04:48 xeyler: my panel driver's probe function is called after the system loads the device tree which requires the driver
04:52 xeyler: but the drm subsystem seems to call the driver's enable function, then disable function almost immediately after
04:53 xeyler: why would the drm subsystem do this?
04:58 xeyler: in fact, it looks like the drm subsystem is running the enable and disable hooks concurrently. this behavior seems to be causing some errors with the driver trying to shutdown the device while it's half-way initialized
05:07 xeyler: i would have expected the drm subsystem to make some sort of guarantees regarding concurrency between enable and disable function invocations. is that not so? am i meant to protect atomic operations in the enable and disable hooks using a mutex?