Mouse doesn't reach left edge of screen (Solved)

[gemk65]

The current problem in Linux Mint is that the mouse pointer does not reach the left edge of the screen; it stops about 2 mm from the left edge. Even so, I am able to select text that extends to the left edge, but I cannot see the pointer taking that action.

With the EndeavourOS, the mouse pointer's behavior is perfectly normal: It reaches the left edge of the screen and is able to select text that extends to the left edge.

The above information suggests to me that there is something amiss in Linux Mint.

I will follow your suggestion and test an Ubuntu distro to see how the mouse behaves. So if the problem appears also with the Ubuntu distro, then, as you say, it would be more correct to say that the fault appears to be with Ubuntu-based software, including but not limited to Mint specifically.

Since I am not so technical, I was not aware that such a distinction could be made. Thank you for the correction.

I will let you know about the results of the Ubuntu test. (I hope that I can find a distro that offers a downloadable ISO file that can be used for creating a live-boot USB stick. This is the only method that is within my limited skill set).

[SMG]

The above information suggests to me that there is something amiss in Linux Mint.

That is a perfectly understandable conclusion. However, it doesn't help narrow down what the problem might be. And we would need to be able to do that to help you.

Ubuntu 20.04.3 LTS (Focal Fossa).

[gemk65]

I have just now completed a test with Ubuntu Desktop 20.04 LTS on a live-boot USB stick.

The problem with the mouse pointer in Ubuntu is identical to the one that I experience in Linux Mint: The pointer does not reach the left edge of the screen but is able to select text that extends to the left edge, though the pointer is not displayed as performing this action (it appears frozen about 2 mm from the left edge).

If there is any other test or method that I can perform in order to diagnose the problem or narrow down its cause, please let me know.

[SMG]

Today the problem fixed itself, just as mysteriously as when it first manifested.

The mouse pointer flickered a bit (disappeared and re-appeared), and then I noticed that it now reached the left edge of the screen, as is normal.

I notice that the mouse does a lot of flickering over certain areas of the screen, such as buttons. I am not sure whether this is normal, or what the issue is, but it is not really a problem, though it might be related to the problem that I had.

Are you still noticing the flickering? Or is that pretty much gone now?

[gemk65]

There was no flickering of the pointer in Ubuntu, and there hasn't been any flickering in Mint in the last few days.
The flickering problem seems to come and go, and not very often, but doesn't really bother me.

[SMG]

There was no flickering of the pointer in Ubuntu, and there hasn't been any flickering in Mint in the last few days.
The flickering problem seems to come and go, and not very often, but doesn't really bother me.

I was thinking it might be a clue as to what might be happening. That is why I asked.

[SMG]

It is possible this may have something to do with what defines the size of a window, what window tiling settings and window settings you use, and the screen resolution. There is obviously a difference between what the mouse is actually doing versus what it looks like it is doing. Apparently, there are a lot of pieces and parts which factor into that and I do not know how to cut down the list to know how to investigate this any further than what I have already done.

[gemk65]

jacobhenry174 replied but his post has disappeared for some reason; he suggested changing the theme.

Thank you both, jacobhenry174 and SMG.

I changed the theme, but there is no change in the mouse's behavior.

Following SMG's comment, I changed Windows settings (Focus mode) and General settings (VSync method, Compositing), but no luck here either. (I don't understand what these 3 settings mean but I don't see any difference after the changes). I have also experimented with the mouse acceleration and sensitivity settings, but the pointer problem remains the same.

Since this mouse pointer problem occurred with the Ubuntu live-boot stick but not with the EndeavourOS live-boot stick, I suppose that it might help if I send the Ubuntu developers a message about this issue.

[gemk65]

I did further experimenting with other operating systems, and I discovered that the problematical mouse pointer behavior occurred with all of them: MX Linux (based on Debian), Manjaro (based on Arch), Ubuntu (based on Debian), and Mint (based on Ubuntu). The result with EndeavourOS, where the pointer behaved normally, was an anomaly.

So now it appears to me that the problem is out-of-date firmware or hardware drivers. My Driver Manager reports that my computer does not need any additional drivers, but perhaps there is a lag time between the actual need for an updated driver versus its availability and discovery in the repositories. Or perhaps the Driver Manager does not involve itself with firmware.

I don't know exactly how this works. When I use Update Manager to perform my weekly updates, does Mint automatically search the repositories for hardware drivers and firmware based on the identified hardware in my system?

If I have a choice between waiting for drivers/firmware to become available to Update Manager versus proceeding on my own to update them, I would prefer to wait.

