For aesthetics, add some whitespace to the subdevice initialization.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
For aesthetics, factor the board reset and detection code out of the
(*attach) function.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
For aesthetics, rename this function to dmm32at_ai_insn_read().
Get the 'chan' and 'range' from the chanspec when declaring the local
variables. Remove the unecessary masking of the 'chan'.
Remove some unnecessary comments.
Change the final return to 'insn->n' to clarify the return value.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Introduce a helper function to read the two's complement analog input
sample from the hardware and munge it to the offset binary (unsigned)
format that comedi expects. Use the comedi_offset_munge() helper to
munge the data.
Use the new helper in the analog input (*insn_read) and in the interrupt
handler for the async command.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Remove the private data member 'ai_scans_left' and use the comedi_async
'scans_done' to detect the end-of-acquisition.
This also removes the artifical max limit on the cmd->stop_arg.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Currently the AI (*cancel) changes a private data member to cause the
interrupt handler to cancel the async command after the next interrupt.
Move the code that disables the interrupt and terminates the acquisition
from the interrunt handler into dmm32at_ai_cancel() so that the async
command is terminated instantly.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The comedi core now counts the number of samples added to the async buffer and
detects the end-of-scan and increments the comedi_async 'scans_done' counter.
Remove the private data member 'stop_counter' and use the 'scans_done' member
to detect the end-of-acquisition.
This fixes a possible interger overflow when calculating the value of the
'stop_counter' and removes the need to accumulate the 'total' number of
samples added to the async buffer in pci9111_handle_fifo_half_full().
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Remove the assumption of the sample size by using the comedi_bytes_to_samples()
helper function to convert the number of bytes to the number of samples.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Absorb this simple helper function into interrupt_pci9118_ai_dma().
Remove the unnecessary local variables 'sampls' and 'm'.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The comedi core now counts the number of samples added to the async buffer and
detects the end-of-scan and increments the comedi_async 'scans_done' counter.
Remove the private data member 'ai_act_scan' and use the 'scans_done' member
to detect the end-of-acquisition.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Use the generic TCW (timer/counter/watchdog) defines for the 8-bit watchdog.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Remove the error return in the (*auto_attach) and allow the driver to
attach to APC-1564 boards that use the PLD Rev 1.0 I/O mapping.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Use the generic TCW (timer/counter/watchdog) defines for the 32-bit counters.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Use the generic TCW (timer/counter/watchdog) defines for the 12-bit timer.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The TCW (timer/counter/watchdog) devices in the various ADDI-DATA drivers
use a common register map definition. Provide a common generic define for
these registers so they don't have to be replicated in each driver.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The timer subdevice is currently broken in this driver.
The Rev 1.0 and 2.x versions of the board both have a 12-bit timer. But only
the Rev 2.x boards have the 3 32-bit counters.
Split the current timer subdevice into two separate subdevices:
1) A single channel 12-bit timer subdevice
2) A three channel 32-bit counter subdevice
This represents the hardware correctly and the counters can be disabled on
the Rev 1.0 boards.
Split up the current (*insn_config), (*insn_write), and (*insn_read) so they
only deal with the hardware associated with the subdevice.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Move the defines for the counter registers from the included source
file to the main driver source file.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The Rev 1.0 APCI-1564 boards do not have counters.
Fix the code in the main driver source so that the I/O accesses to the
counters do not happen if the devpriv->counters member is not initialized.
This does not fix the code in hwdrv_apci1564.c. That code violates the
comedi API and is currently broken.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The APCI-1564 has different I/O mapping depending on if the PLD revision
is Rev 1.0 or Rev 2.x.
Add a member, 'timer', to the private data for the base address of the
12-bit timer. Fix the register defines so they will work for all PLD
revisions and initialize the devpriv->timer appropriately.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The APCI-1564 has different I/O mapping depending on if the PLD revision is
Rev 1.0 or Rev 2.x.
Fix the main register defines so they will work for all PLD revisions and
initialie the dev->iobase appropriately. Move the register defines to the
main driver source file.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The APCI-1564 has different I/O mapping depending on if the PLD revision is
Rev 1.0 or Rev 2.x. The revision can be determined by reading the EEPROM
register at offset 0x00 of PCI BAR 0 and checking the value of bits 7 to 4.
