]> git.dujemihanovic.xyz Git - u-boot.git/log
u-boot.git
5 months agoefi_loader: Fix typo in EFI_RT_VOLATILE_STORE description
Michal Simek [Tue, 16 Jul 2024 13:56:51 +0000 (15:56 +0200)]
efi_loader: Fix typo in EFI_RT_VOLATILE_STORE description

Fix typo in EFI_RT_VOLATILE_STORE description.

Fixes: c28d32f946f0 ("efi_loader: conditionally enable SetvariableRT")
Signed-off-by: Michal Simek <michal.simek@amd.com>
5 months agodoc: move out-of-tree building info to HTML
Heinrich Schuchardt [Sun, 14 Jul 2024 07:18:04 +0000 (09:18 +0200)]
doc: move out-of-tree building info to HTML

Move the information about out-of-tree building
from README to the generated HTML documentation.

Signed-off-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
5 months agoefi_loader: find distro device-path for media devices
Heinrich Schuchardt [Sat, 13 Jul 2024 11:30:29 +0000 (13:30 +0200)]
efi_loader: find distro device-path for media devices

The auto-generated load options for media device do not contain a partition
node. We cannot expect the simple file protocol here.

Get the partition device-path via the loaded image protocol.

Fixes: e91b68fd6b83 ("efi_loader: load distro dtb in bootmgr")
Reported-by: E Shattow <lucent@gmail.com>
Signed-off-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
Tested-by: E Shattow <lucent@gmail.com>
5 months agodoc: add missing table of content links
Sam Povilus [Mon, 8 Jul 2024 16:48:12 +0000 (10:48 -0600)]
doc: add missing table of content links

add missing table of content links, make alphabetical

Signed-off-by: Sam Povilus <sam.povilus@amd.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
5 months agodoc: Remove FIT documentation that is elsewhere
Sam Povilus [Mon, 8 Jul 2024 16:48:11 +0000 (10:48 -0600)]
doc: Remove FIT documentation that is elsewhere

Before 9d0750064e (doc: Move external FIT docs into the main body), the
FIT property data-size was not a mandatory property and still it is not
expected to be set alongside the data property.

Move the data-size property to the "Conditionally mandatory property"
section, where it actually belongs.

Signed-off-by: Sam Povilus <sam.povilus@amd.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
5 months agoMerge patch series "bootstd: Add Android support"
Tom Rini [Thu, 18 Jul 2024 19:52:11 +0000 (13:52 -0600)]
Merge patch series "bootstd: Add Android support"

Mattijs Korpershoek <mkorpershoek@baylibre.com> says:

Android boot flow is a bit different than a regular Linux distro.
Android relies on multiple partitions in order to boot.

A typical boot flow would be:
1. Parse the Bootloader Control Block (BCB, misc partition)
2. If BCB requested bootonce-bootloader, start fastboot and wait.
3. If BCB requested recovery or normal android, run the following:
   a. Get slot (A/B) from BCB
   b. Run AVB (Android Verified Boot) on boot partitions
   c. Load boot and vendor_boot partitions
   d. Load device-tree, ramdisk and boot

The AOSP documentation has more details at [1], [2], [3]

This has been implemented via complex boot scripts such as [4].
However, these boot script are neither very maintainable nor generic.
Moreover, DISTRO_DEFAULTS is being deprecated [5].

Add a generic Android bootflow implementation for bootstd.

For this initial version, only boot image v4 is supported.

This has been tested on sandbox using:
$ ./test/py/test.py --bd sandbox --build -k test_ut

This has also been tested on the AM62X SK EVM using TI's Android SDK[6]
To test on TI board, the following (WIP) patch is needed as well:
https://gitlab.baylibre.com/baylibre/ti/ti-u-boot/-/commit/84cceb912bccd7cdd7f9dd69bca0e5d987a1fd04

[1] https://source.android.com/docs/core/architecture/bootloader
[2] https://source.android.com/docs/core/architecture/partitions
[3] https://source.android.com/docs/core/architecture/partitions/generic-boot
[4] https://source.denx.de/u-boot/u-boot/-/blob/master/include/configs/meson64_android.h
[5] https://lore.kernel.org/r/all/20230914165615.1058529-17-sjg@chromium.org/
[6] https://software-dl.ti.com/processor-sdk-android/esd/AM62X/09_02_00/docs/android/Overview.html

5 months agoMerge patch series "finish using .dtso for overlay source files"
Tom Rini [Thu, 18 Jul 2024 19:51:15 +0000 (13:51 -0600)]
Merge patch series "finish using .dtso for overlay source files"

Rasmus Villemoes <rasmus.villemoes@prevas.dk> says:

This is a followup to the patches that landed in 2024.01 and nearly
made sure that source files for producing .dtbo files use the .dtso
extension. In the same release, a few new .dts files snuck in, and
there was also some test code involving .dtbo -> .dtbo.S -> .dtbo.o I
didn't really know how to handle at the time. This should finish the
job, bring us in sync with linux (at least in this respect), and drop
the .dts -> .dtbo build rule.

5 months agobootstd: Add test for bootmeth_android
Mattijs Korpershoek [Wed, 10 Jul 2024 08:40:06 +0000 (10:40 +0200)]
bootstd: Add test for bootmeth_android

Add a unit test for testing the Android bootmethod.

This requires another mmc image (mmc7) to contain the following partitions:
- misc: contains the Bootloader Control Block (BCB)
- boot_a: contains a fake generic kernel image
- vendor_boot_a: contains a fake vendor_boot image

Also add BOOTMETH_ANDROID as a dependency on sandbox so that we can test
this with:

$ ./test/py/test.py --bd sandbox --build -k test_ut # build the mmc7.img
$ ./test/py/test.py --bd sandbox --build -k bootflow_android

Signed-off-by: Mattijs Korpershoek <mkorpershoek@baylibre.com>
Reviewed-by: Julien Masson <jmasson@baylibre.com>
Reviewed-by: Guillaume La Roque <glaroque@baylibre.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
5 months agobootstd: Add a bootmeth for Android
Mattijs Korpershoek [Wed, 10 Jul 2024 08:40:05 +0000 (10:40 +0200)]
bootstd: Add a bootmeth for Android

Android boot flow is a bit different than a regular Linux distro.
Android relies on multiple partitions in order to boot.

A typical boot flow would be:
1. Parse the Bootloader Control Block (BCB, misc partition)
2. If BCB requested bootonce-bootloader, start fastboot and wait.
3. If BCB requested recovery or normal android, run the following:
3.a. Get slot (A/B) from BCB
3.b. Run AVB (Android Verified Boot) on boot partitions
3.c. Load boot and vendor_boot partitions
3.d. Load device-tree, ramdisk and boot

The AOSP documentation has more details at [1], [2], [3]

This has been implemented via complex boot scripts such as [4].
However, these boot script are neither very maintainable nor generic.
Moreover, DISTRO_DEFAULTS is being deprecated [5].

Add a generic Android bootflow implementation for bootstd.
For this initial version, only boot image v4 is supported.

[1] https://source.android.com/docs/core/architecture/bootloader
[2] https://source.android.com/docs/core/architecture/partitions
[3] https://source.android.com/docs/core/architecture/partitions/generic-boot
[4] https://source.denx.de/u-boot/u-boot/-/blob/master/include/configs/meson64_android.h
[5] https://lore.kernel.org/r/all/20230914165615.1058529-17-sjg@chromium.org/

Reviewed-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Julien Masson <jmasson@baylibre.com>
Tested-by: Guillaume La Roque <glaroque@baylibre.com>
Signed-off-by: Mattijs Korpershoek <mkorpershoek@baylibre.com>
5 months agoandroid: boot: Add set_abootimg_addr() and set_avendor_bootimg_addr()
Mattijs Korpershoek [Wed, 10 Jul 2024 08:40:04 +0000 (10:40 +0200)]
android: boot: Add set_abootimg_addr() and set_avendor_bootimg_addr()

The only way to configure the load addresses for both bootimg and
vendor_bootimg is by using the "abootimg" command.
If we want to use the C API, there is no equivalent.

