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

[Steven Rostedt <rostedt@goodmis.org>]

On Wed, 2009-10-21 at 23:21 +0800, Wu Zhangjin wrote:

> > +unsigned long ftrace_get_parent_addr(unsigned long self_addr,
> > +				     unsigned long parent,
> > +				     unsigned long parent_addr,
> > +				     unsigned long fp)
> > +{
> > +	unsigned long sp, ip, ra;
> > +	unsigned int code;
> > +
> > +	/* move to the instruction "move ra, at" */
> > +	ip = self_addr - 8;
> > +
> > +	/* search the text until finding the "move s8, sp" instruction or
> > +	 * "s{d,w} ra, offset(sp)" instruction */
> > +	do {
> > +		ip -= 4;
> > +		/* read the text we want to match */
> > +		if (probe_kernel_read(&code, (void *)ip, 4)) {
> > +			WARN_ON(1);
> > +			panic("read the text failure\n");
> > +		}
> > +
> > +		/* if the first instruction above "move at, ra" is "move
> > +		 * s8(fp), sp", means the function is a leaf */
> > +		if ((code & MOV_FP_SP) == MOV_FP_SP)
> > +			return parent_addr;
> > +	} while (((code & S_RA) != S_RA));
> > +
> > +	sp = fp + (code & STACK_OFFSET_MASK);
> > +	ra = *(unsigned long *)sp;
> > +
> 
> Seems missed the fault protection here? is there a need? never met fault
> in this place and also the following two places, so, are we safe to
> remove all of the fault protection?

Is that "sp" basically already been check by the above
probe_kernel_read? If so, then it should be fine not to do the check
again.

-- Steve

> 
> Regards
> 	Wu Zhangjin
> 
> > +	if (ra == parent)
> > +		return sp;
> > +	else
> > +		panic
> > +		    ("failed on getting stack address of ra\n: addr: 0x%lx, code: 0x%x\n",
> > +		     ip, code);
> > +}
> > +