Welcome, Guest. Please login or register.
Did you miss your activation email?

Author Topic:  (WORKDED AROUND) xserver-xorg update -- touchpad cursor speed slow  (Read 21031 times)

1sy8

  • Guest
Hi reinob, what is SNA and how did you make it work with your conf file?


Dilb, in /usr/share there are two candidate synaptics conf files. Maybe leftovers from when xorg didn't autoconfigure but maybe fine too. I'll give them a try later in the afternoon by dropping both files in /etc/X11/xorg.conf.d (maybe the prefix numbers 50 and 51 set the priorities?)

Offline dibl

  • siduction community member
  • Global Moderator
  • User
  • *****
  • Posts: 2.403
    • Land of the Buckeye
Dilb, in /usr/share there are two candidate synaptics conf files. Maybe leftovers from when xorg didn't autoconfigure but maybe fine too. I'll give them a try later in the afternoon by dropping both files in /etc/X11/xorg.conf.d (maybe the prefix numbers 50 and 51 set the priorities?)


Right, I saw those, but I don't see reference to mouse acceleration in them.  Since I got a good result with the /etc/X11/xorg.conf.d/synaptics.conf settings, I don't need to try further changes to an otherwise good-performing system.
System76 Oryx Pro, Intel Core i7-11800H, SSD 970 EVO Plus;  Asus ROG STRIX X299-E, Core i7-7740X, Nvidia GTX-1060, dual monitors, SSD 860 EVO

reinob

  • Guest
Hi reinob, what is SNA and how did you make it work with your conf file?

Have a look here: https://wiki.archlinux.org/index.php/Intel_Graphics
("choose acceleration method").

But AFAIK SNA is already the default (maybe not in "stable" distributions..). In my case this is a leftover from an old Ubuntu 12.04 installation. If you want,you can create a /etc/X11/xorg.conf.d/sna.conf file with these contents:

Code: [Select]
Section "Device"
        Identifier "Card0"
        Driver "intel"
        Option "AccelMethod" "sna"
EndSection

If something goes wrong, delete the file and restart X.

Quote
Dilb, in /usr/share there are two candidate synaptics conf files. Maybe leftovers from when xorg didn't autoconfigure but maybe fine too. I'll give them a try later in the afternoon by dropping both files in /etc/X11/xorg.conf.d (maybe the prefix numbers 50 and 51 set the priorities?)

I don't have the 51-file (probably is a siduction-specific file, in view of the name). I'd be careful mixing two separate config files for the same thing (may or may not be a problem). You can always "man synaptics" and tweak all you like.

Offline piper

  • User
  • Posts: 1.786
  • we are the priests ... of the temples of syrinx
Most of what I read about "sna" is not great, in fact, most end up going back to "uxa" which also has 4-5 times less code.

Most (google search of only 7 pages) have said performance degraded

This was probably the best I read

https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=724510

Free speech isn't just fucking saying what you want to say, it's also hearing what you don't want to fucking hear

I either give too many fucks or no fucks at all, it's like I cannot find a middle ground for a moderate fuck distribution, it's like what the fuck

Offline dibl

  • siduction community member
  • Global Moderator
  • User
  • *****
  • Posts: 2.403
    • Land of the Buckeye
My Toshiba netbook has a GMA945, and the sna method provides absolutely no benefit whatever, over the default i915 driver.
System76 Oryx Pro, Intel Core i7-11800H, SSD 970 EVO Plus;  Asus ROG STRIX X299-E, Core i7-7740X, Nvidia GTX-1060, dual monitors, SSD 860 EVO

1sy8

  • Guest
Ok, I also setlled on reinolb's workaround: it works!!! I'll check on next d-u if I can revert.


BTW, dibl, how did you get your system details on the previous page? I need to check that I have a "discrete" Intel graphic (there is supposed to be one on this laptop and I should be able to switch it on and off as needed, to save on the battery life for example). lspci didn't tell me much about it.

Offline devil

  • Administrator
  • User
  • *****
  • Posts: 4.877
inxi -h


greetz
devil

reinob

  • Guest
Quick note: version 1.8.0-1 (bringing new libevdev2 package with it) is available. I dist-upgraded and can report that:

1) touchpad settings (accel) are respected again

EDIT: but still need the synaptics.conf workaround. Without it, the touchpad is again unbearably slow.

