[Top] [All Lists]

Re: [PATCH -v4 4/9] tracing: add static function tracer support for MIPS

To: Adam Nemet <>
Subject: Re: [PATCH -v4 4/9] tracing: add static function tracer support for MIPS
From: Steven Rostedt <>
Date: Thu, 22 Oct 2009 16:52:06 -0400
Cc: David Daney <>,, Richard Sandiford <>,,, Thomas Gleixner <>, Ralf Baechle <>, Nicholas Mc Guire <>
In-reply-to: <19168.49354.525249.654494@ropi.home>
Organization: Kihon Technologies Inc.
Original-recipient: rfc822;
References: <> <> <> <> <> <1256233679.23653.7.camel@falcon> <> <19168.49354.525249.654494@ropi.home>
On Thu, 2009-10-22 at 13:30 -0700, Adam Nemet wrote:
> Maybe scanning the instructions should stop at the beginning of the function
> based on the kernel's symbol table.  I am not sure if we can establish any
> other stopping condition without affecting performance too much.

How would we check the kernel symbol table? Remember, this happens at
_every_ function call.

> Speaking of performance, -pg also affects the instruction scheduling freedom
> of the compiler in the prologue.  With profiling, we limit optimizations not
> to move instructions in and out of the prologue.
> Also note that for functions invoked via tail call you won't get an exit
> event.  E.g. if bar is tail-called from foo:
>   foo entered
>   bar entered
>   foo/bar exited
> However, this is not MIPS-specific and you can always disable tail calls
> with -fno-optimize-sibling-calls.

The question is, would bar have a _mcount call? So far, we have not had
any issues with this on either x86 nor PPC.

/me knocks on wood.

-- Steve

<Prev in Thread] Current Thread [Next in Thread>