ioat3.2 does not support asynchronous error notifications which makes
the driver experience latencies when non-zero pq validate results are
expected. Provide a mechanism for turning off async_xor_val and
async_syndrome_val via Kconfig. This approach is generally useful for
any driver that specifies ASYNC_TX_DISABLE_CHANNEL_SWITCH and would like
to force the async_tx api to fall back to the synchronous path for
certain operations.
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
This restriction prevented ASYNC_TX_DMA from being enabled on platform
configurations where DMA address conversion could not be performed in
place on the stack. Since commit 04ce9ab3 ("async_xor: permit callers
to pass in a 'dma/page scribble' region") the async_tx api now either
uses a caller provided 'scribble' buffer, or performs the conversion in
place when sizeof(dma_addr_t) <= sizeof(struct page *).
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
This supported all DMA channels, and it was tested in SH7722,
SH7780, SH7785 and SH7763.
This can not use with SH DMA API.
Signed-off-by: Nobuhiro Iwamatsu <iwamatsu.nobuhiro@renesas.com>
Reviewed-by: Matt Fleming <matt@console-pimps.org>
Acked-by: Maciej Sosnowski <maciej.sosnowski@intel.com>
Acked-by: Paul Mundt <lethal@linux-sh.org>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Channel switching is problematic for some dmaengine drivers as the
architecture precludes separating the ->prep from ->submit. In these
cases the driver can select ASYNC_TX_DISABLE_CHANNEL_SWITCH to modify
the async_tx allocator to only return channels that support all of the
required asynchronous operations.
For example MD_RAID456=y selects support for asynchronous xor, xor
validate, pq, pq validate, and memcpy. When
ASYNC_TX_DISABLE_CHANNEL_SWITCH=y any channel with all these
capabilities is marked DMA_ASYNC_TX allowing async_tx_find_channel() to
quickly locate compatible channels with the guarantee that dependency
chains will remain on one channel. When
ASYNC_TX_DISABLE_CHANNEL_SWITCH=n async_tx_find_channel() may select
channels that lead to operation chains that need to cross channel
boundaries using the async_tx channel switch capability.
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
This AHB DMA Controller (aka HDMA or DMAC on AT91 systems) is availlable on
at91sam9rl chip. It will be used on other products in the future.
This first release covers only the memory-to-memory tranfer type. This is the
only tranfer type supported by this chip. On other products, it will be used
also for peripheral DMA transfer (slave API support to come).
I used dmatest client without problem in different configurations to test it.
Full documentation for this controller can be found in the SAM9RL datasheet:
http://www.atmel.com/dyn/products/product_card.asp?part_id=4243
Signed-off-by: Nicolas Ferre <nicolas.ferre@atmel.com>
Acked-by: Maciej Sosnowski <maciej.sosnowski@intel.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
On HIGHMEM64G systems dma_addr_t is known to be larger than (void *)
which precludes async_xor from performing dma address conversions by
reusing the input parameter address list. However, other parts of the
dmaengine infrastructure do not suffer this constraint, so the
HIGHMEM64G restriction can be down-levelled.
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
This patch adds support for the integrated DMAC of the TXx9 family.
Signed-off-by: Atsushi Nemoto <anemo@mba.ocn.ne.jp>
Acked-by: Dan Williams <dan.j.williams@intel.com>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
async_xor() needs space to perform dma and page address conversions. In
most cases the code can simply reuse the struct page * array because the
size of the native pointer matches the size of a dma/page address. In
order to support archs where sizeof(dma_addr_t) is larger than
sizeof(struct page *), or to preserve the input parameters, we utilize a
memory region passed in by the caller.
Since the code is now prepared to handle the case where it cannot
perform address conversions on the stack, we no longer need the
!HIGHMEM64G dependency in drivers/dma/Kconfig.
[ Impact: don't clobber input buffers for address conversions ]
Reviewed-by: Andre Noll <maan@systemlinux.org>
Acked-by: Maciej Sosnowski <maciej.sosnowski@intel.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
i.MX3x SoCs contain an Image Processing Unit, consisting of a Control
Module (CM), Display Interface (DI), Synchronous Display Controller (SDC),
Asynchronous Display Controller (ADC), Image Converter (IC), Post-Filter
(PF), Camera Sensor Interface (CSI), and an Image DMA Controller (IDMAC).
CM contains, among other blocks, an Interrupt Generator (IG) and a Clock
and Reset Control Unit (CRCU). This driver serves IDMAC and IG. They are
supported over dmaengine and irq-chip APIs respectively.
