aboutsummaryrefslogtreecommitdiffstats
AgeCommit message (Collapse)AuthorFilesLines
2011-06-19check_needs.pl: Improve the hints codeHEADmasterMauro Carvalho Chehab1-9/+58
For now, it presents only the hints for Fedora/RHEL, as those are the only distros I use. It should be easy to add support for others. Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2011-06-19move all checks to check_needs.plMauro Carvalho Chehab2-4/+19
This helps to add an OS-dependent logic on it Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2011-06-17rmmod.pl: Add a logic to allow removing audio modules with pulseaudioMauro Carvalho Chehab2-2/+71
Pulseaudio keeps audio devices opened forever. In order to be able to remove a device, pulseaudio needs to de-allocate the device. Unfortunately, pulseaudio recognizes alsa drivers as "module" (an integer number, not related to the device nodename), and only allows module removal if running with user matches the console owner. The logic inside rmmod.pl will now take the above into account. So, it will detect if pulseaudio is running. If it is, it will: 1) list the pulseaudio modules with "pacmd list-sinks" and "pacmd list-sources" 2) detect if any of the modules there is provided by a v4l device; 3) If they're provided by a video device, it removes the module with: pactl unload-module 26 Even the above logic is not perfect as, due to a pulseaudio libs bug, pulseaudio can't detect the name of em28xx-alsa driver, as it uses the same interface as the video node. Similar hacks will may be needed for other USB devices, like tm6000 and cx231xx. Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2011-06-17use_dir.pl: Fix the git_log creation logicMauro Carvalho Chehab1-1/+1
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2011-06-16The *keycode_new bits came in .37, not .36Jarod Wilson1-1/+0
Signed-off-by: Jarod Wilson <jarod@redhat.com>
2011-06-10linux/Makefile: Fix aspas escaping at git_logMauro Carvalho Chehab2-3/+3
The aspas scaping logic were breaking the compilation. Fixes it, by just removing the aspas character. While here, use the "-n" git log syntax, instead of a weird logic to just show the last 3 logs. Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2011-06-06IR driver ite-cir needs pnp_driver callback shutdown, which is not in 2.6.31.Hans Verkuil1-0/+1
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
2011-05-31Fix Compilation breakage with 3.0-rc1 patchesMauro Carvalho Chehab1-2/+4
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2011-05-29Remove warnings for macros already definedMauro Carvalho Chehab1-2/+3
On some distro-patched kernels, pr_warn and pr_err were backported. Fix the build logic to avoid warnings like: In file included from <command-line>:0: /home/v4l/media_build/v4l/compat.h:324:1: warning: "pr_warn" redefined Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2011-05-25keycode_new showed up in 2.6.36, apply patch to 36 and 37 tooJarod Wilson2-0/+2
Signed-off-by: Jarod Wilson <jarod@redhat.com>
2011-05-25Use {g,s}etkeycode_new on 2.6.38Jarod Wilson2-0/+35
There was a temporary {g,s}etkeycode_new in the input layer in 2.6.38 as we moved from an older definition to a new one. This partially reverts upstream commit aebd636bd60e2dda0ebc907dd5f7f4a45174411c so that we use the _new variant when building atop 2.6.38. Signed-off-by: Jarod Wilson <jarod@redhat.com>
2011-05-19make_makefile.pl: better handle dependency issuesMauro Carvalho Chehab1-7/+7
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2011-05-19check_needs.pl: Improve script to point to all issues at onceMauro Carvalho Chehab1-2/+6
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2011-05-06Add a new missing include fileMauro Carvalho Chehab1-0/+1
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2011-05-03Add compat bits for IS_ERR_OR_NULL()Mauro Carvalho Chehab2-0/+5
This is defined at include/linux/err.h as: static inline long __must_check IS_ERR_OR_NULL(const void *ptr) { return !ptr || IS_ERR_VALUE((unsigned long)ptr); } To make it work, it is as simple as defining IS_ERR_OR_NULL(ptr) as: (!(ptr) || IS_ERR_VALUE((unsigned long)(ptr))) And add a new check for it: check_file_for_func("include/linux/err.h", "IS_ERR_OR_NULL", "NEED_IS_ERR_OR_NULL"); Signed-off-by: Mauro Carvalho Chehab <maurochehab@gmail.com>
2011-05-02mfd header changes for 2.6.39 prevents timberdale compilation on older kernels.Hans Verkuil1-0/+2
Exclude timberdale drivers from building on pre-2.6.39 kernels. Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
2011-04-20Declare KEY_IMAGES (needed if kernel < 2.6.39-rc4)Mauro Carvalho Chehab1-0/+4
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2011-04-17Remove more trash from ubuntu-proprietary directory layoutMauro Carvalho Chehab1-1/+2
Signed-off-by: Mauro Carvalho Chehab <mchehab@nehalem.(none)>
2011-04-07Fix make tar and make dir targets.Hans Verkuil2-2/+2
The git_log generation is now identical for make todaytar and these other make targets. Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
2011-04-07Properly escape characters '%' '"' and '\' at git_logMauro Carvalho Chehab2-3/+3
Instead of replacing all special characters at git_log, just replace the ones that would cause a trouble for printk. Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2011-04-07fix git_log informationMauro Carvalho Chehab3-8/+4
git_log should contain the last applied patches to a tree. It is there to help developers to analize an user report. Due to two small bugs, it were producing long lists, and outputing control characters like \, % and ". This patch reverts changeset 139af0392c9b3bb0cf13235762931b2f93f89c61 Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2011-04-07Add missing find_next_bit_le and find_next_zero_bit_le for kernels < 2.6.39.Hans Verkuil1-0/+15
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
