Based on DT spec you can have one memory node which multiple ranges or
multiple nodes.
fdt_fixup_memory_banks() is not implemented in a correct way when multiple
memory nodes are present because all ranges are put it to the first memory
node found. And next memory nodes are kept in DT which ends up in the same
range specification in the same DT.

Here is what it is happening.
Origin DT.
memory@0 {
        device_type = "memory";
        reg = <0x0 0x0 0x0 0x80000000>;
};

memory@800000000 {
        device_type = "memory";
        reg = <0x8 0x00000000 0x0 0x80000000>;
};

After fdt_fixup_memory_banks()

memory@0 {
        device_type = "memory";
        reg = <0x0 0x0 0x0 0x80000000>, <0x8 0x00000000 0x0 0x80000000>;
};

memory@800000000 {
        device_type = "memory";
        reg = <0x8 0x00000000 0x0 0x80000000>;
};

As is visible memory@0 node got second range but there is still
memory@800000000 node present and 2G range is listed twice.

The solution can't be that second node is removed because it can be
referenced already that's why it is better for us to disable this option
for now.
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>

Enable cfi flash on zc770-xm012 configuration.
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>

Single configuration is working fine and no issue to enable it.
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>

Replace 'io-standard' property with 'power-source' property in all
zynq dts files to be in sync with Zynq Pinctrl driver.
Signed-off-by: NSai Krishna Potthuri <lakshmi.sai.krishna.potthuri@xilinx.com>
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>

Add missing "jedec, spi-nor" compatible string for QSPI flash node.
Spi-nor framework uses this compatibility string to probe &
initialize flash. With missing compatibility string we are observing
below error:

Zynq> sf probe 0 0 0
 jedec_spi_nor spi_flash@0:0: unrecognized JEDEC id bytes:
00, 00, 00  Failed to initialize SPI flash at 0:0 (error -2)
Signed-off-by: NRaju Kumar Pothuraju <raju.kumar-pothuraju@xilinx.com>
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>

Trivial style patch.
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>

Fix header alignment.
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>

By dt-binding specs led nodes should have -led suffix that's why add it.
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>

Based on thread
https://lists.denx.de/pipermail/u-boot/2021-June/451828.html
especially
"Overall we have a deficiency in the UEFI implementation in that we
cannot deal with block devices added or removed after initialization."
there is a need to deal with removable media as usb/scsi/sata.

That's why bridge this gap in EFI implementation by resetting usb and
scsi resets to get all disks before efi_init_obj_list() is called.

In our standard boot flow, where we use distro boot, order is fixed as
"jtag mmc0 mmc1 qspi0 nand0 usb0 usb1 scsi0 pxe dhcp" with prioritizing
boot device added by commit 2882b39d ("arm64: zynqmp: Setup the first
boot_target at run time").
When device has ESP partition all devices should be detected because then
efi_disk_register() in efi_init_obj_list() is called only once.

The first such a device is sd/emmc(mmc0/mmc1) and then disks on usb/sata
are not handled at all.

The commit 6bb577db ("arm64: zynqmp: Disable
EFI_CAPSULE_ON_DISK_EARLY") also pointed out on this issue but detection of
removable media wasn't solved that's why do it now via preboot command.

I have tested cases without usb and scsi and there is no problem with
calling resets without devices itself.
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>

Do not enable watchdog driver in default zynqmp configuration. The reason
is that not all distributions are handling watchdog properly and then
expires and system resets. If someone needs watchdog in their design please
enable it by hand.
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>

Since set_control_reg is available in sdhci.c, use it and remove
arasan_sdhci_set_control_reg().
Signed-off-by: NAshok Reddy Soma <ashok.reddy.soma@xilinx.com>
Reviewed-by: NJaehoon Chung <jh80.chung@samsung.com>
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>

As per SD spec when SD host controller is reset, it takes 1000msec
to detect the card state. In case, if we enable the sd bus voltage &
card detect state is not stable, then host controller will disable
the sd bus voltage.

In case of warm/subsystem reboot, due to unstable card detect state
host controller is disabling the sd bus voltage to sd card causing
sd card timeout error. So we wait for a maximum of 1000msec to get
the card detect state stable before we enable the sd bus voltage.

This current fix is workaround for now, this needs to be analysed
further. Zynqmp platform should behave the same as Versal, but we
did not encounter this issue as of now. So we are fixing it for
Versal only.
Signed-off-by: NT Karthik Reddy <t.karthik.reddy@xilinx.com>
Signed-off-by: NAshok Reddy Soma <ashok.reddy.soma@xilinx.com>
Reviewed-by: NJaehoon Chung <jh80.chung@samsung.com>
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>

Move tapdelay function calls to zynq_sdhci.c and make them static
inline. zynqmp_tap_delay.h has function prototypes for the functions
defined in tap_delays.c, which will not be needed anymore.

