Alps touchpad is recognized but synaptics clients and scrolling do not work

Also, I lost my ability to scroll with the edge of the touchpad when the kernel was fixed.

Mostly same here on my HP Mini 311. However my device isn't even recognized as a touchpad; it's recognized as a generic mouse, thus there's no "Touchpad" tab in mouse preferences. When I try to activate syndaemon it says that there's no synaptics device found, so there's no way to disable tapping while typing... highly annoying.

Will attach my relevant files since my circumstance seems slightly different from above.

Confirming that this problem also exists in Ubuntu 10.04 Karmic beta 1 (ie, touchpad is recognized as a generic mouse device) on HP Mini 311.

Topher, take a look at Bug 545307. There you will see that Chase recently
developed a patch for the Kernel for recognizing Alps touchpad type listed.
It worked for me. For me, now the tabs show up for me, but the settings
don't adjust. I am not sure if my problem with that is unique to me
(because we were messing around trying to fix it earlier) or whether every
one will have it. But if you have the same type of touchpad and can find the
patch and install it, it is worth a try

On Tue, Mar 30, 2010 at 8:01 PM, Topher <email address hidden> wrote:

> Confirming that this problem also exists in Ubuntu 10.04 Karmic beta 1
> (ie, touchpad is recognized as a generic mouse device) on HP Mini 311.
>
> --
> Alps touchpad is recognized but neither gsynaptics nor mouse preferences
> work
> https://bugs.launchpad.net/bugs/550625
> You received this bug notification because you are a direct subscriber
> of the bug.
>
> Status in “xserver-xorg-input-synaptics” package in Ubuntu: New
>
> Bug description:
> Binary package hint: xserver-xorg-input-synaptics
>
> I have Ubuntu 9.10 and an Alps touchpad (E7: 73 02 64). Chase fixed the
> kernel so that it was recognized and, indeed, the touchpad shows up in both
> gsynaptics and system>preferences>mouse, but when I actually try to adjust
> the settings nothing changes. My goal is to disable the tap-to-click feature
> or disable the touchpad when typing. Chase suggests it may be something
> related to the driver and to open another bug.
>
> ProblemType: Bug
> Architecture: i386
> Date: Sun Mar 28 22:02:25 2010
> DistroRelease: Ubuntu 9.10
> InstallationMedia: Ubuntu 9.10 "Karmic Koala" - Release i386 (20091028.5)
> MachineType: Hewlett-Packard HP Pavilion dm3 Notebook PC
> Package: xserver-xorg-input-synaptics 1.1.2-1ubuntu7
> ProcCmdLine: BOOT_IMAGE=/boot/vmlinuz-2.6.31-21-generic
> root=UUID=f1e2e19e-a0f8-407d-9533-c5ba95a8be50 ro quiet splash
> ProcEnviron:
> LANG=en_US.UTF-8
> SHELL=/bin/bash
> ProcVersionSignature: Ubuntu 2.6.31-21.58~alps4-generic
> RelatedPackageVersions:
> xserver-xorg 1:7.4+3ubuntu10
> libgl1-mesa-glx 7.6.0-1ubuntu4
> libdrm2 2.4.14-1ubuntu1
> xserver-xorg-video-intel 2:2.9.0-1ubuntu2.1
> xserver-xorg-video-ati 1:6.12.99+git20090929.7968e1fb-0ubuntu1
> SourcePackage: xserver-xorg-input-synaptics
> Uname: Linux 2.6.31-21-generic i686
> XorgConf:
> Section "InputDevice"
> Identifier "Macintosh mouse button emulation"
> Driver "synaptics"
> Option "SHMConfig" "true"
> EndSection
> dmi.bios.date: 12/04/2009
> dmi.bios.vendor: Insyde Corp.
> dmi.bios.version: F.22
> dmi.board.asset.tag: Base Board Asset Tag
> dmi.board.name: 3656
> dmi.board.vendor: Hewlett-Packard
> dmi.board.version: 44.1D
> dmi.chassis.type: 10
> dmi.chassis.vendor: Hewlett-Packard
> dmi.chassis.version: N/A
> dmi.modalias:
> dmi:bvnInsydeCorp.:bvrF.22:bd12/04/2009:svnHewlett-Packard:pnHPPaviliondm3NotebookPC:pvr0486110000241F00110320000:rvnHewlett-Packard:rn3656:rvr44.1D:cvnHewlett-Packard:ct10:cvrN/A:
> dmi.product.name: HP Pavilion dm3 Notebook PC
> dmi.product.version: 0486110000241F00110320000
> dmi.sys.vendor: Hewlett-Packard
> fglrx: Not loaded
> system:
> distro: Ubunt...

Thanks for the tip Mary. I installed the patch. No luck! No touchpad tab in mouse prefs, no synaptics driver installed. Perhaps our hardware is a little different. Either way, I hope this bug isn't going to be as thorny as it looks... I'm resorting to a piece of cardboard taped over the touchpad just to type now.

I also have this problem with the ALPS GlidePoint trackpad on my HP Mini 311 on Lucid Beta 2. The buttons and tap to click work but edge scrolling quite working, changes to "Mouse Preferences" have no impact and 2-finger scroll is grayed-out of mouse preferences. This was working until a daily build approximately 2-3 days ago.

@J. McDonald
2-finger scroll is grayed-out because (somehow) the kernel detected that you trackpad doesn't support multitouch (i.e. the last two values of the Synaptics capabilities property should be 0). If it used to work, then it's definitely a kernel issue.

@Topher
I think you're still experiencing a kernel issue and your touchpad is being detected as a mouse. The output of dmesg and your /var/log/Xorg.0.log should help me confirm my theory.

I'm subscribing Chase as he might have an idea about what's happening to J. McDonald's touchpad in the kernel.

@Mary
What's the output of the following command?
xinput list-props "AlpsPS/2 ALPS GlidePoint"

I did try booting into the older kernel (2.6.32-20). Edge scrolling suddenly reappeared but the trackpad tab disappeared from mouse preferences. My Xorg.0 log is attached. I will also attach dmesg in a few minutes

Here is the dmesg

Just to reiterate, with current kernel versions in Lucid beta 2 over the last couple of days, edge scrolling is disabled, it is selectable from "Mouse Preferences" but selecting or deselecting any option in "Mouse Preferences" has no impact on the trackpads function. I don't use 2-finger scrolling, but I did notice it was grayed-out and I though I should include that description for completeness. Edge scrolling returns when reverting to an older kernel version at time of boot using grub, but if I do this the tab for the trackpad disappears and while it seems to function fine, it no longer appears to be recognized by either "Mouse Preferences" or by "GPointing Device Settings". Using the latest kernel versions (2.6.33-20, -21) the trackpad is definitely showing up as an ALPS GlidePoint within GPointing Device, but as with Mouse Settings none of the selectable options appear to change anything with respect to trackpad function. Thanks again for your assistance with this issue.

I have a Sony Vaio VPCF11 which also uses the AlpsPS/2 ALPS GlidePoint (E7: 73 02 64) running Kubuntu 10.04 beta 2. My problems with the touchpad are identical to Mary's. xinput list-props shows a large list of touchpad properties. Furthermore, changes that are made in the KDE touchpad GUI are reflected in the output of xinput list-props. xinput test "AlpsPS/2 ALPS GlidePoint" shows no response from the touchpad while xinput test "PS/2 Mouse" outputs the relative movements and button press events of the touchpad.

There is a duplicate bug here:
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/527890

More info here:
http://ubuntuforums.org/showthread.php?t=1316361&page=2

I can confirm that xinput test "PS/2 Mouse" outputs relative movements in kernel 2.6.32-21 and -20 while xinput test "AlpsPS/2 ALPS GlidePoint" shows no response. In kernel 2.6.31-21, xinput test "ImPS/2 Generic Wheel Mouse" outputs relative movements for the ALPS GlidePoint in the HP Mini 311.

I checked output from lsmod. It does not show synaptics as loaded.

Alberto - sorry I didn't see your request earlier! Here it is.

Mary - your xinput list props looks similar to mine, but in my case xinput test confirms that it's actually "PS/2 Mouse" that is being used instead of "AlpsPS/2 ALPS GlidePoint". Are you able to get any output from xinput test "AlpsPS/2 ALPS GlidePoint" when you use the trackpad?

I think that is something related to driver,

Maybe Absolute mode is not working the way it should.
The Scrolling on Alps is managed by the absolute mode.
So if someone is able to debug the absolute mode please help!!!!

Thanks.

Could this be a bug in evdev?

I'd prefer to leave this bug unduplicated, since I've been following this bug and will be working on it shortly.

Thanks

That sounds good. I probably should not have marked the duplication without fully understanding the implications. Thanks for working on this bug.

J, McDonald:

Just so you know, marking bugs as duplicates is generally a good thing and helps keep bugs manageable. Definitely keep doing what you're doing. It's just that in this case, I have a history with this bug, so I want to work through things here. If this is marked as a duplicate, it gets removed from all the bug lists. It's just a matter of preference and ease for me as I try to work it.

Thanks

No, I don't get any output. (I think I did what you suggested - put xinput
test "AlpsPS/2 ALPS GlidePoint" as a command in the terminal and nothing
came up, although when I tried to close the terminal a message came up
saying a process was running and I would kill it when I closed the
terminal. I waited a long time, but nothing happened, so I closed it.)

On Sat, Apr 17, 2010 at 10:47 AM, J. McDonald <
<email address hidden>> wrote:

> Mary - your xinput list props looks similar to mine, but in my case
> xinput test confirms that it's actually "PS/2 Mouse" that is being used
> instead of "AlpsPS/2 ALPS GlidePoint". Are you able to get any output
> from xinput test "AlpsPS/2 ALPS GlidePoint" when you use the trackpad?
>
> --
> Alps touchpad is recognized but neither gsynaptics nor mouse preferences
> work
> https://bugs.launchpad.net/bugs/550625
> You received this bug notification because you are a direct subscriber
> of the bug.
>
> Status in “xserver-xorg-input-synaptics” package in Ubuntu: New
>
> Bug description:
> Binary package hint: xserver-xorg-input-synaptics
>
> I have Ubuntu 9.10 and an Alps touchpad (E7: 73 02 64). Chase fixed the
> kernel so that it was recognized and, indeed, the touchpad shows up in both
> gsynaptics and system>preferences>mouse, but when I actually try to adjust
> the settings nothing changes. My goal is to disable the tap-to-click feature
> or disable the touchpad when typing. Chase suggests it may be something
> related to the driver and to open another bug.
>
> ProblemType: Bug
> Architecture: i386
> Date: Sun Mar 28 22:02:25 2010
> DistroRelease: Ubuntu 9.10
> InstallationMedia: Ubuntu 9.10 "Karmic Koala" - Release i386 (20091028.5)
> MachineType: Hewlett-Packard HP Pavilion dm3 Notebook PC
> Package: xserver-xorg-input-synaptics 1.1.2-1ubuntu7
> ProcCmdLine: BOOT_IMAGE=/boot/vmlinuz-2.6.31-21-generic
> root=UUID=f1e2e19e-a0f8-407d-9533-c5ba95a8be50 ro quiet splash
> ProcEnviron:
> LANG=en_US.UTF-8
> SHELL=/bin/bash
> ProcVersionSignature: Ubuntu 2.6.31-21.58~alps4-generic
> RelatedPackageVersions:
> xserver-xorg 1:7.4+3ubuntu10
> libgl1-mesa-glx 7.6.0-1ubuntu4
> libdrm2 2.4.14-1ubuntu1
> xserver-xorg-video-intel 2:2.9.0-1ubuntu2.1
> xserver-xorg-video-ati 1:6.12.99+git20090929.7968e1fb-0ubuntu1
> SourcePackage: xserver-xorg-input-synaptics
> Uname: Linux 2.6.31-21-generic i686
> XorgConf:
> Section "InputDevice"
> Identifier "Macintosh mouse button emulation"
> Driver "synaptics"
> Option "SHMConfig" "true"
> EndSection
> dmi.bios.date: 12/04/2009
> dmi.bios.vendor: Insyde Corp.
> dmi.bios.version: F.22
> dmi.board.asset.tag: Base Board Asset Tag
> dmi.board.name: 3656
> dmi.board.vendor: Hewlett-Packard
> dmi.board.version: 44.1D
> dmi.chassis.type: 10
> dmi.chassis.vendor: Hewlett-Packard
> dmi.chassis.version: N/A
> dmi.modalias:
> dmi:bvnInsydeCorp.:bvrF.22:bd12/04/2009:svnHewlett-Packard:pnHPPaviliondm3NotebookPC:pvr0486110000241F00110320000:rvnHewlett-Packard:rn3656:rvr44.1D:cvnHewlett-Packard:ct10:cvrN/A:
> dmi.product.name: HP Pavilion dm3 Notebook PC
> dmi.product.version: 0486110000241F00110320000
> dmi.sys.vendor: Hewlett-Packard
> fglrx: Not loaded
> s...

