diff --git a/plugins/hypervisors/kvm/src/main/java/com/cloud/hypervisor/kvm/resource/LibvirtGpuDef.java b/plugins/hypervisors/kvm/src/main/java/com/cloud/hypervisor/kvm/resource/LibvirtGpuDef.java
index 08086859fb70..d3765c01ccbc 100644
--- a/plugins/hypervisors/kvm/src/main/java/com/cloud/hypervisor/kvm/resource/LibvirtGpuDef.java
+++ b/plugins/hypervisors/kvm/src/main/java/com/cloud/hypervisor/kvm/resource/LibvirtGpuDef.java
@@ -79,28 +79,47 @@ private void generatePciXml(StringBuilder gpuBuilder) {
gpuBuilder.append(" \n");
gpuBuilder.append(" \n");
- // Parse the bus address (e.g., 00:02.0) into domain, bus, slot, function
- String domain = "0x0000";
- String bus = "0x00";
- String slot = "0x00";
- String function = "0x0";
+ // Parse the bus address into domain, bus, slot, function. Two input formats are accepted:
+ // - "dddd:bb:ss.f" full PCI address with domain (e.g. 0000:00:02.0)
+ // - "bb:ss.f" legacy short BDF; domain defaults to 0000
+ // Each segment is parsed as a hex integer and formatted with fixed widths
+ // (domain: 4 hex digits, bus/slot: 2 hex digits, function: 1 hex digit) to
+ // produce canonical libvirt XML values regardless of input casing or padding.
+ int domainVal = 0, busVal = 0, slotVal = 0, funcVal = 0;
if (busAddress != null && !busAddress.isEmpty()) {
String[] parts = busAddress.split(":");
- if (parts.length > 1) {
- bus = "0x" + parts[0];
- String[] slotFunctionParts = parts[1].split("\\.");
- if (slotFunctionParts.length > 0) {
- slot = "0x" + slotFunctionParts[0];
- if (slotFunctionParts.length > 1) {
- function = "0x" + slotFunctionParts[1].trim();
- }
+ try {
+ String slotFunction;
+ if (parts.length == 3) {
+ domainVal = Integer.parseInt(parts[0], 16);
+ busVal = Integer.parseInt(parts[1], 16);
+ slotFunction = parts[2];
+ } else if (parts.length == 2) {
+ busVal = Integer.parseInt(parts[0], 16);
+ slotFunction = parts[1];
+ } else {
+ throw new IllegalArgumentException("Invalid PCI bus address format: '" + busAddress + "'");
}
+ String[] sf = slotFunction.split("\\.");
+ if (sf.length == 2) {
+ slotVal = Integer.parseInt(sf[0], 16);
+ funcVal = Integer.parseInt(sf[1].trim(), 16);
+ } else {
+ throw new IllegalArgumentException("Invalid PCI bus address format: '" + busAddress + "'");
+ }
+ } catch (NumberFormatException e) {
+ throw new IllegalArgumentException("Invalid PCI bus address format: '" + busAddress + "'", e);
}
}
+ String domain = String.format("0x%04x", domainVal);
+ String bus = String.format("0x%02x", busVal);
+ String slot = String.format("0x%02x", slotVal);
+ String function = String.format("0x%x", funcVal);
+
gpuBuilder.append(" \n");
+ .append(slot).append("' function='").append(function).append("'/>\n");
