For the boards that have analog output capability, there are always
4 analog output channels. Convert the 'ao_nchan' member of the boardinfo
into a bit-field, 'has_ao', to save a bit of space and set the analog
output subdevice 'n_chan' to 4 when supported.
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>
This member of the boardinfo is only used as a flag indicating that the
board supports differential analog inputs. Convert the member to a bit-
field to save a bit of space. For aesthetics, rename the member to
'can_do_diff_ai'.
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>
staging: comedi: me4000: remove 'board' from me4000_ai_insn_read()
The 'board' pointer is only used in this function to verify that the
'chan' is valid for an aref of AREF_DIFF. For differential inputs, the
maximum channel is half the subdevice 'n_chan'. Use that instead and
remove the 'board' variable.
Also, the comedi core does not validate the aref flags. Add a check
to ensure that the subdevice actually supports the AREF_DIFF mode.
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>
Neil Horman [Tue, 4 Aug 2015 19:09:47 +0000 (15:09 -0400)]
staging: unisys: visornic: Convert to using napi
Switch the visornic over to use napi. Currently there is a kernel
thread that sits and waits on a wait queue to get notified of incoming
virtual interrupts. It would be nice if we could handle frame reception
using the standard napi processing instead. This patch creates our napi
instance and has the rx thread schedule it
Given that the unisys hypervisor currently requires that queue servicing
be done by a polling loop that wakes up every 2ms, lets instead also
convert that to a timer, which is simpler, and allows us to remove all
the thread starting and stopping code.
Signed-off-by: Neil Horman <nhorman@redhat.com> Signed-off-by: Benjamin Romer <benjamin.romer@unisys.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
drivers: staging: wilc1000: use 'void' for no arguments functions
Added 'void' keyword in the paranthesis of function definitions, when
there are no arguments to the functions. This fixes the checkpatch.pl
error - "Bad function definition 'function()' should probably be
function(void)".
Signed-off-by: Chandra S Gorentla <csgorentla@gmail.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Shraddha Barke [Tue, 4 Aug 2015 19:59:22 +0000 (01:29 +0530)]
Staging: wilc1000: Remove typedefs for struct
The Linux kernel coding style guidelines suggest not using typedefs for
structure and enum types. This patch gets rid of the typedefs for
Ack_session_info_t.
The following Coccinelle semantic patch detects the cases for struct type:
Ever since daemonize was removed in 3.18, there are no longer
any flags passed to kthread_run.
Most of the comments were deleted, but this one lingered on
until now.
staging: lustre: service.c: make local functions static
Makes functions that are not used outside the file in which they are defined
static, as reported by sparse:
drivers/staging/lustre/lustre/ptlrpc/service.c:72:35: warning: symbol
'ptlrpc_alloc_rqbd' was not declared. Should it be static?
1065 drivers/staging/lustre/lustre/ptlrpc/service.c:105:1: warning:
symbol 'ptlrpc_free_rqbd' was not declared. Should it be static?
1066 drivers/staging/lustre/lustre/ptlrpc/service.c:122:1: warning:
symbol 'ptlrpc_grow_req_bufs' was not declared. Should it be static?
1067 drivers/staging/lustre/lustre/ptlrpc/service.c:3055:5: warning:
symbol 'ptlrpc_svcpt_health_check' was not declared. Should it be
static?
Signed-off-by: Zoltán Lajos Kis <zoltan.lajos.kis@gmail.com>" Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
drivers/staging/lustre/lustre/obdecho/echo_client.c:2142:5: warning: symbol 'echo_client_init' was not declared. Should it be static?
drivers/staging/lustre/lustre/obdecho/echo_client.c:2157:6: warning: symbol 'echo_client_exit' was not declared. Should it be static?
Signed-off-by: Zoltán Lajos Kis <zoltan.lajos.kis@gmail.com>" Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Card configuration is stored in SPI EEPROM (93c46 or 93c56)
working in 128|256x16 mode.
Communication is handled using GPIO bitbang.
>From behaviour perspective, delay after read was removed.
It is not needed as we wait after reading GPIO mapped to
PCI-E register - it should have no side effects.
According to sample EEPROM datasheet (AT93Cx6), max frequency for
worst case scenario (1.8V supply) is 250kHZ (vs. 1MHz for 5V).
Driver generates ~50kHZ clock - margin should be big enough
even for devices from other vendors.
