ukify: Switch to JSON HWID description format (#35208)

Fixes #35176

3 files changed, 58 insertions, 65 deletions

diff --git a/man/ukify.xml b/man/ukify.xml
index c634e0bdee..c3b8d920d1 100644
--- a/man/ukify.xml
+++ b/man/ukify.xml
@@ -71,6 +71,8 @@
       <varname>Cmdline=</varname>/<option>--cmdline=</option>,
       <varname>OSRelease=</varname>/<option>--os-release=</option>,
       <varname>DeviceTree=</varname>/<option>--devicetree=</option>,
+      <varname>DeviceTreeAuto=</varname>/<option>--devicetree-auto=</option>,
+      <varname>HWIDs=</varname>/<option>--hwids=</option>,
       <varname>Splash=</varname>/<option>--splash=</option>,
       <varname>PCRPKey=</varname>/<option>--pcrpkey=</option>,
       <varname>Uname=</varname>/<option>--uname=</option>,
@@ -375,6 +377,35 @@
         </varlistentry>
 
         <varlistentry>
+          <term><varname>DeviceTreeAuto=<replaceable>PATH</replaceable>...</varname></term>
+          <term><option>--devicetree-auto=<replaceable>PATH</replaceable></option></term>
+
+          <listitem><para>Zero or more automatically selectable DeviceTree files. In the configuration file, items are separated by
+          whitespace. Each DeviceTree will be in a separate <literal>.dtbauto</literal> section.</para>
+
+          <xi:include href="version-info.xml" xpointer="v257"/></listitem>
+        </varlistentry>
+
+        <varlistentry>
+          <term><varname>HWIDs=<replaceable>PATH</replaceable></varname></term>
+          <term><option>--hwids=<replaceable>PATH</replaceable></option></term>
+
+          <listitem><para>The hardware ID device table (the <literal>.hwids</literal> section). The argument is a
+          path to a directory with JSON HWID device description files. Each file needs to contain a single JSON object with a <literal>name</literal>, <literal>compatible</literal> and <literal>hwids</literal> keys. The <literal>name</literal> and <literal>compatible</literal> keys must have string values and the <literal>hwids</literal> key must have a list of strings as value, where the strings must be valid UUIDs that represent CHIDs/HWIDs.
+          Example:
+          <programlisting><xi:include href="ukify_hwid.json.example" parse="text" /></programlisting>
+          Here <literal>Example Laptop 16 Gen 7</literal> is the device <literal>name</literal> (as defined by the manufacturer),
+          <literal>example,laptop-16-g7</literal> is the <literal>compatible</literal> (as defined by the kernel) and <literal>hwids</literal>
+          is an array of CHIDs/HWIDs (extracted i.e. from <command>fwupdtool hwids</command> output).
+          If not specified, the section will not be present. It is recommended to specify this parameter if automatically
+          selectable DeviceTrees are to be used.
+
+          </para>
+
+          <xi:include href="version-info.xml" xpointer="v257"/></listitem>
+        </varlistentry>
+
+        <varlistentry>
           <term><varname>Uname=<replaceable>VERSION</replaceable></varname></term>
           <term><option>--uname=<replaceable>VERSION</replaceable></option></term>
 
diff --git a/man/ukify_hwid.json.example b/man/ukify_hwid.json.example
new file mode 100644
index 0000000000..83921b5d93
--- /dev/null
+++ b/man/ukify_hwid.json.example
@@ -0,0 +1,8 @@
+{
+    "name": "Example Laptop 16 Gen 7",
+    "compatible": "example,laptop-16-g7",
+    "hwids": [
+        "5dc05bf4-01f6-4089-b464-a08c47ea9295",
+        "3e3f8f3c-2003-46f2-811c-85554f7d5952"
+    ]
+}
diff --git a/src/ukify/ukify.py b/src/ukify/ukify.py
index 06967b02b9..a5719615cc 100755
--- a/src/ukify/ukify.py
+++ b/src/ukify/ukify.py
@@ -42,6 +42,7 @@ import subprocess
 import sys
 import tempfile
 import textwrap
+import uuid
 from collections.abc import Iterable, Iterator, Sequence
 from hashlib import sha256
 from pathlib import Path
@@ -1013,14 +1014,9 @@ def merge_sbat(input_pe: list[Path], input_text: list[str]) -> str:
     )
 
 
-# Keep in sync with EFI_GUID (src/boot/efi.h)
-# uint32_t Data1, uint16_t Data2, uint16_t Data3, uint8_t Data4[8]
-EFI_GUID = tuple[int, int, int, tuple[int, int, int, int, int, int, int, int]]
-EFI_GUID_STRUCT_SIZE = 4 + 2 + 2 + 1 * 8
-
 # Keep in sync with Device (DEVICE_TYPE_DEVICETREE) from src/boot/chid.h
 # uint32_t descriptor, EFI_GUID chid, uint32_t name_offset, uint32_t compatible_offset
-DEVICE_STRUCT_SIZE = 4 + EFI_GUID_STRUCT_SIZE + 4 + 4
+DEVICE_STRUCT_SIZE = 4 + 16 + 4 + 4
 NULL_DEVICE = b'\0' * DEVICE_STRUCT_SIZE
 DEVICE_TYPE_DEVICETREE = 1
 