gpuBuilder.append(" \n");
gpuBuilder.append("\n");
}
diff --git a/plugins/hypervisors/kvm/src/test/java/com/cloud/hypervisor/kvm/resource/LibvirtGpuDefTest.java b/plugins/hypervisors/kvm/src/test/java/com/cloud/hypervisor/kvm/resource/LibvirtGpuDefTest.java
index 0060e1d7ed4d..e6f63e852f85 100644
--- a/plugins/hypervisors/kvm/src/test/java/com/cloud/hypervisor/kvm/resource/LibvirtGpuDefTest.java
+++ b/plugins/hypervisors/kvm/src/test/java/com/cloud/hypervisor/kvm/resource/LibvirtGpuDefTest.java
@@ -115,6 +115,145 @@ public void testGpuDef_withComplexPciAddress() {
assertTrue(gpuXml.contains(""));
}
+ @Test
+ public void testGpuDef_withFullPciAddressDomainZero() {
+ LibvirtGpuDef gpuDef = new LibvirtGpuDef();
+ VgpuTypesInfo pciGpuInfo = new VgpuTypesInfo(
+ GpuDevice.DeviceType.PCI,
+ "passthrough",
+ "passthrough",
+ "0000:00:02.0",
+ "10de",
+ "NVIDIA Corporation",
+ "1b38",
+ "Tesla T4"
+ );
+ gpuDef.defGpu(pciGpuInfo);
+
+ String gpuXml = gpuDef.toString();
+
+ assertTrue(gpuXml.contains(""));
+ }
+
+ @Test
+ public void testGpuDef_withFullPciAddressNonZeroDomain() {
+ LibvirtGpuDef gpuDef = new LibvirtGpuDef();
+ VgpuTypesInfo pciGpuInfo = new VgpuTypesInfo(
+ GpuDevice.DeviceType.PCI,
+ "passthrough",
+ "passthrough",
+ "0001:65:00.0",
+ "10de",
+ "NVIDIA Corporation",
+ "1b38",
+ "Tesla T4"
+ );
+ gpuDef.defGpu(pciGpuInfo);
+
+ String gpuXml = gpuDef.toString();
+
+ assertTrue(gpuXml.contains(""));
+ }
+
+ @Test
+ public void testGpuDef_withNvidiaStyleEightDigitDomain() {
+ // nvidia-smi reports PCI addresses with an 8-digit domain (e.g. "00000001:af:00.1").
+ // generatePciXml must normalize it to the canonical 4-digit form "0x0001".
+ LibvirtGpuDef gpuDef = new LibvirtGpuDef();
+ VgpuTypesInfo pciGpuInfo = new VgpuTypesInfo(
+ GpuDevice.DeviceType.PCI,
+ "passthrough",
+ "passthrough",
+ "00000001:af:00.1",
+ "10de",
+ "NVIDIA Corporation",
+ "1b38",
+ "Tesla T4"
+ );
+ gpuDef.defGpu(pciGpuInfo);
+
+ String gpuXml = gpuDef.toString();
+
+ assertTrue(gpuXml.contains(""));
+ }
+
+ @Test
+ public void testGpuDef_withFullPciAddressVfNonZeroDomain() {
+ LibvirtGpuDef gpuDef = new LibvirtGpuDef();
+ VgpuTypesInfo vfGpuInfo = new VgpuTypesInfo(
+ GpuDevice.DeviceType.PCI,
+ "VF-Profile",
+ "VF-Profile",
+ "0002:81:00.3",
+ "10de",
+ "NVIDIA Corporation",
+ "1eb8",
+ "Tesla T4"
+ );
+ gpuDef.defGpu(vfGpuInfo);
+
+ String gpuXml = gpuDef.toString();
+
+ // Non-passthrough NVIDIA VFs should be unmanaged
+ assertTrue(gpuXml.contains(""));
+ assertTrue(gpuXml.contains(""));
+ }
+
+ @Test
+ public void testGpuDef_withLegacyShortBdfDefaultsDomainToZero() {
+ // Backward compatibility: short BDF with no domain segment must still
+ // produce a valid libvirt address with domain 0x0000.