IDMAC is a specialised DMA controller, its DMA channels cannot be used for
general-purpose operations, even though it might be possible to configure
a memory-to-memory channel for memcpy operation. This driver will not work
with generic dmaengine clients, clients, wishing to use it must use
respective wrapper structures, they also must specify which channels they
require, as channels are hard-wired to specific IPU functions.
Acked-by: Sascha Hauer <s.hauer@pengutronix.de>
Signed-off-by: Guennadi Liakhovetski <lg@denx.de>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
async_tx.ko is a consumer of dma channels. A circular dependency arises
if modules in drivers/dma rely on common code in async_tx.ko. It
prevents either module from being unloaded.
Move dma_wait_for_async_tx and async_tx_run_dependencies to dmaeninge.o
where they should have been from the beginning.
Reviewed-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Modify the Freescale Elo / Elo Plus DMA driver so that it can be compiled as
a module.
The primary change is to stop treating the DMA controller as a bus, and the
DMA channels as devices on the bus. This is because the Open Firmware (OF)
kernel code does not allow busses to be removed, so although we can call
of_platform_bus_probe() to probe the DMA channels, there is no
of_platform_bus_remove(). Instead, the DMA channels are manually probed,
similar to what fsl_elbc_nand.c does.
Cc: Scott Wood <scottwood@freescale.com>
Acked-by: Li Yang <leoli@freescale.com>
Signed-off-by: Timur Tabi <timur@freescale.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
This adds a driver for the Synopsys DesignWare DMA controller (aka
DMACA on AVR32 systems.) This DMA controller can be found integrated
on the AT32AP7000 chip and is primarily meant for peripheral DMA
transfer, but can also be used for memory-to-memory transfers.
This patch is based on a driver from David Brownell which was based on
an older version of the DMA Engine framework. It also implements the
proposed extensions to the DMA Engine API for slave DMA operations.
The dmatest client shows no problems, but there may still be room for
improvement performance-wise. DMA slave transfer performance is
definitely "good enough"; reading 100 MiB from an SD card running at ~20
MHz yields ~7.2 MiB/s average transfer rate.
Full documentation for this controller can be found in the Synopsys
DW AHB DMAC Databook:
http://www.synopsys.com/designware/docs/iip/DW_ahb_dmac/latest/doc/dw_ahb_dmac_db.pdf
The controller has lots of implementation options, so it's usually a
good idea to check the data sheet of the chip it's intergrated on as
well. The AT32AP7000 data sheet can be found here:
http://www.atmel.com/dyn/products/datasheets.asp?family_id=682
Changes since v4:
* Use client_count instead of dma_chan_is_in_use()
* Add missing include
* Unmap buffers unless client told us not to
Changes since v3:
* Update to latest DMA engine and DMA slave APIs
* Embed the hw descriptor into the sw descriptor
* Clean up and update MODULE_DESCRIPTION, copyright date, etc.
Changes since v2:
* Dequeue all pending transfers in terminate_all()
* Rename dw_dmac.h -> dw_dmac_regs.h
* Define and use controller-specific dma_slave data
* Fix up a few outdated comments
* Define hardware registers as structs (doesn't generate better
code, unfortunately, but it looks nicer.)
* Get number of channels from platform_data instead of hardcoding it
based on CONFIG_WHATEVER_CPU.
* Give slave clients exclusive access to the channel
Acked-by: Maciej Sosnowski <maciej.sosnowski@intel.com>,
Signed-off-by: Haavard Skinnemoen <haavard.skinnemoen@atmel.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
This client tests DMA memcpy using various lengths and various offsets
into the source and destination buffers. It will initialize both
buffers with a repeatable pattern and verify that the DMA engine copies
the requested region and nothing more. It will also verify that the
bytes aren't swapped around, and that the source buffer isn't modified.
The dmatest module can be configured to test a specific device, a
specific channel. It can also test multiple channels at the same time,
and it can start multiple threads competing for the same channel.
Changes since v2:
* Support testing multiple channels at the same time
* Support testing with multiple threads competing for the same channel
* Use counting test patterns in order to catch byte ordering issues
Changes since v1:
* Remove extra dashes around "help"
* Remove "default n" from Kconfig
* Turn TEST_BUF_SIZE into a module parameter
* Return DMA_NAK instead of DMA_DUP
* Print unhandled events
* Support testing specific channels and devices
* Move to the end of the Makefile
Acked-by: Maciej Sosnowski <maciej.sosnowski@intel.com>
Signed-off-by: Haavard Skinnemoen <hskinnemoen@atmel.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
The XOR engine found in Marvell's SoCs and system controllers
provides XOR and DMA operation, iSCSI CRC32C calculation, memory
initialization, and memory ECC error cleanup operation support.
