Skip to contentSkip to conversation

bpf: Fixes for kprobe multi on kernel modules

PATCH Applied Jiri Olsa 2 participants 10 messages Oct 25, 2022

Subsystems (9)
Kselftest BPF Tracing Arm STM32 +5 more subsystems
lore
Share & tools
lore

Jiri Olsa

author reviewer
#0 lore raw

v3 cover bpf: Fixes for kprobe multi on kernel modules

hi,
Martynas reported kprobe _multi link does not resolve symbols from kernel modules, which attach by address works.

In addition while fixing that I realized we do not take module reference if the module has kprobe_multi link on top of it and can be removed.

There's mo crash related to this, it will silently disappear from ftrace tables, while kprobe_multi link stays up with no data.

This patchset has fixes for both issues.

v3 changes:
- reorder fields in struct bpf_kprobe_multi_link [Andrii]
- added ack [Andrii]

v2 changes:
- added acks (Song)
- added comment to kallsyms_callback (Song)
- change module_callback realloc logic (Andrii)
- get rid of macros in tests (Andrii)

thanks,
jirka

---
Jiri Olsa (8):
kallsyms: Make module_kallsyms_on_each_symbol generally available
ftrace: Add support to resolve module symbols in ftrace_lookup_symbols
bpf: Rename __bpf_kprobe_multi_cookie_cmp to bpf_kprobe_multi_addrs_cmp
bpf: Take module reference on kprobe_multi link selftests/bpf: Add load_kallsyms_refresh function selftests/bpf: Add bpf_testmod_fentry_* functions selftests/bpf: Add kprobe_multi check to module attach test selftests/bpf: Add kprobe_multi kmod attach api tests

 include/linux/module.h                                             |  9 +++++++++
 kernel/module/kallsyms.c                                           |  2 --
 kernel/trace/bpf_trace.c                                           | 98 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++---
 kernel/trace/ftrace.c                                              | 16 +++++++++++-----
 tools/testing/selftests/bpf/bpf_testmod/bpf_testmod.c              | 24 ++++++++++++++++++++++++
 tools/testing/selftests/bpf/prog_tests/kprobe_multi_testmod_test.c | 89 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
 tools/testing/selftests/bpf/prog_tests/module_attach.c             |  7 +++++++
 tools/testing/selftests/bpf/progs/kprobe_multi.c                   | 50 ++++++++++++++++++++++++++++++++++++++++++++++++++
 tools/testing/selftests/bpf/progs/test_module_attach.c             |  6 ++++++
 tools/testing/selftests/bpf/trace_helpers.c                        | 20 +++++++++++++-------
 tools/testing/selftests/bpf/trace_helpers.h                        |  2 ++
 11 files changed, 306 insertions(+), 17 deletions(-)
 create mode 100644 tools/testing/selftests/bpf/prog_tests/kprobe_multi_testmod_test.c
author reviewer
#1 lore raw

PATCH v3 1/8 kallsyms: Make module_kallsyms_on_each_symbol generally available

Making module_kallsyms_on_each_symbol generally available, so it can be used outside CONFIG_LIVEPATCH option in following changes.

Rather than adding another ifdef option let's make the function generally available (when CONFIG_KALLSYMS and CONFIG_MODULES options are defined).

---
 include/linux/module.h   | 9 +++++++++
 kernel/module/kallsyms.c | 2 --
 2 files changed, 9 insertions(+), 2 deletions(-)
include/linux/module.h +9 -0
index ec61fb53979a..35876e89eb93 100644
@@ -879,8 +879,17 @@ static inline bool module_sig_ok(struct module *module)
 }
 #endif	/* CONFIG_MODULE_SIG */
 
+#if defined(CONFIG_MODULES) && defined(CONFIG_KALLSYMS)
 int module_kallsyms_on_each_symbol(int (*fn)(void *, const char *,
 					     struct module *, unsigned long),
 				   void *data);
+#else
+static inline int module_kallsyms_on_each_symbol(int (*fn)(void *, const char *,
+						 struct module *, unsigned long),
+						 void *data)
+{
+	return -EOPNOTSUPP;
+}
+#endif  /* CONFIG_MODULES && CONFIG_KALLSYMS */
 