Add set_abootimg_addr() and set_avendor_bootimg_addr() so that we can
specify the load address from C.

This can be useful for implementing an Android bootmethod.

Reviewed-by: Igor Opaniuk <igor.opaniuk@gmail.com>
Reviewed-by: Julien Masson <jmasson@baylibre.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Guillaume La Roque <glaroque@baylibre.com>
Tested-by: Guillaume La Roque <glaroque@baylibre.com>
Signed-off-by: Mattijs Korpershoek <mkorpershoek@baylibre.com>
5 months agobootstd: Add bootflow_iter_check_mmc() helper
Mattijs Korpershoek [Wed, 10 Jul 2024 08:40:03 +0000 (10:40 +0200)]
bootstd: Add bootflow_iter_check_mmc() helper

Some bootflows might be able to only boot from MMC devices.

Add a helper function these bootflows can use.

Reviewed-by: Igor Opaniuk <igor.opaniuk@gmail.com>
Reviewed-by: Julien Masson <jmasson@baylibre.com>
Reviewed-by: Guillaume La Roque <glaroque@baylibre.com>
Tested-by: Guillaume La Roque <glaroque@baylibre.com>
Signed-off-by: Mattijs Korpershoek <mkorpershoek@baylibre.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
5 months agoboot: android: Provide vendor_bootimg_addr in boot_get_fdt()
Mattijs Korpershoek [Wed, 10 Jul 2024 08:40:02 +0000 (10:40 +0200)]
boot: android: Provide vendor_bootimg_addr in boot_get_fdt()

When calling android_image_get_dtb_by_index() using boot image v3+,
we should also pass the vendor_boot ramdisk address.

Use get_avendor_bootimg_addr() to do so.

Notes: on boot image v2, this is harmless since get_avendor_bootimg_addr()
       returns -1.
       for legacy implementations that don't have CMD_ABOOTIMG, add a weak
       implementation to avoid linking errors.

Reviewed-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Julien Masson <jmasson@baylibre.com>
Reviewed-by: Igor Opaniuk <igor.opaniuk@gmail.com>
Reviewed-by: Guillaume La Roque <glaroque@baylibre.com>
Tested-by: Guillaume La Roque <glaroque@baylibre.com>
Signed-off-by: Mattijs Korpershoek <mkorpershoek@baylibre.com>
5 months agokbuild: Disallow DTB overlays to built from .dts named source files
Rasmus Villemoes [Wed, 10 Jul 2024 07:16:12 +0000 (09:16 +0200)]
kbuild: Disallow DTB overlays to built from .dts named source files

[equivalent to linux commit 81d362732bac]

As a follow up to the series allowing DTB overlays to built from .dtso
files. Now that all overlays have been renamed, remove the ability to
build from overlays from .dts files to prevent any files with the old
name from accidental being added.

Signed-off-by: Rasmus Villemoes <rasmus.villemoes@prevas.dk>
5 months agotest/py: efi_capsule: rename device tree overlay source to .dtso
Rasmus Villemoes [Wed, 10 Jul 2024 07:16:11 +0000 (09:16 +0200)]
test/py: efi_capsule: rename device tree overlay source to .dtso

Distinguish more clearly between source files meant for producing .dtb
from those meant for producing .dtbo.

Note that in the linux tree, all device tree overlay sources have been
renamed to .dtso, and the .dts->.dtbo rule is gone since v6.5 (commit
81d362732bac). So this is also a step towards staying closer to linux
with respect to both Kbuild and device tree sources, and eventually
eliminating all .dts -> .dtbo instances.

This also matches the documentation update done in commit 4fb7e570d6b.

Cc: Masahisa Kojima <kojima.masahisa@socionext.com>
Signed-off-by: Rasmus Villemoes <rasmus.villemoes@prevas.dk>
5 months agotest: overlay: rename overlay source files to .dtso
Rasmus Villemoes [Wed, 10 Jul 2024 07:16:10 +0000 (09:16 +0200)]
test: overlay: rename overlay source files to .dtso

Distinguish more clearly between source files meant for producing .dtb
from those meant for producing .dtbo.

In this case, the files are really meant to be compiled to .dtbo ->
.dtbo.S -> .dtbo.o that get embedded in the image, which means that
the begin/end symbols generated by the makefile rule changes to
__dtbo_ rather than __dtb, so the consuming .c file needs updating,
but this should not result in any functional change.

Note that in the linux tree, all device tree overlay sources have been
renamed to .dtso, and the .dts->.dtbo rule is gone since v6.5 (commit
81d362732bac). So this is also a step towards staying closer to linux
with respect to both Kbuild and device tree sources.

Cc: Maxime Ripard <maxime.ripard@free-electrons.com>
Cc: Simon Glass <sjg@chromium.org>
Signed-off-by: Rasmus Villemoes <rasmus.villemoes@prevas.dk>
5 months agokbuild: Allow DTB overlays to built into .dtbo.S files
Rasmus Villemoes [Wed, 10 Jul 2024 07:16:09 +0000 (09:16 +0200)]
kbuild: Allow DTB overlays to built into .dtbo.S files

[linux commit 941214a512d8, modified for U-Boot by removing the
include of vmlinux.lds.h and replacing STRUCT_ALIGNMENT by 16.]

DTB files can be built into the kernel by converting them to assembly
files then assembling them into object files. We extend this here
for DTB overlays with the .dtso extensions.

We change the start and end delimiting tag prefix to make it clear that
this data came from overlay files.

Signed-off-by: Rasmus Villemoes <rasmus.villemoes@prevas.dk>
5 months agoarm64: dts: imx8mp: rename DHCOM SoM overlays to .dtso
Rasmus Villemoes [Wed, 10 Jul 2024 07:16:08 +0000 (09:16 +0200)]
arm64: dts: imx8mp: rename DHCOM SoM overlays to .dtso

Distinguish more clearly between source files meant for producing .dtb
from those meant for producing .dtbo. No functional change, as we
currently have rules for producing a foo.dtbo from either foo.dts or
foo.dtso.

Note that in the linux tree, all device tree overlay sources have been
renamed to .dtso, and the .dts->.dtbo rule is gone since v6.5 (commit
81d362732bac). So this is also a step towards staying closer to linux
with respect to both Kbuild and device tree sources.

Cc: Marek Vasut <marex@denx.de>
Signed-off-by: Rasmus Villemoes <rasmus.villemoes@prevas.dk>
Reviewed-by: Marek Vasut <marex@denx.de>
5 months agoARM: dts: remove leftover Amlogic GX/G12 bindings headers
Neil Armstrong [Thu, 18 Jul 2024 08:28:37 +0000 (10:28 +0200)]
ARM: dts: remove leftover Amlogic GX/G12 bindings headers

Remove the leftover Amlogic GX/G12 bindings headers that
maked the v6.10 upstream DT fail to build.

Signed-off-by: Neil Armstrong <neil.armstrong@linaro.org>
5 months agoMerge tag 'u-boot-rockchip-20240718' of https://source.denx.de/u-boot/custodians...
Tom Rini [Thu, 18 Jul 2024 13:49:14 +0000 (07:49 -0600)]
Merge tag 'u-boot-rockchip-20240718' of https://source.denx.de/u-boot/custodians/u-boot-rockchip

- Add boards:
  rk3328: Radxa ROCK Pi E v3;
  rk3588s: FriendlyElec NanoPi R6C/S;
- Remove board: Theobroma Systems RK3368 Lion;
- Add rk3588 pcie support;
- Misc updates for board and config;

CI: https://source.denx.de/u-boot/custodians/u-boot-rockchip/-/pipelines/2163

5 months agobloblist: fix bloblist convention check.
Levi Yun [Wed, 10 Jul 2024 13:53:20 +0000 (14:53 +0100)]
bloblist: fix bloblist convention check.

