Converts qemu x86 machines to boot using distro_config. The intent is to
allow u-boot in qemu to be maximally compatible with many boot methods
without having to change the config. Previously, u-boot would only boot
in a very limited set of circumstances where there was a /boot/vmlinuz
on scsi 0:3 with no ramdisk.
Signed-off-by: Joshua Watt <JPEWhacker@gmail.com> Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
[bmeng: convert doc updates to reST] Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Alex Deymo [Mon, 5 Aug 2019 20:24:45 +0000 (22:24 +0200)]
Update include/android_image.h from AOSP
This takes the latest changes from AOSP from the file
include/bootimg/bootimg.h from the repository:
https://android.googlesource.com/platform/system/tools/mkbootimg
and update the U-Boot version with the latest changes.
This file keeps the changes from AOSP to a minimum:
* Comments were converted from C++ to C style.
* Code inside __cplusplus #ifdef blocks were removed.
* C++11 struct extensions replaced with a single struct.
Signed-off-by: Alex Deymo <deymo@google.com> Reviewed-by: Sam Protsenko <semen.protsenko@linaro.org>
Julius Werner [Fri, 2 Aug 2019 22:52:28 +0000 (15:52 -0700)]
fit: Do not automatically decompress ramdisk images
The Linux ramdisk should always be decompressed by the kernel itself,
not by U-Boot. Therefore, the 'compression' node in the FIT image should
always be set to "none" for ramdisk images, since the only point of
using that node is if you want U-Boot to do the decompression itself.
Yet some systems populate the node to the compression algorithm used by
the kernel instead. This used to be ignored, but now that we support
decompression of all image types it becomes a problem. Since ramdisks
should never be decompressed by U-Boot anyway, this patch adds a special
exception for them to avoid these issues. Still, setting the
'compression' node like that is wrong in the first place, so we still
want to print out a warning so that third-party distributions doing this
can notice and fix it.
Signed-off-by: Julius Werner <jwerner@chromium.org> Reviewed-by: Heiko Schocher <hs@denx.de> Tested-by: Heiko Schocher <hs@denx.de> Reviewed-by: Simon Goldschmidt <simon.k.r.goldschmidt@gmail.com>
console: execute flush on uart when silent is removed
Avoid to flush buffer when silent console is activated as the
console can be reactivate later, after relocation, when the env will
be updated with the saved one.
Solve issue (missing beginning of U-Boot trace) when:
- CONFIG_SILENT_CONSOLE is activated
- silent=1 is defined in default environment (CONFIG_EXTRA_ENV_SETTINGS)
- silent is removed in saved environment with:
> env delete silent; env save
Only functional when PRE_CONSOLE_BUFFER is activated.
Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com> Signed-off-by: Patrice Chotard <patrice.chotard@st.com>
Update the "silent" property with the variable "silent" get from
saved environment, it solves the issue when:
- CONFIG_SILENT_CONSOLE and CONFIG_SYS_CONSOLE_IS_IN_ENV are activated
- silent is not defined in default environment
- silent is requested in saved environment with:
> env set silent 1; env save
On next reboot the console is not disabled as expected after relocation
and the environment load from flash (the callback is not called when
the INSERT is requested in the created hash table)
Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com> Signed-off-by: Patrice Chotard <patrice.chotard@st.com>
Sam Protsenko [Wed, 31 Jul 2019 16:59:09 +0000 (19:59 +0300)]
cmd: avb: Fix compiler warnings
When building U-Boot with AVB enabled, compiler shows next warnings:
cmd/avb.c: In function 'do_avb_read_pvalue':
cmd/avb.c:371:18: warning: format '%ld' expects argument of type
'long int', but argument 2 has type 'size_t'
{aka 'unsigned int'} [-Wformat=]
printf("Read %ld bytes, value = %s\n", bytes_read,
~~^ ~~~~~~~~~~
%d
cmd/avb.c: In function 'do_avb_write_pvalue':
cmd/avb.c:404:19: warning: format '%ld' expects argument of type
'long int', but argument 2 has type '__kernel_size_t'
{aka 'unsigned int'} [-Wformat=]
printf("Wrote %ld bytes\n", strlen(value) + 1);
~~^ ~~~~~~~~~~~~~~~~~
%d
Fix those by using "%zu" specified.
