Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > linux.kernel > #1432057
| From | Petr Mladek <pmladek@suse.com> |
|---|---|
| Newsgroups | linux.kernel |
| Subject | [PATCH 1/4] ftrace/x86: Make sure to modify 5-bite instructions |
| Date | 2016-06-27 16:00 +0200 |
| Message-ID | <rOFeW-XV-33@gated-at.bofh.it> (permalink) |
| References | <rOFeV-XV-7@gated-at.bofh.it> |
| Organization | linux.* mail to news gateway |
The commit 8329e818f14926a604 ("ftrace/x86: Set ftrace_stub to weak to
prevent gcc from using short jumps to it") fixed a situation where we
replaced 2-byte with 5-byte jump instruction.
The original problem crashed the kernel with
# cd /sys/kernel/trace
# echo function_graph > current_tracer
# echo nop > current_tracer
[ 78.122055] invalid opcode: 0000 [#1] SMP
[ 78.125125] Modules linked in: x86_pkg_temp_thermal kvm_intel kvm irqbypass \
crc32c_intel pcspkr iwldvm iwlwifi [ 78.128241] CPU: 2 PID: 17 Comm: migration/2 \
Not tainted 4.6.0-rc4+ #36 [ 78.131310] Hardware name: LENOVO 4286A74/4286A74, \
BIOS 8DET56WW (1.26 ) 12/01/2011 [ 78.134369] task: ffff88040bec4240 ti: \
ffff88040bee4000 task.ti: ffff88040bee4000 [ 78.137412] RIP: \
0010:[<ffffffff818939a8>] [<ffffffff818939a8>] ftrace_stub+0x0/0x8 [ 78.140477] \
RSP: 0018:ffff88040bee7e48 EFLAGS: 00010246 [ 78.143489] RAX: ffff88040bec4240 \
RBX: ffffffff8107a7b0 RCX: 000000000000001f [ 78.146512] RDX: 0000000000000000 \
RSI: ffff88041e2929d8 RDI: ffff88040bee7e50 [ 78.149581] RBP: ffff88040bee7e80 \
R08: ffff88040bee4000 R09: 0000000000000000 [ 78.152647] R10: 00000000000318b7 \
R11: ffff8800d661f800 R12: ffff88040d8011b0 [ 78.155679] R13: ffffffff81e43620 \
R14: ffff88040bda8588 R15: ffffffff81e503e0 [ 78.158675] FS: \
0000000000000000(0000) GS:ffff88041e280000(0000) knlGS:0000000000000000 [ \
78.161699] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 78.164690] CR2: \
00007fadb22dde1d CR3: 00000000d5ce2000 CR4: 00000000000406e0 [ 78.167691] Stack:
[ 78.170658] ffffffff8110b3ee ffffffff8188de90 00000009161d55f6 00000012306fc2e4
[ 78.173710] 0000000000000000 ffff880400000000 ffff88040bec4240 ffff88040bee7ec8
[ 78.176783] ffffffff81893bbd 0000000000000000 ffff88040bec4240 ffffffff81893ba8
[ 78.179863] Call Trace:
[ 78.182853] [<ffffffff8110b3ee>] ? ftrace_return_to_handler+0x8e/0x100
[ 78.185909] [<ffffffff8188de90>] ? __schedule+0xae0/0xae0
[ 78.188941] [<ffffffff81893bbd>] return_to_handler+0x15/0x27
[ 78.192001] [<ffffffff81893ba8>] ? ftrace_graph_caller+0xa8/0xa8
[ 78.195091] [<ffffffff8107a6f0>] ? sort_range+0x30/0x30
[ 78.198138] [<ffffffff810778a9>] kthread+0xc9/0xe0
[ 78.201143] [<ffffffff81891a12>] ret_from_fork+0x22/0x40
[ 78.204138] [<ffffffff810777e0>] ? kthread_worker_fn+0x170/0x170
[ 78.207129] Code: 8b 44 24 48 48 8b 7c 24 70 48 8b 74 24 68 48 8b 54 24 60
48 8b 4c 24 58 48 8b 44 24 50 48 8b 6c 24 20 48 81 c4 d0
00 00 00 e9 fd <ff> ff ff 80 00 00 00 00 9c 55 ff 74 24 18
55 48 89 e5 ff 74 24
[ 78.213997] RIP [<ffffffff818939a8>] ftrace_stub+0x0/0x8
[ 78.217190] RSP <ffff88040bee7e48>
[ 78.220374] ---[ end trace 0af2f0d9f7301011 ]---
, see
https://lkml.kernel.org/g/1463147623-2575-1-git-send-email-namhyung@kernel.org
We already avoid similar crashes by paranoid checks, for example
in add_break() or ftrace_modify_code_direct(). These checks did
not help in the above situation because the old expected code
was read from the location that was modified. See the memcpy()
in update_ftrace_func().
