Thanks everyone who joined this morning, or evening/afternoon depending on where you are! The slides and recording are available on Drive (see bottom), though unfortunately non-Google people will need to request access as our corporate settings don't provide a way to share with literally everyone. This *should* be a one-time pain though, and I'll do my best to grant access immediately. Key Takeaways: - No objections to pursuing a KVM ioctl(), e.g. KVM_CREATE_GUEST_MEMFD. - A KVM ioctl() is cleaner than a generic syscall for pKVM and SNP, and at worst a wash for TDX. Next Steps: - Vitaly to follow-up with Hyper-V folks to see if "Linux as the root partition" needs/wants a generic syscall for managing guest memory. - Sean to post KVM_CREATE_GUEST_MEMFD implementation as v11, and to coordinate with owners of dependent series to sort out how to iterate efficiently. Future Schedule: June 7th - pKVM on x86 (Google + Intel) June 14th - Available! June 21st - No Meeting (Sean OOO) June 28th - Available! July 5th - No Meeting (Sean OOO) Calendar: https://calendar.google.com/calendar/u/0?cid=Y182MWE1YjFmNjQ0NzM5YmY1YmVkN2U1ZWE1ZmMzNjY5Y2UzMmEyNTQ0YzVkYjFjN2M4OTE3MDJjYTUwOTBjN2Q1QGdyb3VwLmNhbGVuZGFyLmdvb2dsZS5jb20 Drive: https://drive.google.com/drive/folders/1aTqCrvTsQI9T4qLhhLs_l986SngGlhPH?resourcekey=0-FDy0ykM3RerZedI8R-zj4A&usp=drive_link
