Simon Glass [Mon, 16 Jan 2017 14:04:10 +0000 (07:04 -0700)]
x86: Don't build call64 and setjmp on 64-bit
These are currently not supported. Calling 64-bit code from 64-bit U-Boot is
much simpler, so this code is not needed. setjmp() is not yet implemented for
64-bit.
Signed-off-by: Simon Glass <sjg@chromium.org> Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Simon Glass [Mon, 16 Jan 2017 14:04:08 +0000 (07:04 -0700)]
x86: ivybridge: Provide a dummy SDRAM init for 64-bit
We don't support SDRAM init in 64-bit mode since it is essentially
impossible to get into that mode before SDRAM set up. Provide dummy functions
for now. At some point we will need to pass the SDRAM parameters through from
SPL.
Signed-off-by: Simon Glass <sjg@chromium.org> Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Simon Glass [Mon, 16 Jan 2017 14:03:54 +0000 (07:03 -0700)]
x86: Add support for 64-bit relocation
Add a 64-bit relocation function. SPL loads U-Boot into RAM at a fixed
address and runs it. U-Boot then relocates itself to the top of RAM using
this relocation function.
Signed-off-by: Simon Glass <sjg@chromium.org> Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Simon Glass [Mon, 16 Jan 2017 14:03:52 +0000 (07:03 -0700)]
x86: Do relocation before clearing BSS
The BSS region may overlap with relocations. If we clear BSS we will
overwrite the start of the relocation area. This doesn't matter when running
from SPI flash, since it is read-only. But when relocating 64-bit U-Boot
from one place in RAM to another, relocation will fail because some of its
relocations have been zeroed.
To fix this, put the ELF fixup call before the BSS clearing call.
Signed-off-by: Simon Glass <sjg@chromium.org> Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Simon Glass [Mon, 16 Jan 2017 14:03:51 +0000 (07:03 -0700)]
x86: board_r: Set the global data pointer after relocation
Since 'gd' is just a normal variable on 64-bit x86, it is relocated by the
time we get to board_init_r(). The old 'gd' variable is passed in as
parameter to board_init_r(), presumably for this situation.
Assign it on 64-bit x86 so that gd points to the correct data.
Options to improve this:
- Make gd a fixed register and remove the board_init_r() parameter
- Make all archs use this board_init_r() parameter
The second has a TODO in the code. The first has a TODO in a future commit
('x86: Support global_data on x86_64')
Signed-off-by: Simon Glass <sjg@chromium.org> Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Simon Glass [Mon, 16 Jan 2017 14:03:49 +0000 (07:03 -0700)]
x86: board_f: Update init sequence for 64-bit startup
Adjust the code so that 64-bit startup works. Since we don't need to do CAR
changes in U-Boot proper anymore (they are done in SPL) we can simplify the
flow and return normally from board_init_f().
Signed-off-by: Simon Glass <sjg@chromium.org> Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Simon Glass [Mon, 16 Jan 2017 14:03:44 +0000 (07:03 -0700)]
x86: Kconfig: Add location options for 16/32-bit init
At present all 16/32-bit init is controlled by CONFIG_X86_RESET_VECTOR. If
this is enabled, then U-Boot is the 'first' boot loader and handles execution
from the reset vector through to U-Boot's command prompt. If it is not
enabled then U-Boot starts at the 32-bit entry and skips most of its init,
assuming that the previous boot loader has done this already.
With the move to suport 64-bit operation, we have more cases to consider.
The 16-bit and 32-bit init may be in SPL rather than in U-Boot proper.
Add Kconfig options which control the location of the 16-bit and the 32-bit
init. These are not intended to be user-setting except for experimentation.
Their values should be determined by whether 64-bit U-Boot is used.
Signed-off-by: Simon Glass <sjg@chromium.org> Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Simon Glass [Mon, 16 Jan 2017 14:03:36 +0000 (07:03 -0700)]
x86: Update mpspec to build on 64-bit machines
At present this uses u32 to store an address. We should use unsigned long
and avoid special types in function return values and parameters unless
necessary. This makes the code more portable.
