David Woodhouse <dw...@infradead.org> writes: > On Thu, 2013-02-21 at 18:10 +0100, Laszlo Ersek wrote: >> Since >> - I like kvm, and >> - I cannot easily change my hardware (which doesn't support UG), and >> - I prefer to run the RHEL-6 kernel which has "old" KVM, >> I depend on David's fix for the PAM registers. > > Which means your suspend/resume is broken. Yay! :) > > Anthony's patchset should hopefully fix that.
Will post a new version later today. Am testing now with Laszlo's mock up from a previous note. I'm also going to write a qtest test case to try and validate all of this behavior (minus the KVM bits). > I wonder what it would take to work around the KVM bug in qemu though; > can we actually *emulate* 16-bit mode instead of using KVM, on the > kernels where KVM is broken? And then transition back to using KVM when > we switch to 32-bit or 64-bit mode? It's in the realm of possibility but ugly enough that we never did it. I even worked on this way back in the day with Xen. It sucked pretty bad. The challenge in support SMP eventually scared me off of it. http://new-wiki.xen.org/old-wiki/xenwiki/HVM/V2E.html Regards, Anthony Liguori > > -- > dwmw2
