Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!watmath!clyde!cbatt!ihnp4!qantel!lll-lcc!styx!twg-ap!amdahl!pyramid!nsc!csi!epimass!oliveb!hplabs!sdcrdcf!randvax!edhall
From: edhall@randvax.UUCP (Ed Hall)
Newsgroups: net.bugs.4bsd,net.unix-wizards
Subject: Re: Interrupting find(1)
Message-ID: <557@randvax.UUCP>
Date: Fri, 26-Sep-86 03:40:08 EDT
Article-I.D.: randvax.557
Posted: Fri Sep 26 03:40:08 1986
Date-Received: Tue, 30-Sep-86 19:49:52 EDT
References: <334@yabbie.rmit.oz> <7566@sun.uucp>
Reply-To: edhall@rand-unix.UUCP (Ed Hall)
Organization: Rand Corp., Santa Monica
Lines: 14
Xref: watmath net.bugs.4bsd:2360 net.unix-wizards:19719

In article <7566@sun.uucp> guy@sun.uucp (Guy Harris) writes:
>> The fix would be simply to change SIG_IGN to SIG_HOLD in both statements.
>
>But not that way.  SIG_HOLD only appears in "/usr/include/sys/signal.h" as a
>historical holdover; setting the signal action to SIG_HOLD will *not* set
>the "p_sigmask" bit for that signal, so it won't hold it.  The correct fix
>is to put a "sigblock"/"sigsetmask" pair around the "wait".

	--or look at the status returned by wait(), and exit appropriately
	if the child process was killed by SIGINT...  This would even work
	on Vanilla V7 systems (assuming any are still around).

		-Ed Hall
		decvax!randvax!edhall