Add support for the QNAP TS-119 and TS-219 Turbo NAS devices.
Signed-off-by: Martin Michlmayr <tbm@cyrius.com>
Signed-off-by: Nicolas Pitre <nico@marvell.com>
With the exception of UART0, all MPP names are uppercase.
Signed-off-by: Martin Michlmayr <tbm@cyrius.com>
Signed-off-by: Nicolas Pitre <nico@marvell.com>
Hook up I2C on Marvell Kirkwood. Tested on a QNAP TS-219 which has
RTC connected through I2C.
Signed-off-by: Martin Michlmayr <tbm@cyrius.com>
Signed-off-by: Nicolas Pitre <nico@marvell.com>
Instead of having various pieces of defconfig files for different
platforms, let's group them into a single one.
Signed-off-by: Eric Miao <eric.miao@marvell.com>
"""The Marvell® PXA168 processor is the first in a family of application
processors targeted at mass market opportunities in computing and consumer
devices. It balances high computing and multimedia performance with low
power consumption to support extended battery life, and includes a wealth
of integrated peripherals to reduce overall BOM cost .... """
See http://www.marvell.com/featured/pxa168.jsp for more information.
1. Marvell Mohawk core is a hybrid of xscale3 and its own ARM core,
there are many enhancements like instructions for flushing the
whole D-cache, and so on
2. Clock reuses Russell's common clkdev, and added the basic support
for UART1/2.
3. Devices are a bit different from the 'mach-pxa' way, the platform
devices are now dynamically allocated only when necessary (i.e.
when pxa_register_device() is called). Description for each device
are stored in an array of 'struct pxa_device_desc'. Now that:
a. this array of device description is marked with __initdata and
can be freed up system is fully up
b. which means board code has to add all needed devices early in
his initializing function
c. platform specific data can now be marked as __initdata since
they are allocated and copied by platform_device_add_data()
4. only the basic UART1/2/3 are added, more devices will come later.
Signed-off-by: Jason Chagas <chagas@marvell.com>
Signed-off-by: Eric Miao <eric.miao@marvell.com>
1. add common GPIO handling code into [arch/arm/plat-pxa]
2. common code in <mach/gpio.h> moved into <plat/gpio.h>, new processors
should implement its own <mach/gpio.h>, provide the following required
definitions and '#include <plat/gpio.h>' in the end:
- GPIO_REGS_VIRT for mapped virtual address of the GPIO registers'
physical I/O memory
- macros of GPLR(), GPSR(), GPDR() for constant optimization for
functions gpio_{set,get}_value() (so that bit-bang code can still
have tolerable performance)
- NR_BUILTIN_GPIO for the number of onchip GPIO
- definitions of __gpio_is_inverted() and __gpio_is_occupied(), they
can be either macros or inlined functions
Signed-off-by: Eric Miao <eric.miao@marvell.com>
1. introduce folder of 'arch/arm/plat-pxa' for common code across different
PXA processor families
2. initially moved DMA code into plat-pxa
3. common code in <mach/dma.h> moved into <plat/dma.h>, new processors
should implement its own <mach/dma.h>, provide the following required
definitions and '#include <plat/dma.h>' in the end:
- DMAC_REGS_VIRT for mapped virtual address of the DMA registers'
physical I/O memory
Signed-off-by: Eric Miao <eric.miao@marvell.com>
This adds LCD functions for Colibri PXA300 and Colibri PXA320 and
configures a LQ043T3DX02 panel.
Original-code-by: Matthias Meier <matthias.j.meier@gmx.net>
Signed-off-by: Daniel Mack <daniel@caiaq.de>
Signed-off-by: Eric Miao <eric.miao@marvell.com>
This adds basic support for Colibri PXA320 modules.
The file colibri-320.c only contains settings specific to this module,
such as the Ethernet interface.
Cc: Matthias Meier <matthias.j.meier@gmx.net>
Signed-off-by: Daniel Mack <daniel@caiaq.de>
Signed-off-by: Eric Miao <eric.miao@marvell.com>
- Move common function for all Colibri PXA3xx boards to the newly
added colibri-pxa3xx.c
- Drop some unnecessary defines from colibri.h
- Make Kconfig reflect the fact that code for colibri 300 module does
also work for the 310 model
- Give up on the huge pin config table which was messed up with lots of
#ifdefs and switch over to locally defined tables for configured
functions
Cc: Matthias Meier <matthias.j.meier@gmx.net>
Signed-off-by: Daniel Mack <daniel@caiaq.de>
Signed-off-by: Eric Miao <eric.miao@marvell.com>
Broaden the AX88796 register mask to allow access to the reset register.
Remove unnecessary value definitions and the second resource block.
Diagnosed-by: Matthias Meier <matthias.j.meier@gmx.net>
Signed-off-by: Daniel Mack <daniel@caiaq.de>
Signed-off-by: Eric Miao <eric.miao@marvell.com>
This adds support for USB OHCI for Toradex' Colibri PXA300 modules as
connected on the evaluation board.
Signed-off-by: Daniel Mack <daniel@caiaq.de>
Signed-off-by: Eric Miao <eric.miao@marvell.com>
This patch add basic support for Toradex' Colibri PXA300 module.
Ethernet is enabled conditionally, depdending on CONFIG_AX88796.
Signed-off-by: Daniel Mack <daniel@caiaq.de>
Signed-off-by: Eric Miao <eric.miao@marvell.com>
Namespace cleanup: rename colibri.c to colibri-pxa270.c and change
some names in colibri.h.
Signed-off-by: Daniel Mack <daniel@caiaq.de>
Signed-off-by: Eric Miao <eric.miao@marvell.com>
The initial version of the DSA driver only supported a single switch
chip per network interface, while DSA-capable switch chips can be
interconnected to form a tree of switch chips. This patch adds support
for multiple switch chips on a network interface.
