Re: 2.6.15-mm2

[Dave Jones <davej@redhat.com>]

On Fri, Jan 13, 2006 at 08:32:05AM +0000, Alan Hourihane wrote:
 > On Fri, 2006-01-13 at 09:02 +1100, Dave Airlie wrote:
 > > > >
 > > > > I've cc'ed Alan Hourihane, but from memory the Intel on-board graphics
 > > > > chips don't advertise the AGP bit on the graphics controllers but work
 > > > > using AGP...
 > > > >
 > > > > I've got an PCIE chipset with Radeon on it, and in that case I could get
 > > > > away without agpgart...
 > > >
 > > > Dave,
 > > >
 > > > You're probably reading too much into that last statement.
 > > >
 > > > I've never seen a pure PCI-e chipset from Intel (i.e. the ones without
 > > > integrated graphics) so that may not be true, but the ones with
 > > > integrated graphics are always treated as AGP based.
 > > >
 > > 
 > > I'll show you one at xdevconf if I can get there, it has just a PCI-E
 > > root bridge no graphics controller, we still init AGP on it but I
 > > don't think there is any need, however for all the integrated
 > > graphics, even if they don't advertise AGP they do use it which is
 > > DaveJ's problem that he was trying not to load AGP if the AGP was
 > > being advertised..
 > 
 > O.k. I didn't see the original thread to this. But yes, all integrated
 > graphics based Intel chipsets are AGP regardless if the chip doesn't
 > advertise it correctly.

FWIW, I've dropped that change from agpgart.git. It caused more problems
than it was worth.

		Dave

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/