Currently, sdhci_prepare_dma() calls flush_cache() regardless of the
DMA direction.

Actually, cache invalidation is enough when reading data from the device.

This is correctly handled by dma_map_single(), which mimics the DMA-API
in Linux kernel. Drivers can be agnostic which cache operation occurs
behind the scene.

This commit also sanitizes the difference between the virtual address
and the dma address.

Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>

Copied from Linux kernel.
include/linux/mmc/host.h

Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Reviewed-by: Jaehoon Chung <jh80.chung@samsung.com>

host->mmc is already (struct mmc *).

memalign() returns an opaque pointer, so there is no need for casting.

Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Reviewed-by: Jaehoon Chung <jh80.chung@samsung.com>

Use {lower,upper}_32_bits() instead of the combination of cast
and shift.

Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Reviewed-by: Jaehoon Chung <jh80.chung@samsung.com>

The same code is run for both SDHCI_QUIRK_32BIT_DMA_ADDR and
define(CONFIG_FIXED_SDHCI_ALIGNED_BUFFER).

Unify the code.

Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Reviewed-by: Jaehoon Chung <jh80.chung@samsung.com>

Using the global variable does not look nice.

Add a new field sthci::align_buffer to point to the bounce buffer.

Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Reviewed-by: Jaehoon Chung <jh80.chung@samsung.com>

To avoid "asm/dma-mapping.h: No such file or directory" error,
we need something.

Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>

The implementation of dma_map_single() and dma_unmap_single() is
exactly the same for all the architectures that support them.

Factor them out to <linux/dma-mapping.h>, and make all drivers to
include <linux/dma-mapping.h> instead of <asm/dma-mapping.h>.

If we need to differentiate them for some architectures, we can
move the generic definitions to <asm-generic/dma-mapping.h>.

Add some comments to the helpers. The concept is quite similar to
the DMA-API of Linux kernel. Drivers are agnostic about what is
going on behind the scene. Just call dma_map_single() before the
DMA, and dma_unmap_single() after it.

Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>

dma_unmap_single() takes the dma address, not virtual address.

Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>

Make dma_map_single() return the dma address, and remove the
pointless volatile.

Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>

The reference manuals for MPC8308 and MPC8309 both say that the
esdhcctl aka DMA Control Register "is implemented as SDHCCR" in the
System configuration registers. Unfortunately, that doesn't mean that
the registers are just mirrors of each other - any write to esdhcctl
is simply ignored. So to actually enable cache snooping, we
unfortunately have to add a little ifdeffery.

There is, naturally, no description of the bit fields of esdhcctl in
the MPC8309 manual, but comparing the description of esdhcctl from the
LS1021A reference manual to the description of the sdhccr in MPC8309,
one also finds that the fields are bit-reversed, so the bit to set is
0x02000000 rather than 0x00000040 - this is also what board_mmc_init()
uses in the two gdsys/mpc8308/ boards.

Signed-off-by: Rasmus Villemoes <rasmus.villemoes@prevas.dk>
Reviewed-by: Yangbo Lu <yangbo.lu@nxp.com>

These functions are defined before the callers.

Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Reviewed-by: Peng Fan <peng.fan@nxp.com>

Since commit 01298da3 ("mmc: Change mode when switching to a boot
partition"), errors in mmc_select_mode_and_width() are ignored.
The return value should be checked.

Fixes: 01298da3 ("mmc: Change mode when switching to a boot partition")
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Reviewed-by: Jaehoon Chung <jh80.chung@samsung.com>

Cadence sent out an errata report to their customers of this IP.
This errata is not so severe, but the tune request should be sent
twice to avoid the potential issue.

Quote from the report:

Problem Summary
---------------
The IP6116 SD/eMMC PHY design has a timing issue on receive data path.
This issue may lead to an incorrect values of read/write pointers of
the synchronization FIFO. Such a situation can happen at the SDR104
and HS200 tuning procedure when the PHY is requested to change a phase
of sampling clock when moving to the next tuning iteration.

Workarounds
-----------
The following are valid workarounds to resolve the issue:

1. In eMMC mode, software sends tune request twice instead of once at
   each iteration. This means that the clock phase is not changed on
   the second request so there is no potential for clock instability.
2. In SD mode, software must not use the hardware tuning and instead
   perform an almost identical procedure to eMMC, using the HRS34 Tune
   Force register.

Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>

erase_grp_size is used, when MMC_WRITE is enabled.
- error: ‘struct mmc’ has no member named ‘erase_grp_size’

Signed-off-by: Jaehoon Chung <jh80.chung@samsung.com>
Reviewed-by: Peng Fan <peng.fan@nxp.com>

If the host has Broken R1B, use only R1 response type.

