pci_virt_to_mem() uses virt_to_phys() to get the physical address.
But pci_virt_to_mem() is also called with uncached addresses which
is wrong according to the documentation of virt_to_phys().
Refactor the PCI_TO_MEM() macro to optionally map an uncached address
back to a cached one before calling pci_virt_to_mem().
Currently pcnet works because virt_to_phys() is incorrectly implemented
on MIPS. With the upcoming asm header file update for MIPS, the
virt_to_phys() implementation will be fixed. Thus this patch is needed
to keep pcnet working on MIPS Malta board.
Signed-off-by: Daniel Schwierzeck <daniel.schwierzeck@gmail.com>
Add Kconfig symbol for L1 cache shift like the kernel does.
The value of CONFIG_SYS_CACHELINE_SIZE is not a reliable source
for ARCH_DMA_MINALIGN anymore, because it is optional on MIPS.
If CONFIG_SYS_CACHELINE_SIZE is not defined by a board, the
cache sizes are automatically detected and ARCH_DMA_MINALIGN
would be set to 128 Bytes.
The default value for CONFIG_MIPS_L1_CACHE_SHIFT is 5 which
corresponds to 32 Bytes. All current MIPS boards already used
that value. While on it, fix the Malta board to use a value of 6
like the kernel port does.
Signed-off-by: Daniel Schwierzeck <daniel.schwierzeck@gmail.com>
Use void pointers as address argument for readl( and writel()).
This is required for the upcoming MIPS asm header file and I/O
accessor update.
Signed-off-by: Daniel Schwierzeck <daniel.schwierzeck@gmail.com>
Use void pointer as address argument for readl(). This is required
for the upcoming MIPS asm header file and I/O accessor update.
Signed-off-by: Daniel Schwierzeck <daniel.schwierzeck@gmail.com>
config.h is used in target and host code and therefore should not
pull in target header files to avoid compile errors. This change
is also required for the MIPS header file sync done in follow-up
patches.
Signed-off-by: Daniel Schwierzeck <daniel.schwierzeck@gmail.com>
Move setup of -march to arch/mips/Makefile and follow the design on ARM.
Also add a possibility to chose specific CPU tune options.
Signed-off-by: Wills Wang <wills.wang@live.com>
Signed-off-by: Daniel Schwierzeck <daniel.schwierzeck@gmail.com>
Refactor machine setup like it is done on ARM. While on it,
also support "include <mach/file.h" for machine specific
header files.
Signed-off-by: Daniel Schwierzeck <daniel.schwierzeck@gmail.com>
Prepare sub-folder for device-tree files. Make support for
device-tree on MIPS available in Kbuild/Kconfig.
Signed-off-by: Daniel Schwierzeck <daniel.schwierzeck@gmail.com>
Signed-off-by: Purna Chandra Mandal <purna.mandal@microchip.com>
SPL binaries are usually linked to a fixed address in SRAM.
Furthermore SPL binaries do not need to relocate itself. Thus
do not build them as position-independent binaries which helps
to largely reduce the size of SPL binaries.
Signed-off-by: Daniel Schwierzeck <daniel.schwierzeck@gmail.com>
Put all functions coded in assembly in sub-sections of
section .text. This allows the linker to garbage collect
unused assembly functions too.
Signed-off-by: Daniel Schwierzeck <daniel.schwierzeck@gmail.com>
Correctly annotate _start and relocate_code as functions to
produce more readable disassembly code generated by objdump.
Signed-off-by: Daniel Schwierzeck <daniel.schwierzeck@gmail.com>
With more recent gcc versions we otherwise get an error like:
note: expected 'const struct sockaddr *' but argument is of type
'struct sockaddr_in *'
and the common solution here is to cast, rather than re-work the code.