An example topology for a 16-port device with an embedded CPU is as
follows:
+-----+ +--------+ +--------+
| |eth0 10| switch |9 10| switch |
| CPU +----------+ +-------+ |
| | | chip 0 | | chip 1 |
+-----+ +---++---+ +---++---+
|| ||
|| ||
||1000baseT ||1000baseT
||ports 1-8 ||ports 9-16
This requires a couple of interdependent changes in the DSA layer:
- The dsa platform driver data needs to be extended: there is still
only one netdevice per DSA driver instance (eth0 in the example
above), but each of the switch chips in the tree needs its own
mii_bus device pointer, MII management bus address, and port name
array. (include/net/dsa.h) The existing in-tree dsa users need
some small changes to deal with this. (arch/arm)
- The DSA and Ethertype DSA tagging modules need to be extended to
use the DSA device ID field on receive and demultiplex the packet
accordingly, and fill in the DSA device ID field on transmit
according to which switch chip the packet is heading to.
(net/dsa/tag_{dsa,edsa}.c)
- The concept of "CPU port", which is the switch chip port that the
CPU is connected to (port 10 on switch chip 0 in the example), needs
to be extended with the concept of "upstream port", which is the
port on the switch chip that will bring us one hop closer to the CPU
(port 10 for both switch chips in the example above).
- The dsa platform data needs to specify which ports on which switch
chips are links to other switch chips, so that we can enable DSA
tagging mode on them. (For inter-switch links, we always use
non-EtherType DSA tagging, since it has lower overhead. The CPU
link uses dsa or edsa tagging depending on what the 'root' switch
chip supports.) This is done by specifying "dsa" for the given
port in the port array.
- The dsa platform data needs to be extended with information on via
which port to reach any given switch chip from any given switch chip.
This info is specified via the per-switch chip data struct ->rtable[]
array, which gives the nexthop ports for each of the other switches
in the tree.
For the example topology above, the dsa platform data would look
something like this:
static struct dsa_chip_data sw[2] = {
{
.mii_bus = &foo,
.sw_addr = 1,
.port_names[0] = "p1",
.port_names[1] = "p2",
.port_names[2] = "p3",
.port_names[3] = "p4",
.port_names[4] = "p5",
.port_names[5] = "p6",
.port_names[6] = "p7",
.port_names[7] = "p8",
.port_names[9] = "dsa",
.port_names[10] = "cpu",
.rtable = (s8 []){ -1, 9, },
}, {
.mii_bus = &foo,
.sw_addr = 2,
.port_names[0] = "p9",
.port_names[1] = "p10",
.port_names[2] = "p11",
.port_names[3] = "p12",
.port_names[4] = "p13",
.port_names[5] = "p14",
.port_names[6] = "p15",
.port_names[7] = "p16",
.port_names[10] = "dsa",
.rtable = (s8 []){ 10, -1, },
},
},
static struct dsa_platform_data pd = {
.netdev = &foo,
.nr_switches = 2,
.sw = sw,
};
Signed-off-by: Lennert Buytenhek <buytenh@marvell.com>
Tested-by: Gary Thomas <gary@mlbassoc.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
It would seem when building kernel modules with modern binutils
(required by modern GCC) for ARM v4T targets (specifically observed
with the Samsung 24xx SoC which is an 920T) R_ARM_V4BX relocations
are emitted for function epilogues.
This manifests at module load time with an "unknown relocation: 40"
error message.
The following patch adds the R_ARM_V4BX relocation to the ARM kernel
module loader. The relocation operation is taken from that within the
binutils bfd library.
Signed-off-by: Simtec Linux Team <linux@simtec.co.uk>
Signed-off-by: Vincent Sanders <vince@simtec.co.uk>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Signed-off-by: Mike Rapoport <mike@compulab.co.il>
Acked-by: Eric Miao <eric.miao@marvell.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
This is a device driver for the NAND flash controller found on the various
DaVinci family chips. It handles up to four SoC chipselects, and some
flavors of secondary chipselect (e.g. based on upper bits of the address
bus) as used with some multichip packages. (Including the 2 GiB chips
used on some TI devel boards.)
The 1-bit ECC hardware is supported (3 bytes ECC per 512 bytes data); but
not yet the newer 4-bit ECC (10 bytes ECC per 512 bytes data), as
available on chips like the DM355 or OMAP-L137 and needed with the more
error-prone MLC NAND chips.
This is a cleaned-up version of code that's been in use for several years
now; sanity checked with the new drivers/mtd/tests.
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Sudhakar Rajashekhara <sudhakar.raj@ti.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
OMAP wishes to pass state to the boot loader upon reboot in order to
instruct it whether to wait for USB-based reflashing or not. There is
already a facility to do this via the reboot() syscall, except we ignore
the string passed to machine_restart().
This patch fixes things to pass this string to arch_reset(). This means
that we keep the reboot mode limited to telling the kernel _how_ to
perform the reboot which should be independent of what we request the
boot loader to do.
Acked-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
clk_add_alias is commonly called for platform devices that are not yet
registered in the device tree. Thus the clock alias is associated with
NULL device name. Fix this by passing the device name instead of just
device pointer.
Signed-off-by: Dmitry Eremin-Solenikov <dbaryshkov@gmail.com>
Signed-off-by: Eric Miao <eric.miao@marvell.com>
Map unused registers at the end of DMA region at 64 MB to allow PCI masters
to cross the boundary when prefetching data from SDRAM.
Signed-off-by: Krzysztof Hałasa <khc@pm.waw.pl>
Ideally, the default should be set to 0 and let the EHCI driver turn
it on as needed. This makes USB usable in the mean time.
