summaryrefslogtreecommitdiff
path: root/linux/drivers/media/video/videodev.c
AgeCommit message (Collapse)Author
2008-07-19backport commit 6606470dd1d628878383c96d10b52a77986ddac7Mauro Carvalho Chehab
From: Mauro Carvalho Chehab <mchehab@infradead.org> Author: Jonathan Corbet <corbet@lwn.net> Date: Fri May 16 14:28:31 2008 -0600 videodev: BKL pushdown Put explicit lock_kernel() calls into videodev_open(). That function itself seems OK, but one never knows about all the open() functions provided by underlying video drivers. Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
2008-07-18videodev/cx18: fix get_index bug and error-handling lock-upsHans Verkuil
From: Hans Verkuil <hverkuil@xs4all.nl> Fix a bug in get_index that was introduced earlier. Also fix two error handling lock-ups in videodev and cx18 that where found thanks to that bug. Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
2008-07-17videodev: add comment and remove magic number.Hans Verkuil
From: Hans Verkuil <hverkuil@xs4all.nl> Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
2008-06-24videodev: simplify get_index()Brandon Philips
Use Hans Verkuil's suggested method of implementing get_index which doesn't depend on class_for_each_device and instead uses the video_device array. This simplifies the code and reduces its memory footprint. Signed-off-by: Brandon Philips <bphilips@suse.de>
2008-07-17videodev: add support for kernels < 2.6.19Hans Verkuil
From: Hans Verkuil <hverkuil@xs4all.nl> Add the necessary compatibility code to handle the struct device/struct class_device differences. It was too much work (with uncertain benefits) to convert several drivers to handle kernels <2.6.19, so they are not built on these older kernels. Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
2008-07-17videodev/dabusb/dvbdev: fix 2.6.16 compilationHans Verkuil
From: Hans Verkuil <hverkuil@xs4all.nl> struct file_operations was made const in 2.6.17. Cast const away when compiling under 2.6.16. Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
2008-07-08v4l-dvb: remove support for kernels < 2.6.16Hans Verkuil
From: Hans Verkuil <hverkuil@xs4all.nl> Phase 4 removes the compatibility support for kernels < 2.6.16. Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
2008-07-08v4l-dvb: remove support for kernels < 2.6.13Hans Verkuil
From: Hans Verkuil <hverkuil@xs4all.nl> Phase 3 of the compat cleanup. Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
2008-06-29Fix whitespace diffs between -git and mercurialMauro Carvalho Chehab
From: Mauro Carvalho Chehab <mchehab@infradead.org> Kernel-sync: Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
2008-06-25videodev: allow PRIVATE_BASE controls when called through VIDIOC_G/S_CTRL.Hans Verkuil
From: Hans Verkuil <hverkuil@xs4all.nl> V4L2_CID_PRIVATE_BASE controls are not allowed when called from VIDIOC_S/G_EXT_CTRL as extended controls use a better mechanism for private controls. But still allow it when called from the VIDIOC_G/S_CTRL to extended control conversion in video_ioctl2() for backwards compatibility. Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
2008-06-25videodev: improve extended control support in video_ioctl2()Hans Verkuil
From: Hans Verkuil <hverkuil@xs4all.nl> - add sanity checks for the extended controls argument. - if the driver only supports extended controls, then convert old-style controls to an extended control callback. Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
2008-06-22videodev: fix/improve ioctl debuggingHans Verkuil
From: Hans Verkuil <hverkuil@xs4all.nl> Various ioctl debugging fixes and improvements: - use %x rather than %d for control IDs and bitmask fields - make two arrays const - show the whole control array for the ext_ctrl ioctls - print pix_fmt for V4L2_BUF_TYPE_VIDEO_OUTPUT - show full type name rather than an integer - fix CROPCAP debugging - fix G/S_TUNER debugging - show error code in case of an error - other small cleanups Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
2008-06-22videodev: fix compilation issue for kernels < 2.6.25Hans Verkuil
From: Hans Verkuil <hverkuil@xs4all.nl> Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
2008-06-22merge: http://www.linuxtv.org/hg/~hverkuil/v4l-dvbMauro Carvalho Chehab
From: Mauro Carvalho Chehab <mchehab@infradead.org> Fixed a small merge conflicts at: merging linux/drivers/media/video/cx18/cx18-cards.h warning: conflicts during merge. merging linux/drivers/media/video/cx18/cx18-cards.h failed! Basically, two new structures were added bellow the same place Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