All the properties look correct to me. The right edge is defined as anything to the right of value 870. I suggest we try to find out what X values are being reported when we attempt to scroll on the right edge of the touchpad. Someone please test the following:

You will want to print out or write down these instructions as we will be changing to a VT for part of it. You can switch back and forth using chvt if you want, but if you are unfamiliar it will be easiest to run through these steps verbatim. If you get stuck in a text terminal and don't know how to get out, run "sudo chvt 7".

$ sudo apt-get install input-utils evtest
$ sudo chvt 1

log in as your normal user

$ sudo lsinput

Find your mouse, should be something like "AlpsPS/2 ALPS GlidePoint". Now copy the "/dev/input/event#" line for the mouse.

$ sudo evtest [/dev/input/event# from previous step] | tee evtest.dump

Now run your finger across the trackpad in three motions:

1. Left to right across the middle of the touchpad
2. Top to bottom across the middle of the touchpad
3. Top to bottom across the right edge as though you are trying to scroll

You will see a ton of output flow across the screen. You may ignore the output, it's just there so you're sure it's working. The output is being saved to evtest.dump. When done:

$ sudo chvt 7

Finally, upload the evtest.dump file here. With this data we should be able to figure out if the edges reported to the synaptics X driver is correct.

Thanks

done this.... My alps has no output (event8), but I also have a ps2 mouse on event7, all data appears on this input.

See attached: events.tgz

@tuxinvader:

Were you in the VT when you ran evtest? If you run evtest while in the desktop X session, you won't receive anything. That's why I have the chvt in the instructions.

I worry that the ps2 mouse events won't be comparable to the synaptics events.

Thanks

I wasn't, I was logged on via ssh.

I have just tested it again on the console (/dev/tty1), and I get exactly the same output. Nothing on the ALPS device, everything on the PS2 mouse device.

I changed to VT and ran lsinput but the event number lines scrolled by so fast that I couldn't see which one corresponded to the ALPS Glidepoint. Is it OK to do lsinput from a terminal window in Gnome to get the event number and then switch to VT to run the test?

I just figured it out:

sudo lsinput | less

enter for each and then q when it reaches the end

I'm slowly learning...

In VT, I did:

$ sudo lsinput | less

and found "AlpsPS/2 ALPS GlidePoint" at

/dev/input/event7

so then I did:

$ sudo evtest /dev/input/event7 | tee evtest.dump

and wound up with a blinking cursor and no output at all corresponding to input from the trackpad. I had to do a forced shutdown to get out of it. I will check to see if evtest.dump was created and post output here if I find it.

@J. McDonald:

I'm sorry, sometimes I forget not everyone is a kernel hacker :). If you run a command at the command line and you need to stop it, hitting ctrl+c usually does the trick. That's what I would have used in this instance. I should have been more clear.

So this is confirmation that there is still an issue with the kernel driver. Even though the driver recognizes the device, it's not processing any packets. I will make a test kernel to determine what is going wrong.

Thanks

I only aspire to be a kernel hacker - but I'll get there someday! I can load up a test kernel and give it a try if that would help with the debugging. Thanks again.

I while in VT I confirmed what Tuxinvader found. This time around "AlpsPS/2 ALPS GlidePoint" was at /dev/input/event9 and "PS/2 Mouse" was at /dev/input/event8.

Like Tuxinvager, I was able to get the trackpad to output but only with

$ sudo evtest /dev/input/event8 | tee evtest2.dump

evtest2.dump is attached. As with the last go round, doing

$ sudo evtest /dev/input/event9 | tee evtest1.dump

gave no output from the trackpad. I quite with CRTL-C (thanks) and the resulting evtest1.dump file was empty. So the ALPS trackpad is using the PS/2 Mouse driver. Surprising that it works at all. Very strange. Are there any hints in the dmesg or Xorg.0.log?

There seems to have been an earlier problem where psmouse loading prior to synaptics would prevent synaptics from loading:

http://ubuntuforums.org/showthread.php?t=1035407

A temporary workaround was to blacklist psmouse and load the module after boot from rc.local

I checked the loaded modules and it does look like synaptics is not loading, so maybe this issue has been around for awhile.

@J. McDonald:

Good find! I would appreciate it if someone could test if blacklisting psmouse solves the issue. Be aware though, that if you blacklist psmouse and the alps driver still isn't working, then you may have difficulties in the desktop session :). If you find yourself in this situation, switch to a VT using 'ctrl+alt+1', log in, and use "sudo rm /etc/modprobe.d/touchpad" to delete the blacklist. Then run 'sudo restart gdm' to start the X desktop again.

The maintainer of the input subsystem of the kernel noted that this could also be due to a new ALPS protocol. If that's the case, we may have difficulties determining how to talk to these new mice.

Heh, I tried blacklisting psmouse in /etc/modprobe.d/touchpad, but it just disabled my touchpad altogether. :-) I checked Mouse preferences just in case Synaptics might have loaded improperly, but there was still no "Touchpad" tab. So I un-blacklisted again.

Fixed it. There's a ubuntu specific patch which adds the following line to the alps kernel driver alps.c:

   { { 0x73, 0x02, 0x64 }, 0xf8, 0xf8, 0 }, /* HP Pavilion dm3 */

Commenting out this new device fixes the ALPs on the HP mini 311c ;-)

Awesome. Is there a standard way to test this fix on my machine? I'm not a kernel hacker either...

You can try my module if you wish.... The only other way would be for you to compile your own kernel....

Download the psmouse.ko file attached, and then in a console (from the directory containing my module) run:

$ sudo rmmode psmouse
$ sudo insmod psmouse.ko

I should warn you about the dangers of installing kernel modules produced by people you don't know on the internet ;-)

You can trust me though :-p

@tuxinvader
You are actually taking a step backwards by commenting out that line in alps.c. Chase put that line in there so the "AlpsPS/2 ALPS GlidePoint" would be recognized in the kernel. I realize that you probably have vertical scrolling back however none of the other functions in Synaptics work properly. Do you have a touchpad tab in mouse prefs?

@Topher
Do you have a touchpad tab in mouse prefs? If not, you probably don't have that line in your kernel anyways.

I was not aware of that, but, granted my touchpad option has disappeared... but I'd rather have a working touchpad that I can't configure, than a broken one which I can ;-)

This is all way beyond my skills, but let me know if you need me to do
anything (you'll have to walk me through it). It sounds like you are making
headway, so I'll be watching for a solution I can easily install! Thank you
all for your work!

On Tue, Apr 20, 2010 at 9:24 AM, Chase Douglas
<email address hidden>wrote:

> @J. McDonald:
>
> Good find! I would appreciate it if someone could test if blacklisting
> psmouse solves the issue. Be aware though, that if you blacklist psmouse
> and the alps driver still isn't working, then you may have difficulties
> in the desktop session :). If you find yourself in this situation,
> switch to a VT using 'ctrl+alt+1', log in, and use "sudo rm
> /etc/modprobe.d/touchpad" to delete the blacklist. Then run 'sudo
> restart gdm' to start the X desktop again.
>
> The maintainer of the input subsystem of the kernel noted that this
> could also be due to a new ALPS protocol. If that's the case, we may
> have difficulties determining how to talk to these new mice.
>
> --
> Alps touchpad is recognized but neither gsynaptics nor mouse preferences
> work
> https://bugs.launchpad.net/bugs/550625
> You received this bug notification because you are a direct subscriber
> of the bug.
>
> Status in “xserver-xorg-input-synaptics” package in Ubuntu: New
>
> Bug description:
> Binary package hint: xserver-xorg-input-synaptics
>
> I have Ubuntu 9.10 and an Alps touchpad (E7: 73 02 64). Chase fixed the
> kernel so that it was recognized and, indeed, the touchpad shows up in both
> gsynaptics and system>preferences>mouse, but when I actually try to adjust
> the settings nothing changes. My goal is to disable the tap-to-click feature
> or disable the touchpad when typing. Chase suggests it may be something
> related to the driver and to open another bug.
>
> ProblemType: Bug
> Architecture: i386
> Date: Sun Mar 28 22:02:25 2010
> DistroRelease: Ubuntu 9.10
> InstallationMedia: Ubuntu 9.10 "Karmic Koala" - Release i386 (20091028.5)
> MachineType: Hewlett-Packard HP Pavilion dm3 Notebook PC
> Package: xserver-xorg-input-synaptics 1.1.2-1ubuntu7
> ProcCmdLine: BOOT_IMAGE=/boot/vmlinuz-2.6.31-21-generic
> root=UUID=f1e2e19e-a0f8-407d-9533-c5ba95a8be50 ro quiet splash
> ProcEnviron:
> LANG=en_US.UTF-8
> SHELL=/bin/bash
> ProcVersionSignature: Ubuntu 2.6.31-21.58~alps4-generic
> RelatedPackageVersions:
> xserver-xorg 1:7.4+3ubuntu10
> libgl1-mesa-glx 7.6.0-1ubuntu4
> libdrm2 2.4.14-1ubuntu1
> xserver-xorg-video-intel 2:2.9.0-1ubuntu2.1
> xserver-xorg-video-ati 1:6.12.99+git20090929.7968e1fb-0ubuntu1
> SourcePackage: xserver-xorg-input-synaptics
> Uname: Linux 2.6.31-21-generic i686
> XorgConf:
> Section "InputDevice"
> Identifier "Macintosh mouse button emulation"
> Driver "synaptics"
> Option "SHMConfig" "true"
> EndSection
> dmi.bios.date: 12/04/2009
> dmi.bios.vendor: Insyde Corp.
> dmi.bios.version: F.22
> dmi.board.asset.tag: Base Board Asset Tag
> dmi.board.name: 3656
> dmi.board.vendor: Hewlett-Packard
> dmi.board.version: 44.1D
> dmi.chassis.type: 10
> dmi.chassis.vendor: Hewlett-Packard
> dmi.chassis.version: N/A
> dmi.modalias:
> dmi:bvnInsydeCorp.:bvrF.22:bd12/04/2009:svnHewlett-Pack...

