Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Mon, 16 May 2022 20:27:40 +0200
From:   Eugene Syromiatnikov <esyr@redhat.com>
To:     Jiri Olsa <jolsa@kernel.org>,
        Masami Hiramatsu <mhiramat@kernel.org>,
        Steven Rostedt <rostedt@goodmis.org>,
        Ingo Molnar <mingo@redhat.com>,
        Alexei Starovoitov <ast@kernel.org>,
        Daniel Borkmann <daniel@iogearbox.net>
Cc:     Andrii Nakryiko <andrii@kernel.org>,
        Martin KaFai Lau <kafai@fb.com>,
        Song Liu <songliubraving@fb.com>, Yonghong Song <yhs@fb.com>,
        John Fastabend <john.fastabend@gmail.com>,
        KP Singh <kpsingh@kernel.org>, netdev@vger.kernel.org,
        bpf@vger.kernel.org, linux-kernel@vger.kernel.org,
        Shuah Khan <shuah@kernel.org>, linux-kselftest@vger.kernel.org
Subject: [PATCH bpf 3/4] bpf_trace: handle compat in kprobe_multi_resolve_syms
Message-ID: <20220516182740.GA29979@asgard.redhat.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
User-Agent: Mutt/1.5.23 (2014-03-12)
Precedence: bulk

For compat processes, userspace pointer size is different.  Since the
copied array is iterated anyway, the simplest fix seems to be copy the
user-supplied array as-is and the iterate as an array of native or
compat pointers, depending on the in_compat_syscall() value.

Fixes: 0dcac272540613d4 ("bpf: Add multi kprobe link")
Signed-off-by: Eugene Syromiatnikov <esyr@redhat.com>
---
 kernel/trace/bpf_trace.c | 20 ++++++++++++++------
 1 file changed, 14 insertions(+), 6 deletions(-)

diff --git a/kernel/trace/bpf_trace.c b/kernel/trace/bpf_trace.c
index d228440..5b0cf54 100644
--- a/kernel/trace/bpf_trace.c
+++ b/kernel/trace/bpf_trace.c
@@ -2352,16 +2352,21 @@ static int
 kprobe_multi_resolve_syms(const void __user *usyms, u32 cnt,
 			  unsigned long *addrs)
 {
-	unsigned long addr, size;
+	unsigned long addr;
+	size_t sym_size;
+	u32 size, elem_size;
 	const char __user **syms;
+	compat_uptr_t __user *compat_syms;
 	int err = -ENOMEM;
 	unsigned int i;
 	char *func;
 
-	if (check_mul_overflow(cnt, sizeof(*syms), &size))
+	elem_size = in_compat_syscall() ? sizeof(*compat_syms) : sizeof(*syms);
+	if (check_mul_overflow(cnt, elem_size, &size))
 		return -EOVERFLOW;
-	size = cnt * sizeof(*syms);
+	size = cnt * elem_size;
 	syms = kvzalloc(size, GFP_KERNEL);
+	compat_syms = (void *)syms;
 	if (!syms)
 		return -ENOMEM;
 
@@ -2375,7 +2380,10 @@ kprobe_multi_resolve_syms(const void __user *usyms, u32 cnt,
 	}
 
 	for (i = 0; i < cnt; i++) {
-		err = strncpy_from_user(func, syms[i], KSYM_NAME_LEN);
+		const char __user *ufunc = in_compat_syscall()
+					? (char __user *)(uintptr_t)compat_syms[i]
+					: syms[i];
+		err = strncpy_from_user(func, ufunc, KSYM_NAME_LEN);
 		if (err == KSYM_NAME_LEN)
 			err = -E2BIG;
 		if (err < 0)
@@ -2384,9 +2392,9 @@ kprobe_multi_resolve_syms(const void __user *usyms, u32 cnt,
 		addr = kallsyms_lookup_name(func);
 		if (!addr)
 			goto error;
-		if (!kallsyms_lookup_size_offset(addr, &size, NULL))
+		if (!kallsyms_lookup_size_offset(addr, &sym_size, NULL))
 			goto error;
-		addr = ftrace_location_range(addr, addr + size - 1);
+		addr = ftrace_location_range(addr, addr + sym_size - 1);
 		if (!addr)
 			goto error;
 		addrs[i] = addr;
-- 
2.1.4