Dean Anderson [Tue, 4 Feb 2014 20:18:03 +0000 (17:18 -0300)]
[media] s2255drv: removal of s2255_dmaqueue structure
Removal of unused and unnecessary s2255dma_queue structure.
Signed-off-by: Dean Anderson <linux-dev@sensoray.com> Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by: Mauro Carvalho Chehab <m.chehab@samsung.com>
Levente Kurusa [Sat, 15 Feb 2014 10:17:11 +0000 (07:17 -0300)]
[media] staging: davinci_vpfe: fix error check
The check would check the pointer, which is never less than 0.
According to the error message, the correct check would be
to check the return value of ipipe_mode. Check that instead.
Laurent Pinchart [Tue, 18 Feb 2014 13:40:47 +0000 (10:40 -0300)]
[media] uvcvideo: Support allocating buffers larger than the current frame size
The queue_setup handler takes an optional format argument that can be
used to allocate buffers for a format different than the current format.
The uvcvideo driver doesn't support changing the format when buffers
have been allocated, but there's no reason not to support allocating
buffers larger than the minimum size.
When the format argument isn't NULL verify that the requested image size
is large enough for the current format and use it for the buffer size.
Laurent Pinchart [Tue, 18 Feb 2014 13:02:50 +0000 (10:02 -0300)]
[media] uvcvideo: Remove duplicate check for number of buffers in queue_setup
videobuf2 already ensures that the number of buffers will not exceed
VIDEO_MAX_FRAME, which is equal to our arbitraty limit of
UVC_MAX_VIDEO_BUFFERS. Remove the duplicate check.
Thomas Pugliese [Fri, 24 Jan 2014 21:17:28 +0000 (18:17 -0300)]
[media] uvcvideo: Update uvc_endpoint_max_bpi to handle USB_SPEED_WIRELESS devices
Isochronous endpoints on devices with speed == USB_SPEED_WIRELESS can
have a max packet size ranging from 1-3584 bytes. Add a case to
uvc_endpoint_max_bpi to handle USB_SPEED_WIRELESS. Otherwise endpoints
for those devices will fall to the default case which masks off any
values > 2047. This causes uvc_init_video to underestimate the
bandwidth available and fail to find a suitable alt setting for high
bandwidth video streams.
[media] ARM: omap2: cm-t35: Add regulators and clock for camera sensor
The camera sensor will soon require regulators and clocks. Register
fixed regulators for its VAA and VDD power supplies and a fixed rate
clock for its master clock.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Acked-by: Igor Grinberg <grinberg@compulab.co.il> Acked-by: Tony Lindgren <tony@atomide.com> Signed-off-by: Mauro Carvalho Chehab <m.chehab@samsung.com>
Laurent Pinchart [Sat, 15 Feb 2014 00:40:48 +0000 (21:40 -0300)]
[media] omap3isp: Don't try to locate external subdev for mem-to-mem pipelines
Memory-to-memory pipelines have no external subdev, we shouldn't try to
locate one and validate its configuration. The driver currently works by
chance due to another bug that results in failure to locate the external
subdev being ignored.
This gets rid of the "omap3isp omap3isp: can't find source, failing now"
error message in the kernel log when operating on a memory-to-memory
pipeline.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Acked-by: Sakari Ailus <sakari.ailus@iki.fi> Tested-by: Peter Meerwald <pmeerw@pmeerw.net> Signed-off-by: Mauro Carvalho Chehab <m.chehab@samsung.com>
Florian Vaussard [Fri, 17 Jan 2014 19:37:38 +0000 (16:37 -0300)]
[media] omap3isp: preview: Fix the crop margins
Commit 3fdfedaaa "[media] omap3isp: preview: Lower the crop margins"
accidentally changed the previewer's cropping, causing the previewer
to miss four pixels on each line, thus corrupting the final image.
Restored the removed setting.
Hans Verkuil [Mon, 10 Feb 2014 11:08:47 +0000 (08:08 -0300)]
[media] vivi: queue_setup improvements
Drop the vid_limit module option: there is really no need to limit this.
No other driver does that. If you try to allocate more buffers then vb2
will automatically reduce the number of buffers anyway.
Also add sanity checks if the size in the fmt argument is going to be
used and drop the code that checks against *nbuffers == 0: this can
never happen (the vb2 framework ensures that) and the code was wrong
anyway since *nbuffers should have been set to the minimum number of
required buffers which is 1 for this driver.
