- 27 Apr, 2007 40 commits
-
-
Trent Piepho authored
One bit wide bitfields need to declared unsigned to have the range 0 to 1, or they have the range -1 to 0. A few techniques to reduce the driver's size by about 1700 bytes on ia32, probably more on x86-64. Put the biggest fields first, less padding is necessary that way. Put fields with a limited range into a smaller type. For example VideoChannels will fit in 3 bits, and TunerType can use 8 bits. Vin_Reg1, Vin_Reg2, and Dvi_yuv define values for 8-bit registers, but they can't just go into an 8-bit field with no changes, since -1 was used as a flag to indicate a value was not present. So what we do is create a one-bit flag for each one to indicate if a value is or is not present. This only takes 9 bits and has the added advantage that when the register isn't overridden (Vin_Reg[12] never are) it doesn't need to appear in the structure definition since the default value for the flag will be zero. Signed-off-by:
Trent Piepho <xyzzy@speakeasy.org> Acked-by:
Thierry MERLE <thierry.merle@free.fr> Signed-off-by:
Mauro Carvalho Chehab <mchehab@infradead.org>
-
Mauro Carvalho Chehab authored
This patch reduces usbvision driver on about 1Kb on i386 over the original version with the old struct: text data bss dec hex filename 52312 11848 60 64220 fadc old/usbvision.ko 52474 10708 60 63242 f70a new/usbvision.ko Acked-by:
-
Mauro Carvalho Chehab authored
Changed usbvision cards table to allow: 1) Not repeat USB ID on two structs; 2) Not need to specify both usb and card description tables at the same order, removing some magic; Some cards had duplicated names. Fixed. A test for an specific board were doing by using a string comparation. The comparation were wrong. Also, it is not a good practice to recognize a board based on his string name. Acked-by:
-
Markus Rechberger authored
fixed hotplugging for dvbnet Signed-off-by:
Michal CIJOML Semler <cijoml@volny.cz> Signed-off-by:
Markus Rechberger <markus.rechberger@amd.com> Signed-off-by:
-
Markus Rechberger authored
fixing hotplug issue for demux[n] and dvr[n] Signed-off-by:
-
Markus Rechberger authored
fixed frontend hotplug issue Signed-off-by:
-
Mike Isely authored
The V4L2 API requires a unique bus_info string returned as part of the v4l2_capability structure. These changes gather up the USB address information, from the underlying device, into a string and report that out through v4l2 and via sysfs (for completeness). Signed-off-by:
Mike Isely <isely@pobox.com> Signed-off-by:
-
Hendrik Borghorst authored
If you modprobe the b2c2-flexcop-pci module you got a hardlock of your system. This is due the usage of spin_lock before spin_lock_init is called. Signed-off-by: -
Trent Piepho authored
sn9c102 is a v4l2 driver, except it used a couple v4l1 helper functions. Stop using those functions and depend on V4L2 in Kconfig. Acked-by:
Luca Risolia <luca.risolia@studio.unibo.it> Signed-off-by:
-
Trent Piepho authored
Makes sure they don't get copied onto the stack. Signed-off-by:
-
Trent Piepho authored
There were many places in the driver which had long sequences of constant register initializations. These were done with one function call per register. The register address and value were immediate values in the function calls. This is very inefficient, as each register and value take twice the space when they are code, as each includes a push instruction to put it on the stack. There there is the overhead, both size and time, for a function call for each register. It's also quite a few lines of C code to do this. The patch creates a function that writes multiple registers from a list, and a macro that makes it easy to construct a such a list as a const static local to send to the function. This gets rid of quite a bit of C code, and shrinks the driver by around 8k, while at the same time being more efficient. Acked-by:
-
Mauro Carvalho Chehab authored
Signed-off-by: -
Scott Alfter authored
SSAI (www.ssai.us) makes several Bt878-based capture cards that get used in our surveillance, conferencing, and medical imaging systems. The attached relatively small patch adds support for these cards, which fall into two broad * boards with one or more Bt878s, one or more composite inputs, and no S-video or tuner inputs * boards with one Bt878, one composite input, one S-video input, and no tuner input Signed-off-by:
Scott Alfter <salfter@ssai.us> Signed-off-by:
-
Trent Piepho authored
It's KERN_WARNING "lgdt....", not "KERN_WARNING lgdt...." Signed-off-by:
-
Jean Delvare authored
The tda7432, tda9875 and tvaudio media drivers don't need to include the linux/i2c-algo-bit.h header file. Signed-off-by:
Jean Delvare <khali@linux-fr.org> Signed-off-by:
-
Jean Delvare authored
Add support for the S-Video and CVBS (composite) analog video inputs of the Leadtek WinFast DTV1000 T adapter. Signed-off-by:
Michael Krufky <mkrufky@linuxtv.org> Signed-off-by:
-
Amit Choudhary authored
Check the return value of kmalloc() in function se401_start_stream(), in file drivers/media/video/se401.c. Signed-off-by:
Amit Choudhary <amit2030@gmail.com> Signed-off-by:
-
Amit Choudhary authored
Signed-off-by:
-
Mauro Carvalho Chehab authored
Signed-off-by: -
Ed Vipas authored
This patch just defines the remote control type. Signed-off-by:
Ed Vipas <epvipas@gmail.com> Signed-off-by:
Hartmut Hackmann <hartmut.hackmann@t-online.de> Signed-off-by:
-
Hartmut Hackmann authored