+ LibvirtGpuDef gpuDef = new LibvirtGpuDef();
+ VgpuTypesInfo pciGpuInfo = new VgpuTypesInfo(
+ GpuDevice.DeviceType.PCI,
+ "passthrough",
+ "passthrough",
+ "af:00.0",
+ "10de",
+ "NVIDIA Corporation",
+ "1b38",
+ "Tesla T4"
+ );
+ gpuDef.defGpu(pciGpuInfo);
+
+ String gpuXml = gpuDef.toString();
+
+ assertTrue(gpuXml.contains(""));
+ }
+
+ @Test
+ public void testGpuDef_withInvalidBusAddressThrows() {
+ String[] invalidAddresses = {
+ "notahex:00.0", // non-hex bus
+ "gg:00:02.0", // non-hex domain
+ "00:02:03:04", // too many colon-separated parts
+ "00", // missing slot/function
+ "00:02", // missing function (no dot)
+ "00:02.0.1", // extra dot in ss.f
+ };
+ for (String addr : invalidAddresses) {
+ LibvirtGpuDef gpuDef = new LibvirtGpuDef();
+ VgpuTypesInfo info = new VgpuTypesInfo(
+ GpuDevice.DeviceType.PCI,
+ "passthrough",
+ "passthrough",
+ addr,
+ "10de",
+ "NVIDIA Corporation",
+ "1b38",
+ "Tesla T4"
+ );
+ gpuDef.defGpu(info);
+ try {
+ String ignored = gpuDef.toString();
+ fail("Expected IllegalArgumentException for address: " + addr + " but got: " + ignored);
+ } catch (IllegalArgumentException e) {
+ assertTrue("Exception message should contain the bad address",
+ e.getMessage().contains(addr));
+ }
+ }
+ }
+
@Test
public void testGpuDef_withNullDeviceType() {
LibvirtGpuDef gpuDef = new LibvirtGpuDef();
diff --git a/scripts/vm/hypervisor/kvm/gpudiscovery.sh b/scripts/vm/hypervisor/kvm/gpudiscovery.sh
index bf6892c898db..524f86241860 100755
--- a/scripts/vm/hypervisor/kvm/gpudiscovery.sh
+++ b/scripts/vm/hypervisor/kvm/gpudiscovery.sh
@@ -357,7 +357,33 @@
# "vgpu_instances":[],
# "vf_instances":[]
-# }
+# },
+# {
+# "pci_address": "0001:65:00.0",
+# "vendor_id": "10de",
+# "device_id": "20B0",
+# "vendor": "NVIDIA Corporation",
+# "device": "A100-SXM4-40GB",
+# "driver": "nvidia",
+# "pci_class": "3D controller",
+# "iommu_group": "20",
+# "sriov_totalvfs": 0,
+# "sriov_numvfs": 0,
+#
+# "full_passthrough": {
+# "enabled": true,
+# "libvirt_address": {
+# "domain": "0x0001",
+# "bus": "0x65",
+# "slot": "0x00",
+# "function": "0x0"
+# },
+# "used_by_vm": "ml-train"
+# },
+#
+# "vgpu_instances": [],
+# "vf_instances": []
+# }
# ]
# }
#
@@ -416,9 +442,18 @@ parse_nvidia_vgpu_profiles() {
store_profile_data
gpu_address="${BASH_REMATCH[1]}"
- # Convert from format like 00000000:AF:00.0 to AF:00.0 and normalize to lowercase
- if [[ $gpu_address =~ [0-9A-Fa-f]+:([0-9A-Fa-f]+:[0-9A-Fa-f]+\.[0-9A-Fa-f]+) ]]; then
- gpu_address="${BASH_REMATCH[1],,}"
+ # nvidia-smi reports addresses in the form "00000000:AF:00.0"
+ # (8-digit domain). Reformat to the canonical 4-digit
+ # "dddd:bb:ss.f" lowercase form so cache keys line up with the
+ # normalize_pci_address output used at lookup time. Short
+ # addresses ("AF:00.0") are widened by prepending domain 0.
+ if [[ $gpu_address =~ ^([0-9A-Fa-f]+):([0-9A-Fa-f]+):([0-9A-Fa-f]+)\.([0-9A-Fa-f]+)$ ]]; then
+ gpu_address=$(printf "%04x:%02x:%02x.%x" \
+ "0x${BASH_REMATCH[1]}" "0x${BASH_REMATCH[2]}" \
+ "0x${BASH_REMATCH[3]}" "0x${BASH_REMATCH[4]}")
+ elif [[ $gpu_address =~ ^([0-9A-Fa-f]+):([0-9A-Fa-f]+)\.([0-9A-Fa-f]+)$ ]]; then
+ gpu_address=$(printf "0000:%02x:%02x.%x" \
+ "0x${BASH_REMATCH[1]}" "0x${BASH_REMATCH[2]}" "0x${BASH_REMATCH[3]}")
else
gpu_address="${gpu_address,,}"
fi
@@ -506,10 +541,58 @@ get_nvidia_profile_info() {
fi
}
-# Get nodedev name for a PCI address (e.g. "00:02.0" -> "pci_0000_00_02_0")
-get_nodedev_name() {
+# Parse a PCI address and assign the libvirt-formatted DOMAIN/BUS/SLOT/FUNC
+# values into the caller's scope. Accepts both full ("0000:00:02.0") and short
+# ("00:02.0") formats; short addresses are assumed to be in PCI domain 0.
