Use a more updated Synaptics driver #2
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This reverts commit 0903d99. Scroll buttons are still present in some modern devices, e.g. the Fujitsu Lifebook E782 and others in the series. Conflicts: include/synaptics.h man/synaptics.man src/synaptics.c
This suite was never really maintained anyway, and it is quite hard to do so anyway. The driver is linked to the server's API too tightly to easily do independent testing. We need to re-implement stubs for the API the driver uses, have to track API changes, etc. Not worth the effort. Signed-off-by: Peter Hutterer <[email protected]>
This reverts commit 3b02e7f. Signed-off-by: Peter Hutterer <[email protected]> Conflicts: man/synaptics.man src/synaptics.c Acked-by: Daniel Stone <[email protected]>
6d47d33 disallows a zero value for horizontal/vertical scroll deltas but the man page wasn't updated. We've added separate toggles to enable/disable scrolling a few years ago, setting the distance to 0 is not recommended. X.Org Bug 75074 <http://bugs.freedesktop.org/show_bug.cgi?id=75074> Signed-off-by: Peter Hutterer <[email protected]>
… area synaptics offers an option to make parts of the touchpad insensitive. This is ie useful to do palm avoidance rather then palm detection (which may be unreliable) by disabling an area of 15% on the right and left side of the touchpad. Currently a motion which has started inside the active area, stops as soon as it moves outside of the active area. If a motion started inside the active area and thus has already generated some move events, this makes no sense. If the user moves outside of the active area in this case, this is very likely because the user wants to continue the motion. This commit allows such motions to continue normally. I would like to thank Juerd Waalboer for the basic idea, some coding and lots of testing for this fix. Cc: Juerd Waalboer <[email protected]> Reported-by: Juerd Waalboer <[email protected]> Tested-by: Juerd Waalboer <[email protected]> Signed-off-by: Hans de Goede <[email protected]> Reviewed-by: Peter Hutterer <[email protected]> Signed-off-by: Peter Hutterer <[email protected]>
New generation of laptops with trackstick do not have physical buttons associated with the trackstick, but instead rely on software buttons at the top of the clickpad. Adding a secondary software button area for this purpose. As we're likely detecting the devices that need it based on udev tags and MatchTag configuration items, this area doesn't need to be exposed through properties. So static configuration is fine. Signed-off-by: Benjamin Tissoires <[email protected]> Reviewed-by: Peter Hutterer <[email protected]> [couple of man-page additions and rewrites] Signed-off-by: Peter Hutterer <[email protected]>
Signed-off-by: Keith Packard <[email protected]> Signed-off-by: Peter Hutterer <[email protected]>
If we try to grab the evdev device before we've set the new file descriptor, libevdev_grab returns -EFAULT, which causes DeviceOn to fail. Signed-off-by: Keith Packard <[email protected]> Signed-off-by: Peter Hutterer <[email protected]>
…licks On a new set of laptops like the Lenovo T440 the trackstick does not have physical buttons. Instead, the touchpad's top edge is supposed to acts software button area. To avoid spurious cursor jumps when the trackstick is in use and the finger is resting on the touchpad, add another mode that disables motion events. Enabled by syndaemon with -t click-only, the default stays unchanged. No specific integration with the traditional disable-while-typing is needed. On such touchpads, disabling motion events is sufficient to avoid spurious events and we don't want to stop HW buttons to send events. Note that this only adds the new state to the driver and to syndaemon, there is nothing hooked up otherwise to actually monitor the trackstick. Special note for syndaemon: optional arguments are a GNU extension, so work around it by messing with an optstring starting with ":" which allows us to manually parse the options. Original version of this patch by John Pham <[email protected]> Signed-off-by: Peter Hutterer <[email protected]> Reviewed-by: Hans de Goede <[email protected]> Signed-off-by: Hans de Goede <[email protected]> Signed-off-by: Peter Hutterer <[email protected]>
