]> git.dujemihanovic.xyz Git - u-boot.git/log
u-boot.git
4 years agox86: irq: Fix some typos
Wolfgang Wallner [Tue, 21 Jul 2020 11:01:45 +0000 (13:01 +0200)]
x86: irq: Fix some typos

Fix some typos in arch/x86/include/asm/irq.h.

Signed-off-by: Wolfgang Wallner <wolfgang.wallner@br-automation.com>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
4 years agoMerge branch '2020-07-31-more-env-updates'
Tom Rini [Fri, 31 Jul 2020 14:13:07 +0000 (10:13 -0400)]
Merge branch '2020-07-31-more-env-updates'

- Fix EFI selftest to not force setting serial# environment (and also
  get the U-Boot prompt dynamically).
- Support for append only environment and other related features.
- Improved ext4 environment support
- Fix the case of fw_setenv being used on flash devices that were not
  already locked.

4 years agofw_setenv: lock the flash only if it was locked before
Ivan Mikhaylov [Fri, 10 Jul 2020 16:54:18 +0000 (19:54 +0300)]
fw_setenv: lock the flash only if it was locked before

With current implementation of fw_setenv, it is always locks u-boot-env
region if lock interface is implemented for such mtd device. You can
not control lock of this region with fw_setenv, there is no option for
it in config or in application itself. Because of this situation may
happen problems like in this thread on xilinx forum:
https://forums.xilinx.com/t5/Embedded-Linux/Flash-be-locked-after-use-fw-setenv-from-user-space
/td-p/1027851

A short summary of that link is: some person has issue with some spi
chip which has lock interface but doesn't locks properly which leads to
lock of whole flash memory on lock of u-boot-env region. As resulted
solution hack was added into spi-nor.c driver for this chip with lock
disablement.

Instead fix this problem by adding logic to fw_setenv only lock the
flash if it was already locked when we attempted to use it.

Signed-off-by: Ivan Mikhaylov <fr0st61te@gmail.com>
4 years agoenv: Add support for explicit write access list
Marek Vasut [Tue, 7 Jul 2020 18:51:39 +0000 (20:51 +0200)]
env: Add support for explicit write access list

This option marks any U-Boot variable which does not have explicit 'w'
writeable flag set as read-only. This way the environment can be locked
down and only variables explicitly configured to be writeable can ever
be changed by either 'env import', 'env set' or loading user environment
from environment storage.

Signed-off-by: Marek Vasut <marex@denx.de>
Reviewed-by: Tom Rini <trini@konsulko.com>
4 years agoenv: Add option to only ever append environment
Marek Vasut [Tue, 7 Jul 2020 18:51:38 +0000 (20:51 +0200)]
env: Add option to only ever append environment

Add configuration option which prevents the environment hash table to be
ever cleared and reloaded with different content. This is useful in case
the first environment loaded into the hash table contains e.g. sensitive
content which must not be dropped or reloaded.

Signed-off-by: Marek Vasut <marex@denx.de>
Reviewed-by: Tom Rini <trini@konsulko.com>
4 years agoenv: Discern environment coming from external storage
Marek Vasut [Tue, 7 Jul 2020 18:51:35 +0000 (20:51 +0200)]
env: Discern environment coming from external storage

Add another custom environment flag which discerns environment coming
from external storage from environment set by U-Boot itself.

Signed-off-by: Marek Vasut <marex@denx.de>
Reviewed-by: Tom Rini <trini@konsulko.com>
4 years agoenv: Add H_DEFAULT flag
Marek Vasut [Tue, 7 Jul 2020 18:51:34 +0000 (20:51 +0200)]
env: Add H_DEFAULT flag

Add another internal environment flag which indicates that the operation
is resetting the environment to the default one. This allows the env code
to discern between import of external environment and reset to default.

Signed-off-by: Marek Vasut <marex@denx.de>
Reviewed-by: Tom Rini <trini@konsulko.com>
4 years agoenv: Warn on force access if ENV_ACCESS_IGNORE_FORCE set
Marek Vasut [Tue, 7 Jul 2020 18:51:33 +0000 (20:51 +0200)]
env: Warn on force access if ENV_ACCESS_IGNORE_FORCE set

If the ENV_ACCESS_IGNORE_FORCE is set, inform user that the variable
cannot be force-set if such attempt happens.

Signed-off-by: Marek Vasut <marex@denx.de>
Reviewed-by: Tom Rini <trini@konsulko.com>
4 years agotest: sandbox: add test for erase command
Patrick Delaunay [Tue, 28 Jul 2020 09:51:27 +0000 (11:51 +0200)]
test: sandbox: add test for erase command

Add test for the erase command tested on ENV in EXT4.

Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
4 years agoenv: ext4: add support of command env erase
Patrick Delaunay [Tue, 28 Jul 2020 09:51:26 +0000 (11:51 +0200)]
env: ext4: add support of command env erase

Add support of opts erase for env in ext4,
this opts is used by command 'env erase'.

This command only fill the env file (CONFIG_ENV_EXT4_FILE)
with 0, the CRC and the saved environment becomes invalid.

Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
4 years agoenv: ext4: introduce new function env_ext4_save_buffer
Patrick Delaunay [Tue, 28 Jul 2020 09:51:25 +0000 (11:51 +0200)]
env: ext4: introduce new function env_ext4_save_buffer

Split the function env_ext4_save to prepare the erase support.

Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
4 years agotest: environment in ext4
Patrick Delaunay [Tue, 28 Jul 2020 09:51:24 +0000 (11:51 +0200)]
test: environment in ext4

Add basic test to persistent environment in ext4:
save and load in host ext4 file 'uboot.env'.

On first execution an empty EXT4 file system is created in
persistent data dir: env.ext4.img.

Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
4 years agoconfigs: sandbox: activate command env select and env load
Patrick Delaunay [Tue, 28 Jul 2020 09:51:23 +0000 (11:51 +0200)]
configs: sandbox: activate command env select and env load

Add support of environment location with the new env command:
'env select' and 'env load'

The ENV backend is selected by priority order
- 0 = "nowhere" (default at boot)
- 1 = "EXT4"

To test EXT4 env support, this backend is selected by name:
> env select EXT4

Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
4 years agoconfigs: sandbox: activate env in ext4 support
Patrick Delaunay [Tue, 28 Jul 2020 09:51:22 +0000 (11:51 +0200)]
configs: sandbox: activate env in ext4 support

Activate ENV in EXT4 support in sandbox.

The sandbox behavior don't change; the default environment with
the nowhere backend (CONFIG_ENV_IS_NOWHERE)is still used:
the weak function env_get_location() return ENVL_NOWHERE for priority 0.

Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
4 years agocmd: env: add env select command
Patrick Delaunay [Tue, 28 Jul 2020 09:51:21 +0000 (11:51 +0200)]
cmd: env: add env select command

Add the new command 'env select' to force the persistent storage
of environment, saved in gd->env_load_prio.

Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
4 years agocmd: env: add env load command
Patrick Delaunay [Tue, 28 Jul 2020 09:51:20 +0000 (11:51 +0200)]
cmd: env: add env load command

Add the new command env load to load the environment from
the current location gd->env_load_prio.

Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
4 years agoenv: the ops driver load becomes mandatory in struct env_driver
Patrick Delaunay [Tue, 28 Jul 2020 09:51:19 +0000 (11:51 +0200)]
env: the ops driver load becomes mandatory in struct env_driver

The ops driver load becomes mandatory in struct env_drive,
change the comment for this ops and remove unnecessary test.

Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
4 years agoenv: nowhere: add .load ops
Patrick Delaunay [Tue, 28 Jul 2020 09:51:18 +0000 (11:51 +0200)]
env: nowhere: add .load ops

Add the ops .load for nowhere ENV backend to load the
default environment.

This ops is needed for the command 'env load'

Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
4 years agoenv: correctly handle env_load_prio
Patrick Delaunay [Tue, 28 Jul 2020 09:51:17 +0000 (11:51 +0200)]
env: correctly handle env_load_prio

Only update gd->env_load_prio in generic function env_load()
and no more in the weak function env_get_location() which is
called in many place (for example in env_driver_lookup, even
for ENVOP_SAVE operation).

This patch is a preliminary step to use env_driver_lookup()/
env_get_location() in new function env_select() without
updating gd->env_load_prio.

Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
4 years agoenv: sf: avoid space in backend name
Patrick Delaunay [Tue, 28 Jul 2020 09:51:16 +0000 (11:51 +0200)]
env: sf: avoid space in backend name

Remove space in ENV backend name for SPI Flash (SF)
to avoid issue with env select command.

Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
4 years agoenv: ext4: set gd->env_valid
Patrick Delaunay [Tue, 28 Jul 2020 09:51:15 +0000 (11:51 +0200)]
env: ext4: set gd->env_valid

Add a missing initialization of gd->env_valid in env_ext4_load
as it is already done in some other env device.

Set gd->env_valid = ENV_VALID in env_ext4_save() and env_ext4_load().

This patch allows to have a correct information in 'env info' command.

Reviewed-by: Tom Rini <trini@konsulko.com>
Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
4 years agoenv: add absolute path at CONFIG_ENV_EXT4_FILE
Patrick Delaunay [Tue, 28 Jul 2020 09:51:14 +0000 (11:51 +0200)]
env: add absolute path at CONFIG_ENV_EXT4_FILE

Add the absolute path to the default value of
CONFIG_ENV_EXT4_FILE = "/uboot.env".

This patch avoid the error :
  Saving Environment to EXT4... File System is consistent
  Please supply Absolute path

Reviewed-by: Tom Rini <trini@konsulko.com>
Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
4 years agoenv: add prototypes for weak function
Patrick Delaunay [Mon, 15 Jun 2020 14:52:17 +0000 (16:52 +0200)]
env: add prototypes for weak function

This patch adds prototypes for several weak functions:
- env_ext4_get_intf
- env_ext4_get_dev_part
- env_get_location

It solves the following warnings when compiling with W=1
on stm32mp1 board:

board/st/stm32mp1/stm32mp1.c:849:19: warning: no previous prototype for 'env_get_location' [-Wmissing-prototypes]
 enum env_location env_get_location(enum env_operation op, int prio)
                   ^~~~~~~~~~~~~~~~
board/st/stm32mp1/stm32mp1.c:876:13: warning: no previous prototype for 'env_ext4_get_intf' [-Wmissing-prototypes]
 const char *env_ext4_get_intf(void)
             ^~~~~~~~~~~~~~~~~
board/st/stm32mp1/stm32mp1.c:889:13: warning: no previous prototype for 'env_ext4_get_dev_part' [-Wmissing-prototypes]
 const char *env_ext4_get_dev_part(void)
             ^~~~~~~~~~~~~~~~~~~~~

Signed-off-by: Patrice Chotard <patrice.chotard@st.com>
Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
4 years agotest: efi_selftest: Do not force serial# setting
Tom Rini [Fri, 31 Jul 2020 13:45:02 +0000 (09:45 -0400)]
test: efi_selftest: Do not force serial# setting

As part of the EFI self test we set and check the serial# variable.
However, we should not be forcing this setting.  In the case where we
are allowed to change the variable it will change, and we will pass the
test.  In the case where we cannot change it, force may or may not be
allowed, depending on further environment restrictions.  Drop the -f
flag here as we do not need it.

Cc: Heinrich Schuchardt <xypron.glpk@gmx.de>
Signed-off-by: Tom Rini <trini@konsulko.com>
4 years agotest: do not rely on => being the prompt
Heinrich Schuchardt [Fri, 24 Jul 2020 18:55:37 +0000 (20:55 +0200)]
test: do not rely on => being the prompt

In our tests we should use the customized prompt for testing.

Reported-by: Tom Rini <trini@konsulko.com>
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Tested-by: Tom Rini <trini@konsulko.com>
4 years agoMerge tag 'dm-pull-28jul20' of git://git.denx.de/u-boot-dm
Tom Rini [Thu, 30 Jul 2020 01:16:08 +0000 (21:16 -0400)]
Merge tag 'dm-pull-28jul20' of git://git.denx.de/u-boot-dm

Use binman instead of one of the Rockchip build scripts
Refactor to allow any arch to create SPI-flash images
New button uclass

4 years agoMerge https://gitlab.denx.de/u-boot/custodians/u-boot-stm
Tom Rini [Wed, 29 Jul 2020 20:30:45 +0000 (16:30 -0400)]
Merge https://gitlab.denx.de/u-boot/custodians/u-boot-stm

- fix SPL boot issue due to early dbgmcu_init() call
- fix SPL boot issue due to dcache memory region configuration
- add support of CONFIG_ENV_IS_IN_MMC
- add specific SD/eMMC partition for U-Boot enviromnent
- enable env in SPL
- use "env info -q" to remove log during boot
- remove env location override for dh_stm32mp1
- update management of misc_read
- check result of find_mmc_device in stm32prog
- use regulator_set_enable_if_allowed for disabling vdd supply in usbphyc
- enable CMD_ADTIMG flag to handle Android images
- device tree alignment with Linux Kernel v5.8-rc1
- remove hnp-srp-disable for usbotg on dk1
- add reset support to uart nodes on stm32mp15x
- use correct weak function name spl_board_prepare_for_linux
- use cd-gpios for ST and DHSOM boards
- add seeed studio odyssey-stm32mp157c board support
- move ethernet PHY into SoM DT
- add DHSOM based DRC02 board support

4 years agoMerge branch '2020-07-29-add-brcm-ns3-support'
Tom Rini [Wed, 29 Jul 2020 18:24:17 +0000 (14:24 -0400)]
Merge branch '2020-07-29-add-brcm-ns3-support'

- Add initial Broadcom NS3 SoC support.

