From: madvenka@linux.microsoft.com
To: broonie@kernel.org, mark.rutland@arm.com, jpoimboe@redhat.com,
jthierry@redhat.com, catalin.marinas@arm.com, will@kernel.org,
linux-arm-kernel@lists.infradead.org,
live-patching@vger.kernel.org, linux-kernel@vger.kernel.org,
madvenka@linux.microsoft.com
Subject: [RFC PATCH v2 7/8] arm64: Detect kretprobed functions in stack trace
Date: Mon, 15 Mar 2021 11:57:59 -0500 [thread overview]
Message-ID: <20210315165800.5948-8-madvenka@linux.microsoft.com> (raw)
In-Reply-To: <20210315165800.5948-1-madvenka@linux.microsoft.com>
From: "Madhavan T. Venkataraman" <madvenka@linux.microsoft.com>
When a kretprobe is active for a function, the function's return address
in its stack frame is modified to point to the kretprobe trampoline. When
the function returns, the frame is popped and control is transferred
to the trampoline. The trampoline eventually returns to the original return
address.
If a stack walk is done within the function (or any functions that get
called from there), the stack trace will only show the trampoline and the
not the original caller. Detect this and mark the stack trace as unreliable.
Also, if the trampoline and the functions it calls do a stack trace,
that stack trace will also have the same problem. Detect this as well.
This is done by looking up the symbol table entry for the trampoline
and checking if the return PC in a frame falls anywhere in the
trampoline function.
Signed-off-by: Madhavan T. Venkataraman <madvenka@linux.microsoft.com>
---
arch/arm64/kernel/stacktrace.c | 43 ++++++++++++++++++++++++++++++++++
1 file changed, 43 insertions(+)
diff --git a/arch/arm64/kernel/stacktrace.c b/arch/arm64/kernel/stacktrace.c
index 358aae3906d7..752b77f11c61 100644
--- a/arch/arm64/kernel/stacktrace.c
+++ b/arch/arm64/kernel/stacktrace.c
@@ -18,6 +18,26 @@
#include <asm/stack_pointer.h>
#include <asm/stacktrace.h>
+#ifdef CONFIG_KRETPROBES
+static bool kretprobe_detected(struct stackframe *frame)
+{
+ static char kretprobe_name[KSYM_NAME_LEN];
+ static unsigned long kretprobe_pc, kretprobe_end_pc;
+ unsigned long pc, offset, size;
+
+ if (!kretprobe_pc) {
+ pc = (unsigned long) kretprobe_trampoline;
+ if (!kallsyms_lookup(pc, &size, &offset, NULL, kretprobe_name))
+ return false;
+
+ kretprobe_pc = pc - offset;
+ kretprobe_end_pc = kretprobe_pc + size;
+ }
+
+ return frame->pc >= kretprobe_pc && frame->pc < kretprobe_end_pc;
+}
+#endif
+
static void check_if_reliable(unsigned long fp, struct stackframe *frame,
struct stack_info *info)
{
@@ -111,6 +131,29 @@ static void check_if_reliable(unsigned long fp, struct stackframe *frame,
frame->reliable = false;
return;
}
+
+#ifdef CONFIG_KRETPROBES
+ /*
+ * The return address of a function that has an active kretprobe
+ * is modified in the stack frame to point to a trampoline. So,
+ * the original return address is not available on the stack.
+ *
+ * A stack trace taken while executing the function (and its
+ * descendants) will not show the original caller. So, mark the
+ * stack trace as unreliable if the trampoline shows up in the
+ * stack trace. (Obtaining the original return address from
+ * task->kretprobe_instances seems problematic and not worth the
+ * effort).
+ *
+ * The stack trace taken while inside the trampoline and functions
+ * called by the trampoline have the same problem as above. This
+ * is also covered by kretprobe_detected() using a range check.
