Hans Verkuil [Thu, 7 May 2015 06:26:21 +0000 (03:26 -0300)]
[media] rcar-vin: use monotonic timestamps
Even though the rcar-vin driver tells userspace that it will give a monotonic
timestamp, it is actually using gettimeofday. Replace this with a proper
monotonic timestamp.
Prashant Laddha [Tue, 5 May 2015 13:16:27 +0000 (10:16 -0300)]
[media] v4l2-dv-timings: fix overflow in gtf timings calculation
The intermediate calculation in the expression for hblank can exceed
32 bit signed range. This overflow can lead to negative values for
hblank. Typecasting intermediate variable to higher precision.
Cc: Martin Bugge <marbugge@cisco.com> Signed-off-by: Prashant Laddha <prladdha@cisco.com>
[hans.verkuil@cisco.com: made the denominator u32, since that's what div_u64 expects] Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>
Antti Palosaari [Mon, 4 May 2015 00:42:02 +0000 (21:42 -0300)]
[media] fc2580: implement V4L2 subdevice for SDR control
Implement V4L2 subdevice for bandwidth and frequency controls of
SDR usage. That driver now implements both DVB frontend and V4L2
subdevice. Driver itself is I2C driver. Lets see how it works.
Signed-off-by: Antti Palosaari <crope@iki.fi> Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>
Antti Palosaari [Wed, 15 Apr 2015 18:01:22 +0000 (15:01 -0300)]
[media] af9035: fix device order in ID list
Driver supports multiple chipset versions. Devices are ordered to
ID table per used chipset type. "ITE 9303 Generic" device uses IT9303
chipset and was added mistakenly between IT9135 IDs.
Signed-off-by: Antti Palosaari <crope@iki.fi> Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>
Antti Palosaari [Wed, 15 Apr 2015 15:10:54 +0000 (12:10 -0300)]
[media] fc2580: improve set params logic
Calculate PLL dividers slightly differently, most likely it is now
correct. Move some register values to innitab. Use jiffies to poll
filter lock. Fix logging.
Signed-off-by: Antti Palosaari <crope@iki.fi> Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>
[media] dib0700: avoid the risk of forgetting to add the adapter's size
For every frontend entry, we need to add the adapter's size. There
are already two patches fixing it. So, it doesn't seem trivial to
keep it there at the right place.
Also, currently, the indentation is wrong on all places.
So, it seems that keeping it right is not too trivial.
Better to use a macro that would do it for us, at least while
this is not converted to dvb-usb-v2.
[media] saa7164: use an MSI interrupt when available
Enhances driver to use an MSI interrupt when available.
Adds the module option 'enable_msi' (type bool) which by default is
enabled. Can be set to 'N' to disable.
Fixes (or can reduce the occurrence of) a crash which is most commonly
reported when both digital tuners of the saa7164 chip is in use. A
reported example can be found here:
http://permalink.gmane.org/gmane.linux.drivers.video-input-infrastructure/83948
David Howells [Thu, 9 Apr 2015 19:36:49 +0000 (16:36 -0300)]
[media] dvb: Document FE_SCALE_DECIBEL units consistently
In comments and in the documentation, the units of properties marked
with the FE_SCALE_DECIBEL scale are specified in terms of 1/1000 dB
or 0.0001 dB. This is inconsistent, however, as 1/1000 is 0.001,
not 0.0001.
Note that the v4l-utils divide the value by 1000 for the signal
strength suggesting that the 1/1000 is correct.
Settle on millidecibels, ie. 1/1000dB or 0.001dB.
Signed-off-by: David Howells <dhowells@redhat.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>
Antti Seppälä [Tue, 31 Mar 2015 17:48:12 +0000 (14:48 -0300)]
[media] rc: nuvoton-cir: Add support for writing wakeup samples via sysfs filter callback
Nuvoton-cir utilizes the encoding capabilities of rc-core to convert
scancodes from user space to pulse/space format understood by the
underlying hardware.
Converted samples are then written to the wakeup fifo along with other
necessary configuration to enable wake up functionality.
Signed-off-by: Antti Seppälä <a.seppala@gmail.com> Signed-off-by: James Hogan <james@albanarts.com> Cc: Jarod Wilson <jarod@redhat.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>
James Hogan [Tue, 31 Mar 2015 17:48:11 +0000 (14:48 -0300)]
[media] rc: rc-loopback: Add loopback of filter scancodes
Add the s_wakeup_filter callback to the rc-loopback driver, which instead of
setting the filter just feeds the scancode back through the input device
so that it can be verified.
