[PATCH v2 4/4] selftests/bpf: add test validating uprobe/uretprobe stack traces
Add a set of tests to validate that stack traces captured from or in the
presence of active uprobes and uretprobes are valid and complete.
For this we use BPF program that are installed either on entry or exit
of user function, plus deep-nested USDT. One of target funtions
(target_1) is recursive to generate two different entries in the stack
trace for the same uprobe/uretprobe, testing potential edge conditions.
Without fixes in this patch set, we get something like this for one of
the scenarios:
caller: 0x758fff - 0x7595ab
target_1: 0x758fd5 - 0x758fff
target_2: 0x758fca - 0x758fd5
target_3: 0x758fbf - 0x758fca
target_4: 0x758fb3 - 0x758fbf
ENTRY #0: 0x758fb3 (in target_4)
ENTRY #1: 0x758fd3 (in target_2)
ENTRY #2: 0x758ffd (in target_1)
ENTRY #3: 0x7fffe000
ENTRY #4: 0x7fffe000
ENTRY #5: 0x6f8f39
ENTRY #6: 0x6fa6f0
ENTRY #7: 0x7f403f229590
Entry #3 and #4 (0x7fffe000) are uretprobe trampoline addresses
which obscure actual target_1 and another target_1 invocations. Also
note that between entry #0 and entry #1 we are missing an entry for
target_3, which is fixed in patch #2.
With all the fixes, we get desired full stack traces:
caller: 0x758fff - 0x7595ab
target_1: 0x758fd5 - 0x758fff
target_2: 0x758fca - 0x758fd5
target_3: 0x758fbf - 0x758fca
target_4: 0x758fb3 - 0x758fbf
ENTRY #0: 0x758fb7 (in target_4)
ENTRY #1: 0x758fc8 (in target_3)
ENTRY #2: 0x758fd3 (in target_2)
ENTRY #3: 0x758ffd (in target_1)
ENTRY #4: 0x758ff3 (in target_1)
ENTRY #5: 0x75922c (in caller)
ENTRY #6: 0x6f8f39
ENTRY #7: 0x6fa6f0
ENTRY #8: 0x7f986adc4cd0
Now there is a logical and complete sequence of function calls.
Signed-off-by: Andrii Nakryiko
---
.../bpf/prog_tests/uretprobe_stack.c | 186 ++
.../selftests/bpf/progs/uretprobe_stack.c | 96 +
2 files changed, 282 insertions(+)
create mode 100644 tools/testing/selftests/bpf/prog_tests/uretprobe_stack.c
create mode 100644 tools/testing/selftests/bpf/progs/uretprobe_stack.c
diff --git a/tools/testing/selftests/bpf/prog_tests/uretprobe_stack.c
b/tools/testing/selftests/bpf/prog_tests/uretprobe_stack.c
new file mode 100644
index ..6deb8d560ddd
--- /dev/null
+++ b/tools/testing/selftests/bpf/prog_tests/uretprobe_stack.c
@@ -0,0 +1,186 @@
+// SPDX-License-Identifier: GPL-2.0
+/* Copyright (c) 2024 Meta Platforms, Inc. and affiliates. */
+
+#include
+#include "uretprobe_stack.skel.h"
+#include "../sdt.h"
+
+/* We set up target_1() -> target_2() -> target_3() -> target_4() -> USDT()
+ * call chain, each being traced by our BPF program. On entry or return from
+ * each target_*() we are capturing user stack trace and recording it in
+ * global variable, so that user space part of the test can validate it.
