On 64bit architectures TLSF_64BIT should be defined. Decide that
depending on the CONFIG_64BIT symbol. When 64bit support is enabled
all allocations will be aligned to 8 byte boundaries which is necessary
for architectures that do not support unaligned accesses or have
a performance penalty for unaligned accesses.
Also changes the undefined tlsf_decl to 'static'.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Commit 2e6a88f210 (add cpu native ordered io accessors) introduced
these macros and then commit be57f20cdd (Fix big endian MMIO
primitives) figured out they are equivalent to __raw_{read,write}*.
They turned out unnecessary after all. Anyway, most source files
use __raw_read* and __raw_write*.
Let's replace a few remaining references and abolish them.
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
- Drop usage of 'ulong', use unsigned long
- remove unnecessary intermediate casting to unsigned long
- Use proper u8/u16/u32 types. Fixes 64bit systems on which ulong is 64bit
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Adding debug statements to this code path does not make much sense
since when enabled it will massively spam the console.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Using xfunctions to allocate log messages is not a good idea. Should
we be out of memory the xfunctions will panic which will cause another
allocation, so we deadlock the system with no message going out.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
To allow init code that depends on the environment being loaded.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
In commit
d384b5639f of: Drop devicetree merge support
the signature of of_unflatten_dtb() was changed, but the inline documentation
was forgotten to adjust. This patch removes the left over "root" argument from
the doc.
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
The variable name misses a CONFIG_ prefix to work as expected.
Fixes: 05a1e4b ARM: i.MX5: Do not register fixed clocks twice
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
This fixes: 80264a8 driver: Call bus->remove instead of driver->remove
On mvebu it happens that:
Upon device shutdown, when iterating through the active device list,
the phy0 device is removed before mdio-mvebu. Then, when the mdio bus
device is removed, the phy0 device is removed again, here:
mdio_bus_remove(on mdio-mvebu)
mvebu_mdio_remove
mdiobus_unregister
unregister_device
mdio_bus_remove(on phy0)
Fix this by setting the mdio busses phy_map[phy->addr] to NULL when
unregistering the phy device, so that mdiobus_unregister no longer
finds a valid phy_device when iterating over the busses device list.
Reported-by: Ezequiel Garcia <ezequiel@vanguardiasur.com.ar>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Tested-by: Ezequiel Garcia <ezequiel@vanguardiasur.com.ar>
This commit makes it possible to have unified debug_ll_*
assembler routines for differrent UART program model.
E.g. we can use debug_ll_putc() routine in common
code without knowledge on exact UART model: ns16550
or ar933x.
Also rename *_check_char -> *_tstc.
Signed-off-by: Antony Pavlov <antonynpavlov@gmail.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
When setting a device parameter fails not only the error code
is of interest, but also what the parameter name is, what the
device name is and whether a device was found for this name.
Add this information to the error message.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
USB requests should be processed within a timeframe of 5s.
The linux kernel also uses this timeout value for the control packets.
For the musb the increase in timeout helps successfully probing USB devices
(at least on one test hardware), but on the other hand doesn't slow down the
system, if there are no devices, i.e. the timeout never hits. That means,
that the USB core correctly aborts instead of timing out.
Why and how the 100ms was chosen in the first place, can not be deduced from
the git history, just that it seems to work fine for most boards.
Change USB_CNTL_TIMEOUT to 5000ms and replace all occurences of
`USB_CNTL_TIMEOUT * x`
with just
`USB_CNTL_TIMEOUT`,
as the timeout is now large enough without it.
Signed-off-by: Steffen Trumtrar <s.trumtrar@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
This fixes: 80264a8 driver: Call bus->remove instead of driver->remove
On mvebu it happens that:
Upon device shutdown, when iterating through the active device list,
the phy0 device is removed before mdio-mvebu. Then, when the mdio bus
device is removed, the phy0 device is removed again, here:
mdio_bus_remove(on mdio-mvebu)
mvebu_mdio_remove
mdiobus_unregister
unregister_device
mdio_bus_remove(on phy0)
Fix this by setting the mdio busses phy_map[phy->addr] to NULL when
unregistering the phy device, so that mdiobus_unregister no longer
finds a valid phy_device when iterating over the busses device list.
Reported-by: Ezequiel Garcia <ezequiel@vanguardiasur.com.ar>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
We have to start searching for the mtd node starting in the given root node not
the defaukt (= barebox internal) one. This means the internal DT is always
fixed up. This leads to booting not fixed up DT when using external DTs.
This patch fixes the problem.
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Acked-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
RGMII-ID just defines transmitter internal delays. Otherwise it is the
same as RGMII. As the differences only influence the behaviour of the
phy, the fec driver should handle it the same way as RGMII.
Signed-off-by: Markus Pargmann <mpa@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
The PCI core fills in the primary bus number, so there is no
need to walk up the PCI hierarchy in the driver manually.
Signed-off-by: Lucas Stach <dev@lynxeye.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
This was erroneously left uninitialzed as nothing was using it.
The i.MX6 PCI driver needs this to be filled properly to decide
if a config space access is allowed for a specific devfn.
This fixes PCI enumeration on the Gateworks GW54xx board.
Signed-off-by: Lucas Stach <dev@lynxeye.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
The bridge device needs to be registered and activated before
the scanning can proceed, as the bridge is the parent for other
devices.
