platform/kernel/u-boot.git
20 months agoMerge branch 'master' of https://source.denx.de/u-boot/custodians/u-boot-sunxi
Tom Rini [Sun, 18 Dec 2022 13:08:55 +0000 (08:08 -0500)]
Merge branch 'master' of https://source.denx.de/u-boot/custodians/u-boot-sunxi

This cleans up each board's defconfig, and fixes the serial console on
some Olimex board. Also we lose another legacy config variable.
The rest are minor cleanups, that actually shouldn't change anything
in the build.

Passed the gitlab CI, plus briefly tested on Pine64-LTS, LicheePi Nano,
and BananaPi M1.

20 months agodoc: update Sphinx requirements for certifi
Heinrich Schuchardt [Fri, 9 Dec 2022 14:33:52 +0000 (15:33 +0100)]
doc: update Sphinx requirements for certifi

Upgrade to version 2022.12.7.

Signed-off-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
20 months agoefi: adjust ebbr to v2.1 in conformance profile
Vincent Stehlé [Fri, 16 Dec 2022 16:55:04 +0000 (17:55 +0100)]
efi: adjust ebbr to v2.1 in conformance profile

The EFI Conformance Profile Table entry for EBBR appears in v2.1.0 of the
EBBR specification[1]. Update naming accordingly.

While at it, update the EBBR version referenced in the documentation.

[1]: https://github.com/ARM-software/ebbr/releases/tag/v2.1.0

Signed-off-by: Vincent Stehlé <vincent.stehle@arm.com>
Cc: Heinrich Schuchardt <xypron.glpk@gmx.de>
Cc: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Reviewed-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
20 months agodoc: fix typos
Maxim Cournoyer [Sat, 17 Dec 2022 02:09:40 +0000 (21:09 -0500)]
doc: fix typos

Fix a few typos spot during a first read of the contribution process.

Signed-off-by: Maxim Cournoyer <maxim.cournoyer@savoirfairelinux.com>
Reviewed-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
Signed-off-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
20 months agoMerge https://source.denx.de/u-boot/custodians/u-boot-x86
Tom Rini [Thu, 15 Dec 2022 18:06:00 +0000 (13:06 -0500)]
Merge https://source.denx.de/u-boot/custodians/u-boot-x86

- Adjust CONFIG_TEXT_BASE for BayTrail based platforms
- 2 cosmetic issue fixes

20 months agosunxi: board: annotate #endif lines
Andre Przywara [Mon, 28 Nov 2022 00:02:56 +0000 (00:02 +0000)]
sunxi: board: annotate #endif lines

The legacy Allwinner code is cluttered with #ifdef's, some of them even
nested, which makes the code hard to read and error prone.
Eventually we will get rid of most of them, but for now let's at least
annotate the #endif lines with the corresponding symbol the bracket
started with.

Reviewed-by: Samuel Holland <samuel@sholland.org>
Tested-by: Samuel Holland <samuel@sholland.org>
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
20 months agosunxi: remove bogus mmc_pinmux_setup() prototype
Andre Przywara [Mon, 28 Nov 2022 00:55:20 +0000 (00:55 +0000)]
sunxi: remove bogus mmc_pinmux_setup() prototype

Since all callers of mmc_pinmux_setup() are located after the definition
of that function, there is no need for a forward declaration (anymore?).

Remove the prototype along with its #ifdef guards.

Reviewed-by: Samuel Holland <samuel@sholland.org>
Tested-by: Samuel Holland <samuel@sholland.org>
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
20 months agosunxi: remove unused CONFIG_MMC_SUNXI_SLOT
Andre Przywara [Mon, 28 Nov 2022 00:03:53 +0000 (00:03 +0000)]
sunxi: remove unused CONFIG_MMC_SUNXI_SLOT

There is a CONFIG_MMC_SUNXI_SLOT definition in our sunxi_common.h config
header, which was used to note the first MMC controller to initialise.
The definition in that header was always set to 0, with no easy way of
overriding this, and certainly none of the existing boards made any use
of that (non-)feature.
Remove that definition and replace it with a constant 0 in the only
user, in board.c. It turns out that this is safe, as this is only used
in the SPL, and the BROM also unconditionally initialises MMC0.
This also removes the last legacy config symbol with SUN*I in it from
the whitelist.

Reviewed-by: Samuel Holland <samuel@sholland.org>
Tested-by: Samuel Holland <samuel@sholland.org>
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
20 months agox86: cosmetic: Fix a typo in the reserve_arch() comments
Bin Meng [Thu, 24 Nov 2022 03:39:23 +0000 (11:39 +0800)]
x86: cosmetic: Fix a typo in the reserve_arch() comments

It should be fsp_continue().

Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
20 months agox86: som-db5800-som-6867: Adjust CONFIG_TEXT_BASE
Bin Meng [Thu, 24 Nov 2022 08:19:54 +0000 (16:19 +0800)]
x86: som-db5800-som-6867: Adjust CONFIG_TEXT_BASE

At present U-Boot no longer builds as a complete rom for som-db5800-som-6867.

    BINMAN  .binman_stamp
  Wrote map file './rom.map' to show errors
  binman: Section '/binman/rom': contents size 0x80302c (8400940) exceeds section size 0x800000 (8388608)

Checking rom.map we see 'intel-vga' section is overlapped with
other sections:

  <none>     fff00000  0009f7c8  u-boot-with-ucode-ptr
  <none>     fff90000  00010000  intel-vga
  <none>     fff9f7c8  00001aae  u-boot-dtb-with-ucode

Let's adjust CONFIG_TEXT_BASE to allow more space for U-Boot codes.

Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
20 months agox86: dfi-bt700: Adjust CONFIG_TEXT_BASE
Bin Meng [Thu, 24 Nov 2022 07:58:17 +0000 (15:58 +0800)]
x86: dfi-bt700: Adjust CONFIG_TEXT_BASE

At present U-Boot no longer builds as a complete rom for all the
configs of dfi-bt700.

    BINMAN  .binman_stamp
  Wrote map file './rom.map' to show errors
  binman: Section '/binman/rom': contents size 0x80e836 (8448054) exceeds section size 0x800000 (8388608)

Checking rom.map we see 'intel-vga' section is overlapped with
other sections:

  <none>     fff00000  000aac90  u-boot-with-ucode-ptr
  <none>     fffa0000  00010000  intel-vga
  <none>     fffaac90  00001df0  u-boot-dtb-with-ucode
  <none>     fffaca80  00019800  u-boot-ucode

Let's adjust CONFIG_TEXT_BASE to allow more space for U-Boot codes.

Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Stefan Roese <sr@denx.de>
20 months agox86: conga-qeval20-qa3-e3845: Adjust CONFIG_TEXT_BASE
Bin Meng [Thu, 24 Nov 2022 07:33:09 +0000 (15:33 +0800)]
x86: conga-qeval20-qa3-e3845: Adjust CONFIG_TEXT_BASE

At present U-Boot no longer builds as a complete rom for all the
configs of conga-qeval20-qa3-e3845.

    BINMAN  .binman_stamp
  Wrote map file './rom.map' to show errors
  binman: Section '/binman/rom': contents size 0x80b680 (8435328) exceeds section size 0x800000 (8388608)

Checking rom.map we see 'intel-vga' section is overlapped with
other sections:

  <none>     fff00000  000a7cb0  u-boot-with-ucode-ptr
  <none>     fffa0000  00010000  intel-vga
  <none>     fffa7cb0  00001c1a  u-boot-dtb-with-ucode
  <none>     fffa98d0  00019800  u-boot-ucode

Let's adjust CONFIG_TEXT_BASE to allow more space for U-Boot codes.

Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
20 months agox86: minnowmax: Adjust CONFIG_TEXT_BASE
Bin Meng [Thu, 24 Nov 2022 03:07:13 +0000 (11:07 +0800)]
x86: minnowmax: Adjust CONFIG_TEXT_BASE

At present U-Boot no longer builds as a complete rom for minnowmax.

    BINMAN  .binman_stamp
  Wrote map file './rom.map' to show errors
  binman: Section '/binman/rom': contents size 0x803146 (8401222) exceeds section size 0x800000 (8388608)

Checking rom.map we see 'fdtmap' section is overlapped with
'intel-vga' section:

  <none>     fffa1390  00019800  u-boot-ucode
  <none>     fffb0000  00010000  intel-vga
  <none>     fffbab90  00000539  fdtmap

Let's adjust CONFIG_TEXT_BASE to allow more space for U-Boot codes.

Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
20 months agox86: bayleybay: Adjust CONFIG_TEXT_BASE
Bin Meng [Thu, 24 Nov 2022 03:24:37 +0000 (11:24 +0800)]
x86: bayleybay: Adjust CONFIG_TEXT_BASE

At present U-Boot no longer builds as a complete rom for bayleybay.

    BINMAN  .binman_stamp
  Wrote map file './rom.map' to show errors
  binman: Section '/binman/rom': contents size 0x814706 (8472326) exceeds section size 0x800000 (8388608)

Checking rom.map we see 'fdtmap' section is overlapped with
'intel-vga' and 'intel-fsp' sections:

  <none>     fffa2150  0002a000  u-boot-ucode
  <none>     fffb0000  00010000  intel-vga
  <none>     fffc0000  00038000  intel-fsp
  <none>     fffcc150  00000539  fdtmap

Let's adjust CONFIG_TEXT_BASE to allow more space for U-Boot codes.

Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
20 months agox86: Fix i8259 ifdef include guard
Alistair Delva [Mon, 26 Sep 2022 22:23:26 +0000 (22:23 +0000)]
x86: Fix i8259 ifdef include guard

When building U-Boot with clang, it notices that the i8259.h include
guard does not work correctly due to a typo. Fix it.

Signed-off-by: Alistair Delva <adelva@google.com>
Cc: Simon Glass <sjg@chromium.org>
Cc: Bin Meng <bmeng.cn@gmail.com>
Cc: Nick Desaulniers <ndesaulniers@google.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
20 months agosunxi: Fix serial console for A10s-OLinuXino-MICRO
Mark Kettenis [Sun, 11 Dec 2022 22:00:54 +0000 (23:00 +0100)]
sunxi: Fix serial console for A10s-OLinuXino-MICRO

On this board CONFIG_CONS_INDEX needs to be 1 unlike other sun5i
boards.  Since this is the default, remove to bogus setting.

Fixes: 7095f8641863 ("sunxi: Convert CONS_INDEX to Kconfig")
Signed-off-by: Mark Kettenis <kettenis@openbsd.org>
Reviewed-by: Andre Przywara <andre.przywara@arm.com>
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
20 months agopinctrl: sunxi: Add P2WI and RSB pinmuxes
Samuel Holland [Fri, 18 Nov 2022 04:22:27 +0000 (22:22 -0600)]
pinctrl: sunxi: Add P2WI and RSB pinmuxes

P2WI and RSB are used to communicate with a PMIC. Most SoCs have only
one possible pinmux. F1C100s has two possibilities, with different mux
values, so omit it until some board needs one of them.

Signed-off-by: Samuel Holland <samuel@sholland.org>
Reviewed-by: Andre Przywara <andre.przywara@arm.com>
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
20 months agosunxi: define SYS_MONITOR_LEN in Kconfig, not _defconfig
Andre Przywara [Sat, 19 Nov 2022 16:10:59 +0000 (16:10 +0000)]
sunxi: define SYS_MONITOR_LEN in Kconfig, not _defconfig

Commit 08574ed339fb ("Convert CONFIG_SYS_MONITOR_LEN to Kconfig") moved
the definition of said config variable from the common sunxi header to
*every board's* defconfig.
This is a platform choice, not board specific, so remove the variable
from there, instead set the one value for all Allwinner boards in
Kconfig.

Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Acked-by: Jernej Skrabec <jernej.skrabec@gmail.com>
20 months agoMerge branch '2022-12-12-build-related-updates' into next
Tom Rini [Tue, 13 Dec 2022 16:34:59 +0000 (11:34 -0500)]
Merge branch '2022-12-12-build-related-updates' into next

- Updates to the keymile platforms for DM_I2C and text based
  environment migration
- Finish migration of MTDPART/MTDIDS_DEFAULT to defconfig
- Disable warning about RWX segments with gcc-12.2

20 months agoRevert "cmd: pxe_utils: Check fdtcontroladdr in label_boot"
Tom Rini [Tue, 13 Dec 2022 14:26:25 +0000 (09:26 -0500)]
Revert "cmd: pxe_utils: Check fdtcontroladdr in label_boot"

With the change here, all extlinux.conf files with only "KERNEL
/fitImage" don't work anymore. One common example of this would be those
files generated by thee Poky/OE WIC bootimg-partition bootloader
partition generator.

This reverts commit d5ba6188dfbf6bb68354bec86e483623f1f6dae2.

Reported-by: Neil Armstrong <neil.armstrong@linaro.org>
Reported-by: Quentin Schulz <quentin.schulz@theobroma-systems.com>
Signed-off-by: Tom Rini <trini@konsulko.com>
20 months agokm/mpc8360: remove unused CONFIG_SYS_PAXE defines
Holger Brunck [Fri, 2 Dec 2022 17:22:43 +0000 (18:22 +0100)]
km/mpc8360: remove unused CONFIG_SYS_PAXE defines

These are unused defines and can be dropped.

Signed-off-by: Holger Brunck <holger.brunck@hitachienergy.com>
20 months agokm/ppc: migrate all mpc83xx to DM_I2C
Holger Brunck [Fri, 2 Dec 2022 17:22:42 +0000 (18:22 +0100)]
km/ppc: migrate all mpc83xx to DM_I2C

Enable DM_I2C and I2C mux to get rid of the usage of the legacy
i2c driver.

Signed-off-by: Holger Brunck <holger.brunck@hitachienergy.com>
20 months agoboard/km: remove obsolete ARCH_KIRKWOOD
Holger Brunck [Fri, 2 Dec 2022 17:22:41 +0000 (18:22 +0100)]
board/km: remove obsolete ARCH_KIRKWOOD

We already removed the support for these boards.

Signed-off-by: Holger Brunck <holger.brunck@hitachienergy.com>
20 months agoboard/km/secu: migrate to use environment text files
Holger Brunck [Fri, 2 Dec 2022 17:22:40 +0000 (18:22 +0100)]
board/km/secu: migrate to use environment text files

Instead of having these defines in a header file, move them to
a simple text file.

Signed-off-by: Holger Brunck <holger.brunck@hitachienergy.com>
20 months agoboard/km/cent2: migrate to environment text file
Holger Brunck [Fri, 2 Dec 2022 17:22:39 +0000 (18:22 +0100)]
board/km/cent2: migrate to environment text file

Use like the other boards a text file for the environment.
As this is the last user of keymile-common.h we can now remove this
file completely.

Signed-off-by: Holger Brunck <holger.brunck@hitachienergy.com>
20 months agokm/powerpc: migrate to env.txt file
Holger Brunck [Fri, 2 Dec 2022 17:22:38 +0000 (18:22 +0100)]
km/powerpc: migrate to env.txt file

Use already present common.env file and add a powerpc specific env
so that we can move all the environment defines to text files.

Signed-off-by: Holger Brunck <holger.brunck@hitachienergy.com>
20 months agoboard/km: move ls102xa boards to environment text files
Holger Brunck [Fri, 2 Dec 2022 17:22:37 +0000 (18:22 +0100)]
board/km: move ls102xa boards to environment text files

Create a common.env which we can use later on also for other boards.

Signed-off-by: Holger Brunck <holger.brunck@hitachienergy.com>
20 months agoconfigs: remove support of MTDIDS_DEFAULT/MTDPARTS_DEFAULT
Patrick Delaunay [Wed, 7 Dec 2022 08:26:44 +0000 (09:26 +0100)]
configs: remove support of MTDIDS_DEFAULT/MTDPARTS_DEFAULT

Complete the migration of MTDPARTS_DEFAULT / MTDIDS_DEFAULT in Kconfig;
this patch removes the support of MTDIDS_DEFAULT / MTDPARTS_DEFAULT
in the configuration files (include/configs/*.h).

Signed-off-by: Patrick Delaunay <patrick.delaunay@foss.st.com>
20 months agoconfigs: SBx81LIFKW: move MTDPART_DEFAULT in defconfig
Patrick Delaunay [Wed, 7 Dec 2022 08:26:43 +0000 (09:26 +0100)]
configs: SBx81LIFKW: move MTDPART_DEFAULT in defconfig

Replace MTDPARTS_DEFAULT in the config include file by
CONFIG_MTDPARTS_DEFAULT in defconfig to complete the Kconfig migration.