While at it also move the enum for the soft button edges out of is_inside_button_area() so that it can be used elsewhere too. Signed-off-by: Hans de Goede <[email protected]> Signed-off-by: Peter Hutterer <[email protected]>
We only use it to store button state which we already have in priv->lastButtons. While at it also properly indent the code block checking the various soft button areas. Signed-off-by: Hans de Goede <[email protected]> Signed-off-by: Peter Hutterer <[email protected]>
Unless the motion has started outside the soft-button area. Note that we must start reporting motions regardless of whether we think we're in the button area or not as soon as we've switched to using cumulative coordinates, since then the coordinates are no longer absolute. This fixes the reporting of unintended motion just before a click in a soft button area which sometimes causes mis-clicks. Signed-off-by: Hans de Goede <[email protected]> Signed-off-by: Peter Hutterer <[email protected]>
This fixes my felipejfc#1 anoyance with clickpads, where 2 out of 3 clicks turn into a click + drag unless I hold my finger really really still. Signed-off-by: Hans de Goede <[email protected]> Replaced property with a hardcoded 100ms. This is not something that we should expose as property, we should find a delay that works best and live with it. Signed-off-by: Peter Hutterer <[email protected]>
When a button click and new coordinates get reported in one go we sync the cumulative coordinates to the old x and y, rather then the newly reported ones. This keeping of the old coordinates causes the following issue: -touch the touchpad in its right click area -let go of the touchpad -rapidly click in the left click area (or middle area), so that the new location and the click get reported in one syn (may require some practicing with evemu-record to reproduce) -the driver registers the click as a right click because it uses the old coordinates from the cumulative coordinates to determine the click location Signed-off-by: Hans de Goede <[email protected]> Signed-off-by: Peter Hutterer <[email protected]>
…lick It is possible for a click to get reported before any related touch events get reported, here is the relevant part of an evemu-record session on a T440s: E: 3.985585 0000 0000 0000 # ------------ SYN_REPORT (0) ---------- E: 3.997419 0003 0039 -001 # EV_ABS / ABS_MT_TRACKING_ID -1 E: 3.997419 0001 014a 0000 # EV_KEY / BTN_TOUCH 0 E: 3.997419 0003 0018 0000 # EV_ABS / ABS_PRESSURE 0 E: 3.997419 0001 0145 0000 # EV_KEY / BTN_TOOL_FINGER 0 E: 3.997419 0000 0000 0000 # ------------ SYN_REPORT (0) ---------- E: 5.117881 0001 0110 0001 # EV_KEY / BTN_LEFT 1 E: 5.117881 0000 0000 0000 # ------------ SYN_REPORT (0) ---------- E: 5.133422 0003 0039 0187 # EV_ABS / ABS_MT_TRACKING_ID 187 E: 5.133422 0003 0035 3098 # EV_ABS / ABS_MT_POSITION_X 3098 E: 5.133422 0003 0036 3282 # EV_ABS / ABS_MT_POSITION_Y 3282 E: 5.133422 0003 003a 0046 # EV_ABS / ABS_MT_PRESSURE 46 E: 5.133422 0001 014a 0001 # EV_KEY / BTN_TOUCH 1 E: 5.133422 0003 0000 3102 # EV_ABS / ABS_X 3102 E: 5.133422 0003 0001 3282 # EV_ABS / ABS_Y 3282 E: 5.133422 0003 0018 0046 # EV_ABS / ABS_PRESSURE 46 E: 5.133422 0001 0145 0001 # EV_KEY / BTN_TOOL_FINGER 1 E: 5.133422 0000 0000 0000 # ------------ SYN_REPORT (0) ---------- Notice the BTN_LEFT event all by itself! If this happens, it may lead to the following problem scenario: -touch the touchpad in its right click area -let go of the touchpad -rapidly click in the middle area, so that BTN_LEFT gets reported before the new coordinates (such as seen in the trace above, this may require some practicing with evemu-record to reproduce) -the driver registers the click as a right