kbuild bot whinges due to print format mistmatch caused by
zram: force disksize setting before using zram.
This patch fixes it.
Reported-by: Wu Fengguang <fengguang.wu@intel.com>
Signed-off-by: Minchan Kim <minchan@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Since 7c65fa2a4c
'staging: vt6656: Remove QWORD from source and replace with u64.'
a new sparse warning showed up:
'drivers/staging/vt6656/card.c:798:26: sparse: constant
0xffffffff00000000U is so big it is unsigned long'
-> Append L to fix the warning.
Reported-by: Fengguang Wu <fengguang.wu@intel.com>
Signed-off-by: Peter Huewe <peterhuewe@gmx.de>
Cc: Forest Bond <forest@alittletooquiet.net>
Cc: Malcolm Priestley <tvboxspy@gmail.com>
Cc: "Justin P. Mattock" <justinmattock@gmail.com>
Cc: Marcos Paulo de Souza <marcos.souza.org@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
MCLKData does not contain any 0x1C value for its field SR28 nor does
XGI340_ECLKData contain any 0x1C or 0x22 value for its field SR2E.
-> the statement always evaluates to false.
-> remove
Signed-off-by: Peter Huewe <peterhuewe@gmx.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
pVBInfo->ram_type is assigned the return value of
XGINew_GetXG20DRAMType which can only be 0, 1 or 2
-> The conditional assignment is not needed here as it always evaluates
to true.
-> remove
Signed-off-by: Peter Huewe <peterhuewe@gmx.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In the lookup table XGI340_ECLKData only the first three 'lines' are
used, thus the remaining entries can be removed.
Rationale:
- vb_init.c uses pVBInfo->ram_type as the index for XGI340_ECLKData.
- pVBInfo->ram_type is assigned the return value of
XGINew_GetXG20DRAMType which can only be 0, 1 or 2
-> only the first three values are used.
-> remove the remeining entries.
Signed-off-by: Peter Huewe <peterhuewe@gmx.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In the lookup tables XGI340New_MCLKData and XGI27New_MCLKData only the
first three 'lines' are used, thus the remaining fields can be removed.
Rationale:
- vb_init.c uses pVBInfo->ram_type as the index for pVBInfo->MCLKData
- pVBInfo->ram_type is assigned the return value of
XGINew_GetXG20DRAMType which can only be 0, 1 or 2
- pVBInfo->MCLKData is assigned to either XGI340New_MCLKData or
XGI27New_MCLKData in vb_setmode.c
-> only the first three values are used, the rest can be removed.
Signed-off-by: Peter Huewe <peterhuewe@gmx.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In the lookup tables XG27_SR13 and XGI340_SR13 only the first three
values are used of each 'line', thus the remaining fields can be
removed.
Rationale:
- vb_init.c uses pVBInfo->ram_type for the 'column' index for
pVBInfo->SR15.
- pVBInfo->ram_type is assigned the return value of
XGINew_GetXG20DRAMType which can only be 0, 1 or 2
- pVBInfo->SR15 is assigned to either XG27_SR13 or XGI340_SR13 in
vb_setmode.c
-> only the first three values are used.
This becomes also evident as values 3-7 are all 0.
Signed-off-by: Peter Huewe <peterhuewe@gmx.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In the lookup tables XGI340_cr41 and XGI27_cr41 only the first three
values are used of each 'line', thus the remaining fields can be
removed.
Rationale:
- vb_init.c uses pVBInfo->ram_type for the 'column' index for
pVBInfo->CR40.
- pVBInfo->ram_type is assigned the return value of
XGINew_GetXG20DRAMType which can only be 0, 1 or 2
- pVBInfo->CR40 is assigned to either XGI340_cr41 or XGI27_cr41 in
vb_setmode.c
-> only the first three values are used.
This becomes also evident as values 3-7 are all 0.
Signed-off-by: Peter Huewe <peterhuewe@gmx.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
After the patch
'staging/xgifb: Don't write the same values x times'
the local variable i is unused, which leads to the following warning:
driverrs/staging/xgifb/vb_init.c: In function
‘XGINew_SetDRAMDefaultRegister340’:
drivers/staging/xgifb/vb_init.c:433:43: warning: unused variable ‘i’
[-Wunused-variable]
This patch fixes this
Signed-off-by: Peter Huewe <peterhuewe@gmx.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
New drivers
1) A driver for ST microelectronics sensors. This driver already covers
a large set of new parts (20 gyros, accelerometer and magnetometers)
not currently covered by the existing drivers. The intent moving forward
is to merge this with the other drivers for similar parts already in tree.