Signed-off-by: Bharat Kumar Reddy Gooty <bharat.gooty@broadcom.com>
Signed-off-by: Vladimir Olovyannikov <vladimir.olovyannikov@broadcom.com>

- add DH Electronics DHCOM SoM and PDK2 board
- DT alignment with kernel v5.5-rc7 for stm32mp1 boards
- fix STM32 image format for big endian hosts in mkimage
- solve warnings in device tree and code for stm32mp1 boards
- remove fdt_high and initrd_high for stm32 and stih boards
- add support of STM32MP15x Rev.Z
- update stm32mp1 readme

K3 J721E/AM65X:
- UART boot fixes for J721E
- Enabling debug uart for AM65x

DRA7xx/AM57xx:
- Preveinting boot from Invalid boot configuration for AM57xx

Keysonte2:
- Linux Boot fixes for K2 platforms.

AM33/AM43/Davinci:
- Ethernt boot fixes for AM43XX
- USB Host boot fixes for AM43XX

- Minor Kconfig fixes
- Ensure we use python3 on all CI in all cases.
- Other minor fixes

For stm32 f4, f7 and h7 boards, remove fdt_high and
initrd_high as they shouldn't be used, this allows the fdt
and initrd relocation.
This implies to set CONFIG_SYS_BOOTMAPSZ to indicate the
amount of memory available to contain kernel, device tree
and initrd for relocation.

Signed-off-by: Patrice Chotard <patrice.chotard@st.com>
Reviewed-by: Patrick DELAUNAY <patrick.delaunay@st.com>

For stm32f4, f7 and h7 boards, reserve:
 - 4MB for kernel
 - 64KB for fdt, boot script, pxefile
 - the remaining memory for ramdisk

Signed-off-by: Patrice Chotard <patrice.chotard@st.com>
Reviewed-by: Patrick DELAUNAY <patrick.delaunay@st.com>

Remove fdt_high and initrd_high as they shouldn't be used,
this allows the fdt and initrd relocation.
This implies to set CONFIG_SYS_BOOTMAPSZ to indicate the
amount of memory available to contain kernel, device tree
and initrd for relocation.

Signed-off-by: Patrice Chotard <patrice.chotard@st.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
Reviewed-by: Patrick DELAUNAY <patrick.delaunay@st.com>

Remove fdt_high and initrd_high (set to 0xffffffff) in stm32mp1 board
enviromnent, and U-Boot always relocate FDT and initrd in bootm command.

This relocation is limited by CONFIG_SYS_BOOTMAPSZ which indicates
the size of the memory region where it is safe to place data passed
to the Linux kernel (DTB, initrd), it is
a) Less than or equal to RAM size.
b) not within the kernel's highmem region

So 256M seems large enough in most circumstances and users
can override this value via environment variable "bootm_mapsize"
if needed.

This modification increases the boot time but avoid assumption
on aligned address for bootm command.

A user can still define this variables themselves if the FDT or
initrd is either left in-place or copied to a specific location.

Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
Reviewed-by: Tom Rini <trini@konsulko.com>

Solve type issue in stm32mp1_lse_enable and stm32mp1_clktree.

This patch solves the warnings when compiling with W=1
on stm32mp1 board:

clk_stm32mp1.c: In function ‘stm32mp1_lse_enable’:
clk_stm32mp1.c:1238:15: warning: comparison of integer expressions
  of different signedness: ‘u32’ {aka ‘unsigned int’} and ‘int’
  [-Wsign-compare]
clk_stm32mp1.c:1239:13: warning: comparison of integer expressions
  of different signedness: ‘u32’ {aka ‘unsigned int’} and ‘int’
  [-Wsign-compare]

clk_stm32mp1.c: In function ‘stm32mp1_clktree’:
clk_stm32mp1.c:1814:17: warning: comparison of integer expressions
  of different signedness: ‘int’ and ‘unsigned int’
  [-Wsign-compare]

Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
Reviewed-by: Patrice Chotard <patrice.chotard@st.com>

Use int as result of dm_i2c_reg_read to avoid warning with
W=1 (warning: comparison is always false due to limited range
of data type [-Wtype-limits])

Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
Reviewed-by: Patrice Chotard <patrice.chotard@st.com>

Change the dfu functions dfu_otp_read and dfu_pmic_read to static,
this patch avoids warning when compiling with W=1.

Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
Reviewed-by: Patrice Chotard <patrice.chotard@st.com>

Add include for dfu, add prototype for set_dfu_alt_info
and avoid warning when compiling with W=1.

Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
Reviewed-by: Patrice Chotard <patrice.chotard@st.com>

Two header fields are not properly converted to little endian
before assignment, resulting in incorrect header while executing
mkimage on big endian hosts.