This driver implements the DMA engine API and supports the following
capabilities:
- memcpy
- xor
- memset
The XOR engine can be used by DMA engine clients implemented in the
kernel, one of those clients is the RAID module. In that case, I
observed 20% improvement in the raid5 write throughput, and 40%
decrease in the CPU utilization when doing array construction, those
results obtained on an 5182 running at 500Mhz.
When enabling the NET DMA client, the performance decreased, so
meanwhile it is recommended to keep this client off.
Signed-off-by: Saeed Bishara <saeed@marvell.com>
Signed-off-by: Lennert Buytenhek <buytenh@marvell.com>
Signed-off-by: Nicolas Pitre <nico@marvell.com>
Acked-by: Maciej Sosnowski <maciej.sosnowski@intel.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
The dependency is redundant since all drivers set their specific arch
dependencies. The NET_DMA option is modified to be enabled only on platforms
where it is known to have a positive effect. HAS_DMA is added as an explicit
dependency for the DMADEVICES menu.
Acked-by: Adrian Bunk <bunk@kernel.org>
Acked-by: Haavard Skinnemoen <haavard.skinnemoen@atmel.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Always enabling the fsl_dma_self_test() to ensure the DMA controller
should works well after the driver probed.
Signed-off-by: Zhang Wei <wei.zhang@freescale.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
The driver implements DMA engine API for Freescale MPC85xx DMA controller,
which could be used by devices in the silicon. The driver supports the
Basic mode of Freescale MPC85xx DMA controller. The MPC85xx processors
supported include MPC8540/60, MPC8555, MPC8548, MPC8641 and so on.
The MPC83xx(MPC8349, MPC8360) are also supported.
[kamalesh@linux.vnet.ibm.com: build fix]
[dan.j.williams@intel.com: merge mm fixes, rebase on async_tx-2.6.25]
Signed-off-by: Zhang Wei <wei.zhang@freescale.com>
Signed-off-by: Ebony Zhu <ebony.zhu@freescale.com>
Acked-by: Kumar Gala <galak@gate.crashing.org>
Cc: Shannon Nelson <shannon.nelson@intel.com>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
The tx_set_src and tx_set_dest methods were originally implemented to allow
an array of addresses to be passed down from async_xor to the dmaengine
driver while minimizing stack overhead. Removing these methods allows
drivers to have all transaction parameters available at 'prep' time, saves
two function pointers in struct dma_async_tx_descriptor, and reduces the
number of indirect branches..
A consequence of moving this data to the 'prep' routine is that
multi-source routines like async_xor need temporary storage to convert an
array of linear addresses into an array of dma addresses. In order to keep
the same stack footprint of the previous implementation the input array is
reused as storage for the dma addresses. This requires that
sizeof(dma_addr_t) be less than or equal to sizeof(void *). As a
consequence CONFIG_DMADEVICES now depends on !CONFIG_HIGHMEM64G. It also
requires that drivers be able to make descriptor resources available when
the 'prep' routine is polled.
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Acked-by: Shannon Nelson <shannon.nelson@intel.com>
This patch corrects recently changed (and now invalid) Kconfig descriptions
for the DMA engine framework:
- Non-Intel(R) hardware also has DMA engines;
- DMA is used for more than memcpy and RAID offloading.
In fact, on most platforms memcpy and RAID aren't factors, and DMA
exists so that peripherals can transfer data to/from memory while
the CPU does other work.
Signed-off-by: Haavard Skinnemoen <hskinnemoen@atmel.com>
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
No reason I can think of of making them default y Most people don't have
the hardware and with default y they just pollute lots of configs during
make oldconfig.
Signed-off-by: Andi Kleen <ak@suse.de>
Acked-by: Jeff Garzik <jeff@garzik.org>
Acked-by: "Nelson, Shannon" <shannon.nelson@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Add code to connect to the DCA driver and provide cpu tags for use by
drivers that would like to use Direct Cache Access hints.
[Adrian Bunk] Several Kconfig cleanup items
[Andrew Morten, Chris Leech] Fix for using cpu_physical_id() even when
built for uni-processor
Signed-off-by: Shannon Nelson <shannon.nelson@intel.com>
Acked-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Continuing the work started in 411f0f3edc ...
This enables code with a dma path, that compiles away, to build without
requiring additional code factoring. It also prevents code that calls
dma_alloc_coherent and dma_free_coherent from linking whereas previously
the code would hit a BUG() at run time. Finally, it allows archs that set
!HAS_DMA to delete their asm/dma-mapping.h file.
