It's not surprising that the transfer request buffer (TRB) physical to
virtual address translation function has bugs in it, since I wrote most of
it at 4am last October. Add a test suite to check the TRB math. This
runs at memory initialization time, and causes the driver to fail to load
if the TRB math fails.
Please excuse the excessively long lines in the test vectors; they can't
really be made shorter and still be readable.
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
In order to giveback a canceled URB, we must ensure that the xHCI
hardware will not access the buffer in an URB. We can't modify the
buffer pointers on endpoint rings without issuing and waiting for a stop
endpoint command. Since URBs can be canceled in interrupt context, we
can't wait on that command. The old code trusted that the host
controller would respond to the command, and would giveback the URBs in
the event handler. If the hardware never responds to the stop endpoint
command, the URBs will never be completed, and we might hang the USB
subsystem.
Implement a watchdog timer that is spawned whenever a stop endpoint
command is queued. If a stop endpoint command event is found on the
event ring during an interrupt, we need to stop the watchdog timer with
del_timer(). Since del_timer() can fail if the timer is running and
waiting on the xHCI lock, we need a way to signal to the timer that
everything is fine and it should exit. If we simply clear
EP_HALT_PENDING, a new stop endpoint command could sneak in and set it
before the watchdog timer can grab the lock.
Instead we use a combination of the EP_HALT_PENDING flag and a counter
for the number of pending stop endpoint commands
(xhci_virt_ep->stop_cmds_pending). If we need to cancel the watchdog
timer and del_timer() succeeds, we decrement the number of pending stop
endpoint commands. If del_timer() fails, we leave the number of pending
stop endpoint commands alone. In either case, we clear the
EP_HALT_PENDING flag.
The timer will decrement the number of pending stop endpoint commands
once it obtains the lock. If the timer is the tail end of the last stop
endpoint command (xhci_virt_ep->stop_cmds_pending == 0), and the
endpoint's command is still pending (EP_HALT_PENDING is set), we assume
the host is dying. The watchdog timer will set XHCI_STATE_DYING, try to
halt the xHCI host, and give back all pending URBs.
Various other places in the driver need to check whether the xHCI host
is dying. If the interrupt handler ever notices, it should immediately
stop processing events. The URB enqueue function should also return
-ESHUTDOWN. The URB dequeue function should simply return the value
of usb_hcd_check_unlink_urb() and the watchdog timer will take care of
giving the URB back. When a device is disconnected, the xHCI hardware
structures should be freed without issuing a disable slot command (since
the hardware probably won't respond to it anyway). The debugging
polling loop should stop polling if the host is dying.
When a device is disconnected, any pending watchdog timers are killed
with del_timer_sync(). It must be synchronous so that the watchdog
timer doesn't attempt to access the freed endpoint structures.
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
The scratchpad_free() function uses xhci->page_size to free some memory
with pci_free_consistent(). However, the page_size is set to zero before
the call, causing kernel oopses on driver unload. Call scratchpad_free()
before setting xhci->page_size to zero.
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Acked-by: John Youn <John.Youn@synopsys.com>
Cc: stable <stable@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
If the xHCI driver fails during the memory initialization, xhci->ir_set
may not be a valid pointer. Check that it points to valid DMA'able memory
before writing to that address during the memory freeing process.
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Cc: stable <stable@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
When setting up a slot context for an address device command, set the
multi-TT field if this is a low or full speed device under a HS hub with
multiple transaction translators.
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
The xHCI driver needs to set the route string in the slot context of all
devices, not just SuperSpeed devices. The route string concept was added
in the USB 3.0 specification, section 10.1.3.2. Each hub in the topology
is expected to have no more than 15 ports in order for the route string of
a device to be unique. SuperSpeed hubs are restricted to only having 15
ports, but FS/LS/HS hubs are not. The xHCI specification says that if the
port number the device is under is greater than 15, that portion of the
route string shall be set to 15.
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Some commands to the xHCI hardware cannot be allowed to fail due to out of
memory issues or the command ring being full.
Add a way to reserve a TRB on the command ring, and make all command
queueing functions indicate whether they are using a reserved TRB.
