Package Details: xf86-input-mtrack-git 0.4.0.r0.gb9f274d-1

Git Clone URL: (read-only)
Package Base: xf86-input-mtrack-git
Description: A multitouch X driver using the kernel MT protocol
Upstream URL:
Licenses: GPL
Conflicts: xf86-input-mtrack, xf86-input-synaptics
Provides: xf86-input-mtrack
Submitter: mikezackles
Maintainer: mikezackles
Last Packager: mikezackles
Votes: 30
Popularity: 0.411176
First Submitted: 2011-04-26 16:31
Last Updated: 2016-10-28 01:54

Latest Comments

mikezackles commented on 2016-10-28 01:56

.SRCINFO has been updated. Apologies for the delayed response gsora.

gsora commented on 2016-03-04 15:17

The Upstream URL is not correct, please modify it accordingly.

viniavila commented on 2016-02-23 16:48

I don't know what to do.

I've installed this package to use mtrack instead of synaptics in my bcm5914 trackpad (Macbook Air 2012) But Xorg insists not to load the mtrack module!

[vinicius@MacBookAir ~]$ cat /var/log/Xorg.0.log | grep "(EE)"
(WW) warning, (EE) error, (NI) not implemented, (??) unknown.
[ 5.223] (EE) Failed to load module "fbdev" (module does not exist, 0)
[ 5.223] (EE) Failed to load module "vesa" (module does not exist, 0)
[ 7.140] (EE) mtrack: cannot configure device
[ 7.140] (EE) Couldn't init device "bcm5974"

I already added my user to input group which is useless since this problem is recorded BEFORE logon in lightdm (I'm using lightdm-gtk-greeter). I tried many settings within 10-mtrack.conf found in internet but none worked. Few days ago in another arch intallation I could make this work flawlessly so I can't figure out what am i doing wrong... Some help wold be appreciated...

mikezackles commented on 2015-10-24 13:52

Thanks epinephrine. I've switched over to p2rkw's fork. Just let me know if anyone has issues, and I'll switch it back.

sebstar commented on 2015-10-22 21:56

Looking at the fork network (, I noticed that apparently p2rkw continues development. I got in touch with him and he confirmed that BlueDragonX doesn't work on the project at the moment and that he continues development.

So maybe we should change the upstream and source url to

mikeokner commented on 2015-08-17 23:03

I'm running into some issues trying to use this driver on a MacBook Pro 11,5 (mid-2015 with Force Trackpad). I tried using both mtrack and synaptics with the latest available kernel via pacman (4.1.5-1) to no avail. Both resulted in an unresponsive cursor -- clicking worked, though. Only the evdev driver worked, but then I didn't have two-finger scrolling or right clicking.

I updated to linux-mainline available via AUR (4.2rc7-1) after seeing a note in the Macbook wiki page. The synaptics driver does work now under 4.2, but the mtrack driver still results in an unresponsive cursor. I'd love to get mtrack working if possible.

mikezackles commented on 2015-07-05 06:13

Thanks for the feedback all. I've added a post_install message that points out the location of the config file and suggests adding yourself to the input group. I've also added xf86-input-synaptics as a conflict.

boterock commented on 2015-07-05 01:55

It should be noted somewhere that the config is in /usr/share/X11/xorg.conf.d/...
I spent my last four days followin internet docs that referred the config file in /etc/X11/xorg.conf.d, and I only realized the config file is in /usr/share because in the comments there is a link to a git diff where I found the config URL

liskevich commented on 2015-05-22 15:18

@hobarrera - I had same issue all of a sudden on my Macbook Pro. You need to add yourself to the input user group so:
sudo gpasswd -a <Yourusername> input

Then just reboot and enjoy

hobarrera commented on 2015-05-14 00:37

Nor this package, nor xf86-input-mtrack worked for me on a 2013 macbook air. No errors on Xorg.0.log. Does anyone else have similar experience on this same hardware?

jyantis commented on 2015-04-20 14:37

Just tested it. Seems to work fine though I did have to remove xf86-input-synaptics to get it to work properly.

mikezackles commented on 2015-02-12 13:08

Updated. Thanks saimn. I can't test this at the moment, so just let me know if there are issues.

saimn commented on 2015-02-12 10:02

There is an extra 'z' at the end of the backup line, which cause a warning.
Also in pkgver() you must use cd "$srcdir/$_pkgname" otherwise it seems that pkgver is not updated.

mikezackles commented on 2015-01-11 16:08

Thanks for your improvements James! I've updated to your PKGBUILD.

I'm not using this package currently, so just let me know if you'd like to become maintainer, and I'll orphan it.

jamesan commented on 2015-01-10 08:51

Left this out -- you can find the diff representing my changes here:

jamesan commented on 2015-01-10 07:42

The PKGBUILD tracks the latest git revision but the latest release version string is wrong. The pkgver() function misses the latest release version as its git tag is not annotated. Adding the '--tags' option to git-describe fetches the correct latest release version string, 0.3.0:
git describe --long --tags

While I was looking at this PKGBUILD, I took the liberty to:

- attribute credits in opening comments;
- parameterise the package name using $_pkgname wherever xf86-input-mtrack is mentioned; and
- moved the 10-mtrack.conf file from /etc to /usr/share, where most other input driver conf files reside. The /etc folder should be left for conf files not part of packages (user-specified overrides).