According to recently firmware handsoff spec [1]'s "Register usage at handoff
boundary", Transfer List's signature value was changed from 0x40_b10b
(3 bytes) to 4a0f_b10b (4 bytes).

As updating of TL's signature, register value of x1/r1 should be:

In aarch32's r1 value should be
    R1[23:0]: set to the TL signature (4a0f_b10b -> masked range value: 0f_b10b)
    R1[31:24]: version of the register convention ==  1

and

In aarch64's x1 value should be
    X1[31:0]: set to the TL signature (4a0f_b10b)
    X1[39:32]: version of the register convention ==  1
    X1[63:40]: MBZ
(See the [2] and [3]).

This patch fix problems:
   1. breaking X1 value with updated specification in aarch64
        - change of length of signature field.

   2. previous error value set in R1 in arm32.
        - length of signature should be 24, but it uses 32bit signature.

This patch is a breaking change. It works only TF-A is updated.

Link: https://github.com/FirmwareHandoff/firmware_handoff
Link: https://github.com/FirmwareHandoff/firmware_handoff/issues/32
Link: https://github.com/FirmwareHandoff/firmware_handoff/commit/5aa7aa1d3a1db75213e458d392b751f0707de027
Signed-off-by: Levi Yun <yeoreum.yun@arm.com>
Reviewed-by: Raymond Mao <raymond.mao@linaro.org>
Reviewed-by: Simon Glass <sjg@chromium.org>
5 months agotest/py/requirements.txt: Bump zipp to current release
Tom Rini [Tue, 9 Jul 2024 23:07:09 +0000 (17:07 -0600)]
test/py/requirements.txt: Bump zipp to current release

A security issue exists with zipp before v3.19.1, and the current
release is now v3.19.2. While the change in versions numbers is large, a
manual inspection of the changelog shows that it's not as big as might
be implied.

Reported-by: GitHub dependabot
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Tom Rini <trini@konsulko.com>
5 months agodriver: rng: Do not check ARM_SMCCC_TRNG_VERSION
Leo Yan [Tue, 9 Jul 2024 16:23:09 +0000 (17:23 +0100)]
driver: rng: Do not check ARM_SMCCC_TRNG_VERSION

As described in the document SMC Calling Convention (ARM DEN 0028 1.5 F),
section 7 "Arm Architecture Calls", the SMC call SMCCC_ARCH_FEATURES is
not expected to support the function ID ARM_SMCCC_TRNG_VERSION. Trusted
Firmware-A follows up the specification in its implementation.

This commit removes the invocation to avoid the failure - which is a
wrong calling in U-boot. The later code invokes ARM_SMCCC_TRNG_VERSION
for retrieving the TRNG version, except it can read back the version
number, it also can be used to detect whether the TRNG is supported or
not.

Signed-off-by: Leo Yan <leo.yan@arm.com>
Reviewed-by: Weizhao Ouyang <o451686892@gmail.com>
5 months agoti: omap: am3517evm: Move environment definition to env file
Adam Ford [Tue, 9 Jul 2024 02:38:57 +0000 (21:38 -0500)]
ti: omap: am3517evm: Move environment definition to env file

Instead of cluttering up a header file with a bunch of defines,
move the default environmental variables to a file called
am3517evm.env and reference it from the defconfig.  Also
remove dead comments.

Signed-off-by: Adam Ford <aford173@gmail.com>
5 months agotimer: mtk: skip setting parent with dummy fixed-clock
Christian Marangi [Sat, 6 Jul 2024 16:13:02 +0000 (18:13 +0200)]
timer: mtk: skip setting parent with dummy fixed-clock

Skip setting parent with dummy fixed-clock. Upstream linux might declare
an additional clock for the mtk timer and that additional clock might
also be a fixed-clock defined in DT. Setting parent of a dummy fixed-clock
resulta in error hence mtk timer fails to probe.

Skip setting parent to permit correct probe of the mtk timer.

Fixes: d3c3606c5cc6 ("timer: MediaTek: add timer driver for MediaTek SoCs")
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
5 months agoMerge patch series "Respin bootstd cros patch into a series of two"
Tom Rini [Tue, 16 Jul 2024 23:09:36 +0000 (17:09 -0600)]
Merge patch series "Respin bootstd cros patch into a series of two"

Vincent Stehlé <vincent.stehle@arm.com> says:

Hi,

This is a respin of this patch [1] after discussion [2]. Thanks to
Simon and Heinrich for their reviews.

To use the guidcmp() function, as suggested by Heinrich, we need to
make it available to bootmeth_cros.c and I think that the cleanest way
to do that is (arguably) to move the guid helper functions to efi.h
near the efi_guid_t definition; this is why the original patch has now
become a series of two patches.

The alternative would be to include efi_loader.h from bootmeth_cros.c
but I think this does not sound "right". If this is in fact the
preferred approach just let me know and I will respin.

There is no difference in the sandbox binaries before/after this
series on Arm and on PC, and all the tests I have run on the sandbox
are unchanged.

5 months agoverdin-am62: add DFU, USB and UUU fastboot support
Vitor Soares [Thu, 4 Jul 2024 15:49:32 +0000 (16:49 +0100)]
verdin-am62: add DFU, USB and UUU fastboot support

Enable USB host as well as USB gadget and DFU support for a53 and r5
configs. Also, enable UUU fastboot support to download files with
the UUU tool from a53.

Additionally, configure usb0 to peripheral mode and add extra
environment for DFU use.

Signed-off-by: Vitor Soares <vitor.soares@toradex.com>
Reviewed-by: Francesco Dolcini <francesco.dolcini@toradex.com>
5 months agoconfigs: rockchip: imply OF_LIBFDT_OVERLAY for rk3308 and rk3328
FUKAUMI Naoki [Wed, 17 Jul 2024 04:47:26 +0000 (13:47 +0900)]
configs: rockchip: imply OF_LIBFDT_OVERLAY for rk3308 and rk3328

for rk3308, all defconfigs have CONFIG_OF_LIBFDT_OVERLAY=y, so enable it
by default.

for rk3328, any defconfig doesn't have it. but there is no strong reason
not to enable it. at least it's required for ROCK Pi E.

Signed-off-by: FUKAUMI Naoki <naoki@radxa.com>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
5 months agoconfigs: rockchip: sync ENV_MEM_LAYOUT_SETTINGS for rk3308, rk3328, and rk3399
FUKAUMI Naoki [Wed, 17 Jul 2024 04:47:25 +0000 (13:47 +0900)]
configs: rockchip: sync ENV_MEM_LAYOUT_SETTINGS for rk3308, rk3328, and rk3399

- add support for compressed kernel for rk3308
- prepare support for fdtoverlay for rk3328

tested on ROCK Pi S 256MB, ROCK Pi E 2GB, and ROCK Pi 4A 4GB with
linux-next-20240613 defconfig kernel.

Signed-off-by: FUKAUMI Naoki <naoki@radxa.com>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
5 months agoconfigs: rockchip: reduce diff for rk3308, rk3328, rk3399, rk3568, and rk3588
FUKAUMI Naoki [Wed, 17 Jul 2024 04:47:24 +0000 (13:47 +0900)]
configs: rockchip: reduce diff for rk3308, rk3328, rk3399, rk3568, and rk3588

this is cosmetic change. no functional change is intended.

- remove redundant white spaces
- replace white spaces with tab
- align position of last letter/word
- sort lines in CFG_EXTRA_ENV_SETTINGS
- add comment after #endif

Signed-off-by: FUKAUMI Naoki <naoki@radxa.com>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
5 months agophy: rockchip: snps-pcie3: Fix clearing PHP_GRF_PCIESEL_CON bits
Sebastian Kropatsch [Sun, 14 Jul 2024 21:23:43 +0000 (23:23 +0200)]
phy: rockchip: snps-pcie3: Fix clearing PHP_GRF_PCIESEL_CON bits

The pcie1ln_sel bits for the RK3588 are getting set but not cleared due
to an incorrect write mask.
Use a newly introduced constant for the write mask to fix this.
Also introduce a GENMASK-based constant for PCIE30_PHY_MODE.