I would like to do some initial debugging to figure out what is going on. I would appreciate it if someone could perform the following steps:

$ sudo sh -c "echo ':mod:psmouse' > /sys/kernel/debug/tracing/set_ftrace_filter"
$ sudo sh -c "echo 1 > /sys/kernel/debug/tracing/tracing_enabled"
$ sudo sh -c "echo 1 > /sys/kernel/debug/tracing/tracing_on"
$ sudo sh -c "echo function_graph > /sys/kernel/debug/tracing/current_tracer"

Now run your finger across the trackpad once.

$ bzip2 -c /sys/kernel/debug/tracing/trace > trace.bz2
$ sudo sh -c "echo nop > /sys/kernel/debug/tracing/current_tracer"

Finally, attach trace.bz2 here.

Thanks

trace attached....

I'm going out in a bit, but I have run a few traces....

official.bz2 == Current Lucid kernel (finger run across pad)
my-mod.bz2 == My reverted driver (finger run across pad)
official-scroll.bz2 == Current Lucid kernel (finger run up/down scroll section)
my-mod-scroll.bz2 == My reverted driver (finger run up/down scroll section)

Hope it helps.

Cheers,
Mark

I'll take a look at the traces.

I just want to thank you all for testing this so quickly! It is much appreciated!

Thanks!

I've made a test i386 psmouse module that will print out the ps/2 packets that are not handled by the alps driver. I would appreciate it if someone would test it:

$ wget http://people.canonical.com/~cndougla/550625/v1/psmouse.ko
$ sudo rmmod psmouse
$ sudo insmod psmouse

Run your finger across the touchpad

$ dmesg > dmesg-debugv1.log
$ sudo rmmod psmouse
$ sudo modprobe psmouse

Finally, attach the dmesg-debugv1.log file here.

If someone is running amd64 and needs a compatible module please let me know.

Thanks

I can do it on my lunch break if you can provide a amd64 module in the next hour.

I've uploaded an amd64 version: http://people.canonical.com/~cndougla/550625/v1/psmouse-amd64.ko.

I followed your instructions with "suod insmod psmouse-amd64.ko" instead of "sudo insmod psmouse" because the file wasn't found. Thanks for all of your help.

I've uploaded a new version of the psmouse module to http://people.canonical.com/~cndougla/550625/v2/. This version of the module adds the ALPS_PASS option for this hardware. The ALPS_PASS option must be enabled or disabled properly or else absolute events will not be enabled. My hope is that by using this option the absolute events will work once again. Please test it out and report whether it is working or not.

If it still fails to work, I would appreciate another set of logs from the instructions I provided in comments #44 and #48.

Thanks

Wow - I go to work for 10 hours and out for a couple of beers afterwards and in the meantime things have progressed at quite a clip. I tried v2 and ran comment 48 first, then 44. I am uploading the two files. Let me know if there is anything else I can do.

Unfortunately, it appears that this touchpad uses a new protocol between the device and the driver. This means that we have to figure out the protocol before we can unlock the full functionality of the touchpad. One way to do this is to watch what the Windows driver does when it talks to the hardware, but this is easier said than done.

In the meantime, the device ID for this touchpad has been backed out of the upstream linux-input tree at my request, and I will shortly ask for it to be backed out of the Ubuntu kernel. Though the touchpad won't be seen in gsynaptics or other synaptics compatible software, scrolling functionality will be restored. When the proper protocol is determined for this touchpad we can look into pulling it into Ubuntu again.

I'm sorry this can't be completely fixed at this time.

Thanks for your understanding.

Could it be something related to the comand setPoll ???Maybe this has another way to set Poll.

Just a comment for clarification - my issue with the ALPS touchpad isn't inability to scroll, it's an inability to use the syndaemon touch disabler (because the synaptics driver isn't loaded), which makes it difficult to type long papers. This was established in the xserver-xorg-input-synaptics bug discussion before I was told that I needed to back down to the kernel level. But to be honest, so far it seems like the same root issue. Just wanted to add that input.

@alabamas:

If you are interested, you can view the code at http://lxr.linux.no/linux+v2.6.33/drivers/input/mouse/alps.c. Without any documentation for this device, we don't really know what to do to make it work.

@Topher:

Yes, without the ability to turn on absolute events for this device, the synaptics userspace functionality won't work. It could be that the synaptics stuff uses the pressure and/or touch area size data that are only available if the device and driver are working right.

@Topher
Take a look at what others have done in this thread: http://ubuntuforums.org/showthread.php?t=1316361
I haven't tried it, but it looks like others have had some success. I am going to wait for a proper fix.

Thanks for everyone's help in trying to resolve this issue.

@Chase
Can you post a diff for the changes that you made in comment 48? I am thinking about trying some tests, and would like to use your work as a model. Thanks.

So in the interim, should I go ahead and run the patch in comment #56? After that, to rid myself of accidental palm-clicks to the touchpad during typing I can run the patch referred to at http://ubuntuforums.org/showthread.php?t=1316361&page=3 and add a delay to touchpad clicks following keyboard activity? This seems like a manageable compromise until the ALPS GlidePoint hardware is figured out.

@tob

Here's the patch I used for printing out some packet information. I really only wanted to know what packets looked like when they weren't getting handled by the bare ps2 handler, since the trace showed everything was either handled there or not handled at all (meaning they weren't a full packet yet and were kept until more data arrived).

Thanks to everyone, particularly Chase, for your work on this!

On Wed, Apr 21, 2010 at 1:31 PM, Chase Douglas
<email address hidden>wrote:

> ** Changed in: linux (Ubuntu)
> Status: In Progress => Fix Committed
>
> ** Summary changed:
>
> - Alps touchpad is recognized but neither gsynaptics nor mouse preferences
> work
> + Alps touchpad is recognized but synaptics clients and scrolling do not
> work
>
> --
> Alps touchpad is recognized but synaptics clients and scrolling do not work
> https://bugs.launchpad.net/bugs/550625
> You received this bug notification because you are a direct subscriber
> of the bug.
>
> Status in “linux” package in Ubuntu: Fix Committed
>
> Bug description:
> SRU Justification:
>
> Impact: The end result of the commit is loss of scroll support, with no
> extra functionality gained.
>
> How addressed: By reverting the commit, the previously enabled scroll
> support is reestablished through the ImPS/2 extension.
>
> Reproduction: Run current Ubuntu kernel on affected hardware. Scroll
> functionality present in previous Ubuntu releases is now missing.
>
> Regression Potential: Since this is a revert of a one-line device id,
> the regression potential is minimal. The device should fall back to
> ImPS/2 protocol without issue.
>
> =====
>
> Binary package hint: xserver-xorg-input-synaptics
>
> I have Ubuntu 9.10 and an Alps touchpad (E7: 73 02 64). Chase fixed the
> kernel so that it was recognized and, indeed, the touchpad shows up in both
> gsynaptics and system>preferences>mouse, but when I actually try to adjust
> the settings nothing changes. My goal is to disable the tap-to-click feature
> or disable the touchpad when typing. Chase suggests it may be something
> related to the driver and to open another bug.
>
> ProblemType: Bug
> Architecture: i386
> Date: Sun Mar 28 22:02:25 2010
> DistroRelease: Ubuntu 9.10
> InstallationMedia: Ubuntu 9.10 "Karmic Koala" - Release i386 (20091028.5)
> MachineType: Hewlett-Packard HP Pavilion dm3 Notebook PC
> Package: xserver-xorg-input-synaptics 1.1.2-1ubuntu7
> ProcCmdLine: BOOT_IMAGE=/boot/vmlinuz-2.6.31-21-generic
> root=UUID=f1e2e19e-a0f8-407d-9533-c5ba95a8be50 ro quiet splash
> ProcEnviron:
> LANG=en_US.UTF-8
> SHELL=/bin/bash
> ProcVersionSignature: Ubuntu 2.6.31-21.58~alps4-generic
> RelatedPackageVersions:
> xserver-xorg 1:7.4+3ubuntu10
> libgl1-mesa-glx 7.6.0-1ubuntu4
> libdrm2 2.4.14-1ubuntu1
> xserver-xorg-video-intel 2:2.9.0-1ubuntu2.1
> xserver-xorg-video-ati 1:6.12.99+git20090929.7968e1fb-0ubuntu1
> SourcePackage: xserver-xorg-input-synaptics
> Uname: Linux 2.6.31-21-generic i686
> XorgConf:
> Section "InputDevice"
> Identifier "Macintosh mouse button emulation"
> Driver "synaptics"
> Option "SHMConfig" "true"
> EndSection
> dmi.bios.date: 12/04/2009
> dmi.bios.vendor: Insyde Corp.
> dmi.bios.version: F.22
> dmi.board.asset.tag: Base Board Asset Tag
> dmi.board.name: 3656
> dmi.board.vendor: Hewlett-Packard
> dmi.board.version: 44.1D
> dmi.chassis.type: 10
> dmi.chassis.vendor: Hewlett-Packard
> dmi.chassis.version: N/A
> dmi.modalias:
> dmi:bvnInsydeCorp.:bvrF.22:bd12/04/2009:svnHewlett-Packard:pnHPPaviliondm3NotebookPC:pvr048611000024...

@J. McDonald:

You can, or you can wait for the first kernel to come from lucid-updates. That will be a while though (a few weeks after release I think).

Chase,

Thanks for the info and thank you for all of your work troubleshooting this issue. How do I go about applying the patch?

@J. McDonald:

Unfortunately, it's not the easiest thing to apply a patch to the kernel. The build system isn't the easiest to work with. What I can do is build a special 2.6.32-21 kernel with the patch to revert the alps signature. You could install it, and when the -22 kernel comes out with the patch included you will be updated automatically. Does this sound good to you? Bear in mind that we can't offer any support for non-Ubuntu kernels, so if you hit any other issues we'll likely ask you to reproduce it on an official Ubuntu kernel.

Chase, thanks, that would be great. I'm not overly concerned about support for this computer - hence the testing, but getting the scroll bar back for the few weeks until -22 comes would make things a bit easier. I also realise that at some point, I should learn to do this for myself.

Actually, we've now started daily pre-proposed kernel builds. These builds will include all the patches we've committed to our kernels, but not yet been released. There's no support for running these kernels, but they should hopefully be safe. You can find them at https://launchpad.net/~kernel-ppa/+archive/pre-proposed/+packages. The current kernel there has the ALPS signature reverted, so it should tide you over until the official kernel.

I would just install the kernel package manually unless you really want to keep up with the daily kernel releases. If you do want the dailies, then add the ppa to your package sources and it will automatically keep you up to date.

Thanks

I went ahead and added the ppa to try out the daily builds. Scrolling has returned to normal. I also added the package "mouseemu" and set a small delay in to disable the trackpad while I type. Everything is working now.

With the latest kernel from the ppa scrolling now works here too. But I cannot configure the touchpad with the GUI anymore. Everything is greyed out. (Kubuntu). And the name of the touchpad is not shown anymore in the GUI.

The final release broke the scrolling again...I'm going to try posting in "main support" forum....

I downgraded linux-image-2.6.32.22-generic from the final release version to version 2.6.32-22.33~pre201004281000 and edge scrolling returned to the ALPS GlidePoint trackpad. Anyone know what's going on with this?

See bug #554050, it seems to be the same as this.

@J. McDonald:

I'm not quite sure how you got a 2.6.32-22-generic kernel already because it hasn't been released yet. However, you should be able to keep updating with the daily pre-proposed builds until the official 2.6.32-22 kernel is released. The official kernel should have the ALPS scrolling fixed.