Signed-off-by: Nicolas Pitre <nico@marvell.com>
Here it is... HIGHMEM for the ARM architecture. :-)
If you don't have enough ram for highmem pages to be allocated and still
want to test this, then the cmdline option "vmalloc=" can be used with
a value large enough to force the highmem threshold down.
Successfully tested on a Marvell DB-78x00-BP Development Board with
2 GB of RAM.
Signed-off-by: Nicolas Pitre <nico@marvell.com>
VIPT aliasing caches have issues of their own which are not yet handled.
Usage of discard_old_kernel_data() in copypage-v6.c is not highmem ready,
kmap/fixmap stuff doesn't take account of cache colouring, etc.
If/when those issues are handled then this could be reverted.
Signed-off-by: Nicolas Pitre <nico@marvell.com>
On xsc3, L2 cache ops are possible only on virtual addresses. The code
is rearranged so to have a linear progression requiring the least amount
of pte setups in the highmem case. To protect the virtual mapping so
created, interrupts must be disabled currently up to a page worth of
address range.
The interrupt disabling is done in a way to minimize the overhead within
the inner loop. The alternative would consist in separate code for
the highmem and non highmem compilation which is less preferable.
Signed-off-by: Nicolas Pitre <nico@marvell.com>
The choice is between looping over the physical range and performing
single cache line operations, or to map highmem pages somewhere, as
cache range ops are possible only on virtual addresses.
Because L2 range ops are much faster, we go with the later by factoring
the physical-to-virtual address conversion and use a fixmap entry for it
in the HIGHMEM case.
Possible future optimizations to avoid the pte setup cost:
- do the pte setup for highmem pages only
- determine a threshold for doing a line-by-line processing on physical
addresses when the range is small
Signed-off-by: Nicolas Pitre <nico@marvell.com>
If a machine class has a custom __virt_to_bus() implementation then it
must provide a __arch_page_to_dma() implementation as well which is
_not_ based on page_address() to support highmem.
This patch fixes existing __arch_page_to_dma() and provide a default
implementation otherwise. The default implementation for highmem is
based on __pfn_to_bus() which is defined only when no custom
__virt_to_bus() is provided by the machine class.
That leaves only ebsa110 and footbridge which cannot support highmem
until they provide their own __arch_page_to_dma() implementation.
But highmem support on those legacy platforms with limited memory is
certainly not a priority.
Signed-off-by: Nicolas Pitre <nico@marvell.com>
This is a helper to be used by the DMA mapping API to handle cache
maintenance for memory identified by a page structure instead of a
virtual address. Those pages may or may not be highmem pages, and
when they're highmem pages, they may or may not be virtually mapped.
When they're not mapped then there is no L1 cache to worry about. But
even in that case the L2 cache must be processed since unmapped highmem
pages can still be L2 cached.
Signed-off-by: Nicolas Pitre <nico@marvell.com>
The kmap virtual area borrows a 2MB range at the top of the 16MB area
below PAGE_OFFSET currently reserved for kernel modules and/or the
XIP kernel. This 2MB corresponds to the range covered by 2 consecutive
second-level page tables, or a single pmd entry as seen by the Linux
page table abstraction. Because XIP kernels are unlikely to be seen
on systems needing highmem support, there shouldn't be any shortage of
VM space for modules (14 MB for modules is still way more than twice the
typical usage).
Because the virtual mapping of highmem pages can go away at any moment
after kunmap() is called on them, we need to bypass the delayed cache
flushing provided by flush_dcache_page() in that case.
The atomic kmap versions are based on fixmaps, and
__cpuc_flush_dcache_page() is used directly in that case.
Signed-off-by: Nicolas Pitre <nico@marvell.com>
This is the minimum fixmap interface expected to be implemented by
architectures supporting highmem.
We have a second level page table already allocated and covering
0xfff00000-0xffffffff because the exception vector page is located
at 0xffff0000, and various cache tricks already use some entries above
0xffff0000. Therefore the PTEs covering 0xfff00000-0xfffeffff are free
to be used.
However the XScale cache flushing code already uses virtual addresses
between 0xfffe0000 and 0xfffeffff.
So this reserves the 0xfff00000-0xfffdffff range for fixmap stuff.
The Documentation/arm/memory.txt information is updated accordingly,
including the information about the actual top of DMA memory mapping
region which didn't match the code.
Signed-off-by: Nicolas Pitre <nico@marvell.com>
* master.kernel.org:/home/rmk/linux-2.6-arm: (23 commits)
[ARM] Fix virtual to physical translation macro corner cases
[ARM] update mach-types
[ARM] 5421/1: ftrace: fix crash due to tracing of __naked functions
MX1 fix include
[ARM] 5419/1: ep93xx: fix build warnings about struct i2c_board_info
[ARM] 5418/1: restore lr before leaving mcount
ARM: OMAP: board-omap3beagle: set i2c-3 to 100kHz
ARM: OMAP: Allow I2C bus driver to be compiled as a module
ARM: OMAP: sched_clock() corrected
ARM: OMAP: Fix compile error if pm.h is included
[ARM] orion5x: pass dram mbus data to xor driver
[ARM] S3C64XX: Fix s3c64xx_setrate_clksrc
[ARM] S3C64XX: sparse warnings in arch/arm/plat-s3c64xx/irq.c
[ARM] S3C64XX: sparse warnings in arch/arm/plat-s3c64xx/s3c6400-clock.c
[ARM] S3C64XX: Fix USB host clock mux list
[ARM] S3C64XX: Fix name of USB host clock.
[ARM] S3C64XX: Rename IRQ_UHOST to IRQ_USBH
[ARM] S3C64XX: Do gpiolib configuration earlier
[ARM] S3C64XX: Staticise s3c64xx_init_irq_eint()
[ARM] SMDK6410: Declare iodesc table static
...
