Mouse pointer constrained by invisible boundaries after xrandr --scale

Deciare · 2011-05-14 21:49:22

A couple of months ago, I learnt that it was possible to approximate a higher screen resolution than my netbook's native resolution by using xrandr. A command line like the following:

xrandr --output LVDS1 --scale 1.25x1.25

Would cause the whole screen to be scaled by 25%, artificially increasing my screen resolution to 1280x750 (up from 1024x600). It worked well at the time.

When I tried doing that again today, the screen looked fine, but I found that I couldn't move the mouse pointer beyond the original 1024x600 boundary. So the bottom quarter and right quarter of the screen were entirely inaccessible, as though the mouse pointer was bumping against an invisible barrier. Windows and menus and stuff could still draw there; it just couldn't be reached with the mouse pointer.

I tried the same command line on a different netbook running Ubuntu Lucid, and it works fine there. No issues with the mouse pointer. Both netbooks I tested with are the same model: Acer Aspire One A110L (AOA110-1302).

So what's happening here? What am I missing? I would appreciate any help.

Last edited by Deciare (2011-05-14 21:52:09)

frony0 · 2011-05-15 14:27:45

I'm not sure, but I think at one point I had the same problem it helped me to set the panning to the same value as the scaled resolution. As in

xrandr --output LVDS1 --scale 1.25x1.25 --panning 1280x750

Correct me if I'm wrong...

Deciare · 2011-05-15 15:37:34

Thank you for the suggestion.

I tried that, along with just --panning without --scale. I'm only able to pan down and right by one pixel before the pointer gets stuck at the same invisible boundary...

Last edited by Deciare (2011-05-15 15:38:01)

w33ky · 2011-05-15 19:45:27

I have the same problem with this invisible boundary.
... I hope this will be fixed soon.

frony0 · 2011-05-15 20:11:05

Ok, sorry, another suggestion. When you run xrandr with no parameters, is the resolution after "current" in the first line your original resolution? If not, try

xrandr --output LVDS1 --fb 1280x750

Though that should be automatic, I think. Beyond that, I've no idea

Deciare · 2011-05-15 20:21:47

I'm sorry to say that --fb doesn't work either. ^^;

xrandr --fb 1280x750 --output LVDS1 --scale 1.25x1.25 --panning 1280x750+0+0/1280x750+0+0/64/64/64/64

The elaborate command above doesn't allow my pointer to escape the invisible wall.

xrandr output from the command above is as follows:

Screen 0: minimum 320 x 200, current 1280 x 750, maximum 4096 x 4096
LVDS1 connected 1280x750+0+0 (normal left inverted right x axis y axis) 195mm x 113mm panning 1280x750+0+0 tracking 1280x750+0+0 border 64/64/64/64
1024x600 60.0*+
800x600 60.3 56.2
640x480 59.9
VGA1 disconnected (normal left inverted right x axis y axis)

w33ky · 2011-05-15 20:28:55

My native resolution is 1024x600.
The resolution I want to reach by scaling is 1376x768.

When I scale up to 1376x768 and use --fb 1376x768 then there is no change, when I use --fb 1024x600 then I get a smaller image of my desktop with black borders on the right and in the bottom. The smaller image of my desktop seams to match the size of the invisible border the mouse pointer can't leave.

NiyaFox · 2011-05-15 22:51:54

Okay, I tried this with the latest version of Ubuntu which is running X server 1.10.1, RandR 1.3 and xrandr 1.3.4 -- same versions as my Arch install.

xrandr --output LVDS1 --scale 1.25x1.25 there not only scales the display but also permits the cursor full access to the screen (ie: the bug is not present)

I wonder if this is helpful?

w33ky · 2011-05-16 06:08:59

Does your pc also have an Intel gma grafics chip?
Maybe the problem is related to it.

I tried on an Asus eeepc 1000H with Intel gma 950.
https://bugs.archlinux.org/task/24274

Last edited by w33ky (2011-05-16 09:40:30)

sublimepua · 2011-05-18 15:24:04

farewell guys. I've tried to fix this myself, but I am not that experienced. I've tried to live with it, but its painful and a little ridiculous. I'm subscribed to the bugtracker, but I will be leaving arch linux until this problem is fixed. I hope that it doesn't take too long. Now I've gotta choose between slackware, ubuntu, and gentoo. Gross.

delcypher · 2011-11-27 09:52:47

This problem bit me today. I tried scaling and panning. Both are broken as the mouse can't move out of the original screen area

It appears to be an Xorg bug:

https://bugs.freedesktop.org/show_bug.cgi?id=39949

Some patches have been submitted but I don't see anyone upstream actually accepting the patch.