This fix is adapted from the upstream Linux commit by Sebastian Reichel:
55491a5fa163 ("phy: rockchip-snps-pcie3: fix clearing PHP_GRF_PCIESEL_CON bits")

Fixes: 50e54e80679b ("phy: rockchip: snps-pcie3: Add support for RK3588")
Signed-off-by: Sebastian Kropatsch <seb-dev@mail.de>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
5 months agophy: rockchip: snps-pcie3: Fix bifurcation for RK3588
Sebastian Kropatsch [Sun, 14 Jul 2024 21:23:42 +0000 (23:23 +0200)]
phy: rockchip: snps-pcie3: Fix bifurcation for RK3588

Misconfigured `PHP_GRF_PCIESEL` values are causing bifurcation issues,
for example on the FriendlyElec CM3588 NAS board which uses bifurcation
on both PCIe PCIe ports (all four lanes) to enable four M.2 NVMe
sockets. Without this fix, NVMe devices do not get recognized.

Correct the `PHP_GRF_PCIESEL` register configuration and simplify the
bifurcation logic, enabling proper PCIe bifurcation based on the
data-lanes property.

This fix is adapted from the upstream Linux commit by Michal Tomek:
f8020dfb311d ("phy: rockchip-snps-pcie3: fix bifurcation on rk3588")

Fixes: 50e54e80679b ("phy: rockchip: snps-pcie3: Add support for RK3588")
Signed-off-by: Sebastian Kropatsch <seb-dev@mail.de>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
5 months agophy: rockchip: snps-pcie3: Fix "rockchip" spelling
Sebastian Kropatsch [Sun, 14 Jul 2024 21:23:41 +0000 (23:23 +0200)]
phy: rockchip: snps-pcie3: Fix "rockchip" spelling

Several identifiers use "rochchip" instead of "rockchip".
Fix this by replacing every instance of "rochchip" with "rockchip".

Signed-off-by: Sebastian Kropatsch <seb-dev@mail.de>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
5 months agophy: rockchip: naneng-combphy: Fix "rockchip" spelling
Sebastian Kropatsch [Sun, 14 Jul 2024 21:23:40 +0000 (23:23 +0200)]
phy: rockchip: naneng-combphy: Fix "rockchip" spelling

Replace "rochchip" by "rockchip" in two instances.

Signed-off-by: Sebastian Kropatsch <seb-dev@mail.de>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
5 months agoboard: rockchip: Add FriendlyElec NanoPi R6S
Sebastian Kropatsch [Thu, 11 Jul 2024 10:15:18 +0000 (12:15 +0200)]
board: rockchip: Add FriendlyElec NanoPi R6S

The NanoPi R6S is a SBC by FriendlyElec based on the Rockchip RK3588s.
It comes with 4GB or 8GB of RAM, a microSD card slot, 32GB eMMC storage,
one RTL8211F 1GbE and two RTL8125 2.5GbE Ethernet ports, one USB 2.0
Type-A and one USB 3.0 Type-A port, a HDMI port, a 12-pin GPIO FPC
connector, a fan connector, IR receiver as well as some buttons and LEDs.

Add initial support for this board using the upstream devicetree sources.

Kernel commit:
f1b11f43b3e9 ("arm64: dts: rockchip: Add support for NanoPi R6S")

Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
Signed-off-by: Sebastian Kropatsch <seb-dev@mail.de>
5 months agoboard: rockchip: Add FriendlyElec NanoPi R6C
Sebastian Kropatsch [Thu, 11 Jul 2024 10:15:17 +0000 (12:15 +0200)]
board: rockchip: Add FriendlyElec NanoPi R6C

The NanoPi R6C is a SBC by FriendlyElec based on the Rockchip RK3588s.
It comes with 4GB or 8GB of RAM, a microSD card slot, optional 32GB eMMC
storage, one M.2 M-Key connector, one RTL8211F 1GbE and one RTL8125
2.5GbE Ethernet port, one USB 2.0 Type-A and one USB 3.0 Type-A port, a
HDMI port, a 30-pin GPIO header as well as multiple buttons and LEDs.

Add initial support for this board using the upstream devicetree sources.

Tested in U-Boot proper:
- Booting from eMMC works
- 1GbE Ethernet works using the eth_eqos driver (tested by ping)
- 2.5GbE Ethernet works using the eth_rtl8169 driver (tested by ping),
  but the status LEDs on this specific port currently aren't working
- NVMe SSD in M.2 socket does get recognized (tested with `nvme scan`
  followed by `nvme details`)

Kernel commit:
d5f1d7437451 ("arm64: dts: rockchip: Add support for NanoPi R6C")

Reviewed-by: Quentin Schulz <quentin.schulz@cherry.de>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
Signed-off-by: Sebastian Kropatsch <seb-dev@mail.de>
5 months agorockchip: tiger-rk3588: add PCIe support
Quentin Schulz [Thu, 4 Jul 2024 12:53:33 +0000 (14:53 +0200)]
rockchip: tiger-rk3588: add PCIe support

This enables PCIe support on Tiger as exposed on
Q7_PCIE[0123]_[RT]X_[PN] signals and more specifically on the `PCI
Express` connector on the Haikou devkit.

This was tested with a PCIe to NVMe adapter (e.g.
https://www.amazon.de/dp/B07RZZ3TJG).

Signed-off-by: Quentin Schulz <quentin.schulz@cherry.de>
Reviewed-by: Heiko Stuebner <heiko@sntech.de>
Tested-by: Heiko Stuebner <heiko@sntech.de>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
5 months agorockchip: jaguar-rk3588: add PCIe M.2 M-KEY NVMe support
Quentin Schulz [Thu, 4 Jul 2024 12:53:32 +0000 (14:53 +0200)]
rockchip: jaguar-rk3588: add PCIe M.2 M-KEY NVMe support

Jaguar has an M.2 M-KEY slot for NVMes, connected to the PCIe3 4-lane
PHY on RK3588.

CONFIG_PHY_ROCKCHIP_NANENG_COMBOPHY=y is technically not necessary since
it's required only for the M.2 E-KEY slot on the main PCB, but that is
used typically for WiFi+BT modules, or on the mezzanine connector but
the features exposed behind that connector aren't supported in U-Boot
(no DT for it right now). However, if the PHY driver is missing, we get
the following error message:
pcie_dw_rockchip pcie@fe170000: failed to get pcie phy (ret=-19)

and you would need to know which PCIe controller that is before deciding
to ignore it. While after enabling the PHY driver, we are greeted with:
pcie_dw_rockchip pcie@fe170000: PCIe-2 Link Fail
which is a bit more acceptable (to me).

The other option would be to disable the PCIe2 PHYs/controllers in the
DT, which I'm not too fond of.

Signed-off-by: Quentin Schulz <quentin.schulz@cherry.de>
Reviewed-by: Heiko Stuebner <heiko@sntech.de>
Tested-by: Heiko Stuebner <heiko@sntech.de>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
5 months agoarm64: dts: rockchip: add PCIe3 support on rk3588-jaguar
Heiko Stuebner [Thu, 4 Jul 2024 12:53:31 +0000 (14:53 +0200)]
arm64: dts: rockchip: add PCIe3 support on rk3588-jaguar

The Jaguar SBC provides an M.2 slot connected to the pcie3 controller.
In contrast to a number of other boards the pcie-refclk is gpio-controlled,
so the necessary clock and is added to the list of pcie3 clocks.