4 years agoarch: arm: use dt and UCLASS_SYSCON to get gic lpi details
Rayagonda Kokatanur [Sun, 26 Jul 2020 17:07:33 +0000 (22:37 +0530)]
arch: arm: use dt and UCLASS_SYSCON to get gic lpi details

Use device tree and UCLASS_SYSCON driver to get
Generic Interrupt Controller (GIC) lpi address and
maximum GIC redistributors count.

Also update Kconfig to select REGMAP and SYSCON when
GIC_V3_ITS is enabled.

Signed-off-by: Rayagonda Kokatanur <rayagonda.kokatanur@broadcom.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
4 years agoarch: arm: use dt and UCLASS_IRQ to get gic details
Rayagonda Kokatanur [Sun, 26 Jul 2020 17:07:32 +0000 (22:37 +0530)]
arch: arm: use dt and UCLASS_IRQ to get gic details

Use device tree and UCLASS_IRQ driver to get following
Generic Interrupt Controller (GIC) details,

-GIC Distributor interface (GICD) base address and
-GIC Redistributors (GICR) base address.

Signed-off-by: Rayagonda Kokatanur <rayagonda.kokatanur@broadcom.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
4 years agoconfigs: ns3: enable tee and optee driver
Rayagonda Kokatanur [Wed, 15 Jul 2020 17:27:20 +0000 (22:57 +0530)]
configs: ns3: enable tee and optee driver

Enable tee and optee drivers.

Signed-off-by: Vikas Gupta <vikas.gupta@broadcom.com>
Signed-off-by: Rayagonda Kokatanur <rayagonda.kokatanur@broadcom.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
4 years agodrivers: tee: broadcom: add optee based bnxt fw load driver
Vikas Gupta [Wed, 15 Jul 2020 17:27:19 +0000 (22:57 +0530)]
drivers: tee: broadcom: add optee based bnxt fw load driver

Add optee based bnxt fw load driver.
bnxt is Broadcom NetXtreme controller Ethernet card.
This driver is used to load bnxt firmware binary using OpTEE.

Signed-off-by: Vikas Gupta <vikas.gupta@broadcom.com>
Signed-off-by: Rayagonda Kokatanur <rayagonda.kokatanur@broadcom.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
4 years agoboard: ns3: add development keys used in FIT
Pramod Kumar [Wed, 15 Jul 2020 17:25:41 +0000 (22:55 +0530)]
board: ns3: add development keys used in FIT

Add development keys used in FIT.

Signed-off-by: Pramod Kumar <pramod.kumar@broadcom.com>
Signed-off-by: Rayagonda Kokatanur <rayagonda.kokatanur@broadcom.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
4 years agoboard: ns3: add FIT image its file
Pramod Kumar [Wed, 15 Jul 2020 17:25:40 +0000 (22:55 +0530)]
board: ns3: add FIT image its file

Add FIT image its file.

Signed-off-by: Pramod Kumar <pramod.kumar@broadcom.com>
Signed-off-by: Rayagonda Kokatanur <rayagonda.kokatanur@broadcom.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
4 years agoconfigs: ns3: enable FIT config
Rayagonda Kokatanur [Wed, 15 Jul 2020 17:25:39 +0000 (22:55 +0530)]
configs: ns3: enable FIT config

Enable FIT config for NS3.

Signed-off-by: Rayagonda Kokatanur <rayagonda.kokatanur@broadcom.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
4 years agoconfigs: ns3: enable sp805 watchdog driver
Rayagonda Kokatanur [Wed, 15 Jul 2020 17:23:07 +0000 (22:53 +0530)]
configs: ns3: enable sp805 watchdog driver

Enable sp805 watchdog driver for ns3.

Signed-off-by: Rayagonda Kokatanur <rayagonda.kokatanur@broadcom.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
4 years agoconfigs: ns3: enable EXT4 and FAT fs support
Rayagonda Kokatanur [Wed, 15 Jul 2020 17:23:06 +0000 (22:53 +0530)]
configs: ns3: enable EXT4 and FAT fs support

Enable EXT4 and FAT fs support for ns3.

Signed-off-by: Rayagonda Kokatanur <rayagonda.kokatanur@broadcom.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
4 years agoconfigs: ns3: enable gpt commands
Rayagonda Kokatanur [Wed, 15 Jul 2020 17:23:05 +0000 (22:53 +0530)]
configs: ns3: enable gpt commands

Enable gpt commands for ns3.

Signed-off-by: Rayagonda Kokatanur <rayagonda.kokatanur@broadcom.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
4 years agoconfigs: ns3: enable mmc commands
Rayagonda Kokatanur [Wed, 15 Jul 2020 17:23:04 +0000 (22:53 +0530)]
configs: ns3: enable mmc commands

Enable mmc commands for NS3.

Signed-off-by: Rayagonda Kokatanur <rayagonda.kokatanur@broadcom.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
4 years agoconfigs: ns3: enable BCM IPROC mmc driver
Rayagonda Kokatanur [Wed, 15 Jul 2020 17:23:03 +0000 (22:53 +0530)]
configs: ns3: enable BCM IPROC mmc driver

Enable BCM IPROC mmc driver ns3.
Enable DMA for MMC Host to have better reads and writes.

Signed-off-by: Rayagonda Kokatanur <rayagonda.kokatanur@broadcom.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
4 years agodt-bindings: pinctrl: add ns3 pads definition
Rayagonda Kokatanur [Wed, 15 Jul 2020 17:23:02 +0000 (22:53 +0530)]
dt-bindings: pinctrl: add ns3 pads definition

Add NS3 pads definitions.

Signed-off-by: Rayagonda Kokatanur <rayagonda.kokatanur@broadcom.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
4 years agoconfigs: ns3: enable pinctrl driver
Rayagonda Kokatanur [Wed, 15 Jul 2020 17:23:01 +0000 (22:53 +0530)]
configs: ns3: enable pinctrl driver

Enable pinctrl driver for ns3.

Signed-off-by: Rayagonda Kokatanur <rayagonda.kokatanur@broadcom.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
4 years agogpio: do not include <asm/arch/gpio.h> on TARGET_BCMNS3
Rayagonda Kokatanur [Tue, 5 May 2020 17:56:47 +0000 (23:26 +0530)]
gpio: do not include <asm/arch/gpio.h> on TARGET_BCMNS3

As no gpio.h is defined for this architecture, to avoid
compilation failure, do not include <asm/arch/gpio.h> for
arch bcmns3.

Signed-off-by: Rayagonda Kokatanur <rayagonda.kokatanur@broadcom.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
4 years agodrivers: gpio: add broadcom iproc gpio driver support
Rayagonda Kokatanur [Tue, 5 May 2020 17:56:46 +0000 (23:26 +0530)]
drivers: gpio: add broadcom iproc gpio driver support

Add gpio driver support for Broadcom iproc-based socs.