Signed-off-by: James Hogan <james@albanarts.com> Signed-off-by: Antti Seppälä <a.seppala@gmail.com> Cc: David Härdeman <david@hardeman.nu> Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>
James Hogan [Tue, 31 Mar 2015 17:48:10 +0000 (14:48 -0300)]
[media] rc: rc-core: Add support for encode_wakeup drivers
Add support in rc-core for drivers which implement the wakeup scancode
filter by encoding the scancode using the raw IR encoders. This is by
way of rc_dev::encode_wakeup which should be set to true to make the
allowed wakeup protocols the same as the set of raw IR encoders.
As well as updating the sysfs interface to know which wakeup protocols
are allowed for encode_wakeup drivers, also ensure that the IR
decoders/encoders are loaded when an encode_wakeup driver is registered.
Signed-off-by: James Hogan <james@albanarts.com> Signed-off-by: Antti Seppälä <a.seppala@gmail.com> Cc: David Härdeman <david@hardeman.nu> Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>
Antti Seppälä [Tue, 31 Mar 2015 17:48:09 +0000 (14:48 -0300)]
[media] rc: ir-rc6-decoder: Add encode capability
Add the capability to encode RC-6 and RC-6A scancodes as raw events.
The protocol is chosen based on the specified protocol mask, and
whether all the required bits are set in the scancode mask, and none of
the unused bits are set in the scancode data.
The Manchester modulation helper is used several times with various
timings so that RC-6 header preamble, the header, header trailing bit
and the data itself can be modulated correctly.
Signed-off-by: Antti Seppälä <a.seppala@gmail.com> Cc: James Hogan <james@albanarts.com> Cc: David Härdeman <david@hardeman.nu> Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>
James Hogan [Tue, 31 Mar 2015 17:48:08 +0000 (14:48 -0300)]
[media] rc: ir-rc5-decoder: Add encode capability
Add the capability to encode RC-5, RC-5X and RC-5-SZ scancodes as raw
events. The protocol is chosen based on the specified protocol mask,
and whether all the required bits are set in the scancode mask, and
none of the unused bits are set in the scancode data. For example a
scancode filter with bit 16 set in both data and mask is unambiguously
RC-5X.
The Manchester modulation helper is used, and for RC-5X it is used twice
with two sets of timings, the first with a short trailer space for the
space in the middle, and the second with no leader so that it can
continue the space.
The encoding in RC-5-SZ first inserts a pulse and then simply utilizes
the generic Manchester encoder available in rc-core.
Signed-off-by: James Hogan <james@albanarts.com> Signed-off-by: Antti Seppälä <a.seppala@gmail.com> Cc: David Härdeman <david@hardeman.nu> Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>
Add a callback to raw ir handlers for encoding and modulating a scancode
to a set of raw events. This could be used for transmit, or for
converting a wakeup scancode filter to a form that is more suitable for
raw hardware wake up filters.
Signed-off-by: James Hogan <james@albanarts.com> Signed-off-by: Antti Seppälä <a.seppala@gmail.com> Cc: David Härdeman <david@hardeman.nu> Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>
Antti Palosaari [Wed, 29 Apr 2015 16:37:34 +0000 (13:37 -0300)]
[media] msi2500: cleanups
Rename state to dev.
Correct some indentations.
Remove FSF address.
Fix some style issues reported by checkpatch.pl.
Correct some style issues I liked.
Signed-off-by: Antti Palosaari <crope@iki.fi> Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>
Antti Palosaari [Wed, 29 Apr 2015 16:13:31 +0000 (13:13 -0300)]
[media] msi2500: revise synthesizer calculation
Update synthesizer calculation to model I prefer nowadays. It is mostly
just renaming some variables, but also minor functionality change how
integer and fractional part are divided (using div_u64_rem()). Also, add
'schematic' of synthesizer following my current understanding.
Signed-off-by: Antti Palosaari <crope@iki.fi> Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>
Antti Palosaari [Wed, 29 Apr 2015 13:10:01 +0000 (10:10 -0300)]
[media] msi001: revise synthesizer calculation
Update synthesizer calculation to model I prefer nowadays. It is mostly
just renaming some variables, but also minor functionality change how
integer and fractional part are divided (using div_u64_rem()). Also, add
'schematic' of synthesizer following my current understanding.