+ *
+ * Note, we put each target function into a custom section to get those
+ * __start_XXX/__stop_XXX symbols, generated by linker for us, which allow us
+ * to know address range of those functions
+ */
+__attribute__((section("uprobe__target_4")))
+__weak int target_4(void)
+{
+ STAP_PROBE1(uretprobe_stack, target, 42);
+ return 42;
+}
+
+extern const void *__start_uprobe__target_4;
+extern const void *__stop_uprobe__target_4;
+
+__attribute__((section("uprobe__target_3")))
+__weak int target_3(void)
+{
+ return target_4();
+}
+
+extern const void *__start_uprobe__target_3;
+extern const void *__stop_uprobe__target_3;
+
+__attribute__((section("uprobe__target_2")))
+__weak int target_2(void)
+{
+ return target_3();
+}
+
+extern const void *__start_uprobe__target_2;
+extern const void *__stop_uprobe__target_2;
+
+__attribute__((section("uprobe__target_1")))
+__weak int target_1(int depth)
+{
+ if (depth < 1)
+ return 1 + target_1(depth + 1);
+ else
+ return target_2();
+}
+
+extern const void *__start_uprobe__target_1;
+extern const void *__stop_uprobe__target_1;
+
+extern const void *__start_uretprobe_stack_sec;
+extern const void *__stop_uretprobe_stack_sec;
+
+struct range {
+ long start;
+ long stop;
+};
+
+static struct range targets[] = {
+ {}, /* we want target_1 to map to target[1], so need 1-based indexing */
+ { (long)&__start_uprobe__target_1, (long)&__stop_uprobe__target_1 },
+ { (long)&__start_uprobe__target_2, (long)&__stop_uprobe__target_2 },
+ { (long)&__start_uprobe__target_3, (long)&__stop_uprobe__target_3 },
+ { (long)&__start_uprobe__target_4, (long)&__stop_uprobe__target_4 },
+};
+
+static struct range caller = {
+ (long)&__start_uretprobe_stack_sec,
+ (long)&__stop_uretprobe_stack_sec,
+};
+
+static void validate_stack(__u64 *ips, int stack_len, int cnt, ...)
+{
+ int i, j;
+ va_list args;
+
+ if (!ASSERT_GT(stack_len, 0, "stack_len"))
+ return;
+
+ stack_len /= 8;
+
+ /* check if we have enough entries to satisfy test expectations */
+ if (!ASSERT_GE(stack_len, cnt, "stack_len2"))
+
[PATCH v2 3/4] perf,x86: avoid missing caller address in stack traces captured in uprobe
When tracing user functions with uprobe functionality, it's common to
install the probe (e.g., a BPF program) at the first instruction of the
function. This is often going to be `push %rbp` instruction in function
preamble, which means that within that function frame pointer hasn't
been established yet. This leads to consistently missing an actual
caller of the traced function, because perf_callchain_user() only
records current IP (capturing traced function) and then following frame
pointer chain (which would be caller's frame, containing the address of
caller's caller).
So when we have target_1 -> target_2 -> target_3 call chain and we are
tracing an entry to target_3, captured stack trace will report
target_1 -> target_3 call chain, which is wrong and confusing.
This patch proposes a x86-64-specific heuristic to detect `push %rbp`
instruction being traced. If that's the case, with the assumption that
applicatoin is compiled with frame pointers, this instruction would be
a strong indicator that this is the entry to the function. In that case,
return address is still pointed to by %rsp, so we fetch it and add to
stack trace before proceeding to unwind the rest using frame
pointer-based logic.
Signed-off-by: Andrii Nakryiko
---
arch/x86/events/core.c | 20
include/linux/uprobes.h | 2 ++
kernel/events/uprobes.c | 2 ++
3 files changed, 24 insertions(+)
diff --git a/arch/x86/events/core.c b/arch/x86/events/core.c
index 5b0dd07b1ef1..82d5570b58ff 100644
--- a/arch/x86/events/core.c
+++ b/arch/x86/events/core.c
@@ -2884,6 +2884,26 @@ perf_callchain_user(struct perf_callchain_entry_ctx
*entry, struct pt_regs *regs
return;
pagefault_disable();
+
+#ifdef CONFIG_UPROBES
+ /*
+* If we are called from uprobe handler, and we are indeed at the very
+* entry to user function (which is normally a `push %rbp` instruction,
+* under assumption of application being compiled with frame pointers),
+* we should read return address from *regs->sp before proceeding
+* to follow frame pointers, otherwise we'll skip immediate caller
+* as %rbp is not yet setup.
