drivers/gpu/drm/nouveau/nvkm/subdev/therm/temp.c: broken emergency poweroff handling

From: Pavel Machek
Date: Sat Sep 23 2017 - 15:04:18 EST


Hi!

Not only it is unneccessarily complex, it is also broken; GFP_ATOMIC
allocation can fail.. and then you fail to shut down the machine.

Someone please fix this.

Thanks,
Pavel

--- a/drivers/gpu/drm/nouveau/nvkm/subdev/therm/temp.c
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/therm/temp.c
@@ -120,6 +120,11 @@ nvkm_therm_sensor_event(struct nvkm_therm *therm, enum nvkm_therm_thrs thrs,
struct work_struct *work;

work = kmalloc(sizeof(*work), GFP_ATOMIC);
+ /* FIXME:
+ 1) this is total overkill, orderly_poweroff() already
+ uses schedule_work internally
+ 2) it would be good to at least printk what is going on
+ */
if (work) {
INIT_WORK(work, nv_poweroff_work);
schedule_work(work);


--
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html

Attachment: signature.asc
Description: Digital signature