After reboot, reset or even short power off, DRAM typically retains
the old stale data for some period of time (for this type of memory,
the bits of data are stored in slowly discharging capacitors).
The current sun6i/sun8i DRAM size detection logic, which is
inherited from the Allwinner code, relies on using a large magic
signature with the hope that it is unique enough and unlikely to
ever accidentally match this leftover garbage data in RAM. But
this approach is inherently unsafe, as can be demonstrated using
the following test program:
/***** A testcase for reproducing the problem ******/
void main(int argc, char *argv[])
{
size_t size, i;
uint32_t *buf;
/* Allocate the buffer */
if (argc < 2 || !(size = (size_t)atoi(argv[1]) * 1048576) ||
!(buf = malloc(size))) {
printf("Need buffer size in MiB as a cmdline argument\n");
exit(1);
}
/* Fill it with the Allwinner DRAM "magic" values */
for (i = 0; i < size / 4; i++)
buf[i] = 0xaa55aa55 + ((uintptr_t)&buf[i] / 4) % 64;
/* Try to reboot */
system("reboot");
/* And wait */
for (;;) {}
}
/***************************************************/
If this test program is run on the device (giving it a large
chunk of memory), then the DRAM size detection logic in u-boot
gets confused after reboot and fails to initialize DRAM properly.
A better approach is not to rely on luck and abstain from making
any assumptions about the properties of the leftover garbage
data in RAM. Instead just use a more reliable code for testing
whether two different addresses refer to the same memory location.
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>
Based on the register / dram_para headers from the Allwinner u-boot / linux
sources + the init sequences from boot0.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Ian Campbell <ijc@hellion.org.uk>
The sun8i boot0 code fills the DRAM with a "random" pattern before comparing
it at different offsets to do columns, etc. detection. The sun6i boot0 code
does not do it, instead relying on the memory contents being random enough
to begin with for the memcmp to properly detect the wrap-around address, iow
it is working purely by chance. Since our sun6i dram code was modelled after
the boot0 code it contained the same issue.
This commit fixes this by filling the memory with a unique, distinct pattern.
The new mctl_mem_fill function this introduces is added as an inline helper
in dram.h, so that it can be shared with the sun8i dram code.
While at it move mctl_mem_matches to dram.h for re-use in sun8i too.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Ian Campbell <ijc@hellion.org.uk>
The await_completion helper is already copy pasted between the sun4i and sun6i
dram code, and we need it for sun8i too, so lets make it an inline helper in
dram.h, rather then adding yet another copy.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Ian Campbell <ijc@hellion.org.uk>
The sun8i dram code sometimes wants to enable sigma delta mode,
add a parameter to allow this.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Ian Campbell <ijc@hellion.org.uk>
sun8i (A23) introduces a new bus for communicating with the pmic, the rsb,
the rsb is also used to communicate with the pmic on the A80, and is
documented in the A80 user manual.
This commit adds support for this based on the rsb driver from the allwinner
u-boot sources.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Ian Campbell <ijc@hellion.org.uk>
The p2wi interface is only available on sun6i, adjust the gpio pinmux and
base address defines for it to reflect this.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Ian Campbell <ijc@hellion.org.uk>
On sun6i the SID is stored in the pmic, rather then in the SoC itself,
add a helper function to abstract this away.
This makes our MAC address generation code also work on sun6i.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Ian Campbell <ijc@hellion.org.uk>
The sunxi mmc controller has both an internal clock divider, as well as
the divider in the mod0-clk for the mmc controller.
The internal divider cannot be used, as it conflicts with the setting of
clock sampling phases which is done in the mod0-clk, so it must be set to
0 (divide by 1).
For some reason while the kernel has had this correct from day one, the
u-boot sunxi mmc code has been using a fixed mod0-clk and setting its
internal divider depending on the desired speed. This is something which
we've inherited from the original Allwinner u-boot sources, but while this
has been fixed in Allwinner's own u-boot code at least for the A23 and later
upstream u-boot was still doing this wrong.
This commit fixes this, thereby also fixing mmc support not working reliable
on the A23 (which seems more sensitive to this) and possible also fixes some
other sunxi mmc issues.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Ian Campbell <ijc@hellion.org.uk>
Add a sunxi_get_ss_bonding_id() function, and use it to differentiate between
the A31s and the A31.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Ian Campbell <ijc@hellion.org.uk>
Switch to using hardware leveling for certain parameters on the EMIF
rather than using precalculated values. Doing this also means we have a
common place now between am437x and am335x for setting
emif_sdram_ref_ctrl with a value for the correct delay length.