2008-06-21videodev: simplify and fix standard enumerationHans Verkuil
From: Hans Verkuil <hverkuil@xs4all.nl> VIDIOC_ENUMSTD did not return all the PAL/SECAM/NTSC variants: it just returned one single PAL/SECAM/NTSC standard without separate entries for the trickier standards like NTSC-JP. Changed the code so that it behaves better. Also simplified the if/switch statements into a common standards lookup table. Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
2008-06-21videodev: zero fields for ENCODER_CMD and VIDIOC_G_SLICED_VBI_CAPHans Verkuil
From: Hans Verkuil <hverkuil@xs4all.nl> This avoids the need of memsets in the ivtv/cx18 drivers. Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
2008-06-20 Introduce "index" attribute for persistent video4linux device nodesMauro Carvalho Chehab
From: brandon@ifup.org <brandon@ifup.org> A number of V4L drivers have a mod param to specify their preferred minors. This is because it is often desirable for applications to have a static /dev name for a particular device. However, using minors has several disadvantages: 1) the requested minor may already be taken 2) using a mod param is driver specific 3) it requires every driver to add a param 4) requires configuration by hand This patch introduces an "index" attribute that when combined with udev rules can create static device paths like this: /dev/v4l/by-path/pci-0000\:00\:1d.2-usb-0\:1\:1.0-video0 /dev/v4l/by-path/pci-0000\:00\:1d.2-usb-0\:1\:1.0-video1 /dev/v4l/by-path/pci-0000\:00\:1d.2-usb-0\:1\:1.0-video2 $ ls -la /dev/v4l/by-path/pci-0000\:00\:1d.2-usb-0\:1\:1.0-video0 lrwxrwxrwx 1 root root 12 2008-04-28 00:02 /dev/v4l/by-path/pci-0000:00:1d.2-usb-0:1:1.0-video0 -> ../../video1 These paths are steady across reboots and should be resistant to rearranging across Kernel versions. video_register_device_index is available to drivers to request a specific index number. Signed-off-by: Brandon Philips <bphilips@suse.de> Signed-off-by: Kees Cook <kees@outflux.net> Signed-off-by: Kay Sievers <kay.sievers@vrfy.org> Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
2008-05-28videodev: renamed the vidioc_*_fmt_* callbacksHans Verkuil
From: Hans Verkuil <hverkuil@xs4all.nl> The naming for the callbacks that handle the VIDIOC_ENUM_FMT and VIDIOC_S/G/TRY_FMT ioctls was very confusing. Renamed it to match the v4l2_buf_type name. Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
2008-05-28videodev: add missing vidioc_try_fmt_sliced_vbi_output and VIDIOC_ENUMOUTPUT ↵Hans Verkuil
handling From: Hans Verkuil <hverkuil@xs4all.nl> There was no vidioc_try_fmt_sliced_vbi_output, instead vidioc_try_fmt_vbi_output was reused. The VIDIOC_ENUMOUTPUT handling was missing altogether, even though the callback existed. Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
2008-05-28videodev: add vidioc_g_std callback.Hans Verkuil
From: Hans Verkuil <hverkuil@xs4all.nl> The default videodev behavior for VIDIOC_G_STD is not correct for all devices. Add a new callback that drivers can use instead. Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
2008-05-27videodev: small fixes for VIDIOC_G_FREQUENCY and VIDIOC_G_FMTHans Verkuil
From: Hans Verkuil <hverkuil@xs4all.nl> __video_do_ioctl incorrectly zeroed the tuner field of v4l2_frequency and did not zero the full fmt union of v4l2_format. Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
2008-05-26 Hardware frequency seek ioctl interfaceMauro Carvalho Chehab
From: Tobias Lorenz <tobias.lorenz@gmx.net> Signed-off-by: Tobias Lorenz <tobias.lorenz@gmx.net> Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
2008-04-21merge: http://linuxtv.org/hg/~dougsland/videodevMauro Carvalho Chehab
From: Mauro Carvalho Chehab <mchehab@infradead.org> Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
2008-04-09videodev: Add default vidioc handlerDouglas Schilling Landgraf
From: Douglas Schilling Landgraf <dougsland@gmail.com> Added default vidioc handler for other private ioctls Signed-off-by: Douglas Schilling Landgraf <dougsland@gmail.com>
2008-04-13videodev: Some printk fixesMauro Carvalho Chehab
From: Mauro Carvalho Chehab <mchehab@infradead.org> Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
2008-04-13videobuf: Improve command output for debug purposesMauro Carvalho Chehab