+#
+# IMPORTANT: bash uses dynamic scoping, so callers that don't want these four
+# values to leak into the global scope MUST declare them `local` before
+# invoking this function, e.g.:
+# local DOMAIN BUS SLOT FUNC
+# parse_pci_address "$addr"
+parse_pci_address() {
local addr="$1"
- echo "pci_$(echo "$addr" | sed 's/[:.]/\_/g' | sed 's/^/0000_/')"
+ if [[ $addr =~ ^([0-9A-Fa-f]+):([0-9A-Fa-f]+):([0-9A-Fa-f]+)\.([0-9A-Fa-f]+)$ ]]; then
+ DOMAIN=$(printf "0x%04x" "0x${BASH_REMATCH[1]}")
+ BUS=$(printf "0x%02x" "0x${BASH_REMATCH[2]}")
+ SLOT=$(printf "0x%02x" "0x${BASH_REMATCH[3]}")
+ FUNC=$(printf "0x%x" "0x${BASH_REMATCH[4]}")
+ elif [[ $addr =~ ^([0-9A-Fa-f]+):([0-9A-Fa-f]+)\.([0-9A-Fa-f]+)$ ]]; then
+ DOMAIN="0x0000"
+ BUS=$(printf "0x%02x" "0x${BASH_REMATCH[1]}")
+ SLOT=$(printf "0x%02x" "0x${BASH_REMATCH[2]}")
+ FUNC=$(printf "0x%x" "0x${BASH_REMATCH[3]}")
+ else
+ DOMAIN="0x0000"
+ BUS="0x00"
+ SLOT="0x00"
+ FUNC="0x0"
+ fi
+}
+
+# Normalize a PCI address to its canonical full domain-qualified form
+# ("dddd:bb:ss.f", lowercase, zero-padded). Both "dddd:bb:ss.f" (full) and
+# "bb:ss.f" (short, domain 0) inputs are accepted.
+normalize_pci_address() {
+ local addr="$1"
+ if [[ $addr =~ ^([0-9A-Fa-f]+):([0-9A-Fa-f]+):([0-9A-Fa-f]+)\.([0-9A-Fa-f]+)$ ]]; then
+ printf "%04x:%02x:%02x.%x\n" \
+ "0x${BASH_REMATCH[1]}" "0x${BASH_REMATCH[2]}" \
+ "0x${BASH_REMATCH[3]}" "0x${BASH_REMATCH[4]}"
+ elif [[ $addr =~ ^([0-9A-Fa-f]+):([0-9A-Fa-f]+)\.([0-9A-Fa-f]+)$ ]]; then
+ printf "0000:%02x:%02x.%x\n" \
+ "0x${BASH_REMATCH[1]}" "0x${BASH_REMATCH[2]}" "0x${BASH_REMATCH[3]}"
+ else
+ echo "$addr"
+ fi
+}
+
+# Get nodedev name for a PCI address (e.g. "0000:00:02.0" -> "pci_0000_00_02_0").
+# Short addresses are widened to domain 0 first.
+get_nodedev_name() {
+ local addr
+ addr=$(normalize_pci_address "$1")
+ echo "pci_$(echo "$addr" | sed 's/[:.]/\_/g')"
}
# Get cached nodedev XML for a PCI address
@@ -572,9 +655,12 @@ get_numa_node() {
echo "${node:--1}"
}
-# Given a PCI address, return its PCI root (the top‐level bridge ID, e.g. "0000:00:03")
+# Given a PCI address, return its PCI root (the top‐level bridge ID, e.g.
+# "0000:00:03.0"). Both full ("0000:00:02.0") and short ("00:02.0") inputs are
+# accepted; output is always domain-qualified.
get_pci_root() {
- local addr="$1"
+ local addr
+ addr=$(normalize_pci_address "$1")
local xml
xml=$(get_nodedev_xml "$addr")
@@ -583,21 +669,23 @@ get_pci_root() {
local parent
parent=$(echo "$xml" | xmlstarlet sel -t -v "/device/parent" 2>/dev/null || true)
if [[ -n "$parent" ]]; then
- # If parent is a PCI device, recursively find its root
- if [[ $parent =~ ^pci_0000_([0-9A-Fa-f]{2})_([0-9A-Fa-f]{2})_([0-9A-Fa-f])$ ]]; then
- local parent_addr="${BASH_REMATCH[1]}:${BASH_REMATCH[2]}.${BASH_REMATCH[3]}"
+ # If parent is a PCI device, recursively find its root.