Neil Horman [Fri, 31 Jul 2015 22:56:32 +0000 (18:56 -0400)]
staging: unisys: visorchannel: Add peek function
According to unisys, the s_par hypervisor has a bug in which it never
triggers an interrupt. That makes the visornic effectively a 2ms poll
loop. In order to just have the rx thread shceduling a napi poll every
2ms, lets instead give it the chance to check the response queue for
data before we schedule. This helper provides that functionality
Signed-off-by: Neil Horman <nhorman@redhat.com> Signed-off-by: Benjamin Romer <benjamin.romer@unisys.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
staging: android: Allow compile test of GPIO consumers if !GPIOLIB
The GPIO subsystem provides dummy GPIO consumer functions if GPIOLIB is
not enabled. Hence drivers that depend on GPIOLIB, but use GPIO consumer
functionality only, can still be compiled if GPIOLIB is not enabled.
Relax the dependency on GPIOLIB if COMPILE_TEST is enabled, where
appropriate.
staging: fbtft: Allow compile test of GPIO consumers if !GPIOLIB
The GPIO subsystem provides dummy GPIO consumer functions if GPIOLIB is
not enabled. Hence drivers that depend on GPIOLIB, but use GPIO consumer
functionality only, can still be compiled if GPIOLIB is not enabled.
Relax the dependency on GPIOLIB if COMPILE_TEST is enabled, where
appropriate.
staging: fbtft: core: Don't set device platform_data
Pass platform_data as an argument to fbtft_framebuffer_alloc()
instead of using dev->platform_data.
This fixes an issue where the device comes from Device Tree
and fbtft_probe_common() sets dev->platform_data to allocated
memory. When the module is reloaded, dev->platform_data points
to freed memory.
Randy Dunlap [Mon, 3 Aug 2015 18:35:47 +0000 (11:35 -0700)]
staging: most: fix aim-sound build errors
Fix build errors: driver uses snd_pcm*() interfaces, so select
SND_PCM.
drivers/built-in.o: In function `audio_rx_completion':
sound.c:(.text+0x3cd376): undefined reference to `snd_pcm_period_elapsed'
drivers/built-in.o: In function `pcm_prepare':
sound.c:(.text+0x3cd3b0): undefined reference to `snd_pcm_format_physical_width'
sound.c:(.text+0x3cd3ce): undefined reference to `snd_pcm_format_big_endian'
sound.c:(.text+0x3cd42c): undefined reference to `snd_pcm_format_big_endian'
drivers/built-in.o: In function `pcm_hw_free':
sound.c:(.text+0x3cd50a): undefined reference to `snd_pcm_lib_free_vmalloc_buffer'
drivers/built-in.o: In function `pcm_hw_params':
sound.c:(.text+0x3cd54c): undefined reference to `_snd_pcm_lib_alloc_vmalloc_buffer'
drivers/built-in.o: In function `playback_thread':
sound.c:(.text+0x3cd6a0): undefined reference to `snd_pcm_period_elapsed'
drivers/built-in.o: In function `audio_probe_channel':
sound.c:(.text+0x3cdc0b): undefined reference to `snd_pcm_new'
sound.c:(.text+0x3cdc2b): undefined reference to `snd_pcm_set_ops'
drivers/built-in.o:(.data+0x952d0): undefined reference to `snd_pcm_lib_ioctl'
drivers/built-in.o:(.data+0x95318): undefined reference to `snd_pcm_lib_get_vmalloc_page'
Signed-off-by: Randy Dunlap <rdunlap@infradead.org> Acked-by: Christian Gromm <christian.gromm@microchip.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Dan Carpenter [Tue, 28 Jul 2015 15:49:25 +0000 (18:49 +0300)]
staging: rtl8192e: don't just print an error and continue
I was looking at how TOTAL_CAM_ENTRY is used and I saw this code. We
print an error but continue writing "EntryNo" to a register as if it
were valid. "EntryNo" is controlled by the user in rtl8192_ioctl() so
it definitely can be invalid. I'm not positive what happens with the
invalid data but it can't be good.
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Dan Carpenter [Tue, 28 Jul 2015 15:48:21 +0000 (18:48 +0300)]
Staging: rtl8192e: array overflow in rtl92e_set_swcam()
"EntryNo" is comes from the user in the ioctl and it's a number between
0-255. The ieee->swcamtable[] array only has 32 elements so it can
result in memory corruption.
