linux-kernel - [PATCH v2] perf,x86: avoid missing caller address in stack traces captured in uprobe

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [thread-next>] [day] [month] [year] [list]

Message-ID: <20240702171858.187562-1-andrii@kernel.org>
Date: Tue,  2 Jul 2024 10:18:58 -0700
From: Andrii Nakryiko <andrii@...nel.org>
To: linux-trace-kernel@...r.kernel.org,
	peterz@...radead.org,
	rostedt@...dmis.org,
	mhiramat@...nel.org
Cc: x86@...nel.org,
	mingo@...hat.com,
	tglx@...utronix.de,
	jpoimboe@...hat.com,
	linux-kernel@...r.kernel.org,
	bpf@...r.kernel.org,
	rihams@...com,
	linux-perf-users@...r.kernel.org,
	Andrii Nakryiko <andrii@...nel.org>
Subject: [PATCH v2] 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`
(`push %ebp` on 32-bit architecture) instruction being traced. Given
entire kernel implementation of user space stack trace capturing works
under assumption that user space code was compiled with frame pointer
register (%rbp/%ebp) preservation, it seems pretty reasonable to use
this instruction as a strong indicator that this is the entry to the
function. In that case, return address is still pointed to by %rsp/%esp,
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 <andrii@...nel.org>
---
v1->v2:
  - use native unsigned long for ret_addr (Peter);
  - add same logic for compat logic in perf_callchain_user32 (Peter).

 arch/x86/events/core.c  | 33 +++++++++++++++++++++++++++++++++
 include/linux/uprobes.h |  2 ++
 kernel/events/uprobes.c |  2 ++
 3 files changed, 37 insertions(+)

diff --git a/arch/x86/events/core.c b/arch/x86/events/core.c
index 5b0dd07b1ef1..60821c1ff2f3 100644
--- a/arch/x86/events/core.c
+++ b/arch/x86/events/core.c
@@ -2833,6 +2833,19 @@ perf_callchain_user32(struct pt_regs *regs, struct perf_callchain_entry_ctx *ent
 
 	fp = compat_ptr(ss_base + regs->bp);
 	pagefault_disable();
+
+#ifdef CONFIG_UPROBES
+	/* see perf_callchain_user() below for why we do this */
+	if (current->utask) {
+		struct arch_uprobe *auprobe = current->utask->auprobe;
+		u32 ret_addr;
+
+		if (auprobe && auprobe->insn[0] == 0x55 /* push %ebp */ &&
+		    !__get_user(ret_addr, (const u32 __user *)regs->sp))
+			perf_callchain_store(entry, ret_addr);
+	}
+#endif
+
 	while (entry->nr < entry->max_stack) {
 		if (!valid_user_frame(fp, sizeof(frame)))
 			break;
@@ -2884,6 +2897,26 @@ 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;
+		unsigned long ret_addr;
+
+		if (auprobe && auprobe->insn[0] == 0x55 /* push %rbp/%ebp */ &&
+		    !__get_user(ret_addr, (const unsigned long __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 b503fafb7fb3..a270a5892ab4 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 int			depth;
 };
diff --git a/kernel/events/uprobes.c b/kernel/events/uprobes.c
index 99be2adedbc0..6e22e4d80f1e 100644
--- a/kernel/events/uprobes.c
+++ b/kernel/events/uprobes.c
@@ -2082,6 +2082,7 @@ static void handler_chain(struct uprobe *uprobe, struct pt_regs *regs)
 	bool need_prep = false; /* prepare return uprobe, when needed */
 
 	down_read(&uprobe->register_rwsem);
+	current->utask->auprobe = &uprobe->arch;
 	for (uc = uprobe->consumers; uc; uc = uc->next) {
 		int rc = 0;
 
@@ -2096,6 +2097,7 @@ 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