Signed-off-by: Patrick Delaunay <patrick.delaunay@foss.st.com>
20 months agoconfigs: SBx81LIFXCAT: move MTDPART_DEFAULT in defconfig
Patrick Delaunay [Wed, 7 Dec 2022 08:26:42 +0000 (09:26 +0100)]
configs: SBx81LIFXCAT: move MTDPART_DEFAULT in defconfig

Replace MTDPARTS_DEFAULT in the config include file by
CONFIG_MTDPARTS_DEFAULT in defconfig to complete the Kconfig migration.

Signed-off-by: Patrick Delaunay <patrick.delaunay@foss.st.com>
20 months agoconfigs: x530: move MTDPART/MTDIDS_DEFAULT in defconfig
Patrick Delaunay [Wed, 7 Dec 2022 08:26:41 +0000 (09:26 +0100)]
configs: x530: move MTDPART/MTDIDS_DEFAULT in defconfig

Replace MTDIDS_DEFAULT and MTDPARTS_DEFAULT in the config include file by
CONFIG_MTDIDS_DEFAULT and CONFIG_MTDPARTS_DEFAULT in defconfig to complete
the Kconfig migration.

Signed-off-by: Patrick Delaunay <patrick.delaunay@foss.st.com>
20 months agoconfigs: am333x_guardian: move MTDIDS_DEFAULT in defconfif
Patrick Delaunay [Wed, 7 Dec 2022 08:26:40 +0000 (09:26 +0100)]
configs: am333x_guardian: move MTDIDS_DEFAULT in defconfif

Replace MTDIDS_DEFAULT in config include file by CONFIG_MTDIDS_DEFAULT
in defonfig to complete the Kconfig migration

Signed-off-by: Patrick Delaunay <patrick.delaunay@foss.st.com>
20 months agoMakefile: link with --no-warn-rwx-segments
Tom Rini [Wed, 7 Dec 2022 20:49:43 +0000 (15:49 -0500)]
Makefile: link with --no-warn-rwx-segments

We borrow from the Linux Kernel 0d362be5b142 ("Makefile: link with -z
noexecstack --no-warn-rwx-segments") here to disable the RWX segment
linking warnings. We do not also bring in -z noexecstack as that
requires auditing and using ".note.GNU-stack" on assembly functions
which do need this feature. Further, we now introduce KBUILD_EFILDFLAGS
so that we can also pass --no-warn-rwx-segments when linking EFI
applications, and those do explicitly pass -z execstack.

Cc: Heinrich Schuchardt <xypron.glpk@gmx.de>
Cc: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Signed-off-by: Tom Rini <trini@konsulko.com>
Reviewed-by: Nick Desaulniers <ndesaulniers@google.com>
Reviewed-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
20 months agoMerge branch '2022-12-12-assorted-improvements' into next
Tom Rini [Mon, 12 Dec 2022 21:32:37 +0000 (16:32 -0500)]
Merge branch '2022-12-12-assorted-improvements' into next

- A number of PXE / sysboot related improvements
- Nuvoton updates
- Small updates to omap4, ARCH_OMAP2PLUS itself, mediatek
  pcie_designware.

20 months agoconfigs: set CONFIG_LMB_MAX_REGIONS=64 for all mt798[16] boards
Daniel Golle [Tue, 29 Nov 2022 19:37:59 +0000 (19:37 +0000)]
configs: set CONFIG_LMB_MAX_REGIONS=64 for all mt798[16] boards

With recently added wireless offloading features in Linux [1] the
number of reserved memory regions with MediaTek SoCs supporting
offloading wireless-to-Ethernet traffic grew beyond the default (8)
which breaks booting Linux:
ERROR: Failed to allocate 0xa6ac bytes below 0xc0000000.
device tree - allocation error
FDT creation failed!
resetting ...

Raise CONFIG_LMB_MAX_REGIONS to 64 like it is already done for other
SoCs which require a larger number of reserved memory regions, eg.
exynos78x0 based a3y17lte, a5y17lte and a7y17lte or dragonboard845c.

[1]: https://lore.kernel.org/netdev/e3489a697b404bd47447190cd2e5adf090ae61c2.1667687249.git.lorenzo@kernel.org/
     https://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next.git/commit/?id=eed4f1ddad8c5ad7596b229caec8bd7b477b81ee

Signed-off-by: Daniel Golle <daniel@makrotopia.org>
20 months agoARM: dts: npcm7xx: add npcm750 gpio node compatible name
Jim Liu [Mon, 28 Nov 2022 02:34:47 +0000 (10:34 +0800)]
ARM: dts: npcm7xx: add npcm750 gpio node compatible name

Add npcm750 gpio node compatible name

Signed-off-by: Jim Liu <JJLIU0@nuvoton.com>
20 months agoARM: dts: npcm8xx: add npcm845 function node
Jim Liu [Mon, 28 Nov 2022 02:32:44 +0000 (10:32 +0800)]
ARM: dts: npcm8xx: add npcm845 function node

1. add usb phy
2. add ehci ohci sdhci
3. add pinctrl node
4. add fiu node

Signed-off-by: Jim Liu <JJLIU0@nuvoton.com>
20 months agoARM: config: enable function for nuvoton npcm845 bmc
Jim Liu [Mon, 28 Nov 2022 02:32:43 +0000 (10:32 +0800)]
ARM: config: enable function for nuvoton npcm845 bmc

Signed-off-by: Jim Liu <JJLIU0@nuvoton.com>
20 months agoomap4: make musb probeable by simple bus
Andreas Kemnade [Sat, 26 Nov 2022 22:30:09 +0000 (23:30 +0100)]
omap4: make musb probeable by simple bus

Like other peripherals important for booting,
do not rely on ti-sysc compatibility alone

Signed-off-by: Andreas Kemnade <andreas@kemnade.info>
20 months agodistro/pxeboot: Handle prompt variable
Manuel Traut [Fri, 18 Nov 2022 08:00:27 +0000 (09:00 +0100)]
distro/pxeboot: Handle prompt variable

Regarding the documentation found here:
https://github.com/u-boot/u-boot/blob/master/common/menu.c#L347

If both timeout and prompt is set to 0 the default entry shall
be booted immediately. However the current behaviour is that
the prompt is shown (tested with distroboot) until the user
selects an entry (no timeout).

This change implements a behaviour as documented. It was tested
with distroboot.

Signed-off-by: Manuel Traut <manuel.traut@mt.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Simon Glass <sjg@chromium.org>
20 months agocmd: pxe: use strdup to copy config
Patrick Delaunay [Fri, 28 Oct 2022 09:01:20 +0000 (11:01 +0200)]
cmd: pxe: use strdup to copy config

Replace malloc and strcpy by strdup in
function parse_label_kernel.

Signed-off-by: Patrick Delaunay <patrick.delaunay@foss.st.com>
Reviewed-by: Neil Armstrong <neil.armstrong@linaro.org>
20 months agocmd: pxe: support INITRD and FDT selection with FIT
Patrick Delaunay [Fri, 28 Oct 2022 09:01:19 +0000 (11:01 +0200)]
cmd: pxe: support INITRD and FDT selection with FIT

Since the commit d5ba6188dfbf ("cmd: pxe_utils: Check fdtcontroladdr
in label_boot") the FDT or the FDTDIR label is required in extlinux.conf
and the fallback done by bootm command when only the device tree present
in this command parameters is no more performed when FIT is used for
kernel.

When the label FDT or FDTDIR are absent or if the device tree file is
absent, the PXE command in U-Boot uses the default U-Boot device tree
selected by fdtcontroladdr = gd->fdt_blob, it is the "Scenario 3".

With this scenario the bootm FIP fallback is no more possible with
the extlinux.conf when only "kernel" label is present and is a FIP:

  kernel <path>#<conf>[#<extra-conf[#...]]

As the U-Boot FDT is always provided in the third bootm argument,
the device tree found in FIP is not used as fallback, it was done
previously in boot_get_fdt().

This patch adds a new field kernel_label to save the full kernel label.
The FDT bootm parameters use the kernel address (to avoid to load a
second time the same FIP) and the config when this full label is reused
for "fdt" or "initrd" label.