Add a way to pre-allocate all the memory a command might need. A command
needs an input context, a variable to store the status, and (optionally) a
completion for the caller to wait on. Change all code that assumes the
input device context, status, and completion for a command is stored in
the xhci virtual USB device structure (xhci_virt_device).
Store pending completions in a FIFO in xhci_virt_device. Make the event
handler for a configure endpoint command check to see whether a pending
command in the list has completed. We need to use separate input device
contexts for some configure endpoint commands, since multiple drivers can
submit requests at the same time that require a configure endpoint
command.
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
The xhci_ring structure contained information that is really related to an
endpoint, not a ring. This will cause problems later when endpoint
streams are supported and there are multiple rings per endpoint.
Move the endpoint state and cancellation information into a new virtual
endpoint structure, xhci_virt_ep. The list of TRBs to be cancelled should
be per endpoint, not per ring, for easy access. There can be only one TRB
that the endpoint stopped on after a stop endpoint command (even with
streams enabled); move the stopped TRB information into the new virtual
endpoint structure. Also move the 31 endpoint rings and temporary ring
storage from the virtual device structure (xhci_virt_device) into the
virtual endpoint structure (xhci_virt_ep).
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Set the max packet size for the default control endpoint on high speed
devices to be 64 bytes. High speed devices always have a max packet size
of 64 bytes. There's no use setting it to eight for the initial 8 byte
descriptor fetch and then issuing (and waiting for) an evaluate context
command to update it to 64 bytes for the subsequent control transfers.
The USB core guesses that the max packet size on a full speed control
endpoint is 64 bytes, and then updates it after the first 8-byte
descriptor fetch. Change the initial setup for the xHCI internal
representation of the full speed device to have a 64 byte max packet size.
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Cc: stable <stable@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Refactor out the code issue, wait for, and parse the event completion code
for a configure endpoint command. Modify it to support the evaluate
context command, which has a very similar submission process. Add
functions to copy parts of the output context into the input context
(which will be used in the evaluate context command).
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Cc: stable <stable@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Different sections of the xHCI 0.95 specification had opposing
requirements for the chain bit in a link transaction request buffer (TRB).
The chain bit is used to designate that adjacent TRBs are all part of the
same scatter gather list that should be sent to the device. Link TRBs can
be in the middle, or at the beginning or end of these chained TRBs.
Sections 4.11.5.1 and 6.4.4.1 both stated the link TRB "shall have the
chain bit set to 1", meaning it is always chained to the next TRB.
However, section 4.6.9 on the stop endpoint command has specific cases for
what the hardware must do for a link TRB with the chain bit set to 0. The
0.96 specification errata later cleared up this issue by fixing the
4.11.5.1 and 6.4.4.1 sections to state that a link TRB can have the chain
bit set to 1 or 0.
The problem is that the xHCI cancellation code depends on the chain bit of
the link TRB being cleared when it's at the end of a TD, and some 0.95
xHCI hardware simply stops processing the ring when it encounters a link
TRB with the chain bit cleared.
Allow users who are testing 0.95 xHCI prototypes to set a module parameter
(link_quirk) to turn on this link TRB work around. Cancellation may not
work if the ring is stopped exactly on a link TRB with chain bit set, but
cancellation should be a relatively uncommon case.
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Cc: stable <stable@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Adds support for controllers that use 64-byte contexts. The following context
data structures are affected by this: Device, Input, Input Control, Endpoint,
and Slot. To accommodate the use of either 32 or 64-byte contexts, a Device or
Input context can only be accessed through functions which look-up and return
pointers to their contained contexts.
Signed-off-by: John Youn <johnyoun@synopsys.com>
Acked-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Make sure the xHCI output device context is 64-byte aligned. Previous
code was using the same structure for both the output device context and
the input control context. Since the structure had 32 bytes of flags
before the device context, the output device context wouldn't be 64-byte
aligned. Define a new structure to use for the output device context and
clean up the debugging for these two structures.
The copy of the device context in the input control context does *not*
need to be 64-byte aligned.