Tested-by: Felipe Balbi <balbi@ti.com>
Tested-by: Tom Rini <trini@ti.com>
Signed-off-by: James Doublesin <doublesin@ti.com>
Signed-off-by: Felipe Balbi <balbi@ti.com>
Resynchronize memcpy/memset with kernel 3.17 and build them in
Thumb2 mode (unified syntax). Those assembler files can be built
and linked in ARM mode too, however when calling them from Thumb2
built code, the stack got corrupted and the copy did not succeed
(the exact details have not been traced back). However, the Linux
kernel builds those files in Thumb2 mode. Hence U-Boot should
build them in Thumb2 mode too when CONFIG_SYS_THUMB_BUILD is set.
To build the files without warning, some assembler instructions
had to be replaced with their UAL compliant variant (thanks
Jeroen for this input).
To build the file in Thumb2 mode the implicit-it=always option need
to be set to generate Thumb2 compliant IT instructions where needed.
We add this option to the general AFLAGS when building for Thumb2.
Reviewed-by: Simon Glass <sjg@chromium.org>
Tested-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Stefan Agner <stefan@agner.ch>
The low-level debugging functions are useful to debug the early boot
stage where the full UART driver is not available.
UniPhier SoCs need to initialize the UART port 0 to use this feature.
The initialization routine is called at the very entry of the
lowlevel_init().
Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
For NAND boot on PH1-LD4, PH1-sLD8, and some other SoCs,
the output of the system bus is disabled by default.
It must be enabled by software to have access to the system bus.
Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
pci ports are used as root complex in Linux. So set this as default
in u-boot for keystone devices
Signed-off-by: Murali Karicheri <m-karicheri2@ti.com>
Part of DMM logic is reuse from commit
47a4bea6af ("ARM: omap4: Update sdram
setting for panda rev A6") Which broke SDP4430 with ES2.3 (uses old
DDR).
So, to maintain support for newer DDR used in Panda ES rev B3, we
should, in addition to the commit
675cc77a3a ("ARM:OMAP4+: panda-es: Support
Rev B3 Elpida DDR2 RAM"), DDR timings, also do DMM configuration
specific to Panda.
Signed-off-by: Nishanth Menon <nm@ti.com>
Commit 8bc347e2ec "ARM: bootm: Allow booting in secure mode on hyp capable
systems" added the capability to select nonsec vs sec mode boot via an
environment var.
There is a subtle gotcha with this functionality, which is that the PSCI nodes
are still created in the fdt (via armv7_update_dt->fdt_psci) even when booting
in secure mode. Which means that if the kernel is PSCI aware then it will fail
to boot because it will try and do PSCI from secure world, which won't work.
This likely didn't get noticed before because the original purpose was to
support booting the legacy linux-sunxi kernels which don't understand PSCI.
To fix expose boot_nonsec (renaming with armv7_ prefix) outside of bootm.c and
use from the virt-dt code.
As well as avoiding the creation of the PSCI nodes we should also avoid
reserving the secure RAM, so do so.
Signed-off-by: Ian Campbell <ijc@hellion.org.uk>
Cc: Hans de Goede <hdegoede@redhat.com>
Cc: Albert ARIBAUD <albert.u.boot@aribaud.net>
Cc: Tom Rini <trini@ti.com>
Acked-by: Hans de Goede <hdegoede@redhat.com>
This training code provides run-time adjustment of DDR PHY parameters
for stable DDR operation.
Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
enable this clock with the following:
clk_usb_otg_enable((void *)HSOTG_BASE_ADDR)
Signed-off-by: Steve Rae <srae@broadcom.com>
Reviewed-by: Felipe Balbi <balbi@ti.com>
Add a board rev entry for the new model A+, and augment the board
rev error handling code to be a bit more verbose.
Signed-off-by: Stephen Warren <swarren@wwwdotorg.org>
There is currently a regression when using newer ARM64 compilers
for semihosting: the way long types are inferred from context
is no longer the same.
The semihosting runtime uses long and size_t, so use this
explicitly in the semihosting code and interface, and voila:
the code now works again.
Tested with aarch64-linux-gnu-gcc: Linaro GCC 4.9-2014.09.
