On 10/13/23 08:57, Dan Carpenter wrote:
On Thu, Oct 12, 2023 at 06:33:20PM +0200, Konrad Dybcio wrote:
@@ -1810,8 +1816,8 @@ int a6xx_gmu_init(struct a6xx_gpu *a6xx_gpu, struct
device_node *node)
return 0;
- if (!IS_ERR_OR_NULL(gmu->qmp))
- qmp_put(gmu->qmp);
+remove_device_link:
+ device_link_del(link);
This patch looks good to me but there is something happening here which
I don't understand. This is not related to you patch, but where does
the device_link_del() happen in the remove() function?
Well, you might have discovered another funny issue with the Adreno driver..
Cleaning up after the GPU failure was always an issue, that is HOPEFULLY
solved today, but perhaps nobody had ever imagined to test rmmod..
I'll take a deeper look later to check if it's not hidden in some API
Konrad
