On Tue, 2013-02-19 at 14:38 +0100, Andreas Färber wrote: > So far QEMU does not distinguish between reset types. Therefore I think > it is a bad idea to place QEMU_RESET_HARD in hw/cuda.c without reviewing > per case whether that is actually a hard or soft reset. > > Better and much less intrusive would be introducing separate functions > with the new argument, leaving the existing code unchanged until > reviewed and decided by their maintainers (which may be in the range of > months). Alternatively a QEMU_RESET_DEFAULT aliased to QEMU_RESET_HARD > to raise awareness where decisions need to be taken.
Yes, agreed. I'd want to set it up in such a fashion that each change from the default to a specific QEMU_RESET_XXX type was in a separate patch, for each device. And the implementation for i440fx would obviously be a separate patch too. -- dwmw2
smime.p7s
Description: S/MIME cryptographic signature