This FIP support in extlinux.conf is restored when the "FDT" label
can be found and select the same FIP (identical file and configuration):

  kernel <path>#<conf>[#<extra-conf[#...]]
  fdt <path>#<conf>[#<extra-conf[#...]]

The patch add also this possibility for initrd.

  initrd <path>#<conf>[#<extra-conf[#...]]

Signed-off-by: Patrick Delaunay <patrick.delaunay@foss.st.com>
Reviewed-by: Neil Armstrong <neil.armstrong@linaro.org>
20 months agocmd: pxe: reorder kernel treatment in label_boot
Patrick Delaunay [Fri, 28 Oct 2022 09:01:18 +0000 (11:01 +0200)]
cmd: pxe: reorder kernel treatment in label_boot

Reorder kernel treatment in label_boot at the beginning of the function.

This patch doesn't change the pxe command behavior, it is only a
preliminary step for next patch, build kernel_addr before parsing
the label initrd and fdt to build the next bootm arguments.

Signed-off-by: Patrick Delaunay <patrick.delaunay@foss.st.com>
Reviewed-by: Neil Armstrong <neil.armstrong@linaro.org>
20 months agodrivers: pci: pcie_dw_common: add upper-limit to iATU
Ben Dooks [Thu, 20 Oct 2022 15:51:09 +0000 (16:51 +0100)]
drivers: pci: pcie_dw_common: add upper-limit to iATU

The 4.6 spec added an upper 32bits to the ATU limit, and since this
driver is already assuming the unrolled feature added in the 4.8
specification this really should be set.

This is causing a bug with testing against the QEMU model as it
defaults the viewports to fully open and not setting this causes
the config viewport to become most of memory (obviously stopping
the emulated system working correctly)

Signed-off-by: Ben Dooks <ben.dooks@sifive.com>
20 months agoarm: mach-omap2: Move common image process functions out of board files
Andrew Davis [Thu, 20 Oct 2022 17:12:19 +0000 (12:12 -0500)]
arm: mach-omap2: Move common image process functions out of board files

The functions board_fit_image_post_process() and board_tee_image_process()
are not actually board specific (despite their names). Any board using the
OMAP2 family can use these functions. Move them to boot-common.c.

Signed-off-by: Andrew Davis <afd@ti.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
20 months agoMerge https://source.denx.de/u-boot/custodians/u-boot-marvell
Tom Rini [Mon, 12 Dec 2022 14:00:58 +0000 (09:00 -0500)]
Merge https://source.denx.de/u-boot/custodians/u-boot-marvell

- mvebu: Espressobin: Fix default env variables (Derek)

20 months agoMerge tag 'u-boot-stm32-20221212' of https://source.denx.de/u-boot/custodians/u-boot-stm
Tom Rini [Mon, 12 Dec 2022 13:59:13 +0000 (08:59 -0500)]
Merge tag 'u-boot-stm32-20221212' of https://source.denx.de/u-boot/custodians/u-boot-stm

phy: usbphyc: use regulator_set_enable_if_allowed for disabling vbus supply
dm: pmic: ignore disabled node in pmic_bind_children

20 months agodm: pmic: ignore disabled node in pmic_bind_children
Patrick Delaunay [Wed, 26 Oct 2022 13:05:10 +0000 (15:05 +0200)]
dm: pmic: ignore disabled node in pmic_bind_children

Ignore the disabled children node in pmic_bind_children() so the
disabled regulators in device tree are not registered.

This patch is based on the dm_scan_fdt_node() code - only the
activated nodes are bound -  and it solves possible issue when a
deactivated regulator is bound, error for duplicated regulator name
for example.

Signed-off-by: Patrick Delaunay <patrick.delaunay@foss.st.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
20 months agophy: usbphyc: use regulator_set_enable_if_allowed for disabling vbus supply
Patrick Delaunay [Tue, 20 Sep 2022 11:39:56 +0000 (13:39 +0200)]
phy: usbphyc: use regulator_set_enable_if_allowed for disabling vbus supply

Use regulator_set_enable_if_allowed() api instead of regulator_set_enable()
while disabling vbus supply. This way the driver doesn't see an error
when it disable an always-on regulator for VBUS.

This patch is needed for STM32MP157C-DK2 board when the regulator
v3v3: buck4 used as the phy vbus supply in kernel device tree
is always on with the next hack for low power use-case:

&usbphyc_port0 {
        ...
/*
 * Hack to keep hub active until all connected devices are suspended
 * otherwise the hub will be powered off as soon as the v3v3 is disabled
 * and it can disturb connected devices.
 */
connector {
compatible = "usb-a-connector";
vbus-supply = <&v3v3>;
};
};

Without this patch and the previous update in DT the command
"usb stop" failed and the next command "usb start" cause a crash.

Signed-off-by: Patrick Delaunay <patrick.delaunay@foss.st.com>
Reviewed-by: Patrice Chotard <patrice.chotard@foss.st.com>
Reviewed-by: Marek Vasut <marex@denx.de>
20 months agoarm: mvebu: Espressobin: Fix default env variables
Derek LaHousse [Mon, 12 Dec 2022 06:34:17 +0000 (07:34 +0100)]
arm: mvebu: Espressobin: Fix default env variables

Default env variables on Espressobin boards are broken since commit c4df0f6f315c
("arm: mvebu: Espressobin: Set default value for $fdtfile env variable") as well
as the 'env default -a' command.

The algorithm to find free space in the default_environment[] array returns
after the first env variable instead of the correct position of the last
variable, where there is allocated free space.

This causes that U-Boot board_late_init() function to overwrite a portion of the
default environment with $ethXaddr and $fdtfile variables immediately after the
first env variable and so it is overwriting other variables.

This patch also adds an additional null byte to terminate the environment array.

But U-Boot board_late_init() function do not fill this nul byte explicitly. And
because of that, U-Boot is later trying to interpret remaining buffer as a
continuation of variable list. Normally buffer should be empty but due to the
above issue, it contains garbage from remaining env variables.

For example 'env default -a' command results in damaging variable names. It was
observed that scritaddr variable name was changed to criptaddr (without leading
's').

This bug was reported and discussed on the Armbian forum:
https://forum.armbian.com/topic/19564-making-espressobin-v7-work-in-2022/?do=findComment&comment=138136

Fix these issues in two steps:

1) Change code which finds free space for dynamic env variables in
default_environment[] array by jumping to the end of the variable list instead
of jumping after the first defined variable. [By Derek]

2) Add code which appends terminating nul byte as indication of the end of the
env list, after the last nul term env string. [By Pali]

Fixes: c4df0f6f315c ("arm: mvebu: Espressobin: Set default value for $fdtfile env variable")
Signed-off-by: Derek LaHousse <derek@seaofdirac.org>
Signed-off-by: Pali Rohár <pali@kernel.org>
Signed-off-by: Stefan Roese <sr@denx.de>
20 months agoMerge tag 'u-boot-nand-20221211' of https://source.denx.de/u-boot/custodians/u-boot...
Tom Rini [Sun, 11 Dec 2022 14:40:25 +0000 (09:40 -0500)]
Merge tag 'u-boot-nand-20221211' of https://source.denx.de/u-boot/custodians/u-boot-nand-flash

Merge tag 'u-boot-nand-20221211' of https://source.denx.de/u-boot/custodians/u-boot-nand-flash

- cmd: nand: Extend nand info to print ecc information
- rawnand: omap_gpmc: driver model support (the first patches of the series)
- mtd: nand: make Samsung SLC NAND usable again
- cmd: mtd: check if a block has to be skipped or erased
- spl: spl_legacy: fix invalid offset in SPL_COPY_PAYLOAD_ONLY

20 months agospl: spl_legacy: fix invalid offset in SPL_COPY_PAYLOAD_ONLY
Dai Okamura [Fri, 9 Dec 2022 11:40:21 +0000 (20:40 +0900)]
spl: spl_legacy: fix invalid offset in SPL_COPY_PAYLOAD_ONLY

This fixes the header offset calculation.

This issue was found on uniphier v7 SoCs with SPL.