From: Mauro Carvalho Chehab <mchehab@infradead.org> Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
2008-04-08media/video/ replace remaining __FUNCTION__ occurrencesMichael Krufky
From: Harvey Harrison <harvey.harrison@gmail.com> __FUNCTION__ is gcc-specific, use __func__ Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com> Signed-off-by: Michael Krufky <mkrufky@linuxtv.org>
2008-02-02Fix Kconfig dependenciesMauro Carvalho Chehab
From: Mauro Carvalho Chehab <mchehab@infradead.org> As pointed by Adrian Bunk, with I2C=m and VIDEO_DEV=y, videodev brokes. This patch moves the functions that videodev needs from v4l2-common. It also fixes some Kconfig changes. After this patch, I2C=m / VIDEO_DEV=y will make v4l2 core statically linked into kernel. v4l2-common will be m, and all V4L drivers will also be m. This approach is very conservative, since it is possible to have V4L drivers that don't need I2C or v4l2-common. The better is to map what drivers really need v4l2-common, making them to select v4l2-common, and allowing the others to be 'y', 'm' and 'n'. Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
2007-12-13video std is a bitmask. Better to print in hexaMauro Carvalho Chehab
From: Mauro Carvalho Chehab <mchehab@infradead.org> Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
2007-10-11Remove obsolete VIDIOC_S/G_MPEGCOMP ioctlsHans Verkuil
From: Hans Verkuil <hverkuil@xs4all.nl> Remove the obsolete VIDIOC_G_MPEGCOMP and VIDIOC_S_MPEGCOMP ioctls from the V4L2 API as per the removal schedule (October 2007). Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
2007-10-08compat: Remove unneeded #ifTrent Piepho
From: Trent Piepho <xyzzy@speakeasy.org> Both true and false branch of the #if had the exact same code, just line wrapped differently. Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
2007-10-08V4L: convert struct class_device to struct deviceMauro Carvalho Chehab
From: Kay Sievers <kay.sievers@vrfy.org> The currently used "struct class_device" will be removed from the kernel. Here is a patch that converts all users in drivers/media/video/ to struct device. Reviewed-by: Thierry Merle <thierry.merle@free.fr> Reviewed-by: Mike Isely <isely@pobox.com> Reviewed-by: Luca Risolia <luca.risolia@studio.unibo.it> Signed-off-by: Kay Sievers <kay.sievers@vrfy.org>
2007-08-06Dev.c: memset fixMauro Carvalho Chehab
From: Mariusz Kozlowski <m.kozlowski@tuxland.pl> Hello, Looks like memset() is zeroing wrong nr of bytes. Signed-off-by: Mariusz Kozlowski <m.kozlowski@tuxland.pl> Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
2007-07-02Fix v4l-dvb backward compatibilityMauro Carvalho Chehab
From: Mauro Carvalho Chehab <mchehab@infradead.org> Due to several internal API changes on kernel, kernel backward compatibility were lost. Basically, compat.h should be the last include for it to work properly. This patch basically reorders kernel headers to allow backward compat to work fine. Also: Some includes were added after some non-include macros, on old drivers. Better to keep all includes at the beginning of the files. Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
2007-05-13Backport kernel changeset e63340ae6b6205fef26b40a75673d1c9c0c8bb90Mauro Carvalho Chehab
From: Mauro Carvalho Chehab kernel-sync: The original patch description, from Randy Dunlap <randy.dunlap@oracle.com>: header cleaning: don't include smp_lock.h when not used Remove includes of <linux/smp_lock.h> where it is not used/needed. Suggested by Al Viro. Builds cleanly on x86_64, i386, alpha, ia64, powerpc, sparc, sparc64, and arm (all 59 defconfigs). Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
2007-05-07V4L2: videodev, allow debuggingMauro Carvalho Chehab
From: Jiri Slaby <jirislaby@gmail.com> videodev, allow debugging fix typo? in videodev.c to allow debugging Signed-off-by: Jiri Slaby <jirislaby@gmail.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
2007-05-01Fix: v4l1_compat should be called only if V4L1_COMPATMauro Carvalho Chehab
From: Mauro Carvalho Chehab <mchehab@infradead.org> Added also some explanations about V4L1 handling Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
2007-05-01VIDIOCGMBUF handling in video_ioctl2()Mauro Carvalho Chehab
From: Sam Revitch <sam.revitch@gmail.com> Correct handling of VIDIOCGMBUF in video_ioctl2(). Signed-off-by: Sam Revitch <samr7@cs.washington.edu> Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
2007-04-24Fix: vidioc_g_parm were not zeroing the memoryMauro Carvalho Chehab