Since vivi is often used as a template driver it is good to have this
driver be as compliant as possible. This broken code was for example
copied to the s2255 driver (which is being fixed as well).
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by: Mauro Carvalho Chehab <m.chehab@samsung.com>
Hans Verkuil [Mon, 10 Feb 2014 11:08:45 +0000 (08:08 -0300)]
[media] vivi: fix sequence counting
The sequence counting was not reset to 0 between each streaming run,
and it was increased only every other frame. This is incorrect behavior:
the confusion is with FIELD_ALTERNATE systems where each field is transmitted
separately and only when both fields have been received is the frame
sequence number increased.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by: Mauro Carvalho Chehab <m.chehab@samsung.com>
Hans Verkuil [Mon, 10 Feb 2014 11:08:44 +0000 (08:08 -0300)]
[media] v4l2-subdev: Allow 32-bit compat ioctls
Add support for 32-bit ioctls with v4l-subdev device nodes.
Rather than keep adding new ioctls to the list in v4l2-compat-ioctl32.c, just check
if the ioctl is a non-private V4L2 ioctl and if so, call the conversion code.
We keep forgetting to add new ioctls, so this is a more robust solution.
In addition extend the subdev API with support for a compat32 function to
convert custom v4l-subdev ioctls.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com> Tested-by: Sakari Ailus <sakari.ailus@linux.intel.com> Signed-off-by: Mauro Carvalho Chehab <m.chehab@samsung.com>
Antti Palosaari [Tue, 11 Feb 2014 03:17:37 +0000 (00:17 -0300)]
[media] tda10071: do not check tuner PLL lock on read_status()
Tuner PLL lock flag was mapped to FE_HAS_SIGNAL, which is wrong. PLL
lock has nothing to do with received signal. In real life that flag
is always set.
Signed-off-by: Antti Palosaari <crope@iki.fi> Signed-off-by: Mauro Carvalho Chehab <m.chehab@samsung.com>
Till Dörges [Tue, 11 Feb 2014 20:27:57 +0000 (17:27 -0300)]
[media] rtl28xxu: add ID [0ccd:00b4] TerraTec NOXON DAB Stick (rev 3)
I've got the following DAB USB stick that also works fine with the
DVB_USB_RTL28XXU driver after I added its USB ID:
Bus 001 Device 009: ID 0ccd:00b4 TerraTec Electronic GmbH
[crope@iki.fi: apply patch partly manually] Signed-off-by: Till Dörges <till@doerges.net> Signed-off-by: Antti Palosaari <crope@iki.fi> Signed-off-by: Mauro Carvalho Chehab <m.chehab@samsung.com>
Add missing m88ts2022 module reference counts as removing that module
is not allowed when it is used by em28xx-dvb module. That same module
was not unregistered correctly, fix it too.
Error cases validated by returning errors from m88ds3103, m88ts2022
and a8293 probe().
Signed-off-by: Antti Palosaari <crope@iki.fi> Signed-off-by: Mauro Carvalho Chehab <m.chehab@samsung.com>
There are several left overs with my old email address.
Remove their occurrences and add myself at CREDITS, to
allow people to be able to reach me on my new addresses.
[media] DocBook: Add a description for the Remote Controller interface
Adds a missing section to describe the remote controller interface.
The DocBook is just addin the same documentation as written at
Documentation/ABI/testing/sysfs-class-rc, using the DocBook's
way, and dropping timestamps/contact info.
While that means that we'll have the same info on two parts, there
are parts of the remote controller interface that doesn't belong
at Documentation/ABI/, and it makes sense to have everything
on the same place. This also means that we'll need to manually
track to be sure that both places will be synchronized, but, as
it is not expected much changes on it, this sync can be done
manually.
It also adds an introduction that states that the IR is a normal
evdev/input interface, plus the sysfs class nodes.
James Hogan [Fri, 17 Jan 2014 13:58:50 +0000 (10:58 -0300)]
[media] media: rc: change 32bit NEC scancode format
Change 32bit NEC scancode format (used by Apple and TiVo remotes) to
encode the data with the correct bit order. Previously the raw bits were
used without being bit reversed, now each 16bit half is bit reversed
compared to before.