 #endif /* _LINUX_MODULE_H */
kernel/module/kallsyms.c +0 -2
index f5c5c9175333..4523f99b0358 100644
@@ -494,7 +494,6 @@ unsigned long module_kallsyms_lookup_name(const char *name)
 	return ret;
 }
 
-#ifdef CONFIG_LIVEPATCH
 int module_kallsyms_on_each_symbol(int (*fn)(void *, const char *,
 					     struct module *, unsigned long),
 				   void *data)
@@ -531,4 +530,3 @@ int module_kallsyms_on_each_symbol(int (*fn)(void *, const char *,
 	mutex_unlock(&module_mutex);
 	return ret;
 }
-#endif /* CONFIG_LIVEPATCH */
Signature · 1 line
2.37.3
author reviewer
#2 lore raw

PATCH v3 2/8 ftrace: Add support to resolve module symbols in ftrace_lookup_symbols

In reply to Jiri Olsa

Currently ftrace_lookup_symbols iterates only over core symbols, adding module_kallsyms_on_each_symbol call to check on modules symbols as well.

Also removing 'args.found == args.cnt' condition, because it's already checked in kallsyms_callback function.

Also removing 'err < 0' check, because both *kallsyms_on_each_symbol functions do not return error.

---
 kernel/trace/ftrace.c | 16 +++++++++++-----
 1 file changed, 11 insertions(+), 5 deletions(-)
kernel/trace/ftrace.c +11 -5
index fbf2543111c0..72de9009a6a0 100644
@@ -8267,6 +8267,10 @@ struct kallsyms_data {
 	size_t found;
 };
 
+/* This function gets called for all kernel and module symbols
+ * and returns 1 in case we resolved all the requested symbols,
+ * 0 otherwise.
+ */
 static int kallsyms_callback(void *data, const char *name,
 			     struct module *mod, unsigned long addr)
 {
@@ -8309,17 +8313,19 @@ static int kallsyms_callback(void *data, const char *name,
 int ftrace_lookup_symbols(const char **sorted_syms, size_t cnt, unsigned long *addrs)
 {
 	struct kallsyms_data args;
-	int err;
+	int found_all;
 
 	memset(addrs, 0, sizeof(*addrs) * cnt);
 	args.addrs = addrs;
 	args.syms = sorted_syms;
 	args.cnt = cnt;
 	args.found = 0;
-	err = kallsyms_on_each_symbol(kallsyms_callback, &args);
-	if (err < 0)
-		return err;
-	return args.found == args.cnt ? 0 : -ESRCH;
+
+	found_all = kallsyms_on_each_symbol(kallsyms_callback, &args);
+	if (found_all)
+		return 0;
+	found_all = module_kallsyms_on_each_symbol(kallsyms_callback, &args);
+	return found_all ? 0 : -ESRCH;
 }
 
 #ifdef CONFIG_SYSCTL
Signature · 1 line
2.37.3
author reviewer
#3 lore raw

PATCH v3 3/8 bpf: Rename __bpf_kprobe_multi_cookie_cmp to bpf_kprobe_multi_addrs_cmp

In reply to Jiri Olsa

Renaming __bpf_kprobe_multi_cookie_cmp to bpf_kprobe_multi_addrs_cmp, because it's more suitable to current and upcoming code.

---
 kernel/trace/bpf_trace.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
kernel/trace/bpf_trace.c +3 -3
index 1ed08967fb97..17ae9e8336db 100644
@@ -2550,7 +2550,7 @@ static void bpf_kprobe_multi_cookie_swap(void *a, void *b, int size, const void
 	swap(*cookie_a, *cookie_b);
 }
 
-static int __bpf_kprobe_multi_cookie_cmp(const void *a, const void *b)
+static int bpf_kprobe_multi_addrs_cmp(const void *a, const void *b)
 {
 	const unsigned long *addr_a = a, *addr_b = b;
 
