2001-10-03 17:10:38 +04:00
|
|
|
/////////////////////////////////////////////////////////////////////////
|
2003-06-20 12:58:12 +04:00
|
|
|
// $Id: proc_ctrl.cc,v 1.72 2003-06-20 08:58:12 sshwarts Exp $
|
2001-10-03 17:10:38 +04:00
|
|
|
/////////////////////////////////////////////////////////////////////////
|
|
|
|
//
|
2001-04-10 06:20:02 +04:00
|
|
|
// Copyright (C) 2001 MandrakeSoft S.A.
|
2001-04-10 05:04:59 +04:00
|
|
|
//
|
|
|
|
// MandrakeSoft S.A.
|
|
|
|
// 43, rue d'Aboukir
|
|
|
|
// 75002 Paris - France
|
|
|
|
// http://www.linux-mandrake.com/
|
|
|
|
// http://www.mandrakesoft.com/
|
|
|
|
//
|
|
|
|
// This library is free software; you can redistribute it and/or
|
|
|
|
// modify it under the terms of the GNU Lesser General Public
|
|
|
|
// License as published by the Free Software Foundation; either
|
|
|
|
// version 2 of the License, or (at your option) any later version.
|
|
|
|
//
|
|
|
|
// This library is distributed in the hope that it will be useful,
|
|
|
|
// but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
|
|
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
|
|
|
|
// Lesser General Public License for more details.
|
|
|
|
//
|
|
|
|
// You should have received a copy of the GNU Lesser General Public
|
|
|
|
// License along with this library; if not, write to the Free Software
|
|
|
|
// Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
|
|
|
|
|
|
|
|
|
2001-05-24 22:46:34 +04:00
|
|
|
#define NEED_CPU_REG_SHORTCUTS 1
|
2001-04-10 05:04:59 +04:00
|
|
|
#include "bochs.h"
|
merge in BRANCH-io-cleanup.
To see the commit logs for this use either cvsweb or
cvs update -r BRANCH-io-cleanup and then 'cvs log' the various files.
In general this provides a generic interface for logging.
logfunctions:: is a class that is inherited by some classes, and also
. allocated as a standalone global called 'genlog'. All logging uses
. one of the ::info(), ::error(), ::ldebug(), ::panic() methods of this
. class through 'BX_INFO(), BX_ERROR(), BX_DEBUG(), BX_PANIC()' macros
. respectively.
.
. An example usage:
. BX_INFO(("Hello, World!\n"));
iofunctions:: is a class that is allocated once by default, and assigned
as the iofunction of each logfunctions instance. It is this class that
maintains the file descriptor and other output related code, at this
point using vfprintf(). At some future point, someone may choose to
write a gui 'console' for bochs to which messages would be redirected
simply by assigning a different iofunction class to the various logfunctions
objects.
More cleanup is coming, but this works for now. If you want to see alot
of debugging output, in main.cc, change onoff[LOGLEV_DEBUG]=0 to =1.
Comments, bugs, flames, to me: todd@fries.net
2001-05-15 18:49:57 +04:00
|
|
|
#define LOG_THIS BX_CPU_THIS_PTR
|
2001-04-10 05:04:59 +04:00
|
|
|
|
2001-05-23 12:16:07 +04:00
|
|
|
#if BX_USE_CPU_SMF
|
|
|
|
#define this (BX_CPU(0))
|
|
|
|
#endif
|
2001-04-10 05:04:59 +04:00
|
|
|
|
2002-09-15 03:17:55 +04:00
|
|
|
|
|
|
|
#if BX_SUPPORT_X86_64==0
|
|
|
|
// Make life easier for merging code.
|
|
|
|
#define RAX EAX
|
|
|
|
#define RBX EBX
|
|
|
|
#define RCX ECX
|
|
|
|
#define RDX EDX
|
|
|
|
#endif
|
|
|
|
|
|
|
|
|
|
|
|
|
2001-04-10 05:04:59 +04:00
|
|
|
void
|
2002-09-18 02:50:53 +04:00
|
|
|
BX_CPU_C::UndefinedOpcode(bxInstruction_c *i)
|
2001-04-10 05:04:59 +04:00
|
|
|
{
|
2002-10-16 21:37:35 +04:00
|
|
|
BX_DEBUG(("UndefinedOpcode: %02x causes exception 6", (unsigned) i->b1()));
|
2002-04-11 05:19:24 +04:00
|
|
|
exception(BX_UD_EXCEPTION, 0, 0);
|
2001-04-10 05:04:59 +04:00
|
|
|
}
|
|
|
|
|
|
|
|
void
|
2002-09-18 02:50:53 +04:00
|
|
|
BX_CPU_C::NOP(bxInstruction_c *i)
|
2001-04-10 05:04:59 +04:00
|
|
|
{
|
2002-10-16 21:37:35 +04:00
|
|
|
}
|
|
|
|
|
|
|
|
void BX_CPU_C::PREFETCH(bxInstruction_c *i)
|
|
|
|
{
|
2003-05-15 20:41:17 +04:00
|
|
|
#if BX_SUPPORT_3DNOW || BX_SUPPORT_SSE >= 1
|
2003-02-13 18:04:11 +03:00
|
|
|
BX_INSTR_PREFETCH_HINT(BX_CPU_ID, i->nnn(), i->seg(), RMAddr(i));
|
2002-10-16 21:37:35 +04:00
|
|
|
#else
|
|
|
|
UndefinedOpcode(i);
|
|
|
|
#endif
|
2001-04-10 05:04:59 +04:00
|
|
|
}
|
|
|
|
|
|
|
|
void
|
2002-09-18 02:50:53 +04:00
|
|
|
BX_CPU_C::HLT(bxInstruction_c *i)
|
2001-04-10 05:04:59 +04:00
|
|
|
{
|
|
|
|
// hack to panic if HLT comes from BIOS
|
|
|
|
if ( BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].selector.value == 0xf000 )
|
2001-11-18 19:32:40 +03:00
|
|
|
BX_PANIC(("HALT instruction encountered in the BIOS ROM"));
|
2001-04-10 05:04:59 +04:00
|
|
|
|
|
|
|
if (CPL!=0) {
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_INFO(("HLT(): CPL!=0"));
|
2002-04-11 05:19:24 +04:00
|
|
|
exception(BX_GP_EXCEPTION, 0, 0);
|
2001-04-10 05:04:59 +04:00
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
2002-09-12 22:10:46 +04:00
|
|
|
if ( ! BX_CPU_THIS_PTR get_IF () ) {
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_INFO(("WARNING: HLT instruction with IF=0!"));
|
2001-04-10 05:04:59 +04:00
|
|
|
}
|
|
|
|
|
|
|
|
// stops instruction execution and places the processor in a
|
|
|
|
// HALT state. An enabled interrupt, NMI, or reset will resume
|
|
|
|
// execution. If interrupt (including NMI) is used to resume
|
|
|
|
// execution after HLT, the saved CS:eIP points to instruction
|
|
|
|
// following HLT.
|
|
|
|
|
|
|
|
// artificial trap bit, why use another variable.
|
|
|
|
BX_CPU_THIS_PTR debug_trap |= 0x80000000; // artificial trap
|
|
|
|
BX_CPU_THIS_PTR async_event = 1; // so processor knows to check
|
|
|
|
// Execution of this instruction completes. The processor
|
|
|
|
// will remain in a halt state until one of the above conditions
|
|
|
|
// is met.
|
2001-11-12 03:45:09 +03:00
|
|
|
|
|
|
|
#if BX_USE_IDLE_HACK
|
2002-10-28 01:26:34 +03:00
|
|
|
bx_gui->sim_is_idle ();
|
2001-11-12 03:45:09 +03:00
|
|
|
#endif /* BX_USE_IDLE_HACK */
|
2001-04-10 05:04:59 +04:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
void
|
2002-09-18 02:50:53 +04:00
|
|
|
BX_CPU_C::CLTS(bxInstruction_c *i)
|
2001-04-10 05:04:59 +04:00
|
|
|
{
|
|
|
|
#if BX_CPU_LEVEL < 2
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_PANIC(("CLTS: not implemented for < 286"));
|
2001-04-10 05:04:59 +04:00
|
|
|
#else
|
|
|
|
|
2001-05-30 22:56:02 +04:00
|
|
|
if (v8086_mode()) BX_PANIC(("clts: v8086 mode unsupported"));
|
2001-04-10 05:04:59 +04:00
|
|
|
|
|
|
|
/* read errata file */
|
|
|
|
// does CLTS also clear NT flag???
