⬢ DragonFlyBSD Kernel Audit
DF-4818 / manifest.json
← back to finding ↓ download raw 
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
{
  "finding_id": "DF-4818",
  "verdict": "inconclusive",
  "impact": "panic",
  "reproduce": {
    "build": "(nothing to build — config-driven trigger; ./build.sh)",
    "run": "findings/poc/DF-4818/run.sh",
    "expected": "phase1: 'nvme0: mapped 7 MSIX IRQs'; phase2: 'nvme0: mapped 1 normal-int IRQs' with disk attached (i=0 vector failure -> clean fallback); phase3: nvme0-20 'mapped 3 MSIX IRQs', nvme21-23 fallback after a mid-loop unwind with a live allocated vector — no nvme panic in any phase (the panic in the logs is vtpci/vtnet, a different driver's bug captured incidentally). The DF-4818 panic itself (nexus_setup_intr/resource_list_release NULL) needs rman_reserve_resource to fail on a freshly granted MSI-X irq and is not forceable on QEMU — see VERDICT.md."
  },
  "kernel_refs": [
    "sys/dev/disk/nvme/nvme_attach.c:166",
    "sys/dev/disk/nvme/nvme_attach.c:239",
    "sys/dev/disk/nvme/nvme_attach.c:193",
    "sys/dev/disk/nvme/nvme.c:823",
    "sys/platform/pc64/x86_64/nexus.c:505",
    "sys/kern/subr_bus.c:2738",
    "sys/bus/pci/pci.c:1644"
  ],
  "artifacts": [
    "README.md",
    "VERDICT.md",
    "run.sh",
    "build.sh",
    "run.log",
    "run.2.log",
    "run.3.log",
    "panic.txt",
    "env.txt",
    "fix.diff",
    "verdict.json",
    "manifest.json"
  ]
}