Signed-off-by: Rayagonda Kokatanur <rayagonda.kokatanur@broadcom.com>
Signed-off-by: Sheetal Tigadoli <sheetal.tigadoli@broadcom.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
4 years agoMAINTAINERS: update maintainers for broadcom ns3 platform
Rayagonda Kokatanur [Wed, 15 Jul 2020 17:19:09 +0000 (22:49 +0530)]
MAINTAINERS: update maintainers for broadcom ns3 platform

Update MAINTAINERS for broadcom ns3 platform (TARGET_NS3).

Signed-off-by: Rayagonda Kokatanur <rayagonda.kokatanur@broadcom.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
4 years agodoc: add README doc for bcmns3 platform
Rayagonda Kokatanur [Wed, 15 Jul 2020 17:19:08 +0000 (22:49 +0530)]
doc: add README doc for bcmns3 platform

Add README doc for bcmns3 platform.

Signed-off-by: Rayagonda Kokatanur <rayagonda.kokatanur@broadcom.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
4 years agoinclude/configs: ns3: add support for flashing images
Bharat Gooty [Wed, 15 Jul 2020 17:19:07 +0000 (22:49 +0530)]
include/configs: ns3: add support for flashing images

Add support for flashing images into QSPI and eMMC.

Signed-off-by: Bharat Gooty <bharat.gooty@broadcom.com>
Signed-off-by: Rayagonda Kokatanur <rayagonda.kokatanur@broadcom.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
4 years agoinclude/configs: ns3: add env variables for Linux boot
Bharat Gooty [Wed, 15 Jul 2020 17:19:06 +0000 (22:49 +0530)]
include/configs: ns3: add env variables for Linux boot

Add env variables and commands for booting Linux.

Signed-off-by: Bharat Gooty <bharat.gooty@broadcom.com>
Signed-off-by: Rayagonda Kokatanur <rayagonda.kokatanur@broadcom.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
4 years agoboard: ns3: limit U-boot relocation within 16MB memory
Bharat Kumar Reddy Gooty [Wed, 15 Jul 2020 17:19:05 +0000 (22:49 +0530)]
board: ns3: limit U-boot relocation within 16MB memory

By default relocation happens to a higher address of DDR,
i.e, DDR start + DDR size.

U-Boot shall be used to collect the ramdump.
Restrict U-Boot to use only the 16MB memory, so that this
memory can be reserved. Limit relocation to happen within
16MB memory, start 0xFF00_0000 and end 0x1_0000_0000

Signed-off-by: Bharat Kumar Reddy Gooty <bharat.gooty@broadcom.com>
Signed-off-by: Rayagonda Kokatanur <rayagonda.kokatanur@broadcom.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
4 years agoboard: ns3: define ddr memory layout
Rayagonda Kokatanur [Wed, 15 Jul 2020 17:19:04 +0000 (22:49 +0530)]
board: ns3: define ddr memory layout

Add both DRAM banks memory information and
the corresponding MMU page table mappings.

Signed-off-by: Bharat Kumar Reddy Gooty <bharat.gooty@broadcom.com>
Signed-off-by: Rayagonda Kokatanur <rayagonda.kokatanur@broadcom.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
4 years agodt-bindings: memory: ns3: add ddr memory definition
Rayagonda Kokatanur [Wed, 15 Jul 2020 17:19:03 +0000 (22:49 +0530)]
dt-bindings: memory: ns3: add ddr memory definition

Add ddr memory definitions.

Signed-off-by: Rayagonda Kokatanur <rayagonda.kokatanur@broadcom.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
4 years agoconfigs: ns3: enable GIC_V3 ITS
Rayagonda Kokatanur [Wed, 15 Jul 2020 17:19:02 +0000 (22:49 +0530)]
configs: ns3: enable GIC_V3 ITS

Enables the Generic Interrupt Controller (GIC) V3
Interrupt Translation Service (ITS) Locality-specific Peripheral
Interrupts (LPI) configuration table and LPI table.

Signed-off-by: Rayagonda Kokatanur <rayagonda.kokatanur@broadcom.com>
Signed-off-by: Bharat Kumar Reddy Gooty <bharat.gooty@broadcom.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
4 years agoboard: ns3: program GIC LPI tables
Rayagonda Kokatanur [Wed, 15 Jul 2020 17:19:01 +0000 (22:49 +0530)]
board: ns3: program GIC LPI tables

U-boot programs the GIC LPI configuration tables and enables
the LPI table.

Signed-off-by: Bharat Kumar Reddy Gooty <bharat.gooty@broadcom.com>
Signed-off-by: Rayagonda Kokatanur <rayagonda.kokatanur@broadcom.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
4 years agoboard: ns3: default reset type to L3
Rayagonda Kokatanur [Wed, 15 Jul 2020 17:19:00 +0000 (22:49 +0530)]
board: ns3: default reset type to L3

Default "reset" from U-Boot to L3 reset.
"reset" command with argument will trigger L1 reset.

Signed-off-by: Rajesh Ravi <rajesh.ravi@broadcom.com>
Signed-off-by: Bharat Kumar Reddy Gooty <bharat.gooty@broadcom.com>
Signed-off-by: Rayagonda Kokatanur <rayagonda.kokatanur@broadcom.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
4 years agoboard: ns3: add api to save boot parameters passed from BL31
Abhishek Shah [Wed, 15 Jul 2020 17:18:59 +0000 (22:48 +0530)]
board: ns3: add api to save boot parameters passed from BL31

Add API to save boot parameters passed from BL31

Use assembly implementation of save_boot_params instead of c function.
Because generally ATF does not set up SP_EL2 on exiting.
Thus, usage of a C function immediately after exiting with no stack
setup done by ATF explicitly, may cause SP_EL2 to be not sane,
which in turn causes a crash if this boot was not lucky to get
an SP_EL2 in valid range. Replace C implementation with assembly one
which does not use stack this early, and let u-boot to set up its stack
later.

Signed-off-by: Abhishek Shah <abhishek.shah@broadcom.com>
Signed-off-by: Rajesh Ravi <rajesh.ravi@broadcom.com>
Signed-off-by: Vladimir Olovyannikov <vladimir.olovyannikov@broadcom.com>
Signed-off-by: Rayagonda Kokatanur <rayagonda.kokatanur@broadcom.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
4 years agodt-bindings: memory: ns3: add memory definitions
Rayagonda Kokatanur [Wed, 15 Jul 2020 17:18:58 +0000 (22:48 +0530)]
dt-bindings: memory: ns3: add memory definitions

Add NS3 memory definitions.

Signed-off-by: Rayagonda Kokatanur <rayagonda.kokatanur@broadcom.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
4 years agoconfigs: ns3: enable clock subsystem
Rayagonda Kokatanur [Wed, 15 Jul 2020 17:18:57 +0000 (22:48 +0530)]
configs: ns3: enable clock subsystem

Enable clock subsystem for ns3.