So for the raw NEC data:
(LSB/First) 0xAAaaCCcc (MSB/Last)
(where traditionally AA=address, aa=~address, CC=command, cc=~command)
We now generate the scancodes:
(MSB) 0x0000AACC (LSB) (normal NEC)
(MSB) 0x00AAaaCC (LSB) (extended NEC, address check wrong)
(MSB) 0xaaAAccCC (LSB) (32-bit NEC, command check wrong)
Note that the address byte order in 32-bit NEC scancodes is different to
that of the extended NEC scancodes. I chose this way as it maintains the
order of the bits in the address/command fields, and CC is clearly
intended to be the LSB of the command if the TiVo codes are anything to
go by so it makes sense for AA to also be the LSB.
The TiVo keymap is updated accordingly.
Signed-off-by: James Hogan <james.hogan@imgtec.com> Cc: Mauro Carvalho Chehab <m.chehab@samsung.com> Cc: Jarod Wilson <jarod@redhat.com> Cc: linux-media@vger.kernel.org Signed-off-by: Mauro Carvalho Chehab <m.chehab@samsung.com>
Add and document a generic sysfs based scancode filtering interface for
making use of IR data matching hardware to filter out uninteresting
scancodes. Two filters exist, one for normal operation and one for
filtering scancodes which are permitted to wake the system from suspend.
The following files are added to /sys/class/rc/rc?/:
- filter: normal scancode filter value
- filter_mask: normal scancode filter mask
- wakeup_filter: wakeup scancode filter value
- wakeup_filter_mask: wakeup scancode filter mask
A new s_filter() driver callback is added which must arrange for the
specified filter to be applied at the right time. Drivers can convert
the scancode filter into a raw IR data filter, which can be applied
immediately or later (for wake up filters).
James Hogan [Wed, 5 Feb 2014 22:15:16 +0000 (19:15 -0300)]
[media] rc: ir-raw: Load ir-sharp-decoder module at init
Commit 1d184b0bc13d ([media] media: rc: add raw decoder for Sharp
protocol) added a new raw IR decoder for the sharp protocol, but didn't
add the code to load the module at init as is done for other raw
decoders, so add that code now.
Signed-off-by: James Hogan <james.hogan@imgtec.com> Signed-off-by: Mauro Carvalho Chehab <m.chehab@samsung.com>
Antti Seppälä [Sat, 25 Jan 2014 09:57:46 +0000 (06:57 -0300)]
[media] nuvoton-cir: Don't touch PS/2 interrupts while initializing
There are reports[1] that on some motherboards loading the nuvoton-cir
disables PS/2 keyboard input. This is caused by an erroneous write of
CIR_INTR_MOUSE_IRQ_BIT to ACPI control register.
According to datasheet the write enables mouse power management event
interrupts which will probably have ill effects if the motherboard has
only one PS/2 port with keyboard in it.
The cir hardware does not need mouse interrupts to function and should
not touch them. This patch removes the illegal writes and registry
definitions.
Joakim Hernberg [Fri, 31 Jan 2014 10:15:48 +0000 (07:15 -0300)]
[media] cx23885: Fix tuning regression for TeVii S471
When tuning to 10818V on Astra 28E2, the system tunes to 11343V
instead. This is a regression in the S471 driver introduced with the
changeset: b43ea8068d2090cb1e44632c8a938ab40d2c7419 [media] cx23885:
Fix TeVii S471 regression since introduction of ts2020.
Sean Young [Mon, 20 Jan 2014 22:10:44 +0000 (19:10 -0300)]
[media] mceusb: improve error logging
A number of recent bug reports involve usb_submit_urb() failing which was
only reported with debug parameter on. In addition, remove custom debug
function.
[m.chehab@samsung.com: patch rebased, as one of the patches on this
series need changes] Signed-off-by: Sean Young <sean@mess.org> Signed-off-by: Mauro Carvalho Chehab <m.chehab@samsung.com>
Frank Schaefer [Sun, 19 Jan 2014 21:48:35 +0000 (18:48 -0300)]
[media] em28xx-i2c: fix the error code for unknown errors
Commit e63b009d6e "em28xx-i2c: Fix error code for I2C error transfers"
changed the code to return -ETIMEDOUT on all unknown errors.
But the proper error code for unknown errors is -EIO.
