Masahiro Yamada [Tue, 24 Feb 2015 13:26:20 +0000 (22:26 +0900)]
kconfig: switch to single .config configuration
When Kconfig for U-boot was examined, one of the biggest issues was
how to support multiple images (Normal, SPL, TPL). There were
actually two options, "single .config" and "multiple .config".
After some discussions and thought experiments, I chose the latter,
i.e. to create ".config", "spl/.config", "tpl/.config" for Normal,
SPL, TPL, respectively.
It is true that the "multiple .config" strategy provided us the
maximum flexibility and helped to avoid duplicating CONFIGs among
Normal, SPL, TPL, but I have noticed some fatal problems:
[1] It is impossible to share CONFIG options across the images.
If you change the configuration of Main image, you often have to
adjust some SPL configurations correspondingly. Currently, we
cannot handle the dependencies between them. It means one of the
biggest advantages of Kconfig is lost.
[2] It is too painful to change both ".config" and "spl/.config".
Sunxi guys started to work around this problem by creating a new
configuration target. Commit cbdd9a9737cc (sunxi: kconfig: Add
%_felconfig rule to enable FEL build of sunxi platforms.) added
"make *_felconfig" to enable CONFIG_SPL_FEL on both images.
Changing the configuration of multiple images in one command is a
generic demand. The current implementation cannot propose any
good solution about this.
[3] Kconfig files are getting ugly and difficult to understand.
Commit b724bd7d6349 (dm: Kconfig: Move CONFIG_SYS_MALLOC_F_LEN to
Kconfig) has sprinkled "if !SPL_BUILD" over the Kconfig files.
[4] The build system got more complicated than it should be.
To adjust Linux-originated Kconfig to U-Boot, the helper script
"scripts/multiconfig.sh" was introduced. Writing a complicated
text processor is a shell script sometimes caused problems.
Now I believe the "single .config" will serve us better. With it,
all the problems above would go away. Instead, we will have to add
some CONFIG_SPL_* (and CONFIG_TPL_*) options such as CONFIG_SPL_DM,
but we will not have much. Anyway, this is what we do now in
scripts/Makefile.spl.
I admit my mistake with my apology and this commit switches to the
single .config configuration.
It is not so difficult to do that:
- Remove unnecessary processings from scripts/multiconfig.sh
This file will remain for a while to support the current defconfig
format. It will be removed after more cleanups are done.
- Adjust some makefiles and Kconfigs
- Add some entries to include/config_uncmd_spl.h and the new file
scripts/Makefile.uncmd_spl. Some CONFIG options that are not
supported on SPL must be disabled because one .config is shared
between SPL and U-Boot proper going forward. I know this is not
a beautiful solution and I think we can do better, but let's see
how much we will have to describe them.
- update doc/README.kconfig
More cleaning up patches will follow this.
Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Simon Glass [Tue, 24 Feb 2015 13:26:19 +0000 (22:26 +0900)]
kconfig: Adjust ordering so that defaults work as expected
At present defaults in arch-specific Kconfig files are ignored if the
top-level item comes ahead of it in include order. This means that it is
not possible to have a U-Boot default that architectures and boards can
override. This does not seem very useful.
Masahiro Yamada [Tue, 24 Feb 2015 13:26:18 +0000 (22:26 +0900)]
ARM: UniPhier: set CONFIG_SYS_MALLOC_F to the global default value
It is true that malloc is necessary for Driver Model before
relocation, but there is no good reason to reserve the malloc
space more than enough. The default value 0x400 works well.
Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com> Reviewed-by: Simon Glass <sjg@chromium.org>
OMAP3: igep0032: Fix regression due commit f3b4bc45.
Commit referenced in subject breaks IGEP0032 build with the following
error:
drivers/misc/status_led.c:30:7: error: 'RED_LED_GPIO' undeclared here (not in a function)
scripts/Makefile.build:275: recipe for target 'drivers/misc/status_led.o' failed
make[2]: *** [drivers/misc/status_led.o] Error 1
scripts/Makefile.build:420: recipe for target 'drivers/misc' failed
make[1]: *** [drivers/misc] Error 2
Makefile:1093: recipe for target 'drivers' failed
make: *** [drivers] Error 2
Fix this by skipping the status led on IGEP0032 machine as is not available
and throw an error for future machines if the status led is not configured
to avoid build breakage.
