- 24 Mar, 2013 40 commits
-
-
Hans Verkuil authored
Signed-off-by:
Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by:
Mauro Carvalho Chehab <mchehab@redhat.com>
-
Hans Verkuil authored
This API is now obsolete and can be removed. Signed-off-by:
-
Hans Verkuil authored
This API is no longer used by any driver and so can be removed from the documentation. Signed-off-by:
-
Hans Verkuil authored
These ioctls are no longer used by any drivers, so remove them. [mchehab@redhat.com: Fix merge conflict] Signed-off-by:
-
Hans Verkuil authored
These ops are no longer used, so it's time to remove them. Signed-off-by:
-
Hans Verkuil authored
It's no longer used, so it can now be removed. Signed-off-by:
-
Hans Verkuil authored
Taken from the 0.9.8-5 version of the out-of-tree wis-go7007 driver. Signed-off-by:
-
Hans Verkuil authored
Taken from the 0.9.8-5 version of the out-of-tree wis-go7007 driver. Signed-off-by:
-
Hans Verkuil authored
The firmware is always the same, so drop it from the board configs. Also make the board configs const. Signed-off-by:
-
Hans Verkuil authored
Actually it's called "Group of VOP" start code in MPEG4 terminology. Taken from the 0.9.8-5 version of the out-of-tree wis-go7007 driver. Signed-off-by:
-
Hans Verkuil authored
But now as a proper standard MPEG control. Signed-off-by:
-
Volokh Konstantin authored
Signed-off-by:
Volokh Konstantin <volokh84@gmail.com> Signed-off-by:
-
Hans Verkuil authored
Signed-off-by:
-
Hans Verkuil authored
Signed-off-by:
-
Hans Verkuil authored
Add support for the other devices that need to load the boot firmware. All firmware files are now placed in a single go7007 directory. Also remove the device_extension_s stuff: this is clearly a left-over from the olden days. Signed-off-by:
-
Hans Verkuil authored
All s2250 names are renamed to go7007. This will be the generic go7007 firmware loader for any go7007 device, not just for the s2250/1. Signed-off-by:
-
Hans Verkuil authored
Signed-off-by:
-
Hans Verkuil authored
Remove these dummy cropping functions: cropping was never implemented. Signed-off-by:
-
Hans Verkuil authored
Do not allocate it, but just embed in the go7007 struct. Signed-off-by:
-
Hans Verkuil authored
Convert this driver to videobuf2 and core locking. Signed-off-by:
-
Hans Verkuil authored
Remove struct go7007_file: all fields contained in that struct are moved to the go7007 struct since they are really global values. The lock has just been deleted (what's the point of a per-fh lock??). Signed-off-by:
-
Hans Verkuil authored
The s_crystal_freq operation has to be called for the saa7115 to set up the audio clock correctly. Signed-off-by:
-
Hans Verkuil authored
There really is no need to split out the board IDs for each tuner. That's what the tuner_type is for, after all. Signed-off-by:
-
Hans Verkuil authored
The go7007 produces elementary streams, so we shouldn't use the STREAM_TYPE control, since that is for multiplexed streams. Instead use V4L2_PIX_FMT_MPEG1/2/4. Initially set up all the values for MPEG-2 dvd-mode, and select the dvd_mode field if those values match that setup precisely. Clean up lots of obsolete code relating to the custom ioctls for the MPEG support. Signed-off-by:
-
Hans Verkuil authored
- The Adlink is initially detected as a sensor board, so the driver config is set up as if it was a sensor. Later the driver discovers that it really is a video capture card and that means that the driver configuration has to be updated for a PAL/NTSC capture card. - Setup the correct initial std based on the TV tuner type. - Remember the exact std the user selected, don't map it to STD_PAL or NTSC. - Use v4l2_disable_ioctl to disable ioctls based on the board config and drop a lot of checking code in each of those ioctls since that is no longer needed. - enum_input should use tvnorms to fill its std field. - configure the input, std and initial tuner frequency at driver initialization. - fix std handling in the s2250 driver. - the code handling scaling for devices without a scaler was broken. This is now fixed. - correctly set readbuffers and capability in g/s_parm. - remove the bogus test for capturemode in s_parm. - correctly implement enum_framesizes/intervals. Signed-off-by:
-
Hans Verkuil authored
Signed-off-by:
-
Hans Verkuil authored
Signed-off-by:
-
Hans Verkuil authored
Signed-off-by:
-
Hans Verkuil authored
Just add a read-only V4L2_CID_JPEG_ACTIVE_MARKER control to replace the JPEGCOMP support. Signed-off-by:
-
Hans Verkuil authored
- use the v4l2_device's release() callback - remove the unnecessary ref_count - don't free usb data structures on disconnect, only do that in the final release callback. This is the correct way in order to safely handle disconnect and removal of modules. Signed-off-by:
-
Hans Verkuil authored
Don't load it everytime you stop encoding. Instead remember it. Another reason for not loading it every time is that this could be called from within the release() file operation, which turns out to be deadly. Signed-off-by:
-
Hans Verkuil authored
- Don't pass data allocated on the stack to usb_control_msg. - Use dma_mapping_error after calling dma_map_page(). Signed-off-by:
-
Hans Verkuil authored
It's deprecated and replaced by g_std. Since this driver already implements g_std the use of current_norm can just be removed. Signed-off-by:
-
Hans Verkuil authored
And don't set the version field, the core does that for you. Signed-off-by:
-
Hans Verkuil authored
The v2 of this function doesn't do DMA to objects on the stack like its predecessor does. Signed-off-by:
-
Hans Verkuil authored
Since we now know what audio inputs there are, we can also get/set and enumerate them. Signed-off-by:
-
Hans Verkuil authored
These are now no longer used and so they can be removed. Signed-off-by:
-
Hans Verkuil authored
Instead of using the wis-* drivers we now use the standard 'proper' subdev drivers. The board configuration tables now also list the possible audio inputs, this will be used later to implement audio inputs. Special mention deserves a little change in set_capture_size() where the height passed to s_mbus_fmt is doubled: that is because the saa7115 driver expects to see the frame height, not the field height as the wis_saa7115 driver did. Another change is that the tuner input is moved from last to the first input, which is consistent with the common practice in other video drivers. Signed-off-by:
-
Volokh Konstantin authored
Do i2c initialization via struct item as tw2804 has a 0x00 i2c address, so we need to use the I2C_CLIENT_TEN flag for validity. Signed-off-by:
-
Volokh Konstantin authored
Switch off all ADC (max 4) with first init, we control it when starting/stopping streaming. Signed-off-by:
-