|
|
|
|
|
|
|
|
// #GP(0) if CPL is not 0
|
|
|
|
if (CPL!=0) {
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_INFO(("CLTS(): CPL!=0"));
|
2002-04-11 05:19:24 +04:00
|
|
|
exception(BX_GP_EXCEPTION, 0, 0);
|
2001-04-10 05:04:59 +04:00
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
BX_CPU_THIS_PTR cr0.ts = 0;
|
|
|
|
BX_CPU_THIS_PTR cr0.val32 &= ~0x08;
|
|
|
|
#endif
|
|
|
|
}
|
|
|
|
|
|
|
|
void
|
2002-09-18 02:50:53 +04:00
|
|
|
BX_CPU_C::INVD(bxInstruction_c *i)
|
2001-04-10 05:04:59 +04:00
|
|
|
{
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_INFO(("---------------"));
|
|
|
|
BX_INFO(("- INVD called -"));
|
|
|
|
BX_INFO(("---------------"));
|
2001-04-10 05:04:59 +04:00
|
|
|
|
|
|
|
#if BX_CPU_LEVEL >= 4
|
|
|
|
invalidate_prefetch_q();
|
|
|
|
|
|
|
|
if (BX_CPU_THIS_PTR cr0.pe) {
|
|
|
|
if (CPL!=0) {
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_INFO(("INVD: CPL!=0"));
|
2002-04-11 05:19:24 +04:00
|
|
|
exception(BX_GP_EXCEPTION, 0, 0);
|
2001-04-10 05:04:59 +04:00
|
|
|
}
|
|
|
|
}
|
2003-02-13 18:04:11 +03:00
|
|
|
BX_INSTR_CACHE_CNTRL(BX_CPU_ID, BX_INSTR_INVD);
|
2001-04-10 05:04:59 +04:00
|
|
|
#else
|
|
|
|
UndefinedOpcode(i);
|
|
|
|
#endif
|
|
|
|
}
|
|
|
|
|
|
|
|
void
|
2002-09-18 02:50:53 +04:00
|
|
|
BX_CPU_C::WBINVD(bxInstruction_c *i)
|
2001-04-10 05:04:59 +04:00
|
|
|
{
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_INFO(("WBINVD: (ignoring)"));
|
2001-04-10 05:04:59 +04:00
|
|
|
|
|
|
|
#if BX_CPU_LEVEL >= 4
|
|
|
|
invalidate_prefetch_q();
|
|
|
|
|
|
|
|
if (BX_CPU_THIS_PTR cr0.pe) {
|
|
|
|
if (CPL!=0) {
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_INFO(("WBINVD: CPL!=0"));
|
2002-04-11 05:19:24 +04:00
|
|
|
exception(BX_GP_EXCEPTION, 0, 0);
|
2001-04-10 05:04:59 +04:00
|
|
|
}
|
|
|
|
}
|
2003-02-13 18:04:11 +03:00
|
|
|
BX_INSTR_CACHE_CNTRL(BX_CPU_ID, BX_INSTR_WBINVD);
|
2001-04-10 05:04:59 +04:00
|
|
|
#else
|
|
|
|
UndefinedOpcode(i);
|
|
|
|
#endif
|
|
|
|
}
|
|
|
|
|
|
|
|
void
|
2002-09-18 02:50:53 +04:00
|
|
|
BX_CPU_C::MOV_DdRd(bxInstruction_c *i)
|
2001-04-10 05:04:59 +04:00
|
|
|
{
|
|
|
|
#if BX_CPU_LEVEL < 3
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_PANIC(("MOV_DdRd: not supported on < 386"));
|
2001-04-10 05:04:59 +04:00
|
|
|
#else
|
|
|
|
Bit32u val_32;
|
|
|
|
|
2001-05-30 22:56:02 +04:00
|
|
|
if (v8086_mode()) BX_PANIC(("MOV_DdRd: v8086 mode unsupported"));
|
2001-04-10 05:04:59 +04:00
|
|
|
|
|
|
|
/* NOTES:
|
|
|
|
* 32bit operands always used
|
|
|
|
* r/m field specifies general register
|
|
|
|
* mod field should always be 11 binary
|
|
|
|
* reg field specifies which special register
|
|
|
|
*/
|
|
|
|
|
2002-09-20 07:52:59 +04:00
|
|
|
if (!i->modC0()) {
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_PANIC(("MOV_DdRd(): rm field not a register!"));
|
2001-04-10 05:04:59 +04:00
|
|
|
}
|
|
|
|
|
|
|
|
invalidate_prefetch_q();
|
|
|
|
|
|
|
|
if (protected_mode() && CPL!=0) {
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_PANIC(("MOV_DdRd: CPL!=0"));
|
2001-04-10 05:04:59 +04:00
|
|
|
/* #GP(0) if CPL is not 0 */
|
2002-04-11 05:19:24 +04:00
|
|
|
exception(BX_GP_EXCEPTION, 0, 0);
|
2001-04-10 05:04:59 +04:00
|
|
|
}
|
|
|
|
|
2002-09-18 02:50:53 +04:00
|
|
|
val_32 = BX_READ_32BIT_REG(i->rm());
|
2001-04-10 05:04:59 +04:00
|
|
|
if (bx_dbg.dreg)
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_INFO(("MOV_DdRd: DR[%u]=%08xh unhandled",
|
2002-09-18 02:50:53 +04:00
|
|
|
(unsigned) i->nnn(), (unsigned) val_32));
|
2001-04-10 05:04:59 +04:00
|
|
|
|
2002-09-18 02:50:53 +04:00
|
|
|
switch (i->nnn()) {
|
2001-04-10 05:04:59 +04:00
|
|
|
case 0: // DR0
|
|
|
|
BX_CPU_THIS_PTR dr0 = val_32;
|
|
|
|
break;
|
|
|
|
case 1: // DR1
|
|
|
|
BX_CPU_THIS_PTR dr1 = val_32;
|
|
|
|
break;
|
|
|
|
case 2: // DR2
|
|
|
|
BX_CPU_THIS_PTR dr2 = val_32;
|
|
|
|
break;
|
|
|
|
case 3: // DR3
|
|
|
|
BX_CPU_THIS_PTR dr3 = val_32;
|
|
|
|
break;
|
|
|
|
|
|
|
|
case 4: // DR4
|
|
|
|
case 6: // DR6
|
|
|
|
// DR4 aliased to DR6 by default. With Debug Extensions on,
|
|
|
|
// access to DR4 causes #UD
|
|
|
|
#if BX_CPU_LEVEL >= 4
|
2002-09-18 02:50:53 +04:00
|
|
|
if ( (i->nnn() == 4) && (BX_CPU_THIS_PTR cr4.get_DE()) ) {
|
2001-04-10 05:04:59 +04:00
|
|
|
// Debug extensions on
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_INFO(("MOV_DdRd: access to DR4 causes #UD"));
|
2001-04-10 05:04:59 +04:00
|
|
|
UndefinedOpcode(i);
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
#if BX_CPU_LEVEL <= 4
|
|
|
|
// On 386/486 bit12 is settable
|
|
|
|
BX_CPU_THIS_PTR dr6 = (BX_CPU_THIS_PTR dr6 & 0xffff0ff0) |
|
|
|
|
(val_32 & 0x0000f00f);
|
|
|
|
#else
|
|
|
|
// On Pentium+, bit12 is always zero
|
|
|
|
BX_CPU_THIS_PTR dr6 = (BX_CPU_THIS_PTR dr6 & 0xffff0ff0) |
|
|
|
|
(val_32 & 0x0000e00f);
|
|
|
|
#endif
|
|
|
|
break;
|
|
|
|
|
|
|
|
case 5: // DR5
|
|
|
|
case 7: // DR7
|
|
|
|
// Note: 486+ ignore GE and LE flags. On the 386, exact
|
|
|
|
// data breakpoint matching does not occur unless it is enabled
|
|
|
|
// by setting the LE and/or GE flags.