Thanks

That was entirely my mistake - I subscribed to pre-proposed daily builds and got a daily build that somehow reverted the change to the earlier 2.6.32-22 daily that you had fixed. I think the military refers to this as SNAFU. I am disabling that repository and I will wait for the official kernel - sorry about adding confusion to this issue.

-----Original Message-----
>From: Chase Douglas <email address hidden>
>Sent: May 3, 2010 10:16 AM
>To: <email address hidden>
>Subject: [Bug 550625] Re: Alps touchpad is recognized but synaptics clients and scrolling do not work
>
>@J. McDonald:
>
>I'm not quite sure how you got a 2.6.32-22-generic kernel already
>because it hasn't been released yet. However, you should be able to keep
>updating with the daily pre-proposed builds until the official 2.6.32-22
>kernel is released. The official kernel should have the ALPS scrolling
>fixed.
>
>Thanks
>
>

Today I updated to linux-image-2.6.32-22-generic (version 2.6.32-22.33) from the official lucid repositories, but edge scrolling doesn't work when I remove the below workaround.
I'm using Sony VAIO EB1E1E (see https://bugs.launchpad.net/bugs/554050) and maybe my touchpad has different device id.
Please let me know how to check the device id.
I tried with dmesg, but I couldn't find anything that looks like the removed line from the patch attached to this bug.

It's good we have this permanent workaround in the mean time;
Add the following to /etc/modprobe.d/psmouse.conf:
options psmouse proto=imps

The psmouse.conf file doesn't exist by default, just create a text file with that name in /etc/modprobe.d/

Regards,
Rocky

There's no edit button :( "mean time" to be read as "meantime"

I'm also affected by this bug (toshiba u500). But i don't know how to install those "pre-proposed" kernels that C. Douglas have mentionned in comment 70.

Not that it was functioning anyway, but today I noticed that a recent update removed the touchpad tab again.

#78's fix worked to re-establish scrolling.

mouseemu delay doesn't seem to work for disabling accidental taps (#71)

My apologies, #78's fix removed the tab, not the update.

Has is problem been fixed? I have a Dell Inspiron with Alps touchpad. I'm using Lucid Lynx with Xubuntu and 2.6.32-22 kernel. I have a problem with my touchpad being really sensitive at times, it's hard to type. Is there anyway to adjust the sensitvity?

Thanks,
Tony

I'm also having this problem with an HP Pavillion DM3 that I bought today - I had to create a /etc/modprobe.d/psmouse.conf file with "options psmouse proto-imps" in order to get edge-scrolling working.

The issue appears to be caused by the touchpad being incorrectly detected as a PS/2 mouse. When I do the following:

xinput test "AlpsPS/2 ALPS GlidePoint"

nothing is registered. When I do this:

xinput test "PS/2 Mouse"

I get co-ordinates for every movement of the trackpad.

Currently on Ubuntu 10.04 64-bit, kernel 2.6.32-22-generic.

Installing gpointing-device-settings fixed my sensitivity adjustment issue.

I can confirm this bug is still present.
Sony Vaio VPCF11C5E
Kernel Ubuntu 2.6.32-22-generic

xinput list
⎡ Virtual core pointer id=2 [master pointer (3)]
⎜ ↳ Virtual core XTEST pointer id=4 [slave pointer (2)]
⎜ ↳ Logitech USB Receiver id=10 [slave pointer (2)] (My Logitech mouse)
⎜ ↳ Logitech USB Receiver id=11 [slave pointer (2)] (My Logitech mouse again)
⎜ ↳ PS/2 Mouse id=13 [slave pointer (2)] [My touchpad, working as a mouse w "tap-to-click"]
⎜ ↳ AlpsPS/2 ALPS GlidePoint id=14 [slave pointer (2)] [My touchpad, not configurable]
⎜ ↳ Macintosh mouse button emulation id=15 [slave pointer (2)]

I get the same results to xinput test as Gareth Halfacree at #84
I _wish_ I could disable tap-to-click!

To recap: the ALPS GlidePoint can in theory be used with a generic PS/2 driver, a generic ImPS/2 driver, or the advanced synaptic/alps driver.

If 0x73, 0x02, 0x64 is added to the alps driver, the kernel will use this to talk to the touchpad. However, the touchpad uses a protocol that is slightly different from the other alps touchpads, and we haven't figured out exactly how it works. Because of that, when the alps driver is used, all the fancy features will appear to be available in the config screens, but will not actually work. Scrolling will not work either.

The generic PS/2 driver doesn't support scrolling. The generic ImPS/2 driver does support scrolling. Until we figure out what protocol this touchpad uses, we should fall back to these, and use the configuration as described by Rocky in #78 to make sure ImPS/2 is used and scrolling works.

Anything I can do to help figure out the protocol for the alps driver with this touchpad?

raboof,

Good summary! FYI, we have reverted the signature in our source tree, but an update with the fix hasn't been released yet. It should be released in the next kernel update.

On how to figure out the new protocol, we probably need to watch what windows does. This is tricky though. The best bet would probably be to fake the device in qemu/kvm, and then watch to see how windows communicates with it. However, this is beyond my knowledge.

The kernel doesn't appear to have much support for multi-touch on Alps touchpads. Currently, there are three basic touchpad packet formats in the kernel: PS/2, Absolute, and Interleaved. None of these formats contain advanced multi-touch information. The Alps GlidePoint is a modern multi-touch touchpad that probably has a new format that is not currently defined in the kernel. I have performed some amateur kernel testing in order to try to determine the new format. However, I am not able to read anything other than basic PS/2 packets whenever the signature is defined. I believe that a new "magic knock" is required in order to place the touchpad into the new multi-touch mode. Trying to determine this new "magic knock" is way beyond my capabilities. I think Chase has the right idea by using Windows to reverse engineer the protocol. Does anyone know any online resources with info on how to sniff and reverse engineer the windows 7 touchpad data? I had to return my laptop due to an unrelated problem, however I would like to look into this further when I receive a new one.

Passing 'proto=imps' when loading the psmouse module does not seem to work for me - xinput still reports a 'PS/2 Generic Mouse' and does not generate scrolling events. This is on Debian Testing, though.

Is there a way to disable the touchpad completely? On my laptop (Dell E6410), both trackpoint and touchpad are registered as PS/2 Mouse. Buttons are designed in a way that make it almost impossible to use trackpoint buttons without simultaneously tapping touchpad. Or is it completely indistinguishable at this point?

Disable and re-enable using "modprobe -r psmouse" and "modprobe psmouse".

I use the Alt+F2 Run thing with gksudo as a quick way to turn it on and off (ie. when using another mouse plugged into usb).

The revert will be included in the next proposed update to Lucid. Chase, could you update the status for current development? Was it reverted upstream as well (mainline) or do we need to revert it there as well until the issue is completely fixed?

Stefan,

The initial patch to add the signature to the driver has been reverted upstream. By reverting the patch in Lucid we will be keeping current with upstream.

Accepted linux into lucid-proposed, the package will build now and be available in a few hours. Please test and give feedback here. See https://wiki.ubuntu.com/Testing/EnableProposed for documentation how to enable and use -proposed. Thank you in advance!

Hello I have a Dell E6410. I have tried the kernel from lucid-proposed. As some have reported, it removes the tab with the alps options from gpointing-device-settings completely. However the vertical (and horizontal) scroll does not work. I does not make any difference to use proto=imps or not when loading psmouse module.

Hello, I have a Dell E6510 and also have tried the kernel from lucid-proposed with no luck on the edge scrolling.

I've the same problem with my DELL Precision M 4500 running Ubuntu
10.04 LTS (Linux opale3 2.6.32-22-generic #36-Ubuntu SMP Thu Jun 3
19:31:57 UTC 2010 x86_64 GNU/Linux).

The touchpad is detected by the kernel, lsinput gives:

/dev/input/event10
   bustype : BUS_I8042
   vendor : 0x2
   product : 0x8
   version : 29478
   name : "AlpsPS/2 ALPS GlidePoint"
   phys : "isa0060/serio1/input0"
   bits ev : EV_SYN EV_KEY EV_REL EV_ABS

It seems more or less "seen" by the X server:

(II) config/udev: Adding input device AlpsPS/2 ALPS GlidePoint (/dev/input/event10)
(**) AlpsPS/2 ALPS GlidePoint: Applying InputClass "evdev pointer catchall"
(**) AlpsPS/2 ALPS GlidePoint: Applying InputClass "evdev touchpad catchall"
(**) AlpsPS/2 ALPS GlidePoint: Applying InputClass "touchpad catchall"
(II) LoadModule: "synaptics"
(II) Loading /usr/lib/xorg/modules/input/synaptics_drv.so
(II) Module synaptics: vendor="X.Org Foundation"
        compiled for 1.7.6, module version = 1.2.2
        Module class: X.Org XInput Driver
        ABI class: X.Org XInput driver, version 7.0
(II) Synaptics touchpad driver version 1.2.2
(**) Option "Device" "/dev/input/event10"
(II) AlpsPS/2 ALPS GlidePoint: x-axis range 0 - 1023
(II) AlpsPS/2 ALPS GlidePoint: y-axis range 0 - 767
(II) AlpsPS/2 ALPS GlidePoint: pressure range 0 - 127
(II) AlpsPS/2 ALPS GlidePoint: finger width range 0 - 0
(II) AlpsPS/2 ALPS GlidePoint: buttons: left right middle
(--) AlpsPS/2 ALPS GlidePoint: touchpad found
(**) AlpsPS/2 ALPS GlidePoint: always reports core events
(II) XINPUT: Adding extended input device "AlpsPS/2 ALPS GlidePoint" (type: TOUCHPAD)
(**) AlpsPS/2 ALPS GlidePoint: (accel) keeping acceleration scheme 1
(**) AlpsPS/2 ALPS GlidePoint: (accel) acceleration profile 0
(**) AlpsPS/2 ALPS GlidePoint: (accel) acceleration factor: 2.000
(**) AlpsPS/2 ALPS GlidePoint: (accel) acceleration threshold: 4
(--) AlpsPS/2 ALPS GlidePoint: touchpad found
(II) config/udev: Adding input device AlpsPS/2 ALPS GlidePoint (/dev/input/mouse3)
(II) No input driver/identifier specified (ignoring)

Thus the synaptics driver is loaded but the last two lines are not very
encouraging. Indeed, the touchpad work like a PS/2 mouse (e.g., no multiple
taps or edge scrolling).

Moreover, with xinput I've:

olivier@opale3:~$ xinput list
Virtual core pointer id=2 [master pointer (3)]
   Virtual core XTEST pointer id=4 [slave pointer (2)]
   AlpsPS/2 ALPS GlidePoint id=13 [slave pointer (2)]
   PS/2 Mouse id=14 [slave pointer (2)]
   Macintosh mouse button emulation id=15 [slave pointer (2)]
Virtual core keyboard id=3 [master keyboard (2)]
 ...

xinput list-props "AlpsPS/2 ALPS GlidePoint" gives a full list of
Synaptics properties (40). However, xinput test "AlpsPS/2 ALPS GlidePoint"
reports NOTHING when I use the touchpad. But, xinput test "PS/2 Mouse" report
things ...

Thanks for working on this problem!

Olivier

Hi, same problem here:
I have a Dell E6510. The AlpsPS/2 touchpad completely disappeared after upgrading to Lucid-proposed (both a PS/2 and a AlpsPS/2 were shown before in gpointing-device-settings). The touchpad can't be configured and scrolling doesn't work.