@@ -2561,7 +2561,7 @@ static int __bpf_kprobe_multi_cookie_cmp(const void *a, const void *b)
 
 static int bpf_kprobe_multi_cookie_cmp(const void *a, const void *b, const void *priv)
 {
-	return __bpf_kprobe_multi_cookie_cmp(a, b);
+	return bpf_kprobe_multi_addrs_cmp(a, b);
 }
 
 static u64 bpf_kprobe_multi_cookie(struct bpf_run_ctx *ctx)
@@ -2579,7 +2579,7 @@ static u64 bpf_kprobe_multi_cookie(struct bpf_run_ctx *ctx)
 		return 0;
 	entry_ip = run_ctx->entry_ip;
 	addr = bsearch(&entry_ip, link->addrs, link->cnt, sizeof(entry_ip),
-		       __bpf_kprobe_multi_cookie_cmp);
+		       bpf_kprobe_multi_addrs_cmp);
 	if (!addr)
 		return 0;
 	cookie = link->cookies + (addr - link->addrs);
Signature · 1 line
2.37.3
author reviewer
#4 lore raw

PATCH v3 4/8 bpf: Take module reference on kprobe_multi link

In reply to Jiri Olsa

Currently we allow to create kprobe multi link on function from kernel module, but we don't take the module reference to ensure it's not unloaded while we are tracing it.

The multi kprobe link is based on fprobe/ftrace layer which takes different approach and releases ftrace hooks when module is unloaded even if there's tracer registered on top of it.

Adding code that gathers all the related modules for the link and takes their references before it's attached. All kernel module references are released after link is unregistered.

Note that we do it the same way already for trampoline probes (but for single address).

---
 kernel/trace/bpf_trace.c | 92 ++++++++++++++++++++++++++++++++++++++++
 1 file changed, 92 insertions(+)
kernel/trace/bpf_trace.c +92 -0
index 17ae9e8336db..ad6c21709dcb 100644
@@ -2452,6 +2452,8 @@ struct bpf_kprobe_multi_link {
 	unsigned long *addrs;
 	u64 *cookies;
 	u32 cnt;
+	u32 mods_cnt;
+	struct module **mods;
 };
 
 struct bpf_kprobe_multi_run_ctx {
@@ -2507,6 +2509,14 @@ static int copy_user_syms(struct user_syms *us, unsigned long __user *usyms, u32
 	return err;
 }
 
+static void kprobe_multi_put_modules(struct module **mods, u32 cnt)
+{
+	u32 i;
+
+	for (i = 0; i < cnt; i++)
+		module_put(mods[i]);
+}
+
 static void free_user_syms(struct user_syms *us)
 {
 	kvfree(us->syms);
@@ -2519,6 +2529,7 @@ static void bpf_kprobe_multi_link_release(struct bpf_link *link)
 
 	kmulti_link = container_of(link, struct bpf_kprobe_multi_link, link);
 	unregister_fprobe(&kmulti_link->fp);
+	kprobe_multi_put_modules(kmulti_link->mods, kmulti_link->mods_cnt);
 }
 
 static void bpf_kprobe_multi_link_dealloc(struct bpf_link *link)
@@ -2528,6 +2539,7 @@ static void bpf_kprobe_multi_link_dealloc(struct bpf_link *link)
 	kmulti_link = container_of(link, struct bpf_kprobe_multi_link, link);
 	kvfree(kmulti_link->addrs);
 	kvfree(kmulti_link->cookies);
+	kfree(kmulti_link->mods);
 	kfree(kmulti_link);
 }
 
@@ -2663,6 +2675,71 @@ static void symbols_swap_r(void *a, void *b, int size, const void *priv)
 	}
 }
 