|
|
|
|
|
|
|
|
// DR5 aliased to DR7 by default. With Debug Extensions on,
|
|
|
|
// access to DR5 causes #UD
|
|
|
|
#if BX_CPU_LEVEL >= 4
|
2002-09-18 02:50:53 +04:00
|
|
|
if ( (i->nnn() == 5) && (BX_CPU_THIS_PTR cr4.get_DE()) ) {
|
2001-04-10 05:04:59 +04:00
|
|
|
// Debug extensions (CR4.DE) on
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_INFO(("MOV_DdRd: access to DR5 causes #UD"));
|
2001-04-10 05:04:59 +04:00
|
|
|
UndefinedOpcode(i);
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
// Some sanity checks...
|
|
|
|
if ( val_32 & 0x00002000 ) {
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_PANIC(("MOV_DdRd: GD bit not supported yet"));
|
2001-04-10 05:04:59 +04:00
|
|
|
// Note: processor clears GD upon entering debug exception
|
|
|
|
// handler, to allow access to the debug registers
|
|
|
|
}
|
|
|
|
if ( (((val_32>>16) & 3)==2) ||
|
|
|
|
(((val_32>>20) & 3)==2) ||
|
|
|
|
(((val_32>>24) & 3)==2) ||
|
|
|
|
(((val_32>>28) & 3)==2) ) {
|
|
|
|
// IO breakpoints (10b) are not yet supported.
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_PANIC(("MOV_DdRd: write of %08x contains IO breakpoint",
|
merge in BRANCH-io-cleanup.
To see the commit logs for this use either cvsweb or
cvs update -r BRANCH-io-cleanup and then 'cvs log' the various files.
In general this provides a generic interface for logging.
logfunctions:: is a class that is inherited by some classes, and also
. allocated as a standalone global called 'genlog'. All logging uses
. one of the ::info(), ::error(), ::ldebug(), ::panic() methods of this
. class through 'BX_INFO(), BX_ERROR(), BX_DEBUG(), BX_PANIC()' macros
. respectively.
.
. An example usage:
. BX_INFO(("Hello, World!\n"));
iofunctions:: is a class that is allocated once by default, and assigned
as the iofunction of each logfunctions instance. It is this class that
maintains the file descriptor and other output related code, at this
point using vfprintf(). At some future point, someone may choose to
write a gui 'console' for bochs to which messages would be redirected
simply by assigning a different iofunction class to the various logfunctions
objects.
More cleanup is coming, but this works for now. If you want to see alot
of debugging output, in main.cc, change onoff[LOGLEV_DEBUG]=0 to =1.
Comments, bugs, flames, to me: todd@fries.net
2001-05-15 18:49:57 +04:00
|
|
|
val_32));
|
2001-04-10 05:04:59 +04:00
|
|
|
}
|
|
|
|
if ( (((val_32>>18) & 3)==2) ||
|
|
|
|
(((val_32>>22) & 3)==2) ||
|
|
|
|
(((val_32>>26) & 3)==2) ||
|
|
|
|
(((val_32>>30) & 3)==2) ) {
|
|
|
|
// LEN0..3 contains undefined length specifier (10b)
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_PANIC(("MOV_DdRd: write of %08x contains undefined LENx",
|
merge in BRANCH-io-cleanup.
To see the commit logs for this use either cvsweb or
cvs update -r BRANCH-io-cleanup and then 'cvs log' the various files.
In general this provides a generic interface for logging.
logfunctions:: is a class that is inherited by some classes, and also
. allocated as a standalone global called 'genlog'. All logging uses
. one of the ::info(), ::error(), ::ldebug(), ::panic() methods of this
. class through 'BX_INFO(), BX_ERROR(), BX_DEBUG(), BX_PANIC()' macros
. respectively.
.
. An example usage:
. BX_INFO(("Hello, World!\n"));
iofunctions:: is a class that is allocated once by default, and assigned
as the iofunction of each logfunctions instance. It is this class that
maintains the file descriptor and other output related code, at this
point using vfprintf(). At some future point, someone may choose to
write a gui 'console' for bochs to which messages would be redirected
simply by assigning a different iofunction class to the various logfunctions
objects.
More cleanup is coming, but this works for now. If you want to see alot
of debugging output, in main.cc, change onoff[LOGLEV_DEBUG]=0 to =1.
Comments, bugs, flames, to me: todd@fries.net
2001-05-15 18:49:57 +04:00
|
|
|
val_32));
|
2001-04-10 05:04:59 +04:00
|
|
|
}
|
|
|
|
if ( ((((val_32>>16) & 3)==0) && (((val_32>>18) & 3)!=0)) ||
|
|
|
|
((((val_32>>20) & 3)==0) && (((val_32>>22) & 3)!=0)) ||
|
|
|
|
((((val_32>>24) & 3)==0) && (((val_32>>26) & 3)!=0)) ||
|
|
|
|
((((val_32>>28) & 3)==0) && (((val_32>>30) & 3)!=0)) ) {
|
|
|
|
// Instruction breakpoint with LENx not 00b (1-byte length)
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_PANIC(("MOV_DdRd: write of %08x, R/W=00b LEN!=00b",
|
merge in BRANCH-io-cleanup.
To see the commit logs for this use either cvsweb or
cvs update -r BRANCH-io-cleanup and then 'cvs log' the various files.
In general this provides a generic interface for logging.
logfunctions:: is a class that is inherited by some classes, and also
. allocated as a standalone global called 'genlog'. All logging uses
. one of the ::info(), ::error(), ::ldebug(), ::panic() methods of this
. class through 'BX_INFO(), BX_ERROR(), BX_DEBUG(), BX_PANIC()' macros
. respectively.
.
. An example usage:
. BX_INFO(("Hello, World!\n"));
iofunctions:: is a class that is allocated once by default, and assigned
as the iofunction of each logfunctions instance. It is this class that
maintains the file descriptor and other output related code, at this
point using vfprintf(). At some future point, someone may choose to
write a gui 'console' for bochs to which messages would be redirected
simply by assigning a different iofunction class to the various logfunctions
objects.
More cleanup is coming, but this works for now. If you want to see alot
of debugging output, in main.cc, change onoff[LOGLEV_DEBUG]=0 to =1.
Comments, bugs, flames, to me: todd@fries.net
2001-05-15 18:49:57 +04:00
|
|
|
val_32));
|
2001-04-10 05:04:59 +04:00
|
|
|
}
|
|
|
|
#if BX_CPU_LEVEL <= 4
|
|
|
|
// 386/486: you can play with all the bits except b10 is always 1
|
|
|
|
BX_CPU_THIS_PTR dr7 = val_32 | 0x00000400;
|
|
|
|
#else
|
|
|
|
// Pentium+: bits15,14,12 are hardwired to 0, rest are settable.
|
|
|
|
// Even bits 11,10 are changeable though reserved.
