sunxi/nand: Add support to the SPL for loading u-boot from internal NAND memory
This commit adds support to the sunxi SPL to load u-boot from the internal
NAND. Note this only adds support to access the boot partitions to load
u-boot, full NAND support to load the kernel, etc. from the nand data
partition will come later.
Signed-off-by: Roy Spliet <r.spliet@ultimaker.com> Reviewed-by: Hans de Goede <hdegoede@redhat.com> Signed-off-by: Hans de Goede <hdegoede@redhat.com>
sunxi/nand: change BLOCK_SIZE in mksunxiboot to match NAND block size
This change is necessary to calculate correct checksum for NAND
boot. Works both for MMC and NAND. Without it BROM rejects boot image
as invalid (bad checksum). (Changes block size from 0x200 to 0x2000).
Signed-off-by: Daniel Kochmański <dkochmanski@turtle-solutions.eu> Signed-off-by: Roy Spliet <r.spliet@ultimaker.com> Reviewed-by: Hans de Goede <hdegoede@redhat.com> Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Hans de Goede [Wed, 20 May 2015 13:27:16 +0000 (15:27 +0200)]
sunxi: Update sunxi-common.h to deal with different A1-SRAM base addr on sun9i
The A1 SRAM Base differs between sun9i and the others, update sunxi-common.h
to deal with this, so that we do not set the initial stack pointer to point
to the BROM.
This avoids the need for the weird undocumented register write I previously
took from the allwiner u-boot sources and which needed #ifdef-ery in start.S
as it needed to be done really early on.
Signed-off-by: Hans de Goede <hdegoede@redhat.com> Acked-by: Ian Campbell <ijc@hellion.org.uk>
Hans de Goede [Tue, 19 May 2015 21:34:00 +0000 (23:34 +0200)]
sunxi: Use axp221 sid on a33
Unlike the A31 and the A23 the A33 actually has a SID inside the SoC again,
but sid[3] is 0 (at least on some SoCs), so it is better to use the axp221
sid.
Signed-off-by: Hans de Goede <hdegoede@redhat.com> Acked-by: Ian Campbell <ijc@hellion.org.uk>
Heiko Schocher [Thu, 28 May 2015 05:27:36 +0000 (07:27 +0200)]
am33xx, spl, siemens: enable debug uart output again
a6b541b090: TI ARMv7: Don't use GD before crt0.S has set it
moves the init of the debug uart at the very end of SPL code.
Enable it for the siemens board earlier, as they print
ddr settings ... all debug output before board_init_r()
is here currently useless. Maybe we must rework this
globally?
Tang Yuantian [Wed, 6 May 2015 03:21:33 +0000 (11:21 +0800)]
fsl/sata: Replace sprintf() with snprintf()
Function 'sprintf' does not check buffer boundaries but outputs
to the buffer of fixed size which could potentially cause buffer
overflow. Use a safer function to replace it.
Matt Porter [Tue, 5 May 2015 19:00:23 +0000 (15:00 -0400)]
common/cmd_boot: keep ARM v7M in thumb mode during do_go_exec()
On ARM v7M, the processor will return to ARM mode when executing
a blx instruction with bit 0 of the address == 0. Always set it
to 1 to stay in thumb mode.
Linus Walleij [Wed, 27 May 2015 07:45:39 +0000 (09:45 +0200)]
vexpress64: use uncompressed kernel by default
The foundation model (FVP) emulator nominally boots using
a clean, uncompressed kernel and the "booti" command. Augment
the default U-Boot script to do this.
Masahiro Yamada [Tue, 26 May 2015 03:51:25 +0000 (12:51 +0900)]
kbuild: define DO_DEPS_ONLY for u-boot.cfg to fix build error
Since 741e58e0fc8e (Create a .cfg file containing the CONFIG options
used to build), all the Blackfin boards fail to build if the parallel
(-j) option is passed.
$ make -s bf506f-ezkit_defconfig
#
# configuration written to .config
#
$ make -j8 CROSS_COMPILE=bfin-elf-
scripts/kconfig/conf --silentoldconfig Kconfig
CHK include/config.h
UPD include/config.h
GEN include/autoconf.mk
GEN include/autoconf.mk.dep
CHK include/config/uboot.release
CHK include/generated/timestamp_autogenerated.h
UPD include/generated/timestamp_autogenerated.h
CFG u-boot.cfg
include/asm-offsets.h:3:43: fatal error:
generated/generic-asm-offsets.h: No such file or directory
compilation terminated.
make: *** [u-boot.cfg] Error 1
When parsing header files for defined CONFIG options, DO_DEPS_ONLY
must be defined to exclude generated headers that might not have
been available yet.