Reported-by: Albert ARIBAUD <albert.u.boot@aribaud.net> Signed-off-by: Enric Balletbo i Serra <enric.balletbo@collabora.com>
Matt Reimer [Thu, 19 Feb 2015 18:22:53 +0000 (11:22 -0700)]
mmc: sdhci: fix bus width switching on Samsung SoCs
Fix bus width switching from 8-bit mode down to 4-bit or 1-bit modes on
Samsung SoCs using SDHCI_QUIRK_USE_WIDE8. These SoCs report controller
version 2.0 yet they support 8-bit bus widths. If 8-bit mode was
previously enabled and then an operation like "mmc dev" caused a switch
back down to 4-bit or 1-bit mode, WIDE8 was left set, causing failures.
This problem was manifested by "mmc dev" timing out.
Signed-off-by: Matt Reimer <mreimer@sdgsystems.com>
mmc: exynos dwmmc: check boot mode before init dwmmc
Before this commit, the mmc devices were always registered
in the same order. So dwmmc channel 0 was registered as mmc 0,
channel 1 as mmc 1, etc.
In case of possibility to boot from more then one device,
the CONFIG_SYS_MMC_ENV_DEV should always point to right mmc device.
This can be achieved by init boot device as first, so it will be
always registered as mmc 0. Thanks to this, the 'saveenv' command
will work fine for all mmc boot devices.
Exynos based boards usually uses mmc host channels configuration:
- 0, or 0+1 for 8 bit - as a default boot device (usually eMMC)
- 2 for 4bit - as an optional boot device (usually SD card slot)
And usually the boot order is defined by OM pin configuration,
which can be changed in a few ways, eg.
- Odroid U3 - eMMC card insertion -> first boot from eMMC
- Odroid X2/XU3 - boot priority jumper
By this commit, Exynos dwmmc driver will check the OM pin configuration,
and then try to init the boot device and register it as mmc 0.
Hans de Goede [Thu, 19 Feb 2015 20:03:21 +0000 (21:03 +0100)]
sunxi: mmc: Always declare High Capacity capability
High Capacity (e)MMC cards work fine on sun4i / sun5i, and not having this
capability set causes u-boot to not recognize the eMMC on an Utoo P66 A13
tablet, so always set it thereby fixing this.
Signed-off-by: Hans de Goede <hdegoede@redhat.com> Acked-by: Ian Campbell <ijc@hellion.org.uk>
Jaehoon Chung [Wed, 4 Feb 2015 06:48:40 +0000 (15:48 +0900)]
mmc: exynos_dw_mmc: use the exynos specific data structure
Clksel value is exynos specific value.
It removed "clksel_val" into dwmci_host and created the
"dwmci_exynos_priv_data" structure for exynos specific data.
Jaehoon Chung [Wed, 4 Feb 2015 06:48:39 +0000 (15:48 +0900)]
mmc: exynos_dw_mmc: set to clksel_val into board-init function
"clksel_val" is assigned to property of mmc or defined value.
But it doesn't write at initial sequence.
There is a reason that get the wrong source-clock value.
This patch fixed it.
The SD/MMC version scheme was buggy when dealing with standard
major.minor.change cases. Fix it by using something similar to
the linux's kernel versioning method.
Signed-off-by: Pantelis Antoniou <pantelis.antoniou@konsulko.com> Tested-by: Jaehoon Chung <jh80.chung@samsung.com> Reported-by: Stephen Warren <swarren@nvidia.com> Tested-by: Stephen Warren <swarren@nvidia.com>
sunxi: Machine id hack to prevent loading buggy sunxi-3.4 kernels
Right now U-Boot supports the CONFIG_OLD_SUNXI_KERNEL_COMPAT option,
which makes it go out of its way in limiting the selection of PLL clock
frequencies and PMIC voltages in order not to upset outdated buggy
sunxi-3.4 kernel releases. And if the CONFIG_OLD_SUNXI_KERNEL_COMPAT
option is not set, then booting such old kernels exhibits various
failures at runtime. This is very user unfriendly, and there were
already several incidents when people wasted their time being hit
by these runtime failures and trying to debug them.