Since now ipaq_model_ops used only for accessing h3600 EGPIOs,
drop it completely and use assign_h3600_egpio() directly.
Signed-off-by: Dmitry Artamonow <mad_soft@inbox.ru>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Replace all occurences with assign_h3600_egpio.
Also simplify code a bit by replacing couple of if-else
statements with one-line equivalents.
Signed-off-by: Dmitry Artamonow <mad_soft@inbox.ru>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Remove unused fields and associated funtions-accesors.
Signed-off-by: Dmitry Artamonow <mad_soft@inbox.ru>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Right now iPaq h3600's default MTD partitions table is a mess. It has
two #ifdefs with #else, giving total 3 variants, depending on your
kernel config. Replace all this with simple two-partitions scheme
(bootloader + rootfs), that used by both shipped WindowsCE and
most of the linux distributions (Familiar, Angstrom)
Signed-off-by: Dmitry Artamonow <mad_soft@inbox.ru>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
There's no actual code for iPAQ sleeves support in kernel that depends
on this config option.
Signed-off-by: Dmitry Artamonow <mad_soft@inbox.ru>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
This patch adds basic support for Dave/DENX QongEVB-LITE i.MX31-based
board. It includes support for clocks initialization, UART1, NOR-flash,
FPGA-attached NAND flash and DNET ethernet controller (inside FPGA).
Signed-off-by: Ilya Yanok <yanok@emcraft.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
This patch adds clkdev support for i.MX31. This is done in a
similar way done previously for i.MX27
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
The UART3 had a copy-paste bug. instead of claiming rxd, txd, rts and
cts pins, cts and rts were claimed twice
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
On MX31 we can't do much without mapping the AIPS1/2 register space.
Move these mappings from individual boards to plat-mxc/mm.c
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
The i.MX35 basically features the same peripherals as the i.MX31 with
some differences:
- The i.MX35 has a FEC ethernet controller
- The NAND controller base addresses are different
- The i.MX35 has only 3 UARTs
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
This patch adds clock support for i.MX35 SoCs. We do not support setting
of clock rates yet, but most interesting clock rates should be reported.
I couldn't test all clock rates and the datasheet contains some obvious
bugs, so expect some bugs in this code.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
We had hardcoded cpu_is_ macros for mxc architectures till now. As we
want to run the same kernel on i.MX31 and i.MX35 this patch adds cpu_is_
macros which expand to 0 or 1 if only one architecture is compiled in and
only check for the cpu type if more than one architecture is compiled
in.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
This patch moves the stuff common to i.MX31 and i.MX35 to mx3x.h and the
specifics to mx31.h/mx35.h. We can build a kernel which runs on i.MX31 and
i.MX35, so always include mx31.h and mx35.h
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
This adds the dma (ipu_dma) and fb devices for the mx31 for which drivers now are
available.
v2: merge the ipu and fb device in the same patch as suggested by Sascha
Signed-off-by: Valentin Longchamp <valentin.longchamp@epfl.ch>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
This enables our mx31moboard to be used on the different baseboards that
we are developping according to the application needs. There are not
many differences between the boards for now, but when other peripherals
are available for mx31 the differences are going to grow.
v2: takes Sascha's comments into account
Signed-off-by: Valentin Longchamp <valentin.longchamp@epfl.ch>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Make sure not to create spurious pulses on GPIOs, when configuring them as
output: first set required level, then switch direction.
Signed-off-by: Guennadi Liakhovetski <lg@denx.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
This driver has been tested on MX27/MX31. It should work on MX1/MX1
aswell, but the actual setting of the PWM is missing so far.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
The in kernel FEC driver has recently been ported to a platform driver.
Add a platform_device for it and register it for pcm038 and mx27ads.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
The i.MX31ADS supports pluggable PMU modules, including the WM835x based
Wolfson Microelectronics 1133-EV1. These boards provide power, audio,
RTC and watchdiog services to the system. This patch adds initial support
for those boards in I2C mode.
Currently support is limited by the available support for the features
of the i.MX31 in the mainline kernel. Some further work will be needed
once other PMU modules are supported and once there is SPI support.
Many of the regulator constraints will be sharable with other PMU
boards.
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
This example takes advantage of the possibility to use tables of iomux
configs.
This is inspired from mx1-mx2 iomux code. It allows a better code
readability.
Signed-off-by: Valentin Longchamp <valentin.longchamp@epfl.ch>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
This new implemenatation avoids that two physical pins are claimed by
the same driver (also with the the gpr hardware modes).
The gpio kernel lib is also called when a capable gpio pin is assigned
its gpio function.
The mxc_iomux_mode function is still here for backward compatibility but
should not be used anymore.
V2:
In the precendent revision, the iomux code was claiming a pin when its
hardware mode was changed. This was uncorrect: when the hardware mode is
changed, the pin must still be claimed through the iomux.
In order to have a pin working in mode hw2, we must fist issue the
mxc_iomux_set_gpr call and then the corresponding mxc_iomux_mode calls
with the FUNC mode (usually done with mxc_iomux_setup_multiple_pins).
The reverse calls must be done to fee the pins.
Signed-off-by: Valentin Longchamp <valentin.longchamp@epfl.ch>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
... from both mx27ads.c and pcm038.c
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Cc: Holger Schurig <hs4233@mail.mn-solutions.de>
Cc: Russell King <rmk+kernel@arm.linux.org.uk>
Cc: Sascha Hauer <s.hauer@pengutronix.de>
Cc: Martin Fuzzey <mfuzzey@gmail.com>
Several of the macros in mx31ads.h depend on mx31.h which is no longer
included in quite so many standard headers as it once was. Include it
directly so we can build.
