Date:   Wed, 25 Jan 2023 12:39:36 +0530
From:   "Naveen N. Rao" <naveen.n.rao@linux.vnet.ibm.com>
Subject: Re: [PATCH] selftests/ftrace: Extend multiple_kprobes.tc to add
 multiple consecutive probes in a function
To:     Masami Hiramatsu <mhiramat@kernel.org>
Cc:     Akanksha J N <akanksha@linux.ibm.com>,
        linux-kernel@vger.kernel.org, linux-kselftest@vger.kernel.org,
        linux-trace-kernel@vger.kernel.org, rostedt@goodmis.org,
        shuah@kernel.org
References: <20230112095600.37665-1-akanksha@linux.ibm.com>
        <1673529279.3c5f8oes3z.naveen@linux.ibm.com>
        <20230113005153.c6ca2f75b9d12627eb63308a@kernel.org>
        <1673601511.tq30r5phea.naveen@linux.ibm.com>
        <20230114002126.a37640f815b74e9e78259a9f@kernel.org>
        <1673856229.a7tekgas75.naveen@linux.ibm.com>
        <20230120085554.ab4dc1b72990a4957c4c88e2@kernel.org>
In-Reply-To: <20230120085554.ab4dc1b72990a4957c4c88e2@kernel.org>
MIME-Version: 1.0
User-Agent: astroid/4d6b06ad (https://github.com/astroidmail/astroid)
Message-Id: <1674629944.vwzovyd4lk.naveen@linux.ibm.com>
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

Hi Masami,

Masami Hiramatsu wrote:
>> >=20
>> > Yes, please make it separate, this test case is for checking whether
>> > the ftrace can define/enable/disable multiple kprobe events. Not for
>> > checking kprobe with different types, nor checking interactions among
>> > different types of kprobes.
>> >=20
>> > (BTW, if you want to test optprobe on x86, you can not put the probes
>> >  within the jump instruction (+5 bytes). It will unoptimize existing
>> >  optimized kprobe in that case)
>>=20
>> Ok, I can see why we won't be able to optimize any of the probes on x86=20
>> with this approach. But, we should be able to do so on powerpc and arm,=20
>> the only other architectures supporting OPTPROBES at this time. For x86,=
=20
>> we may have to extend the test to check kprobes/list.
>=20
> Are there any instruction type specific limitation on those arch for
> using optprobe? I guess the 'call' (branch with link register) will not
> able to be optimized because it leaves the trampoline address on the
> stack.

Yes, at least on powerpc, we only optimize ALU instructions and do not=20
optimize load/store instructions, among many others. This is the reason=20
we try to put a probe uptil 256 offset into a function in the proposed=20
test, which will almost certainly catch an instruction that can be=20
optimized.

>=20
>>=20
>> Crucially, I think trying to place a probe at each byte can still=20
>> exercize interactions across KPROBES_ON_FTRACE and normal kprobes, so=20
>> this test is still a good start. In addition, we get to ensure that=20
>> kprobes infrastructure is rejecting placing probes at non-instruction=20
>> boundaries.
>=20
> The interfere between probes can be happen between kprobes and optprobe
> (*only on x86*), but not with KPORBES_ON_FTRACE. The ftrace replaced NOP
> will be handled as one instruction.=20

Yes.

>=20
>> > And do you really need to run "multiple" kprobes at once?
>> > I think what you need is 'kprobe_opt_types.tc'.
>>=20
>> Yes, enabling those probes is a good stress test to ensure we are only=20
>> accepting valid probe locations.
>>=20
>> multiple_kprobe_types.tc ? :)
>=20
> Please don't mixed it with the concept of 'multiple' probe test.
> It is different that
>  - kprobes can put probes on each instruction boundary.
>  - kprobes can allocate and enable multiple probes at the same time.
>=20
> What the multiple_kprobes.tc tests is the latter one.
> (This is the reason why it chooses different functions so as not to
>  interfere with each other.)

Ok, I was coming from the point of view that both tests end up=20
installing "multiple" kprobes, but I do see your point.

How about adding two new tests:
1. The same test as has been proposed in this thread: trying to add a=20
kprobe at every byte within $FUNCTION_FORK upto an offset of 256 bytes.=20
We can probably call it kprobe_insn_boundary.tc
2. A new test to ensure we can add different kprobe types=20
(kprobe_opt_types.tc). This test will need to enable and check if each=20
probe has been optimized or not and needs arch-specific knowledge so=20
that we can take care of x86.

Would that be ok?


Thanks,
Naveen