aboutsummaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAge
* Handle the alternate spellings of the xkb options.Adam Jackson2008-03-09
|
* Merge branch 'evdev-2.0-branch'Adam Jackson2008-03-08
|\ | | | | | | | | | | | | This resets to before the ad-hoc input hotplug state, ports that to current server ABI, adds input hotplug integration, and re-adds support for absolute coordinate events and XKB. Several other small fixes were cherry-picked from the new 1.2 branch.
| * Force maximum keycode to be 255Adam Jackson2008-03-08
| | | | | | | | | | | | | | | | | | Even though we don't have keycodes for anything above 127, make sure our map always covers up to 255. This ensures that the keycode range never changes. Spiritually cherry-picked from a9e87f29ccdadebb0742317bb57d66eaaca4b593 and 6db4a9fb84f828f745202c3fddc58d389fca220b.
| * Add absolute coordinate event support.Adam Jackson2008-03-08
| | | | | | | | | | | | | | | | There are two major classes here, touchscreens and touchpads. Touchpads are logically more like relative devices, in that your initial touch should not warp the cursor. So attempt to detect touchpads (via the existence of BTN_TOUCH) and translate absolute events from those devices to relative motion.
| * Nuke the keyboard bell code.Adam Jackson2008-03-08
| | | | | | | | It did nothing, and the core won't invoke it anymore anyway.
| * Re-add XKB support.Adam Jackson2008-03-08
| | | | | | | | | | Almost entirely taken from a patch I did ages ago in bug #3912, with slight cleanups, and with the KcCGST options removed as per evdev 1.x.
| * Parse Option "Path" as well as Option "Device".Adam Jackson2008-03-08
| | | | | | | | | | The hotplug core passes Path not Device. With this change, hotplugging works.
| * Ignore button events above BTN_TASK for now.Adam Jackson2008-03-07
| |
| * Support the new DIX motion history APIAdam Jackson2008-03-07
| | | | | | | | Spiritually cherry-picked from 036b457c1b3f7d2d174da890cb8598d907181f8e
| * remove XFree86LOADER usageDaniel Stone2008-03-07
| | | | | | | | Build evdevModuleData, et al, unconditionally.
| * Unlibcwrap.Adam Jackson2008-03-07
| |
| * Branch for evdev 2.0Adam Jackson2008-03-07
| |
* | Updated manpage to reflect current stateSascha Hlusiak2008-03-03
| | | | | | | | Large parts of the manpage were obsolete and confused users. Current state should reflect all available driver options. Some details are still missing.
* | More accurate error messages on device open fail.Sascha Hlusiak2008-02-20
| | | | | | | | | | The message "cannot open input pEvdev" resulted in a lot of bug reports of confused users that did not supply a device path. Now we tell them when it is missing or print out a reason when device open fails.
* | Let kernel autorepeat pass when set on default valuesSascha Hlusiak2008-02-19
| | | | | | | | | | | | xorg-server won't generate soft autorepeat, when interval/delay are at default of 40/660 (see xkb/ddxCtrls.c: XkbDDXUsesSoftRepeat). When we hit the defaults, we let the kernel autorepeat pass, if we differ, we swallow them all and let the server figure out autorepeat in software.
* | Set repeat_delay and repeat_interval to default valuesSascha Hlusiak2008-02-19
| | | | | | | | Set ctrls->repeat_delay and ctrls->repeat_interval so that keys will be automatically repeated in software.
* | Filter all incoming repeated key eventsSascha Hlusiak2008-02-19
| | | | | | | | | | Not just filter modifiers but every repeated key, because the software implementation will take care of what key to repeat and what not. Otherwise the kernel's autorepeated keys interfere with the software generated ones.
* | Don't inform kernel evdev driver about autorepeat changeSascha Hlusiak2008-02-19
| | | | | | | | | | We do autorepeat in software now. Informing the kernel would change the autorepeat speed on virtual terminals, which is probably not wanted so we leave the settings untouched.