So only report -ETIMEDOUT in case of the errors 0x02 and 0x04, which
are according to Mauro Carvalho Chehab's tests related to i2c clock
stretching and return -EIO for the rest.
Signed-off-by: Frank Schäfer <fschaefer.oss@googlemail.com> Signed-off-by: Mauro Carvalho Chehab <m.chehab@samsung.com>
Frank Schaefer [Sun, 19 Jan 2014 21:48:34 +0000 (18:48 -0300)]
[media] em28xx-i2c: fix the i2c error description strings for -ENXIO
Commit d845fb3ae5 "em28xx-i2c: add timeout debug information if i2c_debug enabled"
has added wrong error descriptions for -ENXIO.
The strings are also missing terminating newline characters, which breaks the
output format.
Signed-off-by: Frank Schäfer <fschaefer.oss@googlemail.com> Signed-off-by: Mauro Carvalho Chehab <m.chehab@samsung.com>
Frank Schaefer [Fri, 17 Jan 2014 17:45:31 +0000 (14:45 -0300)]
[media] em28xx-camera: fix return value checks on sensor probing
Since commit e63b009d6e the returned error code in case of not connected/responding
i2c clients is ENXIO isntead of ENODEV, which causes several error messages on
sensor probing.
Fix the i2c return value checks on sensor probing to silence these warnings.
Signed-off-by: Frank Schäfer <fschaefer.oss@googlemail.com> Signed-off-by: Mauro Carvalho Chehab <m.chehab@samsung.com>
James Hogan [Fri, 17 Jan 2014 13:58:48 +0000 (10:58 -0300)]
[media] media: rc: add raw decoder for Sharp protocol
Add a raw decoder for the Sharp protocol. It uses a pulse distance
modulation with a pulse of 320us and a bit period of 2ms for a logical 1
and 1ms for a logical 0. The first part of the message consists of a
5-bit address, an 8-bit command, and two other bits, followed by a 40ms
gap before the echo message which is an inverted version of the main
message except for the address bits.
James Hogan [Thu, 16 Jan 2014 22:56:22 +0000 (19:56 -0300)]
[media] media: rc: only turn on LED if keypress generated
Since v3.12, specifically 153a60bb0fac ([media] rc: add feedback led
trigger for rc keypresses), an LED trigger is activated on IR keydown
whether or not a keypress is generated (i.e. even if there's no matching
keycode). However the repeat and keyup logic isn't used unless there is
a keypress, which results in non-keypress keydown events turning on the
LED and not turning it off again.
On the assumption that the intent was for the LED only to light up on
valid key presses (you probably don't want it lighting up for the wrong
remote control for example), move the led_trigger_event() call inside
the keycode check.
Signed-off-by: James Hogan <james.hogan@imgtec.com> Acked-by: Sean Young <sean@mess.org> Signed-off-by: Mauro Carvalho Chehab <m.chehab@samsung.com>
[media] FE_READ_SNR and FE_READ_SIGNAL_STRENGTH docs
Around 01/14/2014 06:07 PM, Mauro Carvalho Chehab scribbled:
> Em Tue, 14 Jan 2014 17:55:19 +0200
> Georgi Chorbadzhiyski <gf@unixsol.org> escreveu:
>> Around 01/14/2014 05:30 PM, Mauro Carvalho Chehab scribbled:
>>> Em Tue, 14 Jan 2014 17:16:10 +0200
>>> Georgi Chorbadzhiyski <gf@unixsol.org> escreveu:
>>>
>>>> Hi guys, I'm confused the documentation on:
>>>>
>>>> http://linuxtv.org/downloads/v4l-dvb-apis/frontend_fcalls.html#FE_READ_SNR
>>>> http://linuxtv.org/downloads/v4l-dvb-apis/frontend_fcalls.html#FE_READ_SIGNAL_STRENGTH
>>>>
>>>> states that these ioctls return int16_t values but frontend.h states:
>>>>
>>>> https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/tree/include/uapi/linux/dvb/frontend.h
>>>>
>>>> #define FE_READ_SIGNAL_STRENGTH _IOR('o', 71, __u16)
>>>> #define FE_READ_SNR _IOR('o', 72, __u16)
>>>>
>>>> So which one is true?
>>>
>>> Documentation is wrong. The returned values are unsigned. Would you mind send
>>> us a patch fixing it?
>>
>> I would be happy to, but I can't find the repo that holds the documentation.