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Allocates and initializes the scratchpad buffer array (XHCI 4.20). This is an
array of 64-bit DMA addresses to scratch pages that the controller may use
during operation. The number of pages is specified in the "Max Scratchpad
Buffers" field of HCSPARAMS2. The DMA address of this array is written into
slot 0 of the DCBAA.
Signed-off-by: John Youn <johnyoun@synopsys.com>
Acked-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
This is a work around for a bug in the SuperSpeed Endpoint Companion Descriptor
parsing code. It fails in some corner cases, which means ep->ss_ep_comp may be
NULL.
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
The xHCI host controller can be programmed to retry a transfer a certain number
of times per endpoint before it passes back an error condition to the host
controller driver. The xHC will return an error code when the error count
transitions from 1 to 0. Programming an error count of 3 means the xHC tries
the transfer 3 times, programming it with a 1 means it tries to transfer once,
and programming it with 0 means the HW tries the transfer infinitely.
We want isochronous transfers to only be tried once, so set the error count to
one.
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
There are several xHCI data structures that use two 32-bit fields to
represent a 64-bit address. Since some architectures don't support 64-bit
PCI writes, the fields need to be written in two 32-bit writes. The xHCI
specification says that if a platform is incapable of generating 64-bit
writes, software must write the low 32-bits first, then the high 32-bits.
Hardware that supports 64-bit addressing will wait for the high 32-bit
write before reading the revised value, and hardware that only supports
32-bit writes will ignore the high 32-bit write.
Previous xHCI code represented 64-bit addresses with two u32 values. This
lead to buggy code that would write the 32-bits in the wrong order, or
forget to write the upper 32-bits. Change the two u32s to one u64 and
create a function call to write all 64-bit addresses in the proper order.
This new function could be modified in the future if all platforms support
64-bit writes.
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Narrow down time spent holding the xHCI spinlock so that it's only used to
protect the xHCI rings, not as mutual exclusion. Stop allocating memory
while holding the spinlock and calling xhci_alloc_virt_device() and
xhci_endpoint_init().
The USB core should have locking in it to prevent device state to be
manipulated by more than one kernel thread. E.g. you can't free a device
while you're in the middle of setting a new configuration. So removing
the locks from the sections where xhci_alloc_dev() and
xhci_reset_bandwidth() touch xHCI's representation of the device should be
OK.
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
xhci-mem.c includes calls to dma_pool_alloc() and other functions defined
in linux/dmapool.h. Make sure to include that header file.
Reported-by: Randy Dunlap <randy.dunlap@oracle.com>
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Differentiate between SuperSpeed endpoint companion descriptor and the
wireless USB endpoint companion descriptor. Make all structure names for
this descriptor have "ss" (SuperSpeed) in them. David Vrabel asked for
this change in http://marc.info/?l=linux-usb&m=124091465109367&w=2
Reported-by: David Vrabel <david.vrabel@csr.com>
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Make all globally visible functions start with xhci_ and mark functions as
static if they're only called within the same C file. Fix some long lines
while we're at it.
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
The 0.95 xHCI spec says that if the xHCI HW support 64-bit addressing, you
must write the whole 64-bit address as one atomic operation, or write the
low 32 bits, and then the high 32 bits. I had the register writes
swapped in some places.
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Turns out someone never built this code on a 64bit platform.
Someone owes me a beer...
Reported-by: Stephen Rothwell <sfr@canb.auug.org.au>
Cc: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Add URB cancellation support to the xHCI host controller driver. This
currently supports cancellation for endpoints that do not have streams
enabled.
An URB is represented by a number of Transaction Request Buffers (TRBs),
that are chained together to make one (or more) Transaction Descriptors
(TDs) on an endpoint ring. The ring is comprised of contiguous segments,
linked together with Link TRBs (which may or may not be chained into a TD).
To cancel an URB, we must stop the endpoint ring, make the hardware skip
over the TDs in the URB (either by turning them into No-op TDs, or by
moving the hardware's ring dequeue pointer past the last TRB in the last
TD), and then restart the ring.