Fixes: 06377c5a1f ("spl: spl_legacy: Fix NAND boot on OMAP3 BeagleBoard")
Signed-off-by: Dai Okamura <okamura.dai@socionext.com>
Reviewed-By: Michael Trimarchi <michael@amarulasolutions.com>
Acked-by: Michael Trimarchi <michael@amarulasolutions.com>
Link: https://lore.kernel.org/all/20221209114021.3074978-1-okamura.dai@socionext.com
Signed-off-by: Dario Binacchi <dario.binacchi@amarulasolutions.com>
20 months agomtd: nand: mxs_nand_spl: don't read useless pages
Dario Binacchi [Sun, 20 Nov 2022 09:57:04 +0000 (10:57 +0100)]
mtd: nand: mxs_nand_spl: don't read useless pages

The patch prevents pages beyond the last from being unnecessarily read.
This occurs when the last page to be read is not the last page of the
last block. Before this change we would have read all the pages up to
the end of the last block.

Suggested-by: Michael Trimarchi <michael@amarulasolutions.com>
Co-developed-by: Michael Trimarchi <michael@amarulasolutions.com>
Signed-off-by: Michael Trimarchi <michael@amarulasolutions.com>
Signed-off-by: Dario Binacchi <dario.binacchi@amarulasolutions.com>
Acked-by: Michael Trimarchi <michael@amarulasolutions.com>
Link: https://lore.kernel.org/all/20221120095705.3019295-1-dario.binacchi@amarulasolutions.com
20 months agomtd: nand: drop EXPORT_SYMBOL_GPL for nanddev_erase()
Dario Binacchi [Tue, 8 Nov 2022 09:07:19 +0000 (10:07 +0100)]
mtd: nand: drop EXPORT_SYMBOL_GPL for nanddev_erase()

This function is only used within this module, so it is no longer
necessary to use EXPORT_SYMBOL_GPL().

This patch parallels the work done in the following patch:
https://lore.kernel.org/linux-mtd/20221018170205.1733958-1-dario.binacchi@amarulasolutions.com

Signed-off-by: Dario Binacchi <dario.binacchi@amarulasolutions.com>
Reviewed-By: Michael Trimarchi <michael@amarulasolutions.com>
Link: https://lore.kernel.org/all/20221108090719.3631621-1-dario.binacchi@amarulasolutions.com
20 months agocmd: mtd: check if a block has to be skipped or erased
Dario Binacchi [Sun, 30 Oct 2022 14:14:13 +0000 (15:14 +0100)]
cmd: mtd: check if a block has to be skipped or erased

As reported by patch [1], the `mtd erase' command should not erase bad
blocks.
To force bad block erasing you have to use the `mtd erase.dontskipbad'
command.

This patch tries to fix the same issue without modifying code taken
from the linux kernel, in order to make further upgrades easier.

[1] https://lore.kernel.org/all/20221006031501.110290-2-mikhail.kshevetskiy@iopsys.eu/
Suggested-by: Michael Trimarchi <michael@amarulasolutions.com>
Co-developed-by: Michael Trimarchi <michael@amarulasolutions.com>
Signed-off-by: Michael Trimarchi <michael@amarulasolutions.com>
Co-developed-by: Mikhail Kshevetskiy <mikhail.kshevetskiy@iopsys.eu>
Signed-off-by: Mikhail Kshevetskiy <mikhail.kshevetskiy@iopsys.eu>
Tested-by: Mikhail Kshevetskiy <mikhail.kshevetskiy@iopsys.eu>
Signed-off-by: Dario Binacchi <dario.binacchi@amarulasolutions.com>
20 months agomtd: nand: make Samsung SLC NAND usable again
Michael Trimarchi [Fri, 21 Oct 2022 06:05:36 +0000 (08:05 +0200)]
mtd: nand: make Samsung SLC NAND usable again

Upstream linux commit 69fc01296c9281

commit a1286a1fc416 ("mtd: nand: Move Samsung specific init/detection
logic in nand_samsung.c") introduced a regression for Samsung SLC NAND
chips. Prior to this commit chip->bits_per_cell was initialized by calling
nand_get_bits_per_cell() before using nand_is_slc().
With the offending commit this call is skipped, leaving
chip->bits_per_cell cleared to zero when the manufacturer specific
'.detect' function calls nand_is_slc() which in turn interprets
bits_per_cell != 1 as indication for an MLC chip.
The effect is that e.g. a K9F1G08U0F NAND chip is falsely detected as
MLC NAND with 4KiB page size rather than SLC with 2KiB page size.

Add a call to nand_get_bits_per_cell() before calling the .detect hook
function in nand_manufacturer_detect(), so that the nand_is_slc()
calls in the manufacturer specific code will return correct results.

Reported-by: Marcin Gołaś <marcingol30@gmail.com>
Signed-off-by: Michael Trimarchi <michael@amarulasolutions.com>
Reviewed-by: Dario Binacchi <dario.binacchi@amarulasolutions.com>
Link: https://lore.kernel.org/all/20221021060536.11747-1-michael@amarulasolutions.com
Signed-off-by: Dario Binacchi <dario.binacchi@amarulasolutions.com>
20 months agomtd: rawnand: omap_gpmc: Reduce .bss usage
Roger Quadros [Tue, 11 Oct 2022 11:50:06 +0000 (14:50 +0300)]
mtd: rawnand: omap_gpmc: Reduce .bss usage

Allocate omap_ecclayout on the heap as we have
limited .bss space on AM64 R5 SPL configuration.

Reduces .bss usage by 2984 bytes.

Signed-off-by: Roger Quadros <rogerq@kernel.org>
Reviewed-By: Michael Trimarchi <michael@amarulasolutions.com>
Link: https://lore.kernel.org/all/20221011115012.6181-9-rogerq@kernel.org
Signed-off-by: Dario Binacchi <dario.binacchi@amarulasolutions.com>
20 months agomtd: rawnand: nand_spl_loaders: Fix cast type build warning
Roger Quadros [Tue, 11 Oct 2022 11:50:05 +0000 (14:50 +0300)]
mtd: rawnand: nand_spl_loaders: Fix cast type build warning

Fixes the below build warning on 64-bit platforms.

drivers/mtd/nand/raw/nand_spl_loaders.c:26:21: warning: cast from pointer to integer of different size [-Wpointer-to-int-cast]
      dst = (void *)((int)dst - page_offset);

Signed-off-by: Roger Quadros <rogerq@kernel.org>
Reviewed-by: Michael Trimarchi <michael@amarulasolutions.com>
Link: https://lore.kernel.org/all/20221011115012.6181-8-rogerq@kernel.org
Signed-off-by: Dario Binacchi <dario.binacchi@amarulasolutions.com>
21 months agomtd: rawnand: omap_gpmc: Optimize NAND reads
Roger Quadros [Tue, 11 Oct 2022 11:50:02 +0000 (14:50 +0300)]
mtd: rawnand: omap_gpmc: Optimize NAND reads

Rename omap_nand_read() to omap_nand_read_buf() to reflect
actual behaviour.

Use FIFO read address instead of raw read address for reads.

The GPMC automatically converts 32-bit/16-bit reads to NAND
device specific reads (8/16 bit). Use the largest possible
read granularity size for more efficient reads.

Signed-off-by: Roger Quadros <rogerq@kernel.org>
Reviewed-by: Michael Trimarchi <michael@amarulasolutions.com>
Link: https://lore.kernel.org/all/20221011115012.6181-5-rogerq@kernel.org
Signed-off-by: Dario Binacchi <dario.binacchi@amarulasolutions.com>
21 months agomtd: rawnand: omap_gpmc: Fix build warning on 64-bit platforms
Roger Quadros [Tue, 11 Oct 2022 11:50:01 +0000 (14:50 +0300)]
mtd: rawnand: omap_gpmc: Fix build warning on 64-bit platforms

Pointer size cannot be assumed to be 32-bit, so use
use uintptr_t instead of uint32_t.

Fixes the below build warning on 64-bit builds.

drivers/mtd/nand/raw/omap_gpmc.c:439:10: warning: cast from pointer to integer of different size [-Wpointer-to-int-cast]
  head = ((uint32_t) buf) % 4;

Signed-off-by: Roger Quadros <rogerq@kernel.org>
Reviewed-by: Michael Trimarchi <michael@amarulasolutions.com>
Link: https://lore.kernel.org/all/20221011115012.6181-4-rogerq@kernel.org
Signed-off-by: Dario Binacchi <dario.binacchi@amarulasolutions.com>
21 months agomtd: rawnand: omap_gpmc: Enable build for K2/K3 platforms
Roger Quadros [Tue, 11 Oct 2022 11:50:00 +0000 (14:50 +0300)]
mtd: rawnand: omap_gpmc: Enable build for K2/K3 platforms

The GPMC module is present on some K2 and K3 SoCs.
Enable building GPMC NAND driver for K2/K3 platforms.