+struct module_addr_args {
+	unsigned long *addrs;
+	u32 addrs_cnt;
+	struct module **mods;
+	int mods_cnt;
+	int mods_cap;
+};
+
+static int module_callback(void *data, const char *name,
+			   struct module *mod, unsigned long addr)
+{
+	struct module_addr_args *args = data;
+	struct module **mods;
+
+	/* We iterate all modules symbols and for each we:
+	 * - search for it in provided addresses array
+	 * - if found we check if we already have the module pointer stored
+	 *   (we iterate modules sequentially, so we can check just the last
+	 *   module pointer)
+	 * - take module reference and store it
+	 */
+	if (!bsearch(&addr, args->addrs, args->addrs_cnt, sizeof(addr),
+		       bpf_kprobe_multi_addrs_cmp))
+		return 0;
+
+	if (args->mods && args->mods[args->mods_cnt - 1] == mod)
+		return 0;
+
+	if (args->mods_cnt == args->mods_cap) {
+		args->mods_cap = max(16, args->mods_cap * 3 / 2);
+		mods = krealloc_array(args->mods, args->mods_cap, sizeof(*mods), GFP_KERNEL);
+		if (!mods)
+			return -ENOMEM;
+		args->mods = mods;
+	}
+
+	if (!try_module_get(mod))
+		return -EINVAL;
+
+	args->mods[args->mods_cnt] = mod;
+	args->mods_cnt++;
+	return 0;
+}
+
+static int get_modules_for_addrs(struct module ***mods, unsigned long *addrs, u32 addrs_cnt)
+{
+	struct module_addr_args args = {
+		.addrs     = addrs,
+		.addrs_cnt = addrs_cnt,
+	};
+	int err;
+
+	/* We return either err < 0 in case of error, ... */
+	err = module_kallsyms_on_each_symbol(module_callback, &args);
+	if (err) {
+		kprobe_multi_put_modules(args.mods, args.mods_cnt);
+		kfree(args.mods);
+		return err;
+	}
+
+	/* or number of modules found if everything is ok. */
+	*mods = args.mods;
+	return args.mods_cnt;
+}
+
 int bpf_kprobe_multi_link_attach(const union bpf_attr *attr, struct bpf_prog *prog)
 {
 	struct bpf_kprobe_multi_link *link = NULL;
@@ -2773,10 +2850,25 @@ int bpf_kprobe_multi_link_attach(const union bpf_attr *attr, struct bpf_prog *pr
 		       bpf_kprobe_multi_cookie_cmp,
 		       bpf_kprobe_multi_cookie_swap,
 		       link);
+	} else {
+		/*
+		 * We need to sort addrs array even if there are no cookies
+		 * provided, to allow bsearch in get_modules_for_addrs.
+		 */
+		sort(addrs, cnt, sizeof(*addrs),
+		       bpf_kprobe_multi_addrs_cmp, NULL);
+	}
+
+	err = get_modules_for_addrs(&link->mods, addrs, cnt);
+	if (err < 0) {
+		bpf_link_cleanup(&link_primer);
+		return err;
 	}
+	link->mods_cnt = err;
 
 	err = register_fprobe_ips(&link->fp, addrs, cnt);
 	if (err) {
+		kprobe_multi_put_modules(link->mods, link->mods_cnt);
 		bpf_link_cleanup(&link_primer);
 		return err;
 	}
Signature · 1 line
2.37.3
author reviewer
#5 lore raw

PATCH v3 5/8 selftests/bpf: Add load_kallsyms_refresh function

In reply to Jiri Olsa

Adding load_kallsyms_refresh function to re-read symbols from /proc/kallsyms file.

This will be needed to get proper functions addresses from bpf_testmod.ko module, which is loaded/unloaded several times during the tests run, so symbols might be already old when we need to use them.

---
 tools/testing/selftests/bpf/trace_helpers.c | 20 +++++++++++++-------
 tools/testing/selftests/bpf/trace_helpers.h |  2 ++
 2 files changed, 15 insertions(+), 7 deletions(-)
tools/testing/selftests/bpf/trace_helpers.c +13 -7
index 9c4be2cdb21a..09a16a77bae4 100644
@@ -23,7 +23,7 @@ static int ksym_cmp(const void *p1, const void *p2)
 	return ((struct ksym *)p1)->addr - ((struct ksym *)p2)->addr;
 }
 
-int load_kallsyms(void)
+int load_kallsyms_refresh(void)
 {
 	FILE *f;
 	char func[256], buf[256];
@@ -31,12 +31,7 @@ int load_kallsyms(void)
 	void *addr;
 	int i = 0;
 