+*/
+ if (current->utask) {
+ struct arch_uprobe *auprobe = current->utask->auprobe;
+ u64 ret_addr;
+
+ if (auprobe && auprobe->insn[0] == 0x55 /* push %rbp */ &&
+ !__get_user(ret_addr, (const u64 __user *)regs->sp))
+ perf_callchain_store(entry, ret_addr);
+ }
+#endif
+
while (entry->nr < entry->max_stack) {
if (!valid_user_frame(fp, sizeof(frame)))
break;
diff --git a/include/linux/uprobes.h b/include/linux/uprobes.h
index 0c57eec85339..7b785cd30d86 100644
--- a/include/linux/uprobes.h
+++ b/include/linux/uprobes.h
@@ -76,6 +76,8 @@ struct uprobe_task {
struct uprobe *active_uprobe;
unsigned long xol_vaddr;
+ struct arch_uprobe *auprobe;
+
struct return_instance *return_instances;
unsigned intdepth;
};
diff --git a/kernel/events/uprobes.c b/kernel/events/uprobes.c
index 1c99380dc89d..504693845187 100644
--- a/kernel/events/uprobes.c
+++ b/kernel/events/uprobes.c
@@ -2072,6 +2072,7 @@ static void handler_chain(struct uprobe *uprobe, struct
pt_regs *regs)
bool need_prep = false; /* prepare return uprobe, when needed */
down_read(>register_rwsem);
+ current->utask->auprobe = >arch;
for (uc = uprobe->consumers; uc; uc = uc->next) {
int rc = 0;
@@ -2086,6 +2087,7 @@ static void handler_chain(struct uprobe *uprobe, struct
pt_regs *regs)
remove &= rc;
}
+ current->utask->auprobe = NULL;
if (need_prep && !remove)
prepare_uretprobe(uprobe, regs); /* put bp at return */
--
2.43.0
[PATCH v2 2/4] perf,uprobes: fix user stack traces in the presence of pending uretprobes
When kernel has pending uretprobes installed, it hijacks original user
function return address on the stack with a uretprobe trampoline
address. There could be multiple such pending uretprobes (either on
different user functions or on the same recursive one) at any given
time within the same task.
This approach interferes with the user stack trace capture logic, which
would report suprising addresses (like 0x7fffe000) that correspond
to a special "[uprobes]" section that kernel installs in the target
process address space for uretprobe trampoline code, while logically it
should be an address somewhere within the calling function of another
traced user function.
This is easy to correct for, though. Uprobes subsystem keeps track of
pending uretprobes and records original return addresses. This patch is
using this to do a post-processing step and restore each trampoline
address entries with correct original return address. This is done only
if there are pending uretprobes for current task.
This is a similar approach to what fprobe/kretprobe infrastructure is
doing when capturing kernel stack traces in the presence of pending
return probes.