Signed-off-by: Roger Quadros <rogerq@kernel.org>
Reviewed-By: Michael Trimarchi <michael@amarulasolutions.com>
Link: https://lore.kernel.org/all/20221011115012.6181-3-rogerq@kernel.org
Signed-off-by: Dario Binacchi <dario.binacchi@amarulasolutions.com>
21 months agomtd: rawnand: omap_gpmc: Deprecate asm/arch/mem.h
Roger Quadros [Tue, 11 Oct 2022 11:49:59 +0000 (14:49 +0300)]
mtd: rawnand: omap_gpmc: Deprecate asm/arch/mem.h

We want to get rid of <asm/arch/mem.h> so don't
enforce it for new platforms.

This also means GPMC_MAX CS doesn't have to be defined
by platform code.

Define it locally here for now.

Signed-off-by: Roger Quadros <rogerq@kernel.org>
Reviewed-by: Michael Trimarchi <michael@amarulasolutions.com>
Link: https://lore.kernel.org/all/20221011115012.6181-2-rogerq@kernel.org
Signed-off-by: Dario Binacchi <dario.binacchi@amarulasolutions.com>
21 months agocmd: nand: Extend nand info to print ecc information
Michael Trimarchi [Thu, 22 Sep 2022 13:39:37 +0000 (15:39 +0200)]
cmd: nand: Extend nand info to print ecc information

Extract the information about ecc strength and ecc step size
from mtd controller. This information is usefull to check if
what we think as ecc is what we really configured.

Signed-off-by: Michael Trimarchi <michael@amarulasolutions.com>
Reviewed-by: Dario Binacchi <dario.binacchi@amarulasolutions.com>
Link: https://lore.kernel.org/all/20220922133937.277463-1-michael@amarulasolutions.com
Signed-off-by: Dario Binacchi <dario.binacchi@amarulasolutions.com>
21 months agoMerge branch '2022-12-09-platform-updates' into next
Tom Rini [Fri, 9 Dec 2022 22:50:51 +0000 (17:50 -0500)]
Merge branch '2022-12-09-platform-updates' into next

- Assorted TI platform updates
- Add DM_RTC callback functions, and a related x86 clean-up.

21 months agoconfigs: enable OSPI related configs in AM62x
Dhruva Gole [Thu, 27 Oct 2022 14:53:11 +0000 (20:23 +0530)]
configs: enable OSPI related configs in AM62x

Add am62x_evm_r5_defconfig for OSPI Flash support in R5 SPL
and am62x_evm_a53_defconfig for A53 SPL and U-Boot
support.
These configs enable OSPI Flash boot functionality in the board as well
as the usage of OSPI Flash from U-Boot.

Signed-off-by: Dhruva Gole <d-gole@ti.com>
21 months agoarm: dts: Add OSPI support for AM62-SK
Dhruva Gole [Thu, 27 Oct 2022 14:53:10 +0000 (20:23 +0530)]
arm: dts: Add OSPI support for AM62-SK

Add OSPI Support such that this device can boot up using OSPI Flash.
Also can use the flash for other purposes if required from uboot.

Signed-off-by: Dhruva Gole <d-gole@ti.com>
21 months agoarm: dts: k3-am62x: sync dt with linux kernel
Dhruva Gole [Thu, 27 Oct 2022 14:53:09 +0000 (20:23 +0530)]
arm: dts: k3-am62x: sync dt with linux kernel

Sync the DT Files with linux kernel (tag v6.0.3)

Signed-off-by: Dhruva Gole <d-gole@ti.com>
21 months agoconfigs: introduce configs for the am62a
Bryan Brattlof [Fri, 4 Nov 2022 00:13:58 +0000 (19:13 -0500)]
configs: introduce configs for the am62a

Introduce the minimum configs, only SD-MMC and UART boot related
settings, to serve as a good starting point for the am62a as we add more
functionality.

Signed-off-by: Bryan Brattlof <bb@ti.com>
[trini: Disable CONFIG_NET as it's not used, in both platforms]
Signed-off-by: Tom Rini <trini@konsulko.com>
21 months agoboard: ti: introduce the basic files needed to support the am62a
Bryan Brattlof [Fri, 4 Nov 2022 00:13:57 +0000 (19:13 -0500)]
board: ti: introduce the basic files needed to support the am62a

Introduce the bare minimum SD and UART support for the am62a sk.

Signed-off-by: Bryan Brattlof <bb@ti.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
21 months agoarm: mach-k3: am62a: introduce auto-generated SoC data
Bryan Brattlof [Fri, 4 Nov 2022 00:13:56 +0000 (19:13 -0500)]
arm: mach-k3: am62a: introduce auto-generated SoC data

Introduce the auto-generated clock tree and power domain data needed to
attach the am62a into the power-domain and clock frameworks of uboot

Signed-off-by: Bryan Brattlof <bb@ti.com>
21 months agoarm: mach-k3: introduce basic files to support the am62a
Bryan Brattlof [Fri, 4 Nov 2022 00:13:55 +0000 (19:13 -0500)]
arm: mach-k3: introduce basic files to support the am62a

Introduce the mach-k3 files needed to properly boot TI's am62a SoC
family of devices

Signed-off-by: Bryan Brattlof <bb@ti.com>
21 months agosoc: ti: k3-socinfo: add am62a SoC entry
Bryan Brattlof [Fri, 4 Nov 2022 00:13:54 +0000 (19:13 -0500)]
soc: ti: k3-socinfo: add am62a SoC entry

Add identification support for TI's am62ax family of SoCs

Signed-off-by: Bryan Brattlof <bb@ti.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
21 months agoram: k3-ddrss: add am62a controller support
Bryan Brattlof [Fri, 4 Nov 2022 00:13:53 +0000 (19:13 -0500)]
ram: k3-ddrss: add am62a controller support

TI's am62a family of SoCs uses a new 32bit DDR controller that shares
much of the same functionality with the existing am64 and j721e
controllers.

Select this controller by default when u-boot is build for the am62a

Signed-off-by: Bryan Brattlof <bb@ti.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
21 months agoarm: dts: introduce am62a7 u-boot dtbs
Bryan Brattlof [Fri, 4 Nov 2022 00:13:52 +0000 (19:13 -0500)]
arm: dts: introduce am62a7 u-boot dtbs

Introduce the base dts files needed for u-boot or to augment the
linux dtbs for use in the u-boot-spl and u-boot binaries

Signed-off-by: Vignesh Raghavendra <vigneshr@ti.com>
Signed-off-by: Bryan Brattlof <bb@ti.com>
21 months agoarm: dts: introduce am62a7 dtbs from linux kernel
Bryan Brattlof [Fri, 4 Nov 2022 00:13:51 +0000 (19:13 -0500)]
arm: dts: introduce am62a7 dtbs from linux kernel

Introduce the basic am62a7 SoC dtbs from the v6.1-rc3 tag of the linux
kernel along with the new am62a specific pinmux definition that we will
use to generate the dtbs for the u-boot-spl and u-boot binaries

Co-developed-by: Vignesh Raghavendra <vigneshr@ti.com>
Signed-off-by: Vignesh Raghavendra <vigneshr@ti.com>
Signed-off-by: Bryan Brattlof <bb@ti.com>
21 months agoram: k3-ddrss: add auto-generated macros for am62a support
Bryan Brattlof [Mon, 24 Oct 2022 21:53:28 +0000 (16:53 -0500)]
ram: k3-ddrss: add auto-generated macros for am62a support

The new 32bit DDR controller for TI's am62a family of SoCs shares much
of the same functionality with the existing 16bit (am64) and 32bit
(j721e) controllers, so this patch reorganizes the existing
auto-generated macros for the 16bit and 32bit controllers to make room
for the macros for the am62a's controller

This patch consists mostly of header/macro renames and additions with a
new Kconfig option (K3_AM62A_DDRSS) allowing us to select these new
macros during compilation.