You can find my modified PKGBUILD here:

mikezackles commented on 2013-10-20 14:42

Thanks jsteel. The base-devel dependencies have been removed.

jsteel commented on 2013-10-20 10:59

A lot of makedepends on this are included in base-devel so you do not need to specify them.

mikezackles commented on 2013-05-01 22:09

Looks great, thanks Nephyrin! And thanks for making me aware of the cool new pacman feature :)

Nephyrin commented on 2013-05-01 21:29

Here's a version tweaked for pacman 4.1+'s new git/VCS PKGBUILD format:

mikezackles commented on 2013-01-13 19:31

Fixed. Thanks, ReedWood. Also, it looks like the test target is compiling again (probably has been for a while), so I switched back to the regular make/make install targets. Let me know if anyone has issues.

ReedWood commented on 2013-01-13 10:50

Folder m4 in $srcdir/$_gitname-build seems to be missing, leading autoreconf --install to fail. Adding
mkdir m4 directly infront of the call to autoreconf (see patch further down) solves the issue for me.

*** PKGBUILDOrig 2013-01-13 11:44:07.376574592 +0100
--- PKGBUILD 2013-01-13 11:44:36.612759508 +0100
*** 1,5 ****
! pkgver=20121103
pkgdesc="A multitouch X driver using the kernel MT protocol"
arch=('i686' 'x86_64')
--- 1,5 ----
! pkgver=20130113
pkgdesc="A multitouch X driver using the kernel MT protocol"
arch=('i686' 'x86_64')
*** 34,39 ****
--- 34,40 ----
git clone "$srcdir/$_gitname" "$srcdir/$_gitname-build"
cd "$srcdir/$_gitname-build"

+ mkdir m4
autoreconf --install
./configure --prefix=/usr
make Makefile config.h

mikezackles commented on 2012-11-03 13:37

libxss added. Thanks MadFish.

Anonymous comment on 2012-11-03 12:13

libxss should be added to dependencies - otherwise compile error on missing X11/extensions/saver.h

mikezackles commented on 2012-10-13 00:20

No worries :) Glad it's working, and thanks again for letting me know.

frigaut commented on 2012-10-13 00:13

(yes, I contacted him directly.)
Thanks for the prompt reaction and for adding the dependency. I didn't realize that this might have been associated to the fact I am using xorg from testing.

mikezackles commented on 2012-10-11 21:54

OK, I've added resourceproto as a dependency, although it doesn't seem to be necessary using the current stable version of xorg. I don't enable the testing repo on this machine, so I can't directly support building against packages that are still in testing. Thanks for tracking down the issue.

frigaut commented on 2012-10-11 21:28

ok, the bug was not upstream: you are missing the resourceproto package as a dependency in the PKGBUILD.
With it, it builds fine.

mikezackles commented on 2012-10-11 15:52

Thanks, frigaut. I'm not seeing your bug report on github -- did you contact him directly? Please let me know if you hear anything further.

frigaut commented on 2012-10-11 09:50

I have sent a bug report to bluedragonX.

frigaut commented on 2012-10-11 09:43


the built fails with the latest xorg:

make: `Makefile' is up to date.
CC capabilities.lo
In file included from /usr/include/xorg/globals.h:8:0,
from /usr/include/xorg/opaque.h:34,
from /usr/include/xorg/xf86_OSproc.h:127,
from ./include/common.h:32,
from ./include/capabilities.h:25,
from ./src/capabilities.c:22:
/usr/include/xorg/extinit.h:107:38: fatal error: X11/extensions/XResproto.h: No such file or directory
compilation terminated.
make: *** [capabilities.lo] Error 1
==> ERROR: A failure occurred in build().
The build failed.

I'm running the xorg from testing:
525:~ $ pacman -Qs xorg-server
local/xorg-server 1.13.0-2 (xorg)
Xorg X server
local/xorg-server-common 1.13.0-2
Xorg server common files
local/xorg-server-devel 1.13.0-2
Development files for the X.Org X server

This was working fine for the previous 1.12 xorg version, but I had to try to rebuilt due to an incompatiblities as shown in Xorg.log:
[ 13.833] (II) LoadModule: "mtrack"
[ 13.833] (II) Loading /usr/lib/xorg/modules/input/
[ 13.834] (II) Module mtrack: vendor="X.Org Foundation"
[ 13.834] compiled for 1.12.3, module version = 0.1.0
[ 13.834] Module class: X.Org XInput Driver
[ 13.834] ABI class: X.Org XInput driver, version 16.0
[ 13.834] (EE) module ABI major version (16) doesn't match the server's version (18)
[ 13.834] (II) UnloadModule: "mtrack"
[ 13.834] (II) Unloading mtrack
[ 13.834] (EE) Failed to load module "mtrack" (module requirement mismatch, 0)
[ 13.834] (EE) No input driver matching `mtrack'

will check if BlueDragonX has come up with a new version...

