From: Steven Rostedt <rostedt@goodmis.org>
To: Peter Zijlstra <peterz@infradead.org>
Cc: LKML <linux-kernel@vger.kernel.org>, Jessica Yu <jeyu@kernel.org>,
Ingo Molnar <mingo@kernel.org>
Subject: Re: [PATCH] ftrace/module: Allow ftrace to make only loaded module text read-write
Date: Thu, 10 Oct 2019 12:59:15 -0400 [thread overview]
Message-ID: <20191010125915.34a2cfdf@gandalf.local.home> (raw)
In-Reply-To: <20191010105515.5eba7f31@gandalf.local.home>
On Thu, 10 Oct 2019 10:55:15 -0400
Steven Rostedt <rostedt@goodmis.org> wrote:
I'll try to add more detail (mapping to functions)
>
> OK, so basically this moves the enabling of function tracing from
> within the ftrace_module_enable() code without releasing the
> ftrace_lock mutex.
>
> But we have an issue with the state of the module here, as it is still
> set as MODULE_STATE_UNFORMED. Let's look at what happens if we have:
>
>
> CPU0 CPU1
> ---- ----
> echo function > current_tracer
Just need to know that the above will now have all loaded modules have
their functions being traced. That is ftrace_startup > 0.
> modprobe foo
> enable foo functions to be traced
> (foo function records not disabled)
In your code, we have here:
load_module()
foo->state == UNFORMED
ftrace_module_init()
ftrace_process_locs()
mutex_lock(ftrace_lock);
__ftrace_replace_code(rec, 1)
/* Enabling foo's functions */
mutex_unlock(ftrace_lock);
> echo nop > current_tracer
>
> disable all functions being
> traced including foo functions
ftrace_shutdown()
ftrace_run_update_code()
ftrace_arch_code_modify_prepare()
[ on arm and nds32 ]
>
> arch calls set_all_modules_text_rw()
> [skips UNFORMED modules, which foo still is ]
>
> set foo's text to read-only
> foo's state to COMING
complete_formation()
module_enable_ro()
>
> tries to disable foo's functions
> foo's text is read-only
arch_ftrace_update_code()
ftrace_modify_all_code()
[ Includes foo's functions ]
>
> BUG trying to write to ro text!!!
Hope that makes more sense.
-- Steve
>
>
> Like I said, this is very subtle. It may no longer be a bug on x86
> with your patches, but it will bug on ARM or anything else that still
> uses set_all_modules_text_rw() in the ftrace prepare code.
next prev parent reply other threads:[~2019-10-10 16:59 UTC|newest]
Thread overview: 16+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-10-10 2:36 Steven Rostedt
2019-10-10 7:31 ` Peter Zijlstra
2019-10-10 9:26 ` Peter Zijlstra
2019-10-10 9:33 ` Peter Zijlstra
2019-10-10 9:36 ` Peter Zijlstra
2019-10-10 12:29 ` Peter Zijlstra
2019-10-10 14:55 ` Steven Rostedt
2019-10-10 15:03 ` Steven Rostedt
2019-10-10 16:59 ` Steven Rostedt [this message]
2019-10-10 17:01 ` Peter Zijlstra
2019-10-10 17:20 ` Steven Rostedt
2019-10-11 11:09 ` Peter Zijlstra
2019-10-10 12:50 ` Steven Rostedt
2019-10-10 14:11 ` Peter Zijlstra
2019-10-10 12:58 ` Steven Rostedt
2019-10-14 12:31 ` Jessica Yu
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20191010125915.34a2cfdf@gandalf.local.home \
--to=rostedt@goodmis.org \
--cc=jeyu@kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@kernel.org \
--cc=peterz@infradead.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
all inboxes | Powered by JetHome®