Signed-off-by: Bryan Brattlof <bb@ti.com>
21 months agortc: Add fallbacks for dm functions
Sean Anderson [Tue, 22 Nov 2022 17:54:52 +0000 (12:54 -0500)]
rtc: Add fallbacks for dm functions

This adds fallbacks for the various dm_rtc_* functions. This allows
common code to use these functions without ifdefs.

Fixes: c8ce7ba87d1 ("misc: Add support for nvmem cells")
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Sean Anderson <sean.anderson@seco.com>
21 months agox86: fsp: Only compile fsp_save_s3_stack if (SPL_)DM_RTC is enabled
Sean Anderson [Tue, 22 Nov 2022 17:54:51 +0000 (12:54 -0500)]
x86: fsp: Only compile fsp_save_s3_stack if (SPL_)DM_RTC is enabled

This function calls rtc_write32, which has a different signature
depending on if (SPL_)DM_RTC is enabled or not. This could result in a
mismatch in SPL if DM_RTC was enabled but SPL_DM_RTC, as the non-DM
declaration would still be used in SPL even though the implementation
would be for non-DM_RTC. We are switching to the correct definitions in
the next commit, so this will become a compilation error. Since
fsp_save_s3_stack is not called from SPL, avoid compiling it if
(SPL_)DM_RTC is disabled.

Signed-off-by: Sean Anderson <sean.anderson@seco.com>
21 months agoNokia RX-51: Use ENTRY/ENDPROC for save_boot_params
Pali Rohár [Wed, 23 Nov 2022 23:45:00 +0000 (00:45 +0100)]
Nokia RX-51: Use ENTRY/ENDPROC for save_boot_params

ENTRY/ENDPROC macros from linux/linkage.h will make code more readable and
also will properly mark assembly symbol in ELF binary as function symbol.

Signed-off-by: Pali Rohár <pali@kernel.org>
21 months agoMerge branch '2022-12-08-assorted-tooling-and-fs-fixes' into next
Tom Rini [Thu, 8 Dec 2022 20:12:14 +0000 (15:12 -0500)]
Merge branch '2022-12-08-assorted-tooling-and-fs-fixes' into next

- Assorted fixes for squashfs, ext4, a number of host tools (including
  reworking u-boot-initial-env) and then a few unrelated code cleanups /
  fixes.

21 months agoMerge tag 'u-boot-stm32-20221207' of https://source.denx.de/u-boot/custodians/u-boot-stm
Tom Rini [Thu, 8 Dec 2022 16:25:08 +0000 (11:25 -0500)]
Merge tag 'u-boot-stm32-20221207' of https://source.denx.de/u-boot/custodians/u-boot-stm

- Drop MMCI interrupt-names in STM32H743, STM32MP15 and STM322MP13 DT

DHSOM:
  - Enable assorted ST specific commands
  - Add version variable
  - Add boot counter
STM32MP13:
  - Add sdmmc cd-gpios for STM32MP135F-DK
  - Add clock & reset support
STM32 ADC:
  - Split channel init into several routines
  - Add support of generic channels binding

21 months agoMerge https://source.denx.de/u-boot/custodians/u-boot-usb
Tom Rini [Thu, 8 Dec 2022 16:24:50 +0000 (11:24 -0500)]
Merge https://source.denx.de/u-boot/custodians/u-boot-usb

- DFU and RNDIS fixes

21 months agoata: ahci-pci: Replace magic constant by macro
Pali Rohár [Sun, 4 Dec 2022 12:31:08 +0000 (13:31 +0100)]
ata: ahci-pci: Replace magic constant by macro

Replace 0x1b21 by macro PCI_VENDOR_ID_ASMEDIA with the same value.

Signed-off-by: Pali Rohár <pali@kernel.org>
Reviewed-by: Simon Glass <sjg@chromium.org>
21 months agofs/squashfs: use lldiv function for math
Kasper Revsbech [Thu, 1 Dec 2022 15:30:32 +0000 (16:30 +0100)]
fs/squashfs: use lldiv function for math

When compling for x86:
u-boot/fs/squashfs/sqfs.c:90: undefined reference to `__udivmoddi4'

Signed-off-by: Kasper Revsbech <kasper.revsbech.ext@siemensgamesa.com>
Tested-by: Sean Nyekjaer <sean@geanix.com>
21 months agou-boot-initial-env: rework make target
Max Krummenacher [Mon, 28 Nov 2022 08:41:22 +0000 (09:41 +0100)]
u-boot-initial-env: rework make target

With LTO enabled the U-Boot initial environment is no longer stored
in an easy accessible section in env/common.o. I.e. the section name
changes from build to build, its content maybe compressed and it is
annotated with additional data.

Drop trying to read the initial env with elf tools from the compiler
specific object file in favour of adding and using a host tool with
the only functionality of printing the initial env to stdout.

See also:
https://lore.kernel.org/all/927b122e-1f62-e790-f5ca-30bae4332c77@foss.st.com/

Signed-off-by: Max Krummenacher <max.krummenacher@toradex.com>
Acked-by: Pali Rohár <pali@kernel.org>
Reviewed-by: Simon Glass <sjg@chromium.org>
21 months agofirmware: scmi: use protocol node name to bind the scmi regulator driver
Patrick Delaunay [Fri, 25 Nov 2022 11:56:29 +0000 (12:56 +0100)]
firmware: scmi: use protocol node name to bind the scmi regulator driver

In scmi firmware driver, it is better to bind the scmi protocol driver
"scmi_voltage_domain" with the node name of the protocol 17 and not
the sub-node named "regulator", because is a fixed string which doesn't
provide information and because it is not aligned with the other scmi
protocol nodes.

For example on stm32mp135f-dk board with device tree in stm32mp131.dtsi

scmi: scmi {
compatible = "linaro,scmi-optee";
#address-cells = <1>;
#size-cells = <0>;
linaro,optee-channel-id = <0>;
shmem = <&scmi_shm>;
scmi_clk: protocol@14 {
reg = <0x14>;
#clock-cells = <1>;
};
scmi_reset: protocol@16 {
reg = <0x16>;
#reset-cells = <1>;
};
scmi_voltd: protocol@17 {
reg = <0x17>;
scmi_regu: regulators {
#address-cells = <1>;
#size-cells = <0>;
scmi_reg11: voltd-reg11 {
reg = <VOLTD_SCMI_REG11>;
regulator-name = "reg11";
};
scmi_reg18: voltd-reg18 {
reg = <VOLTD_SCMI_REG18>;
regulator-name = "reg18";
};
scmi_usb33: voltd-usb33 {
reg = <VOLTD_SCMI_USB33>;
regulator-name = "usb33";
};
};
};
};

Before the patch:

> dm tree

 scmi_agent    0  [ + ]   scmi-over-optee       |-- scmi
 clk           1  [ + ]   scmi_clk              |   |-- protocol@14
 ...
 reset         1  [   ]   scmi_reset_domain     |   |-- protocol@16
 nop           2  [ + ]   scmi_voltage_domain   |   `-- regulators
 regulator     0  [ + ]   scmi_regulator        |       |-- voltd-reg11
 regulator     1  [ + ]   scmi_regulator        |       |-- voltd-reg18
 regulator     2  [ + ]   scmi_regulator        |       |-- voltd-usb33
 ...

after the patch:

> dm tree

 scmi_agent    0  [ + ]   scmi-over-optee       |-- scmi
 clk           1  [ + ]   scmi_clk              |   |-- protocol@14
 ...
 reset         1  [   ]   scmi_reset_domain     |   |-- protocol@16
 nop           2  [ + ]   scmi_voltage_domain   |   `-- protocol@17
 regulator     0  [ + ]   scmi_regulator        |       |-- voltd-reg11
 regulator     1  [ + ]   scmi_regulator        |       |-- voltd-reg18
 regulator     2  [ + ]   scmi_regulator        |       |-- voltd-usb33
 ...

Signed-off-by: Patrick Delaunay <patrick.delaunay@foss.st.com>
21 months agotools: mkimage: add new image type "fdt_legacy"
Marc Kleine-Budde [Wed, 23 Nov 2022 11:55:33 +0000 (12:55 +0100)]
tools: mkimage: add new image type "fdt_legacy"

If the user select the image type "flat_dt" a FIT image will be build.
This breaks the legacy use case of putting a Flat Device Tree into a
legacy u-boot image.

Add a new image type "fdt_legacy" to build a legacy u-boot image
with a "flat_dt" type.

