On Wed, May 17, 2017 at 11:08:23PM +0200, Nicolai Stange wrote:
I'm experiencing a boot failure on next-20170515:
BUG: unable to handle kernel NULL pointer dereference at 00000000000007cb IP: radeon_driver_load_kms+0xeb/0x230 [radeon]
[snip]
Bisection lead to commit 7ffb0ce31cf9 ("drm/radeon: Don't register Thunderbolt eGPU with vga_switcheroo"). Reverting this commit on top of next-20170515 fixes the issue for me.
My box is a Dell laptop which most certainly hasn't got any Thunderbolt circuitry.
Thanks a lot Nicolai for reporting this, my apologies for the breakage which turns out to be a dereference of rdev->pdev before it's set. :-(
14: e8 e7 c0 0e 00 callq 0xec100 ; radeon_has_atpx() 19: 84 c0 test %al,%al 1b: 74 9a je 0xffffffffffffffb7 1d: 41 f7 c5 00 00 02 00 test $0x20000,%r13d ; flags & RADEON_IS_IGP 24: 75 91 jne 0xffffffffffffffb7 26: 49 8b 44 24 10 mov 0x10(%r12),%rax ; rax = rdev 2b:* 0f b6 90 cb 07 00 00 movzbl 0x7cb(%rax),%edx <-- trapping instruction
Could you verify if the patch below fixes the issue for you?
Thanks!
Lukas
-- >8 --
diff --git a/drivers/gpu/drm/radeon/radeon_kms.c b/drivers/gpu/drm/radeon/radeon_kms.c index e3e7cb1..4761f27 100644 --- a/drivers/gpu/drm/radeon/radeon_kms.c +++ b/drivers/gpu/drm/radeon/radeon_kms.c @@ -116,7 +116,7 @@ int radeon_driver_load_kms(struct drm_device *dev, unsigned long flags) if ((radeon_runtime_pm != 0) && radeon_has_atpx() && ((flags & RADEON_IS_IGP) == 0) && - !pci_is_thunderbolt_attached(rdev->pdev)) + !pci_is_thunderbolt_attached(dev->pdev)) flags |= RADEON_IS_PX;
/* radeon_device_init should report only fatal error