The right solution is not to add hacks and workarounds to the mainline
U-Boot, but to fix these bugs in the sunxi-3.4 kernel. And in fact,
the updated sunxi-3.4 kernels already exist. Still we need to follow
the 'Principle of Least Surprise' and U-Boot needs to ensure that
the old buggy kernels are not getting happily booted when the
CONFIG_OLD_SUNXI_KERNEL_COMPAT option is not set. And this patch
addresses this particular issue.
This patch makes U-Boot store the 'compatibility revision' number in
the top 4 bits of the machine id and pass it to the kernel. The old
buggy kernels will fail to load with a very much googlable error
message on the serial console (the "r1 = 0x100010bb" part of it):
"Error: unrecognized/unsupported machine ID (r1 = 0x100010bb)"
This error message can be documented in the linux-sunxi wiki with
proper explanations about how to resolve this situation and where
to get the necessary bugfixes for the sunxi-3.4 kernel.
The fixed sunxi-3.4 kernels implement a revision compatibility check
and clear the top 4 bits of the machine id if everything is alright.
By accepting the machine id with the bits 31:28 set to 1, the sunxi-3.4
kernel effectively certifies that it has the PLL5 clock speed and
AXP209 DCDC3 voltage fixes applied.
It is still possible to set the CONFIG_OLD_SUNXI_KERNEL_COMPAT option
in U-Boot if the user desires to use an outdated unpatched sunxi-3.4
kernel.
Signed-off-by: Siarhei Siamashka <siarhei.siamashka@gmail.com> Acked-by: Ian Campbell <ijc@hellion.org.uk> Acked-by: Hans de Goede <hdegoede@redhat.com> Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Hans de Goede [Fri, 20 Feb 2015 15:55:12 +0000 (16:55 +0100)]
sunxi: Set the /chosen/stdout-path fdt property for sunxi boards
While discussing with some people how to get the Linux kernel to do the
right thing wrt sending output to both the serial console and the
hdmi out / lcd screen when booting on ARM devices, Grant Likely pointed out
that there already is a solution for this.
All we need to do is set the /chosen/stdout-path fdt property, and if no
console= arguments were specified on the kernel commandline the kernel
will honor this and add this device as a console (next to the primary
video output on hdmi).
And u-boot already has support for setting this, all we need to do is
define OF_STDOUT_PATH and then everything will just work ootb, without
people needing to meddle with adding console= arguments in extlinux.conf .
Signed-off-by: Hans de Goede <hdegoede@redhat.com> Acked-by: Ian Campbell <ijc@hellion.org.uk> Reviewed-by: Tom Rini <trini@ti.com>
Hans de Goede [Thu, 19 Feb 2015 13:46:44 +0000 (14:46 +0100)]
sunxi: Fix sun5i mbus speed when booting old kernels
Older linux-sunxi-3.4 kernels override our PLL6 setting with 300 MHz,
halving the mbus frequency, so set it to 300 MHz ourselves and base the
mbus divider on that.
Signed-off-by: Hans de Goede <hdegoede@redhat.com> Acked-by: Ian Campbell <ijc@hellion.org.uk>
Hans de Goede [Mon, 16 Feb 2015 21:13:43 +0000 (22:13 +0100)]
sunxi: musb: Check Vbus-det before enabling otg port power
Sending out 5V when there is a charger connected to the otg port is not a
good idea, so check for this and error out.
Note this commit currently breaks otg support on the q8h tablets, as we need
to do some magic with the pmic there to get vbus info, this is deliberate
(better safe then sorry), fixing this is on my TODO list.
Signed-off-by: Hans de Goede <hdegoede@redhat.com> Acked-by: Ian Campbell <ijc@hellion.org.uk>
Hans de Goede [Thu, 19 Feb 2015 19:34:32 +0000 (20:34 +0100)]
sunxi: mmc: Always declare High Capacity capability
High Capacity (e)MMC cards work fine on sun4i / sun5i, and not having this
capability set causes u-boot to not recognize the eMMC on an Utoo P66 A13
tablet, so always set it thereby fixing this.