2) the new acceleration scheme is not so nice. seems to "slowly" deccelerate when you lift the finger, so that the precision is (much) lower.

This is probably tweakable (and I'll see if I can improve it). Otherwise I recommend either not upgrading the synaptic driver or at least saving the .deb (/var/cache/..) in case you want to rollback.

Offline dion

  • User
  • Posts: 38
After doing a d-u today, my touchpad stopped functioning completely. An external mouse works.
I copied /usr/share/X11/xorg.conf.d/51-synaptics-siduction.conf to /etc/X11/xorg.conf.d/synapics.conf and rebooted but that didn't help. Any clues on what I could try?

Offline dibl

  • siduction community member
  • Global Moderator
  • User
  • *****
  • Posts: 2.403
    • Land of the Buckeye

EDIT: but still need the synaptics.conf workaround. Without it, the touchpad is again unbearably slow.



Right -- if you look at the changelog file in the updated package, it had nothing to do with acceleration settings or a bug related to certain devices.


With the /etc/X11/xorg.conf.d/synaptics.conf settings as a workaround, the relevant output in /var/log/Xorg.0.log is rather confusing, at least to me.  It appears that two different configuration files are being parsed, or perhaps two different instances of the process are running -- I dunno.  But the settings in /etc/X11/xorg.conf.d/synaptics.conf are being respected in the running X, even though there is another set of settings shown later in the Xorg.0.log file.  So, the workaround is working.
« Last Edit: 2014/07/13, 19:38:50 by dibl »
System76 Oryx Pro, Intel Core i7-11800H, SSD 970 EVO Plus;  Asus ROG STRIX X299-E, Core i7-7740X, Nvidia GTX-1060, dual monitors, SSD 860 EVO

Offline devil

  • Administrator
  • User
  • *****
  • Posts: 4.877
Rumours have it, this was fixed yesterday with a new xserver-xorg-synaptics package.


greetz
devil

Offline dion

  • User
  • Posts: 38
It turns out my problem is related to the kernel. With the d-u I did yesterday, the 3.15-5 kernel was installed and the touchpad does not work with that kernel. If I boot from the 3.14 kernel, the touchpad works perfectly well.

Offline piper

  • User
  • Posts: 1.786
  • we are the priests ... of the temples of syrinx
Quote from: dion
It turns out my problem is related to the kernel. With the d-u I did yesterday, the 3.15-5 kernel was installed and the touchpad does not work with that kernel. If I boot from the 3.14 kernel, the touchpad works perfectly well.
As asked in irc can you try kernel  3.16-rcX
« Last Edit: 2014/07/14, 16:01:24 by piper »
Free speech isn't just fucking saying what you want to say, it's also hearing what you don't want to fucking hear

I either give too many fucks or no fucks at all, it's like I cannot find a middle ground for a moderate fuck distribution, it's like what the fuck

Offline dion

  • User
  • Posts: 38
@piper: That kernel doesn't seem to be in the repo. So I have to download it and install manually I suppose?

reinob

  • Guest
@dion,

I'm running 3.15.5 (from kernel.org, self configured) and this doesn't seem to be the cause of the problem.
Can you check your /proc/config.z and see if the relevant modules are enabled?

# zcat /proc/config.gz |grep -i -e evdev -e synap
CONFIG_INPUT_EVDEV=m
CONFIG_MOUSE_PS2_SYNAPTICS=y
CONFIG_MOUSE_SYNAPTICS_I2C=m
CONFIG_MOUSE_SYNAPTICS_USB=m
CONFIG_USB_VIDEO_CLASS_INPUT_EVDEV=y
CONFIG_USB_PWC_INPUT_EVDEV=y
# CONFIG_TOUCHSCREEN_SYNAPTICS_I2C_RMI4 is not set

I suppose CONFIG_INPUT_EVDEV and CONFIG_MOUSE_PS2_SYNAPTICS and/or CONFIG_MOUSE_SYNAPTICS_USB are critical here.