1 // SPDX-License-Identifier: GPL-2.0
2 #include <linux/compiler.h>
3 #include <linux/rbtree.h>
4 #include <inttypes.h>
5 #include <string.h>
6 #include <ctype.h>
7 #include <stdlib.h>
8 #include "dso.h"
9 #include "map.h"
10 #include "symbol.h"
11 #include <internal/lib.h> // page_size
12 #include "tests.h"
13 #include "debug.h"
14 #include "machine.h"
15 
16 #define UM(x) kallsyms_map->unmap_ip(kallsyms_map, (x))
17 
18 static bool is_ignored_symbol(const char *name, char type)
19 {
20 	/* Symbol names that exactly match to the following are ignored.*/
21 	static const char * const ignored_symbols[] = {
22 		/*
23 		 * Symbols which vary between passes. Passes 1 and 2 must have
24 		 * identical symbol lists. The kallsyms_* symbols below are
25 		 * only added after pass 1, they would be included in pass 2
26 		 * when --all-symbols is specified so exclude them to get a
27 		 * stable symbol list.
28 		 */
29 		"kallsyms_addresses",
30 		"kallsyms_offsets",
31 		"kallsyms_relative_base",
32 		"kallsyms_num_syms",
33 		"kallsyms_names",
34 		"kallsyms_markers",
35 		"kallsyms_token_table",
36 		"kallsyms_token_index",
37 		/* Exclude linker generated symbols which vary between passes */
38 		"_SDA_BASE_",		/* ppc */
39 		"_SDA2_BASE_",		/* ppc */
40 		NULL
41 	};
42 
43 	/* Symbol names that begin with the following are ignored.*/
44 	static const char * const ignored_prefixes[] = {
45 		"$",			/* local symbols for ARM, MIPS, etc. */
46 		".LASANPC",		/* s390 kasan local symbols */
47 		"__crc_",		/* modversions */
48 		"__efistub_",		/* arm64 EFI stub namespace */
49 		"__kvm_nvhe_",		/* arm64 non-VHE KVM namespace */
50 		"__AArch64ADRPThunk_",	/* arm64 lld */
51 		"__ARMV5PILongThunk_",	/* arm lld */
52 		"__ARMV7PILongThunk_",
53 		"__ThumbV7PILongThunk_",
54 		"__LA25Thunk_",		/* mips lld */
55 		"__microLA25Thunk_",
56 		NULL
57 	};
58 
59 	/* Symbol names that end with the following are ignored.*/
60 	static const char * const ignored_suffixes[] = {
61 		"_from_arm",		/* arm */
62 		"_from_thumb",		/* arm */
63 		"_veneer",		/* arm */
64 		NULL
65 	};
66 
67 	/* Symbol names that contain the following are ignored.*/
68 	static const char * const ignored_matches[] = {
69 		".long_branch.",	/* ppc stub */
70 		".plt_branch.",		/* ppc stub */
71 		NULL
72 	};
73 
74 	const char * const *p;
75 
76 	for (p = ignored_symbols; *p; p++)
77 		if (!strcmp(name, *p))
78 			return true;
79 
80 	for (p = ignored_prefixes; *p; p++)
81 		if (!strncmp(name, *p, strlen(*p)))
82 			return true;
83 
84 	for (p = ignored_suffixes; *p; p++) {
85 		int l = strlen(name) - strlen(*p);
86 
87 		if (l >= 0 && !strcmp(name + l, *p))
88 			return true;
89 	}
90 
91 	for (p = ignored_matches; *p; p++) {
92 		if (strstr(name, *p))
93 			return true;
94 	}
95 
96 	if (type == 'U' || type == 'u')
97 		return true;
98 	/* exclude debugging symbols */
99 	if (type == 'N' || type == 'n')
100 		return true;
101 
102 	if (toupper(type) == 'A') {
103 		/* Keep these useful absolute symbols */
104 		if (strcmp(name, "__kernel_syscall_via_break") &&
105 		    strcmp(name, "__kernel_syscall_via_epc") &&
106 		    strcmp(name, "__kernel_sigtramp") &&
107 		    strcmp(name, "__gp"))
108 			return true;
109 	}
110 
111 	return false;
112 }
113 
114 static int test__vmlinux_matches_kallsyms(struct test_suite *test __maybe_unused,
115 					int subtest __maybe_unused)
116 {
117 	int err = -1;
118 	struct rb_node *nd;
119 	struct symbol *sym;
120 	struct map *kallsyms_map, *vmlinux_map, *map;
121 	struct machine kallsyms, vmlinux;
122 	struct maps *maps;
123 	u64 mem_start, mem_end;
124 	bool header_printed;
125 
126 	/*
127 	 * Step 1:
128 	 *
129 	 * Init the machines that will hold kernel, modules obtained from
130 	 * both vmlinux + .ko files and from /proc/kallsyms split by modules.