Reported-by: Riham Selim
Signed-off-by: Andrii Nakryiko
---
kernel/events/callchain.c | 43 ++-
kernel/events/uprobes.c | 9
2 files changed, 51 insertions(+), 1 deletion(-)
diff --git a/kernel/events/callchain.c b/kernel/events/callchain.c
index 1273be84392c..b17e3323f7f6 100644
--- a/kernel/events/callchain.c
+++ b/kernel/events/callchain.c
@@ -11,6 +11,7 @@
#include
#include
#include
+#include
#include "internal.h"
@@ -176,13 +177,51 @@ put_callchain_entry(int rctx)
put_recursion_context(this_cpu_ptr(callchain_recursion), rctx);
}
+static void fixup_uretprobe_trampoline_entries(struct perf_callchain_entry
*entry,
+ int start_entry_idx)
+{
+#ifdef CONFIG_UPROBES
+ struct uprobe_task *utask = current->utask;
+ struct return_instance *ri;
+ __u64 *cur_ip, *last_ip, tramp_addr;
+
+ if (likely(!utask || !utask->return_instances))
+ return;
+
+ cur_ip = >ip[start_entry_idx];
+ last_ip = >ip[entry->nr - 1];
+ ri = utask->return_instances;
+ tramp_addr = uprobe_get_trampoline_vaddr();
+
+ /*
+* If there are pending uretprobes for the current thread, they are
+* recorded in a list inside utask->return_instances; each such
+* pending uretprobe replaces traced user function's return address on
+* the stack, so when stack trace is captured, instead of seeing
+* actual function's return address, we'll have one or many uretprobe
+* trampoline addresses in the stack trace, which are not helpful and
+* misleading to users.
+* So here we go over the pending list of uretprobes, and each
+* encountered trampoline address is replaced with actual return
+* address.
+*/
+ while (ri && cur_ip <= last_ip) {
+ if (*cur_ip == tramp_addr) {
+ *cur_ip = ri->orig_ret_vaddr;
+ ri = ri->next;
+ }
+ cur_ip++;
+ }
+#endif
+}
+
struct perf_callchain_entry *
get_perf_callchain(struct pt_regs *regs, u32 init_nr, bool kernel, bool user,
u32 max_stack, bool crosstask, bool add_mark)
{
struct perf_callchain_entry *entry;
struct perf_callchain_entry_ctx ctx;
- int rctx;
+ int rctx, start_entry_idx;
entry = get_callchain_entry();
if (!entry)
@@ -215,7 +254,9 @@ get_perf_callchain(struct pt_regs *regs, u32 init_nr, bool
kernel, bool user,
if (add_mark)
perf_callchain_store_context(,
PERF_CONTEXT_USER);
+ start_entry_idx = entry->nr;
perf_callchain_user(, regs);
+ fixup_uretprobe_trampoline_entries(entry,
start_entry_idx);
}
}
diff --git a/kernel/events/uprobes.c b/kernel/events/uprobes.c
index d60d24f0f2f4..1c99380dc89d 100644
--- a/kernel/events/uprobes.c
+++ b/kernel/events/uprobes.c
@@ -2149,6 +2149,15 @@ static void handle_trampoline(struct pt_regs *regs)
instruction_pointer_set(regs, ri->orig_ret_vaddr);
do {
+ /* pop current instance from the stack of pending
return instances,
+* as it's not pending anymore: we just fixed up
original
+* instruction pointer in regs and are about to call
handlers;
+* this allows fixup_uretprobe_trampoline_entries() to
properly fix up
+* captured stack traces from uretprobe handlers, in
which pending
+* trampoline addresses on the stack are replaced with
correct
+* original
[PATCH v2 1/4] uprobes: rename get_trampoline_vaddr() and make it global
This helper is needed in another file, so make it a bit more uniquely
named and expose it internally.
Signed-off-by: Andrii Nakryiko
---
include/linux/uprobes.h | 1 +
kernel/events/uprobes.c | 6 +++---
2 files changed, 4 insertions(+), 3 deletions(-)
diff --git a/include/linux/uprobes.h b/include/linux/uprobes.h
index f46e0ca0169c..0c57eec85339 100644
--- a/include/linux/uprobes.h
+++ b/include/linux/uprobes.h
@@ -138,6 +138,7 @@ extern bool arch_uretprobe_is_alive(struct return_instance
*ret, enum rp_check c
extern bool arch_uprobe_ignore(struct arch_uprobe *aup, struct pt_regs *regs);
extern void arch_uprobe_copy_ixol(struct page *page, unsigned long vaddr,
void *src, unsigned long len);
+extern unsigned long uprobe_get_trampoline_vaddr(void);
#else /* !CONFIG_UPROBES */
struct uprobes_state {
};
diff --git a/kernel/events/uprobes.c b/kernel/events/uprobes.c
index 8ae0eefc3a34..d60d24f0f2f4 100644
--- a/kernel/events/uprobes.c
+++ b/kernel/events/uprobes.c
@@ -1827,7 +1827,7 @@ void uprobe_copy_process(struct task_struct *t, unsigned
long flags)
*
* Returns -1 in case the xol_area is not allocated.
