XGetImage returns junk with Xorg/dummy_drv/VTs

[Adam Jackson]

On Mon, 2008-11-24 at 18:17 +0000, Alan Cox wrote:
> > Because when you don't have the VT, you don't have the hardware, so you
> > can't submit commands to it.  We don't preserve the framebuffer when we
> > VT switch away, we just generate exposures for all windows when we
> > restore X.
> > 
> > There was some debate about this way way back in like xfree86 4.0 or so.
> > I don't remember what the motivation was for what we've got now, but
> > it's what we do, so...
> 
> Speed of switching for one. Nothing like copying several megabytes of
> data from a PCI/AGP card via PIO to make life suck.

But we do that anyway for all the offscreen pixmaps, because unlike
windows, we're not allowed to lose their contents.  One more full-screen
pixmap doesn't seem like that big a deal.  Really you're just trading
that against the speed of switching back into X, and I think the
exposure storm ends up being at least comparable if not more expensive.

Posit a PCI card with 10M/sec download speed and a 1280x1024x32
framebuffer.  Downloading the front buffer is half a second.
Re-exposure on the Radeon M7 in the laptop on my left is just about half
a second.  Which one do you punish?

- ajax
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 197 bytes
Desc: This is a digitally signed message part
URL: <http://lists.x.org/archives/xorg/attachments/20081124/75798e28/attachment.pgp>