[SMG]

I did further experimenting with other operating systems, and I discovered that the problematical mouse pointer behavior occurred with all of them: MX Linux (based on Debian), Manjaro (based on Arch), Ubuntu (based on Debian), and Mint (based on Ubuntu). The result with EndeavourOS, where the pointer behaved normally, was an anomaly.

I do not know enough about EndeavourOS to know what might be the difference.

Hardware drivers are included in the kernel. So if those operating systems you tried were using different kernels, then there might be a difference in the drivers which loaded from the kernel. It is also possible this somehow might be related to windows-managers used in each version.

So now it appears to me that the problem is out-of-date firmware or hardware drivers.

Can you explain how you have come to that conclusion?

My Driver Manager reports that my computer does not need any additional drivers, but perhaps there is a lag time between the actual need for an updated driver versus its availability and discovery in the repositories. Or perhaps the Driver Manager does not involve itself with firmware.

Drivers and firmware are two different things.

Driver Manager is only for third-party proprietary drivers. If there are any available for your hardware, you would see them in Driver Manager and be able to install them. Once installed, you would see updates for those drivers in Update Manager. You would not get "updated" drivers in Driver Manager.

Firmware for some hardware is included either in the kernel or in the linux-firmware packages. Both of those are updated through Update Manager.

I am not aware of firmware available for mice or monitors that you would obtain via the operating system.

[gemk65]

gemk65 wrote: ↑
Sun 2021 Oct 24, 11:00 pm
So now it appears to me that the problem is out-of-date firmware or hardware drivers.

SMG wrote:
Can you explain how you have come to that conclusion?

I arrived at this impression (that the problem might be based on outdated computer firmware or hardware drivers) based on a process of elimination:

- My mouse is new (the pointer problem occurred as well with my previous mouse);
- My monitor is new (the pointer problem occurred as well with my previous monitor);
- The pointer problem persists with 4 different operating systems -- MX Linux, Manjaro, Ubuntu, and Mint. If the problem were caused by elements of the Mint operating system (Ubuntu-based), I would expect that changing the operating system to MX Linux (Debian-based) or Manjaro (Arch-based) would make a difference, but there was no difference.
- My computer hardware (launch date 2016) is the only element that I have not changed.

Also, I seem to be one of the few people, if not the only person, that is experiencing this problem. If elements of an operating system were responsible for such a significant issue, I would expect that a large number of people would be reporting it, and I have not found that.

Also, my computer hardware might be atypical -- it is a Gigabyte Brix, with whose reputation I am not familiar, and which arrived at my dealer through one of the uncertain importing strategies in Israel, where I live.

I neglected to mention that when I conducted the test with the live-boot stick of EndeavourOS, the only system in which my mouse pointer successfully reached the left edge of the screen, I noticed mouse pointer instability, just as happened with my Mint system. In EndeavourOS, the mouse pointer reached the left edge of the screen while it was creating multiple duplicate images of the pointer. In Mint, the mouse pointer reached the left edge of the screen while it was flickering. These are clues; they suggest to me that certain software elements responsible for stabilizing the mouse pointer are somehow associated with the pointer problem. So the problem returned when I rebooted Mint.

[SMG]

Also, I seem to be one of the few people, if not the only person, that is experiencing this problem. If elements of an operating system were responsible for such a significant issue, I would expect that a large number of people would be reporting it, and I have not found that.

The only references to this type of issue I found were in specific game interfaces and adjustments to the software related to the game were able to fix the issues.

Also, my computer hardware might be atypical -- it is a Gigabyte Brix,

That might be a possible reason you are seeing this and others are not.

I worked with one other person who has a Brix (different model than yours) and was having issues having their monitor (TV) being recognized when they switched from LM19 to LM20. One of the things I learned during that investigation is the person's HDMI port showed as a DP port in the xrandr output. At the time, I found this link, HDMI stops working when DisplayPort is unplugged, which is a Brix with similar HDMI /DP anomalies.

That may not be an issue in your situation. The Xorg log data you posted earlier indicates your monitor is attached to an HDMI port.

Code: Select all

[    11.163] (II) modeset(0): EDID for output HDMI-3
[    11.163] (II) modeset(0): Manufacturer: LEN  Model: 62ae  Serial#: xxxxxx
[    11.163] (II) modeset(0): Year: 2021  Week: 21
[    11.164] (II) modeset(0): EDID Version: 1.3
[    11.166] (II) modeset(0): Monitor name: S24e-20

But there might be something in how the data gets passed to the monitor (through the computer port) which might be a factor? A Total wild guess on my part.

It may also be that a different monitor cable might make a difference. I've been on enough topics where a cable change made a difference (or even just removing and re-attaching the cable) that I never discount that possibility.

[gemk65]

My computer also has an analog port (D-15 connector) for a VGA cable to the monitor. When I used the VGA cable, the mouse pointer problem was the same.

