vmlinux-kallsyms.c

  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) map__unmap_ip(kallsyms_map, (x))
 17
 18static 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_offsets",
 30		"kallsyms_relative_base",
 31		"kallsyms_num_syms",
 32		"kallsyms_names",
 33		"kallsyms_markers",
 34		"kallsyms_token_table",
 35		"kallsyms_token_index",
 36		/* Exclude linker generated symbols which vary between passes */
 37		"_SDA_BASE_",		/* ppc */
 38		"_SDA2_BASE_",		/* ppc */
 39		NULL
 40	};
 41
 42	/* Symbol names that begin with the following are ignored.*/
 43	static const char * const ignored_prefixes[] = {
 44		"$",			/* local symbols for ARM, MIPS, etc. */
 45		".L",			/* local labels, .LBB,.Ltmpxxx,.L__unnamed_xx,.LASANPC, etc. */
 46		"__crc_",		/* modversions */
 47		"__efistub_",		/* arm64 EFI stub namespace */
 48		"__kvm_nvhe_$",		/* arm64 local symbols in non-VHE KVM namespace */
 49		"__kvm_nvhe_.L",	/* arm64 local symbols in 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
114struct test__vmlinux_matches_kallsyms_cb_args {
115	struct machine kallsyms;
116	struct map *vmlinux_map;
117	bool header_printed;
118};
119
120static int test__vmlinux_matches_kallsyms_cb1(struct map *map, void *data)
121{
122	struct test__vmlinux_matches_kallsyms_cb_args *args = data;
123	struct dso *dso = map__dso(map);
124	/*
125	 * If it is the kernel, kallsyms is always "[kernel.kallsyms]", while
126	 * the kernel will have the path for the vmlinux file being used, so use
127	 * the short name, less descriptive but the same ("[kernel]" in both
128	 * cases.
129	 */
130	struct map *pair = maps__find_by_name(args->kallsyms.kmaps,
131					(dso__kernel(dso) ? dso__short_name(dso) : dso__name(dso)));
132
133	if (pair) {
134		map__set_priv(pair);
135		map__put(pair);
136	} else {
137		if (!args->header_printed) {
138			pr_info("WARN: Maps only in vmlinux:\n");
139			args->header_printed = true;
140		}
141		map__fprintf(map, stderr);
142	}
143	return 0;
144}
145
146static int test__vmlinux_matches_kallsyms_cb2(struct map *map, void *data)
147{
148	struct test__vmlinux_matches_kallsyms_cb_args *args = data;
149	struct map *pair;
150	u64 mem_start = map__unmap_ip(args->vmlinux_map, map__start(map));
151	u64 mem_end = map__unmap_ip(args->vmlinux_map, map__end(map));
152
153	pair = maps__find(args->kallsyms.kmaps, mem_start);
154
155	if (pair != NULL && !map__priv(pair) && map__start(pair) == mem_start) {
156		struct dso *dso = map__dso(map);
157
158		if (!args->header_printed) {
159			pr_info("WARN: Maps in vmlinux with a different name in kallsyms:\n");
160			args->header_printed = true;
161		}
162
163		pr_info("WARN: %" PRIx64 "-%" PRIx64 " %" PRIx64 " %s in kallsyms as",
164			map__start(map), map__end(map), map__pgoff(map), dso__name(dso));
165		if (mem_end != map__end(pair))
166			pr_info(":\nWARN: *%" PRIx64 "-%" PRIx64 " %" PRIx64,
167				map__start(pair), map__end(pair), map__pgoff(pair));
168		pr_info(" %s\n", dso__name(dso));
169		map__set_priv(pair);
170	}
171	map__put(pair);
172	return 0;
173}
174
175static int test__vmlinux_matches_kallsyms_cb3(struct map *map, void *data)
176{
177	struct test__vmlinux_matches_kallsyms_cb_args *args = data;
178
179	if (!map__priv(map)) {
180		if (!args->header_printed) {
181			pr_info("WARN: Maps only in kallsyms:\n");
182			args->header_printed = true;
183		}
184		map__fprintf(map, stderr);
185	}
186	return 0;
187}
188
189static int test__vmlinux_matches_kallsyms(struct test_suite *test __maybe_unused,
190					int subtest __maybe_unused)
191{
192	int err = TEST_FAIL;
193	struct rb_node *nd;
194	struct symbol *sym;
195	struct map *kallsyms_map;
196	struct machine vmlinux;
197	struct maps *maps;
198	u64 mem_start, mem_end;
199	struct test__vmlinux_matches_kallsyms_cb_args args;
200
201	/*
202	 * Step 1:
203	 *
204	 * Init the machines that will hold kernel, modules obtained from
205	 * both vmlinux + .ko files and from /proc/kallsyms split by modules.
206	 */
207	machine__init(&args.kallsyms, "", HOST_KERNEL_ID);
208	machine__init(&vmlinux, "", HOST_KERNEL_ID);
209
210	maps = machine__kernel_maps(&vmlinux);
211
212	/*
213	 * Step 2:
214	 *
215	 * Create the kernel maps for kallsyms and the DSO where we will then
216	 * load /proc/kallsyms. Also create the modules maps from /proc/modules
217	 * and find the .ko files that match them in /lib/modules/`uname -r`/.
218	 */
219	if (machine__create_kernel_maps(&args.kallsyms) < 0) {
220		pr_debug("machine__create_kernel_maps failed");
221		err = TEST_SKIP;
222		goto out;
223	}
224
225	/*
226	 * Step 3:
227	 *
228	 * Load and split /proc/kallsyms into multiple maps, one per module.
229	 * Do not use kcore, as this test was designed before kcore support
230	 * and has parts that only make sense if using the non-kcore code.