Signed-off-by: Sam Protsenko <semen.protsenko@linaro.org> Reviewed-by: Igor Opaniuk <igor.opaniuk@gmail.com>
common/avb_verify.c: In function 'read_persistent_value':
common/avb_verify.c:867:6: warning: implicit declaration of function
'get_open_session'
common/avb_verify.c:870:45: error: 'struct AvbOpsData' has no member
named 'tee'
common/avb_verify.c:894:7: warning: implicit declaration of function
'invoke_func'
common/avb_verify.c: In function 'write_persistent_value':
common/avb_verify.c:931:45: error: 'struct AvbOpsData' has no member
named 'tee'
Guard read_persistent_value() and write_persistent_value() functions
by checking if CONFIG_OPTEE_TA_AVB is enabled (as those are only used in
that case) to fix the build with mentioned configuration.
Signed-off-by: Sam Protsenko <semen.protsenko@linaro.org> Reviewed-by: Igor Opaniuk <igor.opaniuk@gmail.com>
Ramon Fried [Wed, 31 Jul 2019 08:04:26 +0000 (11:04 +0300)]
pci_ep: Fix Coverity warning
Fix the following Coverity warning:
CID 244086: Incorrect expression (BAD_COMPARE)
Comparing pointer "ep_bar" against NULL using anything besides == or
is likely to be incorrect.
autoboot: fix bug using with CAAM and AUTOBOOT_ENCRYPTION
if CONFIG_AUTOBOOT_KEYED, CONFIG_AUTOBOOT_ENCRYPTION and
CONFIG_AUTOBOOT_STOP_STR_SHA256 are enabled in conjunction
with CONFIG_SHA_HW_ACCEL and CONFIG_FSL_CAAM, we get the
Error when pressing a key while waiting for bootdelay:
Error: Address arguments are not aligned
CAAM was not setup properly or it is faulty
Reason is, that used variables are not cache aligned,
so malloc this variables cache aligned.
Probably this is also a bugfix for other hw accelerators
than CAAM.
Stefan Roese [Wed, 3 Jul 2019 05:22:20 +0000 (07:22 +0200)]
watchdog: mtk_wdt: Cosmetic cleanup of latest changes
This patch cleans up some coding style related issues in the mtk_wtd
driver to make this driver comply again with the U-Boot coding style
standards.
The only minimal functional change is that the timeout parameter is now
passed in (u64) instead of (unsigned int) from mtk_wdt_start() to
mtk_wdt_set_timeout(), preserving the original value.
Signed-off-by: Stefan Roese <sr@denx.de> Cc: Shannon Barber <sbarber@dataspeedinc.com> Cc: Ryder Lee <ryder.lee@mediatek.com> Cc: Matthias Brugger <matthias.bgg@gmail.com> Cc: Frank Wunderlich <frank-w@public-files.de> Reviewed-by: Matthias Brugger <matthias.bgg@gmail.com>
Tom Rini [Sun, 4 Aug 2019 23:39:54 +0000 (19:39 -0400)]
Merge tag 'efi-2019-10-rc2' of https://gitlab.denx.de/u-boot/custodians/u-boot-efi
Pull request for UEFI sub-system for v2019.10-rc2
With this pull request a workaround for GRUB on 32bit ARM is re-enabled
and made customizable. Without the patch booting on ARM 32bit with GRUB
prior to version 2.04 or with a cache which is not managed via CP15
fails.
Further work will be needed to achieve a UEFI compliant cache handling.
According to the UEFI spec all caches except those that cannot be
managed via CP15 should be enabled.
An implementation of the ConvertPointer() runtime service is provided.
efi_crt0 is always rebuild to avoid having to call 'make mrproper' when
switching architectures.
As part of copying the logic from Travis to GitLab I kept the TOOLCHAIN
variable. However we don't use that now as the Docker container already
has all toolchains so we don't need to do any downloading. Remove this
variable.
Simon Glass [Sun, 21 Jul 2019 02:51:21 +0000 (20:51 -0600)]
autoboot: Rename CONFIG_MENUKEY to CONFIG_AUTOBOOT_MENUKEY
Since this is part of the autoboot functionality, it makes sense to name
it with an AUTOBOOT prefix. No mainline boards use it so this should be
safe, and downstream boards will need to adjust.
Since this option is just an integer value, it really needs another option
to control whether the feature is enabled or not. Add a new
CONFIG_USE_AUTOBOOT_MENUKEY for that. This fits better with how things are
done with Kconfig, avoiding the need to use a specific value to disable
the feature.