Signed-off-by: Heiko Stuebner <heiko.stuebner@cherry.de>
Reviewed-by: Quentin Schulz <quentin.schulz@theobroma-systems.com>
Link: https://lore.kernel.org/r/20240423074956.2622318-1-heiko@sntech.de
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
[ upstream commit: 0ec7e1096332bc2b9bc881c21cfd234058f747b3 ]

(cherry picked from commit 76a89655ae740dddb57187b5b52071ed99187452)
Tested-by: Heiko Stuebner <heiko@sntech.de>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
5 months agorockchip: add support for Radxa ROCK Pi E v3.0
FUKAUMI Naoki [Mon, 24 Jun 2024 20:45:29 +0000 (05:45 +0900)]
rockchip: add support for Radxa ROCK Pi E v3.0

ROCK Pi E v3.0 uses DDR4 SDRAM instead of DDR3 SDRAM used in v1.2x.

prepare new rk3328-rock-pi-e-v3.dts in u-boot which just includes
upstream rk3328-rock-pi-e.dts.

defconfig still uses
 CONFIG_DEFAULT_FDT_FILE="rockchip/rk3328-rock-pi-e.dtb"

because v3.0 and prior are compatible.

Suggested-by: Jonas Karlman <jonas@kwiboo.se>
Signed-off-by: FUKAUMI Naoki <naoki@radxa.com>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
5 months agophy: rockchip: inno-hdmi: Fix missing readl base addr
Jagan Teki [Sat, 22 Jun 2024 18:21:49 +0000 (23:51 +0530)]
phy: rockchip: inno-hdmi: Fix missing readl base addr

inno_poll passes the reg offset that is used by readl_poll_sleep_timeout
without any base addr.

Fix it.

Bug:
inno_hdmi_phy phy@ff430000: Pre-PLL locking failed
inno_hdmi_phy phy@ff430000: PHY: Failed to power on phy@ff430000: -110.
failed to power on phy (ret=-110)
inno_hdmi_phy phy@ff430000: Pre-PLL locking failed
inno_hdmi_phy phy@ff430000: PHY: Failed to power on phy@ff430000: -110.
failed to power on phy (ret=-110)

Fixes: aa2271184603 ("phy: rockchip: Add Rockchip INNO HDMI PHY driver")
Suggested-by: Jonas Karlman <jonas@kwiboo.se>
Signed-off-by: Jagan Teki <jagan@edgeble.ai>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
5 months agorockchip: remove support for Theobroma Systems RK3368 Lion
Quentin Schulz [Thu, 20 Jun 2024 10:24:51 +0000 (12:24 +0200)]
rockchip: remove support for Theobroma Systems RK3368 Lion

No meaningful changes were made to this SoM since February 2021. Nobody
from Theobroma has booted anything recent on that product since July
2021 at the latest. The product isn't available to buy anymore and
disappeared from our website.

This product is therefore unmaintained and it would be disingenuous to
say the opposite, so drop support for RK3368 Lion.

If you're a user of Lion, feel free to revert this patch or contact our
sales/support department.

Signed-off-by: Quentin Schulz <quentin.schulz@cherry.de>
Acked-by: Heiko Stuebner <heiko@sntech.de>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
5 months agorockchip: theobroma-systems: migrate git URLs to HTTPS
Quentin Schulz [Thu, 20 Jun 2024 10:24:50 +0000 (12:24 +0200)]
rockchip: theobroma-systems: migrate git URLs to HTTPS

It turns out only Puma had a working git:// URL. Though that is now
fixed, having HTTPS URLs make it easier to directly reach our cgit
webserver to check what's up without having to change the URL manually.

Depending on your terminal settings, this also makes it possible to
open the link from it.

Signed-off-by: Quentin Schulz <quentin.schulz@cherry.de>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
5 months agorockchip: puma-rk3399: increase Ethernet PHY timeout to 30s
Quentin Schulz [Wed, 19 Jun 2024 15:44:56 +0000 (17:44 +0200)]
rockchip: puma-rk3399: increase Ethernet PHY timeout to 30s

On Puma v2.1 with a KSZ9031, the Ethernet PHY often doesn't detect the
link in the 4-second default timeout.

After some boot-loop test for about a thousand boots, 70% were below
10s, 90% below 15s and 100% below 30s. Let's play it safe and make it
30s so that hopefully all links that should be detected are.

Signed-off-by: Quentin Schulz <quentin.schulz@cherry.de>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
5 months agorockchip: include cru_rk3588.h and define rockchip_cru for RK3588
FUKAUMI Naoki [Tue, 18 Jun 2024 19:30:44 +0000 (04:30 +0900)]
rockchip: include cru_rk3588.h and define rockchip_cru for RK3588

fix following error found by clang:

  CC      arch/arm/mach-rockchip/cpu-info.o
arch/arm/mach-rockchip/cpu-info.c:23:13: error: incomplete definition of type 'struct rockchip_cru'
   23 |         switch (cru->glb_rst_st) {
      |                 ~~~^
./arch/arm/include/asm/arch-rockchip/clock.h:181:8: note: forward declaration of 'struct rockchip_cru'
  181 | struct rockchip_cru;
      |        ^
1 error generated.

Signed-off-by: FUKAUMI Naoki <naoki@radxa.com>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
5 months agopower: rk8xx: fix display name for RK808
Quentin Schulz [Mon, 17 Jun 2024 16:48:51 +0000 (18:48 +0200)]
power: rk8xx: fix display name for RK808

Commit 2ce40542e0eb ("power: rk8xx: properly print all supported PMICs
name") fixed all PMICs name that were broken but broke the only one that
was not broken already: RK808. This one is a special case because the ID
registers are marked as reserved and always return 0, so the variant
cannot be derived the same way it is done for other PMICs from Rockchip.

Fixes: 2ce40542e0eb ("power: rk8xx: properly print all supported PMICs name")
Signed-off-by: Quentin Schulz <quentin.schulz@cherry.de>
Reviewed-by: Dragan Simic <dsimic@manjaro.org>
5 months agorockchip: rk3399: fix SPI-NOR flash not found in U-Boot pre-reloc
Quentin Schulz [Mon, 17 Jun 2024 13:10:08 +0000 (15:10 +0200)]
rockchip: rk3399: fix SPI-NOR flash not found in U-Boot pre-reloc

In commit 100f489f58a6 ("rockchip: rk3399: Fix loading FIT from SD-card
when booting from eMMC"), the spi1 bootph properties were mistakenly
removed meaning, so re-add them back to fix SPI-NOR flash not being
found in U-Boot pre-reloc as required for RK3399 Puma.

Fixes: 100f489f58a6 ("rockchip: rk3399: Fix loading FIT from SD-card when booting from eMMC")
Signed-off-by: Quentin Schulz <quentin.schulz@cherry.de>
Reviewed-by: Jonas Karlman <jonas@kwiboo.se>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
5 months agorockchip: ringneck-px30: enable IO domain
Quentin Schulz [Fri, 14 Jun 2024 16:54:14 +0000 (18:54 +0200)]
rockchip: ringneck-px30: enable IO domain

Enable the IO domain on Ringneck. Based on the current HW design, this
should do nothing else than making sure vccio6 iodomain is controlled by
the GRF and not GPIO0B6.

Signed-off-by: Quentin Schulz <quentin.schulz@cherry.de>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
5 months agorockchip: io-domain: add support for PX30
Quentin Schulz [Fri, 14 Jun 2024 16:54:13 +0000 (18:54 +0200)]
rockchip: io-domain: add support for PX30

Port the PX30 part of the Rockchip IO Domain driver from Linux.

This differs from linux version in that the io iodomain bit is enabled
in the write ops instead of in an init ops as in linux, this way we can
avoid keeping a full state of all supplies that have been configured.

Signed-off-by: Quentin Schulz <quentin.schulz@cherry.de>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
5 months agorockchip: px30: bind sub-nodes for GRF (syscon)
Quentin Schulz [Fri, 14 Jun 2024 16:54:12 +0000 (18:54 +0200)]
rockchip: px30: bind sub-nodes for GRF (syscon)

There are some sub-nodes under the grf DT nodes, so add bind callback
function in syscon_px30 driver to scan them recursively.

Fixes: e9ccb2f526ed ("rockchip: add px30 architecture core")
Signed-off-by: Quentin Schulz <quentin.schulz@cherry.de>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
5 months agorockchip: puma-rk3399: add button support
Quentin Schulz [Fri, 14 Jun 2024 11:04:56 +0000 (13:04 +0200)]
rockchip: puma-rk3399: add button support

The Haikou Devkit exposes 4 buttons over GPIO so let's enable their
support so their status can be queried from the `button` command from
the CLI.