Signed-off-by: Hans de Goede <hdegoede@redhat.com> Acked-by: Ian Campbell <ijc@hellion.org.uk>
Stephen Warren [Mon, 16 Feb 2015 19:16:15 +0000 (12:16 -0700)]
rpi: add support for Raspberry Pi 2 model B
USB doesn't seem to work yet; the controller detects the on-board Hub/
Ethernet device but can't read the descriptors from it. I haven't
investigated yet.
Signed-off-by: Stephen Warren <swarren@wwwdotorg.org>
Stephen Warren [Mon, 16 Feb 2015 19:16:14 +0000 (12:16 -0700)]
bcm2836 SoC support (used in Raspberry Pi 2 model B)
The bcm2835 and bcm2836 are essentially identical, except:
- The CPU is an ARM1176 v.s. a quad-core Cortex-A7.
- The physical address of many IO controllers has moved.
Rather than introducing a whole new bcm2836 value for $(SOC) or $(ARCH),
update the existing bcm2835 code to handle the minor differences, and
plumb it into the ARMv7 CPU architecture.
Signed-off-by: Stephen Warren <swarren@wwwdotorg.org>
Masahiro Yamada [Fri, 20 Feb 2015 08:04:19 +0000 (17:04 +0900)]
ARM: prepare for including <mach/*.h>
This commit adds $(srctree)/arch/arm/$(machdirs)/include/mach to
the headers search path.
It allows us to replace "#include <asm/arch/foo.h>" with
"#include <mach/foo.h>". As "#include <asm/arch/foo.h>" is still
supported, we can modify each file one by one.
Masahiro Yamada [Fri, 20 Feb 2015 08:04:12 +0000 (17:04 +0900)]
kbuild: prepare for moving headers into mach-*/include/mach
In U-Boot, SoC-specific headers are placed in
arch/$(ARCH)/include/asm/arch-$(SOC) and a symbolic link to that
directory is created at the early stage of the build process.
Creating and removing a symbolic link during the build is not
preferred. In fact, Linux Kernel did away with include/asm-$(ARCH)
directories a long time time ago.
As for ARM, now it is possible to collect SoC sources into
arch/arm/mach-$(SOC). It is also reasonable to move SoC headers
into arch/arm/mach-$(SOC)/include/mach.
This commit prepares for that.
If the directory arch/$(ARCH)/mach-$(SOC)/include/mach exists,
a symbolic to that directory is created. Otherwise, a symbolic link
to arch/$(ARCH)/include/asm/arch-$(SOC) or arch-$(CPU) is created.
Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com> Tested-by: Simon Glass <sjg@chromium.org> [ on nyan-big ] Cc: Stephen Warren <swarren@nvidia.com> Cc: Tom Warren <twarren@nvidia.com>
Masahiro Yamada [Fri, 20 Feb 2015 08:04:02 +0000 (17:04 +0900)]
ARM: prepare for moving SoC sources into mach-*
In U-boot, the directory structure, arch/$(ARCH)/cpu/$(CPU)/$(SOC)/
has been adopted except that $(CPU) is missing from some
architectures and $(SOC) is missing from some CPUs.
This structure did not fit very well in some cases.
[1] AT91
AT91 SoC family have been developed across some ARM processor
generations. Generally speaking, some IPs are often re-used in the
same SoC family (same SoC vendor) even when the main processor is
updated. As a result, a SoC-common directory is needed in the upper
level. Currently, AT91 source files are placed as follows:
Once directories are split, the motivation for refactorings across
CPU directories is lost. Some files in arm920t/at91/ and
arm926ejs/at91/ are so similar that they could be merged.
[2] Tegra
Tegra is a little bit special case where different CPUs are used for
SPL and the main U-boot. To obey the arch/$(ARCH)/cpu/$(CPU)/$(SOC)
structure, the source files must be placed across the CPUs,
again SoC-common directory is necessary in the upper level.
Moreover, there are several families in Tegra: Tegra20, Tegra30,
Tegra114, Tegra124. Here again, the tegra-common directory is needed
to contain commonly-used files.
Tegra directories have been sprinkled in the directory structure.