Cc: Joe Hershberger <joe.hershberger@ni.com>
Cc: Simon Glass <sjg@chromium.org>
Signed-off-by: Tom Rini <trini@konsulko.com>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Acked-by: Joe Hershberger <joe.hershberger@ni.com>
Commit ecc30663 ("Fix board init code to respect the C runtime environment")
breaks x86. This was mentioned on https://patchwork.ozlabs.org/patch/548644
but not addressed. Correct it so that x86 boards boot again.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Tom Rini <trini@konsulko.com>
Adjust pci_rom_load() to return an indication of whether it allocated
memory or not. Adjust the caller to free it. This fixes a memory leak
when PCI_VGA_RAM_IMAGE_START is not used.
Reported-by: Coverity (CID: 134194)
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Tom Rini <trini@konsulko.com>
For this class it is intended to set up the PCI device, so add a comment to
indicate this. This avoids a coverity warning.
Reported-by: Coverity (CID: 134194)
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Tom Rini <trini@konsulko.com>
For SPL we don't really need sprintf() and with tiny-printf this is not
available. Allow this to be dropped in SPL when using tiny-printf.
Signed-off-by: Simon Glass <sjg@chromium.org>
the non-removable property point to sdcard before, it is wrong,
it must point to emmc, correct it.
Signed-off-by: Lin Huang <hl@rock-chips.com>
Acked-by: Simon Glass <sjg@chromium.org>
Tested-by: Simon Glass <sjg@chromium.org>
Rockchip has its own git repository maintained by Simon.
Document this.
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Acked-by: Simon Glass <sjg@chromium.org>
With format-security errors turned on, GCC picks up the use of sprintf with
a format parameter not being a string literal.
Simple uses of sprintf are also converted to use strcpy.
Signed-off-by: Ben Whitten <ben.whitten@gmail.com>
Acked-by: Wolfgang Denk <wd@denx.de>
Reviewed-by: Tom Rini <trini@konsulko.com>
Commit adc421e4ce "arm: move gd handling outside of C code" removed
the call to arch_setup_gd() on ARM and replaced it with assembly code
in crt0.S. However, AArch64 uses a different startup file, and the same
change was not made to it. This leaves gd uninitialized on AArch64, which
typically leads to hangs or crashes. This change fixes that.
Fixes: adc421e4ce ("arm: move gd handling outside of C code")
Signed-off-by: Stephen Warren <swarren@nvidia.com>
Add myself as custodian for the Marvell git repository. Additionally,
add the mach-mvebu directory to the list of files / directories. And
add Armada XP & Armada 38x to the title (not only kirkwood).
Signed-off-by: Stefan Roese <sr@denx.de>
Acked-by: Tom Rini <trini@konsulko.com>
Acked-by: Prafulla Wadaskar <prafulla@marvell.com>
Acked-by: Luka Perkov <luka.perkov@sartura.hr>
Synology keeps per item configuration in a dedicated 'partition' in SPI
flash, namely the one named 'vendor' in DTS file. It contains the two
NICs MAC addresses as well as the item's serial number. I didn't find a
way to have this information extracted automatically, therefore
implemented 'syno populate_env' command which extracts the three values
and puts them into environment. To make things permanent though, one has
to 'saveenv'.
Another command is 'syno clk_gate', which allows to change the clock
gating which is done in DS414 board file.
Signed-off-by: Phil Sutter <phil@nwl.cc>
Acked-by: Stefan Roese <sr@denx.de>
Reviewed-by: Tom Rini <trini@konsulko.com>
This adds support for the MV78230 based DS414 NAS by Synology. The
relevant bits have been extracted from the 'synogpl-5004-armadaxp'
package Synology kindly published, garnished with a fair amount of
trial-and-error.
Sadly, support is far from perfect. The major parts I have failed in
are SATA and XHCI support. Details about these and some other things
follow:
Device Tree
-----------
The device tree file armada-xp-synology-ds414.dts has been copied from
Linux and enhanced by recent U-Boot specific changes to
armada-xp-gp.dts.
SATA Support
------------
There is a Marvell 88SX7042 controller attached to PCIe which is
supported by Linux's sata_mv driver but sadly not U-Boot's sata_mv.