|
|
|
|
BX_CPU_THIS_PTR dr7 = (val_32 & 0xffff2fff) | 0x00000400;
|
|
|
|
#endif
|
|
|
|
break;
|
|
|
|
default:
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_PANIC(("MOV_DdRd: control register index out of range"));
|
2001-04-10 05:04:59 +04:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
}
|
|
|
|
|
|
|
|
void
|
2002-09-18 02:50:53 +04:00
|
|
|
BX_CPU_C::MOV_RdDd(bxInstruction_c *i)
|
2001-04-10 05:04:59 +04:00
|
|
|
{
|
|
|
|
#if BX_CPU_LEVEL < 3
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_PANIC(("MOV_RdDd: not supported on < 386"));
|
2001-04-10 05:04:59 +04:00
|
|
|
#else
|
|
|
|
Bit32u val_32;
|
|
|
|
|
|
|
|
if (v8086_mode()) {
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_INFO(("MOV_RdDd: v8086 mode causes #GP"));
|
2002-04-11 05:19:24 +04:00
|
|
|
exception(BX_GP_EXCEPTION, 0, 0);
|
2001-04-10 05:04:59 +04:00
|
|
|
}
|
|
|
|
|
2002-09-20 07:52:59 +04:00
|
|
|
if (!i->modC0()) {
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_PANIC(("MOV_RdDd(): rm field not a register!"));
|
2001-04-10 05:04:59 +04:00
|
|
|
UndefinedOpcode(i);
|
|
|
|
}
|
|
|
|
|
|
|
|
if (protected_mode() && (CPL!=0)) {
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_INFO(("MOV_RdDd: CPL!=0 causes #GP"));
|
2002-04-11 05:19:24 +04:00
|
|
|
exception(BX_GP_EXCEPTION, 0, 0);
|
2001-04-10 05:04:59 +04:00
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (bx_dbg.dreg)
|
2002-09-18 02:50:53 +04:00
|
|
|
BX_INFO(("MOV_RdDd: DR%u not implemented yet", i->nnn()));
|
2001-04-10 05:04:59 +04:00
|
|
|
|
2002-09-18 02:50:53 +04:00
|
|
|
switch (i->nnn()) {
|
2001-04-10 05:04:59 +04:00
|
|
|
case 0: // DR0
|
|
|
|
val_32 = BX_CPU_THIS_PTR dr0;
|
|
|
|
break;
|
|
|
|
case 1: // DR1
|
|
|
|
val_32 = BX_CPU_THIS_PTR dr1;
|
|
|
|
break;
|
|
|
|
case 2: // DR2
|
|
|
|
val_32 = BX_CPU_THIS_PTR dr2;
|
|
|
|
break;
|
|
|
|
case 3: // DR3
|
|
|
|
val_32 = BX_CPU_THIS_PTR dr3;
|
|
|
|
break;
|
|
|
|
|
|
|
|
case 4: // DR4
|
|
|
|
case 6: // DR6
|
|
|
|
// DR4 aliased to DR6 by default. With Debug Extensions on,
|
|
|
|
// access to DR4 causes #UD
|
|
|
|
#if BX_CPU_LEVEL >= 4
|
2002-09-18 02:50:53 +04:00
|
|
|
if ( (i->nnn() == 4) && (BX_CPU_THIS_PTR cr4.get_DE()) ) {
|
2001-04-10 05:04:59 +04:00
|
|
|
// Debug extensions on
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_INFO(("MOV_RdDd: access to DR4 causes #UD"));
|
2001-04-10 05:04:59 +04:00
|
|
|
UndefinedOpcode(i);
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
val_32 = BX_CPU_THIS_PTR dr6;
|
|
|
|
break;
|
|
|
|
|
|
|
|
case 5: // DR5
|
|
|
|
case 7: // DR7
|
|
|
|
// DR5 aliased to DR7 by default. With Debug Extensions on,
|
|
|
|
// access to DR5 causes #UD
|
|
|
|
#if BX_CPU_LEVEL >= 4
|
2002-09-18 02:50:53 +04:00
|
|
|
if ( (i->nnn() == 5) && (BX_CPU_THIS_PTR cr4.get_DE()) ) {
|
2001-04-10 05:04:59 +04:00
|
|
|
// Debug extensions on
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_INFO(("MOV_RdDd: access to DR5 causes #UD"));
|
2001-04-10 05:04:59 +04:00
|
|
|
UndefinedOpcode(i);
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
val_32 = BX_CPU_THIS_PTR dr7;
|
|
|
|
break;
|
|
|
|
|
|
|
|
default:
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_PANIC(("MOV_RdDd: control register index out of range"));
|
2001-04-10 05:04:59 +04:00
|
|
|
val_32 = 0;
|
|
|
|
}
|
2002-09-18 02:50:53 +04:00
|
|
|
BX_WRITE_32BIT_REGZ(i->rm(), val_32);
|
2001-04-10 05:04:59 +04:00
|
|
|
#endif
|
|
|
|
}
|
|
|
|
|
2002-09-15 03:17:55 +04:00
|
|
|
#if BX_SUPPORT_X86_64
|
|
|
|
void
|
2002-09-18 02:50:53 +04:00
|
|
|
BX_CPU_C::MOV_DqRq(bxInstruction_c *i)
|
2002-09-15 03:17:55 +04:00
|
|
|
{
|
|
|
|
Bit64u val_64;
|
|
|
|
|
|
|
|
if (v8086_mode()) BX_PANIC(("MOV_DqRq: v8086 mode unsupported"));
|
|
|
|
|
|
|
|
/* NOTES:
|
|
|
|
* 64bit operands always used
|
|
|
|
* r/m field specifies general register
|
|
|
|
* mod field should always be 11 binary
|
|
|
|
* reg field specifies which special register
|
|
|
|
*/
|
|
|
|
|
2002-09-20 07:52:59 +04:00
|
|
|
if (!i->modC0()) {
|
2002-09-15 03:17:55 +04:00
|
|
|
BX_PANIC(("MOV_DqRq(): rm field not a register!"));
|
|
|
|
}
|
|
|
|
|
|
|
|
invalidate_prefetch_q();
|
|
|
|
|
|
|
|
if (protected_mode() && CPL!=0) {
|
|
|
|
BX_PANIC(("MOV_DqRq: CPL!=0"));
|
|
|
|
/* #GP(0) if CPL is not 0 */
|
|
|
|
exception(BX_GP_EXCEPTION, 0, 0);
|
|
|
|
}
|
|
|
|
|
2002-09-18 02:50:53 +04:00
|
|
|
val_64 = BX_READ_64BIT_REG(i->rm());
|
2002-09-15 03:17:55 +04:00
|
|
|
if (bx_dbg.dreg)
|
|
|
|
BX_INFO(("MOV_DqRq: DR[%u]=%08xh unhandled",
|
2002-09-18 02:50:53 +04:00
|
|
|
(unsigned) i->nnn(), (unsigned) val_64));
|
2002-09-15 03:17:55 +04:00
|
|
|
|
2002-09-18 02:50:53 +04:00
|
|
|
switch (i->nnn()) {
|
2002-09-15 03:17:55 +04:00
|
|
|
case 0: // DR0
|
|
|
|
BX_CPU_THIS_PTR dr0 = val_64;
|
|
|
|
break;
|
|
|
|
case 1: // DR1
|
|
|
|
BX_CPU_THIS_PTR dr1 = val_64;
|
|
|
|
break;
|
|
|
|
case 2: // DR2
|
|
|
|
BX_CPU_THIS_PTR dr2 = val_64;
|
|
|
|
break;
|
|
|
|
case 3: // DR3
|
|
|
|
BX_CPU_THIS_PTR dr3 = val_64;
|
|
|
|
break;
|
|
|
|
|
|
|
|
case 4: // DR4
|
|
|
|
case 6: // DR6
|
|
|
|
// DR4 aliased to DR6 by default. With Debug Extensions on,
|
|
|
|
// access to DR4 causes #UD
|
|
|
|
#if BX_CPU_LEVEL >= 4
|
2002-09-18 02:50:53 +04:00
|
|
|
if ( (i->nnn() == 4) && (BX_CPU_THIS_PTR cr4.get_DE()) ) {
|
2002-09-15 03:17:55 +04:00
|
|
|
// Debug extensions on
|
|
|
|
BX_INFO(("MOV_DqRq: access to DR4 causes #UD"));
|
|
|
|
UndefinedOpcode(i);
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
#if BX_CPU_LEVEL <= 4
|
|
|
|
// On 386/486 bit12 is settable
|
|
|
|
BX_CPU_THIS_PTR dr6 = (BX_CPU_THIS_PTR dr6 & 0xffff0ff0) |
|
|
|
|
(val_64 & 0x0000f00f);
|
|
|
|
#else
|
|
|
|
// On Pentium+, bit12 is always zero
|
|
|
|
BX_CPU_THIS_PTR dr6 = (BX_CPU_THIS_PTR dr6 & 0xffff0ff0) |
|
|
|
|
(val_64 & 0x0000e00f);
|
|
|
|
#endif
|
|
|
|
break;
|
|
|
|
|
|
|
|
case 5: // DR5
|
|
|
|
case 7: // DR7
|
|
|
|
// Note: 486+ ignore GE and LE flags. On the 386, exact
|
|
|
|
// data breakpoint matching does not occur unless it is enabled
|
|
|
|
// by setting the LE and/or GE flags.
