search for: poisoning_valu

...gt; } Can we put the value in config space instead? > We need "bool page_poisoning" because "poison_value=0" doesn't tell whether > page poising is in use by the guest. Can we use a feature bit for this? > PAGE_POISONING_ZERO sets > "page_poisoning=true, poisoning_value=0", and the host will send the > 0-filled pages to the destination (if not sending 0-filled pages, the > destination host would offer non-zero pages to the guest) Why would it? Linux zeroes all pages on alloc. > The host can discard free pages only when "page_poisoning=false&q...

...gt; } Can we put the value in config space instead? > We need "bool page_poisoning" because "poison_value=0" doesn't tell whether > page poising is in use by the guest. Can we use a feature bit for this? > PAGE_POISONING_ZERO sets > "page_poisoning=true, poisoning_value=0", and the host will send the > 0-filled pages to the destination (if not sending 0-filled pages, the > destination host would offer non-zero pages to the guest) Why would it? Linux zeroes all pages on alloc. > The host can discard free pages only when "page_poisoning=false&q...

On Wed, Nov 29, 2017 at 09:55:26PM +0800, Wei Wang wrote: > The guest free pages should not be discarded by the live migration thread > when page poisoning is enabled with PAGE_POISONING_NO_SANITY=n, because > skipping the transfer of such poisoned free pages will trigger false > positive when new pages are allocated and checked on the destination. > This patch skips the reporting

On Wed, Nov 29, 2017 at 09:55:26PM +0800, Wei Wang wrote: > The guest free pages should not be discarded by the live migration thread > when page poisoning is enabled with PAGE_POISONING_NO_SANITY=n, because > skipping the transfer of such poisoned free pages will trigger false > positive when new pages are allocated and checked on the destination. > This patch skips the reporting

...alloon_free_page_vq_hdr { bool page_poisoning; __virtio32 poison_value; __virtio32 cmd_id; } We need "bool page_poisoning" because "poison_value=0" doesn't tell whether page poising is in use by the guest. PAGE_POISONING_ZERO sets "page_poisoning=true, poisoning_value=0", and the host will send the 0-filled pages to the destination (if not sending 0-filled pages, the destination host would offer non-zero pages to the guest) The host can discard free pages only when "page_poisoning=false". Best, Wei