If a memory bank is not EFI_PAGE_SIZE aligned efi_add_known_memory() the
number of memory pages may be incorrectly calculated.
We have to round up the start address and to round down the end address
to determine which complete pages are provided by the memory bank.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Signed-off-by: Alexander Graf <agraf@suse.de>
The "Devicetree Specification 0.2" does not prescribe that memory
reservations must be EFI page aligned. So let's not make such an
assumption in our code.
Do not carve out the pages for the device tree. This memory area is
already marked as EFI_RUNTIME_SERVICES_DATA.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Signed-off-by: Alexander Graf <agraf@suse.de>
In copy_fdt() we allocate EFI pages for the fdt plus extra 12 KiB as
EFI_RUNTIME_SERVICES_DATA. Afterwards in efi_install_fdt() we overwrite
part of this memory allocation by marking it as EFI_BOOT_SERVICES_DATA.
Remove the code marking the fdt as EFI_BOOT_SERVICES_DATA.
Cf. commit 17ff6f02f5 ("efi_loader: store DT in EFI_RUNTIME_SERVICES_DATA
memory")
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Signed-off-by: Alexander Graf <agraf@suse.de>
Using %zu for efi_intn_t (ssize_t) creates a build warning. Anyway %zu is
not supported by efi_st_error(). So let's convert to int.
Our implementation of StriColl() only returns -1, 0, or 1.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Signed-off-by: Alexander Graf <agraf@suse.de>
When the max_addr parameter of efi_find_free_memory() is within bounds
of an existing map and fits the reservation, we just return that address
as allocation value.
That breaks however if max_addr is not page aligned. So ensure that it
always comes to us page aligned, simplifying the allocation logic.
Without this, I've seen breakage where we were allocating pages at -1U
(32bit) which fits into a region that spans beyond 0x100000000. In that
case, we would return 0xffffffff as a valid memory allocation, although
we usually do guarantee they are all page aligned.
Fix this by aligning the max address argument always.
Signed-off-by: Alexander Graf <agraf@suse.de>
With RELA absolute relocations, the relocation target contains our link
offset which we need to remove from the equation again. We did this
properly in the relative relocation path, but not in the absolute one.
So let's do this for the absolute one as well. That way, u-boot can have
a TEXT_OFFSET of != 0 and still relocate itself properly.
This fixes a bug where efi_loader did not work on the RISC-V QEMU port.
With this patch, I can successfully run UEFI applications on the RISC-V
QEMU port.
Reported-by: Auer, Lukas <lukas.auer@aisec.fraunhofer.de>
Signed-off-by: Alexander Graf <agraf@suse.de>
Tested-by: Bin Meng <bmeng.cn@gmail.com>
Tested-by: Lukas Auer <lukas.auer@aisec.fraunhofer.de>
When you support distro boot and you support USB, you usually want to
also support booting from USB storage.
Reflect that in the Kconfig, so that we don't have to explicitly add
USB storage support to every defconfig individually.
Reported-by: AKASHI Takahiro <takahiro.akashi@linaro.org>
Signed-off-by: Alexander Graf <agraf@suse.de>
The USB storage driver does not compile when CONFIG_BLK is set,
but DM_USB is not set, as we're missing the DM device links for
CONFIG_BLK enabled code paths.
So far it looks like nobody fell into this trap, because no board
enabled CONFIG_BLK and CONFIG_USB_STORAGE while not enabling
CONFIG_DM_USB, but we should still reflect that dependency properly
in Kconfig so that implicit enabling of CONFIG_USB_STORAGE works.
Signed-off-by: Alexander Graf <agraf@suse.de>
Linker generated arrays may be stored in code sections of memory that are
not writable. So let's allocate setup_ok as an array at runtime.
This avoids an illegal memory access observed in the sandbox.
Reported-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Signed-off-by: Alexander Graf <agraf@suse.de>
When an operating system started via bootefi tries to reset or power off
this is done by calling the EFI runtime ResetSystem(). On most ARMv8 system
the actual reset relies on PSCI. Depending on whether the PSCI firmware
resides the hypervisor (EL2) or in the secure monitor (EL3) either an HVC
or an SMC command has to be issued.