Add this check to apci1564_auto_attach(). Currently this driver is coded
to work with the Rev 2.x I/O mapping. For now, fail the attach if a Rev 1.0
PLD is detected. Document the I/O mapping for both revisions.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
According to ADDI-DATA, the PLD Revision 2.x versions of the APCI-1564 use
PCI BAR 0 for the main registers of the board.
Remove the 'amcc_iobase' member of the private data and use the dev->iobase
to store the base address of PCI BAR 1.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
According to ADDI-DATA, only the PLD Revision 2.x versions of the APCI-1564
have the 3 counters. The base address for these counters is found in PCI BAR 1.
For aesthetics, save this base address in the private data. The dev->iobase
can then be used for the main registers of the board.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The APCI1564_DI_INT_STATUS_REG is located in the PCI BAR 0 I/O space. That
base address is stored in devpriv->amcc_iobase. Use that to correctly read
the register.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
According to ADDI-DATA, the APCI-1564 board has a FPGA with a PCI core, it
does not use an AMCC PCI controller chip.
Remove the amcc_s5933.h include as well as the read of the AMCC_OP_REG_INTCSR
register. That offset would actually read APCI1564_WDOG_STATUS_REG register
in the FPGA which has nothing to do with the interrupt status.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
A rather obvious typo in one of the identifier has been found. This patch
fixes the typo and ensures any lines changed do not exceed 80 characters
as indicated by scripts/checkpatch.pl
Signed-off-by: Krzysztof Konopko <kris@konagma.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
rtw_led.* files include Free Software Foundation's mailing address in
the sample GPL notice. This is not desired and picked when running
scripts/checkpatch.pl.
Signed-off-by: Krzysztof Konopko <kris@konagma.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
scripts/checkpatch.pl reports a coding style problem in xmit_linux.c
WARNING:BRACES: braces {} are not necessary for single statement blocks
#61852: FILE: rtl8188eu/os_dep/xmit_linux.c:70:
This patch removes unnecessary braces and simplifies the function to a
single return statement.
Signed-off-by: Krzysztof Konopko <kris@konagma.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
I used some 64 bit instructions when adding the 32 bit getcpu VDSO
function. Fix it.
Fixes: 18ad51dd34 ("powerpc: Add VDSO version of getcpu")
Cc: stable@vger.kernel.org
Signed-off-by: Anton Blanchard <anton@samba.org>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
The flag passed to ioda_eeh_phb_reset() should be EEH_RESET_DEACTIVATE,
which is translated to OPAL_DEASSERT_RESET or something else by the
EEH backend accordingly.
The patch replaces OPAL_DEASSERT_RESET with EEH_RESET_DEACTIVATE for
ioda_eeh_phb_reset().
Cc: stable@vger.kernel.org
Signed-off-by: Gavin Shan <gwshan@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Obviously I had wrong format given to the PE state output from
/sys/bus/pci/devices/xxxx/eeh_pe_state with some typoes, which
was introduced by commit 2013add4ce. The patch fixes it up.
Fixes: 2013add4ce ("powerpc/eeh: Show hex prefix for PE state sysfs")
Signed-off-by: Gavin Shan <gwshan@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
On pseries system (LPAR) xmon failed to enter when running in LE mode,
system is hunging. Inititating xmon will lead to such an output on the
console:
SysRq : Entering xmon
cpu 0x15: Vector: 0 at [c0000003f39ffb10]
pc: c00000000007ed7c: sysrq_handle_xmon+0x5c/0x70
lr: c00000000007ed7c: sysrq_handle_xmon+0x5c/0x70
sp: c0000003f39ffc70
msr: 8000000000009033
current = 0xc0000003fafa7180
paca = 0xc000000007d75e80 softe: 0 irq_happened: 0x01
pid = 14617, comm = bash
Bad kernel stack pointer fafb4b0 at eca7cc4
cpu 0x15: Vector: 300 (Data Access) at [c000000007f07d40]
pc: 000000000eca7cc4
lr: 000000000eca7c44
sp: fafb4b0
msr: 8000000000001000
dar: 10000000
dsisr: 42000000
current = 0xc0000003fafa7180
paca = 0xc000000007d75e80 softe: 0 irq_happened: 0x01
pid = 14617, comm = bash
cpu 0x15: Exception 300 (Data Access) in xmon, returning to main loop
xmon: WARNING: bad recursive fault on cpu 0x15
The root cause is that xmon is calling RTAS to turn off the surveillance
when entering xmon, and RTAS is requiring big endian parameters.
