On 2024/07/21 0:16, Phil Dennis-Jordan wrote:
On Sat 20. Jul 2024 at 16:42, Akihiko Odaki <akihiko.od...@daynix.com <mailto:akihiko.od...@daynix.com>> wrote:> It also became clear in out-of-band communication that Alexander would > probably not end up having the time to see the patch through to inclusion, > and was happy for me to start making changes and to integrate my PCI code. I think you also need to take over the base vmapple change because PVG cannot be tested without it; Only macOS can use PVG, and macOS requires vmapple in my understanding.The PCI device variant works fine with x86-64 macOS as the guest system. (Or technically any UEFI based guest as the bootrom comes with a UEFI frame buffer driver; it just won’t have any acceleration features unless you boot macOS.)
I tried apple-gfx-pci with AArch64 and x86-64 EDK firmware on M2 MacBook Air, but it didn't work. I guess the bootrom does not work with Apple Silicon.
If preferable I can leave out the vmapple/MMIO device variant (the -vmapple.m file) until the rest of the vmapple machine type modifications are ready. There still appears to be some kind of interrupt delivery issue with some devices on vmapple, so USB HID events are very slow. Or I can submit it as is if that’s not a dealbreaker.
The vmapple variant should be omitted for now. It is fine for me to submit the vmapple variant with the other vmapple changes as long as that behavior is documented.
(How do I handle Alex’ unmodified patches though, as git send-email tries to send them from the patch author’s email address, which means the email rapidly gets shot down by DKIM mismatch or whatever?)
I think you can set sendemail.from configuration or specify --from option. Regards, Akihiko Odaki