The lis3l02dq driver currently in staging/iio will be trivial, the lis3
driver in misc more complex as it has a number of additional interfaces.
Any merging in of the lis3 driver will rely on the not currently
merged iio_input bridge driver and handling of freefall notifications
etc.
2) A driver for the itg3200 gyroscope.
Graduations from staging
1) Cleanup and move out of staging of the adxrs450 gyroscope driver. The
cleanup required was all minor but there were a couple of fixes hidden in
there.
Core and driver additions
1) Initial work from Guenter Roeck on device tree support for IIO's provider/
consumer code. Focuses on the iio_hwmon driver and the max1363 adc driver.
The full device tree syntax is currently under discussion but should
follow shortly.
Cleanups and fixes
1) Remove a noop function __iio_update_buffer
2) Couple of small fixes and cleanups for the max1363
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.19 (GNU/Linux)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=E5mK
-----END PGP SIGNATURE-----
Merge tag 'iio-for-3.9c' of git://git.kernel.org/pub/scm/linux/kernel/git/jic23/iio into staging-next
Jonathan writes:
"Third set of IIO new drivers, cleanups and fixes for the 3.9 cycle
New drivers
1) A driver for ST microelectronics sensors. This driver already covers
a large set of new parts (20 gyros, accelerometer and magnetometers)
not currently covered by the existing drivers. The intent moving forward
is to merge this with the other drivers for similar parts already in tree.
The lis3l02dq driver currently in staging/iio will be trivial, the lis3
driver in misc more complex as it has a number of additional interfaces.
Any merging in of the lis3 driver will rely on the not currently
merged iio_input bridge driver and handling of freefall notifications
etc.
2) A driver for the itg3200 gyroscope.
Graduations from staging
1) Cleanup and move out of staging of the adxrs450 gyroscope driver. The
cleanup required was all minor but there were a couple of fixes hidden in
there.
Core and driver additions
1) Initial work from Guenter Roeck on device tree support for IIO's provider/
consumer code. Focuses on the iio_hwmon driver and the max1363 adc driver.
The full device tree syntax is currently under discussion but should
follow shortly.
Cleanups and fixes
1) Remove a noop function __iio_update_buffer
2) Couple of small fixes and cleanups for the max1363
"
Some low-level comedi drivers (incorrectly) point `dev->read_subdev` or
`dev->write_subdev` to a subdevice that does not support asynchronous
commands. Comedi's poll(), read() and write() file operation handlers
assume these subdevices do support asynchronous commands. In
particular, they assume `s->async` is valid (where `s` points to the
read or write subdevice), which it won't be if it has been set
incorrectly. This can lead to a NULL pointer dereference.
Check `s->async` is non-NULL in `comedi_poll()`, `comedi_read()` and
`comedi_write()` to avoid the bug.
Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Cc: stable <stable@vger.kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This module parameter is used to enable the auto config mechanism
in the comedi core. Most of the PCI, PCMCIA, and USB drivers have
been converted to use the auto config mechanism and will not attach
if it is disabled.
Since the 'comedi_autoconfig' parameter is defaulted to true, just
remove it so that the comedi drivers that use auto config will
always be able to attach.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Cc: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Factor the code that programs the analog input scanlist out of the
daqp_ai_insn_read() and daqp_ai_cmd() functions.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Cc: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The (*insn_write) functions are expected to write 'insn->n' number
of samples to the channel. Fix this function so it works as the
comedi core expects.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Cc: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Change the subdevice operation used to write the digital outputs from
a (*insn_write) to a (*insn_bits) function. The (*insn_write) functions
are expected to write 'insn->n' number of samples. The (*insn_bits)
functions just write a single sample (insn->n = 1). Change the return
from '1' to 'insn->n' to clarify what the return is.
Using an (*insn_bits) function to write the digital outputs also allows
the user to read the current state of the output channels.
Fix the io operation used to update the digital outputs. The register is
only 8-bits and should by updated with an outb() not an outw().