There are times when we must drop the xHCI lock during this process, like
when we need to complete cancelled URBs. We must ensure that additional
URBs can be marked as cancelled, and that new URBs can be enqueued (since
the URB completion handlers can do either). The new endpoint ring
variables cancels_pending and state (which can only be modified while
holding the xHCI lock) ensure that future cancellation and enqueueing do
not interrupt any pending cancellation code.
To facilitate cancellation, we must keep track of the starting ring
segment, first TRB, and last TRB for each URB. We also need to keep track
of the list of TDs that have been marked as cancelled, separate from the
list of TDs that are queued for this endpoint. The new variables and
cancellation list are stored in the xhci_td structure.
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Allow device drivers to submit URBs to bulk endpoints on devices under an
xHCI host controller. Share code between the control and bulk enqueueing
functions when it makes sense.
To get the best performance out of bulk transfers, SuperSpeed devices must
have the bMaxBurst size copied from their endpoint companion controller
into the xHCI device context. This allows the host controller to "burst"
up to 16 packets before it has to wait for the device to acknowledge the
first packet.
The buffers in Transfer Request Blocks (TRBs) can cross page boundaries,
but they cannot cross 64KB boundaries. The buffer must be broken into
multiple TRBs if a 64KB boundary is crossed.
The sum of buffer lengths in all the TRBs in a Transfer Descriptor (TD)
cannot exceed 64MB. To work around this, the enqueueing code must enqueue
multiple TDs. The transfer event handler may incorrectly give back the
URB in this case, if it gets a transfer event that points somewhere in the
first TD. FIXME later.
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Since the xHCI host controller hardware (xHC) has an internal schedule, it
needs a better representation of what devices are consuming bandwidth on
the bus. Each device is represented by a device context, with data about
the device, endpoints, and pointers to each endpoint ring.
We need to update the endpoint information for a device context before a
new configuration or alternate interface setting is selected. We setup an
input device context with modified endpoint information and newly
allocated endpoint rings, and then submit a Configure Endpoint Command to
the hardware.
The host controller can reject the new configuration if it exceeds the bus
bandwidth, or the host controller doesn't have enough internal resources
for the configuration. If the command fails, we still have the older
device context with the previous configuration. If the command succeeds,
we free the old endpoint rings.
The root hub isn't a real device, so always say yes to any bandwidth
changes for it.
The USB core will enable, disable, and then enable endpoint 0 several
times during the initialization sequence. The device will always have an
endpoint ring for endpoint 0 and bandwidth allocated for that, unless the
device is disconnected or gets a SetAddress 0 request. So we don't pay
attention for when xhci_check_bandwidth() is called for a re-add of
endpoint 0.
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Allow device drivers to enqueue URBs to control endpoints on devices under
an xHCI host controller. Each control transfer is represented by a
series of Transfer Descriptors (TDs) written to an endpoint ring. There
is one TD for the Setup phase, (optionally) one TD for the Data phase, and
one TD for the Status phase.
Enqueue these TDs onto the endpoint ring that represents the control
endpoint. The host controller hardware will return an event on the event
ring that points to the (DMA) address of one of the TDs on the endpoint
ring. If the transfer was successful, the transfer event TRB will have a
completion code of success, and it will point to the Status phase TD.
Anything else is considered an error.
This should work for control endpoints besides the default endpoint, but
that hasn't been tested.
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
xHCI needs to get a "Slot ID" from the host controller and allocate other
data structures for every USB device. Make usb_alloc_dev() and
usb_release_dev() allocate and free these device structures. After
setting up the xHC device structures, usb_alloc_dev() must wait for the
hardware to respond to an Enable Slot command. usb_alloc_dev() fires off
a Disable Slot command and does not wait for it to complete.
When the USB core wants to choose an address for the device, the xHCI
driver must issue a Set Address command and wait for an event for that
command.
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
xHCI host controllers can optionally implement a no-op test. This
simple test ensures the OS has correctly setup all basic data structures
and can correctly respond to interrupts from the host controller
hardware.
There are two rings exercised by the no-op test: the command ring, and
the event ring.
The host controller driver writes a no-op command TRB to the command
ring, and rings the doorbell for the command ring (the first entry in
the doorbell array). The hardware receives this event, places a command
completion event on the event ring, and fires an interrupt.