131 	 */
132 	machine__init(&kallsyms, "", HOST_KERNEL_ID);
133 	machine__init(&vmlinux, "", HOST_KERNEL_ID);
134 
135 	maps = machine__kernel_maps(&vmlinux);
136 
137 	/*
138 	 * Step 2:
139 	 *
140 	 * Create the kernel maps for kallsyms and the DSO where we will then
141 	 * load /proc/kallsyms. Also create the modules maps from /proc/modules
142 	 * and find the .ko files that match them in /lib/modules/`uname -r`/.
143 	 */
144 	if (machine__create_kernel_maps(&kallsyms) < 0) {
145 		pr_debug("machine__create_kernel_maps ");
146 		goto out;
147 	}
148 
149 	/*
150 	 * Step 3:
151 	 *
152 	 * Load and split /proc/kallsyms into multiple maps, one per module.
153 	 * Do not use kcore, as this test was designed before kcore support
154 	 * and has parts that only make sense if using the non-kcore code.
155 	 * XXX: extend it to stress the kcorre code as well, hint: the list
156 	 * of modules extracted from /proc/kcore, in its current form, can't
157 	 * be compacted against the list of modules found in the "vmlinux"
158 	 * code and with the one got from /proc/modules from the "kallsyms" code.
159 	 */
160 	if (machine__load_kallsyms(&kallsyms, "/proc/kallsyms") <= 0) {
161 		pr_debug("dso__load_kallsyms ");
162 		goto out;
163 	}
164 
165 	/*
166 	 * Step 4:
167 	 *
168 	 * kallsyms will be internally on demand sorted by name so that we can
169 	 * find the reference relocation * symbol, i.e. the symbol we will use
170 	 * to see if the running kernel was relocated by checking if it has the
171 	 * same value in the vmlinux file we load.
172 	 */
173 	kallsyms_map = machine__kernel_map(&kallsyms);
174 
175 	/*
176 	 * Step 5:
177 	 *
178 	 * Now repeat step 2, this time for the vmlinux file we'll auto-locate.
179 	 */
180 	if (machine__create_kernel_maps(&vmlinux) < 0) {
181 		pr_debug("machine__create_kernel_maps ");
182 		goto out;
183 	}
184 
185 	vmlinux_map = machine__kernel_map(&vmlinux);
186 
187 	/*
188 	 * Step 6:
189 	 *
190 	 * Locate a vmlinux file in the vmlinux path that has a buildid that
191 	 * matches the one of the running kernel.
192 	 *
193 	 * While doing that look if we find the ref reloc symbol, if we find it
194 	 * we'll have its ref_reloc_symbol.unrelocated_addr and then
195 	 * maps__reloc_vmlinux will notice and set proper ->[un]map_ip routines
196 	 * to fixup the symbols.
197 	 */
198 	if (machine__load_vmlinux_path(&vmlinux) <= 0) {
199 		pr_debug("Couldn't find a vmlinux that matches the kernel running on this machine, skipping test\n");
200 		err = TEST_SKIP;
201 		goto out;
202 	}
203 
204 	err = 0;
205 	/*
206 	 * Step 7:
207 	 *
208 	 * Now look at the symbols in the vmlinux DSO and check if we find all of them
209 	 * in the kallsyms dso. For the ones that are in both, check its names and
210 	 * end addresses too.
211 	 */
212 	map__for_each_symbol(vmlinux_map, sym, nd) {
213 		struct symbol *pair, *first_pair;
214 
215 		sym  = rb_entry(nd, struct symbol, rb_node);
216 
217 		if (sym->start == sym->end)
218 			continue;
219 
220 		mem_start = vmlinux_map->unmap_ip(vmlinux_map, sym->start);
221 		mem_end = vmlinux_map->unmap_ip(vmlinux_map, sym->end);
222 
223 		first_pair = machine__find_kernel_symbol(&kallsyms, mem_start, NULL);
224 		pair = first_pair;
225 
226 		if (pair && UM(pair->start) == mem_start) {
227 next_pair:
228 			if (arch__compare_symbol_names(sym->name, pair->name) == 0) {
229 				/*
230 				 * kallsyms don't have the symbol end, so we
231 				 * set that by using the next symbol start - 1,
232 				 * in some cases we get this up to a page
233 				 * wrong, trace_kmalloc when I was developing
234 				 * this code was one such example, 2106 bytes
235 				 * off the real size. More than that and we
236 				 * _really_ have a problem.