Also, set the 'maxdata' for the subdevice. For digital io this value
should be '1' (digital io can only be 1 or 0).
Remove the setting of the len_chanlist for the subdevice. This variable
only has meaning for subdevices that support asynchronous commands. The
comedi core will initialize it appropriately during the postconfig.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Cc: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Change the subdevice operation used to read the digital inputs from
a (*insn_read) to a (*insn_bits) function. The (*insn_read) functions
are expected to read 'insn->n' number of samples. The (*insn_bits)
functions just read a single sample (insn->n = 1). Change the return
from '1' to 'insn->n' to clarify what the return is.
Also, set the 'maxdata' for the subdevice. For digital io this value
should be '1' (digital io can only be 1 or 0).
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Cc: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The MAX_DEV define is no longer used in this driver. Remove it.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Cc: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The private data in this driver is associated with the comedi_device
pointer not the pcmcia_device. For aesthetic reasons, rename the
private data struct from local_into_t to daqp_private.
Also, rename the local variables used for the private data from
local to devpriv as that is more common in the comedi drivers.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Cc: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
All the comedi_subdevices in this driver use the same 'local' private data.
Instead of storing the pointer to the private data in each subdevice, just
store it in the comedi_device.
Also, instead of passing the private data pointer to pcmcia_request_irq()
in link->priv, pass the comedi_device pointer. This allows removing the
comedi_device and comedi_subdevice pointers from the private data. We can
get them as needed from the void * in the interrupt function.
Fix the pcmcia suspend/resume functions so they use the comedi_device
pointer stored in link->priv to then get the private data pointer.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Cc: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The analog output subdevice does not support aynchronous commands.
Don't set the dev->write_subdev to this subdevice.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Cc: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The (*detach) function, daqp_detach() is now a simple wrapper around
comedi_pcmcia_disable(). Just use comedi_pcmcia_disable() directly for
the (*detach).
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Cc: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Use the comedi_pcmcia_{enable,disable} helpers to enable/disable
the PCMCIA device.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Cc: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Convert this pcmcia driver to the comedi auto attach mechanism.
This allows getting rid of the static data that held "A list of
"instances" of the device" that was used to pass the pcmcia_device
pointer from the pcmcia_driver to the comedi_driver.
Also, remove the now unused variables in the private data that
were used to pass the pcmcia_device.
Remove the code that pulled the model number of the card out of
the PCMCIA CIS and stashed it in the private data since it is
not used in the driver.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Cc: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
1) The same condition (pVBInfo->IF_DEF_LVDS == 0) was checked in the if
clause directly in front of this one.
2) The same condition pVBInfo->VBType & (VB_SIS301B | VB_SIS302B |
VB_SIS301LV | VB_SIS302LV | VB_XGI301C) was checked in the if
clause directly in front of this one.
Signed-off-by: Peter Huewe <peterhuewe@gmx.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Since XGI_LCDDesStruct is fully contained in XGI330_LCDDataDesStruct2
and the offsets for the first members is identical we can consolidate
the if/else branches here and use XGI330_LCDDataDesStruct2 for
everything.
Signed-off-by: Peter Huewe <peterhuewe@gmx.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
With the previous patch 'Move duplicated code for dram to helper
function' it becomes evident that the code is performing the same
thing 4 or 8 times in a row without changing arguments to the function
and thus writing the same values over and over again.
It was tested that these repeats are unnecessary.
-> we can safely remove them.
Tested-by: Aaro Koskinen <aaro.koskinen@iki.fi>
Signed-off-by: Peter Huewe <peterhuewe@gmx.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
XGINew_SetDRAMDefaultRegister340 uses the same code fragment 4 times
with only a slight variation each time.
-> Move this code to a helper function - this saves some lines and
~450bytes in the .o / .ko
Tested-by: Aaro Koskinen <aaro.koskinen@iki.fi>
Signed-off-by: Peter Huewe <peterhuewe@gmx.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The XGI340_CR6B lookup table consists of the entries
{0xaa, 0xaa, 0xaa, 0xaa} for an index <= 2 and
{0x00, 0x00, 0x00, 0x00} for all other indices.
The only user XGINew_SetDRAMDefaultRegister340 loops over these 4 values
of a line with a for loop and since all entries are the same for each line
we can simply replace the whole lookup table with a simple if/else assignment.