Masahiro Yamada [Fri, 20 Feb 2015 08:04:01 +0000 (17:04 +0900)]
ARM: at91: move board select menu and common settings
The board select menu in arch/arm/Kconfig is still big.
To slim down it, this commit moves AT91 boards to
arch/arm/mach-at91/Kconfig.
Also, consolidate "config SYS_SOC" in each board Kconfig.
The Kconfig files under board/ directory were modified with the
following command:
find board -name Kconfig | xargs sed -i -e '
/config SYS_SOC/ {
N
/default "at91"/ {
N
d
}
}
'
Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com> Acked-by: Andreas Bießmann <andreas.devel@googlemail.co>
Marek Vasut [Wed, 18 Feb 2015 21:36:18 +0000 (22:36 +0100)]
dm: Protect device_unbind() with CONFIG_DM_DEVICE_REMOVE
Since device_unbind() is also defined in device-remove.c,
which is compiled in only in case CONFIG_DM_DEVICE_REMOVE
is defined, protect the device_unbind() prototype with the
same CONFIG_DM_DEVICE_REMOVE check.
Signed-off-by: Marek Vasut <marex@denx.de> Cc: Simon Glass <sjg@chromium.org> Cc: Stefan Roese <sr@denx.de> Cc: Tom Rini <trini@ti.com> Acked-by: Simon Glass <sjg@chromium.org>
Simon Glass [Tue, 27 Jan 2015 01:27:09 +0000 (18:27 -0700)]
serial: ns16550: Support debug UART
Add debug UART functions to permit ns16550 to provide an early debug UART.
Try to avoid using the stack so that this can be called from assembler before
a stack is set up (at least on ARM and PowerPC).
Simon Glass [Tue, 27 Jan 2015 01:27:08 +0000 (18:27 -0700)]
serial: ns16550: Add access functions that don't need platdata
For the debug UART we need to be able to provide any parameters before
driver model is set up. Add parameters to the low-level access functions
to make this possible.
Simon Glass [Tue, 27 Jan 2015 01:27:07 +0000 (18:27 -0700)]
serial: Support an early UART for debugging
This came up in a discussion on the mailing list here:
https://patchwork.ozlabs.org/patch/384613/
My concerns at the time were:
- it doesn't need to be written in assembler
- it doesn't need to be ARM-specific
This patch provides a possible alternative. It works by allowing any serial
driver to export one init function and provide a putc() function. These
can be used to output debug data before the real serial driver is available.
This implementation does not depend on driver model, and it is possible for
it to operate without a stack on some architectures (e.g. PowerPC, ARM). It
provides the same features as the ARM-specific debug.S but with more UART
and architecture support.
Simon Glass [Fri, 13 Feb 2015 19:20:48 +0000 (12:20 -0700)]
dm: Move CONFIG_I2C_COMPAT to Kconfig
Make this option available in Kconfig and clean up the board that uses it.
Note there is also an entry in exynos5-common.h but this affects multiple
boards and should be dropped as part of the Samsung I2C migration to
driver model.
Lubomir Popov [Fri, 30 Jan 2015 17:56:04 +0000 (19:56 +0200)]
cmd_i2c: Provide option for bulk 'i2c write' in one transaction
I2C chips do exist that require a write of some multi-byte data to occur in
a single bus transaction (aka atomic transfer), otherwise either the write
does not come into effect at all, or normal operation of internal circuitry
cannot be guaranteed. The current implementation of the 'i2c write' command
(transfer of multiple bytes from a memory buffer) in fact performs a separate
transaction for each byte to be written and thus cannot support such types of
I2C slave devices.
This patch provides an alternative by allowing 'i2c write' to execute the
write transfer of the given number of bytes in a single bus transaction if
the '-s' option is specified as a final command argument. Else the current
re-addressing method is used.
Signed-off-by: Lubomir Popov <l-popov@ti.com>
hs: adapt to CONFIG_DM_I2C
Joe Hershberger [Thu, 5 Feb 2015 03:56:54 +0000 (21:56 -0600)]
cmd_fdt: Print the control fdt in terms of virtual memory
If you want to inspect the control device tree using the fdt command,
the "fdt address -c" command previously unhelpfully printed the phys
memory address of the device tree. That address could not then be used
to set the fdt address for inspection. Changed the resulting print to
one that can be copied directly to the 'fdt address <addr>' command.