|
|
|
|
|
|
|
|
// DR5 aliased to DR7 by default. With Debug Extensions on,
|
|
|
|
// access to DR5 causes #UD
|
|
|
|
#if BX_CPU_LEVEL >= 4
|
2002-09-18 02:50:53 +04:00
|
|
|
if ( (i->nnn() == 5) && (BX_CPU_THIS_PTR cr4.get_DE()) ) {
|
2002-09-15 03:17:55 +04:00
|
|
|
// Debug extensions (CR4.DE) on
|
|
|
|
BX_INFO(("MOV_DqRq: access to DR5 causes #UD"));
|
|
|
|
UndefinedOpcode(i);
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
// Some sanity checks...
|
|
|
|
if ( val_64 & 0x00002000 ) {
|
|
|
|
BX_PANIC(("MOV_DqRq: GD bit not supported yet"));
|
|
|
|
// Note: processor clears GD upon entering debug exception
|
|
|
|
// handler, to allow access to the debug registers
|
|
|
|
}
|
|
|
|
if ( (((val_64>>16) & 3)==2) ||
|
|
|
|
(((val_64>>20) & 3)==2) ||
|
|
|
|
(((val_64>>24) & 3)==2) ||
|
|
|
|
(((val_64>>28) & 3)==2) ) {
|
|
|
|
// IO breakpoints (10b) are not yet supported.
|
|
|
|
BX_PANIC(("MOV_DqRq: write of %08x contains IO breakpoint",
|
|
|
|
val_64));
|
|
|
|
}
|
|
|
|
if ( (((val_64>>18) & 3)==2) ||
|
|
|
|
(((val_64>>22) & 3)==2) ||
|
|
|
|
(((val_64>>26) & 3)==2) ||
|
|
|
|
(((val_64>>30) & 3)==2) ) {
|
|
|
|
// LEN0..3 contains undefined length specifier (10b)
|
|
|
|
BX_PANIC(("MOV_DqRq: write of %08x contains undefined LENx",
|
|
|
|
val_64));
|
|
|
|
}
|
|
|
|
if ( ((((val_64>>16) & 3)==0) && (((val_64>>18) & 3)!=0)) ||
|
|
|
|
((((val_64>>20) & 3)==0) && (((val_64>>22) & 3)!=0)) ||
|
|
|
|
((((val_64>>24) & 3)==0) && (((val_64>>26) & 3)!=0)) ||
|
|
|
|
((((val_64>>28) & 3)==0) && (((val_64>>30) & 3)!=0)) ) {
|
|
|
|
// Instruction breakpoint with LENx not 00b (1-byte length)
|
|
|
|
BX_PANIC(("MOV_DqRq: write of %08x, R/W=00b LEN!=00b",
|
|
|
|
val_64));
|
|
|
|
}
|
|
|
|
#if BX_CPU_LEVEL <= 4
|
|
|
|
// 386/486: you can play with all the bits except b10 is always 1
|
|
|
|
BX_CPU_THIS_PTR dr7 = val_64 | 0x00000400;
|
|
|
|
#else
|
|
|
|
// Pentium+: bits15,14,12 are hardwired to 0, rest are settable.
|
|
|
|
// Even bits 11,10 are changeable though reserved.
|
|
|
|
BX_CPU_THIS_PTR dr7 = (val_64 & 0xffff2fff) | 0x00000400;
|
|
|
|
#endif
|
|
|
|
break;
|
|
|
|
default:
|
|
|
|
BX_PANIC(("MOV_DqRq: control register index out of range"));
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
#endif // #if BX_SUPPORT_X86_64
|
|
|
|
|
|
|
|
#if BX_SUPPORT_X86_64
|
|
|
|
void
|
2002-09-18 02:50:53 +04:00
|
|
|
BX_CPU_C::MOV_RqDq(bxInstruction_c *i)
|
2002-09-15 03:17:55 +04:00
|
|
|
{
|
|
|
|
Bit64u val_64;
|
|
|
|
|
|
|
|
if (v8086_mode()) {
|
|
|
|
BX_INFO(("MOV_RqDq: v8086 mode causes #GP"));
|
|
|
|
exception(BX_GP_EXCEPTION, 0, 0);
|
|
|
|
}
|
|
|
|
|
2002-09-20 07:52:59 +04:00
|
|
|
if (!i->modC0()) {
|
2002-09-15 03:17:55 +04:00
|
|
|
BX_PANIC(("MOV_RqDq(): rm field not a register!"));
|
|
|
|
UndefinedOpcode(i);
|
|
|
|
}
|
|
|
|
|
|
|
|
if (protected_mode() && (CPL!=0)) {
|
|
|
|
BX_INFO(("MOV_RqDq: CPL!=0 causes #GP"));
|
|
|
|
exception(BX_GP_EXCEPTION, 0, 0);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (bx_dbg.dreg)
|
2002-09-18 02:50:53 +04:00
|
|
|
BX_INFO(("MOV_RqDq: DR%u not implemented yet", i->nnn()));
|
2002-09-15 03:17:55 +04:00
|
|
|
|
2002-09-18 02:50:53 +04:00
|
|
|
switch (i->nnn()) {
|
2002-09-15 03:17:55 +04:00
|
|
|
case 0: // DR0
|
|
|
|
val_64 = BX_CPU_THIS_PTR dr0;
|
|
|
|
break;
|
|
|
|
case 1: // DR1
|
|
|
|
val_64 = BX_CPU_THIS_PTR dr1;
|
|
|
|
break;
|
|
|
|
case 2: // DR2
|
|
|
|
val_64 = BX_CPU_THIS_PTR dr2;
|
|
|
|
break;
|
|
|
|
case 3: // DR3
|
|
|
|
val_64 = BX_CPU_THIS_PTR dr3;
|
|
|
|
break;
|
|
|
|
|
|
|
|
case 4: // DR4
|
|
|
|
case 6: // DR6
|
|
|
|
// DR4 aliased to DR6 by default. With Debug Extensions on,
|
|
|
|
// access to DR4 causes #UD
|
|
|
|
#if BX_CPU_LEVEL >= 4
|
2002-09-18 02:50:53 +04:00
|
|
|
if ( (i->nnn() == 4) && (BX_CPU_THIS_PTR cr4.get_DE()) ) {
|
2002-09-15 03:17:55 +04:00
|
|
|
// Debug extensions on
|
|
|
|
BX_INFO(("MOV_RqDq: access to DR4 causes #UD"));
|
|
|
|
UndefinedOpcode(i);
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
val_64 = BX_CPU_THIS_PTR dr6;
|
|
|
|
break;
|
|
|
|
|
|
|
|
case 5: // DR5
|
|
|
|
case 7: // DR7
|
|
|
|
// DR5 aliased to DR7 by default. With Debug Extensions on,
|
|
|
|
// access to DR5 causes #UD
|
|
|
|
#if BX_CPU_LEVEL >= 4
|
2002-09-18 02:50:53 +04:00
|
|
|
if ( (i->nnn() == 5) && (BX_CPU_THIS_PTR cr4.get_DE()) ) {
|
2002-09-15 03:17:55 +04:00
|
|
|
// Debug extensions on
|
|
|
|
BX_INFO(("MOV_RqDq: access to DR5 causes #UD"));
|
|
|
|
UndefinedOpcode(i);
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
val_64 = BX_CPU_THIS_PTR dr7;
|
|
|
|
break;
|
|
|
|
|
|
|
|
default:
|
|
|
|
BX_PANIC(("MOV_RqDq: control register index out of range"));
|
|
|
|
val_64 = 0;
|
|
|
|
}
|
2002-09-18 02:50:53 +04:00
|
|
|
BX_WRITE_64BIT_REG(i->rm(), val_64);
|
2002-09-15 03:17:55 +04:00
|
|
|
}
|
|
|
|
#endif // #if BX_SUPPORT_X86_64
|
|
|
|
|
|
|
|
|
2001-04-10 05:04:59 +04:00
|
|
|
|
|
|
|
void
|
2002-09-18 02:50:53 +04:00
|
|
|
BX_CPU_C::LMSW_Ew(bxInstruction_c *i)
|
2001-04-10 05:04:59 +04:00
|
|
|
{
|
|
|
|
#if BX_CPU_LEVEL < 2
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_PANIC(("LMSW_Ew(): not supported on 8086!"));
|
2001-04-10 05:04:59 +04:00
|
|
|
#else
|
|
|
|
Bit16u msw;
|
|
|
|
Bit32u cr0;
|
|
|
|
|
2002-09-07 09:21:28 +04:00
|
|
|
invalidate_prefetch_q();
|
|
|
|
|
2002-08-10 16:06:26 +04:00
|
|
|
if (v8086_mode()) BX_PANIC(("proc_ctrl: LMSW in v8086 mode unsupported"));
|