>
> It is in the Kernel tree, under Documentation/DocBook/media/dvb.
The attached file contains the discussed documentation fixes.
Hans Verkuil [Tue, 7 Jan 2014 13:01:27 +0000 (10:01 -0300)]
[media] DocBook media: drop the old incorrect packed RGB table
The old table is most definitely wrong. All applications and all
drivers that I have ever tested follow the corrected table. Furthermore,
that's what all applications expect as well. Any drivers that do not
follow the corrected table are broken and should be fixed.
This patch drops the old table and replaces it with the corrected
table. This should prevent a lot of confusion.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by: Mauro Carvalho Chehab <m.chehab@samsung.com>
Hans Verkuil [Tue, 7 Jan 2014 12:39:57 +0000 (09:39 -0300)]
[media] DocBook media: update three sections of common.xml
Updates for the "Tuners and Modulators", "Video Standards" and
"Digital Video (DV) Timings" sections.
Besides lots of trivial little fixes the main changes are:
- Remove two footnotes from "Video Standards": the first is a discussion
of alternative methods of setting standards, which is pretty pointless
since the standards API is effectively frozen anyway, and the second
points to 'rationale' that makes little or no sense to me.
- Clarify a few things in the "Digital Video (DV) Timings" section.
It was awkwardly formatted as well: there used to be a list with
multiple bullets that has been reduced to a single item, so drop the
list and rewrite that text.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by: Mauro Carvalho Chehab <m.chehab@samsung.com>
Hans Verkuil [Tue, 7 Jan 2014 09:46:37 +0000 (06:46 -0300)]
[media] DocBook media: update copyright years and Introduction
It's now 2014, so update those copyright years. Also fix a typo in the introduction
and mention that this document also covers output, codec and remote control
devices.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by: Mauro Carvalho Chehab <m.chehab@samsung.com>
The VESA standard added a version 2 of the reduced blanking formula.
Note in the comment that this is not yet supported by the v4l2_detect_cvt
function. Obviously this should be implemented eventually, but for now add
this as a reminder.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by: Mauro Carvalho Chehab <m.chehab@samsung.com>
Fixes all style warnings from scripts/checkpatch -f
Signed-off-by: Dean Anderson <linux-dev@sensoray.com> Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by: Mauro Carvalho Chehab <m.chehab@samsung.com>
Martin Bugge [Fri, 24 Jan 2014 13:50:06 +0000 (10:50 -0300)]
[media] adv7842: platform-data for Hotplug Active (HPA) manual/auto
This applies to HDMI-map register 0x69.
So far we have been using HPA manual mode.
This way we had control of HPA which could be
set after EDID had been programmed.
Using a Mac Mini with mini-displayport to DVI-D converter as source
caused the adv7842 to lock up and fail to detect any further signals.
After experimenting with different configurations it was found that
using the HPA auto mode and in addition letting RX-termination
be controlled by HPA prevented this error from occuring.
I was not able to re-create this problem on the adv7604.
Signed-off-by: Martin Bugge <marbugge@cisco.com> Cc: Mats Randgaard <matrandg@cisco.com> Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by: Mauro Carvalho Chehab <m.chehab@samsung.com>
Martin Bugge [Fri, 24 Jan 2014 13:50:04 +0000 (10:50 -0300)]
[media] adv7842: pixelclock read-out
Incorrect registers used for pixelclock read-out.
Same registers as for adv7604 which actually gave an almost
correct read-out, even they are not documented for adv7842.
Corrected deep-color pixel-clock correction.
Signed-off-by: Martin Bugge <marbugge@cisco.com> Cc: Mats Randgaard <matrandg@cisco.com> Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by: Mauro Carvalho Chehab <m.chehab@samsung.com>
Martin Bugge [Fri, 24 Jan 2014 13:50:03 +0000 (10:50 -0300)]
[media] adv7842: adjust gain and offset for DVI-D signals
If the input signal is DVI-D and quantization range is RGB full range,
gain and offset must be adjusted to get the right range on the output.
Copied and adopted from adv7604.
