Re: [PATCH v4 05/15] h8300: Process and signal

[Yoshinori Sato <ysato@users.sourceforge.jp>]

At Sat, 21 Feb 2015 09:51:23 +0000,
Al Viro wrote:
> 
> On Sat, Feb 21, 2015 at 04:53:49PM +0900, Yoshinori Sato wrote:
> 
> > +static inline void *
> > +get_sigframe(struct k_sigaction *ka, struct pt_regs *regs, size_t frame_size)
> > +{
> > +	unsigned long usp;
> > +
> > +	/* Default to using normal stack.  */
> > +	usp = rdusp();
> > +
> > +	/* This is the X/Open sanctioned signal stack switching.  */
> > +	if (ka->sa.sa_flags & SA_ONSTACK) {
> > +		if (!sas_ss_flags(usp))
> > +			usp = current->sas_ss_sp + current->sas_ss_size;
> > +	}
> 
> Use sigsp(), please.  Pass ksig to that sucker and it'll become
> 	return (void *)((sigsp(rdusp, ksig) - frame_size) & -8UL);

OK.

> 
> > +static void do_signal(struct pt_regs *regs)
> > +	 * We want the common case to go fast, which
> > +	 * is why we may in certain cases get here from
> > +	 * kernel mode. Just return without doing anything
> > +	 * if so.
> > +	 */
> > +	if ((regs->ccr & 0x10))
> > +		return;
> 
> Can we get there in kernel mode, actually?

It case checking for entry.S
Removed.

Thanks.

-- 
Yoshinori Sato
<ysato@users.sourceforge.jp>