[lkml]   [1998]   [Apr]   [3]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
Messages in this thread
SubjectRe: GGI and cli/sti in X
On Fri, 27 Mar 1998, Doug Ledford wrote:

> Pavel Machek wrote:
> >
> > Hi!
> >
> > > in any case, Linus's conservative approach to GGI is excellent;
> > > most people are indeed *not* religiously opposed to X. personally,
> > > I've _never_ had an X bug cause video problems or a crash,
> > > and I've been using Linux and X for more than a couple years.
> >
> > Do this for me: Limit your mem to 16Meg, make your machine swap a
> > bit. Now switch consoles from/into X and do it fast. Be creative and
> > you'll see complete freeze or at least complete mess-up of video.
> Although I can't say for sure, this sounds suspiciously like a bug in X
> (well, we know that already, but I'll be more specific). During the switch
> to/from graphical mode, X is responsible for diddling the registers on the
> video card. If X starts getting swapped out due to memory constraints, I
> could see the prospect of X taking a page fault in the middle of one of
> these diddles as being capable of fubarring the display and possibly the

I get 5 or 6 ughs everytime somone runs X onmy system so I'd say your
correct, the ughs are allways in do_page_fault. The ughs are normal on
any system I've seen with < 16 megs ram. The every once and a while
changing consoles will just fail, and leave me with a frozen console.

> machine if they cause the video card to hang the bus. In short, all of the
> code that X loads to diddle with the registers on the video card needs to be
> locked in memory so that it can't be swapped out or else timing critical
> sections in the X server for diddling the registers could get held up as
> code is paged back in. Now, I haven't looked at the X source so I don't
> know if it does this, but this is where I would start looking to see where
> the problem comes from. The other thing I would do is make sure that the X
> server knows that if you switch back and forth quickly, it should lock down
> the video card during mode switches until the switch is complete and
> stable. So, if you switch into graphics mode and then back to text mode, it
> should never start the switch back until all of the diddling for the switch
> to graphics mode is complete. I haven't the faintest clue as to whether or
> not the X server you use enforces this, but I would suspect it already does.
> FWIW, I never see these X hangs, but I don't switch back and forth anyway
> since I prefer my 80x50 xterms to the console vts.

You are most likely correct.


Gerhard Mack

As a computer I find your faith in technology amusing.

To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to

 \ /
  Last update: 2005-03-22 13:42    [W:0.176 / U:0.728 seconds]
©2003-2018 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site