Signed-off-by: Simon Glass <sjg@chromium.org> Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Simon Glass [Mon, 16 Jan 2017 14:03:28 +0000 (07:03 -0700)]
spl: Makefile: Define SPL_ earlier
This Makefile variable can be used in the architecture's main Makefile but
at present it is not set up until later. Set it just before this Makefile is
included.
Signed-off-by: Simon Glass <sjg@chromium.org> Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Simon Glass [Mon, 16 Jan 2017 14:03:26 +0000 (07:03 -0700)]
console: Don't enable CONFIG-CONSOLE_MUX, etc. in SPL
CONFIG_CONSOLE_MUX and CONFIG_SYS_CONSOLE_IS_IN_ENV are not applicable
for SPL. Update the console code to use CONFIG_IS_ENABLED(), so that these
options will be inactive in SPL.
Signed-off-by: Simon Glass <sjg@chromium.org> Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
eLBC IP clock is always a constant divisor of platform clock
pre-defined per SoC. Clock ratio register (LCRR) used in
current implementation governs eLBC IP output cloc.
Update sys_info->freq_localbus to represent eLBC input clock with
value constant divisor of platform clock.
Signed-off-by: Prabhakar Kushwaha <prabhakar.kushwaha@nxp.com> Reviewed-by: York Sun <york.sun@nxp.com>
IFC IP clock is always a constant divisor of platform clock
pre-defined per SoC. Clock control register (CCR) used in
current implementation governs IFC IP output clock.
Update sys_info->freq_localbus to represent IFC input clock with
value constant divisor of platform clock.
Signed-off-by: Prabhakar Kushwaha <prabhakar.kushwaha@nxp.com> Reviewed-by: York Sun <york.sun@nxp.com>
IFC IP clock is always a constant divisor of platform clock
pre-defined per SoC. Clock control register (CCR) used in
current implementation governs IFC IP output clock.
Update sys_info->freq_localbus to represent IFC input clock with
value constant divisor of platform clock.
Signed-off-by: Prabhakar Kushwaha <prabhakar.kushwaha@nxp.com> Reviewed-by: York Sun <york.sun@nxp.com>
Mark Marshall [Tue, 24 Jan 2017 12:47:31 +0000 (13:47 +0100)]
powerpc: mpc5200: Correct return value of memcpy function
The memcpy() function returns a pointer to trg.
Signed-off-by: Mark Marshall <Mark.Marshall@omicron.at> Reviewed-by: Thomas Graziadei <thomas.graziadei@omicronenergy.com> Reviewed-by: Anatolij Gustschin <agust@denx.de> Reviewed-by: York Sun <york.sun@nxp.com>
Mario Six [Wed, 11 Jan 2017 15:01:00 +0000 (16:01 +0100)]
arm: mvebu: Implement secure boot
The patch implements secure booting for the mvebu architecture.
This includes:
- The addition of secure headers and all needed signatures and keys in
mkimage
- Commands capable of writing the board's efuses to both write the
needed cryptographic data and enable the secure booting mechanism
- The creation of convenience text files containing the necessary
commands to write the efuses
The KAK and CSK keys are expected to reside in the files kwb_kak.key and
kwb_csk.key (OpenSSL 2048 bit private keys) in the top-level directory.
Signed-off-by: Reinhard Pfau <reinhard.pfau@gdsys.cc> Signed-off-by: Mario Six <mario.six@gdsys.cc> Reviewed-by: Stefan Roese <sr@denx.de> Reviewed-by: Simon Glass <sjg@chromium.org> Signed-off-by: Stefan Roese <sr@denx.de>
Mario Six [Wed, 11 Jan 2017 15:00:59 +0000 (16:00 +0100)]
tools: kwbimage: Refactor line parsing and fix error
The function image_create_config_parse_oneline is pretty complex, and
since more parameters will be added to support secure booting, we
refactor the function to make it more readable.
Also, when a line contained just a keyword without any parameters,
strtok_r returned NULL, which was then indiscriminately fed into atoi,
causing a segfault. To correct this, we add a NULL check before feeding
the extracted token to atoi, and print an error message in case the
token is NULL.