My HDMI cable is brand new. Following your suggestion, I tried removing and re-inserting it, but that made no difference.

I should mention that when I made a fresh re-installation of Linux Mint 20.2 yesterday, I found that the mouse pointer problem was there right at the start, even before I made any user configurations or customizations to the system.

I hope that Mint's Update Manager scans the repositories for updated firmware and hardware drivers for the Gigabyte Brix (and for Samsung SSD's). Perhaps I will receive a pleasant surprise in a few months' time.

[SMG]

I hope that Mint's Update Manager scans the repositories for updated firmware and hardware drivers for the Gigabyte Brix (and for Samsung SSD's).

If you want updates for your Brix, you would need to install updated BIOS/UEFI versions. The operating system does not update your motherboard.

The operating system can only work with known standards and as long as Brix followed those standards then the operating system drivers can use the hardware properly.

You would have to check with Samsung to see if there are any firmware updates and if the updates can be done on Linux-based distros. I would not expect those types of updates would have anything to do with this issue.

[gemk65]

Thank you again, SMG.

[gemk65]

Here is an update to this problem:

I received an email from Daniel van Vugt (1948528@bugs.launchpad.net) following a bug report (https://bugs.launchpad.net/bugs/1948528) that I made with Ubuntu. He says:

This particular bug has already been reported and is a duplicate of bug 1767654, so it is being marked as such.

https://bugs.launchpad.net/bugs/1767654
[Intel Braswell] Cursor gets stuck on left side of the screen

It's a common problem with the GPUs in Intel Braswell chips.

Package changed: xorg (Ubuntu) => linux (Ubuntu)

[SMG]

Package changed: xorg (Ubuntu) => linux (Ubuntu)

I did not understand what you mean by this statement.

I do see in the linked bug reports Bobby Lawrence mentions:

I'd like to share a slightly improved workaround on this for those still dealing with the issue. The core problem seems to be that the mouse cursor can go into a negative X position by one pixel when moved to the far left of the screen. With either the now obsolete intel proprietary driver or the current intel modesetting driver, this causes issues (exact symptoms are different for each, but seem to be from the same root cause).

My previous workaround got by this by scaling the entire desktop to 99.99% of its former size. This prevented the cursor from getting to the -1 X position, but came at a cost of lost performance and some screen tearing in videos and games. This was due to a bilinear filter being applied by xrandr to the whole desktop to accommodate the slight re-size.

The following command avoids this issue and instead merely shifts the entire desktop right one pixel. It still works around the issue, but seems to avoid the overhead of resizing the whole desktop (which preserves display integrity and performance).

xrandr --output HDMI-3 --transform 1,0,-1,0,1,0,0,0,1

It would be nice to get a root cause fix for this, but this workaround seems to do the trick with very little downside. As before, replace the "HDMI-3" with the name of your active display (found by running xrandr with no options) and place the command in your ~/.xprofile (or make a startup app entry in your desktop manager).

I do not know if perhaps that might help you.

[gemk65]

Thank you, SMG.
The text line that you did not understand was part of the message sent to me about the bug report. I did not understand it either.
I'm grateful that you found that information about what is going on. It is reassuring to know that others have experienced this problem and that the problem is understood. It is unfortunate that there is currently no root cause fix (other than purchasing a new computer).

I would like to try the suggested workaround command:
xrandr --output HDMI-3 --transform 1,0,-1,0,1,0,0,0,1
Do you know how I can later undo this command if I need to?

He also suggests "place the command in your ~/.xprofile"; but I don't have this file. I do see the file ~/.profile. Is their something that I should do with this file?

[SMG]

Thank you, SMG.
The text line that you did not understand was part of the message sent to me about the bug report. I did not understand it either.

In that case, I think it means the problem is not in the xorg package (which is where I think the bug was originally posted), the problem is in the Linux kernel.

I would like to try the suggested workaround command:
xrandr --output HDMI-3 --transform 1,0,-1,0,1,0,0,0,1
Do you know how I can later undo this command if I need to?

If you run this command right now, it will only be good for the current session. To clear it all you have to do is logout/reboot.

He also suggests "place the command in your ~/.xprofile"; but I don't have this file. I do see the file ~/.profile. Is their something that I should do with this file?

Yes, in Mint it is ~/.profile. After testing the line and determining it does work for you, you would just add it to the end of the ~/.profile file (add a new last line to the file with that info). By doing that, the line is automatically run each time you log in.

[gemk65]

Yes! The command fixes the problem! Wow!!

However: Adding this command line to the file ~/.profile did not cause the command to be executed upon reboot.

Can you think of some way to cause this command to be executed automatically for each boot?