Sorry please ignore this patch. It needs more investigation on the cause.
Regards,
Henry
On 07/26/2018 07:04 PM, intel-gvt-dev-bounces@xxxxxxxxxxxxxxxxxxxxx wrote:
From: Hang Yuan <hang.yuan@xxxxxxxxxxxxxxx>
When create one vgpu device and then remove it, oops as below happens.
It's because dmabuf mutex is not initialized if don't open vgpu device.
This patch will initialize dmabuf mutex during creating vgpu to avoid
the oops.
BUG: unable to handle kernel NULL pointer dereference at (null)
IP: [<ffffffff8dd909fb>] __list_add+0x1b/0xc0
PGD 0
Oops: 0000 [#1] SMP
Workqueue: sysfsd sysfs_schedule_callback_work
task: ffff8944fdee1040 ti: ffff89405360c000 task.ti: ffff89405360c000
RIP: 0010:[<ffffffff8dd909fb>] [<ffffffff8dd909fb>] __list_add+0x1b/0xc0
RSP: 0018:ffff89405360fc88 EFLAGS: 00010246
RAX: 00000000ffffffff RBX: ffff89405360fcb0 RCX: 0000000000000000
RDX: ffffab1f4371abb8 RSI: 0000000000000000 RDI: ffff89405360fcb0
RBP: ffff89405360fca0 R08: 0000000000000000 R09: 000000018040002e
R10: 0000000051ab3901 R11: ffffe5658e46acc0 R12: ffffab1f4371abb8
R13: 0000000000000000 R14: 00000000ffffffff R15: ffffab1f4371abb8
FS: 0000000000000000(0000) GS:ffff8945314c0000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 0000000000000000 CR3: 00000001eae0e000 CR4: 00000000003607e0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
Call Trace:
[<ffffffff8e15ecc6>] __mutex_lock_slowpath+0xa6/0x1d0
[<ffffffff8dbfa6d4>] ? __vunmap+0x94/0x100
[<ffffffff8e15e0cf>] mutex_lock+0x1f/0x2f
[<ffffffffc068eb1a>] intel_vgpu_dmabuf_cleanup+0x3a/0xe0 [i915]
[<ffffffffc06780e4>] intel_gvt_destroy_vgpu+0xa4/0x100 [i915]
[<ffffffffc03325d0>] intel_vgpu_remove+0x30/0x40 [kvmgt]
[<ffffffffc0323189>] mdev_device_remove_ops+0x29/0x60 [mdev]
[<ffffffffc0323886>] mdev_device_remove+0x96/0x130 [mdev]
[<ffffffffc0323b34>] remove_callback+0x14/0x30 [mdev]
[<ffffffff8dcc948e>] sysfs_schedule_callback_work+0x1e/0x60
[<ffffffff8dab88af>] process_one_work+0x17f/0x440
[<ffffffff8dab9946>] worker_thread+0x126/0x3c0
[<ffffffff8dab9820>] ? manage_workers.isra.24+0x2a0/0x2a0
[<ffffffff8dac0a31>] kthread+0xd1/0xe0
[<ffffffff8dac0960>] ? insert_kthread_work+0x40/0x40
[<ffffffff8e16cc1d>] ret_from_fork_nospec_begin+0x7/0x21
[<ffffffff8dac0960>] ? insert_kthread_work+0x40/0x40
Code: ff ff ff 66 2e 0f 1f 84 00 00 00 00 00 0f 1f 00 55 48 89 e5 41 55 49 89 f5 41 54 49 89 d4 53 4c 8b 42 08 48 89 fb 49 39 f0 75 2a <4d> 8b 45 00 4d 39 c4 75 68 4c 39 e3 74 3e 4c 39 eb 74 39 49 89
RIP [<ffffffff8dd909fb>] __list_add+0x1b/0xc0
Fixes: e546e281d33d("drm/i915/gvt: Dmabuf support for GVT-g")
Signed-off-by: Hang Yuan <hang.yuan@xxxxxxxxxxxxxxx>
---
drivers/gpu/drm/i915/gvt/kvmgt.c | 1 -
drivers/gpu/drm/i915/gvt/vgpu.c | 1 +
2 files changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/gpu/drm/i915/gvt/kvmgt.c b/drivers/gpu/drm/i915/gvt/kvmgt.c
index a22d539..cbdf915e 100644
--- a/drivers/gpu/drm/i915/gvt/kvmgt.c
+++ b/drivers/gpu/drm/i915/gvt/kvmgt.c
@@ -1615,7 +1615,6 @@ static int kvmgt_guest_init(struct mdev_device *mdev)
kvmgt_protect_table_init(info);
gvt_cache_init(vgpu);
- mutex_init(&vgpu->dmabuf_lock);
init_completion(&vgpu->vblank_done);
info->track_node.track_write = kvmgt_page_track_write;
diff --git a/drivers/gpu/drm/i915/gvt/vgpu.c b/drivers/gpu/drm/i915/gvt/vgpu.c
index f6fa916..0bc1f1e 100644
--- a/drivers/gpu/drm/i915/gvt/vgpu.c
+++ b/drivers/gpu/drm/i915/gvt/vgpu.c
@@ -361,6 +361,7 @@ static struct intel_vgpu *__intel_gvt_create_vgpu(struct intel_gvt *gvt,
vgpu->gvt = gvt;
vgpu->sched_ctl.weight = param->weight;
mutex_init(&vgpu->vgpu_lock);
+ mutex_init(&vgpu->dmabuf_lock);
INIT_LIST_HEAD(&vgpu->dmabuf_obj_list_head);
INIT_RADIX_TREE(&vgpu->page_track_tree, GFP_KERNEL);
idr_init(&vgpu->object_idr);