The current implementation always uses SMC. This results in crashes on
systems where the PSCI firmware is implemented in the hypervisor, e.g.
qemu-arm64_defconfig.
The logic to decide which call is needed based on the device tree is
already implemented in the PSCI firmware driver. During the EFI runtime
the device driver model is not available. But we can minimize code
duplication by merging the EFI runtime reset and poweroff code with
the PSCI firmware driver.
As the same HVC/SMC problem is also evident for the ARMv8 do_poweroff
and reset_misc routines let's move them into the same code module.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Reviewed-by: Sumit Garg <sumit.garg@linaro.org>
Tested-by: Sumit Garg <sumit.garg@linaro.org>
Signed-off-by: Alexander Graf <agraf@suse.de>
vf610twr_defconfig and vf610twr_nand_defconfig suffer from tight memory
restrictions. Disable CONFIG_EFI_LOADER for these boards.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Signed-off-by: Alexander Graf <agraf@suse.de>
Factor out efi_set_bootdev() and extract efi_dp_from_name().
This function will be used to set a boot device in efishell command.
Signed-off-by: AKASHI Takahiro <takahiro.akashi@linaro.org>
Signed-off-by: Alexander Graf <agraf@suse.de>
As in the case of fs_set_blk_dev(), fs_set_blk_dev_with_part() should
maintain and update fs_dev_part whenever called.
Without this patch, a problem will come up when an efi binary associated
with efi's BOOTxxxx variable is invoked via "bootefi bootmgr".
Signed-off-by: AKASHI Takahiro <takahiro.akashi@linaro.org>
Signed-off-by: Alexander Graf <agraf@suse.de>
Per Microsoft PE Format documentation [1], PointerToSymbolTable and
NumberOfSymbols should be zero for an image in the COFF file header.
Currently the COFF file header is hardcoded on RISC-V and these two
members are not zero.
This updates the hardcoded structure to clear these two members, as
well as setting the flag IMAGE_FILE_LOCAL_SYMS_STRIPPED so that we
can generate compliant *.efi images.
[1] https://docs.microsoft.com/zh-cn/windows/desktop/Debug/pe-format
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Signed-off-by: Alexander Graf <agraf@suse.de>
Per Microsoft PE Format documentation [1], PointerToSymbolTable and
NumberOfSymbols should be zero for an image in the COFF file header.
Currently the COFF file header is hardcoded on ARM and these two
members are not zero.
This updates the hardcoded structure to clear these two members, as
well as setting the flag IMAGE_FILE_LOCAL_SYMS_STRIPPED so that we
can generate compliant *.efi images.
[1] https://docs.microsoft.com/zh-cn/windows/desktop/Debug/pe-format
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Signed-off-by: Alexander Graf <agraf@suse.de>
This adds characteristics macros as defined by the Microsoft PE
Format documentation [1].
[1] https://docs.microsoft.com/zh-cn/windows/desktop/Debug/pe-format
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Signed-off-by: Alexander Graf <agraf@suse.de>
Per Microsoft PE Format documentation [1], PointerToSymbolTable and
NumberOfSymbols should be zero for an image in the COFF file header.
Currently U-Boot is generating u-boot-app.efi in which these two
members are not zero.
This updates the build rules to tell linker to remove the symbol
table completely so that we can generate compliant *.efi images.
[1] https://docs.microsoft.com/zh-cn/windows/desktop/Debug/pe-format
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Signed-off-by: Alexander Graf <agraf@suse.de>
Per Microsoft PE Format documentation [1], PointerToSymbolTable and
NumberOfSymbols should be zero for an image in the COFF file header.
Currently U-Boot is generating u-boot-payload.efi image in which
these two members are not zero.
This updates the build rules to tell linker to remove the symbol
table completely so that we can generate compliant *.efi images.
[1] https://docs.microsoft.com/zh-cn/windows/desktop/Debug/pe-format
Reported-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Signed-off-by: Alexander Graf <agraf@suse.de>
Per Microsoft PE Format documentation [1], PointerToSymbolTable and
NumberOfSymbols should be zero for an image in the COFF file header.