This bug was fixed in the package linux - 2.6.32-23.37

---------------
linux (2.6.32-23.37) lucid-proposed; urgency=low

  [ Alex Deucher ]

  * SAUCE: drm/radeon/kms/atom: fix dual-link DVI on DCE3.2/4.0
    - LP: #564559

  [ Andy Whitcroft ]

  * [Config] ports -- build in dm-mod to enable LVM boot
    - LP: #560717
  * tools -- fix perf version extraction for multi-part flavours
    - LP: #555130
  * SAUCE: ACPI: EC: Allow multibyte access to EC (v3)
    - LP: #526354
  * [Config] enforce -- ensure dm_mod is built-in for LVM
    - LP: #560717
  * update to ubuntu-debian:7e708d33054c373faf41da23b73e8b48c342d958
    - LP: #570500, #576274

  [ Chase Douglas ]

  * Revert "(pre-stable): input: ALPS - Add signature for HP Pavilion dm3
    laptops"
    - LP: #550625
  * Enable ftrace function profiler
    - LP: #570389
  * enforce CONFIG_TMPFS_POSIX_ACL=y
    - LP: #575940

  [ Leann Ogasawara ]

  * Revert "staging/comdi -- disable"
    - LP: #563436
  * [Config] Enable multicast routing for sparc
    - LP: #416266
  * [Config] Add ahci.ko to virtual sub-flavour
    - LP: #570542

  [ Stefan Bader ]

  * Revert "SAUCE: drm/i915: Disable FBC on 915GM and 945GM"
    - LP: #588832

  [ Tim Gardner ]

  * ubuntu: rtl8192se -- update to version 0015.0127.2010
    - LP: #567016
  * [Config] Add atl1c to nic-modules udeb
    - LP: #557130

  [ Upstream Kernel Changes ]

  * Revert "(pre-stable) iwlwifi: fix nfreed--"
    - LP: #575853
  * Revert "backlight: mbp_nvidia_bl - add five more MacBook variants"
    - LP: #575853
  * Revert "(pre-stable) pata_via: Add VIA VX900 support"
    - LP: #575853
  * Revert "(pre-stable) x86-32, resume: do a global tlb flush in S4
    resume"
    - LP: #575853
  * Revert "x86: disable IOMMUs on kernel crash"
    - LP: #575853
  * Revert "sunrpc: fix peername failed on closed listener"
    - LP: #575853
  * Revert "sunrpc: move the close processing after do recvfrom method"
    - LP: #575853
  * Revert "(pre-stable) drm/edid: allow certain bogus edids to hit a fixup
    path rather than fail"
    - LP: #575853
  * Revert "drm/radeon/kms: don't print error on -ERESTARTSYS."
    - LP: #575853
  * Revert "ath9k: fix lockdep warning when unloading module" on stable
    kernels
    - LP: #588832
  * Staging: comedi: removed "depricated" from COMEDI_CB_BLOCK
    - LP: #483343
  * fat: fix buffer overflow in vfat_create_shortname()
    - LP: #575853
  * xfs: simplify inode teardown
    - LP: #575853
  * xfs: fix mmap_sem/iolock inversion in xfs_free_eofblocks
    - LP: #575853
  * xfs: I/O completion handlers must use NOFS allocations
    - LP: #575853
  * xfs: Wrapped journal record corruption on read at recovery
    - LP: #575853
  * xfs: Fix error return for fallocate() on XFS
    - LP: #575853
  * xfs: check for not fully initialized inodes in xfs_ireclaim
    - LP: #575853
  * xfs: fix timestamp handling in xfs_setattr
    - LP: #575853
  * xfs: Don't flush stale inodes
    - LP: #575853
  * xfs: Ensure we force all busy extents in range to disk
    - LP: #575853
  * xfs: reclaim inodes under a write lock
    - LP: #575853
  * xfs: Avoid inodes in reclaim when flushing from inode cache
    - LP: #575853
  * xfs: recla...

The bug was marked "Fix released" but the current situation in some Dell Laptops, like mine, the E6410, the E6510, and maybe the M4500 from comment #98 don't have scrolling working. Should we open a new, specific, bug report for these machines?

Same problem like #101 with my E6510. Anybody help the new Kernel?

I'm also affected (Dell Latitude E6510), so far none of the workarounds worked for me at all. Can this bug be reopened or should we open a new one?

Ok, interesting thing here:

no alps glide devices are listed in:
- xinput list
- dmesg | grep touch

But if I do:

xinput test "Mouseemu virtual mouse"

I get a response everytime I touch the touchpad. I.e. the alps device is now linked to the "Mouseemu virtual mouse".

I: Bus=0006 Vendor=001f Product=001e Version=0000
N: Name="Mouseemu virtual mouse"
P: Phys=
S: Sysfs=/devices/virtual/input/input14
U: Uniq=
H: Handlers=mouse4 event13
B: EV=7
B: KEY=70000 0 0 0 0
B: REL=103

How could I force synaptics to load on that device ??

This bug has not been fixed.

The alps driver in the kernel has been reverted to a prior version by removing the Alps GlidePoint touchpad signature (E7: 73 02 64). This forces the kernel to use the basic PS/2 driver for the touchpad which should allow vertical scrolling. However, the basic PS/2 driver doesn't work with synaptics so no advanced features are available.

For me, vertical scrolling doesn't even work with the latest kernel. This bug makes working on a notebook very difficult.

I can confirm the others' comments for a Dell Latitude E6510 on Ubuntu 10.04 (Lucid), even on kernel 2.6.32-23.
The Alps GlidePont entry disppeared from the Pinting Devices settings menu on my laptop as well.
I appreciate your work on fixing this - I'm a linux newbie but willing to do tests if you tell me what i need to do.

Touchpad only works in very basic mode on Dell E6510 (BIOS A03), Ubuntu 10.04 x86-64, kernel 2.6.32-23. No scrolling on the edges, no extra touch pad control panel in mouse prefs.

Further comments on Dell E6510: setting options psmouse proto=imps does not enable touchpad scrolling. This is using upstream 2.6.34 kernel.

Marek Janda wrote on 2010-06-06: #91
>
> Is there a way to disable the touchpad completely? On my laptop (Dell E6410),
> both trackpoint and touchpad are registered as > PS/2 Mouse. Buttons are
> designed in a way that make it almost impossible to use trackpoint buttons
> without simultaneously > tapping touchpad.
> Or is it completely indistinguishable at this point?

Same problem on Dell E6510. If the touchpad cannot work with palm-detection or disable-when-typing, I'd really like to disable it, since it gets in the way both when typing and when pressing the mouse buttons *above* it. Accidental clicks caused by tapping the top of the touchpad is turning into a very real annoyance.

So please, is there at least *any* way to disable the touchpad completely, and only have the stick and its three associated buttons enabled ?

I can confirm that the short term "fix" of the Alps Glidepoint on the HP Mini 311 still gives edge scrolling and is still recognized as a PS/2 Mouse without any Synaptics functionality in the current 2.6.32-23 kernel. Is anyone working on a longer term fix for Synaptics to work properly? Has this been fixed within 2.6.34 or 2.6.35?

The ALPS fix does not work for my ALPS on a VAIO for the latest 2.6.32-23

Please check the files I am attaching with apport-collect

For some reason I can not upload the files

"You are not
the reporter or subscriber of this problem report, or the report is a
duplicate or already closed".

with apport-collect, so here they come manually

You're mentioning Dells E6410, E6510 suffer from this problem too. Both these models are equipped with the Instant ON feature which is AFAIK Linux-kernel based (http://www.splashtop.com/)

If the touchpad scrolling does not work, I'd suggest to contact Dell and insist on a rapid bugfix.
If it does work, IMO it would mean a Linux compatible driver is found in the Instant ON image and could be exported.

In both cases I assume that Linux kernel driver changes must follow GPL and the patched driver should be made available public.

I own the E6510 and as far as I can tell it does not have an 'Instant ON' feature.

raboof, you should have it. One can buy it with the image preloaded from factory or flash it to the system from within the management SW for windows. More info on this technology for Dell E-series is here: http://latitude-on-flash.com
E6510 is listed here: http://latitude-on-flash.com/#tab-buy

Perhaps it is an option depending on Latitude E-series HW configuration. It is definitely Linux based.

I have a Dell E6410 and it has a button for Dell Latitude ON, but I do not have that feature. That is an option that may be ordered when configuring the machine, and I did not choose it. I think the Latitude ON Reader version is available for those who didn't receive the full version (which requires special hardware), but that runs from the hard drive and my HD is gone, replaced with an Intel SSD.

I assume the E6510 also has this feature available. If so, it's a button next to the power button labeled with something that looks like a combination of D, infinity (∞), and the power symbol (IEC5009 Standby symbol). I don't know if the Alps touchpad driver included with the stripped-down SUSE Linux that runs Latitude ON functions completely, but if anyone has that feature, please check it out.

Hey guess what--the source code is available :)

http://support.dell.com/support/downloads/download.aspx?c=us&releaseid=R261970&fileid=384807
http://ftp.us.dell.com/app/Dell_Latitude-ON-Flash-Softw_A05_R261970.exe

I'll be damned - turns out I indeed have this functionality - I had wondered what that button was for :).

Unfortunately, touchpad scrolling does not seem to work with it either.

I suppose most of people would agree a Latitude ON and all the other SplashTop derivatives is a useless feature, purely a marketing thing. However since the manufacturer does ship this feature in their flagship Laptops equipped with the ALPS "0x73, 0x02, 0x64" touchpad, they're obliged to provide a full touchpad experience, not just a crippled compatibility mode - there's even a web browser in the Latitude ON where standard scrolling touchpad features are used heavily.
Once it is confirmed the touchpad really does not work correctly within this environment, I'd suggest to escalate the issue to the Dell support classified as a critical bug subject of a money back guarantee.

I upgraded to 2.6.32-24 this morning, I noticed * Input: ALPS - add interleaved protocol support (Dell E6x00 series) in the changelog. This was an upstream kernel change. Still no scrolling - are we getting any closer?

That seems to be different hardware, see #296610 and http://patchwork.ozlabs.org/patch/42985

It is not clear that this "* Input: ALPS - add interleaved protocol support (Dell E6x00 series)"
patch is in 2.6.32-24 as there is also an entry "Revert "(pre-stable) Input: ALPS - add
interleaved protocol support (Dell E6x00 series)" in the ChangeLog (after if you read the
changelog from the bottom to top).

I upgraded to 2.6.32-24 too (Dell M4500), but the touchpad still does not scroll ... and does
not appear with lsinput and xinput -list (as it does with 2.6.32-22-generic #36-Ubuntu, but
not since 2.6.32-23 ...).

Thus I guess that 2.6.32-24 does not have this "interleaved protocol support".
 Why ?

Thanks for working on this problem,
Olivier

Latest kernel 2.6.35-rc6 is no good either. Touchpad only recognized as generic PS/2 mouse.

Just marked seven other bug reports as duplicates of this one since they are reporting the same basic problem -- the latest ALPS touchpad across an increasing range of hardware doesn't work properly.

I found this bug reported upstream here: https://bugzilla.kernel.org/show_bug.cgi?id=14660

I have the "73 02 64" multitouch/dualpoint device on my Dell E6510. In a very manual process of modifying the virtualbox-ose PS/2 driver, and copying bytes between it and /dev/serio_raw0, I managed to piece together the initialization sequence performed by the Windows driver. The entire sequence, from boot-up, is attached. I also included raw event data for three different finger swipes: left-to-right across the middle, top-to-bottom across the middle, and top-to-bottom across the right-edge scroll area.

Hopefully this helps in decoding the new protocol.