I'm not sure if extending the latter to support PCI devices is worth the
effort at all. Porting sata_mv from Linux exceeded my brain's
capacities. :(
XHCI Support
------------
There is an EtronTech EJ168A XHCI controller attached to PCIe which
drives the two rear USB3 ports. After a bit of playing around I managed
to get it recognized by xhci-pci, but never was able to access any
devices attached to it. Enabling it in ds414 board config shows that it
does not respond to commands for whatever reason. The (somewhat) bright
side to it is that it is not even supported in Synology's customized
U-Boot, but that also means nowhere to steal the relevant bits from.
EHCI Support
------------
This seems functional after issuing 'usb start'. At least it detects USB
storage devices, and IIRC reading from them was OK. OTOH Linux fails to
register the controller if 'usb start' wasn't given before in U-Boot.
According to Synology sources, this board seems to support USB device
(gadget?) mode. Though I didn't play around with it.
PCIe Support
------------
This is fine, but trying to gate the clocks of unused lanes will hang
PCI enum. In addition to that, pci_mvebu seems not to support DM_PCI.
DDR3 Training
-------------
Marvell/Synology uses eight PUPs instead of four. Does not look like
this is meant to be customized in mainline U-Boot at all. OTOH I have
no idea what a "PUP" actually is.
PEX Init
--------
Synology uses different values than mainline U-Boot with this patch:
pex_max_unit_get returns 2, pex_max_if_get returns 7 and
max_serdes_lines is set to 7. Not changing this seems to not have an
impact, although I'm not entirely sure it does not cause issues I am not
aware of.
Static Environment
------------------
This allows to boot stock Synology firmware at least. In order to be a
little more flexible when it comes to booting custom kernels, do not
only load zImage partition, but also rd.gz into memory. This way it is
possible to use about 7MB for kernel with piggyback initramfs.
Signed-off-by: Phil Sutter <phil@nwl.cc>
Acked-by: Stefan Roese <sr@denx.de>
Reviewed-by: Tom Rini <trini@konsulko.com>
This adds basic support for Marvell's MV78230 SoC which belongs to the
Armada XP series.
Signed-off-by: Phil Sutter <phil@nwl.cc>
Acked-by: Stefan Roese <sr@denx.de>
Reviewed-by: Tom Rini <trini@konsulko.com>
This should make it clear that this symbol is meant to be defined by
board headers.
Signed-off-by: Phil Sutter <phil@nwl.cc>
Acked-by: Stefan Roese <sr@denx.de>
Reviewed-by: Tom Rini <trini@konsulko.com>
This patch adds intermediate kconfig symbols which select their SoC
family. Boards then select them instead of the family symbol directly.
Signed-off-by: Phil Sutter <phil@nwl.cc>
Acked-by: Stefan Roese <sr@denx.de>
Reviewed-by: Tom Rini <trini@konsulko.com>
Instead of calling board_sat_r_get() only for those boards providing the
satr11 value via I2C, call it for all boards and return static values
for those not using I2C.
In addition to that, make this a weak function to allow for board code
to override it.
Signed-off-by: Phil Sutter <phil@nwl.cc>
Acked-by: Stefan Roese <sr@denx.de>
Reviewed-by: Tom Rini <trini@konsulko.com>
Armada XP has support for X4 lanes, boards specify this in their
serdes_cfg. During PEX init in high_speed_env_lib.c, the configuration
is stored in GEN_PURP_RES_2_REG.
When enumerating PEX, subsequent interfaces of an X4 lane must be
skipped. Otherwise the enumeration hangs up the board.
The way this is implemented here is not exactly beautiful, but it mimics
how Marvell's BSP does it. Alternatively we could get the information
using board_serdes_cfg_get(), but that won't lead to clean code, either.
Signed-off-by: Phil Sutter <phil@nwl.cc>
Acked-by: Stefan Roese <sr@denx.de>
Reviewed-by: Tom Rini <trini@konsulko.com>
If MV_DEBUG_WL is defined, DEBUG_WL_S and DEBUG_WL_D macros are missing.