+ # libvirt nodedev names look like pci____
+ # where is one or more hex digits (typically 4).
+ if [[ $parent =~ ^pci_([0-9A-Fa-f]+)_([0-9A-Fa-f]{2})_([0-9A-Fa-f]{2})_([0-9A-Fa-f])$ ]]; then
+ local parent_addr="${BASH_REMATCH[1]}:${BASH_REMATCH[2]}:${BASH_REMATCH[3]}.${BASH_REMATCH[4]}"
get_pci_root "$parent_addr"
return
else
- # Parent is not PCI device, so current device is the root
- echo "0000:$addr"
+ # Parent is not a PCI device, so current device is the root
+ echo "$addr"
return
fi
fi
fi
# fallback
- echo "0000:$addr"
+ echo "$addr"
}
# Build VM → hostdev maps:
@@ -613,18 +701,23 @@ for VM in "${VMS[@]}"; do
continue
fi
- # -- PCI hostdevs: use xmlstarlet to extract BDF for all PCI host devices --
- while read -r bus slot func; do
+ # -- PCI hostdevs: use xmlstarlet to extract the full domain:bus:slot.func
+ # for all PCI host devices. libvirt's element may omit the domain
+ # attribute, in which case we default to 0.
+ while IFS=: read -r dom bus slot func; do
[[ -n "$bus" && -n "$slot" && -n "$func" ]] || continue
- # Format to match lspci output (e.g., 01:00.0) by padding with zeros
+ [[ -n "$dom" ]] || dom="0"
+ # Format to match lspci -D output (e.g., 0000:01:00.0) by padding with zeros
+ dom_fmt=$(printf "%04x" "0x$dom")
bus_fmt=$(printf "%02x" "0x$bus")
slot_fmt=$(printf "%02x" "0x$slot")
func_fmt=$(printf "%x" "0x$func")
- BDF="$bus_fmt:$slot_fmt.$func_fmt"
+ BDF="${dom_fmt}:${bus_fmt}:${slot_fmt}.${func_fmt}"
pci_to_vm["$BDF"]="$VM"
done < <(echo "$xml" | xmlstarlet sel -T -t -m "//hostdev[@type='pci']/source/address" \
- -v "substring-after(@bus, '0x')" -o " " \
- -v "substring-after(@slot, '0x')" -o " " \
+ -v "substring-after(@domain, '0x')" -o ":" \
+ -v "substring-after(@bus, '0x')" -o ":" \
+ -v "substring-after(@slot, '0x')" -o ":" \
-v "substring-after(@function, '0x')" -n 2>/dev/null || true)
# -- MDEV hostdevs: use xmlstarlet to extract UUIDs --
@@ -677,7 +770,8 @@ parse_and_add_gpu_properties() {
# Appends JSON strings for each found mdev instance to the global 'vlist' array.
# Arguments:
# $1: mdev_base_path (e.g., /sys/bus/pci/devices/.../mdev_supported_types)
-# $2: bdf (e.g., 01:00.0)
+# $2: bdf — short "bb:ss.f" (e.g. 01:00.0) or full
+# "dddd:bb:ss.f" (e.g. 0001:65:00.0) for non-zero PCI domains
process_mdev_instances() {
local mdev_base_path="$1"
local bdf="$2"
@@ -705,10 +799,8 @@ process_mdev_instances() {
local MDEV_UUID
MDEV_UUID=$(basename "$UDIR")
- local DOMAIN="0x0000"
- local BUS="0x${bdf:0:2}"
- local SLOT="0x${bdf:3:2}"
- local FUNC="0x${bdf:6:1}"
+ local DOMAIN BUS SLOT FUNC
+ parse_pci_address "$bdf"
local raw
raw="${mdev_to_vm[$MDEV_UUID]:-}"
@@ -727,6 +819,10 @@ process_mdev_instances() {
# Parse nvidia-smi vgpu profiles once at the beginning
parse_nvidia_vgpu_profiles
+# `lspci -nnm` keeps the existing output format for devices in PCI domain 0
+# (short "bb:ss.f" form) and includes the domain prefix only for devices on
+# non-zero PCI segments (multi-IIO servers, some ARM systems). The helpers
+# below (parse_pci_address, normalize_pci_address) accept both formats.
mapfile -t LINES < <(lspci -nnm)
echo '{ "gpus": ['
@@ -743,13 +839,18 @@ for LINE in "${LINES[@]}"; do
continue
fi
+ # sysfs paths always require the full domain-qualified form. PCI_ADDR may
+ # be short ("00:02.0") for domain 0 or full ("0001:65:00.0") otherwise, so
+ # we derive a separate SYSFS_ADDR just for filesystem lookups.