Signed-off-by: Antti Palosaari <crope@iki.fi> Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>
Thomas Reitmayr [Fri, 1 May 2015 23:18:04 +0000 (20:18 -0300)]
[media] media: Fix regression in some more dib0700 based devices
Fix an oops during device initialization by correctly setting size_of_priv
instead of leaving it 0.
The regression was introduced by 8abe4a0a3f6d4217b16a ("[media] dib7000:
export just one symbol") and only fixed for one type of dib0700 based
devices in 9e334c75642b6e5bfb95 ("[media] Fix regression in some dib0700
based devices").
Fixes: 8abe4a0a3f6d4217b16a ("[media] dib7000: export just one symbol") Cc: stable@vger.kernel.org # for version 3.17+ Signed-off-by: Thomas Reitmayr <treitmayr@devbase.at> Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>
Olli Salonen [Tue, 5 May 2015 16:33:55 +0000 (13:33 -0300)]
[media] dw2102: resync fifo when demod locks
If the streaming_ctrl is called to enable TS before demod has locked
the TS will be empty. Copied the solution from the dvbsky driver for the
TechnoTrend S2-4600 device: when the state changes from unlock to
lock, call su3000_streaming_ctrl again.
Signed-off-by: Olli Salonen <olli.salonen@iki.fi> Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>
Olli Salonen [Tue, 5 May 2015 16:33:52 +0000 (13:33 -0300)]
[media] dw2102: remove unnecessary printing of MAC address
While reading the MAC address for SU3000-based devices the system was
printing excessive debug information in the logs:
Output before the patch:
[ 1515.780692] bc 00 00 00 00 00
[ 1515.781440] bc ea 00 00 00 00
[ 1515.782251] bc ea 2b 00 00 00
[ 1515.783094] bc ea 2b 46 00 00
[ 1515.783816] bc ea 2b 46 12 00
[ 1515.784565] bc ea 2b 46 12 92
[ 1515.784571] dvb-usb: MAC address: bc:ea:2b:46:12:92
Output after the patch:
[ 3803.495706] dvb-usb: MAC address: bc:ea:2b:46:12:92
Signed-off-by: Olli Salonen <olli.salonen@iki.fi> Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>
Dan Carpenter [Tue, 21 Apr 2015 09:31:10 +0000 (06:31 -0300)]
[media] v4l: xilinx: harmless buffer overflow
My static checker warns that the name of the port can be 15 characters
when you consider the NUL terminator and that's one more than the 14
characters in name[]. Maybe it's an off-by-one?
It's unlikely that we hit the limit and even if we do the overflow will
only affect one of the two bytes of padding so it's harmless. Still
let's fix it and also change the sprintf() to snprintf().
Building for avr32 leads the following build warning:
drivers/media/rc/st_rc.c:270: warning: passing argument 1 of 'IS_ERR' discards qualifiers from pointer target type
drivers/media/rc/st_rc.c:271: warning: passing argument 1 of 'PTR_ERR' discards qualifiers from pointer target type
devm_ioremap_resource() returns void __iomem *, so change 'base' and
'rx_base' definitions accordingly.
Building for avr32 leads the following build warning:
drivers/media/rc/ir-hix5hd2.c:221: warning: passing argument 1 of 'IS_ERR' discards qualifiers from pointer target type
drivers/media/rc/ir-hix5hd2.c:222: warning: passing argument 1 of 'PTR_ERR' discards qualifiers from pointer target type
devm_ioremap_resource() returns void __iomem *, so change 'base' definition
accordingly.
In a rarely hit randconfig case, the r820t tuner driver can
get built when CONFIG_BITREVERSE is not selected by any
other driver, resulting in this error:
drivers/built-in.o: In function `r820t_read.constprop.3':
:(.text+0xa0594): undefined reference to `byte_rev_table'
For consistency, this adds the 'select BITREVERSE' that
all other similar drivers have.
drivers/media/platform/exynos4-is/fimc-is-i2c.c: In function 'fimc_is_i2c_probe':
drivers/media/platform/exynos4-is/fimc-is-i2c.c:58:8: error: implicit declaration of function 'i2c_add_adapter' [-Werror=implicit-function-declaration]
The dependency already exists for exynos-fimc-lite and s5p-fimc,
but is missing for exynos4-fimc.