In addition to that, get rid of debug output printing non-existent
counter variable.
Signed-off-by: Phil Sutter <phil@nwl.cc>
Acked-by: Stefan Roese <sr@denx.de>
Reviewed-by: Tom Rini <trini@konsulko.com>
* There is no boards.cfg anymore, so drop (1).
* Creating flash.c and u-boot.lds seems not mandatory as well.
* Adjusting the enumerators for the above implicitly fixed for
double items numbered (3).
Signed-off-by: Phil Sutter <phil@nwl.cc>
Acked-by: Stefan Roese <sr@denx.de>
Reviewed-by: Tom Rini <trini@konsulko.com>
The debug printing references bar_res, which exists only if
CONFIG_PCI_ENUM_ONLY is not defined. Therefore move it into the ifdef'd
area.
Signed-off-by: Phil Sutter <phil@nwl.cc>
Acked-by: Stefan Roese <sr@denx.de>
Reviewed-by: Tom Rini <trini@konsulko.com>
This patch adds runtime detection of the Marvell UART boot-mode (xmodem
protocol). If this boot-mode is detected, SPL will return to the
BootROM to continue the UART booting.
With this patch its now possible, to generate a U-Boot image that
can be booted either from the strapped boot-device (e.g. SPI NOR, MMC,
etc) or via the xmodem protocol from the UART. In the UART case,
the kwboot tool will dynamically insert the UART boot-device type
into the image. And also patch the load address in the header, so
that the mkimage header will be skipped (as its not expected by the
Marvell BootROM).
This simplifies the development for Armada XP / 38x based boards.
As no special images need to be generated by selecting the
MVEBU_BOOTROM_UARTBOOT Kconfig option.
Since the Kconfig option MVEBU_BOOTROM_UARTBOOT is not needed any
more, its now completely removed.
Signed-off-by: Stefan Roese <sr@denx.de>
Cc: Luka Perkov <luka.perkov@sartura.hr>
Cc: Dirk Eibach <dirk.eibach@gdsys.cc>
Cc: Phil Sutter <phil@nwl.cc>
Cc: Kevin Smith <kevin.smith@elecsyscorp.com>
This patch adds runtime boot-device detection to SPL U-Boot.
Signed-off-by: Stefan Roese <sr@denx.de>
Cc: Luka Perkov <luka.perkov@sartura.hr>
Cc: Dirk Eibach <dirk.eibach@gdsys.cc>
Cc: Phil Sutter <phil@nwl.cc>
Cc: Kevin Smith <kevin.smith@elecsyscorp.com>
As these structs are local only and const, declare them accordingly.
Signed-off-by: Stefan Roese <sr@denx.de>
Cc: Luka Perkov <luka.perkov@sartura.hr>
Cc: Dirk Eibach <dirk.eibach@gdsys.cc>
Cc: Phil Sutter <phil@nwl.cc>
Cc: Kevin Smith <kevin.smith@elecsyscorp.com>
This is preparation for the runtime bootmode detection in spl.c.
Signed-off-by: Stefan Roese <sr@denx.de>
Cc: Luka Perkov <luka.perkov@sartura.hr>
Cc: Dirk Eibach <dirk.eibach@gdsys.cc>
Cc: Phil Sutter <phil@nwl.cc>
Cc: Kevin Smith <kevin.smith@elecsyscorp.com>
This Makefile was not used since quite some time. I only missed to
remove it in the move to mach-mvebu. So lets remove it now so
that the mvebu-common directory is really removed completely.
Signed-off-by: Stefan Roese <sr@denx.de>
Cc: Luka Perkov <luka.perkov@sartura.hr>
Until now, the SoC selection for the ARCH_MVEBU platforms has been done
in the config header. Using CONFIG_ARMADA_XP in a non-clear way. As
it needed to get selected for AXP and A38x based boards. This patch
now changes this to move the SoC selection to Kconfig. And also
uses CONFIG_ARCH_MVEBU as a common define for both AXP and A38x.
