On Mon, Dec 10, 2018 at 11:06:29AM +0530, Selvin Xavier wrote: > > If ib_unregister happens the driver must do the remove_device stuff > > immediately. > > > > So, I think you need to use that stuff I sent to Steve. This patch > > might be an improvement if the above is fixed, maybe, it isn't > > clear. (and Steve, now that I see the same mess in bnxt, I'm more than > > convinced that is the right approach to your problem too) > > > > However, I'd rather see this as part of a series that comprehensively > > fixes all of this. > > This patch is definitely an improvement on the existing driver > issues. I'm happier if we fix all the bug we can see when we start looking at an issue, rather than just the ones testing hits or something... > But i am okay to rework with your patch. I will take your patch, > test it and get back. Was your patch cut against for-next branch? > it was not applying cleanly on for-next. Yes.. It applies on top of e7521d82b33593c9b3ffd1e49a7ea2999ddc2285 Jason