Signed-off-by: Joe Hershberger <joe.hershberger@ni.com> Acked-by: Simon Glass <sjg@chromium.org>
Before avr32 had an extra storage for stack end to have a nice stack printout
on exception. Remove this extra storage and use generic gd->start_addr_sp
instead.
Signed-off-by: Andreas Bießmann <andreas.devel@googlemail.com>
cpu_mmc_init() is required by the init sequence to have a working MMC interface
on avr32. This will not be included in the binary if we omit the avr32 board.c
when building the generic board.
Signed-off-by: Andreas Bießmann <andreas.devel@googlemail.com>
For example on a raspberry pi the u-boot environment can be
saved in a file on the first VFAT partition.
This example illustrates how to use it with fw_printenv/fw_setenv.
KM/IVM: split the IVM reading and parsing in 2 parts
This allows to first read the IVM content (earlier in the boot sequence)
and define the ethaddr env variable thanks to the ivm_read_eepromi().
Later, the IVM content can be parsed and used to define some hush
variables, when the hush subsystem is available thanks to
ivm_analyze_eeprom().
To avoid the HW read to happen twice, the buffer passed to
ivm_read_eeprom() has to be reused by ivm_analyze_eeprom (and thus
allocated before calling ivm_read_eeprom()).
Heiko Schocher [Fri, 6 Feb 2015 08:31:36 +0000 (09:31 +0100)]
spl, spl_nor: fix compiler warning
executing "tools/buildman/buildman mpc5xx" drops this warning:
common/spl/spl_nor.c: In function 'spl_nor_load_image':
common/spl/spl_nor.c:26:10: warning: assignment discards 'const' qualifier from pointer target type [enabled by default]
David Feng [Mon, 2 Feb 2015 08:53:13 +0000 (16:53 +0800)]
PCI: add 64-bit prefetchable memory support
PCI specification allow prefetchable memory to be 32-bit or 64-bit.
PCI express specification states that all memmory bars for prefetchable
memory must be implemented as 64-bit. They all require that 64 bit
prefetchble memory are suported especially when u-boot is ported to
more and more 64bit processors.
Signed-off-by: David Feng <fenghua@phytium.com.cn> Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
sunxi: Support the FEL boot mode in the regular u-boot build
So that the CONFIG_SPL_FEL option is not needed anymore. And the regular
SPL binary, generated by the default u-boot build, is now also bootable
over USB in the FEL mode. The SPL still can boot from the SD card too.
A bunch of system registers need to be saved/restored in order to ensure
that the IRQ handler still works in the BROM FEL code after getting
control back from the SPL. This is done in the sunxi code instead of
abusing ifdefs in 'start.S'.
The decision whether to load the main u-boot binary from the SD card or
return to the FEL code in the BROM is done at runtime.
Signed-off-by: Siarhei Siamashka <siarhei.siamashka@gmail.com>
[hdegoede@redhat.com: Since we now restore various regs before returning to
the FEL BROM code we can drop the sunxi specific #ifdefs in start.S] Acked-by: Hans de Goede <hdegoede@redhat.com> Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Simon Glass [Sat, 7 Feb 2015 17:47:30 +0000 (10:47 -0700)]
sunxi: Normalise FEL support
Make sunxi's FEL code fit with the normal U-Boot boot sequence instead of
creating its own. There are some #ifdefs required in start.S. Future work
will hopefully remove these.
This series is available at u-boot-dm, branch sunxi-working.
Signed-off-by: Simon Glass <sjg@chromium.org> Acked-by: Hans de Goede <hdegoede@redhat.com> Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Simon Glass [Sat, 7 Feb 2015 17:47:29 +0000 (10:47 -0700)]
arm: spl: Provide for a board-specific loader
Some boards have a special way of loading U-Boot that does not fit with
the existing SPL code. For example sunxi uses an 'FEL' mode where U-Boot
is loaded over USB. Add a CONFIG option and boot mode for this.