Simon Glass [Sun, 21 Jul 2019 02:51:20 +0000 (20:51 -0600)]
autoboot: Drop unused CONFIG_MENUPROMPT
This is not defined by any board. We could use CONFIG_AUTOBOOT_PROMPT
instead perhaps, but this depends on CONFIG_AUTOBOOT_KEYED which is not
used for the single-key case.
Simon Glass [Sun, 21 Jul 2019 02:51:19 +0000 (20:51 -0600)]
autoboot: Drop #ifdef CONFIG_AUTOBOOT_KEYED
At present we have two functions named __autoboot() which do different
things. This is confusing. Fix it by using if() instead of #ifdef for
selecting the functions, and renaming them to meaningful names.
Simon Glass [Sun, 21 Jul 2019 02:51:15 +0000 (20:51 -0600)]
autoboot: Use CONFIG_AUTOBOOT_STOP_STR_SHA256 indirectly
This CONFIG option is only present if CONFIG_AUTOBOOT_ENCRYPTION is
enabled so it cannot be used in code without that #ifdef. But we want to
reduce the use of #ifdef in this file and in particular to avoid having
two different functions both named passwd_abort() but which do different
things.
In preparation for this, create an intermediate value which is set to an
empty string if there is no value for CONFIG_AUTOBOOT_STOP_STR_SHA256.
Simon Glass [Sun, 21 Jul 2019 02:51:11 +0000 (20:51 -0600)]
main: Use conditional run_preboot_environment_command()
The function name indicates that it does something, but its entire
operation is actually condition on a CONFIG. Move the condition outside
the function so this is clearer, and use if() instead of #ifdef, like the
reset of the file.
Tom Rini [Wed, 31 Jul 2019 20:06:24 +0000 (16:06 -0400)]
Merge tag 'u-boot-amlogic-20190731' of https://gitlab.denx.de/u-boot/custodians/u-boot-amlogic
- sync Amlogic G12A DT with linux 5.3-rc1
- add support for 4GiB DRAM memory
- add support for Amlogic G12B based Odroid-N2
- small duplicate logic fix for gxbb clock driver
The above patch was designed to shrink code by only pin-muxing items
needed for SPL in SPL and relying on driver model or SPL to mux other
items. Unfortunately, da850evm_direct_nor doesn't use SPL so items
that were only muxed during SPL are not muxed causing the board
to no longer boot.
Adam Ford [Tue, 30 Jul 2019 13:58:58 +0000 (08:58 -0500)]
ARM: da850-evm: Remove references to CONFIG_DA850_AM18X_EVM
With the removal of da850_am18xxevm, there is at least one
whitelisted CONFIG option that can be deleted. This patch
removes CONFIG_DA850_AM18X_EVM since it's not required any more
Adam Ford [Tue, 30 Jul 2019 13:32:37 +0000 (08:32 -0500)]
ARM: dts: da850-evm: Fix MDIO pinmux
In attempts to speed up SPL and reduce size, the MDIO pin muxing
was inadvertently affected. Since the ethernet driver will setup
the pin muxing when ethernet is loaded, this patch will also
pinmux the MDIO pins at the same time. Once an DM compatible
MDIO driver is available, this can be removed.
Fixes: 877ab2423bc2 ("ARM: davinci: da850: Manual pinmux only
when PINCTRL not available")
configs: am335x_boneblack_vboot_defconfig: Add DM for SPI and Flash devices
This patch adds SPI and SPI_FLASH DM support for verified boot on
TI AM335 chipsets. The following compile warning is removed:
===================== WARNING ======================
This board does not use CONFIG_DM_SPI_FLASH. Please update
the board to use CONFIG_SPI_FLASH before the v2019.07 release.
Failure to update by the deadline may result in board removal.
See doc/driver-model/MIGRATION.txt for more info.
====================================================
Built and tested on AM335x device (BeagleboneBlack).
configs: am335x_boneblack_vboot_defconfig: Fix regression by enabling BLK and DM support, disable in SPL
This patch adds BLK and DM support for verified boot on TI AM335x
chipsets. The following compile warnings are removed:
===================== WARNING ======================
This board does not use CONFIG_DM_MMC. Please update
the board to use CONFIG_DM_MMC before the v2019.04 release.
Failure to update by the deadline may result in board removal.
See doc/driver-model/MIGRATION.txt for more info.
====================================================
===================== WARNING ======================
This board does not use CONFIG_DM_USB. Please update
the board to use CONFIG_DM_USB before the v2019.07 release.
