On Mon, 2024-08-19 at 13:24 +0300, Nikolay Borisov wrote: > > I wonder why this patch and patch 9 weren't squashed together. Many changes > > added by patch 9 are removed here. > > As far as I can see this patch depends on the code in patch 10 > (kvm_tdx_caps) so this patch definitely must come after changes > introduced in patch 10. However, patch 9 seems completely independent of > patch 10, so I think patch 10 should become patch 9, and patch 9/11 > should be squashed into one and become patch 10. Yes, thanks. The patch order needs to be cleaned up. This posting was mostly intended to try to settle the whole guest CPU feature API design. I probably should have tagged it RFC instead of just including the coverletter blurb: Please feel free to wait for future revisions to spend time trying to correct smaller code issues. But I would greatly appreciate discussion on the overall design and how we are weighing the tradeoffs for the uAPI.