Signed-off-by: Simon Glass <sjg@chromium.org> Acked-by: Siarhei Siamashka <siarhei.siamashka@gmail.com> Acked-by: Hans de Goede <hdegoede@redhat.com> Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Simon Glass [Sat, 7 Feb 2015 17:47:28 +0000 (10:47 -0700)]
arm: Allow lr to be saved by board code
The link register value can be required on some boards (e.g. FEL mode on
sunxi) so use a branch instruction to jump to save_boot_params() instead
of a branch link.
This requires a branch back to save_boot_params_ret so adjust the users
to deal with this. For exynos just drop the function since it doesn't
do anything.
Signed-off-by: Simon Glass <sjg@chromium.org> Acked-by: Siarhei Siamashka <siarhei.siamashka@gmail.com> Acked-by: Hans de Goede <hdegoede@redhat.com> Signed-off-by: Hans de Goede <hdegoede@redhat.com>
tools: mksunxiboot: Fix problems on big endian systems
Now my PS3 can be also used to build u-boot for sunxi devices.
Signed-off-by: Siarhei Siamashka <siarhei.siamashka@gmail.com> Acked-by: Hans de Goede <hdegoede@redhat.com> Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Hans de Goede [Wed, 4 Feb 2015 11:14:56 +0000 (12:14 +0100)]
sunxi: dram: Un-inline dram helper functions
Move the dram helper functions to a separate C file, rather then having them
as inline helpers in dram.h. This saves 144 bytes in the .text segment for
sun6i builds.
Signed-off-by: Hans de Goede <hdegoede@redhat.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Only use CONFIG_SUNXI_MAX_FB_SIZE to reserve memory at the top when relocating
u-boot, and calculate the actual amount of memory necessary when setting up
the video-mode and use only that, freeing up some additional memory for use
by the kernel.
Signed-off-by: Hans de Goede <hdegoede@redhat.com> Acked-by: Ian Campbell <ijc@hellion.org.uk>
Hans de Goede [Mon, 2 Feb 2015 16:13:29 +0000 (17:13 +0100)]
sunxi: video: Do not use CONFIG_SYS_MEM_TOP_HIDE for the framebuffer
Do not use CONFIG_SYS_MEM_TOP_HIDE for the framebuffer, instead override
board_get_usable_ram_top to make sure that u-boot is not relocated into the
area where we want to use the framebuffer, and patch the devicetree from
sunxi_simplefb_setup() to tell the kernel to not touch the framebuffer.
This makes u-boot properly see the framebuffer as dram, and initalize the
level 2 cache for it, fixing the very slow cfb scrolling problem.
As an added bonus this stops us from reserving the framebuffer when simplefb
is not used because an older kernel is booted, or hdp is used and no hdmi
cable was plugged in, freeing up the memory for kernel use in these cases.
Reported-by: Siarhei Siamashka <siarhei.siamashka@gmail.com> Signed-off-by: Hans de Goede <hdegoede@redhat.com> Acked-by: Ian Campbell <ijc@hellion.org.uk>
Commit a93648d197df48fa46dd55f925ff70468bd81c71 introduced linker generated
lists for imagetool which is the base for some host tools (mkimage, dumpimage,
et al.). Unfortunately some host tool chains do not support the used type of
linker scripts. Therefore this commit broke these host-tools for them, namely
FreeBSD and Darwin (OS/X).
This commit tries to fix this. In order to have a clean distinction between host
and embedded code space we need to introduce our own linker generated list
instead of re-using the available linker_lists.h provided functionality. So we
copy the implementation used in linux kernel script/mod/file2alias.c which has
the very same problem (cause it is a host tool). This code also comes with an
abstraction for Mach-O binary format used in Darwin systems.
Signed-off-by: Andreas Bießmann <andreas.devel@googlemail.com> Cc: Guilherme Maciel Ferreira <guilherme.maciel.ferreira@gmail.com>
Vitaly Andrianov [Wed, 11 Feb 2015 19:05:41 +0000 (14:05 -0500)]
net: keystone_net: move serdes setup to initialization function
On Keystone2 devices serdes must be initialized before accessing MDIO bus.
This commit moves the keystone2_net_serdes_setup() from keystone2_eth_open
to keystone2_emac_initialize to meet that requirement.
This also eliminates unnecessary serdes initializatin every time when the
keystone2_eth_open is being called.
projects / u-boot / log