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Peter Huewe [Fri, 24 Jul 2015 23:59:08 +0000 (01:59 +0200)]
staging/xgifb: fix dumpVGAReg compile error if DEBUG is set
If DEBUG is set dumpVGAReg is called and tries to access
XGISR which is defined as (xgifb_info->dev_info.P3c4)
which is not known within this function.
-> add as parameter to dumpVGAReg so xgifb_info becomes known
Signed-off-by: Peter Huewe <peterhuewe@gmx.de> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The public headers is well known by others and the redundant commits make
the code mess. Remove the comments of the public headers to make the code
tidy.
Signed-off-by: Navy Cheng <navych@126.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The code that tests for high and low xmit watermarks was consolidatedand
simplified. The output behavior should be identical, with the exception
of an off-by-one error being corrected in the tests done when the counters
overflowed.
Signed-off-by: Tim Sell <Timothy.Sell@unisys.com> Signed-off-by: Benjamin Romer <benjamin.romer@unisys.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
* A skb->len error-check was enabled (removed from a "#ifdef DEBUG" block).
* Several unneeded "#ifdef DEBUG" blocks were removed.
* A dev_err() was converted to the more-appropriate netdev_err().
Signed-off-by: Tim Sell <Timothy.Sell@unisys.com> Signed-off-by: Benjamin Romer <benjamin.romer@unisys.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Tim Sell [Tue, 28 Jul 2015 16:29:09 +0000 (12:29 -0400)]
staging: unisys: visornic_resume needs to mirror _serverdown_complete
Previously we simplified the serverdown function to basically turn it
into a dev_close(), but missed the analogous logic in visornic_resume()
(which is essentially the "book-end" of visornic_serverdown_complete()).
As a result, during IO partition recovery, the nic would go closed when
the IO partition went away, but would never be opened again when the IO
partition came back.
This patch changes visornic_resume() to use dev_open(), so that it once
again plays nicely with visornic_serverdown_complete(). Because
dev_open() forces us into the visornic_open() path, other logic in
visornic_resume() was no longer necessary, and lended to simplifying
visornic_resume() even more.
Fixes: 36645d72a377 ("staging: unisys: simplify visornic_serverdown_complete") Signed-off-by: Tim Sell <Timothy.Sell@unisys.com> Signed-off-by: Benjamin Romer <benjamin.romer@unisys.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Tim Sell [Tue, 28 Jul 2015 16:29:08 +0000 (12:29 -0400)]
staging: unisys: visornic - correctly reset flag prior to send_enbdis()
Because devdata->enab_dis_acked is the flag used to determine whether an
enable/disable operation to the IO partition has completed, it should
always be cleared prior to initiating the operation. The call added to
visornic_enable_with_timeout() added in this patch makes the usage there
consistent with visornic_disable_with_timeout().
Signed-off-by: Tim Sell <Timothy.Sell@unisys.com> Signed-off-by: Benjamin Romer <benjamin.romer@unisys.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Tim Sell [Fri, 31 Jul 2015 17:21:33 +0000 (13:21 -0400)]
staging: unisys: visornic - prevent lock recursion after IO recovery
In the patch which changed the serverdown logic to by synchronous, we
were
mistakenly holding on to devdata->priv_lock in the call to
visornic_serverdown_complete(), which ultimately ended up recursively
attempting to grab the same lock via the path:
drivers, staging, unisys Add modalias files to visorbus devices
This patch adds modalias files that export the device UUID type to sysfs
so that udev can autoload the appropriate device driver on demand. Note
that is required a minor name change to the channel device sysfs files
which are currently named visorbus_dev_groups, and are now named
visorbus_channel_groups.
Staging : rtl8188eu : os_dep : Compression of lines for immediate return
This patch compresses two lines into a single line if immediate return is
found. Also remove variable ret as it is no longer needed.
Semantic patch used for this is as follows:
@@
type T;
identifier i,f;
constant C;
@@
- T i;
...when != i
when strict
(
return -C;
|
- i =
+ return
f(...);
- return i;
)
staging/lustre/llite: Don't set page writeback on non-dirty page
New writeback changes in 4.2-RC1 have exposed that we incorrectly
set page_writeback on a page that is being written synchronously,
which aside from this new crash (dereference of NULL inode->i_wb
from set_page_writeback) likely threw off some related page
statistics in the past.
projects / linux-beck.git / log