From: Mauro Carvalho Chehab <mchehab@infradead.org> Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
2007-02-24Add support for VIDIOC_G_CHIP_IDENTHans Verkuil
From: Hans Verkuil <hverkuil@xs4all.nl> VIDIOC_G_CHIP_IDENT improves debugging of card problems: it can be used to detect which chips are on the board and based on that information selected register dumps can be made, making it easy to debug complicated media chips containing tens or hundreds of registers. This ioctl replaces the internal VIDIOC_INT_G_CHIP_IDENT ioctl. Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl> CC: Jonathan Corbet <corbet@lwn.net>
2007-02-21Add support for video output overlays.Hans Verkuil
From: Hans Verkuil <hverkuil@xs4all.nl> Add V4L2_BUF_TYPE_VIDEO_OUTPUT_OVERLAY support. Also add support for local and global alpha overlays. Add new field enums V4L2_FIELD_INTERLACED_TB and V4L2_FIELD_INTERLACED_BT. These changes are needed to support the ivtv On Screen Display features. Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
2007-02-21Backport of git changeset cd354f1ae75e6466a7e31b727faede57a1f89ca5Mauro Carvalho Chehab
From: Mauro Carvalho Chehab <mchehab@infradead.org> Backport a changeset from Schmielau <tim@physik3.uni-rostock.de> that removes uneeded includes for linux/sched.h Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
2007-02-21Backport: make file_operations constMauro Carvalho Chehab
From: Mauro Carvalho Chehab <mchehab@infradead.org> Backport a kernel patch by Arjan van de Ven <arjan@linux.intel.com> that made all file_operations instances const. Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
2007-02-18Add VIDIOC_TRY_ENCODER_CMD and VIDIOC_ENCODER_CMD ioctls.Hans Verkuil
From: Hans Verkuil <hverkuil@xs4all.nl> Add support for starting, stopping, pausing and resuming an MPEG (or similar compressed stream) encoder. Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
2007-02-18Add VIDIOC_G_ENC_INDEX ioctlHans Verkuil
From: Hans Verkuil <hverkuil@xs4all.nl> The VIDIOC_G_ENC_INDEX ioctl can obtain the MPEG index from an MPEG encoder. Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
2007-01-30Add checks for CAP_SYS_ADMIN to VIDIOC_DBG_G_REGISTERTrent Piepho
From: Trent Piepho <xyzzy@speakeasy.org> Before, root privileges were only needed to set hardware registers, not to read them. On some hardware, reading from the wrong place at the wrong time can hang the machine. So, to be consistent, root privileges are required to read registers on all hardware. Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
2007-01-23Make VIDIOC_INT_[SG]_REGISTER ioctls no longer internal onlyTrent Piepho
From: Trent Piepho <xyzzy@speakeasy.org> The direct register access ioctls were defined as kernel internal only, but they are very useful for debugging hardware from userspace and are used as such. Officially export them. VIDIOC_INT_[SG]_REGISTER is renamed to VIDIOC_DBG_[SG]_REGISTER Definition of ioctl and struct v4l2_register is moved from v4l2-common.h to videodev2.h. Types used in struct v4l2_register are changed to the userspace exportable versions (u32 -> __u32, etc). Use of VIDIOC_DBG_S_REGISTER requires CAP_SYS_ADMIN permission, so move the check into the video_ioctl2() dispatcher so it doesn't need to be duplicated in each driver's call-back function. CAP_SYS_ADMIN check is added to pvrusb2 (which doesn't use video_ioctl2). Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
2007-01-22Restore VIDIOC_INT_[SG]_REGISTER callsTrent Piepho
From: Trent Piepho <xyzzy@speakeasy.org> Add support for these ioctls to the video_ioctl2 system and the cx88 driver. Signed-off-by: Trent Piepho <xyzzy@speakeasy.org>
2006-12-16Several stuff backported from 2.6.19-git seriesMauro Carvalho Chehab
From: Mauro Carvalho Chehab <mchehab@infradead.org> - INIT_WORK replaced by INIT_DELAYED_WORK - struct work_struct replaced by struct delayed_work - callback parameters also changed - SLAB_KERNEL replaced by GFP_KERNEL - linux/suspend.h replaced by linux/freezer.h - on cpia: INIT_WORK replaced by INIT_WORK_NAR - file->f_dentry->d_inode replaced by file->f_path.dentry->d_inode Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
2006-12-01Improve debug msgs to show fourcc and buffer length on video_ioctl2Mauro Carvalho Chehab
From: Mauro Carvalho Chehab <mchehab@infradead.org> Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
generated by cgit v1.2.3 (git 2.25.1) at 2025-01-10 07:03:41 +0000