From: "Uwe Kleine-König" <u.kleine-koenig@pengutronix.de>
To: Russell King - ARM Linux <linux@arm.linux.org.uk>
Cc: Ingo Molnar <mingo@elte.hu>,
Frederic Weisbecker <fweisbec@gmail.com>,
Abhishek Sagar <sagar.abhishek@gmail.com>,
Tim Bird <tim.bird@am.sony.com>,
linux-arm-kernel <linux-arm-kernel@lists.arm.linux.org.uk>,
linux kernel <linux-kernel@vger.kernel.org>,
Steven Rostedt <rostedt@goodmis.org>,
Ingo Molnar <mingo@redhat.com>
Subject: Re: Anyone working on ftrace function graph support on ARM?
Date: Wed, 25 Mar 2009 13:09:44 +0100 [thread overview]
Message-ID: <20090325120944.GA30873@pengutronix.de> (raw)
In-Reply-To: <20090325112115.GB31464@n2100.arm.linux.org.uk>
Hi Russell,
> > > Hmm, and it looks like the ftrace code is rather crap:
> > >
> > > ENTRY(mcount)
> > > stmdb sp!, {r0-r3, lr}
> > > ldr r0, =ftrace_trace_function
> > > ldr r2, [r0]
> > > adr r0, ftrace_stub
> > > cmp r0, r2
> > > bne trace
> > > ldr lr, [fp, #-4] @ restore lr
> > > ldmia sp!, {r0-r3, pc}
> > >
> > > trace:
> > > ldr r1, [fp, #-4] @ lr of instrumented routine
> > > mov r0, lr
> > > sub r0, r0, #MCOUNT_INSN_SIZE
> > > mov lr, pc
> > > mov pc, r2
> > > XXX calling a C function results in r0-r3,ip,lr being clobbered XXX
> > >
> > > mov lr, r1 @ restore lr
> > > XXX not necessarily, r1 might be some other random value
> > >
> > > ldmia sp!, {r0-r3, pc}
> > >
> > > In fact, to me the above code looks totally crap, because it's checking
> > > whether the caller is 'ftrace_stub'. It can never be 'ftrace_stub'
> > > because that is an assembly function:
> > >
> > > .globl ftrace_stub
> > > ftrace_stub:
> > > mov pc, lr
> > >
> > > and therefore gcc has no hand in adding a mcount call to it.
> > Hhhm. Isn't the equivalent C-Code ~ as follows:
> >
> > if (ftrace_trace_function != ftrace_stub)
> > trace(some, args);
> > return;
> > ? ftrace_trace_function is initialised to ftrace_stub at compile time
> > and is changed when a tracerfunction is registered.
>
> Correct. But my point is that there's no way for ftrace_stub to ever call
> mcount. So the check there is pointless.
Which check? ftrace_trace_function isn't the caller of mcount but a
function pointer defined in kernel/trace/ftrace.c. And if this pointer
changes, the if-condition becomes true.
> Ok - it would be nice if there was a comment to explain that.
Some comments would be nice, that's right.
> Is someone going to fix the existing ftrace before trying to build stuff
> on top of it?
Mmh, I think you misunderstood the code, so I don't see something to fix
here (at least for non-dynamic ftrace). If the next few mails show,
that it's me who didn't understand the code, I will fix the problems, of
course.
Best regards
Uwe
--
Pengutronix e.K. | Uwe Kleine-König |
Industrial Linux Solutions | http://www.pengutronix.de/ |
next prev parent reply other threads:[~2009-03-25 12:10 UTC|newest]
Thread overview: 37+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-03-24 19:38 Anyone working on ftrace function graph support on ARM? Tim Bird
2009-03-24 20:25 ` Ingo Molnar
2009-03-24 20:48 ` Tim Bird
2009-03-24 20:38 ` Uwe Kleine-König
2009-03-24 21:36 ` Frederic Weisbecker
2009-03-24 21:40 ` Tim Bird
2009-03-24 21:48 ` Ingo Molnar
2009-03-24 21:57 ` Frederic Weisbecker
2009-03-24 22:14 ` Ingo Molnar
2009-03-24 22:54 ` Frederic Weisbecker
2009-03-25 8:36 ` Russell King - ARM Linux
2009-03-25 16:00 ` Steven Rostedt
2009-03-25 17:13 ` Frederic Weisbecker
2009-03-25 20:27 ` [PATCH][GIT PULL] x86, function-graph: only save return values on x86_64 Steven Rostedt
2009-03-25 20:45 ` Jaswinder Singh Rajput
2009-03-25 21:26 ` Steven Rostedt
2009-04-08 16:09 ` Ingo Molnar
2009-04-08 16:37 ` Steven Rostedt
2009-04-08 16:41 ` Ingo Molnar
2009-04-08 17:40 ` Frederic Weisbecker
2009-03-24 22:29 ` Anyone working on ftrace function graph support on ARM? Abhishek Sagar
2009-03-24 22:48 ` Frederic Weisbecker
2009-03-25 8:42 ` Russell King - ARM Linux
2009-03-25 8:54 ` Ingo Molnar
2009-03-25 9:57 ` Russell King - ARM Linux
2009-03-25 10:45 ` Uwe Kleine-König
2009-03-25 11:21 ` Russell King - ARM Linux
2009-03-25 12:09 ` Uwe Kleine-König [this message]
2009-03-25 16:41 ` Tim Bird
2009-03-25 11:41 ` Frederic Weisbecker
2009-03-25 16:34 ` Tim Bird
2009-03-25 17:05 ` Uwe Kleine-König
2009-03-25 17:17 ` Russell King - ARM Linux
2009-03-25 18:37 ` Tim Bird
2009-03-25 18:41 ` Steven Rostedt
2009-03-27 12:58 ` Catalin Marinas
2009-04-09 15:29 ` Daniel Jacobowitz
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=20090325120944.GA30873@pengutronix.de \
--to=u.kleine-koenig@pengutronix.de \
--cc=fweisbec@gmail.com \
--cc=linux-arm-kernel@lists.arm.linux.org.uk \
--cc=linux-kernel@vger.kernel.org \
--cc=linux@arm.linux.org.uk \
--cc=mingo@elte.hu \
--cc=mingo@redhat.com \
--cc=rostedt@goodmis.org \
--cc=sagar.abhishek@gmail.com \
--cc=tim.bird@am.sony.com \
/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;
as well as URLs for NNTP newsgroup(s).