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 = machine__kernel_maps(&vmlinux);
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 	/*
136 	 * Step 2:
137 	 *
138 	 * Create the kernel maps for kallsyms and the DSO where we will then
139 	 * load /proc/kallsyms. Also create the modules maps from /proc/modules
140 	 * and find the .ko files that match them in /lib/modules/`uname -r`/.
141 	 */
142 	if (machine__create_kernel_maps(&kallsyms) < 0) {
143 		pr_debug("machine__create_kernel_maps ");
144 		goto out;
145 	}
146 
147 	/*
148 	 * Step 3:
149 	 *
150 	 * Load and split /proc/kallsyms into multiple maps, one per module.
151 	 * Do not use kcore, as this test was designed before kcore support
152 	 * and has parts that only make sense if using the non-kcore code.
153 	 * XXX: extend it to stress the kcorre code as well, hint: the list
154 	 * of modules extracted from /proc/kcore, in its current form, can't
155 	 * be compacted against the list of modules found in the "vmlinux"
156 	 * code and with the one got from /proc/modules from the "kallsyms" code.
157 	 */
158 	if (machine__load_kallsyms(&kallsyms, "/proc/kallsyms") <= 0) {
159 		pr_debug("dso__load_kallsyms ");
160 		goto out;
161 	}
162 
163 	/*
164 	 * Step 4:
165 	 *
166 	 * kallsyms will be internally on demand sorted by name so that we can
167 	 * find the reference relocation * symbol, i.e. the symbol we will use
168 	 * to see if the running kernel was relocated by checking if it has the
169 	 * same value in the vmlinux file we load.
170 	 */
171 	kallsyms_map = machine__kernel_map(&kallsyms);
172 
173 	/*
174 	 * Step 5:
175 	 *
176 	 * Now repeat step 2, this time for the vmlinux file we'll auto-locate.
177 	 */
178 	if (machine__create_kernel_maps(&vmlinux) < 0) {
179 		pr_debug("machine__create_kernel_maps ");
180 		goto out;
181 	}
182 
183 	vmlinux_map = machine__kernel_map(&vmlinux);
184 
185 	/*
186 	 * Step 6:
187 	 *
188 	 * Locate a vmlinux file in the vmlinux path that has a buildid that
189 	 * matches the one of the running kernel.
190 	 *
191 	 * While doing that look if we find the ref reloc symbol, if we find it
192 	 * we'll have its ref_reloc_symbol.unrelocated_addr and then
193 	 * maps__reloc_vmlinux will notice and set proper ->[un]map_ip routines
194 	 * to fixup the symbols.
195 	 */
196 	if (machine__load_vmlinux_path(&vmlinux) <= 0) {
197 		pr_debug("Couldn't find a vmlinux that matches the kernel running on this machine, skipping test\n");
198 		err = TEST_SKIP;
199 		goto out;
200 	}
201 
202 	err = 0;
203 	/*
204 	 * Step 7:
205 	 *
206 	 * Now look at the symbols in the vmlinux DSO and check if we find all of them
207 	 * in the kallsyms dso. For the ones that are in both, check its names and
208 	 * end addresses too.
209 	 */
210 	map__for_each_symbol(vmlinux_map, sym, nd) {
211 		struct symbol *pair, *first_pair;
212 
213 		sym  = rb_entry(nd, struct symbol, rb_node);
214 
215 		if (sym->start == sym->end)
216 			continue;
217 
218 		mem_start = vmlinux_map->unmap_ip(vmlinux_map, sym->start);
219 		mem_end = vmlinux_map->unmap_ip(vmlinux_map, sym->end);
220 
221 		first_pair = machine__find_kernel_symbol(&kallsyms, mem_start, NULL);
222 		pair = first_pair;
223 
224 		if (pair && UM(pair->start) == mem_start) {
225 next_pair:
226 			if (arch__compare_symbol_names(sym->name, pair->name) == 0) {
227 				/*
228 				 * kallsyms don't have the symbol end, so we
229 				 * set that by using the next symbol start - 1,
230 				 * in some cases we get this up to a page
231 				 * wrong, trace_kmalloc when I was developing
232 				 * this code was one such example, 2106 bytes
233 				 * off the real size. More than that and we
234 				 * _really_ have a problem.
