* Laszlo Ersek (ler...@redhat.com) wrote: > On 06/02/17 17:45, Laszlo Ersek wrote: > > > The patches can cause linker/loader breakage when old firmware is booted > > on new QEMU. However, that's no problem (it's nothing new), the next > > release of QEMU should bundle the new firmware binaries as always. > > Dave made a good point (which I should have realized myself, really!), > namely if you launch old fw on old qemu, then migrate the guest to a new > qemu and then reboot the guest on the target host, within the migrated > VM, things will break.
Yep, sorry it always complicates things; as does the opposite case of back-migrating a VM with an old machine type but newer bios to an old qemu. Dave > So that makes this approach dead in the water. > > Possible mitigations I could think of: > - Make it machine type dependent. Complicated (we don't usually bind > ACPI generation to machine types) and wouldn't help existing devices. > - Let the firmware negotiate these extensions. Very complicated (new > fw-cfg files are needed for negotiation) and wouldn't help existing devices. > > So I guess I'll do what Igor and Gerd suggested: record in advance > whether any pointer field narrower than 8 bytes points into a given > blob, and if so, forbid allocating that blob from 64-bit address space. > This should solve Ard's needs purely within the firmware. > > Regarding the NOACPI hint, I guess I'm dropping that. I only meant > NOACPI for addressing Igor's long-standing dislike for the "ACPI SDT > header probe suppression" in VMGENID (and future similar devices). But, > there's no actual *technical* need to eliminate that (unlike the > technical need for 64-bit blob allocations which should really be > solved), so I guess it's OK to postpone NOACPI indefinitely. > > Self-nack for this set of sets. > > Thanks for the feedback, > Laszlo -- Dr. David Alan Gilbert / dgilb...@redhat.com / Manchester, UK