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
The i.MX I2C driver has not yet been merged into mainline but it is
near to that and the device defintions don't depend directly on it
so we can add the devices now.
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
This patch mimicks what Martin wrote on the mailing list:
* move arch/arm/mach-imx/include/mach/imxfb.h into
arch/arm/mach-mxc/include/mach/imxfb.h
* changes Kconfig so that CONFIG_FB_IMX is selectable
* adds a platform device (copied from some pengutronix
patches)
Signed-off-by: Holger Schurig <hs4233@mail.mn-solutions.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Based on code from "Martin Fuzzey" <mfuzzey@gmail.com>
Signed-off-by: Holger Schurig <hs4233@mail.mn-solutions.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
* define new CONFIG_ARCH_MX21 (this one is currently mutually exclusive to
CONFIG_ARCH_MX27, but this might change)
* splits one header file. Memory definitions, interrupt sources,
DMA channels are split into common part, i.MX27 specific and i.MX21
specific.
* guard access to UART5/UART6, which don't exist on i.MX21
Signed-off-by: Holger Schurig <hs4233@mail.mn-solutions.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Here are some of the warnings that get fixed by this:
> 200 times: warning: cast adds address space to expression (<asn:2>)
twelve times: warning: symbol 'xxx' was not declared. Should it be static
two times: warning: symbol 'clock' shadows an earlier one
five times: warning: incorrect type in initializer (different address spaces)
Signed-off-by: Holger Schurig <hs4233@mail.mn-solutions.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
This patch only adds general clkdev support without actually switching
any MXC architecture to clkdev.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
The context makes it clear already that these are clocks, so there's
no need for such a suffix. This patch only changes the clocks actually
used in the tree. The remaining clocks are renamed in the subsequent
architecture specific patches.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
- rename mxc_clocks_init to architecture specific versions. This
allows us to have more than one architecture compiled in.
- call mxc_timer_init from clock initialisation instead from board
code
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
We had 3 versions of this function in clock support for MX1/2/3
Use a single one instead. I picked the one from the MX3 as it seems
to calculate more accurate as the other ones. Also, on MX27 and MX31 mfn
can be negative, this hasn't been handled correctly on MX27 since now.
This patch has been tested on MX27 and MX31 and produces the same clock
frequencies for me.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
* adds Kconfig variables
* specifies different physical address for i.MX21 because of the
different memory layouts
* disables support for UART5/UART6 in the i.MX serial driver
(the i.MX21 doesn't have those modules)
Based on code from "Martin Fuzzey" <mfuzzey@gmail.com>
Signed-off-by: Holger Schurig <hs4233@mail.mn-solutions.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
* removed iomux-mx1-mx2.h completely
* distributes the former contents to four different files (iomux-mx1.h,
iomux-mx21.h, iomux-mx27.h and the file iomux-mx2x.h, which is common to
both i.MX21 and i.MX27).
* adds all documented IOMUX definitions for i.MX21 and i.MX27
* fixes a few that were wrong (PD14_AOUT_FEC_CLR, PE16_AF_RTCK).
* don't silenly include <linux/io.h>
* and fixes all collateral damage from above
Signed-off-by: Holger Schurig <hs4233@mail.mn-solutions.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Add gpio vbus detection to udc driver, by taking advantage
of the new gpio_vbus driver.
Signed-off-by: Robert Jarzmik <robert.jarzmik@free.fr>
Signed-off-by: Eric Miao <eric.miao@marvell.com>
CSB701 is one of baseboards that can be used with CSB726 SOM.
This currently adds support for button and LED on the board.
More to come later.
Signed-off-by: Dmitry Eremin-Solenikov <dbaryshkov@gmail.com>
Signed-off-by: Eric Miao <eric.miao@marvell.com>
The typo was originally fixed by Mike Rapoport and missed. And is
later reported by Matthias Meier.
Signed-off-by: Matthias Meier <matthias.j.meier@gmx.net>
Signed-off-by: Mike Rapoport <mike@compulab.co.il>
Signed-off-by: Eric Miao <eric.miao@marvell.com>
The current use of these macros works well when the conversion is
entirely linear. In this case, we can be assured that the following
holds true:
__va(p + s) - s = __va(p)
However, this is not always the case, especially when there is a
non-linear conversion (eg, when there is a 3.5GB hole in memory.)
In this case, if 's' is the size of the region (eg, PAGE_SIZE) and
'p' is the final page, the above is most definitely not true.
So, we must ensure that __va() and __pa() are only used with valid
kernel direct mapped RAM addresses. This patch tweaks the code
to achieve this.
Tested-by: Charles Moschel <fred99@carolina.rr.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
This is a fix for the following crash observed in 2.6.29-rc3:
http://lkml.org/lkml/2009/1/29/150
On ARM it doesn't make sense to trace a naked function because then
mcount is called without stack and frame pointer being set up and there
is no chance to restore the lr register to the value before mcount was
called.
Reported-by: Matthias Kaehlcke <matthias@kaehlcke.net>
Tested-by: Matthias Kaehlcke <matthias@kaehlcke.net>
Cc: Abhishek Sagar <sagar.abhishek@gmail.com>
Cc: Steven Rostedt <rostedt@home.goodmis.org>
Cc: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
This patch adds a Non-cacheable Normal ARM executable memory type,
MT_MEMORY_NONCACHED.
On OMAP3, this is used for rapid dynamic voltage/frequency scaling in
the VDD2 voltage domain. OMAP3's SDRAM controller (SDRC) is in the
VDD2 voltage domain, and its clock frequency must change along with
voltage. The SDRC clock change code cannot run from SDRAM itself,
since SDRAM accesses are paused during the clock change. So the
current implementation of the DVFS code executes from OMAP on-chip
SRAM, aka "OCM RAM."