Signed-off-by: Quentin Schulz <quentin.schulz@cherry.de>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
5 months agorockchip: px30: imply ARMV8_CRYPTO
Quentin Schulz [Fri, 14 Jun 2024 11:04:55 +0000 (13:04 +0200)]
rockchip: px30: imply ARMV8_CRYPTO

PX30 supports ARMv8 Cryptography Extensions so let's enable it by
default for all PX30 to make FIT verification when enabled much faster.

While A35 shouldn't be impacted by ARMV8_SET_SMPEN cache coherency
according to the Kconfig help text, let's enable it just in case since
it exists in the documentation[1].

For u-boot part of the FIT image, it is now taking 5ms against currently
35ms. fdt-1 check lowered from 3ms to <1ms. atf-1 from 6ms to <1ms.

[1] https://developer.arm.com/documentation/100236/0100/register-descriptions/aarch64-system-registers/cpu-extended-control-register--el1?lang=en

Signed-off-by: Quentin Schulz <quentin.schulz@cherry.de>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
5 months agorockchip: ringneck-px30: enable FIT verification in SPL
Quentin Schulz [Fri, 14 Jun 2024 11:04:54 +0000 (13:04 +0200)]
rockchip: ringneck-px30: enable FIT verification in SPL

This enables FIT verification in SPL for its payload (bl31, u-boot.itb,
...). This makes PX30 Ringneck match what happens on other Theobroma
boards.

Signed-off-by: Quentin Schulz <quentin.schulz@cherry.de>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
5 months agorockchip: ringneck-px30: disable broken USB gadget
Quentin Schulz [Fri, 14 Jun 2024 11:04:53 +0000 (13:04 +0200)]
rockchip: ringneck-px30: disable broken USB gadget

USB gadget simply doesn't work right now on PX30. Tested on PX30 EVB and
PX30 Ringneck with Linux mainline and Rockchip 5.10 Linux as well as
U-Boot.

We don't want to our users to assume that USB gadget is supported on Q7
USB P1 on Ringneck Haikou, so let's remove its support, which also
removes the ums CLI command, fastboot and Android image booting
support.

Signed-off-by: Quentin Schulz <quentin.schulz@cherry.de>
5 months agobootstd: cros: store partition type in an efi_guid_t
Vincent Stehlé [Wed, 3 Jul 2024 11:37:56 +0000 (13:37 +0200)]
bootstd: cros: store partition type in an efi_guid_t

The scan_part() function uses a struct uuid to store the little-endian
partition type GUID, but this structure should be used only to contain a
big-endian UUID. Use an efi_guid_t instead and use guidcmp() for the
comparison.

Suggested-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Signed-off-by: Vincent Stehlé <vincent.stehle@arm.com>
Cc: Simon Glass <sjg@chromium.org>
Cc: Tom Rini <trini@konsulko.com>
5 months agoefi: move guid helper functions to efi.h
Vincent Stehlé [Wed, 3 Jul 2024 11:37:55 +0000 (13:37 +0200)]
efi: move guid helper functions to efi.h

Move the guidcmp() and guidcpy() functions to efi.h, near the definition of
the efi_guid_t type those functions deal with.

Signed-off-by: Vincent Stehlé <vincent.stehle@arm.com>
Cc: Heinrich Schuchardt <xypron.glpk@gmx.de>
Cc: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Cc: Tom Rini <trini@konsulko.com>
5 months agoMerge patch series "Fix fdtfile for j722s and am62p"
Tom Rini [Tue, 16 Jul 2024 19:01:52 +0000 (13:01 -0600)]
Merge patch series "Fix fdtfile for j722s and am62p"

Manorit Chawdhry <m-chawdhry@ti.com> says:

fdtfile wasn't being populated in these boards in legacy boot using
bootcmd_ti_mmc, migrate these platforms to ti_set_fdt_env.

5 months agoinclude: env: ti_common: Remove findfdt from bootcmd_ti_mmc
Dhruva Gole [Tue, 2 Jul 2024 05:20:04 +0000 (10:50 +0530)]
include: env: ti_common: Remove findfdt from bootcmd_ti_mmc

findfdt is used by bootcmd_ti_mmc by default which used to help populate
the fdtfile. The users of findfdt from bootcmd_ti_mmc have migrated to
ti_set_fdt_env for populating fdtfile; Hence, findfdt can be removed from
bootcmd_ti_mmc having no-impact for any platform.

Remove findfdt to not print out the warning that gets set after calling
ti_set_fdt_env. viz.
"echo WARN: fdtfile already set. Stop using findfdt in script"

Signed-off-by: Dhruva Gole <d-gole@ti.com>
Signed-off-by: Manorit Chawdhry <m-chawdhry@ti.com>
5 months agoinclude: env: ti: mmc: Change name_fdt usage to fdtfile
Manorit Chawdhry [Tue, 2 Jul 2024 05:20:03 +0000 (10:50 +0530)]
include: env: ti: mmc: Change name_fdt usage to fdtfile

name_fdt is kept for backward compatibility but it depends on EEPROM
detection logic and some of the TI K3 platforms don't have that anymore
which causes boot failure in legacy boot flow using bootcmd_ti_mmc.

K2g platforms which uses the same file have their own override causing
this change to be no-impact for them.

Replacing name_fdt usage to fdtfile as fdtfile is populated based on
CONFIG_DEFAULT_DEVICE_TREE after using ti_set_fdt_env.

Signed-off-by: Manorit Chawdhry <m-chawdhry@ti.com>
5 months agoconfigs: am62p|j722s_a53: Add CONFIG_BOARD_LATE_INIT
Manorit Chawdhry [Tue, 2 Jul 2024 05:20:02 +0000 (10:50 +0530)]
configs: am62p|j722s_a53: Add CONFIG_BOARD_LATE_INIT

This is called to set fdtfile based on evm.c code calling ti_set_fdt_env.

Signed-off-by: Manorit Chawdhry <m-chawdhry@ti.com>
5 months agoboard: ti: am62p|j722s: Add ti_set_fdt_env for fdtfile
Manorit Chawdhry [Tue, 2 Jul 2024 05:20:01 +0000 (10:50 +0530)]
board: ti: am62p|j722s: Add ti_set_fdt_env for fdtfile

stdboot and legacy boot depend on fdtfile. Since findfdt is getting
deprecated, move the rest of k3 platforms dependent on findfdt to
ti_set_fdt_env.

Populate fdtfile by calling ti_set_fdt_env in board files.

Signed-off-by: Manorit Chawdhry <m-chawdhry@ti.com>
5 months agoMerge patch series "configs: phycore_am62x_a53: Add more commands"
Tom Rini [Tue, 16 Jul 2024 18:56:46 +0000 (12:56 -0600)]
Merge patch series "configs: phycore_am62x_a53: Add more commands"

5 months agoboard: phytec: phycore_am64x: Move earlycon into own variable
Daniel Schultz [Mon, 1 Jul 2024 06:20:50 +0000 (23:20 -0700)]
board: phytec: phycore_am64x: Move earlycon into own variable

By moving the earlycon definition into a dedicated variable, it's
easier to change these values in case the kernel should print on
a different serial interface.

Signed-off-by: Daniel Schultz <d.schultz@phytec.de>
5 months agoboard: phytec: phycore_am62x: Move earlycon into own variable
Daniel Schultz [Mon, 1 Jul 2024 06:20:49 +0000 (23:20 -0700)]
board: phytec: phycore_am62x: Move earlycon into own variable

By moving the earlycon definition into a dedicated variable, it's
easier to change these values in case the kernel should print on
a different serial interface.