Signed-off-by: Martin Bugge <marbugge@cisco.com> Cc: Mats Randgaard <matrandg@cisco.com> Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by: Mauro Carvalho Chehab <m.chehab@samsung.com>
* tag 'v3.14-rc1': (11781 commits)
Linus 3.14-rc1
hpfs: optimize quad buffer loading
hpfs: remember free space
parisc: add flexible mmap memory layout support
parisc: Make EWOULDBLOCK be equal to EAGAIN on parisc
parisc: convert uapi/asm/stat.h to use native types only
parisc: wire up sched_setattr and sched_getattr
parisc: fix cache-flushing
parisc/sti_console: prefer Linux fonts over built-in ROM fonts
hwmon: Fix SENSORS_TMP102 dependencies to eliminate build errors
hwmon: Fix SENSORS_LM75 dependencies to eliminate build errors
tools/power turbostat: introduce -s to dump counters
tools/power turbostat: remove unused command line option
afs: proc cells and rootcell are writeable
tile: remove compat_sys_lookup_dcookie declaration to fix compile error
Revert "PCI: Remove from bus_list and release resources in pci_release_dev()"
ARM: multi_v7_defconfig: remove redundant entries and re-enable TI_EDMA
ARM: multi_v7_defconfig: add mvebu drivers
clocksource: kona: Add basic use of external clock
drivers: bus: fix CCI driver kcalloc call parameters swap
...
Linus Torvalds [Mon, 3 Feb 2014 00:32:53 +0000 (16:32 -0800)]
Merge branch 'parisc-3.14' of git://git.kernel.org/pub/scm/linux/kernel/git/deller/parisc-linux
Pull parisc updates from Helge Deller:
"The three major changes in this patchset is a implementation for
flexible userspace memory maps, cache-flushing fixes (again), and a
long-discussed ABI change to make EWOULDBLOCK the same value as
EAGAIN.
parisc has been the only platform where we had EWOULDBLOCK != EAGAIN
to keep HP-UX compatibility. Since we will probably never implement
full HP-UX support, we prefer to drop this compatibility to make it
easier for us with Linux userspace programs which mostly never checked
for both values. We don't expect major fall-outs because of this
change, and if we face some, we will simply rebuild the necessary
applications in the debian archives"
* 'parisc-3.14' of git://git.kernel.org/pub/scm/linux/kernel/git/deller/parisc-linux:
parisc: add flexible mmap memory layout support
parisc: Make EWOULDBLOCK be equal to EAGAIN on parisc
parisc: convert uapi/asm/stat.h to use native types only
parisc: wire up sched_setattr and sched_getattr
parisc: fix cache-flushing
parisc/sti_console: prefer Linux fonts over built-in ROM fonts
Mikulas Patocka [Tue, 28 Jan 2014 23:11:33 +0000 (00:11 +0100)]
hpfs: optimize quad buffer loading
HPFS needs to load 4 consecutive 512-byte sectors when accessing the
directory nodes or bitmaps. We can't switch to 2048-byte block size
because files are allocated in the units of 512-byte sectors.
Previously, the driver would allocate a 2048-byte area using kmalloc,
copy the data from four buffers to this area and eventually copy them
back if they were modified.
In the current implementation of the buffer cache, buffers are allocated
in the pagecache. That means that 4 consecutive 512-byte buffers are
stored in consecutive areas in the kernel address space. So, we don't
need to allocate extra memory and copy the content of the buffers there.
This patch optimizes the code to avoid copying the buffers. It checks
if the four buffers are stored in contiguous memory - if they are not,
it falls back to allocating a 2048-byte area and copying data there.
Mikulas Patocka [Tue, 28 Jan 2014 23:10:44 +0000 (00:10 +0100)]
hpfs: remember free space
Previously, hpfs scanned all bitmaps each time the user asked for free
space using statfs. This patch changes it so that hpfs scans the
bitmaps only once, remembes the free space and on next invocation of
statfs it returns the value instantly.
New versions of wine are hammering on the statfs syscall very heavily,
making some games unplayable when they're stored on hpfs, with load
times in minutes.
This should be backported to the stable kernels because it fixes
user-visible problem (excessive level load times in wine).
Helge Deller [Fri, 31 Jan 2014 21:19:52 +0000 (22:19 +0100)]
parisc: add flexible mmap memory layout support
Add support for the flexible mmap memory layout (as described in
http://lwn.net/Articles/91829). This is especially very interesting on
parisc since we currently only support 32bit userspace (even with a
64bit Linux kernel).
