Hi,
I will send three patch series in response to this "anchor" post:
[wave 1 PATCH 0/5] OvmfPkg: AcpiPlatformDxe: enable PCI IO and MMIO
while fetching QEMU tables
[wave 2 PATCH 0/3] OvmfPkg: enable 64-bit PCI MMIO
[wave 3 PATCH 00/15] OvmfPkg, ArmVirtPkg: VirtIo 1.0
The patches are also on github (see the links in the individual blurbs);
the branches are stacked accordingly to the order of waves as well. In
other words, if you check out the branch that corresponds to wave N, you
also get the patches for all earlier waves.
The patches are organized in waves to recognize the following:
- Virtio-1.0 devices really like MMIO BARs (as opposed to IO BARs of
legacy (0.9.5) devices), and they can use 64-bit MMIO BARs as well.
- Using 64-bit MMIO BARs triggers the ordering bug we *still* have in
AcpiPlatformDxe real hard, so that's what the first wave addresses.
I formatted the patches with 7 lines of context. I think they shouldn't
be excessive, and they really helped me re-review the patches in "gitk".
... I'm sure I'm forgetting something, but after ~3 days of pure coding,
testing, plus also chasing that sneaky little QEMU bug (--> see the
blurb of wave 3), I'm not exactly fast. I'll be mostly offline for today
and probably tomorrow, and just lie in the Sun. (It's spring! :))
Thanks
Laszlo
_______________________________________________
edk2-devel mailing list
[email protected]
https://lists.01.org/mailman/listinfo/edk2-devel