Failure to update by the deadline may result in board removal.
See doc/driver-model/MIGRATION.txt for more info.
====================================================
BLK and DM_MMC are enabled by default in SPL as well, which is
making the build to break with an overflow(spl image doesn't
fit into SRAM because of size constraints).
LD spl/drivers/built-in.o
LD spl/u-boot-spl
arm-linux-ld.bfd: u-boot-spl section .u_boot_list will not fit in region .sram
arm-linux-ld.bfd: region .sram overflowed by 116 bytes
make[1]: *** [spl/u-boot-spl] Error 1
make: *** [spl/u-boot-spl] Error 2
For the above reason BLK and DM_MMC is disabled in SPL.
Built and tested on AM335x device (BeagleboneBlack).
The K3 System Controller driver is used for loading and starting
the System Firmware, and is used only on R5 SPL. It need not be
enabled and built for the A72 U-Boot and SPL, so disable it from
the j721e_evm_a72 defconfig.
While at this, also remove the unneeded CONFIG_SPL_REMOTEPROC and
CONFIG_CMD_REMOTEPROC as no remoteprocs are now loaded from A72 SPL.
The K3 System Controller driver is used for loading and starting
the System Firmware, and is used only on R5 SPL. It need not be
enabled and built for the A53 U-Boot and SPL, so disable it from
both the GP and HS AM65x A53 defconfigs.
While at this, also remove the unneeded CONFIG_SPL_REMOTEPROC and
CONFIG_CMD_REMOTEPROC as no remoteprocs are now loaded from A53 SPL.
Signed-off-by: Suman Anna <s-anna@ti.com> Reviewed-by: Lokesh Vutla <lokeshvutla@ti.com>
The commit 355be915ed08 ("arm: dts: k3-am654: Update power-domains
property for each node") has updated the power-domain cells value
and updated power-domains property in various existing dts nodes but
missed updating the cpsw_nuss node. This results in the following
build warning, fix this.
arch/arm/dts/k3-am654-base-board.dtb: Warning (power_domains_property): /interconnect@100000/interconnect@28380000/cpsw_nuss@046000000:power-domains: property size (8) too small for cell size 2
arch/arm/dts/k3-am654-r5-base-board.dtb: Warning (power_domains_property): /interconnect@100000/interconnect@28380000/cpsw_nuss@046000000:power-domains: property size (8) too small for cell size 2
Fixes: 355be915ed08 ("arm: dts: k3-am654: Update power-domains property for each node") Signed-off-by: Suman Anna <s-anna@ti.com> Reviewed-by: Lokesh Vutla <lokeshvutla@ti.com>
Adam Ford [Fri, 26 Jul 2019 22:53:45 +0000 (17:53 -0500)]
ARM: da850evm: Split MTDPARTS into SPL and u-boot
The MTDPARTS currently lists just u-boot.ais as 512k in size.
This works when loading the ais file via serial port, but if one
wanted to update just the u-boot portion, it's not really possible.
This patch splits the MTDPARTS into a 32k SPL partiion and a 480k
u-boot partition which allows u-boot.img to be burned to the u-boot
partition. The remaining partitions are left with the same sizes
and offsets to not break backwards compatibility.
It is essential to have an access to BCB area of 'misc' partition on
Android devices [1]. For BeagleBoard X15 the 'bcb' command will be
further used for reboot reason implementation and booting to recovery.
It can be also used for debugging reasons, like checking RescueParty
messages in BCB area.
[1] doc/android/bcb.txt
Signed-off-by: Sam Protsenko <semen.protsenko@linaro.org>
Sam Protsenko [Thu, 25 Jul 2019 17:11:53 +0000 (20:11 +0300)]
omap: Correct the fastboot product var
"fastboot flashall" expects "fastboot getvar product" value to be one of
values provided in android-info.txt file (in AOSP), from "require
board=" list. Before this patch, "am57xx" is returned for all AM57xx
based boards, as it's set in $board env var from SYS_BOARD in
board/ti/am57xx/Kconfig file, which is used for default implementation
of "fastboot getvar product".
In order to fix that inconsistency, let's do next:
1. In U-Boot: override fastboot.product, reusing the value from
$board_name
2. In AOSP: provide values for all AM57xx boards we can use to
device/ti/beagle_x15/board-info.txt file
This way requirements check in "fastboot flashall" will work as
expected, verifying that user tries to flash images to the board which
those images were built for.