This patch is byte swapping the RTAS arguments when running in LE mode.
Cc: stable@vger.kernel.org
Signed-off-by: Laurent Dufour <ldufour@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
The current HMI event structure is an ABI and carries a version field to
accommodate future changes without affecting/rearranging current structure
members that are valid for previous versions.
The current version check "if (hmi_evt->version != OpalHMIEvt_V1)"
doesn't accomodate the fact that the version number may change in
future.
If firmware starts returning an HMI event with version > 1, this check
will fail and no HMI information will be printed on older kernels.
This patch fixes this issue.
Cc: stable@vger.kernel.org # 3.17+
Signed-off-by: Mahesh Salgaonkar <mahesh@linux.vnet.ibm.com>
[mpe: Reword changelog]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
In commit fadbe0cd52 entitled "staging:
rtl8188eu:Remove rtw_zmalloc(), wrapper for kzalloc()", the author failed
to note that the original code in the wrapper tested whether the caller
could sleep, and set the flags argument to kzalloc() appropriately.
After the patch, GFP_KERNEL is used unconditionally. Unfortunately, several
of the routines may be entered from an interrupt routine and generate
a BUG splat for every such call. Routine rtw_sitesurvey_cmd() is used in the
example below:
BUG: sleeping function called from invalid context at mm/slub.c:1240
in_atomic(): 1, irqs_disabled(): 0, pid: 756, name: wpa_supplicant
INFO: lockdep is turned off.
CPU: 2 PID: 756 Comm: wpa_supplicant Tainted: G WC O 3.18.0-rc4+ #34
Hardware name: TOSHIBA TECRA A50-A/TECRA A50-A, BIOS Version 4.20 04/17/2014
ffffc90005557000 ffff880216fafaa8 ffffffff816b0bbf 0000000000000000
ffff8800c3b58000 ffff880216fafac8 ffffffff8107af77 0000000000000001
0000000000000010 ffff880216fafb18 ffffffff811b06ce 0000000000000000
Call Trace:
[<ffffffff816b0bbf>] dump_stack+0x4e/0x71
[<ffffffff8107af77>] __might_sleep+0xf7/0x120
[<ffffffff811b06ce>] kmem_cache_alloc_trace+0x4e/0x1f0
[<ffffffffa0888226>] ? rtw_sitesurvey_cmd+0x56/0x2a0 [r8188eu]
[<ffffffffa0888226>] rtw_sitesurvey_cmd+0x56/0x2a0 [r8188eu]
[<ffffffffa088f00d>] rtw_do_join+0x22d/0x370 [r8188eu]
[<ffffffffa088f6e8>] rtw_set_802_11_ssid+0x218/0x3d0 [r8188eu]
[<ffffffffa08c3ca5>] rtw_wx_set_essid+0x1e5/0x410 [r8188eu]
[<ffffffffa08c3ac0>] ? rtw_wx_get_rate+0x50/0x50 [r8188eu]
[<ffffffff816938f1>] ioctl_standard_iw_point+0x151/0x3f0
[<ffffffff81693d52>] ioctl_standard_call+0xb2/0xe0
[<ffffffff81597df7>] ? rtnl_lock+0x17/0x20
[<ffffffff816945a0>] ? iw_handler_get_private+0x70/0x70
[<ffffffff81693ca0>] ? call_commit_handler+0x40/0x40
[<ffffffff81693256>] wireless_process_ioctl+0x176/0x1c0
[<ffffffff81693e79>] wext_handle_ioctl+0x69/0xc0
[<ffffffff8159fe79>] dev_ioctl+0x309/0x5e0
[<ffffffff810be9c7>] ? call_rcu+0x17/0x20
[<ffffffff8156a472>] sock_ioctl+0x142/0x2e0
[<ffffffff811e0c70>] do_vfs_ioctl+0x300/0x520
[<ffffffff81101514>] ? __audit_syscall_entry+0xb4/0x110
[<ffffffff81101514>] ? __audit_syscall_entry+0xb4/0x110
[<ffffffff810102bc>] ? do_audit_syscall_entry+0x6c/0x70
[<ffffffff811e0f11>] SyS_ioctl+0x81/0xa0
[<ffffffff816ba1d2>] system_call_fastpath+0x12/0x17
Additional routines that generate this BUG are rtw_joinbss_cmd(),
rtw_dynamic_chk_wk_cmd(), rtw_lps_ctrl_wk_cmd(), rtw_rpt_timer_cfg_cmd(),
rtw_ps_cmd(), report_survey_event(), report_join_res(), survey_timer_hdl(),
and rtw_check_bcn_info().
