This is the mail archive of the systemtap@sourceware.org mailing list for the systemtap project.

Index Nav:	[Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav:	[Date Prev] [Date Next]	[Thread Prev] [Thread Next]
Other format:	[Raw text]

Re: __kbrobes?

From: Masami Hiramatsu <mhiramat at redhat dot com>
To: Jun Koi <junkoi2004 at gmail dot com>
Cc: SystemTAP <systemtap at sources dot redhat dot com>, Ananth N Mavinakayanahalli <ananth at in dot ibm dot com>, Jim Keniston <jkenisto at us dot ibm dot com>
Date: Wed, 30 Jul 2008 13:06:24 -0400
Subject: Re: __kbrobes?
References: <fdaac4d50807300359m329b6ff1k6ec2e2b1c56aab24@mail.gmail.com> <48908579.6050603@redhat.com>

Masami Hiramatsu wrote:
> Hi Jun,
> 
> Jun Koi wrote:
>> Hi,
>>
>> I found some functions use __kprobes as directive (like text_poke()).
>> Closer look points out that it is used to collect them into a section
>> ".kprobes.text". What do we have to do with that? Is that because we
>> dont want to probe those areas?
> 
> The __kprobes means that those functions are invoked from kprobe itself.
> And kprobes doesn't accept to probe at the addresses in ".kprobes.text"
> sectoin for preventing recursive-call of kprobes.
> 
>> If so, why do we want to protect text_poke() against kprobes?
> 
> Indeed. text_poke() seems not to be invoked from kprobes probing
> functions. It is used only for disarming/arming breakpoints from
> registering/disabling kprobes.

Hmm, arch_disarm_kprobe() which invokes text_poke() is invoked in
a special case on x86_64. See reenter_kprobe@arch/x86/kernel/kprobes.c.
So, text_poke() should be marked as __kprobes.

Thank you,

-- 
Masami Hiramatsu

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

e-mail: mhiramat@redhat.com

References:
- __kbrobes?
  - From: Jun Koi
- Re: __kbrobes?
  - From: Masami Hiramatsu

Index Nav:	[Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav:	[Date Prev] [Date Next]	[Thread Prev] [Thread Next]