The debug macros are not properly defined, as they generate warnings
like:
ERROR: Macros with complex values should be enclosed in parentheses
+#define core_dbg(fmt, arg...) if (core_debug) \
+ printk(KERN_DEBUG pr_fmt("core: " fmt), ## arg)
[media] saa7134-ts: use pr_fmt() at the debug macro
instead of using dev->name, let's use pr_fmt() like on the
other parts of saa7134. Also, rename the debug macro to
ts_dbg() to match the namespace for the debug macros.
[media] saa7134-i2c: make debug macros to use pr_fmt()
Converting debug prints to use pr_foo() is not trivial, as the
result will be a way worse than what's provided here, due to the
pieces of the code that prints the I2C transfers. Those use a
lot pr_cont(), and, depending on using either level 1 or 2,
a different set of macros are selected.
So, let's replace d1printk() and d2printk() macros by i2c_dbg()
and i2c_count() adding a debug level there.
On media drivers, debugging messages should be grouped into
categories, as this makes easier to debug the driver.
In the case of saa7134, the core has 2 debug categories, one
for IRQ, and another one for the core itself. The IRQ have
actually 2 levels of debug.
So, instead of using pr_dbg(), where everything would be in
the same box, let's define two macros that use pr_fmt(),
one for the core, and another one for irq.
With that, we can replace the remaining printk() occurrences
at the core to use either core_dbg() or irq_dbg(), depending
on the group of debug macros that need to be enabled.
[media] saa7134: instead of using printk KERN_foo, use pr_foo
Replaces all occurrences of printk with KERN_INFO, KERN_WARNING
and KERN_ERR to pr_info/pr_warning, pr_err, using this small
shell script:
for i in drivers/media/pci/saa7134/*.[ch]; do sed s,'printk(KERN_INFO ','pr_info(',g <$i >a && mv a $i; done
for i in drivers/media/pci/saa7134/*.[ch]; do sed s,'printk(KERN_ERR ','pr_err(',g <$i >a && mv a $i; done
for i in drivers/media/pci/saa7134/*.[ch]; do sed s,'printk(KERN_WARNING ','pr_warn(',g <$i >a && mv a $i; done
[media] rtl28xxu: fix return value check in rtl2832u_tuner_attach()
In case of error, the function platform_device_register_data() returns
ERR_PTR() and never returns NULL. The NULL test in the return value check
should be replaced with IS_ERR().
[media] mn88472: Fix possible leak in mn88472_init()
Commit 307e95c92257 ("[media] mn88472: implement firmware parity check")
introduced the usage of exit paths that do not free the already allocated
firmware data in case the parity handling fails. Go through the correct
exit paths. Detected by Coverity CID 1295989.
Signed-off-by: Christian Engelmayer <cengelma@gmx.at> Acked-by: Benjamin Larsson <benjamin@southpole.se> Signed-off-by: Antti Palosaari <crope@iki.fi> Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>
Olli Salonen [Tue, 5 May 2015 16:54:19 +0000 (13:54 -0300)]
[media] rtl2832: add support for GoTView MasterHD 3 USB tuner
GoTView MasterHD 3 is a DVB-T2/C USB 2.0 tuner.
It's based on the following components:
- USB bridge: RTL2832P (contains also DVB-T demodulator)
- Demodulator: Si2168-A30
- Tuner: Si2148-A20
The demodulator and the tuner will need firmwares. The Si2148 uses Si2158
firmware. Antti has the firmwares available for download:
http://palosaari.fi/linux/v4l-dvb/firmware/
Do note that for DVB-T either of the demodulators can be used. DVB-C and
DVB-T2 are only supported by the Si2168 demodulator. The driver will
register 2 frontends for the same adapter. Frontend 0 will be the RTL2832
demodulator and frontend 1 will be the Si2168 demodulator. The same
tuner is used for both.
As a consequence of the above, it's recommended to use application that
does implement proper DVBv5 support.
For some reason, the old I2C write method sporadically fails. Thus the
need for an option to only use the new I2C write method supported by the
RTL2832.
Signed-off-by: Olli Salonen <olli.salonen@iki.fi> Signed-off-by: Antti Palosaari <crope@iki.fi> Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>
Olli Salonen [Tue, 5 May 2015 16:54:17 +0000 (13:54 -0300)]
[media] si2157: support selection of IF interface
The chips supported by the si2157 driver have two IF outputs (either
pins 12+13 or pins 9+11). Instead of hardcoding the output to be used
add an option to choose which output shall be used.
As this patch changes the default behaviour, the IF interface is
specified in each driver currently using si2157 driver. This is to
keep bisectability.
