dt: Add a check for undocumented compatible strings in kernel

Add a make target, dt_compatible_check, to extract compatible strings
from kernel sources and check if they are documented by a schema.
At least version v2022.08 of dtschema with dt-check-compatible is
required.

This check can also be run manually on specific files or directories:

scripts/dtc/dt-extract-compatibles drivers/clk/ | \
  xargs dt-check-compatible -v -s Documentation/devicetree/bindings/processed-schema.json

Currently, there are about 3800 undocumented compatible strings. Most of
these are cases where the binding is not yet converted (given there
are 1900 .txt binding files remaining).

Link: https://lore.kernel.org/all/20220916012510.2718170-1-robh@kernel.org/
Signed-off-by: Rob Herring <robh@kernel.org>

Documentation/devicetree/bindings/Makefile

@@ -75,3 +75,6 @@ always-$(CHECK_DT_BINDING) += $(patsubst $(srctree)/$(src)/%.yaml,%.example.dtb,
 # build artifacts here before they are processed by scripts/Makefile.clean
 clean-files = $(shell find $(obj) \( -name '*.example.dts' -o \
 			-name '*.example.dtb' \) -delete 2>/dev/null)
+
+dt_compatible_check: $(obj)/processed-schema.json
+	$(Q)$(srctree)/scripts/dtc/dt-extract-compatibles $(srctree) | xargs dt-check-compatible -v -s $<

Makefile

@@ -1419,6 +1419,10 @@ PHONY += dt_binding_check
 dt_binding_check: scripts_dtc
 	$(Q)$(MAKE) $(build)=Documentation/devicetree/bindings
 
+PHONY += dt_compatible_check
+dt_compatible_check: dt_binding_check
+	$(Q)$(MAKE) $(build)=Documentation/devicetree/bindings $@
+
 # ---------------------------------------------------------------------------
 # Modules
 

scripts/dtc/dt-extract-compatibles

@@ -0,0 +1,69 @@
+#!/usr/bin/env python3
+# SPDX-License-Identifier: GPL-2.0-only
+
+import os
+import glob
+import re
+import argparse
+
+
+def parse_of_declare_macros(data):
+	""" Find all compatible strings in OF_DECLARE() style macros """
+	compat_list = []
+	# CPU_METHOD_OF_DECLARE does not have a compatible string
+	for m in re.finditer(r'(?<!CPU_METHOD_)(IRQCHIP|OF)_(DECLARE|MATCH)(_DRIVER)?\(.*?\)', data):
+		try:
+			compat = re.search(r'"(.*?)"', m[0])[1]
+		except:
+			# Fails on compatible strings in #define, so just skip
+			continue
+		compat_list += [compat]
+
+	return compat_list
+
+
+def parse_of_device_id(data):
+	""" Find all compatible strings in of_device_id structs """
+	compat_list = []
+	for m in re.finditer(r'of_device_id\s+[a-zA-Z0-9_]+\[\]\s*=\s*({.*?);', data):
+		compat_list += re.findall(r'\.compatible\s+=\s+"([a-zA-Z0-9_\-,]+)"', m[1])
+
+	return compat_list
+
+
+def parse_compatibles(file):
+	with open(file, 'r', encoding='utf-8') as f:
+		data = f.read().replace('\n', '')
+
+	compat_list = parse_of_declare_macros(data)
+	compat_list += parse_of_device_id(data)
+
+	return compat_list
+
+def print_compat(filename, compatibles):
+	if not compatibles:
+		return
+	if show_filename:
+		compat_str = ' '.join(compatibles)
+		print(filename + ": compatible(s): " + compat_str)
+	else:
+		print(*compatibles, sep='\n')
+
+show_filename = False
+
+if __name__ == "__main__":
+	ap = argparse.ArgumentParser()
+	ap.add_argument("cfile", type=str, nargs='*', help="C source files or directories to parse")
+	ap.add_argument('-H', '--with-filename', help="Print filename with compatibles", action="store_true")
+	args = ap.parse_args()
+
+	show_filename = args.with_filename
+
+	for f in args.cfile:
+		if os.path.isdir(f):
+			for filename in glob.iglob(f + "/**/*.c", recursive=True):
+				compat_list = parse_compatibles(filename)
+				print_compat(filename, compat_list)
+		else:
+			compat_list = parse_compatibles(f)
+			print_compat(f, compat_list)