)]}'
{"/COMMIT_MSG":[{"author":{"_account_id":11604,"name":"sean mooney","email":"smooney@redhat.com","username":"sean-k-mooney"},"change_message_id":"1ba1efe00def39ae77e5abdc1bec85b794521f24","unresolved":true,"context_lines":[{"line_number":12,"context_line":"a VFIO device on the destination cannot handle page faults for"},{"line_number":13,"context_line":"memory pages not yet transferred from the source."},{"line_number":14,"context_line":""},{"line_number":15,"context_line":"When live_migration_permit_post_copy is set to True (the default"},{"line_number":16,"context_line":"in some deployments), Nova unconditionally sets the"},{"line_number":17,"context_line":"VIR_MIGRATE_POSTCOPY flag on all live migrations. This causes"},{"line_number":18,"context_line":"vGPU live migrations to fail with:"},{"line_number":19,"context_line":""}],"source_content_type":"text/x-gerrit-commit-message","patch_set":1,"id":"df8b5e4f_542b501b","line":16,"range":{"start_line":15,"start_character":52,"end_line":16,"end_character":21},"updated":"2026-08-03 14:12:34.000000000","message":"i may be enable by default our of tree such as in our downstream installer\n\nbut its not enabled bv default upstream so we proably should not call this out in genreal.\n\nhttps://docs.openstack.org/nova/latest/configuration/config.html#libvirt.live_migration_permit_post_copy","commit_id":"9995c7b4cd6aa6cac59f2ec254edabfc40dc4436"},{"author":{"_account_id":7166,"name":"Sylvain Bauza","email":"sbauza@redhat.com","username":"sbauza"},"change_message_id":"a564b691fec1acbd0a1e0dbd704fc7ebff04f483","unresolved":false,"context_lines":[{"line_number":12,"context_line":"a VFIO device on the destination cannot handle page faults for"},{"line_number":13,"context_line":"memory pages not yet transferred from the source."},{"line_number":14,"context_line":""},{"line_number":15,"context_line":"When live_migration_permit_post_copy is set to True (the default"},{"line_number":16,"context_line":"in some deployments), Nova unconditionally sets the"},{"line_number":17,"context_line":"VIR_MIGRATE_POSTCOPY flag on all live migrations. This causes"},{"line_number":18,"context_line":"vGPU live migrations to fail with:"},{"line_number":19,"context_line":""}],"source_content_type":"text/x-gerrit-commit-message","patch_set":1,"id":"7634da8b_f6dc8079","line":16,"range":{"start_line":15,"start_character":52,"end_line":16,"end_character":21},"in_reply_to":"df8b5e4f_542b501b","updated":"2026-08-04 09:20:31.000000000","message":"Acknowledged","commit_id":"9995c7b4cd6aa6cac59f2ec254edabfc40dc4436"},{"author":{"_account_id":11604,"name":"sean mooney","email":"smooney@redhat.com","username":"sean-k-mooney"},"change_message_id":"2e89471b0a8caace84de2615f015da3719ed3dd0","unresolved":true,"context_lines":[{"line_number":28,"context_line":"Change-Id: Ibb1f2ace1a1509a8c3e1a12c3e18f397584ab702"},{"line_number":29,"context_line":"Assisted-by: Cursor \u003ccursoragent@cursor.com\u003e"},{"line_number":30,"context_line":"Signed-off-by: Sylvain Bauza \u003csbauza@redhat.com\u003e"},{"line_number":31,"context_line":"Co-authored-by: Cursor \u003ccursoragent@cursor.com\u003e"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":2,"id":"0e60edce_6e99f43a","line":31,"range":{"start_line":31,"start_character":0,"end_line":31,"end_character":2},"updated":"2026-08-04 09:44:45.000000000","message":"this is incorrect.\n\nthere is a cursor settign to disable it addign this or made-with: cursor trailer","commit_id":"73982aa84e100c67692ccb8b9ba204eef54f5191"}],"/PATCHSET_LEVEL":[{"author":{"_account_id":11604,"name":"sean mooney","email":"smooney@redhat.com","username":"sean-k-mooney"},"change_message_id":"1ba1efe00def39ae77e5abdc1bec85b794521f24","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"a2f2255f_f005beac","updated":"2026-08-03 14:12:34.000000000","message":"i guess this is ok i generally prefer avoiding the term vgpu in our docs\ngiven nova offically suprpot generic mdev device passthough ro pci-passthough with vfio-vairant drivers. \n\nvgpu is such an overloaded term that i find it less useful then the underlying technology.","commit_id":"9995c7b4cd6aa6cac59f2ec254edabfc40dc4436"},{"author":{"_account_id":11604,"name":"sean mooney","email":"smooney@redhat.com","username":"sean-k-mooney"},"change_message_id":"2e89471b0a8caace84de2615f015da3719ed3dd0","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":2,"id":"e4e0e070_e8e9f877","updated":"2026-08-04 