lkml.org 
[lkml]   [2013]   [Apr]   [25]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: [PATCH 05/14] perf tools: Introduce new 'ftrace' tool
On Thu, Apr 25, 2013 at 09:05:21AM -0400, Steven Rostedt wrote:
> On Thu, 2013-04-25 at 18:51 +0900, Namhyung Kim wrote:
> > >
> > >>
> > >> Do you mean it should be changed to
> > >>
> > >> if (write_tracing_file("set_ftrace_pid", "-1") < 0)
> > >>
> > >> ? It seems it's the right thing to do for compatibility.
> > >
> > >
> > > looks like it's enough only to open it with O_TRUNC,
> > > so I guess whatever minimal goes write_tracing_file
>
> Correct.
>
> >
> > Yeah, but I can see below from trace-cmd code.
> >
> > trace-cmd/trace-record.c:
> >
> > /*
> > * Older kernels required "-1" to disable pid
> > */
> > if (ret < 0 && !strlen(pid))
> > ret = write(fd, "-1", 2);
> >
> >
> > Could you clarify this, Steve?
>
> As the comment states, older kernels (before 2.6.33) required a -1. As
> trace-cmd is set up to work with at least 2.6.32, it requires the "-1"
> write. It doesn't hurt to have it.
>
> But for perf ftrace (or whatever), I don't think we need the requirement
> to work with 2.6.32 kernels. Perf was added in 2.6.31, but do all
> functionalities still work that far back?

perf maintains backward compatibility, but not sure how far
back it goes.. but as long as it's just writing -1 in here
let's do that

jirka


\
 
 \ /
  Last update: 2013-04-25 18:21    [W:0.595 / U:0.012 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site