*/
-static unsigned long get_trampoline_vaddr(void)
+unsigned long uprobe_get_trampoline_vaddr(void)
{
struct xol_area *area;
unsigned long trampoline_vaddr = -1;
@@ -1878,7 +1878,7 @@ static void prepare_uretprobe(struct uprobe *uprobe,
struct pt_regs *regs)
if (!ri)
return;
- trampoline_vaddr = get_trampoline_vaddr();
+ trampoline_vaddr = uprobe_get_trampoline_vaddr();
orig_ret_vaddr = arch_uretprobe_hijack_return_addr(trampoline_vaddr,
regs);
if (orig_ret_vaddr == -1)
goto fail;
@@ -2187,7 +2187,7 @@ static void handle_swbp(struct pt_regs *regs)
int is_swbp;
bp_vaddr = uprobe_get_swbp_addr(regs);
- if (bp_vaddr == get_trampoline_vaddr())
+ if (bp_vaddr == uprobe_get_trampoline_vaddr())
return handle_trampoline(regs);
uprobe = find_active_uprobe(bp_vaddr, _swbp);
--
2.43.0
[PATCH v2 0/4] Fix user stack traces captured from uprobes
This patch set reports two issues with captured stack traces. First issue, fixed in patch #2, deals with fixing up uretprobe trampoline addresses in captured stack trace. This issue happens when there are pending return probes, for which kernel hijacks some of the return addresses on user stacks. The code is matching those special uretprobe trampoline addresses with the list of pending return probe instances and replaces them with actual return addresses. This is the same fixup logic that fprobe/kretprobe has for kernel stack traces. Second issue, which patch #3 is fixing with the help of heuristic, is having to do with capturing user stack traces in entry uprobes. At the very entrance to user function, frame pointer in rbp register is not yet setup, so actual caller return address is still pointed to by rsp. Patch is using a simple heuristic, looking for `push %rbp` instruction, to fetch this extra direct caller return address, before proceeding to unwind the stack using rbp. Patch #4 adds tests into BPF selftests, that validate that captured stack traces at various points is what we expect to get. This patch, while being BPF selftests, is isolated from any other BPF selftests changes and can go in through non-BPF tree without the risk of merge conflicts. Patches are based on latest linux-trace/probes/for-next. v1->v2: - fixed GCC aggressively inlining test_uretprobe_stack() function (BPF CI); - fixed comments (Peter). Andrii Nakryiko (4): uprobes: rename get_trampoline_vaddr() and make it global perf,uprobes: fix user stack traces in the presence of pending uretprobes perf,x86: avoid missing caller address in stack traces captured in uprobe selftests/bpf: add test validating uprobe/uretprobe stack traces arch/x86/events/core.c| 20 ++ include/linux/uprobes.h | 3 + kernel/events/callchain.c | 43 +++- kernel/events/uprobes.c | 17 +- .../bpf/prog_tests/uretprobe_stack.c | 186 ++ .../selftests/bpf/progs/uretprobe_stack.c | 96 + 6 files changed, 361 insertions(+), 4 deletions(-) create mode 100644 tools/testing/selftests/bpf/prog_tests/uretprobe_stack.c create mode 100644 tools/testing/selftests/bpf/progs/uretprobe_stack.c -- 2.43.0
Re: [PATCH] uprobes: prevent mutex_lock() under rcu_read_lock()
On Mon, May 20, 2024 at 10:30:17PM -0700, Andrii Nakryiko wrote:
> Recent changes made uprobe_cpu_buffer preparation lazy, and moved it
> deeper into __uprobe_trace_func(). This is problematic because
> __uprobe_trace_func() is called inside rcu_read_lock()/rcu_read_unlock()
> block, which then calls prepare_uprobe_buffer() -> uprobe_buffer_get() ->
> mutex_lock(>mutex), leading to a splat about using mutex under
> non-sleepable RCU:
>
> BUG: sleeping function called from invalid context at
> kernel/locking/mutex.c:585
>in_atomic(): 0, irqs_disabled(): 0, non_block: 0, pid: 98231, name:
> stress-ng-sigq
>preempt_count: 0, expected: 0
>RCU nest depth: 1, expected: 0
>...
