Re: kprobes broken since 0d00449c7a28 ("x86: Replace ist_enter() with nmi_enter()")

[Josh Poimboeuf <jpoimboe@redhat.com>]

On Thu, Jan 28, 2021 at 06:45:56PM +0200, Nikolay Borisov wrote:
> On 28.01.21 г. 18:12 ч., Nikolay Borisov wrote:
> > On 28.01.21 г. 5:38 ч., Masami Hiramatsu wrote:
> >> Hi,
> > 
> > <snip>
> >>
> >> Alexei, could you tell me what is the concerning situation for bpf?
> > 
> > Another data point masami is that this affects bpf kprobes which are
> > entered via int3, alternatively if the kprobe is entered via
> > kprobe_ftrace_handler it works as expected. I haven't been able to
> > determine why a particular bpf probe won't use ftrace's infrastructure
> > if it's put at the beginning of the function.  An alternative call chain
> > is :
> > 
> >  => __ftrace_trace_stack
> >  => trace_call_bpf
> >  => kprobe_perf_func
> >  => kprobe_ftrace_handler
> >  => 0xffffffffc095d0c8
> >  => btrfs_validate_metadata_buffer
> >  => end_bio_extent_readpage
> >  => end_workqueue_fn
> >  => btrfs_work_helper
> >  => process_one_work
> >  => worker_thread
> >  => kthread
> >  => ret_from_fork
> > 
> >>
> 
> I have a working theory why I'm seeing this. My kernel (broken) was
> compiled with retpolines off and with the gcc that comes with ubuntu
> (both 9 and 10:
> gcc (Ubuntu 9.3.0-17ubuntu1~20.04) 9.3.0
> gcc-10 (Ubuntu 10.2.0-5ubuntu1~20.04) 10.2.0
> )
> 
> this results in CFI being enabled so functions look like:
> 0xffffffff81493890 <+0>: endbr64
> 0xffffffff81493894 <+4>: callq  0xffffffff8104d820 <__fentry__>
> 
> i.e fentry's thunk is not the first instruction on the function hence
> it's not going through the optimized ftrace handler. Instead it's using
> int3 which is broken as ascertained.
> 
> After testing with my testcase I confirm that with cfi off and
> __fentry__ being the first entry bpf starts working. And indeed, even
> with CFI turned on if I use a probe like :
> 
> bpftrace -e 'kprobe:btrfs_sync_file+4 {printf("kprobe: %s\n",
> kstack());}' &>bpf-output &
> 
> 
> it would be placed on the __fentry__ (and not endbr64) hence it works.
> So perhaps a workaround outside of bpf could essentially detect this
> scenario and adjust the probe to be on the __fentry__ and not preceding
> instruction if it's detected to be endbr64 ?

For now (and the foreseeable future), CET isn't enabled in the kernel.
So that endbr64 shouldn't be there in the first place.  I can make a
proper patch in a bit.

diff --git a/Makefile b/Makefile
index e0af7a4a5598..5ccc4cdf1fb5 100644
--- a/Makefile
+++ b/Makefile
@@ -948,11 +948,8 @@ KBUILD_CFLAGS   += $(call cc-option,-Werror=designated-init)
 # change __FILE__ to the relative path from the srctree
 KBUILD_CPPFLAGS += $(call cc-option,-fmacro-prefix-map=$(srctree)/=)
 
-# ensure -fcf-protection is disabled when using retpoline as it is
-# incompatible with -mindirect-branch=thunk-extern
-ifdef CONFIG_RETPOLINE
+# Intel CET isn't enabled in the kernel
 KBUILD_CFLAGS += $(call cc-option,-fcf-protection=none)
-endif
 
 # include additional Makefiles when needed
 include-y			:= scripts/Makefile.extrawarn