Guy Martin [Thu, 16 Jan 2014 16:17:53 +0000 (17:17 +0100)]
parisc: Make EWOULDBLOCK be equal to EAGAIN on parisc
On Linux, only parisc uses a different value for EWOULDBLOCK which
causes a lot of troubles for applications not checking for both values.
Since the hpux compat is long dead, make EWOULDBLOCK behave the same as
all other architectures.
Signed-off-by: Guy Martin <gmsoft@tuxicoman.be> Signed-off-by: Helge Deller <deller@gmx.de>
Helge Deller [Fri, 31 Jan 2014 22:00:38 +0000 (23:00 +0100)]
parisc: convert uapi/asm/stat.h to use native types only
The stat.h header file is exported to userspace. Some userspace
applications failed to compile due to missing/unknown types, so we
better convert it to use native types only (like it's done on other
architectures too).
Helge Deller [Fri, 31 Jan 2014 20:33:17 +0000 (21:33 +0100)]
parisc: fix cache-flushing
This commit: f8dae00684d678afa13041ef170cecfd1297ed40: parisc: Ensure full cache coherency for kmap/kunmap
caused negative caching side-effects, e.g. hanging processes with expect and
too many inequivalent alias messages from flush_dcache_page() on Debian 5 systems.
This patch now partly reverts it and has been in production use on our debian buildd
makeservers since a week without any major problems.
Signed-off-by: Helge Deller <deller@gmx.de> Signed-off-by: John David Anglin <dave.anglin@bell.net> Cc: stable@vger.kernel.org # v3.9+ Signed-off-by: Helge Deller <deller@gmx.de>
Helge Deller [Fri, 31 Jan 2014 14:39:40 +0000 (15:39 +0100)]
parisc/sti_console: prefer Linux fonts over built-in ROM fonts
The built-in ROM fonts lack many necessary ASCII characters, which is
why it makes sens to prefer the Linux fonts instead if they are
available. This makes consoles on STI graphics cards which are not
supported by the stifb driver (e.g. Visualize FXe) looks much nicer.
Linus Torvalds [Sun, 2 Feb 2014 19:30:08 +0000 (11:30 -0800)]
Merge branch 'slab/next' of git://git.kernel.org/pub/scm/linux/kernel/git/penberg/linux
Pull SLAB changes from Pekka Enberg:
"Random bug fixes that have accumulated in my inbox over the past few
months"
* 'slab/next' of git://git.kernel.org/pub/scm/linux/kernel/git/penberg/linux:
mm: Fix warning on make htmldocs caused by slab.c
mm: slub: work around unneeded lockdep warning
mm: sl[uo]b: fix misleading comments
slub: Fix possible format string bug.
slub: use lockdep_assert_held
slub: Fix calculation of cpu slabs
slab.h: remove duplicate kmalloc declaration and fix kernel-doc warnings
Linus Torvalds [Sun, 2 Feb 2014 19:28:48 +0000 (11:28 -0800)]
Merge branch 'release' of git://git.kernel.org/pub/scm/linux/kernel/git/lenb/linux
Pull turbostat updates from Len Brown.
* 'release' of git://git.kernel.org/pub/scm/linux/kernel/git/lenb/linux:
tools/power turbostat: introduce -s to dump counters
tools/power turbostat: remove unused command line option
turbostat: Add option to report joules consumed per sample
turbostat: run on HSX
turbostat: Add a .gitignore to ignore the compiled turbostat binary
turbostat: Clean up error handling; disambiguate error messages; use err and errx
turbostat: Factor out common function to open file and exit on failure
turbostat: Add a helper to parse a single int out of a file
turbostat: Check return value of fscanf
turbostat: Use GCC's CPUID functions to support PIC
turbostat: Don't attempt to printf an off_t with %zx
turbostat: Don't put unprocessed uapi headers in the include path
Linus Torvalds [Sun, 2 Feb 2014 19:11:06 +0000 (11:11 -0800)]
Merge tag 'fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/arm/arm-soc
Pull ARM SoC fixes from Olof Johansson:
"Here's a set of patches for (hopefully) -rc1. Some of them are fixes,
but a good number of them also do things such as enable new drivers in
the defconfigs for platforms that have such devices, increases
coverage of the multiplatform defconfig and some DTS changes that
plumbs up some of the devices that now have bindings and driver
support.