-	/*
-	 * This is called/used from multiplace places,
-	 * load symbols just once.
-	 */
-	if (sym_cnt)
-		return 0;
+	sym_cnt = 0;
 
 	f = fopen("/proc/kallsyms", "r");
 	if (!f)
@@ -57,6 +52,17 @@ int load_kallsyms(void)
 	return 0;
 }
 
+int load_kallsyms(void)
+{
+	/*
+	 * This is called/used from multiplace places,
+	 * load symbols just once.
+	 */
+	if (sym_cnt)
+		return 0;
+	return load_kallsyms_refresh();
+}
+
 struct ksym *ksym_search(long key)
 {
 	int start = 0, end = sym_cnt;
tools/testing/selftests/bpf/trace_helpers.h +2 -0
index 238a9c98cde2..53efde0e2998 100644
@@ -10,6 +10,8 @@ struct ksym {
 };
 
 int load_kallsyms(void);
+int load_kallsyms_refresh(void);
+
 struct ksym *ksym_search(long key);
 long ksym_get_addr(const char *name);
 
Signature · 1 line
2.37.3
author reviewer
#6 lore raw

PATCH v3 6/8 selftests/bpf: Add bpf_testmod_fentry_* functions

In reply to Jiri Olsa

Adding 3 bpf_testmod_fentry_* functions to have a way to test kprobe multi link on kernel module. They follow bpf_fentry_test* functions prototypes/code.

Adding equivalent functions to all bpf_fentry_test* does not seems necessary at the moment, could be added later.

---
 .../selftests/bpf/bpf_testmod/bpf_testmod.c   | 24 +++++++++++++++++++
 1 file changed, 24 insertions(+)
tools/testing/selftests/bpf/bpf_testmod/bpf_testmod.c +24 -0
index a6021d6117b5..5085fea3cac5 100644
@@ -128,6 +128,23 @@ __weak noinline struct file *bpf_testmod_return_ptr(int arg)
 	}
 }
 
+noinline int bpf_testmod_fentry_test1(int a)
+{
+	return a + 1;
+}
+
+noinline int bpf_testmod_fentry_test2(int a, u64 b)
+{
+	return a + b;
+}
+
+noinline int bpf_testmod_fentry_test3(char a, int b, u64 c)
+{
+	return a + b + c;
+}
+
+int bpf_testmod_fentry_ok;
+
 noinline ssize_t
 bpf_testmod_test_read(struct file *file, struct kobject *kobj,
 		      struct bin_attribute *bin_attr,
@@ -167,6 +184,13 @@ bpf_testmod_test_read(struct file *file, struct kobject *kobj,
 			return snprintf(buf, len, "%d\n", writable.val);
 	}
 
+	if (bpf_testmod_fentry_test1(1) != 2 ||
+	    bpf_testmod_fentry_test2(2, 3) != 5 ||
+	    bpf_testmod_fentry_test3(4, 5, 6) != 15)
+		goto out;
+
+	bpf_testmod_fentry_ok = 1;
+out:
 	return -EIO; /* always fail */
 }
 EXPORT_SYMBOL(bpf_testmod_test_read);
Signature · 1 line
2.37.3
author reviewer
#7 lore raw

PATCH v3 7/8 selftests/bpf: Add kprobe_multi check to module attach test

In reply to Jiri Olsa

Adding test that makes sure the kernel module won't be removed if there's kprobe multi link defined on top of it.

---
 tools/testing/selftests/bpf/prog_tests/module_attach.c | 7 +++++++
 tools/testing/selftests/bpf/progs/test_module_attach.c | 6 ++++++
 2 files changed, 13 insertions(+)
tools/testing/selftests/bpf/prog_tests/module_attach.c +7 -0
index 6d0e50dcf47c..7fc01ff490db 100644
@@ -103,6 +103,13 @@ void test_module_attach(void)
 	ASSERT_ERR(delete_module("bpf_testmod", 0), "delete_module");
 	bpf_link__destroy(link);
 
