisky <troy.kisky@boundarydevices.com>

Signed-off-by: Troy Kisky <troy.kisky@boundarydevices.com>

Signed-off-by: Troy Kisky <troy.kisky@boundarydevices.com>

Signed-off-by: Troy Kisky <troy.kisky@boundarydevices.com>

Signed-off-by: Troy Kisky <troy.kisky@boundarydevices.com>

Signed-off-by: Troy Kisky <troy.kisky@boundarydevices.com>

Signed-off-by: Laci Tele <laci@boundarydevices.com>

Signed-off-by: Troy Kisky <troy.kisky@boundarydevices.com>