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Andrew Bradford [Fri, 22 May 2015 12:30:14 +0000 (08:30 -0400)]
meminfo cmd: Enable printing >=4GiB SDRAM sizes
gd->ram_size is stored as phys_size_t type which can be bigger than an
unsigned long on some architectures. When using unsigned long type,
SDRAM of 4GiB or larger will not print the correct size, but using
phys_size_t will.
Signed-off-by: Andrew Bradford <andrew.bradford@kodakalaris.com>
This introduces a distinction between return codes that are read bytes counts
and errors. Read bytes counts are erroneous when null (no data was read) while
errors are erroneous when non-null.
Signed-off-by: Paul Kocialkowski <contact@paulk.fr> Tested-by: Tim Harvey <tharvey@gateworks.com>
Karl Apsite [Thu, 21 May 2015 13:52:50 +0000 (09:52 -0400)]
add test for two 'loadables'
Nothing too fancy. A simple test that attmpts to load two loadables and
verify that they are properly unpacked in the u-boot sandbox. Signed-off-by: Karl Apsite <Karl.Apsite@dornerworks.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Karl Apsite [Thu, 21 May 2015 13:52:49 +0000 (09:52 -0400)]
Combine bootm_find_<thing> functions together
bootm_find_ramdisk_fdt() renamed to bootm_find_images() for readability.
The function bootm_find_ramdisk_fdt() appears to be a simple wrapper for
bootm_find_ramdisk(), bootm_find_fdt(), and now bootm_find_loadables().
I didn't see any other callers entering a bootm_find<thing>, so removing
the wrapper, and condensing these together hopefully makes the code a
little simpler.
Signed-off-by: Karl Apsite <Karl.Apsite@dornerworks.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Karl Apsite [Thu, 21 May 2015 13:52:48 +0000 (09:52 -0400)]
add boot_get_loadables() to load listed images
Added a trimmed down instance of boot_get_<thing>() to satisfy the
minimum requierments of the added feature. The function follows the
normal patterns set by other boot_get<thing>'s, which should make it a
bit easier to combine them all together into one boot_get_image()
function in a later refactor.
Documentation for the new function can be found in source:
include/image.h
Signed-off-by: Karl Apsite <Karl.Apsite@dornerworks.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Karl Apsite [Thu, 21 May 2015 13:52:47 +0000 (09:52 -0400)]
mkimage will now report information about loadable
Added FIT_LOADABLE_PROP, so the user can identify an optional entry
named "loadables" in their .its configuration. "loadables" is a comma
separated list in the .its
Documentation can be found in doc/uImage.FIT/source_file_format.txt and
doc/uImage.Fit/multi-with-loadables.its
Signed-off-by: Karl Apsite <Karl.Apsite@dornerworks.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Thomas Petazzoni [Sat, 16 May 2015 16:36:15 +0000 (18:36 +0200)]
tools: use pkg-config when available to get SSL flags
Instead of hardcoding -lssl -lcrypto as the flags needed to build
mkimage with FIT signature enabled, use pkg-config when
available. This allows to properly support cases where static linking
is used, which requires linking with -lz, since OpenSSL uses zlib
internally.
We gracefully fallback on the previous behavior of hardcoding -lssl
-lcrypto if pkg-config is not available or fails with an error.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Masahiro Yamada [Fri, 15 May 2015 13:05:18 +0000 (22:05 +0900)]
include: introduce include/linux/io.h as a wrapper of asm/io.h
In the latest Linux coding style, <linux/io.h> should be included
rather than <asm/io.h>. To follow this standard also in U-Boot,
add include/linux/io.h. Currently, it just includes <asm/io.h>.
Linus Walleij [Thu, 14 May 2015 15:38:33 +0000 (17:38 +0200)]
vexpress64: let Juno wait for root device
The Juno reference design typically plugs the root FS
on a USB stick. We need to wait a bit for the root to
appear so tell this on the default command line.
DM3730 CPU used on beagleboard-xm doesn't have any gpio
multiplexed with wait0 pin,
but gpio_63 multiplexed with wait1, so it looks like typo.
At current form it has no sense,
because of wait0 have no function "4".
Signed-off-by: Evgneiy A. Dushistov <dushistov@mail.ru>
Linus Walleij [Mon, 11 May 2015 08:03:57 +0000 (10:03 +0200)]
vexpress64: fix various memory issues
The ARM Trusted Firmware or other security solutions are
eating memory from the top of the physical SDRAM1 space,
moving backward from 0xffffffff, currently occupying e.g.
0xfe000000-0xffffffff with Trusted Firmware.
