On 03.04.2024 23:23, Toke Høiland-Jørgensen wrote: > Heiner Kallweit <hkallweit1@xxxxxxxxx> writes: > >> Using the device-managed versions has no benefit here, because >> resources are released as part of the asynchronous fw loading. >> >> Actual reason why I got here is that I was looking for places with >> dubious use of pcim_pin_device(). > > Could you please expand the commit message to explain what this means in > a bit more detail? That will be helpful to someone who wants to > understand what the difference between these versions is and why this > change makes sense, but who is not that familiar with how these internal > bits are really supposed to work. Someone like myself, in other words :) > Sure. I'll add some more context in a v2. > -Toke Heiner
