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 int test__vmlinux_matches_kallsyms(struct test *test __maybe_unused, int subtest __maybe_unused)
115 {
116 	int err = -1;
117 	struct rb_node *nd;
118 	struct symbol *sym;
119 	struct map *kallsyms_map, *vmlinux_map, *map;
120 	struct machine kallsyms, vmlinux;
121 	struct maps *maps = machine__kernel_maps(&vmlinux);
122 	u64 mem_start, mem_end;
123 	bool header_printed;
124 
125 	/*
126 	 * Step 1:
127 	 *
128 	 * Init the machines that will hold kernel, modules obtained from
129 	 * both vmlinux + .ko files and from /proc/kallsyms split by modules.
130 	 */
131 	machine__init(&kallsyms, "", HOST_KERNEL_ID);
132 	machine__init(&vmlinux, "", HOST_KERNEL_ID);
133 
134 	/*
135 	 * Step 2:
136 	 *
137 	 * Create the kernel maps for kallsyms and the DSO where we will then
138 	 * load /proc/kallsyms. Also create the modules maps from /proc/modules
139 	 * and find the .ko files that match them in /lib/modules/`uname -r`/.
140 	 */
141 	if (machine__create_kernel_maps(&kallsyms) < 0) {
142 		pr_debug("machine__create_kernel_maps ");
143 		goto out;
144 	}
145 
146 	/*
147 	 * Step 3:
148 	 *
149 	 * Load and split /proc/kallsyms into multiple maps, one per module.
150 	 * Do not use kcore, as this test was designed before kcore support
151 	 * and has parts that only make sense if using the non-kcore code.
152 	 * XXX: extend it to stress the kcorre code as well, hint: the list
153 	 * of modules extracted from /proc/kcore, in its current form, can't
154 	 * be compacted against the list of modules found in the "vmlinux"
155 	 * code and with the one got from /proc/modules from the "kallsyms" code.
156 	 */
157 	if (machine__load_kallsyms(&kallsyms, "/proc/kallsyms") <= 0) {
158 		pr_debug("dso__load_kallsyms ");
159 		goto out;
160 	}
161 
162 	/*
163 	 * Step 4:
164 	 *
165 	 * kallsyms will be internally on demand sorted by name so that we can
166 	 * find the reference relocation * symbol, i.e. the symbol we will use
167 	 * to see if the running kernel was relocated by checking if it has the
168 	 * same value in the vmlinux file we load.
169 	 */
170 	kallsyms_map = machine__kernel_map(&kallsyms);
171 
172 	/*
173 	 * Step 5:
174 	 *
175 	 * Now repeat step 2, this time for the vmlinux file we'll auto-locate.
176 	 */
177 	if (machine__create_kernel_maps(&vmlinux) < 0) {
178 		pr_debug("machine__create_kernel_maps ");
179 		goto out;
180 	}
181 
182 	vmlinux_map = machine__kernel_map(&vmlinux);
183 
184 	/*
185 	 * Step 6:
186 	 *
187 	 * Locate a vmlinux file in the vmlinux path that has a buildid that
188 	 * matches the one of the running kernel.
189 	 *
190 	 * While doing that look if we find the ref reloc symbol, if we find it
191 	 * we'll have its ref_reloc_symbol.unrelocated_addr and then
192 	 * maps__reloc_vmlinux will notice and set proper ->[un]map_ip routines
193 	 * to fixup the symbols.
194 	 */
195 	if (machine__load_vmlinux_path(&vmlinux) <= 0) {
196 		pr_debug("Couldn't find a vmlinux that matches the kernel running on this machine, skipping test\n");
197 		err = TEST_SKIP;
198 		goto out;
199 	}
200 
201 	err = 0;
202 	/*
203 	 * Step 7:
204 	 *
205 	 * Now look at the symbols in the vmlinux DSO and check if we find all of them
206 	 * in the kallsyms dso. For the ones that are in both, check its names and
207 	 * end addresses too.
208 	 */
209 	map__for_each_symbol(vmlinux_map, sym, nd) {
210 		struct symbol *pair, *first_pair;
211 
212 		sym  = rb_entry(nd, struct symbol, rb_node);
213 
214 		if (sym->start == sym->end)
215 			continue;
216 
217 		mem_start = vmlinux_map->unmap_ip(vmlinux_map, sym->start);
218 		mem_end = vmlinux_map->unmap_ip(vmlinux_map, sym->end);
219 
220 		first_pair = machine__find_kernel_symbol(&kallsyms, mem_start, NULL);
221 		pair = first_pair;
222 
223 		if (pair && UM(pair->start) == mem_start) {
224 next_pair:
225 			if (arch__compare_symbol_names(sym->name, pair->name) == 0) {
226 				/*
227 				 * kallsyms don't have the symbol end, so we
228 				 * set that by using the next symbol start - 1,
229 				 * in some cases we get this up to a page
230 				 * wrong, trace_kmalloc when I was developing
231 				 * this code was one such example, 2106 bytes
232 				 * off the real size. More than that and we
233 				 * _really_ have a problem.
