Problem--perpetual input from unknown source

[cathbad]

I will begin by saying I do not know if this is hardware or software related, or some mixing of both.

So, for almost a week now I having been experiencing odd input behavior. It started on the 14th; everything was working fine that morning, but that afternoon I began noticing that anytime I visited a website where the page was wider than the browser window the page would immediately scroll to the right.

After further exploration I noted additional behavior:

*Drop-down boxes immediately scroll to the last selection anytime the pointer is moved over the box
*Firefox's search and address bars will not remain expanded while typing (for suggestions) or when clicked
*When [Shift], [Ctrl] or [Alt] key is pressed many programs will respond as if a combination with a directional key has been used, e.g. pressing [Shift] in Firefox will navigate to the previous viewed page as if [Shift]+[Left arrow] was pressed, even when typing into a text box
*Moving the pointer anywhere over a scroll bar causes the view to scroll to the right if over the horizontal scroll bar or the bottom if over the vertical scroll bar
*Clicking on the date/time in the tray then moving the pointer over the calendar causes the calendar to scroll forward through future dates
*While viewing a video in VLC, if the pointer is located within the VLC window the volume increases to maximum
*When the pointer is over the tabs in Boinc the view immediatly scrolls through to the last tab
*Can not move windows more than a few pixels at a time
*And starting today when the Mint icon is clicked the menu will open but I am unable to selected/click an icon to launch a program, but clicking on the Mint icon again will allow programs to lauch, this behavior is intermittent
*Also starting today double clicking on words does not select the word, this is also intermittent

Note that these things are happening with no user input other than moving the mouse or hitting the keys listed above. These are the behaviors I can recall or have been most noticeable, and there may be more to remember or discover.

I have spent the last several days trying to find any information about this but I have come across nothing. Below are the things I have tried so far:

*Scoured through and changed many input-related settings to no avail
*Changed mouse and keyboard - used wireless and wired mouse and keyboards plugged into different USB ports, also occurs with no keyboard attached, and changed batteries in wireless devices
*Behavior occurs in Cinnamon and MATE, whether installed locally or using live USB
*Behavior does NOT occur in Windows 7 guest in VirtualBox
*Upgraded to Mint 18.3 and kernel 4.13.0-19, no change
*Uninstalled and cleaned up old kernels, packages and programs I do not use/need, no change
*Checked SSD and filesystem health/integrity, no change
*Ran multiple virus scans, no threats found
*Checked and rechecked UEFI settings and rebooted too many times to count

I was ready to reinstall Mint but figured that would not help since the behavior also occurred when booting from live USB. At this point I am at a loss. For now I am relegated to using an old Vista laptop for most things.

There have been no problems for the last 8-9 months, then this. The computer specs are:
ASRock Fatal1ty AB350 Gaming K4 Promontory, lasted UEFI 3.30 since last month
Ryzen 1700
CORSAIR Vengeance LPX 16GB (2 x 8GB)
Kingston UV400 SSD

Any help with this is GREATLY appreciated and thank you in advance for any assistance you can offer.

[ClixTrix]

I'm thinking keyboard problem. Could be a problem in keyboard settings or a defective keyboard.

[trapperjohn]

Or perhaps a crapped-out mouse or usb device interference.

I have seen this quite a lot lately with cheap, noisy periferals and poorly QC'd mice.

[rene]

The second item on your "have tried" list seems to be making this a potentially interesting one...

I'm noticing your board has an actual PS/2 mouse/keyboard connector: http://www.asrock.com/mb/AMD/Fatal1ty%2 ... ming%20K4/. The usual thing to do for a motherboard is not enable the PS/2 hardware unless something is in fact connected to the port; you're not usually able to influence that through the BIOS setup but this is the first time I see a board with a PS/2 port and (seemingly) a UEFI BIOS so look around a bit in the options.

I'm wondering if the board got itself in a state of having enabled the PS/2 port without anything attached, yet allowed some other peripheral to grab IRQ12; to have that other peripherals activity interpreted as PS/2 activity. Mind you, more of a hunch than anything really reasoned: I don't know PS/2 hardware well enough at the driver level to know whether or not this would in fact be possible. Still, seems interesting enough to mention as a possibility...

You said to have tried "both wired and wireless" but I am assuming both USB. If you even still have a PS/2 keyboard or mouse available to test with, shutting down the machine, connecting it and booting could be an interesting test. Clearly, if anything IS available in your BIOS concerning the PS/2 port, that would also. As would perhaps reloading BIOS defaults and/or really-really shutting down the machine for a bit: shut down and unplug it from power for a minute or so.

[cathbad]

Thank you for the suggestions.

I initially thought about a faulty device, hence trying so many other devices to no avail. Multiple reboots, shut downs and unpluggings made no difference. I had forgotten about the PS/2 connection since I have not used one in about 15 years. I found a PS/2 mouse and plugged it in and booted. Everything was working correctly and continues to do so after returning to the original mouse and keyboard. It appears that the PS/2 mouse initiated or reverted a setting change somewhere.

This has plagued me for days and you guys were able to prompt a fix in less than a day. THANK YOU!

[JoeFootball]

I found a PS/2 mouse ...

That's the most fascinating part.

Everything was working correctly and continues to do so after returning to the original mouse and keyboard.

Does it remain functional even after powering the system down? i.e., is the fix persistent? Just curious as to what was actually going on and how it got fixed.

Joe

[rene]

Things were likely as suspected; booting with an actual PS/2 mouse detected caused reinitialization of the PS/2 state in the BIOS, and specifically had the BIOS claim (IRQ1 and) IRQ12, putting whatever device was firing away on it previously on another IRQ; this was what was hoped.

Rebooting again without a PS/2 mouse or keyboard connected would generally leave the PS/2 hardware disabled (certainly coming from a really-off state; from having had the power unplugged for a bit) and free up these IRQs for other devices again but even if this much newer than normal for PS/2 BIOS keeps it enabled, the BIOS won't shuffle the other devices around without good cause or without you specifically telling it to do so through the setup; will keep the other devices on non-conflicting IRQs.

For the time being. In that sense it is probably a good idea to check whether or not the PS/2 hardware is still enabled when having booted without a PS/2 mouse or keyboard attached; the easiest way to do that is checking the output of cat /proc/interrupts for any interrupts being handled by "i8042" (i.e., in the column where for for example IRQ0 "timer" will be shown); supposedly 1 and 12 if indeed the hardware is enabled; may on this new hardware also be on different ones. If the PS/2 port is still enabled the issue might return at some point if the BIOS will again put a device on a conflicting IRQ. One definite solution would be to just keep the PS/2 mouse plugged in...

You might try to reproduce by booting with the PS/2 mouse connected, yanking it out (historically you shouldn't do that in an electrical sense but I rather doubt you couldn't these days), reboot into the BIOS and have it "Reset configuration data" in the PCI section -- assuming that is still an identifiable option/place in your UEFI BIOS; it used to be on legacy boards. This might again cause it to put a device on a conflicting IRQ and cause a return of your symptoms.

Or you might not, of course. Also note that this could only be guessed due to everything you had already tried, so that wasn't wasted effort (well... minus the virus scans).