Signed-off-by: Daniel Schultz <d.schultz@phytec.de>
5 months agoconfigs: phycore_am64x_a53: Add more commands
Daniel Schultz [Mon, 1 Jul 2024 06:20:48 +0000 (23:20 -0700)]
configs: phycore_am64x_a53: Add more commands

Add the rtc command for testing this device. Additionally, add smc
and cache commands to boot non-Linux firmwares on the A53.

Signed-off-by: Daniel Schultz <d.schultz@phytec.de>
5 months agoconfigs: phycore_am62x_a53: Add more commands
Daniel Schultz [Mon, 1 Jul 2024 06:20:47 +0000 (23:20 -0700)]
configs: phycore_am62x_a53: Add more commands

Add i2c and rtc commands with all dependencies which are
required for testing.

Additionally, add smc and cache commands to boot Zephyr on the A53.

Signed-off-by: Daniel Schultz <d.schultz@phytec.de>
Reviewed-by: Wadim Egorov <w.egorov@phytec.de>
5 months agoMerge patch series "drivers: bootcount: Add support for ANY filesystem"
Tom Rini [Tue, 16 Jul 2024 18:46:39 +0000 (12:46 -0600)]
Merge patch series "drivers: bootcount: Add support for ANY filesystem"

Vasileios Amoiridis <vassilisamir@gmail.com> says:

This patch adds support to save the bootcount variable in a file located in
FAT filesystem. Up to now, there was support only for EXT filesystem.

5 months agodrivers: bootcount: Fix typo in documentation
Vasileios Amoiridis [Fri, 28 Jun 2024 17:35:43 +0000 (19:35 +0200)]
drivers: bootcount: Fix typo in documentation

The bootcount documentation was using "unattended" while it probably
intending to say "unintended"

Signed-off-by: Vasileios Amoiridis <vasileios.amoiridis@cern.ch>
Reviewed-by: Quentin Schulz <quentin.schulz@cherry.de>
5 months agodoc: api: bootcount: Convert to rST documentation
Vasileios Amoiridis [Fri, 28 Jun 2024 17:35:42 +0000 (19:35 +0200)]
doc: api: bootcount: Convert to rST documentation

Move to the new documentation style with rST formatting.

Signed-off-by: Vasileios Amoiridis <vasileios.amoiridis@cern.ch>
5 months agodrivers: bootcount: Add support for ANY filesystem
Vasileios Amoiridis [Fri, 28 Jun 2024 17:35:41 +0000 (19:35 +0200)]
drivers: bootcount: Add support for ANY filesystem

Add support to save boot count variable in ANY filesystem. Tested with
FAT and EXT.

Reviewed-by: Tom Rini <trini@konsulko.com>
Reviewed-by: Heiko Schocher <hs@denx.de>
Reviewed-by: Philip Oberfichtner <pro@denx.de>
Signed-off-by: Vasileios Amoiridis <vasileios.amoiridis@cern.ch>
5 months agofirmware: psci: register PSCI power domains to stub driver
Neil Armstrong [Fri, 28 Jun 2024 09:54:47 +0000 (11:54 +0200)]
firmware: psci: register PSCI power domains to stub driver

On some SoCs, like Qualcomm SoCs, the PSCI cluster power domain
is used by system-wide firmware interfaces to make sure none
of the CPUs are suspended before submitting requests.

While on U-boot we only use the first core and we never
suspend it, the Device Tree still references it and blocks
those nodes to be probed.

Simply bind the PSCI power-domain subnoded to a stub power
domain driver in order to solve the runtime dependencies.

Signed-off-by: Neil Armstrong <neil.armstrong@linaro.org>
5 months agoMerge branch 'master' of https://source.denx.de/u-boot/custodians/u-boot-sunxi
Tom Rini [Tue, 16 Jul 2024 16:20:14 +0000 (10:20 -0600)]
Merge branch 'master' of https://source.denx.de/u-boot/custodians/u-boot-sunxi

Nothing really major here, some rework of the SPL PMIC drivers, adding
support for the AXP717 on the way, which is showing up on newer boards
now, most prominently some Anbernic handheld gaming devices.  The rest
is enabling Ethernet and SPI boot on the Allwinner V3s SoC, plus two
fixes.  This also updates the "traditional U-Boot" DTs to that of kernel
v6.9.  I will look into upgrading some SoCs to dts/upstream in the
coming cycle, though this will not cover all SoCs for now, as we carry
some non-mainline fix to improve compatibility with older kernels. Will
see how it goes, but for now we stick with the "old way".

The branch survived the gitlab CI run, and Linux boot testing on some
selected boards.

5 months agospi: sunxi: drop max_hz handling
Michael Walle [Fri, 12 Jul 2024 17:14:56 +0000 (19:14 +0200)]
spi: sunxi: drop max_hz handling

The driver is trying to read the "spi-max-frequency" property of the
*controller* driver node. There is no such property. The
"spi-max-frequency" property belongs to the SPI devices on the bus.

Right now, the driver will always fall back to the default value of 1MHz
and thus flash reads are very slow with just about 215kb/s.

In fact, the SPI uclass will already take care of everything and we just
have to clamp the frequency to the values the driver/hardware supports.
Thus, drop the whole max_hz handling.

Signed-off-by: Michael Walle <mwalle@kernel.org>
Reviewed-by: Andre Przywara <andre.przywara@arm.com>
Tested-by: Andre Przywara <andre.przywara@arm.com>
5 months agosunxi: spl: h616: fix booting from high MMC offset
Andre Przywara [Thu, 9 May 2024 23:13:16 +0000 (00:13 +0100)]
sunxi: spl: h616: fix booting from high MMC offset

The BootROM in the Allwinner H616 tries to load the initial boot code
from sector 16 (8KB) of an SD card or eMMC device, but also looks at
sector 512 (256KB). This helps with GPT formatted cards.
A "high" boot offset is also used on previous SoCs, but it's sector 256
(128KB) there instead.

Extend the existing offset calculation code to consider the different
sector offset when running on an H616 SoC. This allows to load U-Boot
on any H616 device when the SPL is not located at 8KB.

Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Reviewed-by: Chen-Yu Tsai <wens@csie.org>
Tested-by: Ryan Walklin <ryan@testtoast.com>
5 months agosunxi: SPL SPI: add support for the V3s SoC
Michael Walle [Mon, 13 May 2024 23:43:19 +0000 (01:43 +0200)]
sunxi: SPL SPI: add support for the V3s SoC

The V3s is identical regarding register layout, clocks and resets to
the sun6i variants. Therefore, we can just add the MACH_SUN8I_V3S to
the sun6i compatible ones.

SPI boot was tested on a custom board with a Gigadevice GD25Q64 8MiB
SPI flash.

Signed-off-by: Michael Walle <mwalle@kernel.org>
Reviewed-by: Andre Przywara <andre.przywara@arm.com>
5 months agonet: sun8i_emac: add support for the V3s
Michael Walle [Mon, 13 May 2024 20:56:09 +0000 (22:56 +0200)]
net: sun8i_emac: add support for the V3s

Add the compatible string for the emac found on the V3s SoC. The SoC
only supports the internal PHY. There are no (R)MII signals on any pins.

Signed-off-by: Michael Walle <mwalle@kernel.org>
Reviewed-by: Andre Przywara <andre.przywara@arm.com>
5 months agoclk: sunxi: add EMAC and EPHY clocks and resets for the V3s SoC
Michael Walle [Mon, 13 May 2024 20:56:08 +0000 (22:56 +0200)]
clk: sunxi: add EMAC and EPHY clocks and resets for the V3s SoC

Add the clock gate registers as well as the reset register bits for the
EMAC and EPHY for the V3s. These are needed by the sun8i network driver.

Signed-off-by: Michael Walle <mwalle@kernel.org>
Reviewed-by: Andre Przywara <andre.przywara@arm.com>
5 months agosunxi: dts: arm/arm64: update devicetree files from Linux-v6.9
Andre Przywara [Fri, 19 Apr 2024 16:59:52 +0000 (17:59 +0100)]
sunxi: dts: arm/arm64: update devicetree files from Linux-v6.9

Sync the devicetree files from the official Linux kernel tree, v6.9.
This is covering Allwinner SoCs with 32-bit and 64-bit ARM cores.