If the OCM RAM pages are marked as Cacheable, the ARM cache controller
will attempt to flush dirty cache lines to the SDRC, so it can fill
those lines with OCM RAM instruction code. The problem is that the
SDRC is paused during DVFS, and so any SDRAM access causes the ARM MPU
subsystem to hang.
TI's original solution to this problem was to mark the OCM RAM
sections as Strongly Ordered memory, thus preventing caching. This is
overkill: since the memory is marked as non-bufferable, OCM RAM writes
become needlessly slow. The idea of "Strongly Ordered SRAM" is also
conceptually disturbing. Previous LAKML list discussion is here:
http://www.spinics.net/lists/arm-kernel/msg54312.html
This memory type MT_MEMORY_NONCACHED is used for OCM RAM by a future
patch.
Cc: Richard Woodruff <r-woodruff2@ti.com>
Signed-off-by: Paul Walmsley <paul@pwsan.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
These changes were included in the S3C audio header move but are not
directly related to it.
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Add the register defines for the sleep and power control
functions in the S3C64XX SYSCON register block.
Signed-off-by: Ben Dooks <ben-linux@fluff.org>
Add definitions for the EINT group registers and move the EINT IRQ
register definitions out of arch/arm/plat-s3c64xx/irq-eint.c so that
they are available for re-use with PM and the other code.
Signed-off-by: Ben Dooks <ben-linux@fluff.org>
Add the modem registers and a virtual mapping for the
modem block. This is is required as there are registers
that control the LCD block that need to be saved over
suspend as well as interrupt controls.
Signed-off-by: Ben Dooks <ben-linux@fluff.org>
As noted by Russell King, the sleep code path is not
elegant and makes use of leaving items on the stack
between calls.
Change the code that does the following:
if (s3c_cpu_save(regs_save) == 0) {
flush_cache_all();
S3C_PMDBG("preparing to sleep\n");
pm_cpu_sleep();
}
to simply call s3c_cpu_save, and let that do the
necessary calls to quiesce and sleep the system.
Signed-off-by: Ben Dooks <ben-linux@fluff.org>
Impact: __per_cpu_load available on all SMP capable archs
Percpu now requires three symbols to be defined - __per_cpu_load,
__per_cpu_start and __per_cpu_end. There were three archs which
didn't have it. Update them as follows.
* powerpc: can use generic PERCPU() macro. Compile tested for
powerpc32, compile/boot tested for powerpc64.
* ia64: can use generic PERCPU_VADDR() macro. __phys_per_cpu_start is
identical to __per_cpu_load. Compile tested and symbol table looks
identical after the change except for the additional __per_cpu_load.
* arm: added explicit __per_cpu_load definition. Currently uses
unified .init output section so can't use the generic macro. Dunno
whether the unified .init ouput section is required by arch
peculiarity so I left it alone. Please break it up and use PERCPU()
if possible.
Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Pat Gefre <pfg@sgi.com>
Cc: Russell King <rmk@arm.linux.org.uk>
The definitions of S3C2412_IISMOD_SDF_MSB and S3C2412_IISMOD_SDF_LSB
are incorrect, being the same S3C2412_IISMOD_SDF_IIS which is the
only correct one in this series.
Signed-off-by: Ben Dooks <ben-linux@fluff.org>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6: (29 commits)
p54: fix race condition in memory management
cfg80211: test before subtraction on unsigned
iwlwifi: fix error flow in iwl*_pci_probe
rt2x00 : more devices to rt73usb.c
rt2x00 : more devices to rt2500usb.c
bonding: Fix device passed into ->ndo_neigh_setup().
vlan: Fix vlan-in-vlan crashes.
net: Fix missing dev->neigh_setup in register_netdevice().
tmspci: fix request_irq race
pkt_sched: act_police: Fix a rate estimator test.
tg3: Fix 5906 link problems
SCTP: change sctp_ctl_sock_init() to try IPv4 if IPv6 fails
IPv6: add "disable" module parameter support to ipv6.ko
sungem: another error printed one too early
aoe: error printed 1 too early
net pcmcia: worklimit reaches -1
net: more timeouts that reach -1
net: fix tokenring license
dm9601: new vendor/product IDs
netlink: invert error code in netlink_set_err()
...
The remaining registers are separated into:
- <mach/regs-ost.h>
- <mach/regs-rtc.h>
- <mach/regs-intc.h>
and then we can remove pxa-regs.h completely. Instead of #include this
file, let's:
1. include the specific <mach/regs-*.h> with care (if that's absolutely
necessary)
2. define the registers in the driver, make cleanly defined API to expose
the register access to external with sufficient reason
Signed-off-by: Eric Miao <eric.miao@marvell.com>
The definitions of PXA_CS<x>_PHYS are really PXA2xx specific and should
be moved out of pxa-regs.h. As an illustration, the PXA3xx static chip
selects definitions are added into pxa3xx-regs.h.
Signed-off-by: Eric Miao <eric.miao@marvell.com>
This isn't perfect but at least solves the problem of pm.c's dependency
on register definitions in <mach/lubbock.h>, which doesn't make much
sense.
Signed-off-by: Eric Miao <eric.miao@marvell.com>
Move the processor specific initialization (largely resources initialization)
out of soc_common_drv_pcmcia_probe() into dedicated sa11xx_drv_pcmcia_probe()
and __pxa2xx_drv_pcmcia_probe().
By doing this, we are now able to move the PCMCIA related definitions out of
pxa-regs.h and back into pxa2xx_base.c.
As a result, remove that reference of _PCMCIA1IO in arch/arm/mach-pxa/viper.c.
Signed-off-by: Eric Miao <eric.miao@marvell.com>
pxa-regs.h and hardware.h are not intended for use directly in driver
code, remove those unnecessary references.