The host controller driver sees the interrupt, and checks the event ring
for TRBs it can process, and sees the command completion event. (See
the rules in xhci-ring.c for who "owns" a TRB. This is a simplified set
of rules, and may not contain all the details that are in the xHCI 0.95
spec.)
A timer fires every 60 seconds to debug the state of the hardware and
command and event rings. This timer only runs if
CONFIG_USB_XHCI_HCD_DEBUGGING is 'y'.
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Instead of keeping a "frame list" like older host controllers, the xHCI
host controller keeps internal representations of the USB devices, with a
transfer ring per endpoint. The host controller queues Transfer Request
Blocks (TRBs) to the endpoint ring, and then "rings the doorbell" for that
device. The host controller processes the transfer, places a transfer
completion event on the event ring, and interrupts the system.
The device context base address array must be allocated by the xHCI host
controller driver, along with the device contexts it points to.
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Allocate basic xHCI host controller data structures. For every xHC, there
is a command ring, an event ring, and a doorbell array.
The doorbell array is used to notify the host controller that work has
been enqueued onto one of the rings. The host controller driver enqueues
commands on the command ring. The HW enqueues command completion events
on the event ring and interrupts the system (currently using PCI
interrupts, although the xHCI HW will use MSI interrupts eventually).
All rings and the doorbell array must be allocated by the xHCI host
controller driver.
Each ring is comprised of one or more segments, which consists of 16-byte
Transfer Request Blocks (TRBs) that can be chained to form a Transfer
Descriptor (TD) that represents a multiple-buffer request. Segments are
linked into a ring using Link TRBs, which means they are dynamically
growable.
The producer of the ring enqueues a TD by writing one or more TRBs in the
ring and toggling the TRB cycle bit for each TRB. The consumer knows it
can process the TRB when the cycle bit matches its internal consumer cycle
state for the ring. The consumer cycle state is toggled an odd amount of
times in the ring.
An example ring (a ring must have a minimum of 16 TRBs on it, but that's
too big to draw in ASCII art):
chain cycle
bit bit
------------------------
| TD A TRB 1 | 1 | 1 |<------------- <-- consumer dequeue ptr
------------------------ | consumer cycle state = 1
| TD A TRB 2 | 1 | 1 | |
------------------------ |
| TD A TRB 3 | 0 | 1 | segment 1 |
------------------------ |
| TD B TRB 1 | 1 | 1 | |
------------------------ |
| TD B TRB 2 | 0 | 1 | |
------------------------ |
| Link TRB | 0 | 1 |----- |
------------------------ | |
| |
chain cycle | |
bit bit | |
------------------------ | |
| TD C TRB 1 | 0 | 1 |<---- |
------------------------ |
| TD D TRB 1 | 1 | 1 | |
------------------------ |
| TD D TRB 2 | 1 | 1 | segment 2 |
------------------------ |
| TD D TRB 3 | 1 | 1 | |
------------------------ |
| TD D TRB 4 | 1 | 1 | |
------------------------ |
| Link TRB | 1 | 1 |----- |
------------------------ | |
| |
chain cycle | |
bit bit | |
------------------------ | |
| TD D TRB 5 | 1 | 1 |<---- |
------------------------ |
| TD D TRB 6 | 0 | 1 | |
------------------------ |
| TD E TRB 1 | 0 | 1 | segment 3 |
------------------------ |
| | 0 | 0 | | <-- producer enqueue ptr
------------------------ |
| | 0 | 0 | |
------------------------ |
| Link TRB | 0 | 0 |---------------
------------------------
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Add PCI initialization code to take control of the xHCI host controller
away from the BIOS, halt, and reset the host controller. The xHCI spec
says that BIOSes must give up the host controller within 5 seconds.
Add some host controller glue functions to handle hardware initialization
and memory allocation for the host controller. The current xHCI
prototypes use PCI interrupts, but the xHCI spec requires MSI-X
interrupts. Add code to support MSI-X interrupts, but use the PCI
interrupts for now.
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