2001-04-10 05:04:59 +04:00
|
|
|
|
|
|
|
if ( protected_mode() ) {
|
|
|
|
if ( CPL != 0 ) {
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_INFO(("LMSW: CPL != 0, CPL=%u", (unsigned) CPL));
|
2002-04-11 05:19:24 +04:00
|
|
|
exception(BX_GP_EXCEPTION, 0, 0);
|
2001-04-10 05:04:59 +04:00
|
|
|
return;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2002-09-20 07:52:59 +04:00
|
|
|
if (i->modC0()) {
|
2002-09-18 02:50:53 +04:00
|
|
|
msw = BX_READ_16BIT_REG(i->rm());
|
2001-04-10 05:04:59 +04:00
|
|
|
}
|
|
|
|
else {
|
2002-09-18 09:36:48 +04:00
|
|
|
read_virtual_word(i->seg(), RMAddr(i), &msw);
|
2001-04-10 05:04:59 +04:00
|
|
|
}
|
|
|
|
|
|
|
|
// LMSW does not affect PG,CD,NW,AM,WP,NE,ET bits, and cannot clear PE
|
|
|
|
|
|
|
|
// LMSW cannot clear PE
|
|
|
|
if ( ((msw & 0x0001)==0) && BX_CPU_THIS_PTR cr0.pe ) {
|
|
|
|
msw |= 0x0001; // adjust PE bit to current value of 1
|
|
|
|
}
|
|
|
|
|
|
|
|
msw &= 0x000f; // LMSW only affects last 4 flags
|
|
|
|
cr0 = (BX_CPU_THIS_PTR cr0.val32 & 0xfffffff0) | msw;
|
|
|
|
SetCR0(cr0);
|
|
|
|
#endif /* BX_CPU_LEVEL < 2 */
|
|
|
|
}
|
|
|
|
|
|
|
|
void
|
2002-09-18 02:50:53 +04:00
|
|
|
BX_CPU_C::SMSW_Ew(bxInstruction_c *i)
|
2001-04-10 05:04:59 +04:00
|
|
|
{
|
|
|
|
#if BX_CPU_LEVEL < 2
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_PANIC(("SMSW_Ew: not supported yet!"));
|
2001-04-10 05:04:59 +04:00
|
|
|
#else
|
|
|
|
Bit16u msw;
|
|
|
|
|
|
|
|
|
|
|
|
#if BX_CPU_LEVEL == 2
|
|
|
|
msw = 0xfff0; /* 80286 init value */
|
|
|
|
msw |= (BX_CPU_THIS_PTR cr0.ts << 3) |
|
|
|
|
(BX_CPU_THIS_PTR cr0.em << 2) |
|
|
|
|
(BX_CPU_THIS_PTR cr0.mp << 1) |
|
|
|
|
BX_CPU_THIS_PTR cr0.pe;
|
|
|
|
#else /* 386+ */
|
|
|
|
/* reserved bits 0 ??? */
|
|
|
|
/* should NE bit be included here ??? */
|
|
|
|
// should ET bit be included here (AW)
|
|
|
|
msw = (BX_CPU_THIS_PTR cr0.ts << 3) |
|
|
|
|
(BX_CPU_THIS_PTR cr0.em << 2) |
|
|
|
|
(BX_CPU_THIS_PTR cr0.mp << 1) |
|
|
|
|
BX_CPU_THIS_PTR cr0.pe;
|
|
|
|
#endif
|
|
|
|
|
|
|
|
|
2002-09-20 07:52:59 +04:00
|
|
|
if (i->modC0()) {
|
2002-09-18 09:36:48 +04:00
|
|
|
if (i->os32L()) {
|
2002-09-18 02:50:53 +04:00
|
|
|
BX_WRITE_32BIT_REGZ(i->rm(), msw); // zeros out high 16bits
|
2001-04-10 05:04:59 +04:00
|
|
|
}
|
|
|
|
else {
|
2002-09-18 02:50:53 +04:00
|
|
|
BX_WRITE_16BIT_REG(i->rm(), msw);
|
2001-04-10 05:04:59 +04:00
|
|
|
}
|
|
|
|
}
|
|
|
|
else {
|
2002-09-18 09:36:48 +04:00
|
|
|
write_virtual_word(i->seg(), RMAddr(i), &msw);
|
2001-04-10 05:04:59 +04:00
|
|
|
}
|
|
|
|
|
|
|
|
#endif
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
void
|
2002-09-18 02:50:53 +04:00
|
|
|
BX_CPU_C::MOV_CdRd(bxInstruction_c *i)
|
2001-04-10 05:04:59 +04:00
|
|
|
{
|
|
|
|
// mov general register data to control register
|
|
|
|
#if BX_CPU_LEVEL < 3
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_PANIC(("MOV_CdRd: not supported on < 386"));
|
2001-04-10 05:04:59 +04:00
|
|
|
#else
|
|
|
|
Bit32u val_32;
|
|
|
|
|
|
|
|
|
2002-08-10 16:06:26 +04:00
|
|
|
/* if (v8086_mode()) BX_PANIC(("proc_ctrl: MOV_CdRd in v8086 mode unsupported"));*/
|
2001-04-10 05:04:59 +04:00
|
|
|
|
|
|
|
/* NOTES:
|
|
|
|
* 32bit operands always used
|
|
|
|
* r/m field specifies general register
|
|
|
|
* mod field should always be 11 binary
|
|
|
|
* reg field specifies which special register
|
|
|
|
*/
|
|
|
|
|
2002-09-20 07:52:59 +04:00
|
|
|
if (!i->modC0()) {
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_PANIC(("MOV_CdRd(): rm field not a register!"));
|
2001-04-10 05:04:59 +04:00
|
|
|
}
|
|
|
|
|
|
|
|
invalidate_prefetch_q();
|
|
|
|
|
2002-08-10 16:06:26 +04:00
|
|
|
if ((protected_mode() || v8086_mode()) && CPL!=0) {
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_PANIC(("MOV_CdRd: CPL!=0"));
|
2001-04-10 05:04:59 +04:00
|
|
|
/* #GP(0) if CPL is not 0 */
|
2002-04-11 05:19:24 +04:00
|
|
|
exception(BX_GP_EXCEPTION, 0, 0);
|
2001-04-10 05:04:59 +04:00
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
2002-09-18 02:50:53 +04:00
|
|
|
val_32 = BX_READ_32BIT_REG(i->rm());
|
2001-04-10 05:04:59 +04:00
|
|
|
|
2002-09-18 02:50:53 +04:00
|
|
|
switch (i->nnn()) {
|
2001-04-10 05:04:59 +04:00
|
|
|
case 0: // CR0 (MSW)
|
2002-09-15 03:17:55 +04:00
|
|
|
// BX_INFO(("MOV_CdRd:CR0: R32 = %08x @CS:EIP %04x:%04x ",
|
2001-04-10 05:04:59 +04:00
|
|
|
// (unsigned) val_32,
|
|
|
|
// (unsigned) BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].selector.value,
|
2002-09-13 04:15:23 +04:00
|
|
|
// (unsigned) EIP));
|
2001-04-10 05:04:59 +04:00
|
|
|
SetCR0(val_32);
|
|
|
|
break;
|
|
|
|
|
|
|
|
case 1: /* CR1 */
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_PANIC(("MOV_CdRd: CR1 not implemented yet"));
|
2001-04-10 05:04:59 +04:00
|
|
|
break;
|
|
|
|
case 2: /* CR2 */
|
2001-08-25 01:02:37 +04:00
|
|
|
BX_DEBUG(("MOV_CdRd: CR2 not implemented yet"));
|
2002-10-04 21:04:33 +04:00
|
|
|
BX_DEBUG(("MOV_CdRd: CR2 = reg"));
|
2001-04-10 05:04:59 +04:00
|
|
|
BX_CPU_THIS_PTR cr2 = val_32;
|
|
|
|
break;
|
|
|
|
case 3: // CR3
|
|
|
|
if (bx_dbg.creg)
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_INFO(("MOV_CdRd:CR3 = %08x", (unsigned) val_32));
|
2001-04-10 05:04:59 +04:00
|
|
|
// Reserved bits take on value of MOV instruction
|
|
|
|
CR3_change(val_32);
|
2003-02-13 18:04:11 +03:00
|
|
|
BX_INSTR_TLB_CNTRL(BX_CPU_ID, BX_INSTR_MOV_CR3, val_32);
|
2002-09-07 09:21:28 +04:00
|
|
|
// Reload of CR3 always serializes.