+ SYSFS_ADDR=$(normalize_pci_address "$PCI_ADDR")
+
# If this is a VF, skip it. It will be processed under its PF.
- if [[ -e "/sys/bus/pci/devices/0000:$PCI_ADDR/physfn" ]]; then
+ if [[ -e "/sys/bus/pci/devices/$SYSFS_ADDR/physfn" ]]; then
continue
fi
- # Only process GPU classes (3D controller)
- if [[ ! "$PCI_CLASS" =~ (3D\ controller|Processing\ accelerators) ]]; then
+ # Only process PCI classes - 3D controller, Processing accelerators, Display controller
+ if [[ ! "$PCI_CLASS" =~ (3D\ controller|Processing\ accelerators|Display\ controller) ]]; then
continue
fi
@@ -761,7 +862,7 @@ for LINE in "${LINES[@]}"; do
DEVICE_ID=$(sed -E 's/.*\[([0-9A-Fa-f]{4})\]$/\1/' <<<"$DEVICE_FIELD")
# Kernel driver
- DRV_PATH="/sys/bus/pci/devices/0000:$PCI_ADDR/driver"
+ DRV_PATH="/sys/bus/pci/devices/$SYSFS_ADDR/driver"
if [[ -L $DRV_PATH ]]; then
DRIVER=$(basename "$(readlink "$DRV_PATH")")
else
@@ -781,7 +882,7 @@ for LINE in "${LINES[@]}"; do
read -r TOTALVFS NUMVFS < <(get_sriov_counts "$PCI_ADDR")
# Get Physical GPU properties from its own description file, if available
- PF_DESC_PATH="/sys/bus/pci/devices/0000:$PCI_ADDR/description"
+ PF_DESC_PATH="/sys/bus/pci/devices/$SYSFS_ADDR/description"
parse_and_add_gpu_properties "$PF_DESC_PATH"
# Save physical function's properties before they are overwritten by vGPU/VF processing
PF_MAX_INSTANCES=$MAX_INSTANCES
@@ -791,25 +892,27 @@ for LINE in "${LINES[@]}"; do
PF_MAX_RESOLUTION_Y=$MAX_RESOLUTION_Y
# === full_passthrough usage ===
- raw="${pci_to_vm[$PCI_ADDR]:-}"
+ # pci_to_vm is keyed by the full domain-qualified form, so normalize the
+ # lspci-style PCI_ADDR (which may be short for domain 0) before lookup.
+ raw="${pci_to_vm[$(normalize_pci_address "$PCI_ADDR")]:-}"
FULL_USED_JSON=$(to_json_vm "$raw")
# === vGPU (MDEV) instances ===
VGPU_ARRAY="[]"
declare -a vlist=()
# Process mdev on the Physical Function
- MDEV_BASE="/sys/bus/pci/devices/0000:$PCI_ADDR/mdev_supported_types"
+ MDEV_BASE="/sys/bus/pci/devices/$SYSFS_ADDR/mdev_supported_types"
process_mdev_instances "$MDEV_BASE" "$PCI_ADDR"
# === VF instances (SR-IOV / MIG) ===
VF_ARRAY="[]"
declare -a flist=()
if ((TOTALVFS > 0)); then
- for VF_LINK in /sys/bus/pci/devices/0000:"$PCI_ADDR"/virtfn*; do
+ for VF_LINK in /sys/bus/pci/devices/"$SYSFS_ADDR"/virtfn*; do
[[ -L $VF_LINK ]] || continue
VF_PATH=$(readlink -f "$VF_LINK")
- VF_ADDR=${VF_PATH##*/} # e.g. "0000:65:00.2"
- VF_BDF="${VF_ADDR:5}" # "65:00.2"
+ # Keep the full domain-qualified address (e.g. "0000:65:00.2")
+ VF_BDF=${VF_PATH##*/}
# For NVIDIA SR-IOV, check for vGPU (mdev) on the VF itself
if [[ "$VENDOR_ID" == "10de" ]]; then
@@ -817,10 +920,7 @@ for LINE in "${LINES[@]}"; do
process_mdev_instances "$VF_MDEV_BASE" "$VF_BDF"
fi
- DOMAIN="0x0000"
- BUS="0x${VF_BDF:0:2}"
- SLOT="0x${VF_BDF:3:2}"
- FUNC="0x${VF_BDF:6:1}"
+ parse_pci_address "$VF_BDF"
# Determine vf_profile using nvidia-smi information
VF_PROFILE=""
@@ -835,8 +935,10 @@ for LINE in "${LINES[@]}"; do
# For NVIDIA GPUs, check current vGPU type
current_vgpu_type=$(get_current_vgpu_type "$VF_PATH")
if [[ "$current_vgpu_type" != "0" ]]; then
- # Get profile info from nvidia-smi cache
- profile_info=$(get_nvidia_profile_info "$PCI_ADDR" "$current_vgpu_type")
+ # nvidia_vgpu_profiles is keyed by the canonical full
+ # "dddd:bb:ss.f" form, so widen PCI_ADDR (which may be
+ # short for domain 0) before the cache lookup.