Besides some cosmectic changes, this adds S/PDIF, DMA, and thermal
support to the H616 .dtsi. Nothing that U-Boot needs for itself, but
helpful to pass on to kernels.
This also adds new devicetrees, for some LonganPi H616 board, and the
Remix Mini PC. The latter will get a defconfig next.

As before, this omits the non-backwards compatible changes to the R_INTC
controller, to remain compatible with older kernels.

Signed-off-by: Andre Przywara <andre.przywara@arm.com>
5 months agoMerge patch series "api: Remove duplicate newlines"
Tom Rini [Mon, 15 Jul 2024 18:28:22 +0000 (12:28 -0600)]
Merge patch series "api: Remove duplicate newlines"

This removes a number of duplicate newlines throughout the codebase.

5 months agoxtensa: Remove duplicate newlines
Marek Vasut [Sat, 13 Jul 2024 13:19:36 +0000 (15:19 +0200)]
xtensa: Remove duplicate newlines

Drop all duplicate newlines. No functional change.

Signed-off-by: Marek Vasut <marek.vasut+renesas@mailbox.org>
5 months agox86: Remove duplicate newlines
Marek Vasut [Sat, 13 Jul 2024 13:19:35 +0000 (15:19 +0200)]
x86: Remove duplicate newlines

Drop all duplicate newlines. No functional change.

Signed-off-by: Marek Vasut <marek.vasut+renesas@mailbox.org>
5 months agotools: Remove duplicate newlines
Marek Vasut [Sat, 13 Jul 2024 13:19:34 +0000 (15:19 +0200)]
tools: Remove duplicate newlines

Drop all duplicate newlines. No functional change.

Signed-off-by: Marek Vasut <marek.vasut+renesas@mailbox.org>
5 months agotest: Remove duplicate newlines
Marek Vasut [Sat, 13 Jul 2024 13:19:33 +0000 (15:19 +0200)]
test: Remove duplicate newlines

Drop all duplicate newlines. No functional change.

Signed-off-by: Marek Vasut <marek.vasut+renesas@mailbox.org>
5 months agosh: Remove duplicate newlines
Marek Vasut [Sat, 13 Jul 2024 13:19:32 +0000 (15:19 +0200)]
sh: Remove duplicate newlines

Drop all duplicate newlines. No functional change.

Signed-off-by: Marek Vasut <marek.vasut+renesas@mailbox.org>
5 months agoscripts: Remove duplicate newlines
Marek Vasut [Sat, 13 Jul 2024 13:19:31 +0000 (15:19 +0200)]
scripts: Remove duplicate newlines

Drop all duplicate newlines. No functional change.

Signed-off-by: Marek Vasut <marek.vasut+renesas@mailbox.org>
5 months agosandbox: Remove duplicate newlines
Marek Vasut [Sat, 13 Jul 2024 13:19:30 +0000 (15:19 +0200)]
sandbox: Remove duplicate newlines

Drop all duplicate newlines. No functional change.

Signed-off-by: Marek Vasut <marek.vasut+renesas@mailbox.org>
5 months agoppc: Remove duplicate newlines
Marek Vasut [Sat, 13 Jul 2024 13:19:29 +0000 (15:19 +0200)]
ppc: Remove duplicate newlines

Drop all duplicate newlines. No functional change.

Signed-off-by: Marek Vasut <marek.vasut+renesas@mailbox.org>
5 months agopost: Remove duplicate newlines
Marek Vasut [Sat, 13 Jul 2024 13:19:28 +0000 (15:19 +0200)]
post: Remove duplicate newlines

Drop all duplicate newlines. No functional change.

Signed-off-by: Marek Vasut <marek.vasut+renesas@mailbox.org>
5 months agonios2: Remove duplicate newlines
Marek Vasut [Sat, 13 Jul 2024 13:19:27 +0000 (15:19 +0200)]
nios2: Remove duplicate newlines

Drop all duplicate newlines. No functional change.

Signed-off-by: Marek Vasut <marek.vasut+renesas@mailbox.org>
5 months agonet: Remove duplicate newlines
Marek Vasut [Sat, 13 Jul 2024 13:19:26 +0000 (15:19 +0200)]
net: Remove duplicate newlines

Drop all duplicate newlines. No functional change.

Signed-off-by: Marek Vasut <marek.vasut+renesas@mailbox.org>
5 months agomips: Remove duplicate newlines
Marek Vasut [Sat, 13 Jul 2024 13:19:25 +0000 (15:19 +0200)]
mips: Remove duplicate newlines

Drop all duplicate newlines. No functional change.

Signed-off-by: Marek Vasut <marek.vasut+renesas@mailbox.org>
5 months agomicroblaze: Remove duplicate newlines
Marek Vasut [Sat, 13 Jul 2024 13:19:24 +0000 (15:19 +0200)]
microblaze: Remove duplicate newlines

Drop all duplicate newlines. No functional change.

Signed-off-by: Marek Vasut <marek.vasut+renesas@mailbox.org>
5 months agom68k: Remove duplicate newlines
Marek Vasut [Sat, 13 Jul 2024 13:19:23 +0000 (15:19 +0200)]
m68k: Remove duplicate newlines

Drop all duplicate newlines. No functional change.

Signed-off-by: Marek Vasut <marek.vasut+renesas@mailbox.org>
5 months agolib: Remove duplicate newlines
Marek Vasut [Sat, 13 Jul 2024 13:19:22 +0000 (15:19 +0200)]
lib: Remove duplicate newlines

Drop all duplicate newlines. No functional change.

Signed-off-by: Marek Vasut <marek.vasut+renesas@mailbox.org>
5 months agofs: Remove duplicate newlines
Marek Vasut [Sat, 13 Jul 2024 13:19:21 +0000 (15:19 +0200)]
fs: Remove duplicate newlines

Drop all duplicate newlines. No functional change.

Signed-off-by: Marek Vasut <marek.vasut+renesas@mailbox.org>
5 months agoexamples: Remove duplicate newlines
Marek Vasut [Sat, 13 Jul 2024 13:19:20 +0000 (15:19 +0200)]
examples: Remove duplicate newlines

Drop all duplicate newlines. No functional change.

Signed-off-by: Marek Vasut <marek.vasut+renesas@mailbox.org>
5 months agoenv: Remove duplicate newlines
Marek Vasut [Sat, 13 Jul 2024 13:19:19 +0000 (15:19 +0200)]
env: Remove duplicate newlines

Drop all duplicate newlines. No functional change.

Signed-off-by: Marek Vasut <marek.vasut+renesas@mailbox.org>
5 months agodisk: Remove duplicate newlines
Marek Vasut [Sat, 13 Jul 2024 13:19:18 +0000 (15:19 +0200)]
disk: Remove duplicate newlines

Drop all duplicate newlines. No functional change.

Signed-off-by: Marek Vasut <marek.vasut+renesas@mailbox.org>
5 months agocommon: Remove duplicate newlines
Marek Vasut [Sat, 13 Jul 2024 13:19:17 +0000 (15:19 +0200)]
common: Remove duplicate newlines

Drop all duplicate newlines. No functional change.

Signed-off-by: Marek Vasut <marek.vasut+renesas@mailbox.org>
5 months agocmd: Remove duplicate newlines
Marek Vasut [Sat, 13 Jul 2024 13:19:16 +0000 (15:19 +0200)]
cmd: Remove duplicate newlines

Drop all duplicate newlines. No functional change.

Signed-off-by: Marek Vasut <marek.vasut+renesas@mailbox.org>
5 months agoboot: Remove duplicate newlines
Marek Vasut [Sat, 13 Jul 2024 13:19:15 +0000 (15:19 +0200)]
boot: Remove duplicate newlines

Drop all duplicate newlines. No functional change.

Signed-off-by: Marek Vasut <marek.vasut+renesas@mailbox.org>