Signed-off-by: Mario Six <mario.six@gdsys.cc> Reviewed-by: Stefan Roese <sr@denx.de> Signed-off-by: Stefan Roese <sr@denx.de>
Mario Six [Wed, 11 Jan 2017 15:00:55 +0000 (16:00 +0100)]
tools: kwbimage: Fix arithmetic with void pointers
Arithmetic with void pointers, e.g. a - b where both a and b are void
pointers, is undefined in the C standard. Since we are operating with
byte data here, we switch the void pointers to uint8_t pointers, and add
the necessary casts.
Signed-off-by: Mario Six <mario.six@gdsys.cc> Reviewed-by: Stefan Roese <sr@denx.de> Signed-off-by: Stefan Roese <sr@denx.de>
Mario Six [Wed, 11 Jan 2017 15:00:54 +0000 (16:00 +0100)]
tools: kwbimage: Fix style violations
Fix some style violations:
- nine instances of missing blank lines after declarations
- one overly long line
- one split string (which also rewords an error message more concisely)
- two superfluous else
Signed-off-by: Mario Six <mario.six@gdsys.cc> Reviewed-by: Stefan Roese <sr@denx.de> Signed-off-by: Stefan Roese <sr@denx.de>
Mario Six [Wed, 11 Jan 2017 15:00:53 +0000 (16:00 +0100)]
tools: kwbimage: Fix dest addr
To enable secure boot, we need to jump back into the BootROM to continue
the SoC's boot process instead of letting the SPL load and run the main
U-Boot image.
But, since the u-boot-spl.img (including the 64 byte header) is loaded
by the SoC as the main image, we need to compensate for the header
length to get a correct entry point.
Thus, we subtract the header size from the destination address, so that
the execution address points at the actual entry point of the image.
The current boards ignore both parameters anyway, so this change shouldn't
concern them.
Signed-off-by: Mario Six <mario.six@gdsys.cc> Reviewed-by: Stefan Roese <sr@denx.de> Signed-off-by: Stefan Roese <sr@denx.de>
Mario Six [Wed, 11 Jan 2017 15:00:50 +0000 (16:00 +0100)]
lib: tpm: Add command to flush resources
This patch adds a function to the TPM library, which allows U-Boot to
flush resources, e.g. keys, from the TPM.
Signed-off-by: Mario Six <mario.six@gdsys.cc> Reviewed-by: Stefan Roese <sr@denx.de> Reviewed-by: Simon Glass <sjg@chromium.org> Signed-off-by: Stefan Roese <sr@denx.de>
Reinhard Pfau [Wed, 11 Jan 2017 15:00:48 +0000 (16:00 +0100)]
arm: mvebu: spl.c: Remove useless gd declaration
ddaa905 ("arm: mvebu: Add DM (driver model) support") removed the
assignment of the gd pointer, but kept the (now superfluous) declaration
of the gd pointer.
Remove this declaration.
Signed-off-by: Reinhard Pfau <pfau@gdsys.de> Reviewed-by: Stefan Roese <sr@denx.de> Signed-off-by: Stefan Roese <sr@denx.de>
Mario Six [Wed, 11 Jan 2017 15:00:47 +0000 (16:00 +0100)]
mvebu: Add board_pex_config()
Allow boards to do some initialization when PCIe comes up.
Signed-off-by: Dirk Eibach <dirk.eibach@gdsys.cc> Signed-off-by: Mario Six <mario.six@gdsys.cc> Reviewed-by: Stefan Roese <sr@denx.de> Signed-off-by: Stefan Roese <sr@denx.de>
Add support for Marvell 88E1680 Integrated Octal
10/100/1000 Mbps Energy Efficient Ethernet Transceiver.
Signed-off-by: Dirk Eibach <dirk.eibach@gdsys.cc> Signed-off-by: Mario Six <mario.six@gdsys.cc> Reviewed-by: Stefan Roese <sr@denx.de> Acked-by: Joe Hershberger <joe.hershberger@ni.com> Signed-off-by: Stefan Roese <sr@denx.de>
Dirk Eibach [Wed, 11 Jan 2017 15:00:45 +0000 (16:00 +0100)]
pci: mvebu: Fix Armada 38x support
Armada 38x has four PCI ports, not three.