Currently U-Boot is generating *.efi images (eg: helloworld.efi) in
which these two members are not zero.
This updates the build rules to tell linker to remove the symbol
table completely so that we can generate compliant *.efi images.
[1] https://docs.microsoft.com/zh-cn/windows/desktop/Debug/pe-format
Reported-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Signed-off-by: Alexander Graf <agraf@suse.de>
If UninstallMultipleProtocolInterfaces fails, we sometimes return the wrong
status code. The UEFI spec mandates to always return EFI_INVALID_PARAMETER.
Update unit test.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Signed-off-by: Alexander Graf <agraf@suse.de>
The UEFI spec defines that a line feed moves the cursor to the next line
and (only) a carriage return moves the cursor to the beginning of the line.
So we should issue CR LF when we want to get to the start of the next line.
Add some comments.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Signed-off-by: Alexander Graf <agraf@suse.de>
Rename the component parent of some EFI objects to header. This avoids
misunderstandings.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Signed-off-by: Alexander Graf <agraf@suse.de>
A pointer to a struct efi_object is a handle. We do not need any handle
member in this structure. Let's eliminate it.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Signed-off-by: Alexander Graf <agraf@suse.de>
All our handles point to a struct efi_object. So let's define the
efi_handle_t accordingly. This helps us to discover coding errors much
more easily. This becomes evident by the corrections to the usage of
handles in this patch.
Rename variable image_handle to image_obj where applicable.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Signed-off-by: Alexander Graf <agraf@suse.de>
fix typos
correct the header comment of efi_selftest_variables.c
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Signed-off-by: Alexander Graf <agraf@suse.de>
When the last protocol interface has been uninstalled remove the handle.
Adjust ReinstallProtocol so that it does not remove the handle.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Signed-off-by: Alexander Graf <agraf@suse.de>
Allow an EFI application to shut down the system.
If EFI_RESET_SHUTDOWN is issued call do_poweroff().
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Signed-off-by: Alexander Graf <agraf@suse.de>
Test the handling of execptions by trying to execute an undefined
instruction. For 32bit ARM we expect \selftest to be listed as loaded
image.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Signed-off-by: Alexander Graf <agraf@suse.de>
We should not make anything in lib/efi_selftest if
CONFIG_CMD_BOOTEFI_SELFTEST is not defined.
We can make that test in lib/Makefile
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Signed-off-by: Alexander Graf <agraf@suse.de>
Variable 'days' must be defined as signed int. Otherwise the conversion
fails for some dates, e.g. 2004-08-25. Cf function rtc_time64_to_tm() in
the Linux kernel source.
Fixes: 992c1db455 "drivers: rtc: resolve year 2038 problem in rtc_to_tm"
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Convert TPM fully to DM
Tidy up sandbox I2C emulation
Add a 'make qcheck' target for faster testing
A few other misc things
(dropped the final patch which breaks clang for some reason)
-----BEGIN PGP SIGNATURE-----
iQEzBAABCgAdFiEEslwAIq+Gp8wWVbYnfxc6PpAIreYFAlwBjx8ACgkQfxc6PpAI
rebWEAf+PMc/M3X6cCKm+8P9xTNFv0gv234HVke1eTDG3eyZB6WQnJn5CS0LwyPj
JUfwk4xK0Sqev/sQ6RuzmxmPAgla6UrEFB355tiZRBtFsMgf8bfevOqbxVqMbZl/
bOe4YsFJ3sAgQDuhi2C5xzm2speGtaB8Qy8s1Zlv2sx3zEnLgRqVgeEVHzulZsk3
gAlWKN3ys6EueYnOladjrrvrfOX8SQKYbThI+ACy7suZbtasGRITwP/4fQoXtmOl
39e9WsiW6DBRW1sxNxpixrOS5trgWjOOFiKlY3GNWP+N34+9vYASwuBWwRWqfgXo
cmelT6AcGiPUbHyr4k0CJHKCJlSbaQ==
=p83K
-----END PGP SIGNATURE-----
Merge tag 'pull-30nov18' of git://git.denx.de/u-boot-dm
Fix sound on sandbox
Convert TPM fully to DM
Tidy up sandbox I2C emulation
Add a 'make qcheck' target for faster testing
A few other misc things
(dropped the final patch which breaks clang for some reason)
Sync with other rockchip SoCs, use board_debug_uart_init() to
init default UART iomux.