>Call Trace:
>
> dump_stack_lvl+0x3d/0xe0
> __might_resched+0x24c/0x270
> ? prepare_uprobe_buffer+0xd5/0x1d0
> __mutex_lock+0x41/0x820
> ? ___perf_sw_event+0x206/0x290
> ? __perf_event_task_sched_in+0x54/0x660
> ? __perf_event_task_sched_in+0x54/0x660
> prepare_uprobe_buffer+0xd5/0x1d0
> __uprobe_trace_func+0x4a/0x140
> uprobe_dispatcher+0x135/0x280
> ? uprobe_dispatcher+0x94/0x280
> uprobe_notify_resume+0x650/0xec0
> ? atomic_notifier_call_chain+0x21/0x110
> ? atomic_notifier_call_chain+0xf8/0x110
> irqentry_exit_to_user_mode+0xe2/0x1e0
> asm_exc_int3+0x35/0x40
>RIP: 0033:0x7f7e1d4da390
>Code: 33 04 00 0f 1f 80 00 00 00 00 f3 0f 1e fa b9 01 00 00 00 e9 b2 fc ff
> ff 66 90 f3 0f 1e fa 31 c9 e9 a5 fc ff ff 0f 1f 44 00 00 0f 1e fa b8 27
> 00 00 00 0f 05 c3 0f 1f 40 00 f3 0f 1e fa b8 6e
>RSP: 002b:7ffd2abc3608 EFLAGS: 0246
>RAX: RBX: 76d325f1 RCX:
>RDX: 76d325f1 RSI: 000a RDI: 7ffd2abc3690
>RBP: 000a R08: 00017fb7 R09: 00017fb7
>R10: 00017fb7 R11: 0246 R12: 00017ff2
>R13: 7ffd2abc3610 R14: R15: 7ffd2abc3780
>
>
> Luckily, it's easy to fix by moving prepare_uprobe_buffer() to be called
> slightly earlier: into uprobe_trace_func() and uretprobe_trace_func(), outside
> of RCU locked section. This still keeps this buffer preparation lazy and helps
> avoid the overhead when it's not needed. E.g., if there is only BPF uprobe
> handler installed on a given uprobe, buffer won't be initialized.
>
> Note, the other user of prepare_uprobe_buffer(), __uprobe_perf_func(), is not
> affected, as it doesn't prepare buffer under RCU read lock.
>
> Fixes: 1b8f85defbc8 ("uprobes: prepare uprobe args buffer lazily")
> Reported-by: Breno Leitao
> Signed-off-by: Andrii Nakryiko
Tested-by: Breno Leitao
Re: [PATCH] uprobes: prevent mutex_lock() under rcu_read_lock()
On 05/20, Andrii Nakryiko wrote:
>
> Fixes: 1b8f85defbc8 ("uprobes: prepare uprobe args buffer lazily")
> Reported-by: Breno Leitao
> Signed-off-by: Andrii Nakryiko
> ---
> kernel/trace/trace_uprobe.c | 14 +-
> 1 file changed, 9 insertions(+), 5 deletions(-)
Reviewed-by: Oleg Nesterov
