[PATCH 2/2] drm/amdgpu: Don't call amd_powerplay_destroy() if we don't have powerplay

[Christian König]

Am 01.07.2017 um 19:13 schrieb John Brooks:
> amd_powerplay_destroy() expects a handle pointing to a struct pp_instance.
> On chips without PowerPlay, pp_handle points to a struct amdgpu_device. The
> resulting attempt to kfree() fields of the wrong struct ends in fire:
>
> [   91.560405] BUG: unable to handle kernel paging request at ffffebe000000620
> [   91.560414] IP: kfree+0x57/0x160
> [   91.560416] PGD 0
> [   91.560416] P4D 0
>
> [   91.560420] Oops: 0000 [#1] SMP
> [   91.560422] Modules linked in: tun x86_pkg_temp_thermal crc32_pclmul ghash_clmulni_intel efivarfs amdgpu(-) i2c_algo_bit drm_kms_helper syscopyarea sysfillrect sysimgblt fb_sys_fops ttm drm
> [   91.560438] CPU: 6 PID: 3598 Comm: rmmod Not tainted 4.12.0-rc5-drm-next-4.13-ttmpatch+ #1
> [   91.560443] Hardware name: Gigabyte Technology Co., Ltd. Z97X-UD3H-BK/Z97X-UD3H-BK-CF, BIOS F6 06/17/2014
> [   91.560448] task: ffff8805063d6a00 task.stack: ffffc90003400000
> [   91.560451] RIP: 0010:kfree+0x57/0x160
> [   91.560454] RSP: 0018:ffffc90003403cc0 EFLAGS: 00010286
> [   91.560457] RAX: 000077ff80000000 RBX: 00000000000186a0 RCX: 0000000180400035
> [   91.560460] RDX: 0000000180400036 RSI: ffffea001418e740 RDI: ffffea0000000000
> [   91.560463] RBP: ffffc90003403cd8 R08: 000000000639d201 R09: 0000000180400035
> [   91.560467] R10: ffffebe000000600 R11: 0000000000000300 R12: ffff880500530030
> [   91.560470] R13: ffffffffa01e70fc R14: 00000000ffffffff R15: ffff880500530000
> [   91.560473] FS:  00007f7e500c3700(0000) GS:ffff88051ed80000(0000) knlGS:0000000000000000
> [   91.560478] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> [   91.560480] CR2: ffffebe000000620 CR3: 0000000503103000 CR4: 00000000001406e0
> [   91.560483] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
> [   91.560487] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
> [   91.560489] Call Trace:
> [   91.560530]  amd_powerplay_destroy+0x1c/0x60 [amdgpu]
> [   91.560558]  amdgpu_pp_late_fini+0x44/0x60 [amdgpu]
> [   91.560575]  amdgpu_fini+0x254/0x490 [amdgpu]
> [   91.560593]  amdgpu_device_fini+0x58/0x1b0 [amdgpu]
> [   91.560610]  amdgpu_driver_unload_kms+0x4f/0xa0 [amdgpu]
> [   91.560622]  drm_dev_unregister+0x3c/0xe0 [drm]
> [   91.560638]  amdgpu_pci_remove+0x19/0x30 [amdgpu]
> [   91.560643]  pci_device_remove+0x39/0xc0
> [   91.560648]  device_release_driver_internal+0x155/0x210
> [   91.560651]  driver_detach+0x38/0x70
> [   91.560655]  bus_remove_driver+0x4c/0xa0
> [   91.560658]  driver_unregister+0x2c/0x40
> [   91.560662]  pci_unregister_driver+0x21/0x90
> [   91.560689]  amdgpu_exit+0x15/0x406 [amdgpu]
> [   91.560694]  SyS_delete_module+0x1a8/0x270
> [   91.560698]  ? exit_to_usermode_loop+0x92/0xa0
> [   91.560702]  entry_SYSCALL_64_fastpath+0x13/0x94
> [   91.560705] RIP: 0033:0x7f7e4fc118e7
> [   91.560708] RSP: 002b:00007fff978ca118 EFLAGS: 00000206 ORIG_RAX: 00000000000000b0
> [   91.560713] RAX: ffffffffffffffda RBX: 000055afe21bc200 RCX: 00007f7e4fc118e7
> [   91.560716] RDX: 000000000000000a RSI: 0000000000000800 RDI: 000055afe21bc268
> [   91.560719] RBP: 0000000000000003 R08: 0000000000000000 R09: 1999999999999999
> [   91.560722] R10: 0000000000000883 R11: 0000000000000206 R12: 00007fff978c9100
> [   91.560725] R13: 0000000000000000 R14: 0000000000000000 R15: 000055afe21bc200
> [   91.560728] Code: 00 00 00 80 ff 77 00 00 48 bf 00 00 00 00 00 ea ff ff 49 01 da 48 0f 42 05 57 33 bd 00 49 01 c2 49 c1 ea 0c 49 c1 e2 06 49 01 fa <49> 8b 42 20 48 8d 78 ff a8 01 4c 0f 45 d7 49 8b 52 20 48 8d 42
> [   91.560759] RIP: kfree+0x57/0x160 RSP: ffffc90003403cc0
> [   91.560761] CR2: ffffebe000000620
> [   91.560765] ---[ end trace 08a9f3cd82223c1d ]---
>
> Fixes: 1c8638024846 (drm/amd/powerplay: refine powerplay interface.)
> Signed-off-by: John Brooks <john at fastquake.com>

Acked-by: Christian König <christian.koenig at amd.com> if Alex doesn't 
objects I'm going to pick those up next week into our internal branches.

Christian.

> ---
>   drivers/gpu/drm/amd/amdgpu/amdgpu_powerplay.c | 3 ++-
>   1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_powerplay.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_powerplay.c
> index 72c03c7..93ffb85 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_powerplay.c
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_powerplay.c
> @@ -209,7 +209,8 @@ static void amdgpu_pp_late_fini(void *handle)
>   	if (adev->pp_enabled && adev->pm.dpm_enabled)
>   		amdgpu_pm_sysfs_fini(adev);
>   
> -	amd_powerplay_destroy(adev->powerplay.pp_handle);
> +	if (adev->pp_enabled)
> +		amd_powerplay_destroy(adev->powerplay.pp_handle);
>   }
>   
>   static int amdgpu_pp_suspend(void *handle)