linux-kernel - Re: [PATCH 0/7] tracing: Add support for in-kernel synthetic event API

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] [thread-next>] [day] [month] [year] [list]

Message-Id: <20191219234511.bb499b3d1590059506db6982@kernel.org>
Date:   Thu, 19 Dec 2019 23:45:11 +0900
From:   Masami Hiramatsu <mhiramat@...nel.org>
To:     Tom Zanussi <zanussi@...nel.org>
Cc:     rostedt@...dmis.org, artem.bityutskiy@...ux.intel.com,
        mhiramat@...nel.org, linux-kernel@...r.kernel.org,
        linux-rt-users@...r.kernel.org
Subject: Re: [PATCH 0/7] tracing: Add support for in-kernel synthetic event
 API

Hello Tom,

On Wed, 18 Dec 2019 09:27:36 -0600
Tom Zanussi <zanussi@...nel.org> wrote:

> Hi,
> 
> I've recently had several requests and suggestions from users to add
> support for the creation and generation of synthetic events from
> kernel code such as modules, and not just from the available command
> line commands.

This reminds me my recent series of patches.

Could you use synth_event_run_command() for this purpose as I did
in boot-time tracing series?

https://lkml.kernel.org/r/157528179955.22451.16317363776831311868.stgit@devnote2

Above example uses a command string as same as command string, but I think
we can introduce some macros to construct the command string for easier
definition.

Or maybe it is possible to pass the const char *args[] array to that API,
instead of single char *cmd. (for dynamic event definiton)

Maybe we should consider more generic APIs for modules to create/delete
dynamic-event including synthetic and probes, and those might reuse
existing command parsers.

> This patchset adds support for that.  The first three patches add some
> minor preliminary setup, followed by the two main patches that add the
> ability to create and generate synthetic events from the kernel.  The
> next patch adds a test module that demonstrates actual use of the API
> and verifies that it works as intended, followed by Documentation.

Could you also check the locks are correctly acquired? It seems that
your APIs doesn't lock it.

Thank you,

-- 
Masami Hiramatsu <mhiramat@...nel.org>