Signed-off-by: Kever Yang <kever.yang@rock-chips.com>
Reviewed-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
This patch sets the PLL of CPU cluster B (BPLL) to 600 MHz.
This decreases the boot time of Linux 4.19 by about 8%.
The 600 MHz are inspired by the 600 MHz used for LPLL initialization
(came in with commit 9f636a249c).
Tested on RK3399-Q7 on Haikou base board.
Signed-off-by: Christoph Muellner <christoph.muellner@theobroma-systems.com>
Reviewed-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
The "Flamingo" carrier-board for the RK3399-Q7 has a RV3029 populated
and the application will use the off-module RV3029 RTC including the
battery backed SRAM.
To support this use case, this commit includes the following changes:
* updates the rv3029 driver to use DM
* implements the read8/write8 operations
This syncs the implementation with the Linux code (based on 4.17),
porting the trickle-charger support from there (with improvements to
avoid unnecessary EEPROM updates) and adheres to the Linux DTS
binding.
Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Tested-by: Klaus Goger <klaus.goger@theobroma-systems.com>
The MicroCrystal RV3029 driver didn't have a Kconfig entry and was not used
anywhere. Add it to Kconfig to make it selectable.
Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Tested-by: Klaus Goger <klaus.goger@theobroma-systems.com>
This allow easier integration of RV1108 based boards on generic
distributions and build systems.
To avoid behavior change, we make evb-rv1108 to use the existing
environment as it boots from its SPI NOR.
Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Reviewed-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
This adds the definitions need to use the USB OTG in rv1108
board. This has been tested using USB Mass Storage to export and
program a eMMC device.
Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Reviewed-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Like it is done for other Rockchip SoCs, introduce a board_usb_init()
function so that USB OTG can be functional on rv1108 too.
Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Reviewed-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
This adds the pinctrl handles to enable the use of eMMC on custom
boards (as minievk) and makes it easier for later addition.
Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Reviewed-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
In order to be able to build the Rockchip eMMC driver on rv1108, the
BOUNCE_BUFFER option needs to be selected. Select it like it is done
on the other Rockchip SoC common files.
Reviewed-by: Andy Yan <andy.yan@rock-chips.com>
Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Reviewed-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Make adjustments to the rv1108 clock driver in order to align it
with the internal Rockchip version.
Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Reviewed-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Some SanDisk Ultra cards trigger intermittent errors on detection
resulting in an -EOPNOTSUPP, when running at 50MHz.
Waveform analysis suggest that the level shifters that are used on the
RK3399-Q7 module (for voltage translation between the on-module
voltages and the 3.3V required on the card-edge) don't handle clock
rates at or above 48MHz properly. This change reduces the maximum
frequency on the external SD-interface to 40MHz (for a safety margin
of 20%).
Reported-by: Jakob Unterwurzacher <jakob.unterwurzacher@theobroma-systems.com>
Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Tested-by: Christoph Muellner <christoph.muellner@theobroma-systems.com>
The RK3188 rock board does not need TPL: remove TPL_TINY_MEMSET from
config.
Signed-off-by: Kever Yang <kever.yang@rock-chips.com>
Reviewed-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
[Fixed up commit message:]
Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
The RK3399 SPL has two cases that may end in a hard-stop: if either
the pinctrl can not be initialised or if the DRAM fails to initialise.
Both have previously not triggered an error message unless DEBUG was
defined (i.e. both used debug() to print the error).
This converts both error messages to be printed using pr_err() to
ensure that some output points to the cause of the hard-stop.
Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
There was an incorrect check when looping and finding the first
fast enough frequency in the freq_rang table. The code did
actually return the first that was either exactly correct or
too slow.
Signed-off-by: Richard Röjfors <richard@puffinpack.se>
Reviewed-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