@Andrew Skalski

Maybe you should also post the same message on the linux bugzilla:

https://bugzilla.kernel.org/show_bug.cgi?id=14660

Let's hope your effort helps at getting the protocol decoded!!

@Andrew Skalski

Great job!

Could you elaborate a little more on the procedure that you used to read the raw data from the Windows driver?

Thanks

Awesome. I'm sure documenting how you managed to convince virtualbox to expose this information would be tremendously useful to many people trying to get their devices supported!

I am not sure how this patch affects what you guys are doing here, but it is new and impacts this device:

https://patchwork.kernel.org/patch/118834/

Johnny Hughes <email address hidden> writes:

> I am not sure how this patch affects what you guys are doing here, but
> it is new and impacts this device:
>
> https://patchwork.kernel.org/patch/118834/

That's really interesting. A patch coming from Dell with magic init
sequences is certainly the most promising progress I've seen so far on
this issue. What's the chance of this patch ending up in the default
Lucid kernel, if it works ? Would be great to get a slightly less
crippled touchpad.

I applied the Dell patch (I had to modify it somewhat to make it apply cleanly), and verified that it successfully enables vertical scrolling. It still doesn't support syndaemon, but it's a big improvement from what we had before. Attached is a 64-bit psmouse.ko that I built against 2.6.32-24-generic.

When I have some time, I'll see about putting together a description of how I used virtualbox to manipulate the PS/2 driver. Here's the short version:

# Enable logging in the kbd driver. Output will be written to ~/.VirtualBox/Machines/YourMachineName/Logs/Vbox.log
export VBOX_RELEASE_LOG=dev_kbd.e.l.l2.l3.f
VirtualBox [--args]

Snapshot your VM, just in case.

In the VM settings, disable "VT-x/AMD-V" (System -> Acceleration), then start up your VM.

The logs will look like:

00:00:10.810 kbd_write_mouse: kbd: write mouse 0xff
00:00:10.810 void kbd_queue(KBDState*, int, int): mouse command response: 0xfa
00:00:10.810 void kbd_queue(KBDState*, int, int): mouse command response: 0xaa
00:00:10.810 void kbd_queue(KBDState*, int, int): mouse command response: 0x00

This is the Windows driver conversing with VirtualBox's mouse emulation code (virtualbox-ose-3.1.6-dfsg/src/VBox/Devices/Input/DevPS2.cpp).

From that point, I replaced the kbd_write_mouse() function with my own version which, instead of emulating a mouse, would blindly reply to mouse commands by returning bytes from a hard-coded array. When it runs out of array data, reply to all further commands with 0xFA (OK).

At first, I started with an empty array of responses. Then I went through many iterations of:

1) Boot up the VM. Figure out what the Windows driver is trying to do next, after exhausting my array of reply data. Power off the VM.

2) Switch to Linux, send that new command to my touchpad via the serio_raw driver (see the script I posted to https://bugzilla.kernel.org/show_bug.cgi?id=14660). Read the touchpad's response, and add it to the end of my array.

3) Rebuild VirtualBox (debian/rules build), install the updated .so files (VBoxDD*.so) to /usr/lib/virtualbox.

4) Repeat.

Andrew Skalski <email address hidden> writes:

> I applied the Dell patch (I had to modify it somewhat to make it apply
> cleanly), and verified that it successfully enables vertical scrolling.
> It still doesn't support syndaemon, but it's a big improvement from what
> we had before. Attached is a 64-bit psmouse.ko that I built against
> 2.6.32-24-generic.

I tested it by rmmod'ing psmouse and insmod'ing the psmouse.ko file you
attached. It works well here, touchpad scrolling is enabled.

[10506.261681] alps.c: Enabled hardware quirk, falling back to psmouse-core
[10506.274889] input: ImPS/2 ALPS GlidePoint as /devices/platform/i8042/serio1/input/input16

(Dell Latitude E6510)

I did the same as Øyvind Stegard in #141 to test the psmouse.ko posted by Andrew Skalski in #140 (https://bugs.launchpad.net/ubuntu/+source/linux/+bug/550625/+attachment/1488476/+files/psmouse.ko) on my Dell E6410. Touchpad scrolling is enabled. Thanks Andrew!

Unfortunately, Øyvind Stegard's module in #141 is not compatible with the patched version of 2.6.32-24-generic (https://bugs.launchpad.net/ubuntu/+source/linux/+bug/578673/comments/65) created by Manoj to address the E6410 Intel video card suspend/resume Bug #578673. Upon resume, the mouse becomes erratic and unusable. So, I can either scroll or suspend but not both :-(.

Sorry, in comment #143 I should have said "Andrew Skalski's module in #140" instead of "Øyvind Stegard's module in #141".

Unfortunately, Andrew Skalski's module in #140 is not compatible with the patched version of 2.6.32-24-generic (https://bugs.launchpad.net/ubuntu/+source/linux/+bug/578673/comments/65) created by Manoj to address the E6410 Intel video card suspend/resume Bug #578673. Upon resume, the mouse becomes erratic and unusable. So, I can either scroll or suspend but not both :-(.

I installed the patched psmouse on my e6410 as well and it works. I haven't tried suspend resume lately but I'll give that a shot too.

Good to finally see progress on this.

@pvillela

I tried suspending, and got the same erratic pointer behavior. Looking at the code, the patch bails out of alps_init() before setting up reconnect/disconnect handlers. The hardware initialization is called from the wrong place (alps_init, instead of alps_hw_init) too, so simply assigning the reconnect handler wouldn't help without further changes. I checked back at the patchwork site, but there were no updates.

Possible workaround is to unload and reload the module on resume. In Ubuntu create the following new file named e.g. "90-psmouse-reload" in "/etc/pm/sleep.d/" with the following content:

#!/bin/sh
case "$1" in
    thaw|resume)
        rmmod psmouse; modprobe psmouse
    ;;
esac
exit 0

However I still consider this Dell contributed driver being the first step. I consider features like horizontal scrolling and sensitivity settings essential. IMO the touchpad by default is hypersensitive and it happens very often than one accidentally invokes a double-click action. While clicking the mouse button (right above the touchpad surface) the thumb accidentally touches the touchpad surface. The resulting action is two subsequent clicks, first one wanted, second one unwanted.

I can verify the same on my E6510, vertical scrolling works but erratic mouse after hibernation or suspend. Glad to see the progress. Thanks...

Skalsi's psmouse.ko file (post #140) works wonderfully on kernel 2.6.32-24-generic, 64-bit on e6510. Unfortunately, I also had the wild cursor problem with resume from suspend. Fortunately, mikeos's solution (post #147) solved that problem as well. However, don't forget to chmod +x the 90-psmouse-reload file to make it executable.

Thank you all for the help. If and when this touchpad is fully functional (tapping, syndaemon, etc) I'll be much happier.

I am pretty new to ubuntu and linux in general, so I dont know how to do any of the fancy stuff you guys are talking about here to report bugs. But I would just like to say that I have a very similar problem with my Gateway i5 NV-Series Laptop.

The touchpad is recognized as a Imps/2 generic mouse wheel. I cannot scroll down using the touchpad, but if i scroll up it automatically scrolls to the bottom of the page. It's a very odd problem. Furthermore, I am able to lock the touchpad when I hit the lock button on the laptop, which complicates things further seeing as it's not even recognized as a touchpad.

Hope this gets solved soon.

Back from vacation. I rewrote the Dell patch, and fixed the suspend/resume bug. I couldn't find any way to enable hardware tapping or scrolling within absolute mode -- those features are apparently handled by the driver for this device. So, until software support is written for Linux, we're stuck with IMPS emulation mode. I did, however, figure out how to disable/enable hardware tapping in IMPS mode, and left it as a comment inside the patch.

The attached patch is against the Ubuntu-2.6.32-24.40 tag.

Attached updated psmouse.ko for x86_64

Andrew, your patch does not apply cleanly on Ubuntu Maverick kernel 2.6.35. As you wrote in #140 you had to modify the Dell patch, it was not the case with 2.6.35 where it did apply cleanly right away. The problem is in "alps.c", especially alps_hw_init function which is significantly different. Could you please have a look at this kernel?

Here's a patch against 2.6.35. I moved the reset out of initialization and into disconnect, and rewrote the "PSMOUSE_IMPS" hack a little differently (the original Dell patch would change psmouse->type, even when it wasn't supposed to do so.)

I wonder if this driver should be in its own .c file, separate from alps.c, since the new protocol is so different. Maybe that's something to think about if somebody decides to implement the 6-byte absolute protocol.

I tested docking/undocking from a station with a PS/2 mouse attached, and it behaved more or less sanely. While docked, the touchpad falls back to generic PS/2 mode, however.

Hi Andrew,

I think your work on getting this trackpad to function based on the original patch from Dell is great!

Would you feel comfortable sending your patch on to <email address hidden>? That list is where the Ubuntu kernel team manages patches that go into the official kernel. If you are interested, the sign-up page is at: https://lists.ubuntu.com/mailman/listinfo/kernel-team.

Thanks!

P.S.: I realized I'm still assigned to this bug, but I don't have time to shepherd it through right now. I'm unassigning myself.

Would it be possible to get Andrew's patch rolled into kernel-ppa? I'm running Lucid with backported kernel maverick-2.6.35-17-generic-pae, and this driver still doesn't work.

Attached psmouse.ko module based on latest patch by Andrew; this version has tapping disabled via modified extra bits. I suppose mainly GlidePoint users are negatively impacted by touchpad's hypersensitivity (#147), so we'd better turn tapping off.

If this patch is going to go to the mainline kernel, it should be made clear how to control the extra bits of IMPS mode settings from a userspace program (synclient?).

Confirming this issue on a Dell E5410.

Andrew - Thanks for your work on this patch! It solved vertical scrolling for me.

Now to figure out suspend-resume/wireless connection display

@togume #158:

Check out Bug 594837. That might be your suspend/resume problem on a Dell E5410 (was for me on an E6410). That fix should be released very soon, though.

Andrew's latest patch and psmouse module build for 2.6.32-24 x86-64 works great on my E6510. Suspend/resume OK as well. Thanks.

Could somebody build Andrew's psmouse module for 32bit, or give me instructions on how to do it myself? Thanks!

I will try that out for sure on my 32bit build... So when I have the time and it works well, I can provide you a 32bit kernel build with that patched module.

I you do not like to wait, her is a full kernel Compile HowTo, which should answer all questions:
http://kernel-handbook.alioth.debian.org/

ubuntu documentation on this issue is here: https://help.ubuntu.com/community/Kernel/Compile

In short and easy:
1. You need some tools to compile, which ones is explained in the docs I linked.
2. You need kernel Sources for the kernel you like to build. Stable kernel sources can be downloaded via apt-get. mainline Builds with sources can be found here: http://kernel.ubuntu.com/~kernel-ppa/mainline/
3. You need to apply the patches to the kernel you like to recompile. You can do that manually by reading the patch file. It gives a path/name of a file in the kernel package and lines to add to that file (marked with ++) and lines to be deleted (marked with --). Do the changes.
4. Build the kernel with the patched module from source.
5. Restart with the newly compiled kernel.

Hopefully this gives you an overview. There are Forums which can answer your questions better, this is for bugtracking ;)
Have Fun!

This bug also affects my Acer Timeline 1830T-3721.

I'm running the 64-bit version of Lucid (Kubuntu & KDE); uname -a shows:
Linux xxx 2.6.32-24-generic #41-Ubuntu SMP Thu Aug 19 01:38:40 UTC 2010 x86_64 GNU/Linux

Applying Andrew Skalski's psmouse.ko in comment #154 is a partial, and very much appreciated, fix.