Signed-off-by: Olli Salonen <olli.salonen@iki.fi> Signed-off-by: Antti Palosaari <crope@iki.fi> Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>
Olli Salonen [Sat, 21 Mar 2015 20:42:23 +0000 (17:42 -0300)]
[media] dw2102: TeVii S482 support
TeVii S482 is a PCIe device with two tuners that actually contains two
USB devices. The devices are visible in the lsusb printout.
Bus 006 Device 002: ID 9022:d483 TeVii Technology Ltd.
Bus 007 Device 002: ID 9022:d484 TeVii Technology Ltd.
The device itself works exactly with the same settings as TechnoTrend
TT-connect S2-4600. Firmware for DS3103 demodulator is required:
http://palosaari.fi/linux/v4l-dvb/firmware/M88DS3103/
This patch should be applied on top of the TT S2-4600 patch:
https://patchwork.linuxtv.org/patch/28818/
Signed-off-by: Olli Salonen <olli.salonen@iki.fi> Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>
Hans Verkuil [Fri, 24 Apr 2015 09:52:47 +0000 (06:52 -0300)]
[media] marvell-ccic: fix RGB444 format
The RGB444 format swapped the red and blue components, fix this.
Rather than making a new BGR444 format (as I proposed initially), Jon prefers
to just fix this and return the colors in the right order. I think that makes
sense in this case.
Since the RGB444 pixel format is deprecated due to the ambiguous specification
of the alpha component we use the XRGB444 pixel format instead (specified as having
no alpha channel).
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com> Acked-by: Jonathan Corbet <corbet@lwn.net> Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>
Philipp Zabel [Mon, 4 May 2015 10:51:07 +0000 (07:51 -0300)]
[media] coda: Set last buffer flag and fix EOS event
Setting the last buffer flag causes the videobuf2 core to return -EPIPE from
DQBUF calls on the capture queue after the last buffer is dequeued.
This patch also fixes the EOS event to conform to the specification. It now is
sent right after the last buffer has been decoded instead of when the last
buffer is dequeued.
Philipp Zabel [Mon, 4 May 2015 10:51:06 +0000 (07:51 -0300)]
[media] videobuf2: return -EPIPE from DQBUF after the last buffer
If the last buffer was dequeued from a capture queue, let poll return
immediately and let DQBUF return -EPIPE to signal there will no more
buffers to dequeue until STREAMOFF.
The driver signals the last buffer by setting the V4L2_BUF_FLAG_LAST.
To reenable dequeuing on the capture queue, the driver must explicitly
call vb2_clear_last_buffer_queued. The last buffer queued flag is
cleared automatically during STREAMOFF.
Signed-off-by: Philipp Zabel <p.zabel@pengutronix.de> Acked-by: Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by: Kamil Debski <k.debski@samsung.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>
Peter Seiderer [Mon, 4 May 2015 10:51:05 +0000 (07:51 -0300)]
[media] videodev2: Add V4L2_BUF_FLAG_LAST
This v4l2_buffer flag can be used by drivers to mark a capture buffer
as the last generated buffer, for example after a V4L2_DEC_CMD_STOP
command was issued.
Signed-off-by: Peter Seiderer <ps.report@gmx.net> Signed-off-by: Philipp Zabel <p.zabel@pengutronix.de> Acked-by: Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by: Kamil Debski <k.debski@samsung.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>
Document the interaction between VIDIOC_DECODER_CMD V4L2_DEC_CMD_STOP and
VIDIOC_ENCODER_CMD V4L2_ENC_CMD_STOP to start the draining, the V4L2_EVENT_EOS
event signalling all capture buffers are finished and ready to be dequeud,
the new V4L2_BUF_FLAG_LAST buffer flag indicating the last buffer being dequeued
from the capture queue, and the poll and VIDIOC_DQBUF ioctl return values once
the queue is drained.
