[m-dev.] for review: cleanup signal setting.

[Tyson Dowd]

On 15-May-1998, Fergus Henderson <fjh at cs.mu.OZ.AU> wrote:
> On 14-May-1998, Tyson Dowd <trd at cs.mu.OZ.AU> wrote:
> > -static void
> > -checked_signal(int sig, void (*handler)(int))
> > -{
> > -	/*
> > -	** We really need sigaction and SA_RESTART, otherwise profiling signals
> > -	** might interrupt I/O, causing a profiled program to get I/O errors.
> > -	** But if we haven't got it, I guess we just have to punt...
> > -	*/
> > -#ifndef SA_RESTART
> > -#define SA_RESTART 0
> > -#endif
> 
> I don't think your log message provided any rationale for removing that
> bit of code.  Did that just get unintentionally lost in the reorganization?

No, it didn't include any rationale, but it wasn't unintentional.

My rationale was:

	- The given comment did not justify defining SA_RESTART.
	- The code for profiling included this definition, but it was
	  not used in setup_signals for the memory zones. 
	- Its abscence never caused problems in setup_signals.
	- I assume this defintion was added because of a missing #include
	  or somesuch, which has since gone away.
	- If there was a more serious problem, I expect it would be
	  commented or occur elsewhere.
	- If removing this code causes a problem, I'll at least be able
	  to diagnose it, and then explain the SA_RESTART defintion
	  in a comment.

It should have been included in the log message (and cvs should
have a facility for adding extra log messages to changes).
Anyway, it's documented here now.

-- 
       Tyson Dowd           # So I asked Sarah: what's the serial number on
                            # your computer? She replied:
     trd at cs.mu.oz.au        #          A-C-2-4-0-V-/-5-0-H-Z
http://www.cs.mu.oz.au/~trd #