Module Name: src Committed By: tsutsui Date: Sat May 19 08:29:32 UTC 2012
Modified Files: src/sys/arch/x68k/x68k: locore.s Log Message: Fix leftover comment derived from hp300 but not applicable to x68k. To generate a diff of this commit: cvs rdiff -u -r1.109 -r1.110 src/sys/arch/x68k/x68k/locore.s Please note that diffs are not public domain; they are subject to the copyright notices on the relevant files.
Modified files: Index: src/sys/arch/x68k/x68k/locore.s diff -u src/sys/arch/x68k/x68k/locore.s:1.109 src/sys/arch/x68k/x68k/locore.s:1.110 --- src/sys/arch/x68k/x68k/locore.s:1.109 Sun Apr 29 07:17:12 2012 +++ src/sys/arch/x68k/x68k/locore.s Sat May 19 08:29:32 2012 @@ -1,4 +1,4 @@ -/* $NetBSD: locore.s,v 1.109 2012/04/29 07:17:12 tsutsui Exp $ */ +/* $NetBSD: locore.s,v 1.110 2012/05/19 08:29:32 tsutsui Exp $ */ /* * Copyright (c) 1988 University of Utah. @@ -789,13 +789,7 @@ Lstart2: /* * Prepare to enable MMU. - * Since the kernel is not mapped logical == physical we must insure - * that when the MMU is turned on, all prefetched addresses (including - * the PC) are valid. In order guarantee that, we use the last physical - * page (which is conveniently mapped == VA) and load it up with enough - * code to defeat the prefetch, then we execute the jump back to here. - * - * Is this all really necessary, or am I paranoid?? + * Since the kernel is mapped logical == physical, we just turn it on. */ RELOC(Sysseg_pa, %a0) | system segment table addr movl %a0@,%d1 | read value (a PA)