235 				 */
236 				s64 skew = mem_end - UM(pair->end);
237 				if (llabs(skew) >= page_size)
238 					pr_debug("WARN: %#" PRIx64 ": diff end addr for %s v: %#" PRIx64 " k: %#" PRIx64 "\n",
239 						 mem_start, sym->name, mem_end,
240 						 UM(pair->end));
241 
242 				/*
243 				 * Do not count this as a failure, because we
244 				 * could really find a case where it's not
245 				 * possible to get proper function end from
246 				 * kallsyms.
247 				 */
248 				continue;
249 			} else {
250 				pair = machine__find_kernel_symbol_by_name(&kallsyms, sym->name, NULL);
251 				if (pair) {
252 					if (UM(pair->start) == mem_start)
253 						goto next_pair;
254 
255 					pr_debug("WARN: %#" PRIx64 ": diff name v: %s k: %s\n",
256 						 mem_start, sym->name, pair->name);
257 				} else {
258 					pr_debug("WARN: %#" PRIx64 ": diff name v: %s k: %s\n",
259 						 mem_start, sym->name, first_pair->name);
260 				}
261 
262 				continue;
263 			}
264 		} else if (mem_start == kallsyms.vmlinux_map->end) {
265 			/*
266 			 * Ignore aliases to _etext, i.e. to the end of the kernel text area,
267 			 * such as __indirect_thunk_end.
268 			 */
269 			continue;
270 		} else if (is_ignored_symbol(sym->name, sym->type)) {
271 			/*
272 			 * Ignore hidden symbols, see scripts/kallsyms.c for the details
273 			 */
274 			continue;
275 		} else {
276 			pr_debug("ERR : %#" PRIx64 ": %s not on kallsyms\n",
277 				 mem_start, sym->name);
278 		}
279 
280 		err = -1;
281 	}
282 
283 	if (verbose <= 0)
284 		goto out;
285 
286 	header_printed = false;
287 
288 	maps__for_each_entry(maps, map) {
289 		struct map *
290 		/*
291 		 * If it is the kernel, kallsyms is always "[kernel.kallsyms]", while
292 		 * the kernel will have the path for the vmlinux file being used,
293 		 * so use the short name, less descriptive but the same ("[kernel]" in
294 		 * both cases.
295 		 */
296 		pair = maps__find_by_name(&kallsyms.kmaps, (map->dso->kernel ?
297 								map->dso->short_name :
298 								map->dso->name));
299 		if (pair) {
300 			pair->priv = 1;
301 		} else {
302 			if (!header_printed) {
303 				pr_info("WARN: Maps only in vmlinux:\n");
304 				header_printed = true;
305 			}
306 			map__fprintf(map, stderr);
307 		}
308 	}
309 
310 	header_printed = false;
311 
312 	maps__for_each_entry(maps, map) {
313 		struct map *pair;
314 
315 		mem_start = vmlinux_map->unmap_ip(vmlinux_map, map->start);
316 		mem_end = vmlinux_map->unmap_ip(vmlinux_map, map->end);
317 
318 		pair = maps__find(&kallsyms.kmaps, mem_start);
319 		if (pair == NULL || pair->priv)
320 			continue;
321 
322 		if (pair->start == mem_start) {
323 			if (!header_printed) {
324 				pr_info("WARN: Maps in vmlinux with a different name in kallsyms:\n");
325 				header_printed = true;
326 			}
327 
328 			pr_info("WARN: %" PRIx64 "-%" PRIx64 " %" PRIx64 " %s in kallsyms as",
329 				map->start, map->end, map->pgoff, map->dso->name);
330 			if (mem_end != pair->end)
331 				pr_info(":\nWARN: *%" PRIx64 "-%" PRIx64 " %" PRIx64,
332 					pair->start, pair->end, pair->pgoff);
333 			pr_info(" %s\n", pair->dso->name);
334 			pair->priv = 1;
335 		}
336 	}
337 
338 	header_printed = false;
339 
340 	maps = machine__kernel_maps(&kallsyms);
341 
342 	maps__for_each_entry(maps, map) {
343 		if (!map->priv) {
344 			if (!header_printed) {
345 				pr_info("WARN: Maps only in kallsyms:\n");
346 				header_printed = true;
347 			}
348 			map__fprintf(map, stderr);
349 		}
350 	}
351 out:
352 	machine__exit(&kallsyms);
353 	machine__exit(&vmlinux);
354 	return err;
355 }
356 
357 DEFINE_SUITE("vmlinux symtab matches kallsyms", vmlinux_matches_kallsyms);
358