Signed-off-by: Larry Finger <Larry.Finger@lwfinger.net>
Cc: navin patidar <navin.patidar@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The powersave tasklet is created in rtllib_softmac_init and
it is not removed while unloading the module.
Cc: Larry Finger <Larry.Finger@lwfinger.net>
Signed-off-by: Devendra Naga <devendra.aaru@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This patch fixes the checkpatch.pl warning:
WARNING:Missing a blank line after declaration
Signed-off-by: Anjana Sasindran <anjanasasindran123@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The backlight_device_unregister() function tests whether its argument is NULL
and then returns immediately. Thus the test around the call is not needed.
This issue was detected by using the Coccinelle software.
Signed-off-by: Markus Elfring <elfring@users.sourceforge.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The oz_free_urb_link() function tests whether its argument is NULL
and then returns immediately. Thus the test around the call is not needed.
This issue was detected by using the Coccinelle software.
Signed-off-by: Markus Elfring <elfring@users.sourceforge.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This patch fixes the multiple blank lines issue complained by checkpatch.pl
by removing useless blank lines.
Signed-off-by: Le Tan <tamlokveer@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Give DEFAULT_KEYPAD and DEFAULT_LCD defines better names,
so that their meaning is emphasized.
Signed-off-by: Mariusz Gorski <marius.gorski@gmail.com>
Acked-By: Willy Tarreau <w@1wt.eu>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Change the order of the module parameter declarations
so that it matches the Kconfig order.
Signed-off-by: Mariusz Gorski <marius.gorski@gmail.com>
Acked-By: Willy Tarreau <w@1wt.eu>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Change the order of the DEFAULT_* values redefines so that it
matches the Kconfig order.
Signed-off-by: Mariusz Gorski <marius.gorski@gmail.com>
Acked-By: Willy Tarreau <w@1wt.eu>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Change the order of the initial DEFAULT_* defines so that it
matches the Kconfig order.
Signed-off-by: Mariusz Gorski <marius.gorski@gmail.com>
Acked-By: Willy Tarreau <w@1wt.eu>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Eliminate cvmx_dprintf() usage and use dev_err() instead.
Signed-off-by: Aaro Koskinen <aaro.koskinen@iki.fi>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Move helper functions to make them visible to all functions.
Signed-off-by: Aaro Koskinen <aaro.koskinen@iki.fi>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Octeon Ethernet driver is currently trying to wake up multiple
CPU cores for parallel NAPI processing when needed, to cope
with the fact that all network interfaces are handled through
the same receive queue processed by a single core/interrupt
(see http://marc.info/?l=linux-kernel&m=137209784914565&w=2).
This partially overlaps with generic RPS functionality.
Commit a6a39a7fcf ("staging: octeon-ethernet: disable load balance
for receiving packet when CONFIG_RPS is enabled") already disabled this
when RPS is enabled. Let's delete this code altogether as it has issues
like packet reordering and potential problems with CPU offlining/onlining.
Tested on EdgeRouter Lite.
Signed-off-by: Aaro Koskinen <aaro.koskinen@iki.fi>
Acked-By: David Daney <david.daney@cavium.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The functions ion_heap_destroy() and vfree() perform also input
parameter validation. Thus the test around the call is not needed.
This issue was detected by using the Coccinelle software.
Signed-off-by: Markus Elfring <elfring@users.sourceforge.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
John W. Linville says:
====================
pull request: wireless 2014-11-26
Please pull this little batch of fixes intended for the 3.18 stream...
For the iwlwifi one, Emmanuel says:
"Not all the firmware know how to handle the HOT_SPOT_CMD.
Make sure that the firmware will know this command before
sending it. This avoids a firmware crash."
Along with that, Larry sends a pair of rtlwifi fixes to address some
discrepancies from moving drivers out of staging. Larry says:
"These two patches are needed to fix a regression introduced when
driver rtl8821ae was moved from staging to the regular wireless tree."
Please let me know if there are problems!
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
H Hartley Sweeten