Re: [PATCH -tip v5 2/7] kprobes: checks probe address is instruction boudary on x86

From: Masami Hiramatsu <mhiramat@redhat.com>
To: Steven Rostedt <rostedt@goodmis.org>
Cc: Ingo Molnar <mingo@elte.hu>, lkml <linux-kernel@vger.kernel.org>,
	systemtap <systemtap@sources.redhat.com>,
	kvm <kvm@vger.kernel.org>,
	Ananth N Mavinakayanahalli <ananth@in.ibm.com>,
	Jim Keniston <jkenisto@us.ibm.com>
Subject: Re: [PATCH -tip v5 2/7] kprobes: checks probe address is instruction boudary on x86
Date: Mon, 11 May 2009 11:14:36 -0400	[thread overview]
Message-ID: <4A0840DC.7080507@redhat.com> (raw)
In-Reply-To: <4A083DAD.8000009@redhat.com>

Masami Hiramatsu wrote:
>>> +	if (!kallsyms_lookup(paddr, NULL, &offset, NULL, __dummy_buf))
>>> +		return 0;
>>> +
>>> +	/* Decode instructions */
>>> +	addr = paddr - offset;
>>> +	while (addr < paddr) {
>>> +		insn_init_kernel(&insn, (void *)addr);
>>> +		insn_get_opcode(&insn);
>>> +		if (OPCODE1(&insn) == BREAKPOINT_INSTRUCTION) {
>>> +			ret = recover_probed_instruction(buf, addr);
>> Oh, the above puts back the original op code. That is why it is OK?
> 
> Oops, no. I have to use get_kprobe() instead. Thanks!

Ah, I forgot another possibility. There might be another subsystem,
like kgdb, will put their break point on the kernel.
In that case, decoder will decode the instruction is a break point
instruction and the first opcode is int3. So, this part is correct.
In the future, we need to add a generic recover_instruction() code
for those text modification subsystems.

Thank you,
-- 
Masami Hiramatsu

Software Engineer
Hitachi Computer Products (America) Inc.
Software Solutions Division

e-mail: mhiramat@redhat.com