+ */
+ if (kretprobe_detected(frame)) {
+ frame->reliable = false;
+ return;
+ }
+#endif
}
/*
--
2.25.1
_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
next prev parent reply other threads:[~2021-03-15 17:02 UTC|newest]
Thread overview: 55+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <5997dfe8d261a3a543667b83c902883c1e4bd270>
2021-03-15 16:57 ` [RFC PATCH v2 0/8] arm64: Implement reliable stack trace madvenka
2021-03-15 16:57 ` [RFC PATCH v2 1/8] arm64: Implement stack trace termination record madvenka
2021-03-18 15:09 ` Mark Brown
2021-03-18 20:26 ` Madhavan T. Venkataraman
2021-03-19 12:30 ` Mark Brown
2021-03-19 14:29 ` Madhavan T. Venkataraman
2021-03-19 18:19 ` Madhavan T. Venkataraman
2021-03-19 22:03 ` Madhavan T. Venkataraman
2021-03-23 10:24 ` Mark Rutland
2021-03-23 12:39 ` Madhavan T. Venkataraman
2021-03-15 16:57 ` [RFC PATCH v2 2/8] arm64: Implement frame types madvenka
2021-03-18 17:40 ` Mark Brown
2021-03-18 22:22 ` Madhavan T. Venkataraman
2021-03-19 13:22 ` Mark Brown
2021-03-19 14:40 ` Madhavan T. Venkataraman
2021-03-19 15:02 ` Madhavan T. Venkataraman
2021-03-19 16:20 ` Mark Brown
2021-03-19 16:27 ` Madhavan T. Venkataraman
2021-03-23 10:34 ` Mark Rutland
2021-03-15 16:57 ` [RFC PATCH v2 3/8] arm64: Terminate the stack trace at TASK_FRAME and EL0_FRAME madvenka
2021-03-18 18:26 ` Mark Brown
2021-03-18 20:29 ` Madhavan T. Venkataraman
2021-03-23 10:36 ` Mark Rutland
2021-03-23 12:40 ` Madhavan T. Venkataraman
2021-03-15 16:57 ` [RFC PATCH v2 4/8] arm64: Detect an EL1 exception frame and mark a stack trace unreliable madvenka
2021-03-23 10:42 ` Mark Rutland
2021-03-23 12:46 ` Madhavan T. Venkataraman
2021-03-23 13:04 ` Mark Rutland
2021-03-23 13:31 ` Madhavan T. Venkataraman
2021-03-23 14:33 ` Mark Rutland
2021-03-23 15:22 ` Madhavan T. Venkataraman
2021-03-15 16:57 ` [RFC PATCH v2 5/8] arm64: Detect an FTRACE " madvenka
2021-03-23 10:51 ` Mark Rutland
2021-03-23 12:56 ` Madhavan T. Venkataraman
2021-03-23 13:36 ` Mark Rutland
2021-03-23 13:38 ` Madhavan T. Venkataraman
2021-03-23 14:15 ` Madhavan T. Venkataraman
2021-03-23 14:57 ` Mark Rutland
2021-03-23 15:26 ` Madhavan T. Venkataraman
2021-03-23 16:20 ` Madhavan T. Venkataraman
2021-03-23 17:02 ` Mark Rutland
2021-03-23 17:23 ` Madhavan T. Venkataraman
2021-03-23 17:27 ` Madhavan T. Venkataraman
2021-03-23 18:27 ` Mark Brown
2021-03-23 20:23 ` Madhavan T. Venkataraman
2021-03-23 18:30 ` Mark Rutland
2021-03-23 20:24 ` Madhavan T. Venkataraman
2021-03-23 21:04 ` Madhavan T. Venkataraman
2021-03-23 16:48 ` Mark Rutland
2021-03-23 16:53 ` Madhavan T. Venkataraman
2021-03-23 17:09 ` Mark Rutland
2021-03-15 16:57 ` [RFC PATCH v2 6/8] arm64: Check the return PC of every stack frame madvenka
2021-03-15 16:57 ` madvenka [this message]
2021-03-15 16:58 ` [RFC PATCH v2 8/8] arm64: Implement arch_stack_walk_reliable() madvenka
2021-03-15 19:01 ` [RFC PATCH v2 0/8] arm64: Implement reliable stack trace Madhavan T. Venkataraman
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20210315165800.5948-8-madvenka@linux.microsoft.com \
--to=madvenka@linux.microsoft.com \
--cc=broonie@kernel.org \
--cc=catalin.marinas@arm.com \
--cc=jpoimboe@redhat.com \
--cc=jthierry@redhat.com \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=linux-kernel@vger.kernel.org \
--cc=live-patching@vger.kernel.org \
--cc=mark.rutland@arm.com \
--cc=will@kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).