click because it uses the old coordinates from the cumulative coordinates to determine the click location This commit fixes this by: 1) Resetting the cumulative coordinates not only when no button is pressed, but also when there is no finger touching the touchpad, so that when we do get a touch the cumulative coordinates start at the right place 2) Delaying processing the BTN_LEFT down transition if there is no finger touching the touchpad This approach has one downside, if we wrongly identify a touchpad as a clickpad, then the left button won't work unless the user touches the touchpad while clicking the left button. If we want we can fix this by doing something like this: 1) Making update_hw_button_state return a delay; and 2) Tracking that we've delayed BTN_LEFT down transition processing; and 3) When we've delayed BTN_LEFT down transition return a small delay value; and 4) If when we're called again we still don't have a finger down, just treat the click as a BTN_LEFT But this is not worth the trouble IMHO, the proper thing to do in this scenario is to fix the mis-identification of the touchpad as a clickpad. Signed-off-by: Hans de Goede <[email protected]> Signed-off-by: Peter Hutterer <[email protected]>
leftover from an earlier revision Signed-off-by: Peter Hutterer <[email protected]>
Signed-off-by: Hans de Goede <[email protected]> Signed-off-by: Peter Hutterer <[email protected]>
This was originally intended as a fixed xorg.conf option only (and still largely is seen as such). Secondary software button are required only on a specific series of touchpads and should be pre-configured by the system and/or the distribution. As such, the property will not be initialized if it is not set in the xorg.conf and will thus not respond to runtime changes. Exposing the property in this way gives clients a chance of detecting if a top software button area is present and thus adjust their behaviour accordingly. Signed-off-by: Peter Hutterer <[email protected]> Reviewed-by: Hans de Goede <[email protected]>
This was supposed to emulate a SYN_REPORT event so that the upper layers process what's in the queue. Signed-off-by: Peter Hutterer <[email protected]> Reviewed-by: Hans de Goede <[email protected]> Reviewed-by: Benjamin Tissoires <[email protected]>
The kernel guarantees slots start at 0 Signed-off-by: Peter Hutterer <[email protected]> Reviewed-by: Hans de Goede <[email protected]> Reviewed-by: Benjamin Tissoires <[email protected]>
When checking the device don't open a new mtdev instance, use the existing libevdev struct. Signed-off-by: Peter Hutterer <[email protected]> Reviewed-by: Hans de Goede <[email protected]> Reviewed-by: Benjamin Tissoires <[email protected]>
All kernel touchpad devices now support slots, there isn't really a need to support protocol A devices in synaptics. If such devices exist, we just treat them as non-multitouch devices. Signed-off-by: Peter Hutterer <[email protected]> Reviewed-by: Hans de Goede <[email protected]> Reviewed-by: Benjamin Tissoires <[email protected]>
Signed-off-by: Stephen Chandler Paul <[email protected]> Signed-off-by: Peter Hutterer <[email protected]>
Clicking in the top soft button area causes the trackpad to begin registering motion, even if the finger never leaves the top soft button area. We don't want this kind of behavior for the top soft button area, since it makes clicking and dragging items much more difficult when using a pointing stick. Signed-off-by: Stephen Chandler Paul <[email protected]> Signed-off-by: Peter Hutterer <[email protected]>
…button clicks" This third state is not needed, the behaviour of the touchpad driver is now good enough to not need an external syndaemon instance to toggle this third state. This reverts commit eea7335. Conflicts: man/synaptics.man src/synaptics.c Signed-off-by: Peter Hutterer <[email protected]>
Signed-off-by: Peter Hutterer <[email protected]>
Bad fdi file, type="string" is missing and it wouldn't merge properly. This reverts commit a35b2d6.
