On 06/12/16 05:07 AM, Alex Deucher wrote:
Since commit 8a357d10043c ("drm: Nerf DRM_CONTROL nodes"), a struct drm_device's ->control member is always NULL.
In the case of CONFIG_DEBUG_FS=y, amdgpu_debugfs_add_files() accesses ->control->debugfs_root though. This results in the following Oops:
[ 9.627636] BUG: unable to handle kernel NULL pointer dereference at 0000000000000018 [ 9.628274] IP: amdgpu_debugfs_add_files+0x8d/0x100 [amdgpu] [ 9.628867] PGD 2514c7067 [ 9.628876] PUD 2514c8067 [ 9.629448] PMD 0 [ 9.630026] Oops: 0000 [#1] PREEMPT SMP [...] [ 9.639906] Call Trace: [ 9.640538] amdgpu_fence_driver_init+0x1e/0x40 [amdgpu] [ 9.641186] amdgpu_device_init+0xa6d/0x1410 [amdgpu] [ 9.641900] ? kmalloc_order_trace+0x2e/0x100 [ 9.642587] amdgpu_driver_load_kms+0x5b/0x200 [amdgpu] [ 9.643355] drm_dev_register+0xa7/0xd0 [ 9.644016] drm_get_pci_dev+0xde/0x1d0 [ 9.644659] amdgpu_pci_probe+0xbe/0xf0 [amdgpu] [...]
Fix this by omitting the drm_debugfs_create_files() call for the control minor debugfs directory which is now non-existent anyway.
Port of Nicolai Stange's radeon patch to amdgpu.
Not sure about the kernel, but in other projects the convention is
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=98915
Fixes: 8a357d10043c ("drm: Nerf DRM_CONTROL nodes") Signed-off-by: Alex Deucher alexander.deucher@amd.com
Regardless of the above,
Reviewed-by: Michel Dänzer michel.daenzer@amd.com