Remove tap_delays.c and zynqmp_tap_delay.h files.
Signed-off-by: NAshok Reddy Soma <ashok.reddy.soma@xilinx.com>
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>
Reviewed-by: NJaehoon Chung <jh80.chung@samsung.com>

Currently xilinx sdhci driver is using zynqmp_mmio_write() to set
tapdelay values and DLL resets. Continue to use this for SPL and mini
U-Boot where U-Boot will be executed at EL3 level.

Use firmware call xilinx_pm_request() using appropriate arguments to
set input/output tapdelays and also for DLL resets in regular flow(EL2).

Host driver should explicitly request DLL reset before ITAP (assert DLL)
and after OTAP (release DLL) to avoid issues in some cases. Also handle
error return where possible.
Signed-off-by: NT Karthik Reddy <t.karthik.reddy@xilinx.com>
Signed-off-by: NAshok Reddy Soma <ashok.reddy.soma@xilinx.com>
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>
Reviewed-by: NJaehoon Chung <jh80.chung@samsung.com>

Add enums for pm node id's, pm ioctl id's, tapdelay types, dll reset types
Signed-off-by: NT Karthik Reddy <t.karthik.reddy@xilinx.com>
Signed-off-by: NAshok Reddy Soma <ashok.reddy.soma@xilinx.com>
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>

set_delay() has return type as void. If there are any errors while
setting tapdelay's it won't be able to return them.

Change the prototype of set_delay() in sdhci_ops structure and return
the errors from wherever it is called.
Signed-off-by: NAshok Reddy Soma <ashok.reddy.soma@xilinx.com>
Reviewed-by: NJaehoon Chung <jh80.chung@samsung.com>
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>

Change return type of arasan_sdhci_set_tapdelay() to int, to facilitate
returning errors. Get return values from input and output set clock phase
functions inside arasan_sdhci_set_tapdelay() and return those errors.
Signed-off-by: NT Karthik Reddy <t.karthik.reddy@xilinx.com>
Signed-off-by: NAshok Reddy Soma <ashok.reddy.soma@xilinx.com>
Reviewed-by: NJaehoon Chung <jh80.chung@samsung.com>
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>

Add firmware based reset controller for Xilinx ZynqMP SoC to let other
drivers to call reset functions. Driver is only tested on Xilinx ZynqMP but
support for Xilinx Versal can be simply added. That's why reset_id and
nr_reset are assigned in probe folder.
Driver is inpired by driver from Linux kernel.
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>

Enable RTC command to be able to check available.
And also enable ZynqMP RTC driver to be possible to use by default.

Here is the list when both drivers are enabled:
ZynqMP> rtc list
RTC #0 - rtc_emul
RTC #1 - rtc@ffa60000
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>

The whole driver logic is taken from Linux kernel but only set/get/reset
functions are implemented. When device is power off RTC is power out of
battery.
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>

RTCs are using sequence number defined in aliases node. Date command with
DM_RTC enabled is looking for the first RTC with index 0. But when
RTC_EMULATION is enabled it gets likely most of the time index 0 even when
system has rtc0 device via aliases node and gets sequence number 0.
That's why extend the code to look for sequence 0 number first. If this
fails continue to use existing device with index 0.
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>

RTCs are using aliases for sequences. That's why enable
DM_UC_FLAG_SEQ_ALIAS for exact RTC indentification.
The same flag is used by a lot of other uclasses like mmc, pci, serial,
spi, timer, tpm, etc.
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>

Based on discussion at
https://lore.kernel.org/r/20200318125003.GA2727094@kroah.com we got
recommendation to use explicit values for all enum values.
The patch is following this recommendation.
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>
Link: https://lore.kernel.org/r/daeb67ded45d8a8f6a96717d1fb9c84439dd2ae8.1612361627.git.michal.simek@xilinx.com

The main reason for this to be implemented is capsule update.
Two memories are supported and tested which is MMC FAT based and QSPI
based.

For creating capsule these commands are used:
./tools/mkeficapsule --raw spl/boot.bin --index 1 capsule1.bin
./tools/mkeficapsule --raw u-boot.itb --index 2 capsule2.bin

Then transfer to SD card where these commands run:
load mmc 0 10000000 capsule1.bin
efidebug capsule update -v 10000000
load mmc 0 10000000 capsule2.bin
efidebug capsule update -v 10000000

Depends on the boot device used are binaries loaded to qspi or mmc fat
partition.
Also multiboot register is handled to make sure that the same location(id)
is used as image which is upgraded.

Two locations are used by purpose for SPL flow. If only boot.bin is used
create only one capsule.
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>

This variable is pointing to offset is qspi where u-boot image is placed.
In our case it is location of u-boot.itb file. Offset is the same as is
used by Xilinx Zynq SoC.
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>

When U-Boot runs in EL2 there is no access to csu_base registers that's why
this has to be done via firmware interface to find out multi boot register
value. Till now this function is called only from SPL in EL3.
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>

Change multi_boot() to return multiboot value and move print out of this
function and let this function to be used by other functions without
duplicating message.
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>

When env_set() is called there is no need to allocate memory for variable
which is already saved that's why free it.
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>

- Assorted fixes

Modified the help text of 'chpart' command ,mentioning that it is
for MTD devices.
Signed-off-by: NAdarsh Babu Kalepalli <opensource.kab@gmail.com>
Reviewed-by: NSimon Glass <sjg@chromium.org>

HELP description is provided for ‘configure’ sub-command
of ‘blkcache’.
Signed-off-by: NAdarsh Babu Kalepalli <opensource.kab@gmail.com>
Reviewed-by: NSimon Glass <sjg@chromium.org>

Symbol CONFIG_SYS_ID_EEPROM is defined in include/configs/MPC8548CDS.h
but never used. Remove it here and from the whitelist.
Signed-off-by: NHeinrich Schuchardt <xypron.glpk@gmx.de>

Start out by documenting general expectations on when CI is run, how
anyone can run Azure pipelines, and how GitLab CI pipelines can be run.
Signed-off-by: NTom Rini <trini@konsulko.com>

In SquashFS, the contents of a directory is stored by
squashfs_directory_entry structures which contain the file's name, inode
and position within the filesystem.

The inode number is not stored directly; instead each directory has one
or more headers which set a base inode number, and files store the
offset from that to the file's inode number.

In mksquashfs, each inode is allocated a number in the same order as
they are written to the directory table; thus the offset from the
header's base inode number to the file's inode number is usually
positive.

Hardlinks are simply stored with two directory entries referencing the
same file. This means the second entry will thus have an inode number
much lower than the surrounding files. Since the header's base inode
number comes from the first entry that uses the header, this delta will
usually be negative.

Previously, U-Boot's squashfs_directory_entry.inode_offset field was
declared as an unsigned value. Thus when a negative value was found, it
would either resolve to an invalid inode number or to that of an
unrelated file.

A squashfs image to test this can be created like so:

    echo hi > sqfs_test_files/001-root-file
    mkdir     sqfs_test_files/002-subdir
    touch     sqfs_test_files/002-subdir/003-file
    ln        sqfs_test_files/{001-root-file,002-subdir/004-link}
    mksquashfs sqfs_test_files/ test.sqfs -noappend

Note that squashfs sorts the files ASCIIbetacally, so we can use the
names to control the order they appear in. The ordering is important -
the first reference to the file must have a lower inode number than the
directory in which the second reference resides, and the second
reference cannot be the first file in the directory.

Listing this sample image in U-Boot results in:

=> sqfsls virtio 2 002-subdir
         0   003-file
Inode not found.
         0   004-link
Signed-off-by: NCampbell Suter <campbell@snapit.group>
Reviewed-by: NMiquel Raynal <miquel.raynal@bootlin.com>

The "-E" option to mkimage generates a FIT with external data using the
data-size and data-offset properties which must both be ignored when
verifying a signature.

Add "data-offset" to the list of excluded properties for signature
verification; since the line is now too long, re-format the list to
one-per-line and make it static since the data is constant.
Signed-off-by: NJohn Keeping <john@metanate.com>
Reviewed-by: NSimon Glass <sjg@chromium.org>

The current stable release of LLVM is 12, update to that.  While at it,
fix that we had not correctly upgraded to LLVM 11 previously.
Signed-off-by: NTom Rini <trini@konsulko.com>

- SPI-NOR fix (Big Meng)
- XMC XM25QH64C flash (Reto Schneider)

The DT bindings of "jedec,spi-nor" [1] defines "m25p,fast-read" property
to indicate that "fast read" opcode can be used to read data from the
chip instead of the usual "read" opcode.

If this property is not present in DT, mask out fast read in
spi_nor_init_params(). This change mirrors the same logic in
spi_nor_info_init_params() in drivers/mtd/spi-nor/core.c in
the Linux kernel v5.14-rc3.

[1] Documentation/devicetree/bindings/mtd/jedec,spi-nor.yaml in the kernel tree
Signed-off-by: NBin Meng <bmeng.cn@gmail.com>
Reviewed-by: NJagan Teki <jagan@amarulasolutions.com>

The smart spi_nor_adjust_hwcaps() does not respect the SPI flash's
hwcaps, and only looks to the controller on what can be supported.

The flash's hwcaps needs to be AND'ed before checking.

Fixes: 71025f01 ("mtd: spi-nor-core: Rework hwcaps selection")
Signed-off-by: NBin Meng <bmeng.cn@gmail.com>
Reviewed-by: NPratyush Yadav <p.yadav@ti.com>
Reviewed-by: NJagan Teki <jagan@amarulasolutions.com>