Cc: Darwin Rambo <drambo@broadcom.com>
Cc: AKASHI Takahiro <takahiro.akashi@linaro.org>
Cc: Mark Hambleton <mark.hambleton@arm.com>
Cc: Tom Rini <trini@ti.com>
Acked-by: Steve Rae <srae@broadcom.com>
Suggested-by: Mark Hambleton <mark.hambleton@arm.com>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
The semihosting code exposes internal file handle handling
functions to read(), open(), close() and get the length of
a certain file handle.
However the code using it is only interested in either
reading and entire named file into memory or getting the
file length of a file referred by name. No file handles
are used.
Thus make the file handle code internal to this file by
removing these functions from the semihosting header file
and staticize them.
This gives us some freedom to rearrange the semihosting
code without affecting the external interface.
Cc: Darwin Rambo <drambo@broadcom.com>
Cc: AKASHI Takahiro <takahiro.akashi@linaro.org>
Cc: Mark Hambleton <mark.hambleton@arm.com>
Cc: Tom Rini <trini@ti.com>
Acked-by: Steve Rae <srae@broadcom.com>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Not all exynos 5420 based devices with an LCD also have a parade LVDS
bridge. So make sure compilation doesn't break if CONFIG_LCD is enabled
and CONFIG_VIDEO_PARADE is not.
As a side-effect move the parade functions from the exynos system header
file to its own file.
Signed-off-by: Sjoerd Simons <sjoerd.simons@collabora.co.uk>
Signed-off-by: Minkyu Kang <mk7.kang@samsung.com>
Implement an API that can be used by drivers to allocate memory from a
pool that is mapped uncached. This is useful if drivers would otherwise
need to do extensive cache maintenance (or explicitly maintaining the
cache isn't safe).
The API is protected using the new CONFIG_SYS_NONCACHED_MEMORY setting.
Boards can set this to the size to be used for the non-cached area. The
area will typically be right below the malloc() area, but architectures
should take care of aligning the beginning and end of the area to honor
any mapping restrictions. Architectures must also ensure that mappings
established for this area do not overlap with the malloc() area (which
should remain cached for improved performance).
While the API is currently only implemented for ARM v7, it should be
generic enough to allow other architectures to implement it as well.
Signed-off-by: Thierry Reding <treding@nvidia.com>
Signed-off-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Tom Warren <twarren@nvidia.com>
This controller was introduced on Tegra114 to handle XUSB pads. On
Tegra124 it is also used for PCIe and SATA pin muxing and PHY control.
Only the Tegra124 PCIe and SATA functionality is currently implemented,
with weak symbols on Tegra114.
Tegra20 and Tegra30 also provide weak symbols for these functions so
that drivers can use the same API irrespective of which SoC they're
being built for.
Signed-off-by: Thierry Reding <treding@nvidia.com>
Signed-off-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Tom Warren <twarren@nvidia.com>
Implement the powergate API that allows various power partitions to be
power up and down.
Signed-off-by: Thierry Reding <treding@nvidia.com>
Signed-off-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Tom Warren <twarren@nvidia.com>
This reset is required for PCIe and the corresponding ID therefore needs
to be defined. The enumeration value for this was properly defined on
some SoCs but not on others. Similarly, some contained it in the mapping
of peripheral IDs to clock IDs, other didn't. This patch defines it
consistently for all supported SoC generations.
Acked-by: Stephen Warren <swarren@nvidia.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
Signed-off-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Tom Warren <twarren@nvidia.com>
This function is required by PCIe and SATA. This patch implements it on
Tegra20, Tegra30 and Tegra124. It isn't implemented for Tegra114 because
it doesn't support PCIe or SATA.
Acked-by: Stephen Warren <swarren@nvidia.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
Signed-off-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Tom Warren <twarren@nvidia.com>
Implement an API that can be used by drivers to allocate memory from a
pool that is mapped uncached. This is useful if drivers would otherwise
need to do extensive cache maintenance (or explicitly maintaining the
cache isn't safe).
The API is protected using the new CONFIG_SYS_NONCACHED_MEMORY setting.
Boards can set this to the size to be used for the non-cached area. The
area will typically be right below the malloc() area, but architectures
should take care of aligning the beginning and end of the area to honor
any mapping restrictions. Architectures must also ensure that mappings
established for this area do not overlap with the malloc() area (which
should remain cached for improved performance).