This solution to reserving memory for secure world is not
optimal, so we need to think of how the secure world and
earlier boot stages should communicate to U-Boot what
memory they are eating up. For now let's just put 16MB
aside.
Also enable the memory test command and define start and
end of the test range so we can check that we actually have
all that memory available and working.
Suggested-by: Axel Haslam <ahaslam@baylibre.com> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Tim Harvey [Fri, 8 May 2015 21:52:09 +0000 (14:52 -0700)]
env_mmc: add error message to pass to set_default_env
Add an error message that gets passed to set_default_env() like
env_nand implements. This message is displayed to the user as the
reason for falling back to the default environment.
Debugging showed that the problem could be solved by preventing u-boot
from scanning different flash widths by specifying
CONFIG_SYS_FLASH_CFI_WIDTH so that only using the 'correct' width was
used.
Signed-off-by: Ryan Harkin <ryan.harkin@linaro.org> Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Lokesh Vutla [Fri, 8 May 2015 11:06:28 +0000 (16:36 +0530)]
ARM: BeagleBoard-x15: Rev A2x5: Change UART3 mux
UART3 RX and TX pins are now moved to support uart download capability
on the platform. Old pins are now GPIOs routed to expansion ports.
All original boards should either have been returned for modifications
or already modified for the required change and maintaining
compatibility for older boards are no longer needed.
Previously, if the menu activated by the 'sysboot' command gets
interrupted by a Ctrl-C, the behaviour is as if the menu timeout was
reached - i.e. boot the default menu entry. This patch fixes that
so a Ctrl-C now terminates the command as the user would expect.
Signed-off-by: Tuomas Tynkkynen <tuomas.tynkkynen@iki.fi> Reviewed-by: Simon Glass <sjg@chromium.org>
Tom Rini [Wed, 27 May 2015 17:24:10 +0000 (13:24 -0400)]
Revert "Revert "stm32f4: fix serial output""
Due to a misunderstanding, in 698a12b we reverted changes that we in
fact wanted to keep. So lets fix that mistake and bring the code back
to how it should have been.
Joe Hershberger [Tue, 19 May 2015 18:21:24 +0000 (13:21 -0500)]
moveconfig: Add a switch to enable printing errors
In some cases the build for the autoconf breaks. This outputs the errors
following the status so that action can be taken without building again
manually.
Signed-off-by: Joe Hershberger <joe.hershberger@ni.com> Acked-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Joe Hershberger [Tue, 19 May 2015 18:21:23 +0000 (13:21 -0500)]
moveconfig: Print a message for missing compiler
A common case for failed builds is a missing compiler. Print a message
for that case to tell the user concisely which compiler was expected
that was not found.
This patch also has the effect of not printing build errors any longer.
The next patch will add a switch to optionally bring that back.
Signed-off-by: Joe Hershberger <joe.hershberger@ni.com> Acked-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Joe Hershberger [Tue, 19 May 2015 18:21:18 +0000 (13:21 -0500)]
moveconfig: Ignore duplicate configs when moving
When moving configs, it is important to know what was defined in the
config header even if it duplicates the configs coming from Kconfig.
This is specifically needed for the case where a config is set to
default 'y' in the Kconfig. This would previously cause the actual value
from the include config to be filtered out, and moveconfig.py would
think that it was 'n'... This means that the value that should be 'y'
is now (in every defconfig) set to 'not set'.
tools/moveconfig.py now defines KCONFIG_IGNORE_DUPLICATES to prevent the
filtering from happening and selecting wrong values for the defconfig.
Signed-off-by: Joe Hershberger <joe.hershberger@ni.com> Acked-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Joe Hershberger [Tue, 19 May 2015 18:21:17 +0000 (13:21 -0500)]
moveconfig: Always run savedefconfig on the moved config
This will ensure that the order of the defconfig entries will always
match that of the Kconfig files. After one slightly painful (but
still early in the process) pass over all boards, this should keep
the defconfigs clean from here on.
Users must edit the Kconfig first to add the menu entries and then run
moveconfig.py to update the defconfig files and the include configs.
As such, moveconfig.py cannot compare against the '.config' contents.
Signed-off-by: Joe Hershberger <joe.hershberger@ni.com> Acked-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Masahiro Yamada [Wed, 20 May 2015 02:36:07 +0000 (11:36 +0900)]
tools: moveconfig: a tool to move CONFIGs from headers to defconfigs
This tool was originally written for my local use to ease the task
of tons of CONFIG moves, but there have been some requests for
mainlining it.
So, I have tidied up the code with nicer comments, and here it is.
See the comment block of the script for usage.