@@ -1029,29 +1025,21 @@ def device_make_descriptor(device_type: int, size: int) -> int:
     return (size) | (device_type << 28)
 
 
-def pack_device(offsets: dict[str, int], name: str, compatible: str, chids: list[EFI_GUID]) -> bytes:
+DEVICETREE_DESCRIPTOR = device_make_descriptor(DEVICE_TYPE_DEVICETREE, DEVICE_STRUCT_SIZE)
+
+
+def pack_device(offsets: dict[str, int], name: str, compatible: str, chids: set[uuid.UUID]) -> bytes:
     data = b''
 
-    for data1, data2, data3, data4 in chids:
-        data += struct.pack(
-            '<IIHH8BII',
-            device_make_descriptor(DEVICE_TYPE_DEVICETREE, DEVICE_STRUCT_SIZE),
-            data1,
-            data2,
-            data3,
-            *data4,
-            offsets[name],
-            offsets[compatible],
-        )
+    for chid in sorted(chids):
+        data += struct.pack('<I', DEVICETREE_DESCRIPTOR)
+        data += chid.bytes_le
+        data += struct.pack('<II', offsets[name], offsets[compatible])
 
     assert len(data) == DEVICE_STRUCT_SIZE * len(chids)
     return data
 
 
-def hex_pairs_list(string: str) -> list[int]:
-    return [int(string[i : i + 2], 16) for i in range(0, len(string), 2)]
-
-
 def pack_strings(strings: set[str], base: int) -> tuple[bytes, dict[str, int]]:
     blob = b''
     offsets = {}
@@ -1064,56 +1052,22 @@ def pack_strings(strings: set[str], base: int) -> tuple[bytes, dict[str, int]]:
 
 
 def parse_hwid_dir(path: Path) -> bytes:
-    hwid_files = path.rglob('*.txt')
+    hwid_files = path.rglob('*.json')
 
     strings: set[str] = set()
-    devices: collections.defaultdict[tuple[str, str], list[EFI_GUID]] = collections.defaultdict(list)
-
-    uuid_regexp = re.compile(
-        r'\{[0-9a-f]{8}-[0-9a-f]{4}-[0-5][0-9a-f]{3}-[089ab][0-9a-f]{3}-[0-9a-f]{12}\}', re.I
-    )
+    devices: collections.defaultdict[tuple[str, str], set[uuid.UUID]] = collections.defaultdict(set)
 
     for hwid_file in hwid_files:
-        content = hwid_file.open().readlines()
-
-        data: dict[str, str] = {
-            'Manufacturer': '',
-            'Family': '',
-            'Compatible': '',
-        }
-        uuids: list[EFI_GUID] = []
+        data = json.loads(hwid_file.read_text(encoding='UTF-8'))
 
-        for line in content:
-            for k in data:
-                if line.startswith(k):
-                    data[k] = line.split(':')[1].strip()
-                    break
-            else:
-                uuid = uuid_regexp.match(line)
-                if uuid is not None:
-                    d1, d2, d3, d4, d5 = uuid.group(0)[1:-1].split('-')
-
-                    data1 = int(d1, 16)
-                    data2 = int(d2, 16)
-                    data3 = int(d3, 16)
-                    data4 = cast(
-                        tuple[int, int, int, int, int, int, int, int],
-                        tuple(hex_pairs_list(d4) + hex_pairs_list(d5)),
-                    )
-
-                    uuids.append((data1, data2, data3, data4))
-
-        for k, v in data.items():
-            if not v:
+        for k in ['name', 'compatible', 'hwids']:
+            if k not in data:
                 raise ValueError(f'hwid description file "{hwid_file}" does not contain "{k}"')
 
-        name = data['Manufacturer'] + ' ' + data['Family']
-        compatible = data['Compatible']
-
-        strings |= set([name, compatible])
+        strings |= {data['name'], data['compatible']}
 
-        # (compatible, name) pair uniquely identifies the device
-        devices[(compatible, name)] += uuids
+        # (name, compatible) pair uniquely identifies the device
+        devices[(data['name'], data['compatible'])] |= {uuid.UUID(u) for u in data['hwids']}
 
     total_device_structs = 1
     for dev, uuids in devices.items():
@@ -1122,7 +1076,7 @@ def parse_hwid_dir(path: Path) -> bytes:
     strings_blob, offsets = pack_strings(strings, total_device_structs * DEVICE_STRUCT_SIZE)
 
     devices_blob = b''
-    for (compatible, name), uuids in devices.items():
+    for (name, compatible), uuids in devices.items():
         devices_blob += pack_device(offsets, name, compatible, uuids)
 
     devices_blob += NULL_DEVICE