Cc: Cornelia Huck <cornelia.huck@de.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: John W. Linville <linville@tuxdriver.com>
Cc: Kyle McMartin <kyle@parisc-linux.org>
Cc: James Bottomley <James.Bottomley@SteelEye.com>
Cc: Tejun Heo <htejun@gmail.com>
Cc: Jeff Garzik <jeff@garzik.org>
Cc: <geert@linux-m68k.org>
Cc: <zippel@linux-m68k.org>
Cc: <spyro@f2s.com>
Cc: <ysato@users.sourceforge.jp>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The Intel(R) IOP series of i/o processors integrate an Xscale core with
raid acceleration engines. The capabilities per platform are:
iop219:
(2) copy engines
iop321:
(2) copy engines
(1) xor and block fill engine
iop33x:
(2) copy and crc32c engines
(1) xor, xor zero sum, pq, pq zero sum, and block fill engine
iop34x (iop13xx):
(2) copy, crc32c, xor, xor zero sum, and block fill engines
(1) copy, crc32c, xor, xor zero sum, pq, pq zero sum, and block fill engine
The driver supports the features of the async_tx api:
* asynchronous notification of operation completion
* implicit (interupt triggered) handling of inter-channel transaction
dependencies
The driver adapts to the platform it is running by two methods.
1/ #include <asm/arch/adma.h> which defines the hardware specific
iop_chan_* and iop_desc_* routines as a series of static inline
functions
2/ The private platform data attached to the platform_device defines the
capabilities of the channels
20070626: Callbacks are run in a tasklet. Given the recent discussion on
LKML about killing tasklets in favor of workqueues I did a quick conversion
of the driver. Raid5 resync performance dropped from 50MB/s to 30MB/s, so
the tasklet implementation remains until a generic softirq interface is
available.
Changelog:
* fixed a slot allocation bug in do_iop13xx_adma_xor that caused too few
slots to be requested eventually leading to data corruption
* enabled the slot allocation routine to attempt to free slots before
returning -ENOMEM
* switched the cleanup routine to solely use the software chain and the
status register to determine if a descriptor is complete. This is
necessary to support other IOP engines that do not have status writeback
capability
* make the driver iop generic
* modified the allocation routines to understand allocating a group of
slots for a single operation
* added a null xor initialization operation for the xor only channel on
iop3xx
* support xor operations on buffers larger than the hardware maximum
* split the do_* routines into separate prep, src/dest set, submit stages
* added async_tx support (dependent operations initiation at cleanup time)
* simplified group handling
* added interrupt support (callbacks via tasklets)
* brought the pending depth inline with ioat (i.e. 4 descriptors)
* drop dma mapping methods, suggested by Chris Leech
* don't use inline in C files, Adrian Bunk
* remove static tasklet declarations
* make iop_adma_alloc_slots easier to read and remove chances for a
corrupted descriptor chain
* fix locking bug in iop_adma_alloc_chan_resources, Benjamin Herrenschmidt
* convert capabilities over to dma_cap_mask_t
* fixup sparse warnings
* add descriptor flush before iop_chan_enable
* checkpatch.pl fixes
* gpl v2 only correction
* move set_src, set_dest, submit to async_tx methods
* move group_list and phys to async_tx
Cc: Russell King <rmk@arm.linux.org.uk>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
The async_tx api provides methods for describing a chain of asynchronous
bulk memory transfers/transforms with support for inter-transactional
dependencies. It is implemented as a dmaengine client that smooths over
the details of different hardware offload engine implementations. Code
that is written to the api can optimize for asynchronous operation and the
api will fit the chain of operations to the available offload resources.
I imagine that any piece of ADMA hardware would register with the
'async_*' subsystem, and a call to async_X would be routed as
appropriate, or be run in-line. - Neil Brown
async_tx exploits the capabilities of struct dma_async_tx_descriptor to
provide an api of the following general format:
struct dma_async_tx_descriptor *
async_<operation>(..., struct dma_async_tx_descriptor *depend_tx,
dma_async_tx_callback cb_fn, void *cb_param)
{
struct dma_chan *chan = async_tx_find_channel(depend_tx, <operation>);
struct dma_device *device = chan ? chan->device : NULL;
int int_en = cb_fn ? 1 : 0;
struct dma_async_tx_descriptor *tx = device ?
device->device_prep_dma_<operation>(chan, len, int_en) : NULL;
if (tx) { /* run <operation> asynchronously */
...
tx->tx_set_dest(addr, tx, index);
...
tx->tx_set_src(addr, tx, index);
...
async_tx_submit(chan, tx, flags, depend_tx, cb_fn, cb_param);
} else { /* run <operation> synchronously */
...