+	link = bpf_program__attach(skel->progs.kprobe_multi);
+	if (!ASSERT_OK_PTR(link, "attach_kprobe_multi"))
+		goto cleanup;
+
+	ASSERT_ERR(delete_module("bpf_testmod", 0), "delete_module");
+	bpf_link__destroy(link);
+
 cleanup:
 	test_module_attach__destroy(skel);
 }
tools/testing/selftests/bpf/progs/test_module_attach.c +6 -0
index 08628afedb77..8a1b50f3a002 100644
@@ -110,4 +110,10 @@ int BPF_PROG(handle_fmod_ret,
 	return 0; /* don't override the exit code */
 }
 
+SEC("kprobe.multi/bpf_testmod_test_read")
+int BPF_PROG(kprobe_multi)
+{
+	return 0;
+}
+
 char _license[] SEC("license") = "GPL";
Signature · 1 line
2.37.3
author reviewer
#8 lore raw

PATCH v3 8/8 selftests/bpf: Add kprobe_multi kmod attach api tests

In reply to Jiri Olsa

Adding kprobe_multi kmod attach api tests that attach bpf_testmod functions via bpf_program__attach_kprobe_multi_opts.

Running it as serial test, because we don't want other tests to reload bpf_testmod while it's running.

---
 .../prog_tests/kprobe_multi_testmod_test.c    | 89 +++++++++++++++++++
 .../selftests/bpf/progs/kprobe_multi.c        | 50 +++++++++++
 2 files changed, 139 insertions(+)
 create mode 100644 tools/testing/selftests/bpf/prog_tests/kprobe_multi_testmod_test.c
tools/testing/selftests/bpf/prog_tests/kprobe_multi_testmod_test.c new file +89 -0
new file mode 100644
index 000000000000..1fbe7e4ac00a
@@ -0,0 +1,89 @@
+// SPDX-License-Identifier: GPL-2.0
+#include <test_progs.h>
+#include "kprobe_multi.skel.h"
+#include "trace_helpers.h"
+#include "bpf/libbpf_internal.h"
+
+static void kprobe_multi_testmod_check(struct kprobe_multi *skel)
+{
+	ASSERT_EQ(skel->bss->kprobe_testmod_test1_result, 1, "kprobe_test1_result");
+	ASSERT_EQ(skel->bss->kprobe_testmod_test2_result, 1, "kprobe_test2_result");
+	ASSERT_EQ(skel->bss->kprobe_testmod_test3_result, 1, "kprobe_test3_result");
+
+	ASSERT_EQ(skel->bss->kretprobe_testmod_test1_result, 1, "kretprobe_test1_result");
+	ASSERT_EQ(skel->bss->kretprobe_testmod_test2_result, 1, "kretprobe_test2_result");
+	ASSERT_EQ(skel->bss->kretprobe_testmod_test3_result, 1, "kretprobe_test3_result");
+}
+
+static void test_testmod_attach_api(struct bpf_kprobe_multi_opts *opts)
+{
+	struct kprobe_multi *skel = NULL;
+
+	skel = kprobe_multi__open_and_load();
+	if (!ASSERT_OK_PTR(skel, "fentry_raw_skel_load"))
+		return;
+
+	skel->bss->pid = getpid();
+
+	skel->links.test_kprobe_testmod = bpf_program__attach_kprobe_multi_opts(
+						skel->progs.test_kprobe_testmod,
+						NULL, opts);
+	if (!skel->links.test_kprobe_testmod)
+		goto cleanup;
+
+	opts->retprobe = true;
+	skel->links.test_kretprobe_testmod = bpf_program__attach_kprobe_multi_opts(
+						skel->progs.test_kretprobe_testmod,
+						NULL, opts);
+	if (!skel->links.test_kretprobe_testmod)
+		goto cleanup;
+
+	ASSERT_OK(trigger_module_test_read(1), "trigger_read");
+	kprobe_multi_testmod_check(skel);
+
+cleanup:
+	kprobe_multi__destroy(skel);
+}
+
+static void test_testmod_attach_api_addrs(void)
+{
+	LIBBPF_OPTS(bpf_kprobe_multi_opts, opts);
+	unsigned long long addrs[3];
+
+	addrs[0] = ksym_get_addr("bpf_testmod_fentry_test1");
+	ASSERT_NEQ(addrs[0], 0, "ksym_get_addr");
+	addrs[1] = ksym_get_addr("bpf_testmod_fentry_test2");
+	ASSERT_NEQ(addrs[1], 0, "ksym_get_addr");
+	addrs[2] = ksym_get_addr("bpf_testmod_fentry_test3");
+	ASSERT_NEQ(addrs[2], 0, "ksym_get_addr");
+
+	opts.addrs = (const unsigned long *) addrs;
+	opts.cnt = ARRAY_SIZE(addrs);
+
+	test_testmod_attach_api(&opts);
+}
+
+static void test_testmod_attach_api_syms(void)
+{
+	LIBBPF_OPTS(bpf_kprobe_multi_opts, opts);
+	const char *syms[3] = {
+		"bpf_testmod_fentry_test1",
+		"bpf_testmod_fentry_test2",
+		"bpf_testmod_fentry_test3",
+	};
+
+	opts.syms = syms;
+	opts.cnt = ARRAY_SIZE(syms);
+	test_testmod_attach_api(&opts);
+}
+
+void serial_test_kprobe_multi_testmod_test(void)
+{
+	if (!ASSERT_OK(load_kallsyms_refresh(), "load_kallsyms_refresh"))
+		return;
+
+	if (test__start_subtest("testmod_attach_api_syms"))
+		test_testmod_attach_api_syms();
+	if (test__start_subtest("testmod_attach_api_addrs"))
+		test_testmod_attach_api_addrs();
+}
tools/testing/selftests/bpf/progs/kprobe_multi.c +50 -0

