On Sep 14,  5:43pm, enami@but-b.or.jp (enami tsugutomo) wrote:
-- Subject: Re: more siginfo q;s

| > +static void	ksiginfo_del(struct proc *p, ksiginfo_t *);
| 
| The p isn't necessary.

How do I clear the pointer so sigctx?

| > +	if (
| > +#ifdef notyet	/* XXX: QUEUING */
| > +	    ksi->ksi_signo >= SIGRTMIN ||
| > +#endif
| > +	    (kpool = ksiginfo_get(p, ksi->ksi_signo)) == NULL) {
| > +		if ((kpool = pool_get(&ksiginfo_pool, PR_NOWAIT)) == NULL)
| > +			return;
| > +		*kpool = *ksi;
| > +		ksiginfo_put(p, kpool);
| > +	} else
| > +		*kpool = *ksi;
| 
| As the ksiginfo_get() returns an element with leaving it on the list,
| I guess the assignment in the else-clause break the list.

Yes, this is a problem. I'll fix it.
 
| > +        exithook_establish(ksiginfo_exithook, NULL);
| 
| Use tab to indent.

Fixed, simon caught it.

| > +			kpsendsig(l, ksi, returnmask);
| > +			ksiginfo_del(p, ksi);
| > +			pool_put(&ksiginfo_pool, ksi);
| 
| Hmm, it looks like kpsendsig() may sleep but are you sure that bad
| thing won't occur?  (I don't know since I just read this diff through
| peephole).

What kind of bad thing you are thinking of?
 
| > +	struct ksiginfo *ksi_next;
| > +	struct ksiginfo *ksi_prev;
| 
| Why needs to be doublly linked?

I had it a singly linked list, but I thought that the performance/storage
trade-off is better for the doubly linked list, specially for queued signals.

christos