This makes things a bit clearer - especially for new board additions.
Additionally the defines CONFIG_SYS_MVEBU_DDR_AXP and
CONFIG_SYS_MVEBU_DDR_A38X are replaced with the already available
CONFIG_ARMADA_38X and CONFIG_ARMADA_XP.
And CONFIG_DDR3 is removed, as its not referenced anywhere.
Signed-off-by: Stefan Roese <sr@denx.de>
Cc: Luka Perkov <luka.perkov@sartura.hr>
Merging all the board specific Kconfig options into the main Kconfig file
for mach-mvebu makes things easier to maintain.
Signed-off-by: Stefan Roese <sr@denx.de>
Cc: Luka Perkov <luka.perkov@sartura.hr>
With this patch, the CPU and the DDR frequencies will get printed in the
U-Boot startup messages. Resulting in such a log:
U-Boot 2016.01-rc2-00188-gb8eeaec-dirty (Dec 21 2015 - 12:32:35 +0100)
SoC: MV78460-B0 at 1600 MHz
I2C: ready
DRAM: 4 GiB (800 MHz, ECC not enabled)
...
Signed-off-by: Stefan Roese <sr@denx.de>
Cc: Luka Perkov <luka.perkov@sartura.hr>
Change some of the PEX configuration output lines from always output to
only ouput upon specific debug enabling.
This changes the SPL output from:
U-Boot SPL 2016.01-rc2-00037-g9353a7f (Dec 10 2015 - 10:27:42)
High speed PHY - Version: 2.1.5 (COM-PHY-V20)
Update Device ID PEX0782611ab
Update Device ID PEX1782611ab
Update Device ID PEX2782611ab
Update Device ID PEX3782611ab
Update Device ID PEX8782611ab
Update PEX Device ID 0x78260
High speed PHY - Ended Successfully
DDR3 Training Sequence - Ver 5.7.4
DDR3 Training Sequence - Ended Successfully
to:
U-Boot SPL 2016.01-rc2-00037-g9353a7f-dirty (Dec 10 2015 - 10:32:04)
High speed PHY - Version: 2.1.5 (COM-PHY-V20)
High speed PHY - Ended Successfully
DDR3 Training Sequence - Ver 5.7.4
DDR3 Training Sequence - Ended Successfully
Resulting in a little faster bootup time.
Signed-off-by: Stefan Roese <sr@denx.de>
Cc: Luka Perkov <luka.perkov@sartura.hr>
This is not needed any more since the switch to DM / DTS network
initialization on MVEBU. Lets remove it, as it otherwise leads
to compilation warning when CONFIG_NET is not enabled.
Signed-off-by: Stefan Roese <sr@denx.de>
Cc: Luka Perkov <luka.perkov@sartura.hr>
Currently, ECC support is enabled for all Armada XP boards. So the
DDR3 driver tries to configure the controller with ECC support, even
on boards without ECC. This patch makes this ECC optional which now
can be configured on a board-per-board basis.
Signed-off-by: Stefan Roese <sr@denx.de>
Cc: Luka Perkov <luka.perkov@sartura.hr>
Cc: Phil Sutter <phil@nwl.cc>
This patch adds support for the dual core Armada XP variant, the
MV78260. It has some minor differences to the 4-core MV78460,
e.g. only 12 serdes lanes.
Signed-off-by: Stefan Roese <sr@denx.de>
Cc: Luka Perkov <luka.perkov@sartura.hr>
Cc: Phil Sutter <phil@nwl.cc>
Using board ID 0 is reserved for the non-Marvell "custom" boards. So
move the board ID's to reflect this.
Signed-off-by: Stefan Roese <sr@denx.de>
Cc: Luka Perkov <luka.perkov@sartura.hr>
Cc: Phil Sutter <phil@nwl.cc>
Use the clrbits() / setbits() functions instead of clrsetbits() when
bit are only cleared or set.
Signed-off-by: Stefan Roese <sr@denx.de>
Cc: Luka Perkov <luka.perkov@sartura.hr>
