kern/44011: [PATCH] PT_CONTINUE, PT_KILL and PT_DETACH are broken

[Mark Kettenis]

The following reply was made to PR kern/44011; it has been noted by GNATS.

From: Mark Kettenis <kettenis at chello.nl>
To: freebsd-gnats-submit at FreeBSD.org, kettenis at chello.nl
Cc:  
Subject: Re: kern/44011: [PATCH] PT_CONTINUE, PT_KILL and PT_DETACH are broken
Date: Sat, 12 Apr 2003 19:37:25 +0200 (CEST)

 Since the patch no longer applies cleanly, I updated it.  Here's a
 version against revision 1.244 of kern_sig.c.  I also updated the file
 at
 
   http://members.chello.nl/~m.m.kettenis/FreeBSD/5-current/ptrace.patch
 
 I hope somebody can look into this issue.
 
 Mark
 
 --- /usr/src/sys/kern/kern_sig.c.orig	Sat Apr 12 04:54:46 2003
 +++ /usr/src/sys/kern/kern_sig.c	Sat Apr 12 18:47:41 2003
 @@ -1918,7 +1918,6 @@ issignal(td)
  		if (SIGISEMPTY(sigpending))	/* no signal to send */
  			return (0);
  		sig = sig_ffs(&sigpending);
 -		prop = sigprop(sig);
  
  		_STOPEVENT(p, S_SIG, sig);
  
 @@ -1950,14 +1949,6 @@ issignal(td)
  			PROC_LOCK(p);
  
  			/*
 -			 * If the traced bit got turned off, go back up
 -			 * to the top to rescan signals.  This ensures
 -			 * that p_sig* and ps_sigact are consistent.
 -			 */
 -			if ((p->p_flag & P_TRACED) == 0)
 -				continue;
 -
 -			/*
  			 * If parent wants us to take the signal,
  			 * then it will leave it in p->p_xstat;
  			 * otherwise we just look for signals again.
 @@ -1968,6 +1959,14 @@ issignal(td)
  				continue;
  
  			/*
 +			 * If the traced bit got turned off, go back up
 +			 * to the top to rescan signals.  This ensures
 +			 * that p_sig* and ps_sigact are consistent.
 +			 */
 +			if ((p->p_flag & P_TRACED) == 0)
 +				continue;
 +
 +			/*
  			 * Put the new signal into td_siglist.  If the
  			 * signal is being masked, look for other signals.
  			 */
 @@ -1976,6 +1975,8 @@ issignal(td)
  				continue;
  			signotify(td);
  		}
 +
 +		prop = sigprop(sig);
  
  		/*
  		 * Decide whether the signal should be returned.