Signed-off-by: Rayagonda Kokatanur <rayagonda.kokatanur@broadcom.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
4 years agoarm: cpu: armv8: add L3 memory flush support
Rayagonda Kokatanur [Wed, 15 Jul 2020 17:18:56 +0000 (22:48 +0530)]
arm: cpu: armv8: add L3 memory flush support

Add L3 memory flush support for NS3.

Signed-off-by: Rayagonda Kokatanur <rayagonda.kokatanur@broadcom.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
4 years agoboard: ns3: add support for Broadcom Northstar 3
Rayagonda Kokatanur [Wed, 15 Jul 2020 17:18:55 +0000 (22:48 +0530)]
board: ns3: add support for Broadcom Northstar 3

Add support for Broadcom Northstar 3 SoC.
NS3 is a octo-core 64-bit ARMv8 Cortex-A72 processors
targeting a broad range of networking applications.

Signed-off-by: Rayagonda Kokatanur <rayagonda.kokatanur@broadcom.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
4 years agoMerge branch '2020-07-28-misc-soc-improvements'
Tom Rini [Wed, 29 Jul 2020 13:26:11 +0000 (09:26 -0400)]
Merge branch '2020-07-28-misc-soc-improvements'

- Assorted MediaTek improvements
- s5p4418 support
- QEMU ARM platform improvements
- Qualcomm IPQ40xx support

4 years agoboard: mediatek: Add support for UniElec U7623 board
David Woodhouse [Sun, 12 Jul 2020 22:33:03 +0000 (23:33 +0100)]
board: mediatek: Add support for UniElec U7623 board

This is an MT7623A-based board, very similar to the Banana Pi R2.

http://www.unielecinc.com/q/news/cn/p/product/detail.html?qd_guid=OjXwKCaRlN

Signed-off-by: David Woodhouse <dwmw2@infradead.org>
4 years agoboard: mediatek: Use CONFIG_DEFAULT_FDT_FILE for default environment
David Woodhouse [Sun, 12 Jul 2020 22:33:02 +0000 (23:33 +0100)]
board: mediatek: Use CONFIG_DEFAULT_FDT_FILE for default environment

Rather than hard-coding it to the Banana Pi R2.

Signed-off-by: David Woodhouse <dwmw2@infradead.org>
4 years agoboard: mediatek: fix mmc_get_boot_dev() for platforms without external SD
David Woodhouse [Sun, 12 Jul 2020 22:33:01 +0000 (23:33 +0100)]
board: mediatek: fix mmc_get_boot_dev() for platforms without external SD

On the UniElec U7623 board there is no external SD slot and the preloader
doesn't fill in the magic field at 0x81dffff0 to indicate that it was
booted from eMMC.

Signed-off-by: David Woodhouse <dwmw2@infradead.org>
4 years agoarm: add (default) config for nanopi2 board
Stefan Bosch [Fri, 10 Jul 2020 17:07:38 +0000 (19:07 +0200)]
arm: add (default) config for nanopi2 board

Changes in relation to FriendlyARM's U-Boot nanopi2-v2016.01:
- Configuration changed, mainly several "CONFIG_..." moved from
  s5p4418_nanopi2.h to s5p4418_nanopi2_defconfig and USB related
  configs removed because USB is not supported yet.
- s5p4418_nanopi2.h: "CONFIG_" removed from several s5p4418/nanopi2
  specific defines because the appropriate values do not need to be
  configurable.
- pinctrl is supported now, therefore "CONFIG_PINCTRL=y" added to
  s5p4418_nanopi2_defconfig.

Signed-off-by: Stefan Bosch <stefan_b@posteo.net>
4 years agoarm: add support for SoC s5p4418 (cpu) / nanopi2 board
Stefan Bosch [Fri, 10 Jul 2020 17:07:37 +0000 (19:07 +0200)]
arm: add support for SoC s5p4418 (cpu) / nanopi2 board

Changes in relation to FriendlyARM's U-Boot nanopi2-v2016.01:
- SPL not supported yet --> no spl-dir in arch/arm/cpu/armv7/s5p4418/.
  Appropriate line in Makefile removed.
- cpu.c: '#include <cpu_func.h>' added.
- arch/arm/cpu/armv7/s5p4418/u-boot.lds removed, is not required
  anylonger.
- "obj-$(CONFIG_ARCH_NEXELL) += s5p-common/" added to
  arch/arm/cpu/armv7/Makefile since s5p-common/pwm.c is used instead
  of drivers/pwm/pwm-nexell.c.
- s5p4418.dtsi: '#include "../../../include/generated/autoconf.h"'
  removed, is not necessary, error at out-of-tree building.
  '#ifdef CONFIG_CPU_NXP4330'-blocks (2x) removed. Some minor changes
  regarding mmc. 'u-boot,dm-pre-reloc' added to dp0 because of added
  DM_VIDEO support.
- board/s5p4418/ renamed to board/friendlyarm/
- All s5p4418-boards except nanopi2 removed because there is no
  possibility to test the other boards.
- Kconfig: Changes to have a structure like mach-bcm283x (RaspberryPi),
  e.g. "config ..." entries moved from/to other Kconfig.
- "CONFIG_" removed from several s5p4418/nanopi2 specific defines
  because the appropriate values do not need to be configurable.
- nanopi2/board.c: All getenv(), getenv_ulong(), setenv() and saveenv()
  renamed to env_get(), env_get_ulong(), env_set() and env_save(),
  respectively. MACH_TYPE_S5P4418 is not defined anymore, therefore
  appropriate code removed (not necessary for DT-kernels).
- nanopi2/onewire.c: All crc8() renamed to crc8_ow() because crc8() is
  already defined in lib/crc8.c (with different parameters).
- dts: "nexell,s5pxx18-i2c" used instead of "i2c-gpio", i2c0 and
  i2c1 added. gmac-, ehci- and dwc2otg-entries removed because the
  appropriate functionality is not supported yet. New mmc-property
  "mmcboost" added.
  s5p4418-pinctrl.dtsi: gmac-entries removed, mmc- and i2c-entries
  added.
- '#ifdef CONFIG...' changed to 'if (IS_ENABLED(CONFIG...))' where
  possible (and similar).

Signed-off-by: Stefan Bosch <stefan_b@posteo.net>
4 years agovideo: add nexell video driver (display/video driver)
Stefan Bosch [Fri, 10 Jul 2020 17:07:36 +0000 (19:07 +0200)]
video: add nexell video driver (display/video driver)

Changes in relation to FriendlyARM's U-Boot nanopi2-v2016.01:
- nexell_display.c: Changed to DM, CONFIG_FB_ADDR can not be used
  anymore because framebuffer is allocated by video_reserve() in
  video-uclass.c. Therefore code changed appropriately.
- '#ifdef CONFIG...' changed to 'if (IS_ENABLED(CONFIG...))' where
  possible (and similar).
- livetree API (dev_read_...) is used instead of fdt one (fdt...).