…ersion) Signed-off-by: Peter Hutterer <[email protected]>
Signed-off-by: Peter Hutterer <[email protected]>
Enabling clicks in off mode also allows for the new Lenovo *40 series to use the top software buttons while the touchpad is disabled. This benefits those that usually disable touchpads altogether but still need the buttons for the trackstick. This changes existing behaviour, but TouchpadOff was always intended to stop erroneous events while typing. Physical button presses are hard to trigger accidentally. On the touchpads that TouchpadOff concept was originally designed for the buttons are nowhere near the keyboard and are physically separated from the touchpad anyway. On Clickpads, triggering a physical click requires more force than accidentally touching the surface. https://bugs.freedesktop.org/show_bug.cgi?id=76156 Signed-off-by: Peter Hutterer <[email protected]> Reviewed-by: Hans de Goede <[email protected]>
Signed-off-by: Peter Hutterer <[email protected]>
This was required when we started supporting hotplugging to avoid duplicate events. These days the drawback of not being able to record events in the case of a bug is significant. Check the configuration source on init. If the device was hotplugged through a a server config backend, disable the grab. If the device was statically configured through an xorg.conf then leave the default grab enabled to avoid a duplicate device. Signed-off-by: Peter Hutterer <[email protected]> Reviewed-by: Hans de Goede <[email protected]>
last_mt_vals_slot is only used in one location and there we can just use cur_slot Signed-off-by: Peter Hutterer <[email protected]> Reviewed-by: Hans de Goede <[email protected]>
Once the sync finishes, we get -EAGAIN. This only indicates the sync is done, but some events may still be waiting in the pipe for us to read. We must read those now, otherwise select may not trigger on further data. Signed-off-by: Peter Hutterer <[email protected]> Reviewed-by: Hans de Goede <[email protected]>
This is a bit problematic: libevdev only has one global log handler. So if we have another driver use libevdev, we'll either overwrite that handler or get overwritten, whichever comes first. So we need to re-set the handler every time we get an event to make sure we log through our handler. Likewise, if we ever drop the device, we need to unset the log handler back to NULL because we may unload the module and our handler may disappear. Use the lowest logging priority, let the server filter based on the verbosity level instead. Signed-off-by: Peter Hutterer <[email protected]> Reviewed-by: Hans de Goede <[email protected]>
Everytime I look at this I get confused about OPEN_EMPTY vs EMPTY. Let's fix that. Signed-off-by: Peter Hutterer <[email protected]> Reviewed-by: Hans de Goede <[email protected]>
And expand DMI strings to more precise matches Signed-off-by: Peter Hutterer <[email protected]> Reviewed-by: Hans de Goede <[email protected]>
When trying to do a 3 fingerclick on a touchpad which only tracks 2 touches, this may register as a 3 or 2 fingerclick depending on the order in which the touchpad detects the fingers. If the 2 outer fingers of the 3 get seen first, then the 2 touches will be too far apart for the heuristic to see them as being close together, and the click gets counted as a 2 finger click. A user will likely never do a 2 finger click with a 3th finger resting somewhere else on the pad, where-as the above misdetection of the clicks is a real issue, so simply always count a click with trippletap set as a 3 finger click on pads which track less then 3 touches. https://bugzilla.redhat.com/show_bug.cgi?id=1086218 Signed-off-by: Hans de Goede <[email protected]> Reviewed-by: Peter Hutterer <[email protected]> Signed-off-by: Peter Hutterer <[email protected]>
Signed-off-by: Peter Hutterer <[email protected]>
Add a HasSecondaryButtons boolean config option which defaults to true for devices with the INPUT_PROP_TOPBUTTONPAD and false for all other devices. Only parse the SecondarySoftButtonAreas when this option is true, effectively disabling the top buttons when it is false. Likewise, only initialize the SecondarySoftButtonAreas property if we enable support for it. This means that it is now safe to always set a SecondarySoftButtonAreas default in 50-synaptics.conf, and that he section which was intended for use with future pnp-id matching can be dropped, as that is now all handled in the kernel. While at also remove the comment about disabling the bottom edge area, as that is now done automatically. Signed-off-by: Hans de Goede <[email protected]> Signed-off-by: Peter Hutterer <[email protected]> Reviewed-by: Hans de Goede <[email protected]>