Signed-off-by: Sam Protsenko <semen.protsenko@linaro.org> Acked-by: Andrew F. Davis <afd@ti.com>
Sam Protsenko [Thu, 25 Jul 2019 13:20:41 +0000 (16:20 +0300)]
env: ti: boot: Handle reboot reason from BCB
In case of Android boot, reboot reason can be written into BCB (usually
it's an area in 'misc' partition). U-Boot then can obtain that reboot
reason from BCB and handle it accordingly to achieve correct Android
boot flow, like it was suggested in [1]:
- if it's empty: perform normal Android boot from eMMC
- if it contains "bootonce-bootloader": get into fastboot mode
- if it contains "boot-recovery": perform recovery boot
The latter is not implemented yet, as it depends on some features that
are not implemented on TI platforms yet (in AOSP and in U-Boot).
First thing we check in meson_clk_set_rate_by_id() is current_rate == rate.
There is not need to check it again.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Acked-by: Neil Armstrong <narmstrong@baylibre.com> Signed-off-by: Neil Armstrong <narmstrong@baylibre.com>
The comment about init op being NULL used to be next to the NULL check
code. Commit 8ca51e51c182 ("dm: mmc: Add a way to use driver model for
MMC operations") separated the comment from the code. Put them back
together.
Fixes: 8ca51e51c182 ("dm: mmc: Add a way to use driver model for MMC operations") Signed-off-by: Baruch Siach <baruch@tkos.co.il> Reviewed-by: Peng Fan <peng.fan@nxp.com>
mmc: mv_sdhci: fix uninitialized pointer deref on probe
Since commit 3d296365e4e8 ("mmc: sdhci: Add support for
sdhci-caps-mask") sdhci_setup_cfg() expects a valid sdhci_host mmc
field. Move the mmc field initialization before sdhci_setup_cfg()
call to avoid crash on mmc pointer dereference.
Fixes: 3d296365e4e8 ("mmc: sdhci: Add support for sdhci-caps-mask") Cc: Faiz Abbas <faiz_abbas@ti.com> Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Add a standalone MMC block write test. This allows direct testing of MMC
access rather than relying on doing so as a side-effect of e.g. DFU or
UMS testing, which may not be enabled on all platforms.
Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com> Reviewed-by: Stephen Warren <swarren@nvidia.com>
Peng Fan [Wed, 31 Jul 2019 07:01:49 +0000 (07:01 +0000)]
dm: clk: ignore default settings when node not valid
When the device not binded with a node, we need ignore
the parents and rate settings.
Cc: Simon Glass <sjg@chromium.org> Cc: Jagan Teki <jagan@amarulasolutions.com> Cc: Philipp Tomsich <philipp.tomsich@theobroma-systems.com> Cc: Neil Armstrong <narmstrong@baylibre.com> Cc: Andreas Dannenberg <dannenberg@ti.com> Signed-off-by: Peng Fan <peng.fan@nxp.com>
Peng Fan [Wed, 31 Jul 2019 07:01:23 +0000 (07:01 +0000)]
clk: introduce clk_dev_binded
When support Clock Common Framework, U-Boot use dev for
clk tree information, there is no clk->parent. When
support composite clk, it contains mux/gate/divider,
but the mux/gate/divider is not binded with device.
So we could not use dev_get_uclass_priv to get the correct
clk_mux/gate/divider. So add clk_dev_binded to let
choose the correct method.
Tom Rini [Tue, 30 Jul 2019 23:19:54 +0000 (19:19 -0400)]
Merge tag 'video-for-2019.10-rc1' of https://gitlab.denx.de/u-boot/custodians/u-boot-video
- fix EDID mode filtering
- extend mxc_ipuv3_fb to enable backlight/display
- include fb_base in global_data for DM_VIDEO
- show frame buffer address via board info
as used to be with legacy VIDEO support
We read the address map before assigning the memory for the pages that will
be mapped to virtual addresses. So these pages will overlap with the entry
for EFI_CONVENTIONAL_MEMORY. We have to ensure that every page is described
at most once in the map.
Remove EFI_CONVENTIONAL_MEMORY from the map that we pass to
SetVirtualAddressMap().
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
When changing the architecture without calling 'make clean' a subsequent
make fails with
lib/efi_loader/efi_crt0.o: file not recognized: File format not recognized
Force efi_crt0.o to be always rebuild.
Reported-by: Kever Yang <kever.yang@rock-chips.com> Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>