Signed-off-by: Stefan Bosch <stefan_b@posteo.net>
4 years agovideo: add nexell video driver (soc: dpc, makefile)
Stefan Bosch [Fri, 10 Jul 2020 17:07:35 +0000 (19:07 +0200)]
video: add nexell video driver (soc: dpc, makefile)

Low level functions for DPC (Display Controller) and Makefile for all
nexell video low level functions.

Signed-off-by: Stefan Bosch <stefan_b@posteo.net>
4 years agovideo: add nexell video driver (soc: lvds, hdmi)
Stefan Bosch [Fri, 10 Jul 2020 17:07:34 +0000 (19:07 +0200)]
video: add nexell video driver (soc: lvds, hdmi)

Low level functions for LVDS and HDMI display interfaces.

Signed-off-by: Stefan Bosch <stefan_b@posteo.net>
4 years agovideo: add nexell video driver (soc: mlc, mipi)
Stefan Bosch [Fri, 10 Jul 2020 17:07:33 +0000 (19:07 +0200)]
video: add nexell video driver (soc: mlc, mipi)

Low level functions for MLC (Multi Layer Control) and MIPI (Mobile
Industry Processor Interface).

Signed-off-by: Stefan Bosch <stefan_b@posteo.net>
4 years agovideo: add nexell video driver (soc: displaytop)
Stefan Bosch [Fri, 10 Jul 2020 17:07:32 +0000 (19:07 +0200)]
video: add nexell video driver (soc: displaytop)

Low level functions for DisplayTop (Display Topology).

Signed-off-by: Stefan Bosch <stefan_b@posteo.net>
4 years agopwm: add driver for nexell
Stefan Bosch [Fri, 10 Jul 2020 17:07:31 +0000 (19:07 +0200)]
pwm: add driver for nexell

Changes in relation to FriendlyARM's U-Boot nanopi2-v2016.01:
- Since drivers/pwm/pwm-nexell.c is an adapted version of
  s5p-common/pwm.c an appropriately changed version of s5p-common/pwm.c
  is used instead. Therefore arch/arm/mach-s5pc1xx/include/mach/pwm.h
  copied to arch/arm/mach-nexell/include/mach and s5p-common/Makefile
  changed appropriately.
- '#ifdef CONFIG...' changed to 'if (IS_ENABLED(CONFIG...))' where
  possible (and similar).

Signed-off-by: Stefan Bosch <stefan_b@posteo.net>
4 years agopinctrl: add nexell driver
Stefan Bosch [Fri, 10 Jul 2020 17:07:30 +0000 (19:07 +0200)]
pinctrl: add nexell driver

Changes in relation to FriendlyARM's U-Boot nanopi2-v2016.01:
- livetree API (dev_read_...) is used instead of fdt one (fdt...).
- doc/device-tree-bindings/pinctrl/nexell,s5pxx18-pinctrl.txt added.

Signed-off-by: Stefan Bosch <stefan_b@posteo.net>
4 years agommc: add nexell driver
Stefan Bosch [Fri, 10 Jul 2020 17:07:29 +0000 (19:07 +0200)]
mmc: add nexell driver

Changes in relation to FriendlyARM's U-Boot nanopi2-v2016.01:
- driver changed to DM.
- pinctrl-driver/dt is used now instead of configuring the mmc I/O-pins
  in the mmc-driver.
- nexell_dwmmc_ofdata_to_platdata() reworked, i.e. valid default values
  are used now (where possible) and the appropriate if-blocks have
  been removed.
- new dt-property "mmcboost" is used now instead of "CONFIG_BOOST_MMC"
  which was not defined anywhere.

Signed-off-by: Stefan Bosch <stefan_b@posteo.net>
4 years agoi2c: add nexell driver
Stefan Bosch [Fri, 10 Jul 2020 17:07:28 +0000 (19:07 +0200)]
i2c: add nexell driver

Changes in relation to FriendlyARM's U-Boot nanopi2-v2016.01:
- i2c/nx_i2c.c: Some adaptions mainly because of changes in
  "struct udevice".
- several Bugfixes in nx_i2c.c.
- the driver has been for s5p6818 only. Code extended appropriately
  in order s5p4418 is also working.
- "probe_chip" added.
- pinctrl-driver/dt is used instead of configuring the i2c I/O-pins
  in the i2c-driver.
- '#ifdef CONFIG...' changed to 'if (IS_ENABLED(CONFIG...))' where
  possible (and similar).
- livetree API (dev_read_...) is used instead of fdt one (fdt...).

Signed-off-by: Stefan Bosch <stefan_b@posteo.net>
4 years agogpio: add nexell driver
Stefan Bosch [Fri, 10 Jul 2020 17:07:27 +0000 (19:07 +0200)]
gpio: add nexell driver

Changes in relation to FriendlyARM's U-Boot nanopi2-v2016.01:
- livetree API (dev_read_...) is used instead of fdt one (fdt...).

Signed-off-by: Stefan Bosch <stefan_b@posteo.net>
4 years agoarm: add mach-nexell (all files except header files)
Stefan Bosch [Fri, 10 Jul 2020 17:07:26 +0000 (19:07 +0200)]
arm: add mach-nexell (all files except header files)

Changes in relation to FriendlyARM's U-Boot nanopi2-v2016.01:
- SPL not supported yet --> no spl-directory in arch/arm/mach-nexell.
  Appropriate line in Makefile removed.
- clock.c: 'section(".data")' added to declaration of clk_periphs[] and
  core_hz.
- Kconfig: Changes to have a structure like in mach-bcm283x/Kconfig,
  e.g. "config ..." entries moved from other Kconfig.
- timer.c: 'section(".data")' added to declaration of timestamp and
  lastdec.
- arch/arm/mach-nexell/serial.c removed because this is for the UARTs
  of the S5P6818 SoC which is not supported yet. S5P4418 UARTs are
  different, here the (existing) PL011-code is used.
- '#ifdef CONFIG...' changed to 'if (IS_ENABLED(CONFIG...))' where
  possible (and similar).

Signed-off-by: Stefan Bosch <stefan_b@posteo.net>
4 years agoarm: add mach-nexell (header files)
Stefan Bosch [Fri, 10 Jul 2020 17:07:25 +0000 (19:07 +0200)]
arm: add mach-nexell (header files)

Changes in relation to FriendlyARM's U-Boot nanopi2-v2016.01:
- DM_VIDEO support (display_dev.h).
- boot0.h added, handles NSIH --> tools/nexell obsolete.
- gpio.h: Include-path to errno.h changed.

