rhgb freezing computer

[Adam Jackson]

Alan Cox wrote:
> On Tue, Sep 26, 2006 at 04:08:40PM -0400, Adam Jackson wrote:
>> VT_WAITACTIVE only ever gets called from xf86OpenConsole, which is in 
>> the server generation path (and _not_ in normal EnterVT, which is sort 
>> of disturbing).  This appears to only be possible to trigger during 
>> server init, and specifically in the handoff from rhgb's X server to 
>> gdm's.
> 
> That makes sense if the timing is ugly
> 
> 
> Can this occur
> 
> 	rhgb X server -> decides to go away
> 	gdm X server starts
> 	gdm X server issues switch to screen
> 	rhgb X server cleanup issues switch back to text mode screen
> 	gdm X server waits for its screen

I think it can.  I'm fairly sure gdm launches X while rhgb's server is 
still active, since otherwise we wouldn't have to make sure rhgb got DRI 
disabled (old unrelated bug where only one X server can have working DRI 
at a time).  And xf86CloseConsole will do:

1: KDSETMODE to KD_TEXT
2: VT_GETMODE
3: VT_SETMODE to VT_AUTO
4: VT_ACTIVATE to the VT that was current when the server started

So would the race look like:

rhgb X				gdm X
				VT_ACTIVATE 7
VT_ACTIVATE 1
				VT_WAITACTIVE 7

?

> Also does this ever occur without selinux ?

It doesn't appear to happen with selinux=0.  Suspect that's more timing 
than policy though.

- ajax