* | Clear EV_ABS_V_USE_TOUCH for devices without a touch buttonVinay Bondhugula2008-01-29
| | | | | | | | | | | | | | VMware's absolute pointing device does not have a pressure/touch button. Although evdev correctly recognizes this, it sets "use_touch" for X and Y axes regardless of the touch button. This patch clears the USE_TOUCH flag for the X and Y axes if the device does not have a touch button.
* | Initialise b_map_data to correct size. Bug #13991Peter Hutterer2008-01-29
| | | | | | | | X.Org Bug 13991 <http://bugs.freedesktop.org/show_bug.cgi?id=13991>
* | Don't flush buttons on init (bug#12630)Julien Cristau2007-12-07
| |
* | Fix one additional typo in evdev.man (axis named BRAKE, not BREAK).xf86-input-evdev-1.2.0Peter Samuelson2007-10-11
| |
* | evdev.man minor fixesPeter Samuelson2007-10-11
| | | | | | | | | | | | | | | | | | | | | | * Use \- (ASCII dash) instead of - (hyphen), where appropriate * Use a roff escape instead of a UTF-8 character. * Consistent use of quote characters. * Consistent use of typographic conventions: italic "n" in place of "<N>" or "integer" or "number", bold for option names, bit maps, etc. Debian bug #446118
* | Add *~ to .gitignore to skip patch/emacs droppingsJames Cloos2007-09-03
| |
* | Options: Accept differently-named optionsDaniel Stone2007-08-07
| | | | | | | | | | Take some differently-named options for device and the XKB options. These should really be better-handled, but eh.
* | Multiple typos in evdev.manBrice Goglin2007-07-26
| | | | | | | | | | Reported by "A. Costa" <agcosta@gis.net> in http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=432063
* | Clean up a little bit the code.Tiago Vignatti2007-06-15
| |
* | Remove inotify things.Tiago Vignatti2007-06-15
| |
* | Update version using PACKAGE_VERSION_*.Zephaniah E. Hull2007-06-13
| |
* | Move the mode option handling to the Init stage.Zephaniah E. Hull2007-06-12
| | | | | | | | (Sorta ugly placement wise, but the other segfaults.)
* | evdev.c:Zephaniah E. Hull2007-06-11
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Set and use the X mode control, letting XInput see it properly. Kill some cruft. evdev.h: Set and use the X mode control, letting XInput see it properly. evdev_axes.c: Set and use the X mode control, letting XInput see it properly. evdev_key.c: Remove some cruft, the bell func can be NULL now, so do so. Don't bother looking for keys above 0xF7 until we can actually do something with them.
* | Handle default remapping around existing button targets better.Zephaniah E. Hull2007-06-06
| |
* | Move MapButtons_t closer to where it's used.Zephaniah E. Hull2007-06-06
| |
* | Wheel mice work again!Zephaniah E. Hull2007-06-06
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | (Old configs don't.) evdev.c: Add EvdevParseMapToButton and EvdevParseMapToButtons to evdev_map_parsers. Add EvdevParseMapOption to search through evdev_map_parsers. Fix up EvdevTokenize to handle the evdev_option_token_t changes. EvdevAxesNew0 after BtnNew0 instead of before now. (This isn't the right fix.) evdev.h: EVDEV_MAXBUTTONS -> BTN_MAX. Redid evdevBtnRec with the new mapping goodness. Removed v_min and v_max from evdevAbsRec. Reworked evdev_option_token_t, no union, no is_chain. If it's a chain, it still has a string, but the chain pointer is set. EvdevParseMapToButton, EvdevParseMapToButtons, and EvdevParseMapOption. evdev_axes.c: Kill off EvdevAxesMapButton, a variant lives in evdev_btn.c now. Changes for the evdev_option_token_t changes. Use EvdevParseMapOption instead of repeating the contents twice. Disable EV_ABS_V_INVERT for the moment. (Better fix maybe needed.) evdev_btn.c: s/Ptr /Rec */g EvdevMapButton and parser. EvdevMapButtons and parser. Nuke EvdevBtnCalcRemap as a whole. Move everything but the alloc to New1 from New0. New mapping code, same guts os the axes mapping code even.