Deciare · 2011-11-28 04:55:28

Ah, perfect! Reverting http://cgit.freedesktop.org/xorg/xserve … 44eb1a0a11 (referenced from https://bugs.freedesktop.org/show_bug.cgi?id=40063) is exactly the workaround I needed. Thank you for pointing out that bug report.

Last edited by Deciare (2011-11-28 04:55:48)

josvanr · 2012-01-31 11:08:05

Hello, I'm trying to do this also but no succes until now. When I try to revert the patch I get something like 'detected previously unreverted patch' or somethig. Could you briefliy tell what you did; which files did you download and what did you do to revert the patch?

Deciare · 2012-01-31 20:02:33

Since you are reverting the patch, you need to indicate that the patch should be reversed on the command line. So if you saved the patch from git as constrain-cursor-harder.patch, for instance, you would need to modify the PKGBUILD so that:

constrain-cursor-harder.patch is in the source array
The patch's md5sum is in the md5sums array
The patch command line specifies that it's a reverse patch (the -R is important):
```
patch -RNp1 -i "${srcdir}/constrain-cursor-harder.patch"
```

Unfortunately, the patch from git no longer reverses cleanly against xorg-server 1.11.4.

I've made an updated version of the patch that should apply cleanly, without a need to be reversed, here: http://moe.isisview.org/patches/xorg-se … rder.patch

karol · 2012-01-31 20:06:52

Should this go upstream?

Deciare · 2012-01-31 20:27:30

Can do. Attached patch to upstream bug report: https://bugs.freedesktop.org/show_bug.cgi?id=39949

igge · 2012-07-13 19:48:27

Does anyone now if there is a solution to this problem?
Is still can't get it to work. I want my netbook to use 1024*768 and the
screen resolution is at that but the mouse pointer is still constrained
by invisible boundaries.

//Igge

Gusar · 2012-07-13 19:59:38

This seems to work if you go by comments: https://bugs.freedesktop.org/show_bug.cgi?id=39949#c20

Though I also recall AaronP of Nvidia posting some patch not long ago, but I'm not sure. I'll do a search, see what I can find. Edit: Hmm, for now all I can find is this: http://article.gmane.org/gmane.comp.fre … 138/match= <- quite useful in that you don't have to calculate scale factors yourself. But I have some vague memory he also posted another patch too. Seems not, though. There is another randr patch he posted, but it's unrelated to this issue.

Last edited by Gusar (2012-07-13 20:14:42)

karol · 2012-10-17 23:33:22

This seems to be fixed now https://bugs.archlinux.org/task/24274
I've tried it and it works for me.

Arch Linux

#1 2011-05-14 21:49:22

Mouse pointer constrained by invisible boundaries after xrandr --scale

#2 2011-05-15 14:27:45

Re: Mouse pointer constrained by invisible boundaries after xrandr --scale

#3 2011-05-15 15:37:34

Re: Mouse pointer constrained by invisible boundaries after xrandr --scale

#4 2011-05-15 19:45:27

Re: Mouse pointer constrained by invisible boundaries after xrandr --scale

#5 2011-05-15 20:11:05

Re: Mouse pointer constrained by invisible boundaries after xrandr --scale

#6 2011-05-15 20:21:47

Re: Mouse pointer constrained by invisible boundaries after xrandr --scale

#7 2011-05-15 20:28:55

Re: Mouse pointer constrained by invisible boundaries after xrandr --scale

#8 2011-05-15 22:51:54

Re: Mouse pointer constrained by invisible boundaries after xrandr --scale

#9 2011-05-16 06:08:59

Re: Mouse pointer constrained by invisible boundaries after xrandr --scale

#10 2011-05-18 15:24:04

Re: Mouse pointer constrained by invisible boundaries after xrandr --scale

#11 2011-11-27 09:52:47

Re: Mouse pointer constrained by invisible boundaries after xrandr --scale

#12 2011-11-28 04:55:28

Re: Mouse pointer constrained by invisible boundaries after xrandr --scale

#13 2012-01-31 11:08:05

Re: Mouse pointer constrained by invisible boundaries after xrandr --scale

#14 2012-01-31 20:02:33

Re: Mouse pointer constrained by invisible boundaries after xrandr --scale

#15 2012-01-31 20:06:52

Re: Mouse pointer constrained by invisible boundaries after xrandr --scale

#16 2012-01-31 20:27:30

Re: Mouse pointer constrained by invisible boundaries after xrandr --scale

#17 2012-07-13 19:48:27

Re: Mouse pointer constrained by invisible boundaries after xrandr --scale

#18 2012-07-13 19:59:38

Re: Mouse pointer constrained by invisible boundaries after xrandr --scale

#19 2012-10-17 23:33:22

Re: Mouse pointer constrained by invisible boundaries after xrandr --scale

Board footer