237 				 */
238 				s64 skew = mem_end - UM(pair->end);
239 				if (llabs(skew) >= page_size)
240 					pr_debug("WARN: %#" PRIx64 ": diff end addr for %s v: %#" PRIx64 " k: %#" PRIx64 "\n",
241 						 mem_start, sym->name, mem_end,
242 						 UM(pair->end));
243 
244 				/*
245 				 * Do not count this as a failure, because we
246 				 * could really find a case where it's not
247 				 * possible to get proper function end from
248 				 * kallsyms.
249 				 */
250 				continue;
251 			} else {
252 				pair = machine__find_kernel_symbol_by_name(&kallsyms, sym->name, NULL);
253 				if (pair) {
254 					if (UM(pair->start) == mem_start)
255 						goto next_pair;
256 
257 					pr_debug("WARN: %#" PRIx64 ": diff name v: %s k: %s\n",
258 						 mem_start, sym->name, pair->name);
259 				} else {
260 					pr_debug("WARN: %#" PRIx64 ": diff name v: %s k: %s\n",
261 						 mem_start, sym->name, first_pair->name);
262 				}
263 
264 				continue;
265 			}
266 		} else if (mem_start == kallsyms.vmlinux_map->end) {
267 			/*
268 			 * Ignore aliases to _etext, i.e. to the end of the kernel text area,
269 			 * such as __indirect_thunk_end.
270 			 */
271 			continue;
272 		} else if (is_ignored_symbol(sym->name, sym->type)) {
273 			/*
274 			 * Ignore hidden symbols, see scripts/kallsyms.c for the details
275 			 */
276 			continue;
277 		} else {
278 			pr_debug("ERR : %#" PRIx64 ": %s not on kallsyms\n",
279 				 mem_start, sym->name);
280 		}
281 
282 		err = -1;
283 	}
284 
285 	if (verbose <= 0)
286 		goto out;
287 
288 	header_printed = false;
289 
290 	maps__for_each_entry(maps, map) {
291 		struct map *
292 		/*
293 		 * If it is the kernel, kallsyms is always "[kernel.kallsyms]", while
294 		 * the kernel will have the path for the vmlinux file being used,
295 		 * so use the short name, less descriptive but the same ("[kernel]" in
296 		 * both cases.
297 		 */
298 		pair = maps__find_by_name(kallsyms.kmaps, (map->dso->kernel ?
299 								map->dso->short_name :
300 								map->dso->name));
301 		if (pair) {
302 			pair->priv = 1;
303 		} else {
304 			if (!header_printed) {
305 				pr_info("WARN: Maps only in vmlinux:\n");
306 				header_printed = true;
307 			}
308 			map__fprintf(map, stderr);
309 		}
310 	}
311 
312 	header_printed = false;
313 
314 	maps__for_each_entry(maps, map) {
315 		struct map *pair;
316 
317 		mem_start = vmlinux_map->unmap_ip(vmlinux_map, map->start);
318 		mem_end = vmlinux_map->unmap_ip(vmlinux_map, map->end);
319 
320 		pair = maps__find(kallsyms.kmaps, mem_start);
321 		if (pair == NULL || pair->priv)
322 			continue;
323 
324 		if (pair->start == mem_start) {
325 			if (!header_printed) {
326 				pr_info("WARN: Maps in vmlinux with a different name in kallsyms:\n");
327 				header_printed = true;
328 			}
329 
330 			pr_info("WARN: %" PRIx64 "-%" PRIx64 " %" PRIx64 " %s in kallsyms as",
331 				map->start, map->end, map->pgoff, map->dso->name);
332 			if (mem_end != pair->end)
333 				pr_info(":\nWARN: *%" PRIx64 "-%" PRIx64 " %" PRIx64,
334 					pair->start, pair->end, pair->pgoff);
335 			pr_info(" %s\n", pair->dso->name);
336 			pair->priv = 1;
337 		}
338 	}
339 
340 	header_printed = false;
341 
342 	maps = machine__kernel_maps(&kallsyms);
343 
344 	maps__for_each_entry(maps, map) {
345 		if (!map->priv) {
346 			if (!header_printed) {
347 				pr_info("WARN: Maps only in kallsyms:\n");
348 				header_printed = true;
349 			}
350 			map__fprintf(map, stderr);
351 		}
352 	}
353 out:
354 	machine__exit(&kallsyms);
355 	machine__exit(&vmlinux);
356 	return err;
357 }
358 
359 DEFINE_SUITE("vmlinux symtab matches kallsyms", vmlinux_matches_kallsyms);
360