Signed-off-by: Eric Miao <eric.miao@marvell.com>
Let's handle GPIOs by banks, each bank covers up to 32 GPIOs with one set
of registers, and each set of registers start from different offsets.
GPLR GPDR GPSR GPCR GRER GFER GEDR
BANK 0 - 0x0000 0x000C 0x0018 0x0024 0x0030 0x003C 0x0048
BANK 1 - 0x0004 0x0010 0x001C 0x0028 0x0034 0x0040 0x004C
BANK 2 - 0x0008 0x0014 0x0020 0x002C 0x0038 0x0044 0x0050
BANK 3 - 0x0100 0x010C 0x0118 0x0124 0x0130 0x013C 0x0148
BANK 4 - 0x0104 0x0110 0x011C 0x0128 0x0134 0x0140 0x014C
BANK 5 - 0x0108 0x0114 0x0120 0x012C 0x0138 0x0144 0x0150
NOTE:
BANK 3 is only available on PXA27x and later processors.
BANK 4 and 5 are only available on PXA935
1. introduce GPIO_BANK(n) for the offset base of each bank
2. 'struct pxa_gpio_chip' is expanded to include IRQ edge and mask
setings, and saved register values as well, and is dynamically
allocated due to possible bank number ranging from 3 to 6
3. all accesses to GPIO registers are made through 'regbase' within
'pxa_gpio_chip', and register offset
4. introduce several inline functions to simplify the code a bit
5. change IRQ demux handler to base on gpio chips
Signed-off-by: Mike Rapoport <mike@compulab.co.il>
Signed-off-by: Eric Miao <eric.miao@marvell.com>
This makes gpio.c fully independent of pxa-regs.h (except for the
virtual address of the registers).
Signed-off-by: Mike Rapoport <mike@compulab.co.il>
Signed-off-by: Eric Miao <eric.miao@marvell.com>
Looks like we have to live with pxa_gpio_mode() for a while, giving
its presence is actually making gpio.c not generic enough, let's
move it temporarily outside before it can be fully purged.
Signed-off-by: Eric Miao <eric.miao@marvell.com>
This is part of the work making gpio.c generic enough, the changes
include:
1. move IRQ handling of GPIO 0 and 1 outside (and back into irq.c)
2. pxa_init_gpio() accepts a range for muxed GPIO IRQs, and an IRQ
number for the muxed GPIOs
3. __gpio_is_occupied() and __gpio_is_inverted() are made inline,
and are moved into <mach/gpio.h> instead of generic gpio.c
Signed-off-by: Eric Miao <eric.miao@marvell.com>
1. Driver code where pxa_request_dma() is called will most likely
reference DMA registers as well, and it is really unnecessary
to include pxa-regs.h just for this. Move the definitions into
<mach/dma.h> and make relevant drivers include it instead of
<mach/pxa-regs.h>.
2. Introduce DMAC_REGS_VIRT as the virtual address base for these
DMA registers. This allows later processors to re-use the same
IP while registers may start at different I/O address.
Signed-off-by: Eric Miao <eric.miao@marvell.com>
Considering the header mess ATM, it is not always possible to include
the correct header files within board code. Let's keep this simple:
<mach/pxa25x.h> - for pxa25x based platforms
<mach/pxa27x.h> - for pxa27x based platforms
<mach/pxa300.h> - for pxa300 based platforms
<mach/pxa320.h> - for pxa320 based platforms
<mach/pxa930.h> - for pxa930 based platforms
NOTE:
1. one header one board file, they are not compatible (i.e. they have
conflicting definitions which won't compile if included together).
2. Unless strictly necessary, the following header files are considered
to be SoC files use _only_, and is not recommended to be included in
board code:
<mach/hardware.h>
<mach/pxa-regs.h>
<mach/pxa2xx-regs.h>
<mach/pxa3xx-regs.h>
<mach/mfp.h>
<mach/mfp-pxa2xx.h>
<mach/mfp-pxa25x.h>
<mach/mfp-pxa27x.h>
<mach/mfp-pxa3xx.h>
<mach/mfp-pxa300.h>
<mach/mfp-pxa320.h>
<mach/mfp-pxa930.h>
Signed-off-by: Eric Miao <eric.miao@marvell.com>
Default to the same behaviour as the shipped WinCE system.
Signed-off-by: Philipp Zabel <philipp.zabel@gmail.com>
Signed-off-by: Eric Miao <eric.miao@marvell.com>
The PM CRC checking code kmallocs an area to save a set of
CRC values during suspend. This triggers a warning due to the
call of a function that might sleep whilst the system is not
in a valid state to do so.
Move the allocation and free to points in the suspend and resume
process where they can call a function that might-sleep.
Signed-off-by: Ben Dooks <ben-linux@fluff.org>
When doing the CRC check of the memory, avoid checking
the page that our stack is residing in as this changes
during the execution of the suspend and resume.
Signed-off-by: Ben Dooks <ben-linux@fluff.org>
Add the definition for S3C_GPIO_END to allow the PM code to build.
This means moving the GPIO bank numbers to a separate file to allow
the gpio and regs-gpio to include them. Including regs-gpio.h into
gpio.h causes too many build problems and adding gpio.h would mean
editing a large number of files.
Signed-off-by: Ben Dooks <ben-linux@fluff.org>
Change the way the UART state is saved over suspend to allow the s3c64xx
code to modify the settings on resume to avoid any illegal state changes
to the UART clocks. This will also allow us to save the UDIVSLOT register
on newer SoCs.
Move to using a structure for the UART use the extant Kconfig configuration
specifying the number of UARTs.
Signed-off-by: Ben Dooks <ben-linux@fluff.org>
Rename s3c2410_cpu_resume to s3c_cpu_resume and s3c2410_cpu_save to
s3c_cpu_save to remove the CPU specific naming of these functions
which are now in the generic PM code.