Tested-by: Aaro Koskinen <aaro.koskinen@iki.fi>
Signed-off-by: Peter Huewe <peterhuewe@gmx.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The (*detach) function, daq700_detach() is now a simple wrapper around
comedi_pcmcia_disable(). Just use comedi_pcmcia_disable() directly for
the (*detach).
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Cc: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Use the comedi_pcmcia_{enable,disable} helpers to enable/disable
the PCMCIA device.
This driver does not use interrupts and is not an audio device,
remove CONF_ENABLE_IRQ and CONF_AUTO_AUDIO from the link->config_flags.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Cc: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Use the comedi_pcmcia_{enable,disable} helpers to enable/disable
the PCMCIA device.
This driver does not use interrupts so remove CONF_ENABLE_IRQ from
the link->config_flags.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Cc: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The (*detach) function, das16cs_detach() is a simple wrapper around
comedi_pcmcia_disable(). Just use comedi_pcmcia_disable() directly
for the (*detach).
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Cc: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Use pcmcia_request_irq() instead of request_irq() to request the
irq for the PCMCIA device. This allows the PCMCIA core to clean
up the registration in pcmcia_disable_device().
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Cc: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Use the comedi_pcmcia_{enable,disable} helpers to enable/disable
the PCMCIA device.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Cc: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Introduce some helper functions to enable/disable the PCMCIA device.
This will allow removing some of the boilerplate code in the comedi
PCMCIA drivers.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Cc: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This patch simplifies the code of XGISetModeNew by reordering the
if/else if/case conditions when both branches are doing exactly the
same.
Signed-off-by: Peter Huewe <peterhuewe@gmx.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
These two functions share the same code except one line - thus we can
simply merge them and add a parameter to switch between both variants.
Signed-off-by: Peter Huewe <peterhuewe@gmx.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Since data can only be 0x0000, 0x0035 or 0x0048 we can simply skip the
bit shifting and masking as data & 0xFF is always equal to data and
data & 0xFF00 is always 0.
So we simply use data and 0 directly and save the assignment.
Signed-off-by: Peter Huewe <peterhuewe@gmx.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Instead of subtracting one and then assign a different name and add 1
again we simply use HDE directly. HDE wasn't used directly before, so no
change in functionality.
Same applies to VDE.
-> now we can remove the variable with the very descriptive name E ;)
Signed-off-by: Peter Huewe <peterhuewe@gmx.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
We don't need to use this switch-case here for a simple two case
if-else.
Signed-off-by: Peter Huewe <peterhuewe@gmx.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Since SR[0] in the (readonly) XGI330_StandTable is always 0x01 we can
skip or'ing with 0x01 and make the code simpler by removing the if
statements.
Since this function is the only user of the XGI330_StandTable we can
also include the unconditional |= 0x20 into the input data and move the
assignment to SR1 into the loop, which I prefer to start at 0.
Signed-off-by: Peter Huewe <peterhuewe@gmx.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Clang warning:
drivers/staging/xgifb/XGI_main_26.c: warning: variable 'ret' is
used uninitialized whenever 'if' condition:
if (xgifb_info->mode_idx < 0) {
evaluates to true.
drivers/staging/xgifb/XGI_main_26.c:
note: uninitialized use occurs here
return ret;
This patch initializes the variable in this case.
Signed-off-by: Peter Huewe <peterhuewe@gmx.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The mttr field must be declared as signed int (as in every other fb
driver) for the mttr functions to work properly.
Moreover the value should be initialized with -1.
Signed-off-by: Peter Huewe <peterhuewe@gmx.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This patch removes some comparisons that always evaluate to false since
xoffset and yoffset are defined as __u32 in fb_var_screeninfo in
include/linux/fb.h and thus can never be negative.
Signed-off-by: Peter Huewe <peterhuewe@gmx.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Since the smaller LCDRefreshIndex is contained identically in LCDARefreshIndex
we can simply use LCDARefreshIndex and skip the if/else.
Since LCDARefreshIndex is only used readonly and contains only small
unsigned values we also change its declaration to const u8.
In order to prevent an out-of-bounds access I changed the mask from 0x0F
to 0x07 and added a dummy value.
