Re: [PATCH RFC 0/9] qemu: Support mapped-ram migration capability

On 8/7/24 09:45, Daniel P. Berrangé wrote: > Annoyingly we already have a 'save_image_formt' in qemu.conf though > taking  'raw', 'zstd', 'lzop', etc to choose the compression type. > So we have a terminology clash. Thinking about this more, and your previous idea about abusing the header 'compressed' field [1], I'm beginning to warm to the idea of adding a new item to the list currently accepted by save_image_format. I'm also warming to Fabiano's suggestion to call the new item 'sparse' :-). AFAICT, from a user perspective, save_image_format does not imply compression. The implication primarily stems from variable and function naming in the code :-). The current documentation of save_image_format may slightly imply compression, but we can easily improve that. The setting already accepts 'raw' for the existing (non-compressed) stream format.

[1] https://lists.libvirt.org/archives/list/devel@lists.libvirt.org/message/P...