The optimization in pci_init_board() seems to assume that every port has
three lanes. This is obviously wrong, and breaks support for Armada 38x.
Signed-off-by: Dirk Eibach <dirk.eibach@gdsys.cc> Signed-off-by: Mario Six <mario.six@gdsys.cc> Reviewed-by: Stefan Roese <sr@denx.de> Signed-off-by: Stefan Roese <sr@denx.de>
Stefan Roese [Thu, 12 Jan 2017 15:37:49 +0000 (16:37 +0100)]
phy: comphy_a3700: Change SD/MMC compatible DT node to match the updates
Now that the SD/SDIO/MMC DT properties are updated in the Marvell
A3700 and A7/8k DT files, we need to match the checks for compatible
node in the PHY driver as well.
Mark Marshall [Tue, 24 Jan 2017 14:40:23 +0000 (15:40 +0100)]
powerpc: mpc85xx: Use symbolic names for cache control bits
We should use the symbolic names for the cache control bits.
Signed-off-by: Mark Marshall <Mark.Marshall@omicron.at> Reviewed-by: Thomas Graziadei <thomas.graziadei@omicronenergy.com> Reviewed-by: York Sun <york.sun@nxp.com>
The r1 register is modified several times during the cache-ram setup of
the MPC83xx SoCs.
Since this SP modification confuses debuggers, we use a general purpose
register to compute the new stack pointer value, and only set the SP
once after all computations are done.
Signed-off-by: Mario Six <mario.six@gdsys.cc> Reviewed-by: Joakim Tjernlund <Joakim.Tjernlund@infinera.com> Reviewed-by: York Sun <york.sun@nxp.com>
Fixup port_mac_address property in MC DPC with values from the u-boot
environment. Since u-boot already reads the environment MAC addresses
when probing the PHYs, use these values.
The u-boot environment MAC addresses take precedence over any eventual
ones defined in the DPC, except for the case where they are randomly
assigned (no u-boot env value declared for port).
The patch assumes the "/board_info/ports/" node is present in the DPC.
Signed-off-by: Bogdan Purcareata <bogdan.purcareata@nxp.com> Reviewed-by: Prabhakar Kushwaha <prabhakar.kushwaha@nxp.com>
[York S: Fix several indentations] Reviewed-by: York Sun <york.sun@nxp.com>
Masahiro Yamada [Mon, 30 Jan 2017 10:46:50 +0000 (19:46 +0900)]
mmc: sandbox: rename CONFIG, fix dependency, and use it in Makefile
[1] Rename CONFIG_SANDBOX_MMC to CONFIG_MMC_SANDBOX for consistency
I want all MMC driver options prefixed with CONFIG_MMC_.
[2] Fix dependency
Add necessary depends on to avoid compile error.
Instead "depends on MMC" is unneeded because this config entry
resides inside of "if MMC".
[3] Currently, this config symbol is not referenced at all.
Use it to enable/disable the driver in Makefile.
Masahiro Yamada [Mon, 30 Jan 2017 10:46:49 +0000 (19:46 +0900)]
mmc: move CONFIG_GENERIC_MMC to Kconfig
Now, CONFIG_GENERIC_MMC seems equivalent to CONFIG_MMC.
Let's create an entry for "config GENERIC_MMC" with "default MMC",
then convert all macro defines in headers to Kconfig. Almost all
of the defines will go away.
I see only two exceptions:
configs/blanche_defconfig
configs/sandbox_noblk_defconfig
They define CONFIG_GENERIC_MMC, but not CONFIG_MMC. Something
might be wrong with these two boards, so should be checked later.
Anyway, this is the output of the moveconfig tool.