09:44:45.000000000","message":"ok i think the doc change is fine now but we shoudl not use Co-Authored-By for llms\ntehre are copyright implcitions to that that do not hold for ai agents\n\nonce htat is fixed im happy to proceed with this","commit_id":"73982aa84e100c67692ccb8b9ba204eef54f5191"}],"doc/source/admin/live-migration-usage.rst":[{"author":{"_account_id":11604,"name":"sean mooney","email":"smooney@redhat.com","username":"sean-k-mooney"},"change_message_id":"1ba1efe00def39ae77e5abdc1bec85b794521f24","unresolved":true,"context_lines":[{"line_number":318,"context_line":"     - Post-copy may lead to an increased page fault rate during migration,"},{"line_number":319,"context_line":"       which can slow the instance down."},{"line_number":320,"context_line":""},{"line_number":321,"context_line":"     - Post-copy is incompatible with VFIO device migration (including"},{"line_number":322,"context_line":"       vGPU/mediated devices). QEMU will reject any live migration that"},{"line_number":323,"context_line":"       combines a VFIO device with postcopy capabilities. If your deployment"},{"line_number":324,"context_line":"       uses vGPU live migration, set ``live_migration_permit_post_copy\u003dfalse``"}],"source_content_type":"text/x-rst","patch_set":1,"id":"b083c2ab_a5ac0f6c","line":321,"range":{"start_line":321,"start_character":38,"end_line":321,"end_character":60},"updated":"2026-08-03 14:12:34.000000000","message":"is this true for both vfio variant driver and vfio-mediated devices.","commit_id":"9995c7b4cd6aa6cac59f2ec254edabfc40dc4436"},{"author":{"_account_id":9708,"name":"Balazs Gibizer","display_name":"gibi","email":"gibizer@gmail.com","username":"gibi"},"change_message_id":"52e76a1987112b495ace0dece7fb3a9a06e0aaf6","unresolved":true,"context_lines":[{"line_number":318,"context_line":"     - Post-copy may lead to an increased page fault rate during migration,"},{"line_number":319,"context_line":"       which can slow the instance down."},{"line_number":320,"context_line":""},{"line_number":321,"context_line":"     - Post-copy is incompatible with VFIO device migration (including"},{"line_number":322,"context_line":"       vGPU/mediated devices). QEMU will reject any live migration that"},{"line_number":323,"context_line":"       combines a VFIO device with postcopy capabilities. If your deployment"},{"line_number":324,"context_line":"       uses vGPU live migration, set ``live_migration_permit_post_copy\u003dfalse``"}],"source_content_type":"text/x-rst","patch_set":1,"id":"d03a8930_c7dd78f3","line":321,"range":{"start_line":321,"start_character":38,"end_line":321,"end_character":60},"in_reply_to":"b083c2ab_a5ac0f6c","updated":"2026-08-04 08:36:14.000000000","message":"I\u0027m not sure based on https://lists.nongnu.org/archive/html/qemu-devel/2023-08/msg04794.html if this is only SRIOV + vfio variant driver, or mdev or both.\n\nAs far as I see https://review.opendev.org/c/openstack/nova/+/999607/1/doc/source/admin/virtual-gpu.rst#470 is only added to the mdev based GPU migration not for the SRIOV + variant driver case.","commit_id":"9995c7b4cd6aa6cac59f2ec254edabfc40dc4436"},{"author":{"_account_id":7166,"name":"Sylvain Bauza","email":"sbauza@redhat.com","username":"sbauza"},"change_message_id":"a564b691fec1acbd0a1e0dbd704fc7ebff04f483","unresolved":false,"context_lines":[{"line_number":318,"context_line":"     - Post-copy may lead to an increased page fault rate during migration,"},{"line_number":319,"context_line":"       which can slow the instance down."},{"line_number":320,"context_line":""},{"line_number":321,"context_line":"     - Post-copy is incompatible with VFIO device migration (including"},{"line_number":322,"context_line":"       vGPU/mediated devices). QEMU will reject any live migration that"},{"line_number":323,"context_line":"       combines a VFIO device with postcopy capabilities. If your deployment"},{"line_number":324,"context_line":"       uses vGPU live migration, set ``live_migration_permit_post_copy\u003dfalse``"}],"source_content_type":"text/x-rst","patch_set":1,"id":"fbd190b3_5107f34e","line":321,"range":{"start_line":321,"start_character":38,"end_line":321,"end_character":60},"in_reply_to":"d03a8930_c7dd78f3","updated":"2026-08-04 09:20:31.000000000","message":"Done","commit_id":"9995c7b4cd6aa6cac59f2ec254edabfc40dc4436"}]}