<operation>
...
async_tx_sync_epilog(flags, depend_tx, cb_fn, cb_param);
}
return tx;
}
async_tx_find_channel() returns a capable channel from its pool. The
channel pool is organized as a per-cpu array of channel pointers. The
async_tx_rebalance() routine is tasked with managing these arrays. In the
uniprocessor case async_tx_rebalance() tries to spread responsibility
evenly over channels of similar capabilities. For example if there are two
copy+xor channels, one will handle copy operations and the other will
handle xor. In the SMP case async_tx_rebalance() attempts to spread the
operations evenly over the cpus, e.g. cpu0 gets copy channel0 and xor
channel0 while cpu1 gets copy channel 1 and xor channel 1. When a
dependency is specified async_tx_find_channel defaults to keeping the
operation on the same channel. A xor->copy->xor chain will stay on one
channel if it supports both operation types, otherwise the transaction will
transition between a copy and a xor resource.
Currently the raid5 implementation in the MD raid456 driver has been
converted to the async_tx api. A driver for the offload engines on the
Intel Xscale series of I/O processors, iop-adma, is provided in a later
commit. With the iop-adma driver and async_tx, raid456 is able to offload
copy, xor, and xor-zero-sum operations to hardware engines.
On iop342 tiobench showed higher throughput for sequential writes (20 - 30%
improvement) and sequential reads to a degraded array (40 - 55%
improvement). For the other cases performance was roughly equal, +/- a few
percentage points. On a x86-smp platform the performance of the async_tx
implementation (in synchronous mode) was also +/- a few percentage points
of the original implementation. According to 'top' on iop342 CPU
utilization drops from ~50% to ~15% during a 'resync' while the speed
according to /proc/mdstat doubles from ~25 MB/s to ~50 MB/s.
The tiobench command line used for testing was: tiobench --size 2048
--block 4096 --block 131072 --dir /mnt/raid --numruns 5
* iop342 had 1GB of memory available
Details:
* if CONFIG_DMA_ENGINE=n the asynchronous path is compiled away by making
async_tx_find_channel a static inline routine that always returns NULL
* when a callback is specified for a given transaction an interrupt will
fire at operation completion time and the callback will occur in a
tasklet. if the the channel does not support interrupts then a live
polling wait will be performed
* the api is written as a dmaengine client that requests all available
channels
* In support of dependencies the api implicitly schedules channel-switch
interrupts. The interrupt triggers the cleanup tasklet which causes
pending operations to be scheduled on the next channel
* Xor engines treat an xor destination address differently than a software
xor routine. To the software routine the destination address is an implied
source, whereas engines treat it as a write-only destination. This patch
modifies the xor_blocks routine to take a an explicit destination address
to mirror the hardware.
Changelog:
* fixed a leftover debug print
* don't allow callbacks in async_interrupt_cond
* fixed xor_block changes
* fixed usage of ASYNC_TX_XOR_DROP_DEST
* drop dma mapping methods, suggested by Chris Leech
* printk warning fixups from Andrew Morton
* don't use inline in C files, Adrian Bunk
* select the API when MD is enabled
* BUG_ON xor source counts <= 1
* implicitly handle hardware concerns like channel switching and
interrupts, Neil Brown
* remove the per operation type list, and distribute operation capabilities
evenly amongst the available channels
* simplify async_tx_find_channel to optimize the fast path
* introduce the channel_table_initialized flag to prevent early calls to
the api
* reorganize the code to mimic crypto
* include mm.h as not all archs include it in dma-mapping.h
* make the Kconfig options non-user visible, Adrian Bunk
* move async_tx under crypto since it is meant as 'core' functionality, and
the two may share algorithms in the future
* move large inline functions into c files
* checkpatch.pl fixes
* gpl v2 only correction
Cc: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Acked-By: NeilBrown <neilb@suse.de>
Disable some more menus in the configuration files that are of no
interest to a s390 machine.
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Attempts to allocate per-CPU DMA channels
Signed-off-by: Chris Leech <christopher.leech@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Provides an API for offloading memory copies to DMA devices
Signed-off-by: Chris Leech <christopher.leech@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Dan Williams
Nobuhiro Iwamatsu
Nicolas Ferre
Atsushi Nemoto
Guennadi Liakhovetski
Timur Tabi
Haavard Skinnemoen
Haavard Skinnemoen
Saeed Bishara
Zhang Wei
Andi Kleen
Shannon Nelson
Martin Schwidefsky
Chris Leech