Signed-off-by: Peter Huewe <peterhuewe@gmx.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
1) User of zram normally do mkfs.xxx or mkswap before using
the zram block device(ex, normally, do it at booting time)
It ends up allocating such metadata of zram before real usage so
benefit of lazy initialzation would be mitigated.
2) Some user want to use zram when memory pressure is high.(ie, load zram
dynamically, NOT booting time). It does make sense because people don't
want to waste memory until memory pressure is high(ie, where zram is really
helpful time). In this case, lazy initialzation could be failed easily
because we will use GFP_NOIO instead of GFP_KERNEL for avoiding deadlock.
So the benefit of lazy initialzation would be mitigated, too.
3) Metadata overhead is not critical and Nitin has a plan to diet it.
4K : 12 byte(64bit machine) -> 64G : 192M so 0.3% isn't big overhead
If insane user use such big zram device up to 20, it could consume 6% of ram
but efficieny of zram will cover the waste.
So this patch gives up lazy initialization and instead we initialize metadata
at disksize setting time.
Acked-by: Jerome Marchand <jmarchand@redhat.com>
Acked-by: Nitin Gupta <ngupta@vflare.org>
Signed-off-by: Minchan Kim <minchan@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Now zram document syas "set disksize is optional"
but partly it's wrong. When you try to use zram firstly after
booting, you must set disksize, otherwise zram can't work because
zram gendisk's size is 0. But once you do it, you can use zram freely
after reset because reset doesn't reset to zero paradoxically.
So in this time, disksize setting is optional.:(
It's inconsitent for user behavior and not straightforward.
This patch forces always setting disksize firstly before using zram.
Yes. It changes current behavior so someone could complain when
he upgrades zram. Apparently it could be a problem if zram is mainline
but it still lives in staging so behavior could be changed for right
way to go. Let them excuse.
Acked-by: Jerome Marchand <jmarchand@redhat.com>
Acked-by: Nitin Gupta <ngupta@vflare.org>
Acked-by: Dan Magenheimer <dan.magenheimer@oracle.com>
Signed-off-by: Minchan Kim <minchan@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Now zram allocates new page with GFP_KERNEL in zram I/O path
if IO is partial. Unfortunately, It may cause deadlock with
reclaim path like below.
write_page from fs
fs_lock
allocation(GFP_KERNEL)
reclaim
pageout
write_page from fs
fs_lock <-- deadlock
This patch fixes it by using GFP_NOIO. In read path, we
reorganize code flow so that kmap_atomic is called after the
GFP_NOIO allocation.
Cc: stable@vger.kernel.org
Acked-by: Jerome Marchand <jmarchand@redhat.com>
Acked-by: Nitin Gupta <ngupta@vflare.org>
[ penberg@kernel.org: don't use GFP_ATOMIC ]
Signed-off-by: Pekka Enberg <penberg@kernel.org>
Signed-off-by: Minchan Kim <minchan@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
There are a number of pci vendor ids that are used in multiple
drivers in the comedi subsystem. Move these ids to pci_ids.h.
This also fixes some build warnings reported by the kbuild test
robot about PCI_VENDOR_ID_AMPLICON being undeclared.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Tuomas <tvainikk _at_ gmail _dot_ com> reported problems getting
meaningful output from a Lab-PC+ in differential mode for AI cmds, but
AI insn reads gave correct readings. He tracked it down to two
problems, one of which is addressed by this patch.
It seems the setting of the channel bits for particular scanning modes
was incorrect for differential mode. (Only half the number of channels
are available in differential mode; comedi refers to them as channels 0,
1, 2 and 3, but the hardware documentation refers to them as channels 0,
2, 4 and 6.) In differential mode, the setting of the channel enable
bits in the command1 register should depend on whether the scan enable
bit is set. Effectively, we need to double the comedi channel number
when the scan enable bit is not set in differential mode. The scan
enable bit gets set when the AI scan mode is `MODE_MULT_CHAN_UP` or
`MODE_MULT_CHAN_DOWN`, and gets cleared when the AI scan mode is
`MODE_SINGLE_CHAN` or `MODE_SINGLE_CHAN_INTERVAL`. The existing test
for whether the comedi channel number needs to be doubled in
differential mode is incorrect in `labpc_ai_cmd()`. This patch corrects
the test.
Thanks to Tuomas for suggesting the fix.
Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Cc: <stable@vger.kernel.org> # 3.7.x, 3.8.x
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