While the API is currently only implemented for ARM v7, it should be
generic enough to allow other architectures to implement it as well.
Signed-off-by: Thierry Reding <treding@nvidia.com>
Signed-off-by: Simon Glass <sjg@chromium.org>
This converts all Tegra boards over to use driver model for I2C. The driver
is adjusted to use driver model and the following obsolete CONFIGs are
removed:
- CONFIG_SYS_I2C_INIT_BOARD
- CONFIG_I2C_MULTI_BUS
- CONFIG_SYS_MAX_I2C_BUS
- CONFIG_SYS_I2C_SPEED
- CONFIG_SYS_I2C
This has been tested on:
- trimslice (no I2C)
- beaver
- Jetson-TK1
It has not been tested on Tegra 114 as I don't have that board.
Acked-by: Heiko Schocher <hs@denx.de>
Signed-off-by: Simon Glass <sjg@chromium.org>
LS1 has 4 SMMUs for address translation of the masters. All the
SMMUs' stream IDs are 8-bit. The address translation depends on the
stream ID of the incoming transaction.
Each master has unique stream ID assigned to it and is configurable
through SCFG registers. The stream ID for the masters is identical
and share the same register field of STREAM ID registers.
Signed-off-by: Xiubo Li <Li.Xiubo@freescale.com>
Reviewed-by: York Sun <yorksun@freescale.com>
The Central Security Unit (CSU) allows secure world software to
change the default access control policies of peripherals/bus
slaves, determining which bus masters may access them. This
allows peripherals to be separated into distinct security domains.
Combined with SMMU configuration of the system masters privileges,
these features provide protection against indirect unauthorized
access to data.
For now we configure all the peripheral access permissions as R/W.
Signed-off-by: Xiubo Li <Li.Xiubo@freescale.com>
Reviewed-by: York Sun <yorksun@freescale.com>
Enable hypervisors utilizing the ARMv7 virtualization extension
on the LS1021A-QDS/TWR boards with the A7 core tile, we add the
required configuration variable.
Signed-off-by: Xiubo Li <Li.Xiubo@freescale.com>
Reviewed-by: York Sun <yorksun@freescale.com>
This patch adds NAND boot support for LS1021AQDS board. SPL
framework is used. PBL initialize the internal RAM and copy
SPL to it, then SPL initialize DDR using SPD and copy u-boot
from NAND flash to DDR, finally SPL transfer control to u-boot.
Signed-off-by: Prabhakar Kushwaha <prabhakar@freescale.com>
Signed-off-by: Alison Wang <alison.wang@freescale.com>
Reviewed-by: York Sun <yorksun@freescale.com>
This patch adds QSPI boot support for LS1021AQDS/TWR board.
The QSPI boot image need to be programmed into the QSPI flash
first. Then the booting will start from QSPI memory space.
Signed-off-by: Alison Wang <alison.wang@freescale.com>
Reviewed-by: York Sun <yorksun@freescale.com>
This patch adds SD boot support for LS1021AQDS board. SPL
framework is used. PBL initialize the internal RAM and copy
SPL to it, then SPL initialize DDR using SPD and copy u-boot
from SD card to DDR, finally SPL transfer control to u-boot.
Signed-off-by: Alison Wang <alison.wang@freescale.com>
Signed-off-by: Jason Jin <jason.jin@freescale.com>
Reviewed-by: York Sun <yorksun@freescale.com>
The patch changes PCIe dts node status to 'disabled' if the
corresponding controller is disabled according to serdes protocol.
Signed-off-by: Minghuan Lian <Minghuan.Lian@freescale.com>
Reviewed-by: York Sun <yorksun@freescale.com>
After wakeup from deep sleep, Clear EPU registers as early as possible
to prevent from possible issue. It's also safe to clear at normal boot.
Signed-off-by: Chenhui Zhao <chenhui.zhao@freescale.com>
Reviewed-by: York Sun <yorksun@freescale.com>
R-Mobile and R-Car ARM SoCs use sh_mmcif as MMC host driver.
This adds arch-rmobile/mmc.h that defines mmcif_mmc_init().
