drm/amdgpu: fix ref count leak in amdgpu_driver_open_kms
authorNavid Emamdoost <navid.emamdoost@gmail.com>
Sun, 14 Jun 2020 07:12:29 +0000 (02:12 -0500)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Thu, 3 Sep 2020 09:21:16 +0000 (11:21 +0200)
[ Upstream commit 9ba8923cbbe11564dd1bf9f3602add9a9cfbb5c6 ]

in amdgpu_driver_open_kms the call to pm_runtime_get_sync increments the
counter even in case of failure, leading to incorrect
ref count. In case of failure, decrement the ref count before returning.

Signed-off-by: Navid Emamdoost <navid.emamdoost@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c

index ab5134d920d9620b4ebcd92c12818d2f23e870c8..96fc1566f28e5fa26cddbb7df1b9f6641b2081d2 100644 (file)
@@ -543,7 +543,7 @@ int amdgpu_driver_open_kms(struct drm_device *dev, struct drm_file *file_priv)
 
        r = pm_runtime_get_sync(dev->dev);
        if (r < 0)
-               return r;
+               goto pm_put;
 
        fpriv = kzalloc(sizeof(*fpriv), GFP_KERNEL);
        if (unlikely(!fpriv)) {
@@ -566,6 +566,7 @@ int amdgpu_driver_open_kms(struct drm_device *dev, struct drm_file *file_priv)
 
 out_suspend:
        pm_runtime_mark_last_busy(dev->dev);
+pm_put:
        pm_runtime_put_autosuspend(dev->dev);
 
        return r;