Signed-off-by: Philipp Zabel <p.zabel@pengutronix.de> Acked-by: Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by: Kamil Debski <k.debski@samsung.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>
* tag 'v4.1-rc3': (381 commits)
Linux 4.1-rc3
drm: Zero out invalid vblank timestamp in drm_update_vblank_count.
m32r: make flush_cpumask non-volatile.
mnt: Fix fs_fully_visible to verify the root directory is visible
path_openat(): fix double fput()
namei: d_is_negative() should be checked before ->d_seq validation
ARM: dts: Add keep-power-in-suspend to WiFi SDIO node for exynos5250-snow
ARM: dts: Fix typo in trip point temperature for exynos5420/5440
ARM: dts: add 'rtc_src' clock to rtc node for exynos4412-odroid boards
ARM: dts: Make DP a consumer of DISP1 power domain on Exynos5420
MAINTAINERS: add Conexant Digicolor machines entry
MAINTAINERS: socfpga: update the git repo for SoCFPGA
drm/tegra: Don't use vblank_disable_immediate on incapable driver.
mmc: dw_mmc: dw_mci_get_cd check MMC_CAP_NONREMOVABLE
mmc: dw_mmc: init desc in dw_mci_idmac_init
ARM: multi_v7_defconfig: Select more FSL SoCs
MAINTAINERS: replace an AT91 maintainer
drivers: CCI: fix used_mask init in validate_group()
drm/radeon: stop trying to suspend UVD sessions
drm/radeon: more strictly validate the UVD codec
...
Linus Torvalds [Sun, 10 May 2015 21:58:53 +0000 (14:58 -0700)]
Merge branch 'drm-fixes' of git://people.freedesktop.org/~airlied/linux
Pull drm fixes from Dave Airlie:
"I really need to get back to sending these on my Friday, instead of my
Monday morning, but nothing too amazing in here: a few amdkfd fixes, a
few radeon fixes, i915 fixes, one tegra fix and one core fix"
* 'drm-fixes' of git://people.freedesktop.org/~airlied/linux:
drm: Zero out invalid vblank timestamp in drm_update_vblank_count.
drm/tegra: Don't use vblank_disable_immediate on incapable driver.
drm/radeon: stop trying to suspend UVD sessions
drm/radeon: more strictly validate the UVD codec
drm/radeon: make UVD handle checking more strict
drm/radeon: make VCE handle check more strict
drm/radeon: fix userptr lockup
drm/radeon: fix userptr BO unpin bug v3
drm/amdkfd: Initialize sdma vm when creating sdma queue
drm/amdkfd: Don't report local memory size
drm/amdkfd: allow unregister process with queues
drm/i915: Drop PIPE-A quirk for 945GSE HP Mini
drm/i915: Sink rate read should be saved in deca-kHz
drm/i915/dp: there is no audio on port A
drm/i915: Add missing MacBook Pro models with dual channel LVDS
drm/i915: Assume dual channel LVDS if pixel clock necessitates it
drm/radeon: don't setup audio on asics that don't support it
drm/radeon: disable semaphores for UVD V1 (v2)
Dave Airlie [Sun, 10 May 2015 20:06:22 +0000 (06:06 +1000)]
Merge tag 'drm-intel-fixes-2015-05-08' of git://anongit.freedesktop.org/drm-intel into drm-fixes
misc i915 fixes.
* tag 'drm-intel-fixes-2015-05-08' of git://anongit.freedesktop.org/drm-intel:
drm/i915: Drop PIPE-A quirk for 945GSE HP Mini
drm/i915: Sink rate read should be saved in deca-kHz
drm/i915/dp: there is no audio on port A
drm/i915: Add missing MacBook Pro models with dual channel LVDS
drm/i915: Assume dual channel LVDS if pixel clock necessitates it
Mario Kleiner [Tue, 7 Apr 2015 04:31:09 +0000 (06:31 +0200)]
drm: Zero out invalid vblank timestamp in drm_update_vblank_count.
Since commit 844b03f27739135fe1fed2fef06da0ffc4c7a081 we make
sure that after vblank irq off, we return the last valid
(vblank count, vblank timestamp) pair to clients, e.g., during
modesets, which is good.
An overlooked side effect of that commit for kms drivers without
support for precise vblank timestamping is that at vblank irq
enable, when we update the vblank counter from the hw counter, we
can't update the corresponding vblank timestamp, so now we have a
totally mismatched timestamp for the new count to confuse clients.
Restore old client visible behaviour from before Linux 3.17, but
zero out the timestamp at vblank counter update (instead of disable
as in original implementation) if we can't generate a meaningful
timestamp immediately for the new vblank counter. This will fix
this regression, so callers know they need to retry again later
if they need a valid timestamp, but at the same time preserves
the improvements made in the commit mentioned above.
Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com> Cc: <stable@vger.kernel.org> #v3.17+ Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Cc: Daniel Vetter <daniel@ffwll.ch> Signed-off-by: Dave Airlie <airlied@redhat.com>
projects / mirror_ubuntu-kernels.git / log