Add check for ras pointers.
Issues caused by this commit: be5c7eb104067d61
[ 2312.987618] BUG: kernel NULL pointer dereference, address: 00000000000000e8
[ 2312.987622] #PF: supervisor read access in kernel mode
[ 2312.987624] #PF: error_code(0x0000) - not-present page
[ 2312.987625] PGD 0 P4D 0
[ 2312.987627] Oops: 0000 [#1] PREEMPT SMP NOPTI
[ 2312.987630] CPU: 9 PID: 1749 Comm: modprobe Not tainted 6.3.7-38fc8aadcfb2 #1
[ 2312.987632] Hardware name: AMD Celadon-CZN/Celadon-CZN, BIOS TLD1001Bb 12/01/2020
[ 2312.987634] RIP: 0010:amdgpu_ras_reset_error_count+0x126/0x140 [amdgpu]
[ 2312.987852] Code: 10 48 c7 c1 ec 6a 54 c1 77 08 4a 8b 0c ed c0 35 59 c1 48 8b 33 48 c7 c2 78 a7 4d c1 48 c7 c7 60 a4 5c c1 e8 8c 9e ca d0 eb bf <41> 8b 86 e8 00 00 00 85 c0 0f 84 37 ff ff ff e9 26 ff ff ff 31 c0
[ 2312.987855] RSP: 0018:ffffa40402e378e0 EFLAGS: 00010246
[ 2312.987856] RAX: 0000000000000000 RBX: ffff90cf09580000 RCX: 0000000000000000
[ 2312.987858] RDX: 0000000000000000 RSI: 0000000000000006 RDI: ffff90cf09580000
[ 2312.987859] RBP: ffffa40402e37908 R08: 0000000000000000 R09: c0000000fffeffff
[ 2312.987860] R10: 0000000000000000 R11: ffffa40402e37640 R12: ffffffffc1593d80
[ 2312.987861] R13: 0000000000000006 R14: 0000000000000000 R15: 0000000000000000
[ 2312.987862] FS: 00007fb5d3b33c40(0000) GS:ffff90d006840000(0000) knlGS:0000000000000000
[ 2312.987864] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[ 2312.987865] CR2: 00000000000000e8 CR3: 000000010ae24000 CR4: 0000000000750ee0
[ 2312.987867] PKRU: 55555554
[ 2312.987868] Call Trace:
[ 2312.987870] <TASK>
[ 2312.987872] ? show_regs+0x5b/0x70
[ 2312.987877] ? __die_body+0x1f/0x70
[ 2312.987879] ? __die+0x2a/0x40
[ 2312.987881] ? page_fault_oops+0x156/0x470
[ 2312.987884] ? dev_printk_emit+0x87/0xc0
[ 2312.987889] ? do_user_addr_fault+0x34a/0x720
[ 2312.987891] ? exc_page_fault+0x75/0x180
[ 2312.987895] ? asm_exc_page_fault+0x27/0x30
[ 2312.987898] ? amdgpu_ras_reset_error_count+0x126/0x140 [amdgpu]
[ 2312.987980] gmc_v9_0_late_init+0x7f/0xc0 [amdgpu]
[ 2312.988064] amdgpu_device_ip_late_init+0x49/0x2b0 [amdgpu]
[ 2312.988144] ? mutex_lock+0x12/0x40
[ 2312.988148] amdgpu_device_init+0x2253/0x24e0 [amdgpu]
[ 2312.988225] ? pci_read_config_word+0x23/0x40
[ 2312.988230] amdgpu_driver_load_kms+0x1a/0x1a0 [amdgpu]
[ 2312.988278] amdgpu_pci_probe+0x16b/0x4a0 [amdgpu]
[ 2312.988278] local_pci_probe+0x4a/0xb0
[ 2312.988278] pci_device_probe+0xd9/0x240
[ 2312.988278] really_probe+0x116/0x3e0
[ 2312.988278] ? pm_runtime_barrier+0x55/0xa0
[ 2312.988278] __driver_probe_device+0x81/0x160
[ 2312.988278] driver_probe_device+0x24/0xb0
[ 2312.988278] __driver_attach+0x10e/0x170
[ 2312.988278] ? __device_attach_driver+0x120/0x120
[ 2312.988278] bus_for_each_dev+0x7b/0xd0
[ 2312.988278] driver_attach+0x1e/0x30
[ 2312.988278] bus_add_driver+0x11d/0x220
[ 2312.988278] ? 0xffffffffc0b56000
[ 2312.988278] driver_register+0x5e/0x120
[ 2312.988278] ? 0xffffffffc0b56000
[ 2312.988278] __pci_register_driver+0x68/0x70
[ 2312.988278] amdgpu_init+0x74/0x1000 [amdgpu]
[ 2312.988278] do_one_initcall+0x48/0x210
[ 2312.988278] ? kmalloc_trace+0x2a/0xa0
[ 2312.988278] do_init_module+0x4f/0x1f3
[ 2312.988278] load_module+0x21fe/0x23f0
[ 2312.988278] ? kernel_read_file+0x291/0x310
[ 2312.988278] __do_sys_finit_module+0xc0/0x130
[ 2312.988278] ? __do_sys_finit_module+0xc0/0x130
[ 2312.988278] __x64_sys_finit_module+0x1a/0x20
[ 2312.988278] do_syscall_64+0x3a/0x90
[ 2312.988278] entry_SYSCALL_64_after_hwframe+0x63/0xcd
Signed-off-by: Jesse Zhang <Jesse.Zhang@xxxxxxx>
---
drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
index 303fbb6a48b6..33801a5bb460 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
@@ -1223,7 +1223,7 @@ int amdgpu_ras_reset_error_count(struct amdgpu_device *adev,
struct amdgpu_ras *ras = amdgpu_ras_get_context(adev);
const struct amdgpu_mca_smu_funcs *mca_funcs = adev->mca.mca_funcs;
- if (!block_obj || !block_obj->hw_ops) {
+ if (!block_obj || !block_obj->hw_ops || !ras) {
dev_dbg_once(adev->dev, "%s doesn't config RAS function\n",
ras_block_str(block));
return -EOPNOTSUPP;
--
2.25.1