2011-04-07For now do not print $media_build_version.Hans Verkuil1-1/+5
$media_build_version can be huge (particularly after a merge commit) and full of weird characters. This should be done differently, but for now just don't use it. Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
2011-04-06The header i2c-id.h doesn't exist anymore on .39-rc. Remove it.Mauro Carvalho Chehab1-1/+0
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2011-04-06Skip commented patches in the backports/*_series files.Hans Verkuil1-2/+2
Makes it easier when debugging patch failures. Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
2011-04-04Fix compilation with kernels older than 2.6.36Mauro Carvalho Chehab8-10/+160
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2011-04-04Refresh patches to avoid breakageMauro Carvalho Chehab10-224/+225
After pulling from 2.6.39 changes, two patches broke, one due to just context changes, and the other one due to a change that were already merged upstream. While fixing them, I did a quilt refresh on the patches that used fuzz logic. Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2011-03-28Remove yaVDR files at rminstall targetMauro Carvalho Chehab1-5/+8
Ubuntu's based yaVDR puts driver stuff also at the wrong dir Signed-off-by: Mauro Carvalho Chehab <mchehab@nehalem.(none)>
2011-03-23Add altera driver to stagingMauro Carvalho Chehab2-1/+2
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2011-03-21Escape " character inside patch name on version_patch.plMauro Carvalho Chehab1-0/+1
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2011-03-21Fix its usage with make dir DIR=Mauro Carvalho Chehab1-1/+1
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2011-03-21check_needs.pl: Fix error messageMauro Carvalho Chehab1-1/+0
Instead of a simple "die" with Can't find, provide a more useful way: $ ./check_needs.pl please install "foo", otherwise, build won't work. This program is generally found at "bar" package. need foo at ./check_needs.pl line 20. Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2011-03-21Verify if everything needed to compile is thereMauro Carvalho Chehab2-1/+33
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2011-03-21Improve warning messagesMauro Carvalho Chehab1-4/+4
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2011-03-21Fix addition of git_log into the generated tarballMauro Carvalho Chehab2-4/+4
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2011-03-21Update list of files to ignoreMauro Carvalho Chehab3-1278/+21
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2011-03-21Be sure that, with untar, it will use the downloaded stuffMauro Carvalho Chehab1-1/+1
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2011-03-21Add a printk messages for media buildsMauro Carvalho Chehab3-2/+100
When someone reports an error, we need to know what are the applied patches. Also, users need to be warned that using the backported tree is not the same as running a stable kernel. This patch adds a logic that patches the subsystem core in order to indicate that an experimental tree is being used. It also prints the latest applied patches, in order to help developers when a bug report is filled. Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2011-03-21use_dir.pl: update linux/git_log fileMauro Carvalho Chehab1-0/+1
git_log is used to help developers to know what's the latest .git patch applied from the tree. Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2011-03-11Don't build radio-wl1273 for kernels < 2.6.39.Hans Verkuil1-0/+4
MFD header changes cause errors when building this driver. Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
2011-03-11Fix 2.6.31 rc-main.c warning.Hans Verkuil2-0/+12
The field 'groups' in struct device_type isn't const in 2.6.31. Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
2011-03-11Update the rc-main.c patch.Hans Verkuil1-18/+29
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
2011-03-11Add new includes.Hans Verkuil1-0/+3
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
2011-03-07Remove temporary firewire workaround: no longer needed.Hans Verkuil1-7/+0
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
2011-03-07Add missing linux/ti_wilink_st.h include.Hans Verkuil1-0/+1
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
2011-03-06Fix 2.6.31 compile warnings (vm_ops isn't const in 2.6.31)Hans Verkuil2-1/+190
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
2011-03-06Fix pvrusb2 compile warnings.Hans Verkuil1-1/+1
media_build/v4l/pvrusb2-sysfs.c: In function 'pvr2_sysfs_add_control': media_build/v4l/pvrusb2-sysfs.c:327:2: warning: statement with no effect media_build/v4l/pvrusb2-sysfs.c:332:2: warning: statement with no effect media_build/v4l/pvrusb2-sysfs.c:337:2: warning: statement with no effect media_build/v4l/pvrusb2-sysfs.c:342:2: warning: statement with no effect media_build/v4l/pvrusb2-sysfs.c:347:2: warning: statement with no effect media_build/v4l/pvrusb2-sysfs.c:352:2: warning: statement with no effect media_build/v4l/pvrusb2-sysfs.c:356:2: warning: statement with no effect media_build/v4l/pvrusb2-sysfs.c:360:2: warning: statement with no effect media_build/v4l/pvrusb2-sysfs.c:365:2: warning: statement with no effect sysfs_attr_init(a) should define to nothing instead of '0'. Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
2011-03-06Fix 2.6.37 via-camera buildHans Verkuil1-0/+1
via-camera needs linux/via-core.h. Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
2011-03-05Add backport support for usleep_rangeMauro Carvalho Chehab2-0/+5
This is not a perfect backport, but it should work Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2011-03-03Add altera.h header to the building systemMauro Carvalho Chehab1-0/+1
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

Privacy Policy