Johny Mattsson <[email protected]> writes: > On 21 August 2014 19:48, Markus Armbruster <[email protected]> wrote: >> >> Doesn't this break migration to/from unfixed versions? >> > > It might. It is not something we've tried here, and I'm certainly not > familiar enough with QEMU internals myself to say one way or another. We > encountered this issue in what I can best describe as a short-and-furious > proof-of-concept project, and as part of the wrapping up of this I'm going > over our change logs to see if anything can be contributed back upstream. > It seemed like a trivial fix in the tree, but I acknowledge it may have > ramifications we didn't consider. > > >> I suspect you need a subsection. Initialize ->romd = 1, then let the >> subsection override it. >> > > And here I'm well and truly out of my depth! If someone who actually knows > what they're doing would like to run with this, please do. > > Sorry to effectively do a dump-and-run here - I hope its a net positive at > least.
Sure it is! Imperfect patches can be a rather nice kind of bug report.