This fixes a NULL ptr derefernce when scanning PCI hierarchies
with bridges behind bridges.
Signed-off-by: Lucas Stach <dev@lynxeye.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Since 2011 barebox' of_device_id struct uses unsigned long type for data field:
struct of_device_id {
char *compatible;
unsigned long data;
};
Almost always struct of_device_id.data field are used as pointer
and need 'unsigned long' casting.
E.g. see 'git grep -A 4 of_device_id drivers/' output:
drivers/ata/sata-imx.c:static __maybe_unused struct of_device_id imx_sata_dt_ids[] = {
drivers/ata/sata-imx.c- {
drivers/ata/sata-imx.c- .compatible = "fsl,imx6q-ahci",
drivers/ata/sata-imx.c- .data = (unsigned long)&data_imx6,
drivers/ata/sata-imx.c- }, {
Here is of_device_id struct in linux kernel v4.0:
struct of_device_id {
char name[32];
char type[32];
char compatible[128];
const void *data;
};
Changing of_device_id.data type to 'const void *data' will increase
barebox' linux kernel compatibility and decrease number of 'unsigned
long' casts.
Part of the patch was done using the 'coccinelle' tool with the
following semantic patch:
@rule1@
identifier dev;
identifier type;
identifier func;
@@
func(...) {
<...
- dev_get_drvdata(dev, (unsigned long *)&type)
+ dev_get_drvdata(dev, (const void **)&type)
...>
}
@rule2@
identifier dev;
identifier type;
identifier func;
identifier data;
@@
func(...) {
<...
- dev_get_drvdata(dev, (unsigned long *)&type->data)
+ dev_get_drvdata(dev, (const void **)&type->data)
...>
}
Signed-off-by: Antony Pavlov <antonynpavlov@gmail.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
The e1000 driver uses the dma coherent functions
and thus can only be build for architectures implementing them.
Here is mips malta build error log:
drivers/built-in.o: In function `e1000_transmit':
drivers/net/e1000.c:(.text.e1000_transmit+0x80): undefined reference to
`dma_sync_single_for_device'
drivers/built-in.o: In function `e1000_poll':
drivers/net/e1000.c:(.text.e1000_poll+0x60): undefined reference to
`dma_sync_single_for_cpu'
Signed-off-by: Antony Pavlov <antonynpavlov@gmail.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
This adds a simple check that any compatible strings in DeviceTree dts
files are present in Documentation/devicetree/bindings and
in dts/Bindings/.
Vendor prefixes are also checked for existing in vendor-prefixes.txt.
This code is based on linux v4.0-rc6 checkpatch.pl dt-checking code by
Rob Herring & Florian Vaussard:
Rob Herring (2):
checkpatch: add DT compatible string documentation checks
checkpatch: fix wildcard DT compatible string checking
Florian Vaussard (3):
checkpatch: fix spurious vendor compatible warnings
checkpatch: check compatible strings in .c and .h too
checkpatch: improve the compatible vendor match
Signed-off-by: Antony Pavlov <antonynpavlov@gmail.com>
Cc: Rob Herring <robh@kernel.org>
Cc: Florian Vaussard <florian.vaussard@epfl.ch>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
To get the number of address cells and size cells we have to use
the newly created partition node, not the parent device node. The
parent device node returns the address/size cells of the controller
node, not the partition node.
On an am335x machine this fixes the device tree passed to Linux.
The situation there is:
...
gpmc@50000000 {
...
#address-cells = <0x2>;
#size-cells = <0x1>;
ranges = <0x0 0x0 0x8000000 0x10000000>;
...
nand@0,0 {
reg = <0x0 0x0 0x0>;
#address-cells = <0x1>;
#size-cells = <0x1>;
...
partition@7 {
label = "system";
reg = <0x220000 0x7de0000>;
};
};
};
and without this patch barebox passes three bytes for
partition@7's reg property instead of only two which naturally
confuses Linux and yields to the system partition to start at 0
with a size of 0x220000.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Reported-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Tested-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
amba_{g,s}et_drvdata macros from linux kernel use
dev_{g,s}et_drvdata functions.
But in linux dev_get_drvdata() takes only one argument
while in barebox dev_get_drvdata() takes two arguments.
There is no dev_set_drvdata() in barebox at all.
So amba_{g,s}et_drvdata macros from linux are
unusable for barebox.
Signed-off-by: Antony Pavlov <antonynpavlov@gmail.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Hello!
This patch adds support for USB ethernet adapter D-Link DUB-E100 H/W Rev C1. I tested this on
Beagleboneblack Rev A6A and our custom h/w based on i.mx6 Phytec phycore module. More
information about the adapter is here:
http://us.dlink.com/products/connect/high-speed-usb-2-0-fast-ethernet-adapter/
Signed-off-by: Mayur Nande <mayur.nande@medel.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Update mvebu_defconfig to cover recently introduced Lenovo ix4 and
its related drivers.
Signed-off-by: Sebastian Hesselbarth <sebastian.hesselbarth@gmail.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
This adds support for Marvell Armada XP based 4-bay NAS Lenovo
Iomega ix4-300d.
Signed-off-by: Sebastian Hesselbarth <sebastian.hesselbarth@gmail.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Before adding new Armada XP based boards becomes messier than necessary,
sort Armada XP based board Kconfig and image Makefile alphabetically.
Signed-off-by: Sebastian Hesselbarth <sebastian.hesselbarth@gmail.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