53 lines not on this page · raw message

Signature · 1 line
2.37.3
#9 lore raw
In reply to Jiri Olsa
This series was applied to bpf/bpf-next.git (master) by Alexei Starovoitov <[email protected]>:

Hello:

This series was applied to bpf/bpf-next.git (master) by Alexei Starovoitov <[email protected]>:

Quoted 9 lines · On Tue, 25 Oct 2022 15:41:40 +0200 you wrote:

hi,
Martynas reported kprobe _multi link does not resolve symbols from kernel modules, which attach by address works.

In addition while fixing that I realized we do not take module reference if the module has kprobe_multi link on top of it and can be removed.

[...]

Here is the summary with links:
- [PATCHv3,bpf-next,1/8] kallsyms: Make module_kallsyms_on_each_symbol generally available

    https://git.kernel.org/bpf/bpf-next/c/73feb8d5fa3b
  - [PATCHv3,bpf-next,2/8] ftrace: Add support to resolve module symbols in ftrace_lookup_symbols
    https://git.kernel.org/bpf/bpf-next/c/3640bf8584f4
  - [PATCHv3,bpf-next,3/8] bpf: Rename __bpf_kprobe_multi_cookie_cmp to bpf_kprobe_multi_addrs_cmp
    https://git.kernel.org/bpf/bpf-next/c/1a1b0716d36d
  - [PATCHv3,bpf-next,4/8] bpf: Take module reference on kprobe_multi link
    https://git.kernel.org/bpf/bpf-next/c/e22061b2d309
  - [PATCHv3,bpf-next,5/8] selftests/bpf: Add load_kallsyms_refresh function
    https://git.kernel.org/bpf/bpf-next/c/10705b2b7a8e
  - [PATCHv3,bpf-next,6/8] selftests/bpf: Add bpf_testmod_fentry_* functions
    https://git.kernel.org/bpf/bpf-next/c/fee356ede980
  - [PATCHv3,bpf-next,7/8] selftests/bpf: Add kprobe_multi check to module attach test
    https://git.kernel.org/bpf/bpf-next/c/e697d8dcebd2
  - [PATCHv3,bpf-next,8/8] selftests/bpf: Add kprobe_multi kmod attach api tests
    https://git.kernel.org/bpf/bpf-next/c/b2440443a64f

You are awesome, thank you!

Signature · 2 lines
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html