linux-kernel - Re: [PATCH resend 6/8] tracing/ftrace: Add might

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-prev] [day] [month] [year] [list]

Message-ID: <f6caee5c-9d4d-449c-b697-a0a27993bd33@efficios.com>
Date: Mon, 28 Oct 2024 15:02:26 -0400
From: Mathieu Desnoyers <mathieu.desnoyers@...icios.com>
To: Thomas Gleixner <tglx@...utronix.de>, Steven Rostedt
 <rostedt@...dmis.org>, Masami Hiramatsu <mhiramat@...nel.org>
Cc: linux-kernel@...r.kernel.org, Peter Zijlstra <peterz@...radead.org>,
 Alexei Starovoitov <ast@...nel.org>, Yonghong Song <yhs@...com>,
 "Paul E . McKenney" <paulmck@...nel.org>, Ingo Molnar <mingo@...hat.com>,
 Arnaldo Carvalho de Melo <acme@...nel.org>,
 Mark Rutland <mark.rutland@....com>,
 Alexander Shishkin <alexander.shishkin@...ux.intel.com>,
 Namhyung Kim <namhyung@...nel.org>,
 Andrii Nakryiko <andrii.nakryiko@...il.com>, bpf@...r.kernel.org,
 Joel Fernandes <joel@...lfernandes.org>, linux-trace-kernel@...r.kernel.org,
 Michael Jeanson <mjeanson@...icios.com>
Subject: Re: [PATCH resend 6/8] tracing/ftrace: Add might_fault check to
 syscall probes

On 2024-10-28 13:42, Thomas Gleixner wrote:
> On Mon, Sep 30 2024 at 15:23, Mathieu Desnoyers wrote:
>> diff --git a/kernel/trace/trace_syscalls.c b/kernel/trace/trace_syscalls.c
>> index a3d8ac00793e..0430890cbb42 100644
>> --- a/kernel/trace/trace_syscalls.c
>> +++ b/kernel/trace/trace_syscalls.c
>> @@ -303,6 +303,7 @@ static void ftrace_syscall_enter(void *data, struct pt_regs *regs, long id)
>>   	 * Syscall probe called with preemption enabled, but the ring
>>   	 * buffer and per-cpu data require preemption to be disabled.
>>   	 */
>> +	might_fault();
>>   	guard(preempt_notrace)();
> 
> I find it odd that the might_fault() check is in all the implementations
> and not in the tracepoint itself:
> 
>      if (syscall) {
>          might_fault();
>   	rcu_read_unlock_trace();
>     } else ...
> 
> That's where I would have expected it to be.

You raise a good point: we should also add a might_fault() check in
__DO_TRACE() in the syscall case, so we can catch incorrect use of the
syscall tracepoint even if no probes are registered to it.

I've added the might_fault() in each tracer syscall probe to make sure
a tracer don't end up registering a faultable probe on a tracepoint
protected with preempt_disable by mistake. It validates that the tracers
are using the tracepoint registration as expected.

I'll prepare separate a patch adding this and will add it to this
series.

Thanks,

Mathieu

> 
> Thanks,
> 
>          tglx

-- 
Mathieu Desnoyers
EfficiOS Inc.
https://www.efficios.com