Signed-off-by: Nobuhiro Iwamatsu <nobuhiro.iwamatsu.yj@renesas.com>
Signed-off-by: Nobuhiro Iwamatsu <iwamatsu@nigauri.org>
stv0991 architecture support added. It contains the support for
following blocks
- Timer
- uart
Signed-off-by: Vikas Manocha <vikas.manocha@st.com>
[trini: Add arch/arm/cpu/armv7/Makefile hunk]
Signed-off-by: Tom Rini <trini@ti.com>
Some UniPhier boards are equipped with an expansion slot that
some optional SRAM/NOR-flash cards can be attached to. So, run-time
detection of the number of flash banks would be more user-friendly.
Until this commit, UniPhier boards have achieved this by (ab)using
board_flash_wp_on() because the boot failed if flash_size got zero.
Fortunately, this problem was solved by commit 70879a9256 (flash:
do not fail even if flash_size is zero).
Now it is possible to throw away such a tricky workaround. This
commit also enables CONFIG_SYS_MAX_FLASH_BANKS_DETECT for further
refactoring.
Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
Detect the board revision early during boot, and print the decoded
model name.
Eventually, this information can be used for tasks such as:
- Allowing/preventing USB device mode; some models have a USB device on-
board so only host mode makes sense. Others connect the SoC directly
to the USB connector, so device-mode might make sense.
- The on-board USB hub/Ethernet requires different GPIOs to enable it,
although luckily the default appears to be fine so far.
- The compute module contains an on-board eMMC device, so we could store
the environment there. Other models use an SD card and so don't support
saving the environment (unless we store it in a file on the FAT boot
partition...)
Set $fdtfile based on this information. At present, the mainline Linux
kernel doesn't contain a separate DTB for most models, but I hope that
will change soon.
Signed-off-by: Stephen Warren <swarren@wwwdotorg.org>
Reviewed-by: Simon Glass <sjg@chromium.org>
Tested-by: Simon Glass <sjg@chromium.org>
Correctly increment the base address of the freeze controller. And since
SYSMGR_FRZCTRL_VIOCTRL_SHIFT is not needed, remove it from the include file.
Signed-off-by: Dinh Nguyen <dinguyen@opensource.altera.com>
Cc: Vince Bridgers <vbridger@opensource.altera.com>
Cc: Chin Liang See <clsee@altera.com>
Cc: Marek Vasut <marex@denx.de>
Acked-by: Pavel Machek <pavel@denx.de>
Cc: Wolfgang Denk <wd@denx.de>
socfpga_scan_manager structure was missing a data member.
Signed-off-by: Dinh Nguyen <dinguyen@opensource.altera.com>
Cc: Vince Bridgers <vbridger@opensource.altera.com>
Cc: Chin Liang See <clsee@altera.com>
Cc: Marek Vasut <marex@denx.de>
Acked-by: Pavel Machek <pavel@denx.de>
Cc: Wolfgang Denk <wd@denx.de>
Re-map NAND&I2C boot-device to the "normal" NAND boot-device.
Otherwise the SPL boot IF can't handle this device correctly.
Somehow booting with Hynix 4GBit NAND H27U4G8 on Siemens
Draco leads to this boot-device passed to SPL from the BootROM.
With this change, Draco boots just fine into main U-Boot.
Signed-off-by: Stefan Roese <sr@denx.de>
Cc: Tom Rini <trini@ti.com>
Cc: Roger Meier <r.meier@siemens.com>
Cc: Samuel Egli <samuel.egli@siemens.com>
The infomation of module control register for R-Car ARM SoC (r8a7790, r8a7791,
r8a7793 and r8a7794) are almost the same, they can be combined into one
structure. This provides structure that summarizes infomation of module control
register and default register values.
And this structure is the module control use of the kernel at boot time.
Signed-off-by: Hisashi Nakamura <hisashi.nakamura.ak@renesas.com>
Signed-off-by: Nobuhiro Iwamatsu <nobuhiro.iwamatsu.yj@renesas.com>
Signed-off-by: Nobuhiro Iwamatsu <iwamatsu@nigauri.org>
Siarhei Siamashka
Hans de Goede
James Doublesin
Stefan Agner
Masahiro Yamada
Karicheri, Muralidharan
Nishanth Menon
Ian Campbell
Peng Fan
Steve Rae
Stephen Warren
Linus Walleij
Sjoerd Simons
Thierry Reding
Simon Glass
Xiubo Li
Alison Wang
Minghuan Lian
chenhui zhao
Nobuhiro Iwamatsu
Vikas Manocha
Dinh Nguyen
Stefan Roese