Normally, the expected code is generated according to the expected
state of the modified location. update_ftrace_func() is called
in several situations. It is not easy to generate the expected code
a reasonable way. But the function seems to be called only when
we modify a call or a jump instruction. We could at least check
that it is the 5-byte variant.
Signed-off-by: Petr Mladek <pmladek@suse.com>
---
arch/x86/kernel/ftrace.c | 9 +++++++++
1 file changed, 9 insertions(+)
diff --git a/arch/x86/kernel/ftrace.c b/arch/x86/kernel/ftrace.c
index d036cfb4495d..42ea69d35dfd 100644
--- a/arch/x86/kernel/ftrace.c
+++ b/arch/x86/kernel/ftrace.c
@@ -235,6 +235,15 @@ static int update_ftrace_func(unsigned long ip, void *new)
memcpy(old, (void *)ip, MCOUNT_INSN_SIZE);
+ /*
+ * Make sure that we replace 5-byte instruction that
+ * is either a call or a jump.
+ */
+ if (old[0] != 0xe8 && old[0] != 0xe9) {
+ pr_warn("Expected e8 or e9, got %x\n", old[0]);
+ return -EINVAL;
+ }
+
ftrace_update_func = ip;
/* Make sure the breakpoints see the ftrace_update_func update */
smp_wmb();
--
1.8.5.6
Back to linux.kernel | Previous | Next — Previous in thread | Next in thread | Find similar | Unroll thread
[PATCH 0/4] ftrace: One more check on x86 and some small fixes Petr Mladek <pmladek@suse.com> - 2016-06-27 16:00 +0200
[PATCH 3/4] ftrace: Always destroy trampoline when shutting down the trace Petr Mladek <pmladek@suse.com> - 2016-06-27 16:00 +0200
[PATCH 2/4] ftrace/x86: Do not crash when reading wrong ftrace func Petr Mladek <pmladek@suse.com> - 2016-06-27 16:00 +0200
Re: [PATCH 2/4] ftrace/x86: Do not crash when reading wrong ftrace func Steven Rostedt <rostedt@goodmis.org> - 2016-06-27 16:50 +0200
Re: [PATCH 2/4] ftrace/x86: Do not crash when reading wrong ftrace func Petr Mladek <pmladek@suse.com> - 2016-06-28 12:10 +0200
[PATCH 1/4] ftrace/x86: Make sure to modify 5-bite instructions Petr Mladek <pmladek@suse.com> - 2016-06-27 16:00 +0200
[PATCH 4/4] ftrace: Fixup trace_selftest_ops() Petr Mladek <pmladek@suse.com> - 2016-06-27 16:00 +0200
Re: [PATCH 0/4] ftrace: One more check on x86 and some small fixes Namhyung Kim <namhyung@gmail.com> - 2016-06-28 07:30 +0200
Re: [PATCH 0/4] ftrace: One more check on x86 and some small fixes Steven Rostedt <rostedt@goodmis.org> - 2016-06-28 21:20 +0200
Re: [PATCH 0/4] ftrace: One more check on x86 and some small fixes Petr Mladek <pmladek@suse.com> - 2016-06-29 10:30 +0200
csiph-web