After adding it (sudo rmmod psmouse; sudo insmod psmouse.ko), xinput show that the ALPS touchpad is recognized:
xinput list
⎡ Virtual core pointer id=2 [master pointer (3)]
⎜ ↳ Virtual core XTEST pointer id=4 [slave pointer (2)]
⎜ ↳ Macintosh mouse button emulation id=12 [slave pointer (2)]
⎜ ↳ ImPS/2 ALPS GlidePoint id=11 [slave pointer (2)]
⎣ Virtual core keyboard id=3 [master keyboard (2)]
    ↳ Virtual core XTEST keyboard id=5 [slave keyboard (3)]
    ↳ Power Button id=6 [slave keyboard (3)]
    ↳ Video Bus id=7 [slave keyboard (3)]
    ↳ Sleep Button id=8 [slave keyboard (3)]
    ↳ 1.3M WebCam id=9 [slave keyboard (3)]
    ↳ AT Translated Set 2 keyboard id=10 [slave keyboard (3)]

However, the touchpad settings in System Settings | Keyboard & Mouse | Touchpad continue to be grayed-out.

Prior to adding Mr. Skalski's psmouse.ko, the touchpad supported only mouse movement and tap-to-click; after adding it, it additionally supports one-finger vertical scrolling in the right-edge vertical scrolling area. Two-finger anything (scroll, horizontal scroll, tap) is not supported; unfortunately, these are the features I most desire to have.

Oddly, it supports right-click if tapped at one spot along the top edge of the touchpad, immediately to the left of the right-edge scroll-area (?!); this is unexpected and mostly useless, as it's hard to consistently hit this precise spot. (I also set this area up as a right-click "hotspot" under under MS-Windows; is it possible the MS-Wndows driver communicates settings to the touchpad hardware, and the touchpad hardware retains settings across reboots??)

Running xinput list-properties on the touchpad shows the following:

xinput list-props 11
Device 'ImPS/2 ALPS GlidePoint':
        Device Enabled (125): 1
        Device Accel Profile (242): 0
        Device Accel Constant Deceleration (243): 1.000000
        Device Accel Adaptive Deceleration (245): 1.000000
        Device Accel Velocity Scaling (246): 10.000000
        Evdev Reopen Attempts (240): 10
        Evdev Axis Inversion (247): 0, 0
        Evdev Axes Swap (249): 0
        Axis Labels (250): "Rel X" (133), "Rel Y" (134)
        Button Labels (251): "Button Left" (126), "Button Middle" (127), "Button Right" (128), "Button Wheel Up" (129), "Button Wheel Down" (130), "Button Horiz Wheel Left" (131), "Button Horiz Wheel Right" (132)
        Evdev Middle Button Emulation (252): 2
        Evdev Middle Button Timeout (253): 50
        Evdev Wheel Emulation (254): 0
        Evdev Wheel Emulation Axes (255): 0...

Correction: I wrote "Applying Andrew Skalski's psmouse.ko in comment #154", I should have written comment #152.

Hi Guys,

I'm not an Ubuntu user. But my wife's laptop (Gateway NV59) has a newer Synaptics/Alps touch pad that isn't recognized by alps.c, just shows up as a regular ps2 mouse. I applied the patch from dell against kernel 2.6.35.2. After which the touch pad was not functional, but /proc/bus/input/devices did list it as a Synaptics/Alps touch pad. This was with psmouse built into the kernel. Tried the patch from comment 154 with kernel 2.6.35.4 and built it as a module this time. Now the touch pad on the NV59 is functional and is listed in /proc/bus/input/devices but seems about the same as without the patch. I also tried this same kernel on a Dell 1545, which is well supported by the normal module and Xorg synaptic driver. The touch pad shows up in /proc/bus/input/devices but the Xorg synaptics driver can't work with it. Might want to do some more testing before dropping this in a release kernel.

Oh well, at least Dell is doing something. Thanks for everyone's efforts.

Stupid noob question here: every time I reboot, to install Andrew Skalski's modified psmouse.ko, I have to sudo rmmod psmouse; sudo insmod psmouse.ko.

How can I get Skalski's psmouse.ko loaded automatically?

TP Diffenbach <email address hidden> writes:

> Stupid noob question here: every time I reboot, to install Andrew
> Skalski's modified psmouse.ko, I have to sudo rmmod psmouse; sudo insmod
> psmouse.ko.
>
> How can I get Skalski's psmouse.ko loaded automatically?

Well, you could put these two lines in /etc/rc.local:

modprobe -r psmouse || true
insmod /path/to/hacked/psmouse.ko || echo 'Failed to load hacked psmouse.ko, probably needs recompiling.'

Haven't tested it, though. I'll rather wait till the fixes are
incorporated into the default kernel. You'll need to be aware when
kernel [ABI] upgrades happen. This approach will fail in that scenario
(if new kernel version requires module recompile).

Øyvind
--
< Øyvind Stegard
 < http://folk.uio.no/oyvinst/

Øyvind Stegard <email address hidden> writes:
> modprobe -r psmouse || true
> insmod /path/to/hacked/psmouse.ko || echo 'Failed to load hacked
> psmouse.ko, probably needs recompiling.'

Perhaps slightly better is to fallback to the regular psmouse module, if
the other one fails to load:

modprobe -r psmouse || true
insmod /path/to/hacked/psmouse.ko || { echo 'Failed to load hacked psmouse.ko, falling back to standard.'; modprobe psmouse; true; }

Øyvind
--
< Øyvind Stegard
 < http://folk.uio.no/oyvinst/

@Brian #159 - Thanks. This was indeed fixed with the latest kernel release 2.6.32-24. I was tracking a duplicate one https://bugs.launchpad.net/ubuntu/+source/linux/+bug/578673

@TP Diffenbach - If it's working well enough you might consider replacing the default psmouse.ko with the fixed one. Just remember to rename the old one and keep it close in case you need to fall back on it. This way it will be enabled on reboot.

Again, thanks for the work on this!

For those not running 64-bit Ubuntu, here's a 32-bit version of the driver, compiled against 2.6.32-24-generic-pae.

This is based on Andrew Skalski's patch in #154, with a couple of changes to work on 2.6.32 (probably the same as his patch in #151, which I missed). This patch also disables tap-to-click as well as tapping the top right corner as a right-click.

I'm using this on a Dell E6510 without any problems.

Cheers ............................... JD

Tested Andrew's psmouse.ko (see #152) on my E6410 (i5, Intel video) with standard Ubuntu kernel 2.6.32-24-generic. Scrolling is enabled and suspend-resume works (with the known inconvenience that the machine resumes with a very dim screen -- workaround: increase screen brightness manually). Thanks to Andrew for the great work.

No trackpad scrolling in 10.10 beta on my Latitude E6410 for me

As a Debian Squeeze user, I wanted to let you all know that Andrew's patch from #151 did the trick. I've got vertical scrolling agian. I can't configure the touchpad using gsynaptics, but I don't really care as long as vertical scrolling is back.

I manually patched the 2.6.32 source using the 'make' instructions from https://confluence.nau.edu/display/\~cmg238\@nau.edu/Recognize+ALPS+Touchpad+on+Dell+E6510+in+Ubuntu

On a side note, this kind of rapid development has me wondering if I should switch back to Ubuntu. I couldn't find *anyone* on the Debian side doing anything to try and fix this ALPS Touchpad issue. Kudos to Andrew and everyone else in this thread!

Being part of the largest Linux desktop community does have its advantages. Of course, we're all part of the even larger Debian-based community and the upstream work shouldn't be minimalized either.

Anyone know if this workaround patch is slated to make its way into the auto-updates or should I hold my breath for a real fix to the problem?

Manoj Iyer from Canonical has recently introduced the original Dell patch to the Maverick kernel 2.6.35-22. It means there exists the above discussed issue with erratic mouse behaviour on resume. There are no extra features implemented which were brought by Andrew to control things like tapping disable.

The faulty patch:
http://kernel.ubuntu.com/git?p=ubuntu/ubuntu-maverick.git;a=commitdiff;h=bd79df42218fe1db5e683b567791d4bb97b60685

I notified Manoj about the existence of Andrew's patch. So attention, Andrew's patch currently does not apply cleanly since 2.6.35-22. Meanwhile I was able to successfully apply and compile the patch against 2.6.35-22 by patching drivers/input/mouse sources from 2.6.35-21. Andrew's patch also applied cleanly and works correctly on 2.6.36rc4.

The Dell patch has been reverted in 2.6.35-22.33. As a result Andrew's patch applies cleanly again. It's already the period of kernel freeze for Maverick so it's unlikely that Maverick would get the support out of the box :/ Blame Alps for not willing to publish the new protocol specifications.

Did anyone actually manage to contact someone at ALPS?

raboof: http://kernel.ubuntu.com/git?p=ubuntu/ubuntu-maverick.git;a=commitdiff;h=bd79df42218fe1db5e683b567791d4bb97b60685

mikeos <email address hidden> writes:

> raboof: http://kernel.ubuntu.com/git?p=ubuntu/ubuntu-
> maverick.git;a=commitdiff;h=bd79df42218fe1db5e683b567791d4bb97b60685

Hope the Dell patch (or derived/improved works) ends up getting included
with Ubuntu anyway, if it doesn't break anything.. Scrolling is better
than nothing.

Please disregard my comment (#165). Idiot error on my part. I had changed 0x0005 to 02, thinking that would disable tapping, but that disabled IMPS too. Set it to 03, now everything works great. Sorry for the bad report. Thanks for the patch Andrew!

psmouse.ko, compiled against 2.6.36-rc5 with Andrew's patch applied.

I have a similar problem, I think. I have an ASUS Eee 1015PED. The touchpad works fine, even all the capacitive multitouch things, like the two-fingered scrolling and the three-fingered right click, but there is no Touchpad Tab in Preferences -> Mouse, so I can't turn on the disable-touchpad-while-typing thing, or turn off tap-to-click...both problems are incredibly annoying to me, because my mouse flies around when I'm trying to type, and I always inadvertently click things with the touchpad while moving my finger.

I tried running syndaemon, but it says it can't find anything. I tried gsynaptics, but it says it needs 'SHMClient' 'true' enabled in xorg.conf. That's not going to fly because there is no xorg.conf in 10.04 because there's no hal; it's all udev now, which I have no idea how to configure. I tried fiddling with synclient but it said there wasn't any driver to connect to.

I think that, at least on my machine, the touchpad is operating on a lower level than udev is aware of, because my scrolling is working, but the Touchpad tab in Mouse is absent. This suggests udev is seeing it as a generic mouse, and some other driver or something is passing the touchpad-type stuff to the higher levels. For example, to get the Fn keys working on my 1015, I had to use a kernel parameter. Does getting the touchpad to work require a kernel parameter?

I was directed to come here by looking at this: http://markmail.org/message/y5zvwpemhd63dfj5

Some outputs:

lsmod | grep psmouse
psmouse 63245 0
-
dmesg | grep mouse
[ 0.392123] input: Macintosh mouse button emulation as /devices/virtual/input/input4
[ 0.586860] mice: PS/2 mouse device common for all mice
[ 14.467520] psmouse serio1: ID: 10 00 64

Thanks to Andrew #154 for the patch.
I applied it and fiddled with the codes on line 175 of alps.c

with:
alps_ec_write(psmouse, 0x0005, 0x01)

scrolling works, tapping as well.

although:
alps_ec_write(psmouse, 0x0005, 0x02)
which should disable tapping leads to unusable touchpad behaviour. It clicks randomly all the time. :-S

Have I missed something? This is a Dell E6510 with 2.6.35-22-generic-pae (32-bit)

@John (#170) How did you disable tapping entirely for the 10.04-kernel?