The first draft was
http://patchwork.ozlabs.org/patch/430422/
Main updates are:
- Adapted to the single .config configuration
- Support colored log
- Support moving multiple options at once
(and take configs via input file only)
- Continue even if some boards fail
(Idea provided by Joe Hershberger)
- Add more options
- More comments and code cleanups
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com> Signed-off-by: Joe Hershberger <joe.hershberger@ni.com>
Peter Robinson [Fri, 22 May 2015 16:30:45 +0000 (17:30 +0100)]
imx6: move all standard includes to mx6_common.h
The linux/sizes.h, asm/arch/imx-regs.h, asm/imx-common/gpio.h,
config_cmd_default.h includes are used fairly universally across imx6 boards
so include them in mx6_common.h by default.
We define CONFIG_SYS_NO_FLASH before config_cmd_default.h so that we
don't have to undef CONFIG_CMD_FLASH / CONFIG_CMD_IMLS everywhere.
Signed-off-by: Peter Robinson <pbrobinson@gmail.com>
Tim Harvey [Thu, 21 May 2015 15:40:06 +0000 (08:40 -0700)]
thermal: imx_thermal: increase critical temperature threshold
The CPU temperature grade from OTP is now used to define the critical
threshold at which point we busyloop until we are below, however this
threshold is still too low.
Instead of 20C below the max CPU temperature, change it to 5C defined now
by TEMPERATURE_HOT_DETLA for clarity. Rename 'passive' to 'critical'
as that better defines our use case here. Additionally change the output
of the busyloop message to show the max CPU temperature as well as current.
Before:
CPU Temperature is 101 C, too hot to boot, waiting...
CPU Temperature is 101 C, too hot to boot, waiting...
After:
CPU Temperature (101C) too close to max (105C) waiting...
CPU Temperature (101C) too close to max (105C) waiting...
Cc: Stefan Roese <sr@denx.de> Cc: Eric Nelson <eric.nelson@boundarydevices.com> Cc: Heiko Schocher <hs@denx.de> Cc: Nikita Kiryanov <nikita@compulab.co.il> Cc: Jon Nettleton <jon.nettleton@gmail.com> Cc: Jason Liu <r64343@freescale.com> Cc: Ye Li <b37916@freescale.com> Cc: Fabio Estevam <fabio.estevam@freescale.com> Cc: Christian Gmeiner <christian.gmeiner@gmail.com> Cc: Markus Niebel <Markus.Niebel@tq-group.com> Cc: Peng Fan <b51431@freescale.com> Signed-off-by: Tim Harvey <tharvey@gateworks.com>
Peng Fan [Wed, 20 May 2015 02:28:48 +0000 (10:28 +0800)]
imx: dma: correct MXS_DMA_ALIGNMENT
We should not hardcode MXS_DMA_ALIGNMENT to 32, since we can not guarantee
that socs' cache line size is 32 bytes.
If on chips whose cache line size is 64 bytes, error occurs:
"
NAND: ERROR: v7_dcache_inval_range - start address is not aligned - 0xbdf1d1a0
ERROR: v7_dcache_inval_range - stop address is not aligned - 0xbdf1f4a0
ERROR: v7_dcache_inval_range - start address is not aligned - 0xbdf1d1a0
"
Align MXS_DMA_ALIGNMENT with ARCH_DMA_MINALIGN whose value is same to
CONFIG_SYS_CACHELINE_SIZE if CONFIG_SYS_CACHELINE_SIZE defined.
Signed-off-by: Peng Fan <Peng.Fan@freescale.com> Acked-by: Marek Vasut <marex@denx.de>
Masahiro Yamada [Mon, 18 May 2015 06:31:52 +0000 (15:31 +0900)]
ARM: zynq: add default ps7_init_gpl.c/h for Zed, MicroZed, ZC70x
Due to licensing issues, the files ps7_init.c/h are not able to be
distributed with U-Boot source code. Recent Xilinx tools also
provide the GPL variants (ps7_init_gpl.c/h), compatible with U-Boot
license.
Prior to this commit, we had to copy ps7_init files into
board/xilinx/zynq/ before the compile.
To be more user-friendly, let's include ps7_init_gpl.c/h for
Zedboard, MicroZed, ZC702, ZC706.
These init code have been taken from the hwplatform_templates
directory of Xilinx SDK 2014.4.
You can still use customized ps7_init_gpl.c/h by enabling
CONFIG_ZYNQ_CUSTOM_INIT. The recommended directory for storing them
is now board/xilinx/zynq/custom_hw_platform, but board/xilinx/zynq
is still supported for backward compatibility. The latter emits
a warning message to prompt users to gradually switch to the new
directory.
