kern/49758: i386 radeondrmkms hangs machine on attachment

[jdbaker%mylinuxisp.com@localhost]

>Number:         49758
>Category:       kern
>Synopsis:       i386 radeondrmkms hangs machine on attachment
>Confidential:   no
>Severity:       critical
>Priority:       high
>Responsible:    kern-bug-people
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Wed Mar 18 21:50:00 +0000 2015
>Originator:     John D. Baker
>Release:        NetBSD/i386-7.0_BETA (and 7.99.6+)
>Organization:
>Environment:
NetBSD slab.technoskunk.fur 7.0_BETA NetBSD 7.0_BETA (SLAB) #3: Wed Mar 18 12:20:54 CDT 2015  sysop%verthandi.technoskunk.fur@localhost:/d0/build/netbsd-7/obj/i386/sys/arch/i386/compile/SLAB i386

>Description:
With recent -current and netbsd-7 (as of 18 March 2015), radeondrmkms
on i386 still hangs my ThinkPad A31p on about 3 out of 4 boot attempts.

At the point when radeondrmkmsfb attaches, the screen will either be
completely black, or tiled with green hash (that would have been
kernel messages).

It will also do this when using serial console.  In all cases, the
machine is wedged hard and the only way to regain control is to
power-cycle the machine and try again.

Following successfull boot, relevant dmesg items:

NetBSD 7.0_BETA (SLAB) #3: Wed Mar 18 12:20:54 CDT 2015
	sysop%verthandi.technoskunk.fur@localhost:/d0/build/netbsd-7/obj/i386/sys/arch/i386/compile/SLAB
[...]
IBM 2653H4U (Not Available)
[...]
acpivga0 at acpi0 (VID): ACPI Display Adapter
acpiout0 at acpivga0 (LCD0, 0x0110): ACPI Display Output Device
acpiout1 at acpivga0 (CRT0, 0x0100): ACPI Display Output Device
acpiout2 at acpivga0 (TV0, 0x0200): ACPI Display Output Device
acpiout3 at acpivga0 (DVI0, 0x0210): ACPI Display Output Device
acpivga0: connected output devices:
acpivga0:   0x0110 (acpiout0): LCD Panel, head 0
acpivga0:   0x0100 (acpiout1): Ext. Monitor, head 0
acpivga0:   0x0200 (acpiout2): TV, head 0
acpivga0:   0x0210 (acpiout3): Unknown Output Device, head 0
[...]
pchb0 at pci0 dev 0 function 0: Intel 82845 Host (rev. 0x04)
agp0 at pchb0: aperture at 0xe0000000, size 0x4000000
ppb0 at pci0 dev 1 function 0: Intel 82845 AGP (rev. 0x04)
pci1 at ppb0 bus 1
pci1: i/o space, memory space enabled
radeon0 at pci1 dev 0 function 0: ATI Technologies FireGL Mobility 7800 M7 LX (rev. 0x00)
radeon0: unable to reserve VGA registers for i386 radeondrmkms hack
[...]
drm: initializing kernel modesetting (RV200 0x1002:0x4C58 0x1014:0x0518).
drm: register mmio base: 0xd0100000
drm: register mmio size: 65536
radeon0: info: GTT: 64M 0xE0000000 - 0xE3FFFFFF
radeon0: info: VRAM: 128M 0x00000000E8000000 - 0x00000000EFFFFFFF (64M used)
drm: Detected VRAM RAM=80M, BAR=128M
drm: RAM width 128bits DDR
Zone  kernel: Available graphics memory: 802208 kiB
drm: radeon: 64M of VRAM memory ready
drm: radeon: 64M of GTT memory ready.
radeon0: info: WB disabled
radeon0: info: fence driver on ring 0 use gpu addr 0x00000000e0000000 and cpu addr 0x0xdb3e9000
drm: Supports vblank timestamp caching Rev 2 (21.10.2013).
drm: Driver supports precise vblank timestamp query.
radeon0: interrupting at irq 9 (radeon)
drm: radeon: irq initialized.
drm: Loading R100 Microcode
drm: radeon: ring at 0x00000000E0001000
drm: ring test succeeded in 0 usecs
drm: ib test succeeded in 0 usecs
drm: Panel ID String: 1600x1200               
drm: Panel Size 1600x1200
drm: No TV DAC info found in BIOS
drm: Radeon Display Connectors
drm: Connector 0:
drm:   VGA-1
drm:   DDC: 0x60 0x60 0x60 0x60 0x60 0x60 0x60 0x60
drm:   Encoders:
drm:     CRT1: INTERNAL_DAC1
drm: Connector 1:
drm:   DVI-D-1
drm:   HPD1
drm:   DDC: 0x64 0x64 0x64 0x64 0x64 0x64 0x64 0x64
drm:   Encoders:
drm:     DFP1: INTERNAL_TMDS1
drm: Connector 2:
drm:   LVDS-1
drm:   Encoders:
drm:     LCD1: INTERNAL_LVDS
drm: Connector 3:
drm:   SVIDEO-1
drm:   Encoders:
drm:     TV1: INTERNAL_DAC2
radeondrmkmsfb0 at radeon0
radeon0: info: registered panic notifier
radeondrmkmsfb0: framebuffer at 0xdb5f0000, size 1600x1200, depth 32, stride 6400
wsdisplay0 at radeondrmkmsfb0 kbdmux 1: console (default, vt100 emulation), using wskbd0


The message about "radeon0: unable to reserve VGA registers for i386
radeondrmkms hack" had me worried that I'd left an outdated local patch
in place, but a quick check revealed that I had not.
>How-To-Repeat:
With netbsd-7 or -current i386 GENERIC (or kernel derived therefrom)
attempt to boot a machine which attaches radeondrmkms and friends.
>Fix: