Subject: The radeon framebuffer device panics init (on i386)
To: None <tech-kern@NetBSD.org>
From: Vincent <email@example.com>
Date: 12/18/2006 22:17:55
I suddently realized there was a Radeon framebuffer device that could be
added to the kernel, and had the violent desire to test it, because I do
tinker with X these days. Bad luck, it broke my brand new HEAD kernel.
After spitting that:
NetBSD 4.99.6 (PRESARIO) #2: Mon Dec 18 19:08:07 CET 2006
radeonfb0 at pci1 dev 0 function 0: vendor 0x1002 product 0x4c57
RADEON_BIOS_4_SCRATCH = 01000004
RADEON_FP_GEN_CNTL = 01430000
RADEON_FP2_GEN_CNTL = 00000008
RADEON_TMDS_CNTL = 01000000
RADEON_TMDS_TRANSMITTER_CNTL = 10000082
RADEON_TMDS_PLL_CNTL = 00000a1b
RADEON_LVDS_GEN_CNTL = 083dffa1
RADEON_FP_HORZ_STRETCH = 0e7f09ff
RADEON_FP_VERT_STRETCH = 0e2ff855
radeonfb0: unable to map BIOS!
radeonfb0: BIOS inititialization failed
the kernels fails to fork init, because:
panic: cnopen: no console device
Well, I am wrong somewhere, is this expected behaviour or the effect of
a vicious hidden and nasty bug?
PS: obviously, removing radeonfb* makes the kernel work okay at first sight.