Convert the value of the header fields image_checksum and
edcsa_algorithm to little endian before the assignment.

Signed-off-by: Antonio Borneo <antonio.borneo@st.com>
Reviewed-by: Patrick DELAUNAY <patrick.delaunay@st.com>
Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>

Add support for Rev.Z of STM32MP15x cpu.

Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
Reviewed-by: Patrice Chotard <patrice.chotard@st.com>

Update readme:
- list the supported SOC and change family to STM32MP15x
- add warning on OTP write and prerequisite:
  check if MAC address is not yet provisioned.
- Use filesize for mmc write command (avoid to write all partition
  with ${partsize}). ${filesize} and ${partsize} are set by previous
  load command.

Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
Reviewed-by: Patrice Chotard <patrice.chotard@st.com>

Fix the following DT dtc warnings for stm32mp1 boards:

Warning (unit_address_vs_reg): /soc/rcc@50000000/st,pll@0:
  node has a unit name, but no reg property
Warning (unit_address_vs_reg): /soc/rcc@50000000/st,pll@1:
  node has a unit name, but no reg property
Warning (unit_address_vs_reg): /soc/rcc@50000000/st,pll@2:
  node has a unit name, but no reg property
Warning (unit_address_vs_reg): /soc/rcc@50000000/st,pll@3:
  node has a unit name, but no reg property

Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
Reviewed-by: Patrice Chotard <patrice.chotard@st.com>

This patch fix the warning:
dt.dts: Warning (simple_bus_reg): Node /soc/ddr@5A003000
simple-bus unit address format error, expected "5a003000"

Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
Reviewed-by: Patrice Chotard <patrice.chotard@st.com>

LTDC modifies the clock frequency to adapt it to the display. Such
frequency change is not detected by the FDCAN driver that instead
cache the value at probe and pretend to use it later.

Keep the LTDC alone on PLL4_Q by moving the FDCAN to PLL4_R.

Signed-off-by: Antonio Borneo <antonio.borneo@st.com>
Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
Reviewed-by: Patrice Chotard <patrice.chotard@st.com>

Device tree and binding alignment with kernel v5.5-rc7

Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
Reviewed-by: Patrice Chotard <patrice.chotard@st.com>

Update the driver to use the latest binding from kernel v5.5-rc1:
no more use syscon or regmap to access to pwr register and
only one pwr_regulators node with the compatibility "st,stm32mp1,pwr-reg"
is available.

Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
Reviewed-by: Patrice Chotard <patrice.chotard@st.com>

Device tree and binding alignment with kernel v5.4

Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
Reviewed-by: Patrice Chotard <patrice.chotard@st.com>

Activate early console functionality on AM65x devices to allow for
early diagnostic messages until the main console is ready
to get activated.

Signed-off-by: Andreas Dannenberg <dannenberg@ti.com>
Signed-off-by: Lokesh Vutla <lokeshvutla@ti.com>

The boot parameter table index memory address for J721E was configured
to an incorrect value which prevented the use of this definition to
determine which boot parameter table is active which is needed to be
able to distinguish between primary and backup boot modes. Fix this
issue by updating the value to the correct one also in alignment with
the J721E Technical Reference Manual (TRM).

Signed-off-by: Andreas Dannenberg <dannenberg@ti.com>
Signed-off-by: Lokesh Vutla <lokeshvutla@ti.com>

Keystone2 u-boot loads the initrd image into non-LPAE addressed memory
but linux kernel is running in LPAE. This causes a conflict as kernel
detects that non-memory address is passed and kernel ignores initrd.
There is an existing fixup logic to modify the address in the proper
configuration, but this is disabled at the moment. Enable the fixup
by setting the env variable for this so that initrd can be used
properly.

Signed-off-by: Tero Kristo <t-kristo@ti.com>
Signed-off-by: Lokesh Vutla <lokeshvutla@ti.com>

On am571x-idk there can be following configurations based on Jumper J51
and LCD panel detected.

1) J51 removed (6port): 6 port Ethernet. Disable LCD panel.
2) J51 placed (LCD) + Panel detected: 4 port Ethernet with appropriate LCD.
3) J51 placed (LCD) + Panel not detected/not supported.

Configuration 3 is considered invalid as we can't use display nor ICSS1
ethernet ports due to hardware muxing. Alert the user to fix the
configuration and prevent boot.

Alternative was to allow boot and limit to 4 port Ethernet with no display
but this involved introduction of another DTB for the kernel and was
considered not worth the hassle.

Signed-off-by: Roger Quadros <rogerq@ti.com>
Acked-by: Suman Anna <s-anna@ti.com>
Signed-off-by: Lokesh Vutla <lokeshvutla@ti.com>