btnx: Send keyboard and mouse combination events with mouse buttons

[daou]

daou:

I solved my problem with the F2! I traced the issue to this function:

It was never returning the fd because it always fell through to the "both tests failed" printf. I hardcoded the compare to have the bus id of my keyboard, 0x0011, and then return without doing the test_bit stuff. There is another device with the same product and vendor, the mac mouse emulation thing, so I had to add the bus compare.

Anyway, thanks for the help with this, supporting a feature that isn't really part of btnx yet. My thumb thanks you!

No problem. Glad you got it working. Never hurts to know a little C .

[daou]

I tried the svn version as of 19 Aug 2007 on Ubuntu Gutsy (AMD64), and it failed. I have a VX Revolution. When I ran btnx-config, it would only detect events for left button, right button, scroll wheel forward/back. It detected the forward thumb button as the same as button 1.

...
Everything works as expected. I'm not sure what else I can do to help getting the new version working on my machine (thus making it more robust). I'm happy to help with getting this fixed if you give me some guidance.

Note that this looks to be a similar issue that DarwinsTheory was having, but I never saw a resolution for it.

Could you send me your /proc/bus/input/devices output, especially the section with your mouse. And also your xorg.conf InputSection.

[daou]

I tried sudo hexdump /dev/input/event2, but when I moved with my mouse (and also pressed every button) nothing showed, but when I typed sudo hexdump /dev/input/event3, everything worked except my [+] and [-] buttons ... when I pressed [+] or [-] no hexadecimal code showed ... but when I pressed side buttons or scroll everything worked properly.

It's strange ...

This means btnx cannot do anything with the buttons because they aren't sending any events. It could be a setting somewhere that we have overlooked... if you changed your xorg.conf InputSection like I recommended, then I'm a little bit at a loss as to what to try.

I'll keep this in mind and try to figure something out. But if you manage to fix it, tell me what you did.

[d.code]

Here are the seemingly relevant portions of those 2 files. In addition to the input section I have for my mouse, I'm also using the wacom driver for my tablet screen. These aren't using evdev, though, so I don't see how it would interfere. The btnx devices are what I'm currently using with the older version. Also, I've re-inserted my USB receiver a couple times since firing up X. I guess that's why I'm using the larger Sysfs indices.

------------------------------------------
 /proc/bus/input/devices
------------------------------------------
I: Bus=0003 Vendor=046d Product=c518 Version=0111
N: Name="Logitech USB Receiver"
P: Phys=usb-0000:00:1a.7-4.3/input0
S: Sysfs=/class/input/input9
U: Uniq=
H: Handlers=mouse2 event9 
B: EV=7
B: KEY=ffff0000 0 0 0 0
B: REL=143

I: Bus=0003 Vendor=046d Product=c518 Version=0111
N: Name="Logitech USB Receiver"
P: Phys=usb-0000:00:1a.7-4.3/input1
S: Sysfs=/class/input/input10
U: Uniq=
H: Handlers=kbd event10 
B: EV=f
B: KEY=7fff002c3027 bf00444000000000 1 f808837c000 667bfad9415fed 8e000000000000 0
B: REL=40
B: ABS=100000000

I: Bus=0000 Vendor=b216 Product=0002 Version=0000
N: Name="btnx keyboard"
P: Phys=
S: Sysfs=/class/input/input11
U: Uniq=
H: Handlers=kbd event11 
B: EV=100003
B: KEY=7fffffffffffffff ffffffffffffffff ffffffff00000000 0 ffffffffffffffff ffffffffffffffff ffffffffffffffff ffffffffffffffff

I: Bus=0000 Vendor=b216 Product=0001 Version=0000
N: Name="btnx mouse"
P: Phys=
S: Sysfs=/class/input/input12
U: Uniq=
H: Handlers=mouse3 event12 
B: EV=7
B: KEY=ffffffff ffffffffffffffff 0 0 0 0
B: REL=103

------------------------------------------
 xorg.conf (Relevant)
------------------------------------------
Section "InputDevice"
	Identifier	"Configured Mouse"
	Driver		"mouse"
	Option		"CorePointer"
	Option		"Device"		"/dev/psaux"
	Option		"Protocol"		"auto"
	Option		"EmulateWheel"		"true"
	Option		"EmulateWheelButton"	"2"
	Option		"Buttons"		"3"
EndSection

[fixx2000]

Is anyone having problems with downloading this file or opening the website?

[Tim Perry]

Aye, seems to be down. Waiting in hope...

[daou]

The site is back online. There was a lightning storm a couple of days ago which knocked out power. Forgot to boot up the server (an old IBM Thinkpad dusting in the corner).

[daou]

Here are the seemingly relevant portions of those 2 files. In addition to the input section I have for my mouse, I'm also using the wacom driver for my tablet screen. These aren't using evdev, though, so I don't see how it would interfere. The btnx devices are what I'm currently using with the older version. Also, I've re-inserted my USB receiver a couple times since firing up X. I guess that's why I'm using the larger Sysfs indices.

The event handler detection changed in 0.3.0, and I believe this is what is causing the problem. It's done in a much better way (from a programmer's perspective) in 0.3.x. The method in 0.2.x worked, but its implementation looks like a desperate hack. It looks like there is a bug involved.

It probably has to do with the detection process trying to distinguish between mouse and keyboard handlers. This is done by observing return values from ioctl commands on the event handlers in the function find_handler() in btnx.c.

A mouse is detected if the ioctl command returns EV_REL bits (related to mouse motion and wheel scrolls). A keyboard is detected if EV_KEY (keys) and EV_ABS (absolute motion values) bits are detected. It looks like the mouse check fails for you. I'm not sure how to fix this at the moment. I'll have to look into it further.

[po0f]

daou,

Great work here! I really appreciate you taking to time to create a tool like this for all to enjoy. I had previously used your other HOWTO to get my MX Revolution up and running and am actually glad that there is now a more generic solution. Thanks again!

[phyrewall]

BTNX-config doesn't detect my USB G5, any ideas?