The following reply was made to PR port-m68k/36848; it has been noted by GNATS.

From: SAITOH Masanobu <msaitoh@netbsd.org>
To: gnats-bugs@NetBSD.org
Cc: 
Subject: PR/36848 CVS commit: [netbsd-4] src/sys/arch
Date: Tue, 11 Sep 2007 08:01:39 +0000 (UTC)

 Module Name:	src
 Committed By:	msaitoh
 Date:		Tue Sep 11 08:01:39 UTC 2007
 
 Modified Files:
 	src/sys/arch/amiga/amiga [netbsd-4]: trap.c
 	src/sys/arch/atari/atari [netbsd-4]: trap.c
 	src/sys/arch/cesfic/cesfic [netbsd-4]: trap.c
 	src/sys/arch/hp300/hp300 [netbsd-4]: trap.c
 	src/sys/arch/luna68k/luna68k [netbsd-4]: trap.c
 	src/sys/arch/mac68k/mac68k [netbsd-4]: trap.c
 	src/sys/arch/mvme68k/mvme68k [netbsd-4]: trap.c
 	src/sys/arch/news68k/news68k [netbsd-4]: trap.c
 	src/sys/arch/next68k/next68k [netbsd-4]: trap.c
 	src/sys/arch/sun3/sun3 [netbsd-4]: trap.c
 	src/sys/arch/x68k/x68k [netbsd-4]: trap.c
 
 Log Message:
 Pull up following revision(s) (requested by mhitch in ticket #866):
 	sys/arch/atari/atari/trap.c:		1.92
 	sys/arch/mac68k/mac68k/trap.c:		1.129
 	sys/arch/amiga/amiga/trap.c:		1.116
 	sys/arch/mvme68k/mvme68k/trap.c:	1.91
 	sys/arch/news68k/news68k/trap.c:	1.54
 	sys/arch/sun3/sun3/trap.c:		1.131
 	sys/arch/next68k/next68k/trap.c:	1.68
 	sys/arch/luna68k/luna68k/trap.c:	1.47
 	sys/arch/cesfic/cesfic/trap.c:		1.36
 	sys/arch/x68k/x68k/trap.c:		1.90
 	sys/arch/hp300/hp300/trap.c:		1.132
 68030 and 68040 processors consider the read portion of a read-modify-write
 transfer as a write to ensure the memory is writable before starting any
 transfer.  The fault status information does not reflect this in the 'read'
 status bit (i.e. it shows up as a read access), so faults with a RMW access
 to non-writable memory was not getting the correct protection.  The page would
 be read-only and the instruction would fault over and over.
 A specific example is when a process forks, and the child process attempts
 to execute a RMW access to a data page, which is read-only because it's CoP
 Copy-On-Write.
 When checking if the page needs to be writablek, also check the locked transfer
 and treat any locked transfer as a write.
 68060 already handled this correctly, since it has separate read and write
 fault bits, and both are set on a RMW access and the trap code was checking
 the write status bit.
 Fixes PR#36848.
 
 
 To generate a diff of this commit:
 cvs rdiff -r1.110.8.1 -r1.110.8.2 src/sys/arch/amiga/amiga/trap.c
 cvs rdiff -r1.84.8.1 -r1.84.8.2 src/sys/arch/atari/atari/trap.c
 cvs rdiff -r1.29.8.1 -r1.29.8.2 src/sys/arch/cesfic/cesfic/trap.c
 cvs rdiff -r1.125.8.1 -r1.125.8.2 src/sys/arch/hp300/hp300/trap.c
 cvs rdiff -r1.39.8.1 -r1.39.8.2 src/sys/arch/luna68k/luna68k/trap.c
 cvs rdiff -r1.121.8.1 -r1.121.8.2 src/sys/arch/mac68k/mac68k/trap.c
 cvs rdiff -r1.85.8.1 -r1.85.8.2 src/sys/arch/mvme68k/mvme68k/trap.c
 cvs rdiff -r1.47.8.1 -r1.47.8.2 src/sys/arch/news68k/news68k/trap.c
 cvs rdiff -r1.61.8.1 -r1.61.8.2 src/sys/arch/next68k/next68k/trap.c
 cvs rdiff -r1.126.8.1 -r1.126.8.2 src/sys/arch/sun3/sun3/trap.c
 cvs rdiff -r1.82.8.1 -r1.82.8.2 src/sys/arch/x68k/x68k/trap.c
 
 Please note that diffs are not public domain; they are subject to the
 copyright notices on the relevant files.