* | Fix REL mode.Zephaniah E. Hull2007-06-05
| | | | | | | | | | | | Make buttons mostly work again. (Short term fix.) NOTE: Wheel to button remapping is still missing!
* | Update the copyright notices for me to include 2007.Zephaniah E. Hull2007-06-05
| | | | | | | | Finish merging copyright notices. (evdev_key.c was missed.)
* | Merge branch 'master' of ↵Zephaniah E. Hull2007-06-05
|\ \ | | | | | | | | | | | | | | | | | | | | | git+ssh://git.freedesktop.org/git/xorg/driver/xf86-input-evdev Conflicts: src/evdev.h
| * | Fix header file #ifdef guard.Kristian Høgsberg2007-06-05
| | |
| * | Merge copyright notices.Kristian Høgsberg2007-06-05
| | |
* | | Alright, this is a really big commit that breaks stuff.Zephaniah E. Hull2007-06-05
|/ / | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | evdev.h: Switch to flags in the abs and rel structs. Add the axes struct, and defines. Rework the abs and rel structs, moving stuff to the axes struct and moving everything to the new mapping handling. Add the structs and function declarations for the new tokenization stuff, parsing stuff, and mapping stuff. evdev.c: Add EvdevTokenize, and the evdev_map_parsers list. evdev_axes.c: Basicly a full rewrite, big, messy. We now use a completely different mapping setup for axes, and mapping to buttons is currently missing. However we now handle ABS_CALIB and ABS_AREA, including rotation in both rel and abs modes. evdev_btn.c: Disable lots of code and break things horribly, we compile but we don't work very well. Fixing this is next on my todo list.
* | Disable and remove the device when a read error occurs.Peter Hutterer2007-04-24
| |
* | Try harder to make unplugging work.Zephaniah E. Hull2007-04-24
| |
* | Kill some more unused bits.Zephaniah E. Hull2007-04-24
| |
* | Handle driver removal cleanly, we now xfree everything we Xcalloc.Zephaniah E. Hull2007-04-24
| | | | | | | | | | NOTE: Someone needs to check for memory leaks, I'd be willing to bet a fair bit that we've got tons in the mouse and xkb code, but.
* | Remove the evil xf86ActivateDevice declaration, don't need it anymore.Zephaniah E. Hull2007-04-24
| | | | | | | | | | | | Fix the crash on device unplug. NOTE: We still don't properly get rid of ourselves.
* | Remove all of evdev_brain.c and all callers.Zephaniah E. Hull2007-04-24
| | | | | | | | | | | | | | | | | | | | | | | | Remove the non-functional 2.4 kernel compiling compatibility. Remove the XInput 1.4+ SendCoreEvents/DontSendCoreEvents define replacements. Remove evdevDriver, some stuff from evdevDevice, all users. NOTE: This WILL BREAK all existing configurations. NOTE: Device read failure segfaults, don't unplug stuff.
* | Updated configure to 1.2.0, no changes yet.Zephaniah E. Hull2007-04-24
| | | | | | | | | | | | | | | | | | | | This branch is the great xserver input-hotplug rewrite, consider it unstable but usable with the 'current' (as of this commit) git xserver. (Once it's more done that is.) NOTE: This branch will _not_ remain source or binary compatable with pre-input-hotplug xservers, you have been warned.
* | Remove driver from list if no device is associated any more.Peter Hutterer2007-04-23
| | | | | | | | | | This effectively stops the driver from hotplugging new devices. Devices have to be added with the dbus hotplugging events.
* | Remove direct call to DEVICE_INIT, ActivateDevice() takes care of that.Peter Hutterer2007-04-19
| |
* | Add some documentation to evdev.c and evdev_brain.c.Peter Hutterer2007-04-19
| |
generated by cgit v1.2.3 (git 2.46.0) at 2025-05-01 16:09:17 +0000