+ profile_info=$(get_nvidia_profile_info "$(normalize_pci_address "$PCI_ADDR")" "$current_vgpu_type")
IFS='|' read -r VF_PROFILE_NAME VF_MAX_INSTANCES VF_VIDEO_RAM VF_MAX_HEADS VF_MAX_RESOLUTION_X VF_MAX_RESOLUTION_Y <<< "$profile_info"
VF_PROFILE="$VF_PROFILE_NAME"
fi
@@ -853,12 +955,17 @@ for LINE in "${LINES[@]}"; do
fi
VF_PROFILE_JSON=$(json_escape "$VF_PROFILE")
- # Determine which VM uses this VF_BDF
- raw="${pci_to_vm[$VF_BDF]:-}"
+ # Determine which VM uses this VF_BDF (normalize for map lookup)
+ raw="${pci_to_vm[$(normalize_pci_address "$VF_BDF")]:-}"
USED_JSON=$(to_json_vm "$raw")
+ # For backward-compat JSON output, strip the default "0000:" domain
+ # prefix so domain-0 VFs still print as "65:00.2" rather than the
+ # full "0000:65:00.2". Non-zero domains are preserved unchanged.
+ VF_DISPLAY_ADDR="${VF_BDF#0000:}"
+
flist+=(
- "{\"vf_pci_address\":\"$VF_BDF\",\"vf_profile\":$VF_PROFILE_JSON,\"max_instances\":$VF_MAX_INSTANCES,\"video_ram\":$VF_VIDEO_RAM,\"max_heads\":$VF_MAX_HEADS,\"max_resolution_x\":$VF_MAX_RESOLUTION_X,\"max_resolution_y\":$VF_MAX_RESOLUTION_Y,\"libvirt_address\":{\"domain\":\"$DOMAIN\",\"bus\":\"$BUS\",\"slot\":\"$SLOT\",\"function\":\"$FUNC\"},\"used_by_vm\":$USED_JSON}")
+ "{\"vf_pci_address\":\"$VF_DISPLAY_ADDR\",\"vf_profile\":$VF_PROFILE_JSON,\"max_instances\":$VF_MAX_INSTANCES,\"video_ram\":$VF_VIDEO_RAM,\"max_heads\":$VF_MAX_HEADS,\"max_resolution_x\":$VF_MAX_RESOLUTION_X,\"max_resolution_y\":$VF_MAX_RESOLUTION_Y,\"libvirt_address\":{\"domain\":\"$DOMAIN\",\"bus\":\"$BUS\",\"slot\":\"$SLOT\",\"function\":\"$FUNC\"},\"used_by_vm\":$USED_JSON}")
done
if [ ${#flist[@]} -gt 0 ]; then
VF_ARRAY="[$(
@@ -882,10 +989,9 @@ for LINE in "${LINES[@]}"; do
if [[ ${#vlist[@]} -eq 0 && ${#flist[@]} -eq 0 ]]; then
FP_ENABLED=1
fi
- DOMAIN="0x0000"
- BUS="0x${PCI_ADDR:0:2}"
- SLOT="0x${PCI_ADDR:3:2}"
- FUNC="0x${PCI_ADDR:6:1}"
+
+ # Sets global DOMAIN/BUS/SLOT/FUNC for JSON output below
+ parse_pci_address "$PCI_ADDR"
# Emit JSON
if $first_gpu; then