Some Macbooks are being tagged as MODEL_UNIBODY_MACBOOKs when they should not be. This causes the default sensitivity to be very low for them, making the touchpad almost unusable. This change puts those devices into the correct bucket again. Signed-off-by: Clinton Sprain <[email protected]> Reviewed-by: Peter Hutterer <[email protected]> Signed-off-by: Peter Hutterer <[email protected]>
Signed-off-by: Peter Hutterer <[email protected]>
Rely on INPUT_PROP_TOP_BUTTONPAD and default button areas as well. Signed-off-by: Peter Hutterer <[email protected]>
Signed-off-by: Peter Hutterer <[email protected]>
We had reports that the top software button area is hard to hit for those using the trackpoint and clicking the buttons with their thumb. Analysis of event recordings (3 different people) for left, right and middle clicks shows that there is a significant amount of events up to about 10mm (with outliers up to 12mm) from the top of the touchpad. That maps to 15%. Interestingly, the middle button does not seem to need this, presumably the haptic feedback of the little dots sticking out from the surface make hitting the button easier. Its size is increased to 15% anyway, for simplicity and because a sample set of 3 is too small to be definitive about this. Signed-off-by: Peter Hutterer <[email protected]> Reviewed-by: Hans de Goede <[email protected]> (cherry picked from commit 730101223432f60397c61f74a5e6789b3ee34ecd)
When two fingers are used, the coordinates of only one of them is taken into account. This can lead to sudden variations of the absolute coordinates when two-fingers taps are performed if the finger considered changes. Take into account coordinates variations to prevent unwanted taps only if the number of fingers doesn't change. Reviewed-by: Peter Hutterer <[email protected]> Signed-off-by: Peter Hutterer <[email protected]> (cherry picked from commit 7d0ff39519e4d3760722b914883bee276035061c)
In file included from /usr/include/string.h:634:0,
from /usr/include/xorg/os.h:53,
from /usr/include/xorg/misc.h:115,
from /usr/include/xorg/xf86str.h:37,
from /usr/include/xorg/xf86Xinput.h:54,
from synproto.h:36,
from synproto.c:24:
/usr/include/xorg/os.h:579:1: error: expected identifier or '(' before '__extension__'
strndup(const char *str, size_t n);
See http://lists.freedesktop.org/archives/xorg-devel/2014-July/043070.html
Signed-off-by: Peter Hutterer <[email protected]>
(cherry picked from commit 96e60a4ea242d2decf109835981ae186cc36f642)
Default on evdev devices is CLOCK_REALTIME. If that clock falls behind the server's CLOCK_MONOTONIC, motion after a clickpad click may be delayed by the difference in the clocks. In detail: When the timer func is triggered, GetTimeInMillis() which is CLOCK_MONOTONIC, is stored as hwState->millis. The eventcomm backend uses struct input_event time (CLOCK_REALTIME). When we read events from the device, if the evdev time is less than the server time, the fix for (#48777) sets the current event time to hwState->millis. Until the evdev time overtakes that stored time, all events have the hwState->millis time. If during that time a clickpad triggers a physical click, clickpad_click_millis is set to hwState->millis + the ignore-motion timeout. Thus, all motion is ignored until the event time overtakes that stored time. The whole issue is further enhanced by us unconditionally setting the timer func if we get any events, which is a separate issue anyway. Signed-off-by: Peter Hutterer <[email protected]> Reviewed-by: Hans de Goede <[email protected]> (cherry picked from commit 90d19302306f49722e210227b2fb5161e6f51880)
And warn when we run out of labels. Signed-off-by: Peter Hutterer <[email protected]> Reviewed-by: Keith Packard <[email protected]> (cherry picked from commit d239f831f17ccf5468f5dc6b2f199a9c1f6e35af)
open_slots holds the slot index, resetting it to 0 is a bad idea. And make sure that we do reset after DEVICE_INIT. We already do so on DEVICE_CLOSE, but after the first DEVICE_ON the data could still be random. Signed-off-by: Peter Hutterer <[email protected]> Reviewed-by: Keith Packard <[email protected]> (cherry picked from commit afbbcfa10eb3a2295823720907f35bb59972dd82)
Default resolution is 1, don't allow setting 0 to avoid divisions by 0 or just general weirdness. Signed-off-by: Peter Hutterer <[email protected]> (cherry picked from commit 049611bd7f04e285909c55807478306cce83385f)
Signed-off-by: Peter Hutterer <[email protected]>
…once again Merged and solved all conflicts
|
have you tested it? |
|
I have successfully tested it on Ubuntu 15.04
|
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This fixes #1