Signed-off-by: Stefan Bosch <stefan_b@posteo.net>
4 years agoarm: qemu: override flash accessors to use virtualizable instructions
Ard Biesheuvel [Tue, 7 Jul 2020 10:07:11 +0000 (12:07 +0200)]
arm: qemu: override flash accessors to use virtualizable instructions

Some instructions in the ARM ISA have multiple output registers, such
as ldrd/ldp (load pair), where two registers are loaded from memory,
but also ldr with indexing, where the memory base register is incremented
as well when the value is loaded to the destination register.

MMIO emulation under KVM is based on using the architecturally defined
syndrome information that is provided when an exception is taken to the
hypervisor. This syndrome information describes whether the instruction
that triggered the exception is a load or a store, what the faulting
address was, and which register was the destination register.

This syndrome information can only describe one destination register, and
when the trapping instruction is one with multiple outputs, KVM throws an
error like

  kvm [615929]: Data abort outside memslots with no valid syndrome info

on the host and kills the QEMU process with the following error:

  U-Boot 2020.07-rc3-00208-g88bd5b179360-dirty (Jun 06 2020 - 11:59:22 +0200)

  DRAM:  1 GiB
  Flash: error: kvm run failed Function not implemented
  R00=00000001 R01=00000040 R02=7ee0ce20 R03=00000000
  R04=7ffd9eec R05=00000004 R06=7ffda3f8 R07=00000055
  R08=7ffd9eec R09=7ef0ded0 R10=7ee0ce20 R11=00000000
  R12=00000004 R13=7ee0cdf8 R14=00000000 R15=7ff72d08
  PSR=200001d3 --C- A svc32
  QEMU: Terminated

This means that, in order to run U-Boot in QEMU under KVM, we need to
avoid such instructions when accessing emulated devices. For the flash
in particular, which is a hybrid between a ROM (backed by a read-only
KVM memslot) when in array mode, and an emulated MMIO device (when in
write mode), we need to take care to only use instructions that KVM can
deal with when they trap.

So override the flash read accessors that are used when running on QEMU
under KVM. Note that the the 64-bit wide read and write accessors have
been omitted: they are never used when running under QEMU given that it
does not emulate CFI flash that supports it.

Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Reviewed-by: Andre Przywara <andre.przywara@arm.com>
4 years agoarm: qemu: disable the EFI workaround for older GRUB
Ard Biesheuvel [Tue, 7 Jul 2020 10:07:10 +0000 (12:07 +0200)]
arm: qemu: disable the EFI workaround for older GRUB

The QEMU/mach-virt targeted port of u-boot currently only runs on
QEMU under TCG emulation, which does not model the caches at all,
and so no users can exist that are relying on the GRUB hack for
EFI boot.

We will shortly enable support for running under KVM, but the GRUB
hack (which disables all caches without doing cache cleaning by VA
during ExitBootServices()) is likely to cause more problems than it
solves, given that KVM hosts require correct maintenance if they
incorporate non-architected system caches.

So let's disable the GRUB hack by default on the QEMU/mach-virt
port.

Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Reviewed-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
4 years agoarm: qemu: implement enable_caches()
Ard Biesheuvel [Tue, 7 Jul 2020 10:07:09 +0000 (12:07 +0200)]
arm: qemu: implement enable_caches()

Add an override for enable_caches to enable the I and D caches, along
with the cached 1:1 mapping of all of DRAM. This is needed for running
U-Boot under virtualization with QEMU/kvm.

Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Reviewed-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Reviewed-by: Andre Przywara <andre.przywara@arm.com>
4 years agoarm: qemu: enable LPAE on 32-bit
Ard Biesheuvel [Tue, 7 Jul 2020 10:07:08 +0000 (12:07 +0200)]
arm: qemu: enable LPAE on 32-bit

QEMU's mach-virt machine only supports selecting CPU models that
implement the virtualization extensions, and are therefore guaranteed
to support LPAE as well.

Initially, QEMU would not allow emulating these CPUs running in HYP
mode (or EL2, for AArch64), but today, it also contains a complete
implementation of the virtualization extensions themselves.

This means we could be running U-Boot in HYP mode, in which case the
LPAE long descriptor page table format is the only format that is
supported. If we are not running in HYP mode, we can use either.

So let's enable CONFIG_ARMV7_LPAE for qemu_arm_defconfig so that we
get the best support for running with the MMU and caches enabled at
any privilege level.

Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Acked-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
4 years agoarm: enable allocate-on-read for LPAE's DCACHE_WRITEBACK/_WRITETHROUGH
Ard Biesheuvel [Tue, 7 Jul 2020 10:07:07 +0000 (12:07 +0200)]
arm: enable allocate-on-read for LPAE's DCACHE_WRITEBACK/_WRITETHROUGH

The LPAE versions of DCACHE_WRITEBACK and DCACHE_WRITETHROUGH are currently
defined as no-allocate for both reads and writes, which deviates from the
non-LPAE definition, and mostly defeats the purpose of enabling the caches
in the first place.

So align LPAE with !LPAE, and enable allocate-on-read for both. And while
at it, add some clarification about the meaning of the chosen values.

Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Reviewed-by: Andre Przywara <andre.przywara@arm.com>
4 years agomsm_gpio: Add support for Qualcomm IPQ40xx
Robert Marko [Mon, 6 Jul 2020 08:37:56 +0000 (10:37 +0200)]
msm_gpio: Add support for Qualcomm IPQ40xx

Snapdragon SoCs and IPQ40xx use common TLMM IP,
so existing driver supports IPQ40xx as well.

So lets simply add a compatible for IPQ40xx.

Signed-off-by: Robert Marko <robert.marko@sartura.hr>
Reviewed-By: Ramon Fried <rfried.dev@gmail.com>
4 years agomsm_serial: Read bit rate register value from DT
Robert Marko [Mon, 6 Jul 2020 08:37:55 +0000 (10:37 +0200)]
msm_serial: Read bit rate register value from DT

IPQ40xx and currently supported Snapdragon boards don't use the same one
so enable reading it from DT, if no DT property is found default value
is the same as the previous define.

Signed-off-by: Robert Marko <robert.marko@sartura.hr>
Reviewed-By: Ramon Fried <rfried.dev@gmail.com>
4 years agoarm: Add support for Qualcomm IPQ40xx family
Robert Marko [Mon, 6 Jul 2020 08:37:54 +0000 (10:37 +0200)]
arm: Add support for Qualcomm IPQ40xx family

This introduces initial support for the popular Qualcomm
IPQ40x8 and IPQ40x9 WiSoC series.

IPQ40xx series have 4x Cortex A7 ARM-v7A cores.
Supported are: IPQ4018, IPQ4019, IPQ4028 and IPQ4029.

IPQ40x8 and IPQ40x9 use the same cores, but differ in
addressable RAM size (1GB for IPQ40x9 and 256MB for IPQ40x8)
and supported peripherals (IPQ40x8 lacks RGMII, LCD controller
and EMMC/SDHCI controllers).