234 				 */
235 				s64 skew = mem_end - UM(pair->end);
236 				if (llabs(skew) >= page_size)
237 					pr_debug("WARN: %#" PRIx64 ": diff end addr for %s v: %#" PRIx64 " k: %#" PRIx64 "\n",
238 						 mem_start, sym->name, mem_end,
239 						 UM(pair->end));
240 
241 				/*
242 				 * Do not count this as a failure, because we
243 				 * could really find a case where it's not
244 				 * possible to get proper function end from
245 				 * kallsyms.
246 				 */
247 				continue;
248 			} else {
249 				pair = machine__find_kernel_symbol_by_name(&kallsyms, sym->name, NULL);
250 				if (pair) {
251 					if (UM(pair->start) == mem_start)
252 						goto next_pair;
253 
254 					pr_debug("WARN: %#" PRIx64 ": diff name v: %s k: %s\n",
255 						 mem_start, sym->name, pair->name);
256 				} else {
257 					pr_debug("WARN: %#" PRIx64 ": diff name v: %s k: %s\n",
258 						 mem_start, sym->name, first_pair->name);
259 				}
260 
261 				continue;
262 			}
263 		} else if (mem_start == kallsyms.vmlinux_map->end) {
264 			/*
265 			 * Ignore aliases to _etext, i.e. to the end of the kernel text area,
266 			 * such as __indirect_thunk_end.
267 			 */
268 			continue;
269 		} else if (is_ignored_symbol(sym->name, sym->type)) {
270 			/*
271 			 * Ignore hidden symbols, see scripts/kallsyms.c for the details
272 			 */
273 			continue;
274 		} else {
275 			pr_debug("ERR : %#" PRIx64 ": %s not on kallsyms\n",
276 				 mem_start, sym->name);
277 		}
278 
279 		err = -1;
280 	}
281 
282 	if (verbose <= 0)
283 		goto out;
284 
285 	header_printed = false;
286 
287 	maps__for_each_entry(maps, map) {
288 		struct map *
289 		/*
290 		 * If it is the kernel, kallsyms is always "[kernel.kallsyms]", while
291 		 * the kernel will have the path for the vmlinux file being used,
292 		 * so use the short name, less descriptive but the same ("[kernel]" in
293 		 * both cases.
294 		 */
295 		pair = maps__find_by_name(&kallsyms.kmaps, (map->dso->kernel ?
296 								map->dso->short_name :
297 								map->dso->name));
298 		if (pair) {
299 			pair->priv = 1;
300 		} else {
301 			if (!header_printed) {
302 				pr_info("WARN: Maps only in vmlinux:\n");
303 				header_printed = true;
304 			}
305 			map__fprintf(map, stderr);
306 		}
307 	}
308 
309 	header_printed = false;
310 
311 	maps__for_each_entry(maps, map) {
312 		struct map *pair;
313 
314 		mem_start = vmlinux_map->unmap_ip(vmlinux_map, map->start);
315 		mem_end = vmlinux_map->unmap_ip(vmlinux_map, map->end);
316 
317 		pair = maps__find(&kallsyms.kmaps, mem_start);
318 		if (pair == NULL || pair->priv)
319 			continue;
320 
321 		if (pair->start == mem_start) {
322 			if (!header_printed) {
323 				pr_info("WARN: Maps in vmlinux with a different name in kallsyms:\n");
324 				header_printed = true;
325 			}
326 
327 			pr_info("WARN: %" PRIx64 "-%" PRIx64 " %" PRIx64 " %s in kallsyms as",
328 				map->start, map->end, map->pgoff, map->dso->name);
329 			if (mem_end != pair->end)
330 				pr_info(":\nWARN: *%" PRIx64 "-%" PRIx64 " %" PRIx64,
331 					pair->start, pair->end, pair->pgoff);
332 			pr_info(" %s\n", pair->dso->name);
333 			pair->priv = 1;
334 		}
335 	}
336 
337 	header_printed = false;
338 
339 	maps = machine__kernel_maps(&kallsyms);
340 
341 	maps__for_each_entry(maps, map) {
342 		if (!map->priv) {
343 			if (!header_printed) {
344 				pr_info("WARN: Maps only in kallsyms:\n");
345 				header_printed = true;
346 			}
347 			map__fprintf(map, stderr);
348 		}
349 	}
350 out:
351 	machine__exit(&kallsyms);
352 	machine__exit(&vmlinux);
353 	return err;
354 }
355