Anonymous comment on 2012-05-14 11:09

Is this the right one to use for testing XInput 2.2 with an Apple bluetooth trackpad?

mikezackles commented on 2012-02-25 17:52

You're welcome!

Anonymous comment on 2012-02-25 17:44

Hi Mike, just gave it a test and it's building / installing fine now.
Thanks a bunch for your efforts.

mikezackles commented on 2012-02-24 17:15

Thanks paulez. I'm not having that problem, but I went ahead and switched it to https anyway.

paulez commented on 2012-02-24 17:07

Git repo url should be using https protocol, if not it returns an HTTP 403 error while cloning the repo.

mikezackles commented on 2012-02-20 14:05

OK, hopefully it should work now. Looks like make install was building the test target during the package() step.

mikezackles commented on 2012-02-18 21:58

Drat. Thanks for letting me know. Worst case I'll get to it when I get home Monday or Tuesday. (I only have my phone right now).

Anonymous comment on 2012-02-18 21:23

I can confirm that the lastest change to the PKGBUILD is still building the mtrack-test which is failing with the testing repo server. With that said, commenting out the various mtrack-test elements from the Makefile does allow the package to build.

mikezackles commented on 2012-02-16 20:20

I see. I don't think I'm in a position to enable the testing repo on this machine right now, but I just uploaded a version that shouldn't build mtrack-test. Please let me know if there are issues. I'll investigate a more robust fix once the problem code hits extra.

Nephyrin commented on 2012-02-16 18:49

I'm using xorg-server-devel from the testing repo, which is probably the cause.

mikezackles commented on 2012-02-16 04:35

Just had a chance to glance at this. I added a backup section including 10-mtrack.conf, but I'm not having a problem building mtrack-test.c (xorg-server-devel 1.11.4-1). Thanks for the feedback -- please let me know if I'm missing something.

mikezackles commented on 2012-02-15 21:52

Ah, got it. Sorry about the conf file. I'll get it sorted when I get back.

Nephyrin commented on 2012-02-15 21:22

Also - 10-mtrack.conf should be marked as a config file so updating the package doesn't wipe out my custom config :(

Nephyrin commented on 2012-02-15 21:19

This compiles if you remove the 'test' target from the makefile, which isn't required for normal operation. The newest broke the test (header define conflicts)

mikezackles commented on 2012-02-15 13:33

Unfortunately I'm on vacation right now, so I may not be able to take a look until next week.

nbryskin commented on 2012-02-15 05:50

This package fails to build with such error:
CC mtrack_test-mtrack-test.o
./tools/mtrack-test.c:35:5: error: conflicting types for ‘xf86SetIntOption’
/usr/include/xorg/xf86Opt.h:73:22: note: previous declaration of ‘xf86SetIntOption’ was here
./tools/mtrack-test.c:40:5: error: conflicting types for ‘xf86SetBoolOption’
/usr/include/xorg/xf86Opt.h:76:22: note: previous declaration of ‘xf86SetBoolOption’ was here
./tools/mtrack-test.c:45:8: error: conflicting types for ‘xf86SetRealOption’
/usr/include/xorg/xf86Opt.h:74:25: note: previous declaration of ‘xf86SetRealOption’ was here
make[1]: *** [mtrack_test-mtrack-test.o] Error 1
make[1]: Leaving directory `/tmp/yaourt-tmp-nbryskin/aur-xf86-input-mtrack-git/src/mtrack-build'
make: *** [all] Error 2

mikezackles commented on 2011-11-08 13:33

Done. Thanks corion.

corion commented on 2011-11-08 09:31

I found that it needs pkg-config, so maybe it is worth adding it to makedepends.

mikezackles commented on 2011-09-12 10:23

I'm not having any trouble with the build on my machine, but I've added libtool to makedepends in case that was the problem. Please let me know if there are still issues.

mikezackles commented on 2011-09-10 16:13

Thanks Hunner. I'm out for the day, but I'll have a look tomorrow.

Anonymous comment on 2011-09-10 15:33

Looks like there's something wrong with mtrack's autotools too. error: possibly undefined macro: AC_DISABLE_STATIC
If this token and others are legitimate, please use m4_pattern_allow.
See the Autoconf documentation. error: possibly undefined macro: AC_PROG_LIBTOOL
autoreconf: /usr/bin/autoconf failed with exit status: 1

mikezackles commented on 2011-06-06 01:39

Thanks, frigaut. Looks like the author switched to autotools. I haven't installed the latest yet on my system, but I can verify that the build works for me now.

Sorry about the delayed response -- I was lazy about checking email this weekend.

frigaut commented on 2011-06-04 02:43

there seems to be an issue...

==> GIT checkout done or server timeout
==> Starting make...
Cloning into /home/frigaut/abs/aur/xf86-input-mtrack-git/src/mtrack-build...
make: *** No targets specified and no makefile found. Stop.
==> ERROR: A failure occurred in build().
poliahu:xf86-input-mtrack-git $