Sorry that I haven't had the time to follow up on this in a while. If upstream is holding out for multitouch specs, then I wouldn't hold my breath waiting. From what I could tell, when I looked at it, the multitouch support in Windows is handled by the user-space driver, and not by the hardware. So if we want that support, we're on our own. I've decoded enough of the protocol that a full-featured driver can be written (with the possible exception of sensitivity adjustment; though based on the Windows driver initialization, that *might* be the job of registers 0x14d and 0x163.) But the implementation will need to handle basically everything: tapping, dragging, scrolling, gestures, etc.

@mättu (#183): Try setting register 0x0005 to the value 0x03 (IMPS emulation + Disable hardware tap)

@mättu (#183): I used:
    alps_ec_write(psmouse, 0x0005, 0x07)
to disable the tapping completely (this also disables the corner tap as a right click).

One of the earlier patches (from @Andrew?) had notes about what the different bits controlled - I based mine on that documentation.

works, great.

I edited the patch according your inscructions (line 175) and changed the git diff lilnes so that it can be applied directly in ~/linux-2.6.35/drivers/input/mouse/

Here is what I had to do on delle6510 with ubuntu 10.10 2.6.35-22-generic-pae:
(to remind myself of my first kernel module patching session) ;-)

Install kernel headers:
 sudo apt-get build-dep --no-install-recommends linux-image-$(uname -r)
 apt-get source linux-image-$(uname -r)
this puts a "linux-[your-version]" directory into your home

install git:
 sudo apt-get install git

change into it:
 cd src/drivers/input/mouse

backup alps.c, alps.h, psmouse.h

apply patch:
 patch -p0 < patch-dell-e6510

make & install:
 make -C /lib/modules/`uname -r`/build M=`pwd` psmouse.ko
 sudo rmmod psmouse
 sudo cp psmouse.ko /lib/modules/`uname -r`/kernel/drivers/input/mouse/
 sudo modprobe psmouse

thanks everybody very much!
:m)

.. and here ist the compiled psmouse.ko for kernel 2.6.35-22-generic-pae (32-bit)

 sudo rmmod psmouse
 sudo cp psmouse.ko /lib/modules/`uname -r`/kernel/drivers/input/mouse/
 sudo modprobe psmouse

I did just that but then the touchpad doesn't work at all. What did I do wrong?
jc@jc-netbook:~$ cd ./Skrivbord
jc@jc-netbook:~/Skrivbord$ sudo cp psmouse.ko /lib/modules/`uname -r`/kernel/drivers/input/mouse/
jc@jc-netbook:~/Skrivbord$ sudo modprobe psmouse
FATAL: Error inserting psmouse (/lib/modules/2.6.35-22-generic/kernel/drivers/input/mouse/psmouse.ko): Invalid module format
jc@jc-netbook:~/Skrivbord$

What do you mean, "just that"? Did you download mättu's module or did you compile it yourself? The psmouse.ko module posted by mättu is compiled for kernel 2.6.35-22-generic-pae 32bit and might not work with other kernels. I hope you have a backup of /lib/modules/`uname -r`/kernel/drivers/input/mouse/psmouse.ko so you can place the original one back.

Just that: I downloaded Mättu's module and the kernel I have is 2.6.35-22
and obviously I should have made a backup of /lib/modules/`uname -r`/kernel/drivers/input/mouse/psmouse.ko

But I'm quite new to Ubuntu, so I didn't think of it...

Note that Mättu's kernel is a PAE kernel (generic-pae), so that might be why it doesn't work for you. Anyways, here's my psmouse.ko for 2.6.35-7, you might give it a try but I don't know if it will work.

well I upgraded to 10:10 past Sunday, and check the kernel and got the answer that it was 2.6.35-22 before following Mattus' advice. But today the update system told me that there were new kernel headers. After updating my touchpad is back but with the same problem that gpointing devices doesn't recognise it and I can't disable it when typing....
My laptop is a HP Pavilion dm1-2010so.

xinput -list
⎡ Virtual core pointer id=2 [master pointer (3)]
⎜ ↳ Virtual core XTEST pointer id=4 [slave pointer (2)]
⎜ ↳ PS/2 Generic Mouse id=12 [slave pointer (2)]
⎜ ↳ Macintosh mouse button emulation id=13 [slave pointer (2)]
⎣ Virtual core keyboard id=3 [master keyboard (2)]
    ↳ Virtual core XTEST keyboard id=5 [slave keyboard (3)]
    ↳ Power Button id=6 [slave keyboard (3)]
    ↳ Video Bus id=7 [slave keyboard (3)]
    ↳ Power Button id=8 [slave keyboard (3)]
    ↳ Sleep Button id=9 [slave keyboard (3)]
    ↳ HP Webcam-50 id=10 [slave keyboard (3)]
    ↳ AT Translated Set 2 keyboard id=11 [slave keyboard (3)]

Dell Latitude E4310 same Problem.

Linux grisu-nb 2.6.35-22-generic #35-Ubuntu SMP Sat Oct 16 20:45:36 UTC 2010 x86_64 GNU/Linux

xinput -list
⎡ Virtual core pointer id=2 [master pointer (3)]
⎜ ↳ Virtual core XTEST pointer id=4 [slave pointer (2)]
⎜ ↳ Logitech USB Receiver id=11 [slave pointer (2)]
⎜ ↳ PS/2 Generic Mouse id=14 [slave pointer (2)]
⎣ Virtual core keyboard id=3 [master keyboard (2)]
    ↳ Virtual core XTEST keyboard id=5 [slave keyboard (3)]
    ↳ Power Button id=6 [slave keyboard (3)]
    ↳ Video Bus id=7 [slave keyboard (3)]
    ↳ Power Button id=8 [slave keyboard (3)]
    ↳ Sleep Button id=9 [slave keyboard (3)]
    ↳ Logitech USB Receiver id=10 [slave keyboard (3)]
    ↳ Laptop_Integrated_Webcam_2M id=12 [slave keyboard (3)]
    ↳ AT Translated Set 2 keyboard id=13 [slave keyboard (3)]
    ↳ Dell WMI hotkeys id=15 [slave keyboard (3)]

Works great for me on E6410 Maverick 2.6.35-22-generic X86

Here is my psmouse.ko

After almost any Kernel related updates one needs to redo the steps described in #186. I'm saying this as a heads-up to people who don't have success trying the psmouse.ko's attached here, and as an explanation to why it stopped working "all of a sudden".

What I don't understand is how is it that such a long standing bug is not addressed by Ubuntu devs even if it's a temporary fix as this one. Fingers crossed for a solid fix though...

Thank you to all who contributed in this!

Does anyone have a psmouse.ko for 2.6.35-22-generic 64bits ?

And, for now, we have to do this manipulation at each kernel update ?

Here is my psmouse.ko for 2.6.35-22-generic 64bits.

Scrolling is back.
But I've lost click by tapping...

and I just find that file and (backup/)replace it with yours ?

"
aftertaf@tranquility:/lib/modules/2.6.36-020636rc8-generic/kernel/drivers/input/mouse$ sudo modprobe psmouse
FATAL: Error inserting psmouse (/lib/modules/2.6.36-020636rc8-generic/kernel/drivers/input/mouse/psmouse.ko): Invalid module format
"

Fixable?
or not compatible with this kernel ?

As few people seem to be capable of compiling the KO (amongst them is ME !)

Could somebody who knows well sends us a link or brief explanation how to do
that

Could somebody do that ? I could then compile mine on a:

2.6.32-25-generic ....... x86_64 GNU/Linux

Good work for the "patchers" !!

Hi,

This post from @remco on a related bug (https://bugs.launchpad.net/ubuntu/+source/linux/+bug/606238) explains it all:
http://confluence.nau.edu/display/~<email address hidden>/Recognize+ALPS+Touchpad+on+Dell+E6510+in+Ubuntu

BR,
Steven

in my case (latitude e6510) the patch and the instructions provided here fix the scrolling issue: http://<email address hidden>/msg2560539.html

In my dell precision M4500 with ubuntu 10.10 64bits problem is still present.

I tried also kernel 2.6.36 from:
http://kernel.ubuntu.com/~kernel-ppa/mainline/v2.6.36-maverick/
without luck. The above kernel solves other suspend/resume problems on dell M4500 (bug 563318).

Applying the patch from Andrew Walski #154 solves the scrolling problem, even if the touchpad is still not recognized as a touchpad in the "mouse" control panel.
I attach here the psmouse.ko module compiled against kernel 2.6.36 64bits

a side note.

i can see that in 64bit lucid with 2.6.32-21-generic dmesg reports the ALPS pad, but in 2.6.32-24-generic and beyond (2.6.32-25-generic) the pad is not recognised.

So there has been a regression within the kernel.

Now, while the pads is recognised in 2.6.32-21-generic, and the track pad configuration tab appears correctly within mouse configuration, it still does not support scrolling.

hsiboy

in 2.6.36 64 bit maverick, I confirm that the touchpad is broken: not recognised and usable just as a mouse (moving cursor, tap = click) but no scrolling.

and I also confirm that psmouse.ko patched above by bagl0312 enables scrolling :)

I also confirm that psmouse.ko patched above by bagl0312 enables scrolling on my HP pavilion dm1-2010so. But still the GP pointing device settings doesn't reconise the touchpad...

I also confirm that on my Dell E4310 the patched module proposed by bagl0312 in #203 does not chenge anything. The pointing device is still not recognised properly.

xinput -list
⎡ Virtual core pointer id=2 [master pointer (3)]
⎜ ↳ Virtual core XTEST pointer id=4 [slave pointer (2)]
⎜ ↳ PS/2 Generic Mouse id=11 [slave pointer (2)]
⎣ Virtual core keyboard id=3 [master keyboard (2)]
    ↳ Virtual core XTEST keyboard id=5 [slave keyboard (3)]
    ↳ Power Button id=6 [slave keyboard (3)]
    ↳ Video Bus id=7 [slave keyboard (3)]
    ↳ Power Button id=8 [slave keyboard (3)]
    ↳ Sleep Button id=9 [slave keyboard (3)]
    ↳ AT Translated Set 2 keyboard id=10 [slave keyboard (3)]
    ↳ Dell WMI hotkeys id=12 [slave keyboard (3)]

It works with the basic functionalities but I cannot set scrolling nor change its sensitivity. Actually until the bug is fixed what I would like to do is to switch the touchpad off and leave only the pointer in the middle of the keyboard. Any suggestion on how this can be made? The GPointing device setting windows shows me just one mouse (PS2).

Thanks

Luca

Sorry!!! Wrong message. I had a wrong options for the module. With the patch the mouse is recognised as an ALPS Glidepoint and scrolling works fine.

I do apologise for the wrong message.

Luca

I can confirm that the patch in https://bugs.launchpad.net/ubuntu/+source/linux/+bug/550625/comments/154 works for my Dell Latitude e4310 with the latest 10.10 kernel.

I am using Ubuntu 10.04 (lucid) on a DELL PRECISION M4500 (company notebook). Due to company policy I cannot upgrade to maverick (in fact I am grateful not to be forced to used plain lenny like many others :-) ) ...

Patch from #154 doesn't work on my present 2.6.32-25-generic kernel - something causes panic.

However, this patch:
https://patchwork.kernel.org/patch/118834/
does, with limited functionality.

Attached is a clean version of that patch that works against lucid kernels.