The commit dates are recent; we've mostly collected these fixes in the
last few days but I also had to rebuild the branch yesterday to sort
out some internal conflicts which reset the timestamps. The changes
should have been tested by each platform maintainer already (and few
of them have cross-platform impact) so I'm personally not too
concerned by it at this time"
* tag 'fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/arm/arm-soc: (23 commits)
ARM: multi_v7_defconfig: remove redundant entries and re-enable TI_EDMA
ARM: multi_v7_defconfig: add mvebu drivers
clocksource: kona: Add basic use of external clock
drivers: bus: fix CCI driver kcalloc call parameters swap
ARM: dts: bcm28155-ap: Fix Card Detection GPIO
ARM: multi_v7_defconfig: Select CONFIG_AT803X_PHY
ARM: keystone: config: fix build warning when CONFIG_DMADEVICES is not set
MAINTAINERS: ARM: SiRF: use regex patterns to involve all SiRF drivers
ARM: dts: zynq: Add SDHCI nodes
ARM: hisi: don't select SMP
ARM: tegra: rebuild tegra_defconfig to add DEBUG_FS
ARM: multi_v7: copy most options from tegra_defconfig
ARM: iop32x: fix power off handling for the EM7210 board
ARM: integrator: restore static map on the CP
ARM: msm_defconfig: Enable MSM clock drivers
ARM: dts: msm: Add clock controller nodes and hook into uart
ARM: OMAP4+: move errata initialization to omap4_pm_init_early
ARM: OMAP4460: cpuidle: Extend PM_OMAP4_ROM_SMP_BOOT_ERRATUM_GICD on cpuidle
ARM: mvebu: fix compilation warning on Armada 370 (i.e. non-SMP)
ARM: shmobile: r8a7790.dtsi: ficx i2c[0-3] clock reference
...
Jean Delvare [Sun, 2 Feb 2014 16:59:07 +0000 (17:59 +0100)]
hwmon: Fix SENSORS_TMP102 dependencies to eliminate build errors
Similar to what was done for the lm75 driver.
Add depends on THERMAL since that is what provides the
register/unregister functions above, but only if THERMAL_OF was
selected as this is an optional feature of the driver.
Signed-off-by: Jean Delvare <khali@linux-fr.org> Cc: Randy Dunlap <rdunlap@infradead.org> Acked-by: Eduardo Valentin <eduardo.valentin@ti.com> Reviewed-by: Guenter Roeck <linux@roeck-us.net>
Jean Delvare [Sun, 2 Feb 2014 16:59:07 +0000 (17:59 +0100)]
hwmon: Fix SENSORS_LM75 dependencies to eliminate build errors
Based on an earlier attempt by Randy Dunlap.
Fix SENSORS_LM75 dependencies to eliminate build errors:
drivers/built-in.o: In function `lm75_remove':
lm75.c:(.text+0x12bd8c): undefined reference to `thermal_zone_of_sensor_unregister'
drivers/built-in.o: In function `lm75_probe':
lm75.c:(.text+0x12c123): undefined reference to `thermal_zone_of_sensor_register'
Add depends on THERMAL since that is what provides the
register/unregister functions above, but only if THERMAL_OF was
selected as this is an optional feature of the driver.
Signed-off-by: Jean Delvare <khali@linux-fr.org> Cc: Randy Dunlap <rdunlap@infradead.org> Acked-by: Eduardo Valentin <eduardo.valentin@ti.com> Reviewed-by: Guenter Roeck <linux@roeck-us.net>
Linus Torvalds [Sat, 1 Feb 2014 19:03:16 +0000 (11:03 -0800)]
Merge branch 'misc' of git://git.kernel.org/pub/scm/linux/kernel/git/mmarek/kbuild
Pull misc kbuild changes from Michal Marek:
"The non-critical part of kbuild is small this time:
- Three fixes for make deb-pkg
- A new coccinelle check
One of the deb-pkg fixes is a leftover from the last merge window,
hence the merge commit"
* 'misc' of git://git.kernel.org/pub/scm/linux/kernel/git/mmarek/kbuild:
deb-pkg: Fix building for MIPS big-endian or ARM OABI
deb-pkg: Fix cross-building linux-headers package
scripts: Coccinelle script for pm_runtime_* return checks with IS_ERR_VALUE
deb-pkg: Inhibit initramfs builders if CONFIG_BLK_DEV_INITRD is not set
projects / karo-tx-linux.git / log