This commit was created as follows:
[1] create a config entry in drivers/mmc/Kconfig
[2] tools/moveconfig.py -r HEAD GENERIC_MMC
[3] manual clean-up of garbage comments in doc/README.* and
include/configs/*.h
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com> Reviewed-by: Tom Rini <trini@konsulko.com>
Patrick Bruenn [Mon, 23 Jan 2017 14:11:27 +0000 (15:11 +0100)]
arm: dts: imx53-cx9020: fix packetloss on fec_mxc
The pinmuxing for i.MX53 FEC ethernet copied from
<kernel>/arch/arm/boot/dts/imx53-qsb-common.dtsi (at least until v4.9)
was bad. It is different from the manual pinmuxing in
<u-boot>/board/freescale/mx53loco/mx53loco.c which was used in
cx9020 implementation previously before mainlining into u-boot.
It seems the bug in imx53-qsb kernel device tree is hidden for so long,
because it was never used, by the kernel driver.
Signed-off-by: Patrick Bruenn <p.bruenn@beckhoff.com>
Tom Rini [Sun, 29 Jan 2017 13:01:06 +0000 (08:01 -0500)]
Merge branch 'master' of git://git.denx.de/u-boot-uniphier
- Fix clk driver
- Optimize DRAM init code for LD20 SoC
- Get DRAM information from more reliable source
- Clean up SoC init code
- Allow to use Image.gz for booting ARM64 Linux
- Tidy up environments to use with ATF
- Clean up I2C drivers
Masahiro Yamada [Fri, 27 Jan 2017 21:53:54 +0000 (06:53 +0900)]
ARM: uniphier: make update commands more flexible for ATF
Currently, SPL (u-boot-spl.bin) and U-Boot (u-boot.bin) are stored
in non-volatile devices, and some environments are defined to update
the images easily.
When ARM Trusted Firmware is fully used, SPL is not used. U-Boot
proper is contained as BL33 into FIP (Firmware Image Package), which
is standard container used by ATF. Allow to use it.
Masahiro Yamada [Fri, 27 Jan 2017 21:53:53 +0000 (06:53 +0900)]
ARM: uniphier: change CONFIG_SPL_PAD_TO to 128KB
The Boot ROM supports authentication feature to prevent malformed
software from being run on products. The signature is added at the
tail of the second stage loader (= SPL in U-boot terminology).
The size of the second stage loader was 64KB, and it was consistent
across SoCs. The situation changed when LD20 SoC appeared; it loads
80KB second stage loader, and it is the only exception.
Currently, CONFIG_SPL_PAD_TO is set to 64KB and U-Boot proper is
loaded from the 64KB offset of non-volatile devices. This means the
signature of LD20 SoC (located at 80KB offset) corrupts the U-Boot
proper image.
Let's move the U-Boot proper image to 128KB offset. It uses 48KB
for nothing but padding, and we could actually locate the U-Boot
proper at 80KB offset. However, the power of 2 generally seems a
better choice for the offset address.
Masahiro Yamada [Fri, 27 Jan 2017 21:53:52 +0000 (06:53 +0900)]
ARM: uniphier: change the offset to environment storage area
When ARM Trusted Firmware is used, bl1.bin + fip.bin exceeds 512KB,
so the boot image and the current environment area will overlap.
Move the environment storage to 1MB offset.
Masahiro Yamada [Fri, 27 Jan 2017 21:53:51 +0000 (06:53 +0900)]
ARM: uniphier: set initrd_high environment to skip initrd relocation
The boot_ramdisk_high() checks the environment "initrd_high" and,
if it is set to (ulong)-1, skip the initrd relocation. This is
useful for faster booting when we know the initrd is already located
within the reach of the kernel.
Change "norboot" to copy images in order to make it work without
depending on the automatic relocation.
Masahiro Yamada [Fri, 27 Jan 2017 21:53:50 +0000 (06:53 +0900)]
ARM: uniphier: use Image.gz instead Image for booting ARM64 Linux
The ARM64 Linux raw image now amounts to 15MB and it is getting
bigger and bigger. Using Image.gz saves about 8MB. The cost of
unzip is smaller than what we get by saving the kernel loading
from non-volatile devices.
The ARM32 Linux still uses zImage, a self-decompressor image,
so it should not be affected.