|
|
|
|
// invalidate_prefetch_q(); // Already done.
|
2001-04-10 05:04:59 +04:00
|
|
|
break;
|
|
|
|
case 4: // CR4
|
Integrated patches for:
- Paging code rehash. You must now use --enable-4meg-pages to
use 4Meg pages, with the default of disabled, since we don't well
support 4Meg pages yet. Paging table walks model a real CPU
more closely now, and I fixed some bugs in the old logic.
- Segment check redundancy elimination. After a segment is loaded,
reads and writes are marked when a segment type check succeeds, and
they are skipped thereafter, when possible.
- Repeated IO and memory string copy acceleration. Only some variants
of instructions are available on all platforms, word and dword
variants only on x86 for the moment due to alignment and endian issues.
This is compiled in currently with no option - I should add a configure
option.
- Added a guest linear address to host TLB. Actually, I just stick
the host address (mem.vector[addr] address) in the upper 29 bits
of the field 'combined_access' since they are unused. Convenient
for now. I'm only storing page frame addresses. This was the
simplest for of such a TLB. We can likely enhance this. Also,
I only accelerated the normal read/write routines in access.cc.
Could also modify the read-modify-write versions too. You must
use --enable-guest2host-tlb, to try this out. Currently speeds
up Win95 boot time by about 3.5% for me. More ground to cover...
- Minor mods to CPUI/MOV_CdRd for CMOV.
- Integrated enhancements from Volker to getHostMemAddr() for PCI
being enabled.
2002-09-02 00:12:09 +04:00
|
|
|
{
|
2001-04-10 05:04:59 +04:00
|
|
|
#if BX_CPU_LEVEL == 3
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_PANIC(("MOV_CdRd: write to CR4 of 0x%08x on 386",
|
merge in BRANCH-io-cleanup.
To see the commit logs for this use either cvsweb or
cvs update -r BRANCH-io-cleanup and then 'cvs log' the various files.
In general this provides a generic interface for logging.
logfunctions:: is a class that is inherited by some classes, and also
. allocated as a standalone global called 'genlog'. All logging uses
. one of the ::info(), ::error(), ::ldebug(), ::panic() methods of this
. class through 'BX_INFO(), BX_ERROR(), BX_DEBUG(), BX_PANIC()' macros
. respectively.
.
. An example usage:
. BX_INFO(("Hello, World!\n"));
iofunctions:: is a class that is allocated once by default, and assigned
as the iofunction of each logfunctions instance. It is this class that
maintains the file descriptor and other output related code, at this
point using vfprintf(). At some future point, someone may choose to
write a gui 'console' for bochs to which messages would be redirected
simply by assigning a different iofunction class to the various logfunctions
objects.
More cleanup is coming, but this works for now. If you want to see alot
of debugging output, in main.cc, change onoff[LOGLEV_DEBUG]=0 to =1.
Comments, bugs, flames, to me: todd@fries.net
2001-05-15 18:49:57 +04:00
|
|
|
val_32));
|
2001-04-10 05:04:59 +04:00
|
|
|
UndefinedOpcode(i);
|
|
|
|
#else
|
|
|
|
// Protected mode: #GP(0) if attempt to write a 1 to
|
|
|
|
// any reserved bit of CR4
|
|
|
|
|
2002-09-15 03:17:55 +04:00
|
|
|
SetCR4(val_32);
|
2001-04-10 05:04:59 +04:00
|
|
|
#endif
|
Integrated patches for:
- Paging code rehash. You must now use --enable-4meg-pages to
use 4Meg pages, with the default of disabled, since we don't well
support 4Meg pages yet. Paging table walks model a real CPU
more closely now, and I fixed some bugs in the old logic.
- Segment check redundancy elimination. After a segment is loaded,
reads and writes are marked when a segment type check succeeds, and
they are skipped thereafter, when possible.
- Repeated IO and memory string copy acceleration. Only some variants
of instructions are available on all platforms, word and dword
variants only on x86 for the moment due to alignment and endian issues.
This is compiled in currently with no option - I should add a configure
option.
- Added a guest linear address to host TLB. Actually, I just stick
the host address (mem.vector[addr] address) in the upper 29 bits
of the field 'combined_access' since they are unused. Convenient
for now. I'm only storing page frame addresses. This was the
simplest for of such a TLB. We can likely enhance this. Also,
I only accelerated the normal read/write routines in access.cc.
Could also modify the read-modify-write versions too. You must
use --enable-guest2host-tlb, to try this out. Currently speeds
up Win95 boot time by about 3.5% for me. More ground to cover...
- Minor mods to CPUI/MOV_CdRd for CMOV.
- Integrated enhancements from Volker to getHostMemAddr() for PCI
being enabled.
2002-09-02 00:12:09 +04:00
|
|
|
}
|
2001-04-10 05:04:59 +04:00
|
|
|
break;
|
|
|
|
default:
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_PANIC(("MOV_CdRd: control register index out of range"));
|
2001-04-10 05:04:59 +04:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
}
|
|
|
|
|
|
|
|
void
|
2002-09-18 02:50:53 +04:00
|
|
|
BX_CPU_C::MOV_RdCd(bxInstruction_c *i)
|
2001-04-10 05:04:59 +04:00
|
|
|
{
|
|
|
|
// mov control register data to register
|
|
|
|
#if BX_CPU_LEVEL < 3
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_PANIC(("MOV_RdCd: not supported on < 386"));
|
2001-04-10 05:04:59 +04:00
|
|
|
#else
|
|
|
|
Bit32u val_32;
|
|
|
|
|
2002-08-10 16:06:26 +04:00
|
|
|
/* if (v8086_mode()) BX_PANIC(("proc_ctrl: MOV_RdCd in v8086 mode unsupported"));*/
|
2001-04-10 05:04:59 +04:00
|
|
|
|
|
|
|
/* NOTES:
|
|
|
|
* 32bit operands always used
|
|
|
|
* r/m field specifies general register
|
|
|
|
* mod field should always be 11 binary
|
|
|
|
* reg field specifies which special register
|
|
|
|
*/
|
|
|
|
|
2002-09-20 07:52:59 +04:00
|
|
|
if (!i->modC0()) {
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_PANIC(("MOV_RdCd(): rm field not a register!"));
|
2001-04-10 05:04:59 +04:00
|
|
|
}
|
|
|
|
|
2002-08-10 16:06:26 +04:00
|
|