231	 * XXX: extend it to stress the kcorre code as well, hint: the list
232	 * of modules extracted from /proc/kcore, in its current form, can't
233	 * be compacted against the list of modules found in the "vmlinux"
234	 * code and with the one got from /proc/modules from the "kallsyms" code.
235	 */
236	if (machine__load_kallsyms(&args.kallsyms, "/proc/kallsyms") <= 0) {
237		pr_debug("machine__load_kallsyms failed");
238		err = TEST_SKIP;
239		goto out;
240	}
241
242	/*
243	 * Step 4:
244	 *
245	 * kallsyms will be internally on demand sorted by name so that we can
246	 * find the reference relocation * symbol, i.e. the symbol we will use
247	 * to see if the running kernel was relocated by checking if it has the
248	 * same value in the vmlinux file we load.
249	 */
250	kallsyms_map = machine__kernel_map(&args.kallsyms);
251
252	/*
253	 * Step 5:
254	 *
255	 * Now repeat step 2, this time for the vmlinux file we'll auto-locate.
256	 */
257	if (machine__create_kernel_maps(&vmlinux) < 0) {
258		pr_info("machine__create_kernel_maps failed");
259		goto out;
260	}
261
262	args.vmlinux_map = machine__kernel_map(&vmlinux);
263
264	/*
265	 * Step 6:
266	 *
267	 * Locate a vmlinux file in the vmlinux path that has a buildid that
268	 * matches the one of the running kernel.
269	 *
270	 * While doing that look if we find the ref reloc symbol, if we find it
271	 * we'll have its ref_reloc_symbol.unrelocated_addr and then
272	 * maps__reloc_vmlinux will notice and set proper ->[un]map_ip routines
273	 * to fixup the symbols.
274	 */
275	if (machine__load_vmlinux_path(&vmlinux) <= 0) {
276		pr_info("Couldn't find a vmlinux that matches the kernel running on this machine, skipping test\n");
277		err = TEST_SKIP;
278		goto out;
279	}
280
281	err = 0;
282	/*
283	 * Step 7:
284	 *
285	 * Now look at the symbols in the vmlinux DSO and check if we find all of them
286	 * in the kallsyms dso. For the ones that are in both, check its names and
287	 * end addresses too.
288	 */
289	map__for_each_symbol(args.vmlinux_map, sym, nd) {
290		struct symbol *pair, *first_pair;
291
292		sym  = rb_entry(nd, struct symbol, rb_node);
293
294		if (sym->start == sym->end)
295			continue;
296
297		mem_start = map__unmap_ip(args.vmlinux_map, sym->start);
298		mem_end = map__unmap_ip(args.vmlinux_map, sym->end);
299
300		first_pair = machine__find_kernel_symbol(&args.kallsyms, mem_start, NULL);
301		pair = first_pair;
302
303		if (pair && UM(pair->start) == mem_start) {
304next_pair:
305			if (arch__compare_symbol_names(sym->name, pair->name) == 0) {
306				/*
307				 * kallsyms don't have the symbol end, so we
308				 * set that by using the next symbol start - 1,
309				 * in some cases we get this up to a page
310				 * wrong, trace_kmalloc when I was developing
311				 * this code was one such example, 2106 bytes
312				 * off the real size. More than that and we
313				 * _really_ have a problem.
314				 */
315				s64 skew = mem_end - UM(pair->end);
316				if (llabs(skew) >= page_size)
317					pr_debug("WARN: %#" PRIx64 ": diff end addr for %s v: %#" PRIx64 " k: %#" PRIx64 "\n",
318						 mem_start, sym->name, mem_end,
319						 UM(pair->end));
320
321				/*
322				 * Do not count this as a failure, because we
323				 * could really find a case where it's not
324				 * possible to get proper function end from
325				 * kallsyms.
326				 */
327				continue;
328			} else {
329				pair = machine__find_kernel_symbol_by_name(&args.kallsyms,
330									   sym->name, NULL);
331				if (pair) {
332					if (UM(pair->start) == mem_start)
333						goto next_pair;
334
335					pr_debug("WARN: %#" PRIx64 ": diff name v: %s k: %s\n",
336						 mem_start, sym->name, pair->name);
337				} else {
338					pr_debug("WARN: %#" PRIx64 ": diff name v: %s k: %s\n",
339						 mem_start, sym->name, first_pair->name);
340				}
341
342				continue;
343			}
344		} else if (mem_start == map__end(args.kallsyms.vmlinux_map)) {
345			/*
346			 * Ignore aliases to _etext, i.e. to the end of the kernel text area,
347			 * such as __indirect_thunk_end.
348			 */
349			continue;
350		} else if (is_ignored_symbol(sym->name, sym->type)) {
351			/*
352			 * Ignore hidden symbols, see scripts/kallsyms.c for the details
353			 */
354			continue;
355		} else {
356			pr_debug("ERR : %#" PRIx64 ": %s not on kallsyms\n",
357				 mem_start, sym->name);
358		}
359
360		err = -1;
361	}
362
363	if (verbose <= 0)
364		goto out;
365
366	args.header_printed = false;
367	maps__for_each_map(maps, test__vmlinux_matches_kallsyms_cb1, &args);
368
369	args.header_printed = false;
370	maps__for_each_map(maps, test__vmlinux_matches_kallsyms_cb2, &args);
371
372	args.header_printed = false;
373	maps = machine__kernel_maps(&args.kallsyms);
374	maps__for_each_map(maps, test__vmlinux_matches_kallsyms_cb3, &args);
375
376out:
377	machine__exit(&args.kallsyms);
378	machine__exit(&vmlinux);
379	return err;
380}
381
382DEFINE_SUITE("vmlinux symtab matches kallsyms", vmlinux_matches_kallsyms);