When the tuner is attached, the tda10046 is not initilized yet, so it is searching for its firmware. If the tuner is attached to the tda10046 silent i2c port, a bus collision can occur. Now the version is probed during the first init or sleep call. Signed-off-by:
-
Hartmut Hackmann authored
Set tda8290 to analog mode after init, otherwise the tuner driver will not accept i.e. the standby command. Signed-off-by:
-
Jarod Wilson authored
This patch causes the bttv driver to automatically load the dvb-bt8xx module for bttv/dvb hybrid cards. Successfully tested with a pcHDTV HD-2000 card. This patch is based on the recent patches to enable autoloading of cx88-dvb, cx88-blackbird and saa7134-dvb. Signed-off-by:
Jarod Wilson <jwilson@redhat.com> Signed-off-by:
-
Damian Minkov authored
Fix audio input source for capturing(playing) audio on AverTv Go 007 cards. Signed-off-by:
Damian Minkov <damencho@damencho.com> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
-
Mauro Carvalho Chehab authored
cx88_print_irqbits were expecting a string pointer with 32 bytes. Better to pass the string size and use ARRAY_SIZE on its calls. Signed-off-by: -
Mauro Carvalho Chehab authored
Signed-off-by: -
Mauro Carvalho Chehab authored
ATI HDTV Wonder needs to initialize some registers before allowing the tuner to start working. The current logic have lots of magic. This patch makes the code cleaner, using ARRAY_SIZE() for the initialization array and using a bidimensional array, instead of doing some stuff like: &buffer[i+2] Signed-off-by: -
Mauro Carvalho Chehab authored
Signed-off-by: -
Mauro Carvalho Chehab authored
there were several "magic" for loops, addressing gpiomux array size (4). Adrian Bunk showed that one of the loops were wrong, going from 0 to 4. This patch provides the right fix for this trouble, by using ARRAY_SIZE on all places where we have a for loop using gpiomux. Thanks to: Adrian Bunk <bunk@stusta.de> for pointing me about this trouble. Signed-off-by: -
Luca Risolia authored
@ Don't assume that SOF headers can't cross packets boundaries @ Fix compression quality selection + Add support for MI-0360 image sensor * Documentation updates @ Fix sysfs @ MI0343 rewritten * HV7131R color fixes and add new ABLC control * Rename the archive from "sn9c102" to "sn9c1xx" * fix typos * better support for TAS5110D @ fix OV7630 wrong colors @ Don't return an error if no input buffers are enqueued yet on VIDIOC_STREAMON * Add informations about colorspaces * More appropriate error codes in case of failure of some system calls * More precise hardware detection * Add more informations about supported hardware in the documentation + More supported devices + Add support for HV7131R image sensor Signed-off-by:
-
Adrian Bunk authored
After looking at a section bug (in the non-modular case, clearly non-init code referenced the __initdata parport_nr[]), I thought it was time to convert this driver to module_{init,exit}. Signed-off-by:Adrian Bunk <bunk@stusta.de> Signed-off-by:
-
Jonathan Corbet authored
Add raw bayer support to the ov7670 driver Signed-off-by:
Jonathan Corbet <corbet@lwn.net> Signed-off-by:
-
Jonathan Corbet authored
Fix up Cafe/ov7670 copyrights and maintainer entries Signed-off-by:
-
Jonathan Corbet authored
Don't mirror ov7670 images by default. The ov7670 sensor driver sets the mirror bit by default, which is not the desired mode. OLPC has been running with this patch for a while. Signed-off-by:
-
Jonathan Corbet authored
Add suspend/resume support to the Cafe CCIC driver. Signed-off-by:
-
Jonathan Corbet authored
Fix up unsociable Cafe CCIC delays. The Cafe CCIC driver contains some lengthy delays, some of which are unnecessary and some of which are done under lock. Some were marked with comments, but the comments somehow failed to make the issue go away. So fix it for real. Signed-off-by:
-
Jonathan Corbet authored
Remove an obsolete PCI ID. The CAFE driver includes three PCI IDs, one of which corresponds to a development board which is no longer in use. Signed-off-by:
-
Jonathan Corbet authored
Set the PCI dev in the V4L2 dev so that the proper sysfs link gets made Signed-off-by:
Dan Williams <dcbw@redhat.com> Signed-off-by:
-
Mauro Carvalho Chehab authored
Logitech ViewPort AV 100 has the same internals as Cisco VT Camera. Fixing Pwc driver to handle it properly. Also, fixed the comments for both cameras. Thanks to Martin Rubli for pointing me this. Signed-off-by: -
Jean Tourrilhes authored
I have a Cisco VT Camera, and it was just collecting dust. I decided to try connecting it to my Linux box at home. Just a disgression about the product. The Cisco VT Camera is a webcam Cisco sold to work with their IP phone hardware and software. It's mostly useless on Windows, as it interfaces only to Cisco software. You can find some for cheap on eBay... Physically, it's just a Logitech Pro 4000. The only difference with the Pro 4000 is the Cisco logo and that it's grey like the Pro 3000. I believe Cisco is now selling the Cisco VT Camera II, which look to be something else... So, assuming that it was a Pro 4000 inside, I created the little patch attached. I'm new to webcam under Linux, but I managed to get an image from it using xawtv, and the image looked all right, so I consider that a success. The imaged seemed a bit small and I could not get the microphone driver loaded, but I assume it's my lack of experience. Note that I did not try any other type_id, but this one works great. Signed-off-by:
-
