On Wed, Oct 12, 2016 at 02:26:26PM +0100, George Dunlap wrote:
> >> There is a version field in gcov_info, so we can compare that and reject
> >> incompatible version.
> >> We need to use hooks in livepatching to call the constructor /
> >> destructor when applying / reverting a live-patch. We might need to be
> >> cautious about locks or whatever, but I'm sure it can be figured out.
> >> But I have no idea how useful it would be to use gcov and livepatching
> >> together. For now the easiest thing to do is to
> >> depends on !LIVEPATCH
> >> in Kconfig.
> > Wouldn't it be just as easy, and more useful, to set a "has been
> > livepatched" flag, and return errors for all gcov hypercalls if its' set?
> > I would expect most users to want to build a single hypervisor that can
> > be used for both gcov testing and live patching (under different
> > circumstances).
> I mean software provider, not user, of course. That's what I would want
> for CentOS, and I'm sure that's what the XenServer (and probably Oracle)
> guys want as well.
The usefulness of such build is not as big as you might think I'm
afraid -- it wouldn't be useful until we figure out how to apply a
livepatch generated with gcov support built in.
Xen-devel mailing list