IQP4028/IPQ4029 models differ from IPQ4018/IPQ4019 only
by their rated temperatures rates with IPQ402X models being
rated for wider temperature ranges.

Initially this supports:
* Simple clock driver (Only for UART1 now, will be extended)
* Pinctrl driver (Supports UARTX and GPIO now, will be extended)
* GPIOs already supported by msm_gpio driver with updates
* UARTs already supported by serial_msm driver with updates

Further peripherals will come in later patches.

Signed-off-by: Robert Marko <robert.marko@sartura.hr>
4 years agopinctrl: mediatek: add PUPD/R0/R1 support for MT7623
David Woodhouse [Fri, 19 Jun 2020 11:40:20 +0000 (12:40 +0100)]
pinctrl: mediatek: add PUPD/R0/R1 support for MT7623

The pins for the MMC controller weren't being set up correctly because the
pinctrl driver only sets the GPIO pullup/pulldown config and doesn't
handle the special cases with PUPD/R0/R1 control.

Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Tested-by: Frank Wunderlich <frank-w@public-files.de>
4 years agoarm: dts: mediatek: enable sgmii mode and mt7531 switch for mt7629
MarkLee [Fri, 19 Jun 2020 11:17:17 +0000 (19:17 +0800)]
arm: dts: mediatek: enable sgmii mode and mt7531 switch for mt7629

This patch enable sgmii mode and mt7531 switch support in mt7629
ethernet dts node

Signed-off-by: MarkLee <Mark-MC.Lee@mediatek.com>
4 years agoeth: mtk-eth: enable mt7629 sgmii mode support in mediatek eth driver
MarkLee [Fri, 19 Jun 2020 11:17:16 +0000 (19:17 +0800)]
eth: mtk-eth: enable mt7629 sgmii mode support in mediatek eth driver

The sgmii mode init flow is almost the same for all mediatek SoC, the
only difference is the register offset(SGMSYS_GEN2_SPEED) is 0x2028
in the old chip(mt7622) but changed to 0x128 for the newer chip(mt7629
and the following chips).

Signed-off-by: MarkLee <Mark-MC.Lee@mediatek.com>
4 years agodm: remove superfluous comment for union ofnode_union
Heinrich Schuchardt [Fri, 24 Jul 2020 16:39:43 +0000 (18:39 +0200)]
dm: remove superfluous comment for union ofnode_union

"future live tree" does not make sense anymore as we have CONFIG_OF_LIVE.

Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
4 years agotest: dm: add a test for class button
Philippe Reynes [Fri, 24 Jul 2020 16:19:52 +0000 (18:19 +0200)]
test: dm: add a test for class button

Add a test to confirm that we can read button state
using the button-gpio driver.

Signed-off-by: Philippe Reynes <philippe.reynes@softathome.com>
4 years agotest/py: add tests for the button commands
Philippe Reynes [Fri, 24 Jul 2020 16:19:51 +0000 (18:19 +0200)]
test/py: add tests for the button commands

Adds tests for the button commands.

Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Philippe Reynes <philippe.reynes@softathome.com>
4 years agosandbox: enable button
Philippe Reynes [Fri, 24 Jul 2020 16:19:50 +0000 (18:19 +0200)]
sandbox: enable button

Enable the support of button (driver and command) on sandbox.

Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Philippe Reynes <philippe.reynes@softathome.com>
4 years agosandbox64: enable button
Philippe Reynes [Fri, 24 Jul 2020 16:19:49 +0000 (18:19 +0200)]
sandbox64: enable button

Enable the support of button (driver and command) on sandbox64.

Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Philippe Reynes <philippe.reynes@softathome.com>
4 years agosandbox: dtsi: add buttons
Philippe Reynes [Fri, 24 Jul 2020 16:19:48 +0000 (18:19 +0200)]
sandbox: dtsi: add buttons

Adds two buttons on sandbox so button framework may be tested.

Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Philippe Reynes <philippe.reynes@softathome.com>
4 years agocmd: button: add a new 'button' command
Philippe Reynes [Fri, 24 Jul 2020 16:19:47 +0000 (18:19 +0200)]
cmd: button: add a new 'button' command

Adds a command 'button' that provides the list of buttons
supported by the board, and the state of a button.

Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Philippe Reynes <philippe.reynes@softathome.com>
4 years agodm: button: add a driver for button driven by gpio
Philippe Reynes [Fri, 24 Jul 2020 16:19:46 +0000 (18:19 +0200)]
dm: button: add a driver for button driven by gpio

Add a simple driver which allows use of buttons attached to GPIOs.

Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Philippe Reynes <philippe.reynes@softathome.com>
4 years agodm: button: add an uclass for button
Philippe Reynes [Fri, 24 Jul 2020 16:19:45 +0000 (18:19 +0200)]
dm: button: add an uclass for button

Add a new uclass for button that implements two functions:
- button_get_by_label
- button_get_status

Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Philippe Reynes <philippe.reynes@softathome.com>
4 years agosandbox, test: change hog gpio
Philippe Reynes [Fri, 24 Jul 2020 13:51:53 +0000 (15:51 +0200)]
sandbox, test: change hog gpio

Since commit 9ba84329dc45 ("sandbox, test: add test for GPIO_HOG
function"), the gpio_a 0,1,2 and 3 are used by hog in test.dts.
But 2 leds 'sandbox:red' and 'sandbox:green' are using gpio_a 0
and 1. As hog always request his gpios, the led command on both
led is broken:

=> led sandbox:red
LED 'sandbox:red' not found (err=-16)

The gpio is already requested by hog, so it can't be enabled
for led 'sandbox:red'.

This commit change the gpio used by hog to 10, 11, 12 and 13,
so the led command could be used again with 'sandbox:red' and
'sandbox:green'.

Signed-off-by: Philippe Reynes <philippe.reynes@softathome.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Heiko Schocher <hs@denx.de>
4 years agocmd: host: return value of do_host_bind()
Heinrich Schuchardt [Fri, 24 Jul 2020 06:29:55 +0000 (08:29 +0200)]
cmd: host: return value of do_host_bind()

When trying to bind to a non-existent file the following output is
displayed:

=> host bind 0 non_existent
Failed to access host backing file 'non_existent'
exit not allowed from main input shell.

The last line is quite unexpected and due to an incorrect return value.

If do_host_bind() fails, return CMD_RET_FAILURE (= 1).

Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
4 years agodm: Fix build error when OF_CONTROL is not set
Dan Murphy [Thu, 23 Jul 2020 12:01:38 +0000 (07:01 -0500)]
dm: Fix build error when OF_CONTROL is not set

With OF_CONTROL disabled the build fails for
include/dm/read.h:932:10: error: ‘ENOTSUPP’ undeclared (first use in this function)
  932 |  return -ENOTSUPP;

Fixes: 45224e8f2691 ("dm: core: gracefully handle alias seq without of")
Signed-off-by: Dan Murphy <dmurphy@ti.com>