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com> Acked-by: Sören Brinkmann <soren.brinkmann@xilinx.com> Acked-by: Joe Hershberger <joe.hershberger@ni.com> Signed-off-by: Michal Simek <michal.simek@xilinx.com>
Masahiro Yamada [Mon, 18 May 2015 06:31:51 +0000 (15:31 +0900)]
ARM: zynq: add separate configuration for ZC702 and ZC706
Prior to this commit, ZC702 and ZC706 shared the same configuration
and were built as follows:
ZC702: make zynq_zc70x_defconfig && make
ZC706: make zynq_zc70x_defconfig && make DEVICE_TREE=zynq-zc706
This commit introduces separate configuration for them, which makes
the next commit much easier.
Going forward, the recommended build commands are:
ZC702: make zynq_zc702_defconfig && make
ZC706: make zynq_zc706_defconfig && make
Although the old work flow is still supported, CONFIG_TARGET_ZC70X
has been marked as deprecated. If used, the warning message is
shown to prompt users to switch to the new scheme.
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com> Acked-by: Joe Hershberger <joe.hershberger@ni.com> Signed-off-by: Michal Simek <michal.simek@xilinx.com>
Stefan Agner [Fri, 8 May 2015 17:07:13 +0000 (19:07 +0200)]
mtd: vf610_nfc: enable ONFI detection
This changes enable ONFI detection. The Read ID command now allows
one address byte which is needed for ONFI detection. To read the
ONFI parameter page, the NAND_CMD_PARAM need to be supported. The
CMD code enables one command and one address byte along with reading
data from flash using R/B#, as specified by ONFI.
Stefan Agner [Fri, 8 May 2015 17:07:12 +0000 (19:07 +0200)]
mtd: vf610_nfc: add 32-error correction option for HW ECC
Add option to choose between current 24-error correction and 32-error
correction through Kconfig. 32-error correction allow to use NAND
chips which require up to 8-bit error correction per 512 byte (when
using 2K pages).
Stefan Agner [Fri, 8 May 2015 17:07:11 +0000 (19:07 +0200)]
mtd: vf610_nfc: add Freescale NFC controller configs to Kconfig
This commit allows users to enable/disable the Freescale NFC
controller found in systems like Vybrid (VF610), MPC5125, MCF54418
or Kinetis K70 via Kconfig with more detailed help docs.
Signed-off-by: Stefan Agner <stefan@agner.ch> Acked-by: Stefano Babic <sbabic@denx.de>
[scottwood: updated vf610twr_nand_defconfig] Signed-off-by: Scott Wood <scottwood@freescale.com>
Stefan Agner [Fri, 8 May 2015 17:07:10 +0000 (19:07 +0200)]
mtd: vf610_nfc: use in-band bad block table
Use in-band bad block table (NAND_BBT_NO_OOB) which allows to
use the full OOB for hardare ECC purposes. Since there is no
ECC correction on the OOB it is also safer to use in-band area
to store the bad block table marker.
Stefan Agner [Fri, 8 May 2015 17:07:09 +0000 (19:07 +0200)]
mtd: vf610_nfc: implement OOB only read
Implement read of OOB area only. When using column and sector size
properties, only parts of the page can be read. However, this works
only when hardware ECC is disabled, otherwise the ECC engine would
ruin the data in the buffer. To allow OOB only reads, three points
had to be addressed:
- Set ECC mode per command.
- Handle NAND_CMD_READOOB seperate. Make sure column and sector
size is correctly set up, while disabling ECC.
- Now, the OOB data end up at the beginning of the buffer. Remove
the special handling of OOB (spareonly).
Especially bad block scans benefit from this change. On a 512MiB
SLC NAND device, the bad block scan took 1.5s less than before.
Stefan Agner [Fri, 8 May 2015 17:07:07 +0000 (19:07 +0200)]
mtd: vf610_nfc: remove read on SEQIN
Since we do not support sub-page writes anyway, reading the page
back to the controller on SEQIN command is not required. Remove
the page read on SEQIN.
However, the column/page values relevant to the SEQIN command, hence
set the column/row address on SEQIN command.
Stefan Agner [Fri, 8 May 2015 17:07:06 +0000 (19:07 +0200)]
mtd: vf610_nfc: remove caching of page in buffer
To improve performance we remember the current page in the buffer
and avoid reading it twice. This implicit page cache increases
complexity while does not increase performance in real world cases.
This patch removes that feature.
Acked-by: Bill Pringlemeir <bpringlemeir@nbsps.com> Signed-off-by: Stefan Agner <stefan@agner.ch>
projects / u-boot.git / log