Link: https://lore.kernel.org/all/20221028155205.ojw6tcso2fofgnhm@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Reviewed-by: Sean Anderson <seanga2@gmail.com>
21 months agotools: mkimage: don't print error message "Success" in case of failure
Marc Kleine-Budde [Wed, 23 Nov 2022 11:55:32 +0000 (12:55 +0100)]
tools: mkimage: don't print error message "Success" in case of failure

In case there's no struct image_type_params::set_header callback, no
"errno" will be set. Don't fail with an error message, followed by
"Success". Remove the printing of the human readable "errno" value.

Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Reviewed-by: Simon Glass <sjg@chromium.org>
21 months agotools: fdtgrep: Fix handle leak
Mikhail Ilin [Wed, 23 Nov 2022 11:31:03 +0000 (14:31 +0300)]
tools: fdtgrep: Fix handle leak

The handle "fd" was created in fdtgrep.c:708 by calling the
 "open" function and is lost in fdtgrep.c:716 and fdtgrep.c:723.
 Close file descriptor 'fd' before exiting with an error from function
 utilfdt_read_err_len(const char *filename, char **buffp, off_t *len).

Fixes: 1043d0a0296a ("fdt: Add fdtgrep tool")
Signed-off-by: Mikhail Ilin <ilin.mikhail.ol@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
21 months agotools: mkimage: Fix nullptr at strchr()
Mikhail Ilin [Wed, 23 Nov 2022 09:39:36 +0000 (12:39 +0300)]
tools: mkimage: Fix nullptr at strchr()

The copy_datafile(ifd, params.datafile) function has been
 implemented to copy data by reducing the number of lines in the main
 function.

Signed-off-by: Mikhail Ilin <ilin.mikhail.ol@gmail.com>
21 months agofs: ext4: Fix free(NULL)
Mikhail Ilin [Tue, 22 Nov 2022 08:00:55 +0000 (11:00 +0300)]
fs: ext4: Fix free(NULL)

The 'depth_dirname', 'ptr', 'parent_inode' and 'first_inode' pointers
may be null. Thus, it is necessary to check them before using free() to
avoid free(NULL) cases.

Fixes: 934b14f2bb30 ("ext4: free allocations by parse_path()")
Signed-off-by: Mikhail Ilin <ilin.mikhail.ol@gmail.com>
21 months agotool: ifwitool: Fix buffer overflow
Mikhail Ilin [Fri, 18 Nov 2022 11:50:24 +0000 (14:50 +0300)]
tool: ifwitool: Fix buffer overflow

An incorrect 1st parameter is passed to the fix_member()
 function. Should use a pointer to the beginning of the parent structure
 (bpdt or subpart_dir, because are boxed), not to their fields. Otherwise,
 this leads to an overrun of the structure boundary, since in the
 fix_member() function, an 'offset' is made, relative to the 1st argument,
 which itself is an 'offset' from the beginning of the structure.

Signed-off-by: Mikhail Ilin <ilin.mikhail.ol@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
21 months agotool: ifwitool: The function localtime() can return NULL.
Mikhail Ilin [Fri, 18 Nov 2022 08:50:02 +0000 (11:50 +0300)]
tool: ifwitool: The function localtime() can return NULL.

This will cause the local_time pointer is passed as the 4th argument
 to function strftime() to also point to NULL. This result in a
 segmentation fault. Thus, it's necessary to add a check of the local_time
 pointer to NULL.

Found by Linux Verification Center (linuxtesting.org) with SVACE.

Signed-off-by: Mikhail Ilin <ilin.mikhail.ol@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
21 months agotools: env: fw_env: Fix unused-result warning
Jaehoon Chung [Thu, 13 Oct 2022 08:41:21 +0000 (17:41 +0900)]
tools: env: fw_env: Fix unused-result warning

Fix unused-result warning about fread.

tools/env/fw_env.c: In function ‘find_nvmem_device’:
tools/env/fw_env.c:1751:3: warning: ignoring return value of ‘fread’, declared with attribute warn_unused_result [-Wunused-result]
 1751 |   fread(buf, sizeof(buf), 1, fp);
      |   ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Signed-off-by: Jaehoon Chung <jh80.chung@samsung.com>
21 months agousb: gadget: rndis: Prevent InformationBufferOffset manipulation
Szymon Heidrich [Mon, 5 Dec 2022 09:28:23 +0000 (10:28 +0100)]
usb: gadget: rndis: Prevent InformationBufferOffset manipulation

Prevent access to arbitrary memory locations in gen_ndis_set_resp
via manipulation of buf->InformationBufferOffset. Original
implementation permits manipulation of InformationBufferOffset to
exploit OID_GEN_CURRENT_PACKET_FILTER to set arbitrary memory contents
within a 32byte offset as the devices packet filter. The packet filter
value may be next retrieved using gen_ndis_query_resp so it is possible
to extract specific memory regions two bytes a time.

The rndis_query_response was not modified as neither the buffer offset
nor length passed to gen_ndis_query_resp is used.

Signed-off-by: Szymon Heidrich <szymon.heidrich@gmail.com>
21 months agodfu: Make DFU virtual backend SPL friendly
Patrick Delaunay [Wed, 30 Nov 2022 10:42:02 +0000 (11:42 +0100)]
dfu: Make DFU virtual backend SPL friendly

Define stub for dfu_*_virt function in SPL, because
CONFIG_SPL_DFU_VIRT is not defined.

This patch avoids compilation issue in dfu_fill_entity() when
CONFIG_SPL_DFU is activated because the dfu_fill_entity_virt()
function is not available.

Fixes: ec44cace4b8d2 ("dfu: add DFU virtual backend")
Reported-by: Marek Vasut <marex@denx.de>
Signed-off-by: Patrick Delaunay <patrick.delaunay@foss.st.com>
Reviewed-by: Marek Vasut <marex@denx.de>
Tested-by: Marek Vasut <marex@denx.de>
21 months agousb: gadget: dfu: Fix check of transfer direction
Hugo SIMELIERE [Wed, 30 Nov 2022 08:29:16 +0000 (09:29 +0100)]
usb: gadget: dfu: Fix check of transfer direction

Commit fbce985e28eaca3af82afecc11961aadaf971a7e to fix CVE-2022-2347
blocks DFU usb requests.
The verification of the transfer direction was done by an equality
but it is a bit mask.

Signed-off-by: Hugo SIMELIERE <hsimeliere.opensource@witekio.com>
Reviewed-by: Fabio Estevam <festevam@denx.de>
Reviewed-by: Sultan Qasim Khan <sultan.qasimkhan@nccgroup.com>
Reviewed-by: Marek Vasut <marex@denx.de>
Tested-by: Marek Vasut <marex@denx.de>
21 months agoMerge https://source.denx.de/u-boot/custodians/u-boot-riscv
Tom Rini [Thu, 8 Dec 2022 13:28:14 +0000 (08:28 -0500)]
Merge https://source.denx.de/u-boot/custodians/u-boot-riscv

- Kautuk's semihosting patch:
  move semihosting library from arm directory to common place and add
  RISC-V support
- Zong's Kconfig patch:
  use "imply" instead of "select" to allow user to decide if
  SPL_SEPARATE_BSS should be selected

21 months agoMerge tag 'u-boot-at91-fixes-2023.01-b' of https://source.denx.de/u-boot/custodians...
Tom Rini [Thu, 8 Dec 2022 13:27:50 +0000 (08:27 -0500)]
Merge tag 'u-boot-at91-fixes-2023.01-b' of https://source.denx.de/u-boot/custodians/u-boot-at91

Second set of u-boot-at91 fixes for the 2023.01 cycle:

This is a single tiny fix that allows the correct name for one pin on
sama7g5 device. People with DT coming from Linux will have build errors
without this if they add NAND device.

21 months agoriscv: use imply instead of select for SPL_SEPARATE_BSS
Zong Li [Wed, 16 Nov 2022 07:08:39 +0000 (07:08 +0000)]
riscv: use imply instead of select for SPL_SEPARATE_BSS

Use imply instead of select, then it can still be disabled by
board-specific defconfig, or be set to n manually.

Signed-off-by: Zong Li <zong.li@sifive.com>
Reviewed-by: Rick Chen <rick@andestech.com>
Reviewed-by: Bin Meng <bmeng@tinylab.org>