Re: [drm:drm_init] *ERROR* Cannot initialize the agpgart module.

[Dave Jones <davej@codemonkey.org.uk>]

On Fri, Dec 20, 2002 at 06:29:18PM -0500, Ed Tomlinson wrote:
 > Dave, with the pull from this morning (8am EST), it almost works modular.
 > I get:
 > 
 > Dec 20 18:20:19 oscar upsd[636]: Communication established
 > Dec 20 18:20:47 oscar kernel: Linux agpgart interface v0.100 (c) Dave Jones
 > Dec 20 18:20:47 oscar kernel: agpgart: Detected VIA MVP3 chipset
 > Dec 20 18:20:47 oscar kernel: agpgart: AGP aperture is 64M @ 0xe0000000
 > Dec 20 18:20:58 oscar kernel: [drm] Initialized mga 3.1.0 20021029 on minor 0
 > Dec 20 18:20:58 oscar kernel: Module agpgart cannot be unloaded due to unsafe usage in drivers/char/ag
 > p/backend.c:58

This one is due to the way AGPGART does (or has done for the last 3
years) its module locking. It does a MOD_INC_USE_COUNT as soon as
someone calls the acquire routines. (So you can't unload agpgart
whilst you've a 3d using app (like X) open).
This seems quite sensible, but these days you can't unload agpgart.ko
anyway because the chipset module (via-agp.ko in your case) already
has it 'in use', so I'm tempted to drop those bits.

 > but find this in the X startup log.
 > (EE) MGA: Failed to load module "mga_hal" (module does not exist, 0)

That's matrox's binary only X blob. Not my fault.

 > (EE) MGA(0): [agp] Out of memory (-1014)

This one is. But it may be a knock-on effect from the bug above.
I'll nail that one first.

 > (EE) MGA(0): [drm] failed to remove DRM signal handler
 > DRIUnlock called when not locked

That one's a problem for the DRI folks.

		Dave

-- 
| Dave Jones.        http://www.codemonkey.org.uk