Re: [PATCH v15 2/4] trace/objtrace: Get the value of the object

[Steven Rostedt <rostedt@goodmis.org>]

On Sun, 25 Sep 2022 13:27:10 +0800
Jeff Xie <xiehuan09@gmail.com> wrote:

> 
> It also looks like the "add" command field has been removed,
> Masami previously suggested I to add a reserved command field for the
> future (see [1] and [2]):
> 
> [1]:
> https://lore.kernel.org/lkml/20211026225234.549ec8e9eb59f1fd1671edbc@kernel.org/
> 
> [2]:
> https://lore.kernel.org/lkml/20211105004704.3f4aa456c2c537a56f64e65a@kernel.org/

Ah, I see. add means enable the tracing and del means disable.

Masami, perhaps we should call it "enable/disable" instead? Because, to
me "del" means to remove it (permanently), where as "enable/disable"
matches other triggers like the "enable/disable event" that starts and
stops another event.

> 
> 
> > This code is available in the kprobe code that eprobes also uses:
> >
> > See process_fetch_insn() in kernel/trace/trace_eprobe.c
> >
> > and the parsing of the string is in kernel/trace/trace_probe.c:
> >
> >   parse_probe_arg()
> >  
> 
> Thanks for your guidance, but the next merge window is coming soon,
> and since I'm not doing full-time kernel development,
> I can't guarantee that the development and testing of this change will
> be completed before the next window merge.
> 
>  Anyone familiar with this code providing a patch is welcome ;-)

If I get a chance, I may see if I can quickly implement it.

> 
> >
> > I think doing this will make it much more extensive, not to mention it
> > will match the syntax of other code in the tracing infrastructure.  
> 
> I super agree.

Thanks for you patience, and I'm sorry it took so long for me to review
it.

-- Steve