On Mon, 17 Jun, at 12:02:05PM, Jan Beulich wrote: > >>> On 17.06.13 at 12:17, Matt Fleming <matt@xxxxxxxxxxxxxxxxx> wrote: > > > > What value would you suggest for the retry? > > I'd be considering something like 5...10, but this to some degree > depends on what odd kinds of behavior this in fact needs to > cover. I genuinely don't see how picking numbers (seemingly) at random is an improvement over the simple case of "if ExitBootServices() returns an error, you get one more try". Unless you're aware of firmware that requires calling ExitBootServices() multiple times? The grub ChangeLog that Joey linked to is not particularly enlightening. The scenario that this patch addresses is a very clear one, where the firmware event handlers that respond to the initial ExitBootServices event allocate/free memory, thereby invalidating the memory map cookie we passed to ExitBootServices(), and requiring us to call ExitBootServices() a second time. I'm not saying that retrying once is the only solution that will ever make sense. But certainly until we start seeing reports of problems and understand why firmware might want us to invoke ExitBootServices() multiple times, it seems like the most straight forward option. -- Matt Fleming, Intel Open Source Technology Center -- To unsubscribe from this list: send the line "unsubscribe stable" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html