New install of Xorg is enabling AEI

That can't be correct. In a new install I see the following line in Xorg.0.log
Code:
(WW) AllowEmptyInput is on, devices using drivers 'kbd', 'mouse' or 'vmmouse' will be disabled.
Xorg is disabling kbd and mouse plus there seems to be a hard lock (sometimes) when exiting Xorg.

That is without any xorg.conf file. I've created and modified an xorg.conf to include the following section
Code:
Section "ServerFlags"
        Option "AllowEmptyInput" "Off"
        Option "AutoAddDevices" "Off"
EndSection

With that added Xorg works correctly. Without the ServerFlags section AEI is enabled.
 
davidgurvich said:
That can't be correct. In a new install I see the following line in Xorg.0.log
Code:
(WW) AllowEmptyInput is on, devices using drivers 'kbd', 'mouse' or 'vmmouse' will be disabled.

Step 1: don't panic.
Step 2: ignore it, don't mess with AllowEmptyInput. Leave it alone. Seriously.[1]
Step 3: if you want to prevent X from using HAL for input device detection, all that is needed is
Code:
Options "AutoAddDevices" "Off"

There's no need for a separate ServerFlags section. Like all things in ServerFlags, it can go in the ServerLayout section.

[1] AllowEmptyInput, FreeBSD, and Xorg Input
 
I'm not messing with AEI. I installed FreeBSD and xorg. Xorg disabled both the keyboard and mouse. The new install of xorg enabled AEI as the default.
 
What I'm trying to say is that the warning message can be ignored. Don't set a value for AllowEmptyInput, let it default. Keyboard and mouse will still be detected on startup by X. If you run hal, nothing else is needed. If hal isn't running, turn off AutoAddDevices as above. Mouse and keyboard mentions in ServerLayout can be removed, and InputDevice sections can be removed.
 
The default on FreeBSD is to run HAL and have xorg-server use it. The "disabled" messages are because HAL has been disabled by AutoAddDevices Off. The mouse and keyboard should still work, even without being declared in InputDevice sections.

If you're not using HAL, x11-servers/xorg-server can be rebuilt with that option disabled.
 
That is very interesting but not the situation. On a new install Xorg enabled AEI without having any configuration file and does not vary regardless of HAL or DBUS running.

After creating a bare xorg.conf that disables AEI there are no issues with the mouse and keyboard.

Let me repeat: Xorg starts with AEI enabled on a new install without any configuration file and regardless of HAL.
 
Xorg, just installed from ports last night, without an xorg.conf file, with hald and dbus enabled, and there's no mention of AllowEmptyInput anywhere in the Xorg log file.

Remove your xorg.conf file, restart X, and show us the full Xorg log file that is generated.

Adam
 
The thing I don't understand is why you think this is a problem. AEI being enabled is the default, it's supposed to be that way.

Turning AEI off was used, wrongly, as a workaround for a bug in the FreeBSD xorg-server port. For one day, in 2009. It has been misused ever since. The article link in post #2 goes into more detail.
 
AEI enabled is fine and I don't care about that one way or another. The keyboard and mouse being disabled regardless of hald and dbus is another matter.

There should not be a requirement to create an xorg.conf file just to get a keyboard and mouse. I make the assumption that nobody has a working keyboard and mouse on a new install of FreeBSD9 with hald and dbus enabled and it's not just a thinkpad T42 specific issue.

Why are the keyboard and mouse drivers disabled? Shouldn't using AEI test for the existence of a keyboard and mouse when they are present instead of simply disabling them?
 
I did not install from ports. The install was done using the FreeBSD 9.0 memstick image and Xorg was added using pkg_add. Perhaps there is a problem with some binary package. I'll do another test install after RELEASE.
 
davidgurvich said:
I make the assumption that nobody has a working keyboard+mouse on a new install of FreeBSD9 with hald+dbus enabled and it's not just a thinkpad T42 specific issue.

I'm pretty certain we would have had more complaints on here and IRC if that assumption were true.
 
I'm pretty sure you're right. Everyone likes to complain about something. Just look at all the occupy something groups. There's probably some driver issue with the radeon mobility 7500 that I haven't seen before.

Incidentally, the firefox7 binary segfaults while other browsers work.
 
I actually had a similar problem with FF7 one one of my machines recently. I then decided to update the machine from i386 to amd64, so I never really looked into the crashes.

Adam
 
I actually have a T42 with 8-STABLE. The Radeon 7500 isn't fast, but it's always worked well. It's building 9-STABLE right now. That and a ports rebuild is going to take a while, but I'll report back after it finishes.

Update: 9.0-RC1 installed, but still with old ports built for 8-STABLE and xorg-server built without HAL support. Works with or without xorg.conf.
 
Update: 9.0-RC1, all ports rebuilt, xorg-server still built with HAL option disabled: works without xorg.conf. Even runs ioquake3, but the game dies after a short time, probably due to limited video RAM.

After installing hal, rebuilding xorg-server with the HAL option enabled, and starting hald: mouse and keyboard work without xorg.conf. Again, ioquake3 works, although there's some screen corruption after quitting it. If hald isn't running, xorg.conf with AutoAddDevices set to off is needed.

Now I'm going to remove hal and rebuild xorg-server without it. Here's the xorg.conf I normally use on the T42, comments removed for conciseness:
Code:
Section "ServerLayout"
	Identifier   "Manually Configured"
	Screen       0  "Screen0" 0 0
	Option       "AutoAddDevices" "Off"   # not needed if xorg-server is built without hal
	Option       "DontZap" "Off"
EndSection

Section "Files"
	ModulePath   "/usr/local/lib/xorg/modules"
	FontPath     "/usr/local/lib/X11/fonts/bitstream-vera/"
EndSection

Section "DRI"
	Group 0
	Mode  0660
EndSection

Section "Extensions"
	Option       "Composite" "Enable"
EndSection

Section "Device"
	Identifier  "Card0"
	Driver      "radeon"
	VendorName  "ATI Technologies Inc"
	BoardName   "Radeon Mobility M7 LW [Radeon Mobility 7500]"
	BusID       "PCI:1:0:0"
EndSection

Section "Screen"
	Identifier "Screen0"
	Device     "Card0"
	Monitor    "Monitor0"
	SubSection "Display"
		Virtual 1024 768
		Modes "1024x768"
	EndSubSection
EndSection
 
Back
Top