> From: Peter Crosthwaite [mailto:crosthwaitepe...@gmail.com] > Sent: Wednesday, 9 December 2015 10:55 > > Yes, that's the same change I made. Is it going to be applied? > > > > It missed the boat for 2.5, but you could help by putting a tested-by > or reviewed-by to the patch.
Ok. > > The complete diff needed to boot Windows appears below. The first hunk > avoids re-triggering the insert interrupt on card reset, > > Is this a runtime reset or an initial reset that is causing you grief? > You patch is also patching the controller reset so is it more a case > of the interrupt trigger on controller reset that is causing you > issues? It's the software-issued controller reset... at board power on, the insert interrupt bit is set (and UEFI depends on seeing it set), but when the Windows driver starts, it issues a reset command which clears the interrupt status. It also has a bug which livelocks the system if an insert interrupt is seen after that reset (which never happens on hardware, obviously). Andrew