Signed-off-by: Ben Dooks <ben-linux@fluff.org>
Since we have moved a large proportion of the PM code to the common
support area, remove the cpu specific name from the initialisation
function.
Signed-off-by: Ben Dooks <ben-linux@fluff.org>
There's a bug in calculation of IRQ_EINT_BIT introduced on the test
branch for pm changes for s3c by Ben Dooks fixed in this patch.
There's also a bit of a mystery about how wake gets to wake EINT
set of interrupts, I added a couple of lines that make it work for
EINT4+ but not sure what's meant to be there for EINT0-3.
Still, this gets GTA02 resume working again.
Signed-off-by: Andy Green <andy@openmoko.com>
[ben-linux@fluff.org: remove irq-pm.c change]
Signed-off-by: Ben Dooks <ben-linux@fluff.org>
Move the IRQ_EINT sleep control to be available to all
s3c impelmentations. Since s3c_irqext_wake is not large,
place it in arch/arm/plat-s3c/pm.c as adding it to a new
file would be a waste of compile time.
Signed-off-by: Ben Dooks <ben-linux@fluff.org>
Split the PM code out of arch/arm/plat-s3c24xx/irq.c to
remove some of the #ifdefs being used. Also fix a couple
of places where the absecnce of a function was redefined
to the same thing.
Signed-off-by: Ben Dooks <ben-linux@fluff.org>
Fix warnings from struct resource being bigger than unsigned long by
forcing the type. We are only a 32bit platform so no physical memory
addresses will be too big to fit in this.
Signed-off-by: Ben Dooks <ben-linux@fluff.org>
Split the optional memory check code out of the pm.c file
as it is quite a big #ifdef block and as-such can be moved
out and simply compiled when the configuration is set.
Signed-off-by: Ben Dooks <ben-linux@fluff.org>
Start moving the PM code by moving all the common support functions
to a common location in arch/arm/plat-s3c. With the move we rename
the functions from s3cxxx_ to s3c_ to fit the new location.
Signed-off-by: Ben Dooks <ben-linux@fluff.org>
Move the <plat/pm.h> header to plat-s3c as preparation
for moving parts of the s3c24xx pm support which are
common into the plat-s3c support.
Signed-off-by: Ben Dooks <ben-linux@fluff.org>
Includes missed irqs.h in devices.c and mx1ads.c.
Signed-off-by: Darius Augulis <augulis.darius@gmail.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Add two more bitfields for the PSP register. As they seem to exist
for PXA3xx only, define them conditionally.
Signed-off-by: Daniel Mack <daniel@caiaq.de>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
A bit in PXA's SSCR0 register was erroneously named ADC but its name is
in fact ACS (audio clock select).
Signed-off-by: Daniel Mack <daniel@caiaq.de>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Add the initial code to support the S3C64XX I2S hardware using the
s3c-i2s-v2 core code.
Signed-off-by: Ben Dooks <ben@simtec.co.uk>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Fix build warnings due to struct i2c_board_info in <mach/platform.h>
Patch "5311/1: add core support for built in i2c bus" is causing 11 of
39 the build warnings with Kautobuild for ep93xx_defconfig on kernel
2.6.29-rc5-git4. This patch fixes it.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Signed-off-by: Stanislaw Gruszka <stf_xl@wp.pl>
Cc: Andrew Victor <avictor.za@gmail.com>
Acked-by: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Acked-by: Andrew Victor <linux@maxim.org.za>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
This is IDE host driver for AT91 (SAM9, CAP9, AT572D940HF) Static Memory
Controller with Compact Flash True IDE Mode logic.
Driver have to switch 8/16 bit bus width when accessing Task Tile or Data
Register. Moreover some extra things need to be done when setting PIO mode.
Only PIO mode is used, hardware have no DMA support. If interrupt line is
connected through GPIO extra quirk is needed to cope with fake interrupts.
Signed-off-by: Stanislaw Gruszka <stf_xl@wp.pl>
Cc: Andrew Victor <avictor.za@gmail.com>
Acked-by: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
gcc seems to expect that lr isn't clobbered by mcount, because for a
function starting with:
static int func(void)
{
void *ra = __builtin_return_address(0);
printk(KERN_EMERG "__builtin_return_address(0) = %pS\n", ra)
...
the following assembler is generated by gcc 4.3.2:
0: e1a0c00d mov ip, sp
4: e92dd810 push {r4, fp, ip, lr, pc}
8: e24cb004 sub fp, ip, #4 ; 0x4
c: ebfffffe bl 0 <mcount>
10: e59f0034 ldr r0, [pc, #52]
14: e1a0100e mov r1, lr
18: ebfffffe bl 0 <printk>
Without this patch obviously __builtin_return_address(0) yields
func+0x10 instead of the return address of the caller.
Note this patch fixes a similar issue for the routines used with dynamic
ftrace even though this isn't currently selectable for ARM.
Cc: Abhishek Sagar <sagar.abhishek@gmail.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Martin Michlmayr
Nicolas Pitre
Russell King
Eric Miao
Daniel Mack
Lennert Buytenhek
Daniel Silverstone
Mike Rapoport
David Brownell
Dmitry Eremin-Solenikov
Krzysztof Hałasa
Ingo Molnar
Shadi Ammouri
Linus Torvalds
Dmitry Artamonow
Ilya Yanok
Sascha Hauer
Valentin Longchamp
Guennadi Liakhovetski
Holger Schurig
Mark Brown
Uwe Kleine-König
Robert Jarzmik
Paul Walmsley
Ben Dooks
Tejun Heo
Philipp Zabel
Andy Green
Darius Augulis
Ben Dooks
Hartley Sweeten
Stanislaw Gruszka