)]}'
{
  "commit": "bf7ef7a2da3e61dc104f26c679c9465e3fbe7dde",
  "tree": "8ae92216df15dcd7d8a1215c298758fcad122fad",
  "parents": [
    "08fc4cb51774f763dcc6fd74637aa9e00eb6a0ba"
  ],
  "author": {
    "name": "Avihai Horon",
    "email": "avihaih@nvidia.com",
    "time": "Wed Sep 06 18:08:52 2023 +0300"
  },
  "committer": {
    "name": "Cédric Le Goater",
    "email": "clg@redhat.com",
    "time": "Mon Sep 11 08:34:06 2023 +0200"
  },
  "message": "vfio/migration: Block VFIO migration with postcopy migration\n\nVFIO migration is not compatible with postcopy migration. A VFIO device\nin the destination can\u0027t handle page faults for pages that have not been\nsent yet.\n\nDoing such migration will cause the VM to crash in the destination:\n\nqemu-system-x86_64: VFIO_MAP_DMA failed: Bad address\nqemu-system-x86_64: vfio_dma_map(0x55a28c7659d0, 0xc0000, 0xb000, 0x7f1b11a00000) \u003d -14 (Bad address)\nqemu: hardware error: vfio: DMA mapping failed, unable to continue\n\nTo prevent this, block VFIO migration with postcopy migration.\n\nReported-by: Yanghang Liu \u003cyanghliu@redhat.com\u003e\nSigned-off-by: Avihai Horon \u003cavihaih@nvidia.com\u003e\nTested-by: Yanghang Liu \u003cyanghliu@redhat.com\u003e\nReviewed-by: Peter Xu \u003cpeterx@redhat.com\u003e\nSigned-off-by: Cédric Le Goater \u003cclg@redhat.com\u003e\n",
  "tree_diff": [
    {
      "type": "modify",
      "old_id": "71855468fe985291e2d009b81c6efd29abcbe755",
      "old_mode": 33188,
      "old_path": "hw/vfio/migration.c",
      "new_id": "20994dc1d60b1606728415fec17c19cfd00c4dee",
      "new_mode": 33188,
      "new_path": "hw/vfio/migration.c"
    }
  ]
}