This will allow to have hw driver or asm optimised driver.
Use a priority level to determine which one to use at runtime.
The generic one will be 0.
Signed-off-by: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
GXemul supports MIPS Malta but with some limitations.
The most notable problems:
* PCI-enabled barebox just hangs under GXemul;
* by default gxemul uses 5Kc processor,
so we have to force MIPS32R1 processor usage.
To fix these problems the special gxemul-malta_defconfig
is introduced and the board's documentation is updated.
Signed-off-by: Antony Pavlov <antonynpavlov@gmail.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
qemu-malta.dts uses clock-frequency = <1843200>,
but debug_ll.h for Malta uses another value.
This patch fixes this inconsistency.
Signed-off-by: Antony Pavlov <antonynpavlov@gmail.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
This file originates in Linux. Linux has it under include/linux/
directory since commit dccd2304cc90.
Let's move it to the same place as well in barebox.
This commit was generated by the following commands:
find -name '*.[chS]' | xargs sed -i -e 's:<sizes.h>:<linux/sizes.h>:'
git mv include/sizes.h include/linux/
Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Currently, MIPS is the only architecture that needs
include/generated/asm-offsets.h, but we have got ./Kbuild file now.
It is a good reason to move asm-offsets.h rule from arch/mips/Makefile
to ./Kbuild and add dummy asm-offsets.c for the other architectures.
asm-offsets.h would be useful for all the architectures.
This commit does not implement include/generated/bounds.h,
but if necessary, it is easy to implement it.
Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
We can't just enable SYS_SUPPORTS_LITTLE_ENDIAN for successful
little-endian qemu-malta barebox build. Some byte swapping-related
macros are missed, e.g.:
arch/mips/mach-malta/pci.c: In function 'pcibios_init':
arch/mips/mach-malta/pci.c:218:28: error: 'GT_PCI0_CMD_MBYTESWAP_BIT'
undeclared (first use in this function)
GT_WRITE(GT_PCI0_CMD_OFS, GT_PCI0_CMD_MBYTESWAP_BIT |
^
This patch adds necessary macros definition.
Signed-off-by: Antony Pavlov <antonynpavlov@gmail.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
The latest Ingenic CPUs (e.g. JZ4780) use new vendor ID.
Based on commit from https://github.com/MIPS/CI20_linux/tree/ci20-v3.16
commit 00b672aa52f299f1d67ab18274c3f5e5d5a15767
Author: Paul Burton <paul.burton@imgtec.com>
Date: Mon Jul 8 12:14:28 2013 +0100
MIPS: add (another) Ingenic vendor ID
Ingenic have switched to a new vendor ID for the Xburst core used in
their current SoCs such as the jz4780. Add this vendor ID and handle it
in addition to their former vendor ID.
Signed-off-by: Antony Pavlov <antonynpavlov@gmail.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Old Ingenic SoCs (JZ4755, JZ4740) use MIPS32r1 ISA CPU cores
than latest ones (JZ4770, JZ4780) use MIPS32r2.
Signed-off-by: Antony Pavlov <antonynpavlov@gmail.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
With IOMEM() adapted for MIPS we can use physical addresses
in device tree reg property.
Signed-off-by: Antony Pavlov <antonynpavlov@gmail.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
To build an object file from a device tree currently we currently
have to add them to the Makefile twice, once to dtb-y and once to
obj-y. This patch introduces obj-dtb-y and pbl-dtb-y to directly
compile a device tree object file for inclusion in the barebox binary
or the pbl respectively.
The now unneeded dtb-y targets are removed.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
We have our file helper functions in several places. Move them
all to lib/libfile.c.
With this we no longer have file helpers in fs/fs.c which contains
the core fs functions and no functions in lib/libbb.c which are
not from busybox.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Also move reset_cpu() for jz4755 SoC from platform code
into the new driver code.
At the moment mach-xburst lacks clk support so jz4740 watchdog
driver looks like a template. We can improve jz4740 watchdog
driver later after adding clk support.
Signed-off-by: Antony Pavlov <antonynpavlov@gmail.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
We have many variants of the same bitops.h file. Consistently
use the same file for all architectures which completely use
the generic bitops versions.
Some architectures had static inline versions of functions
identically to the generic versions, these are removed and
the generic versions are used directly now.
Also several architectures depend on the generic find_*_bit
functions but didn't have the GENERIC_FIND_NEXT_BIT Kconfig
option selected. This is added where needed.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
The doxygen documentation is long outdated. Remove it. It will
be replaced with sphinx based documentation later.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
We already have 'select COMMON_CLK' so 'select HAVE_CLK'
is redundant.
Signed-off-by: Antony Pavlov <antonynpavlov@gmail.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Also we can enable m24c02 eeprom chip in dts-file e.g.
&i2c0 {
status = "okay";
eeprom: m24c02@50 {
compatible = "spd";
reg = <0x50>;
};
};
Alas! qemu mips malta spd m24c02 eeprom chip emulation is not perfect:
the block read operation does not work properly.
Here is an example.
If we read eeprom content byte-by-byte then there is no problem:
barebox:/ for i in 0x08 0x09 0x0a 0x0b 0x0c 0x0d 0x0e 0x0f ;
> do i2c_read -b 0 -a 0x50 -r $i -c 1 ; done
0x01
0x75
0x54
0x00
0x82
0x08
0x00
0x01
Compare this output with content of qemu.git/hw/mips/mips_malta.c:
static eeprom24c0x_t spd_eeprom = {
.contents = {
...
/* 00000008: */ 0x01,0x75,0x54,0x00,0x82,0x08,0x00,0x01,
But if we read several bytes at once the we have data corruption:
barebox:/ i2c_read -b 0 -a 0x50 -r 0x8 -c 8
0x01 0xff 0xff 0xff 0xff 0xff 0xff 0xff
Signed-off-by: Antony Pavlov <antonynpavlov@gmail.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>