|
if ((protected_mode() || v8086_mode()) && CPL!=0) {
|
|
|
|
BX_INFO(("MOV_RdCd: CPL!=0"));
|
2002-04-11 05:19:24 +04:00
|
|
|
exception(BX_GP_EXCEPTION, 0, 0);
|
2001-04-10 05:04:59 +04:00
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
2002-09-18 02:50:53 +04:00
|
|
|
switch (i->nnn()) {
|
2001-04-10 05:04:59 +04:00
|
|
|
case 0: // CR0 (MSW)
|
|
|
|
val_32 = BX_CPU_THIS_PTR cr0.val32;
|
|
|
|
#if 0
|
2002-09-15 03:17:55 +04:00
|
|
|
BX_INFO(("MOV_RdCd:CR0: R32 = %08x @CS:EIP %04x:%04x",
|
2001-04-10 05:04:59 +04:00
|
|
|
(unsigned) val_32,
|
|
|
|
(unsigned) BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].selector.value,
|
2002-09-13 04:15:23 +04:00
|
|
|
(unsigned) EIP));
|
2001-04-10 05:04:59 +04:00
|
|
|
#endif
|
|
|
|
break;
|
|
|
|
case 1: /* CR1 */
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_PANIC(("MOV_RdCd: CR1 not implemented yet"));
|
2001-04-10 05:04:59 +04:00
|
|
|
val_32 = 0;
|
|
|
|
break;
|
|
|
|
case 2: /* CR2 */
|
|
|
|
if (bx_dbg.creg)
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_INFO(("MOV_RdCd: CR2"));
|
2001-04-10 05:04:59 +04:00
|
|
|
val_32 = BX_CPU_THIS_PTR cr2;
|
|
|
|
break;
|
|
|
|
case 3: // CR3
|
|
|
|
if (bx_dbg.creg)
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_INFO(("MOV_RdCd: reading CR3"));
|
2001-04-10 05:04:59 +04:00
|
|
|
val_32 = BX_CPU_THIS_PTR cr3;
|
|
|
|
break;
|
|
|
|
case 4: // CR4
|
|
|
|
#if BX_CPU_LEVEL == 3
|
|
|
|
val_32 = 0;
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_INFO(("MOV_RdCd: read of CR4 causes #UD"));
|
2001-04-10 05:04:59 +04:00
|
|
|
UndefinedOpcode(i);
|
|
|
|
#else
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_INFO(("MOV_RdCd: read of CR4"));
|
2002-09-14 23:21:41 +04:00
|
|
|
val_32 = BX_CPU_THIS_PTR cr4.getRegister();
|
2001-04-10 05:04:59 +04:00
|
|
|
#endif
|
|
|
|
break;
|
|
|
|
default:
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_PANIC(("MOV_RdCd: control register index out of range"));
|
2001-04-10 05:04:59 +04:00
|
|
|
val_32 = 0;
|
|
|
|
}
|
2002-09-18 02:50:53 +04:00
|
|
|
BX_WRITE_32BIT_REGZ(i->rm(), val_32);
|
2002-09-15 03:17:55 +04:00
|
|
|
#endif
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
#if BX_SUPPORT_X86_64
|
|
|
|
void
|
2002-09-18 02:50:53 +04:00
|
|
|
BX_CPU_C::MOV_CqRq(bxInstruction_c *i)
|
2002-09-15 03:17:55 +04:00
|
|
|
{
|
|
|
|
// mov general register data to control register
|
|
|
|
#if BX_CPU_LEVEL < 3
|
|
|
|
BX_PANIC(("MOV_CqRq: not supported on < 386"));
|
|
|
|
#else
|
|
|
|
Bit64u val_64;
|
|
|
|
|
|
|
|
|
|
|
|
if (v8086_mode()) BX_PANIC(("proc_ctrl: v8086 mode unsupported"));
|
|
|
|
|
|
|
|
/* NOTES:
|
|
|
|
* 64bit operands always used
|
|
|
|
* r/m field specifies general register
|
|
|
|
* mod field should always be 11 binary
|
|
|
|
* reg field specifies which special register
|
|
|
|
*/
|
|
|
|
|
2002-09-20 07:52:59 +04:00
|
|
|
if (!i->modC0()) {
|
2002-09-15 03:17:55 +04:00
|
|
|
BX_PANIC(("MOV_CqRq(): rm field not a register!"));
|
|
|
|
}
|
|
|
|
|
|
|
|
invalidate_prefetch_q();
|
|
|
|
|
|
|
|
if (protected_mode() && CPL!=0) {
|
|
|
|
BX_PANIC(("MOV_CqRq: CPL!=0"));
|
|
|
|
/* #GP(0) if CPL is not 0 */
|
|
|
|
exception(BX_GP_EXCEPTION, 0, 0);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
2002-09-18 02:50:53 +04:00
|
|
|
val_64 = BX_READ_64BIT_REG(i->rm());
|
2002-09-15 03:17:55 +04:00
|
|
|
|
2002-09-18 02:50:53 +04:00
|
|
|
switch (i->nnn()) {
|
2002-09-15 03:17:55 +04:00
|
|
|
case 0: // CR0 (MSW)
|
|
|
|
// BX_INFO(("MOV_CqRq:CR0: R64 = %08x @CS:EIP %04x:%04x ",
|
|
|
|
// (unsigned) val_64,
|
|
|
|
// (unsigned) BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].selector.value,
|
|
|
|
// (unsigned) EIP));
|
|
|
|
SetCR0(val_64);
|
|
|
|
break;
|
|
|
|
|
|
|
|
case 1: /* CR1 */
|
|
|
|
BX_PANIC(("MOV_CqRq: CR1 not implemented yet"));
|
|
|
|
break;
|
|
|
|
case 2: /* CR2 */
|
|
|
|
BX_DEBUG(("MOV_CqRq: CR2 not implemented yet"));
|
2002-10-04 21:04:33 +04:00
|
|
|
BX_DEBUG(("MOV_CqRq: CR2 = reg"));
|
2002-09-15 03:17:55 +04:00
|
|
|
BX_CPU_THIS_PTR cr2 = val_64;
|
|
|
|
break;
|
|
|
|
case 3: // CR3
|
|
|
|
if (bx_dbg.creg)
|
|
|
|
BX_INFO(("MOV_CqRq:CR3 = %08x", (unsigned) val_64));
|
|
|
|
// Reserved bits take on value of MOV instruction
|
|
|
|
CR3_change(val_64);
|
2003-02-13 18:04:11 +03:00
|
|
|
BX_INSTR_TLB_CNTRL(BX_CPU_ID, BX_INSTR_MOV_CR3, val_64);
|
2002-09-15 03:17:55 +04:00
|
|
|
break;
|
|
|
|
case 4: // CR4
|
|
|
|
#if BX_CPU_LEVEL == 3
|
|
|
|
BX_PANIC(("MOV_CqRq: write to CR4 of 0x%08x on 386",
|
|
|
|
val_64));
|
|
|
|
UndefinedOpcode(i);
|
|
|
|
#else
|
|
|
|
// Protected mode: #GP(0) if attempt to write a 1 to
|
|
|
|
// any reserved bit of CR4
|
|
|
|
|
|
|
|
BX_INFO(("MOV_CqRq: ignoring write to CR4 of 0x%08x",
|
|
|
|
val_64));
|
|
|
|
if (val_64) {
|
|
|
|
BX_INFO(("MOV_CqRq: (CR4) write of 0x%08x not supported!",
|
|
|
|
val_64));
|
|
|
|
}
|
|
|
|
// Only allow writes of 0 to CR4 for now.
|
|
|
|
// Writes to bits in CR4 should not be 1s as CPUID
|
|
|
|
// returns not-supported for all of these features.
|
|
|
|
SetCR4(val_64);
|
|
|
|
#endif
|
|
|
|
break;
|
|
|
|
default:
|
|
|
|
BX_PANIC(("MOV_CqRq: control register index out of range"));
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
}
|
|
|
|
#endif // #if BX_SUPPORT_X86_64
|
|
|
|
|
|
|
|
|
|
|
|
#if BX_SUPPORT_X86_64
|
|
|
|
void
|
2002-09-18 02:50:53 +04:00
|
|
|
BX_CPU_C::MOV_RqCq(bxInstruction_c *i)
|
2002-09-15 03:17:55 +04:00
|
|
|
{
|
|
|
|
// mov control register data to register
|
|
|
|
#if BX_CPU_LEVEL < 3
|
|
|
|
BX_PANIC(("MOV_RqCq: not supported on < 386"));
|
|
|
|
#else
|
|
|
|
Bit64u val_64;
|
|
|
|
|
|
|
|
if (v8086_mode()) BX_PANIC(("proc_ctrl: v8086 mode unsupported"));
|
|
|
|
|
|
|
|
/* NOTES:
|
|
|
|
* 64bit operands always used
|
|
|
|
* r/m field specifies general register
|
|
|
|
* mod field should always be 11 binary
|
|
|
|
* reg field specifies which special register
|
|
|
|
*/
|
|
|
|
|
2002-09-20 07:52:59 +04:00
|
|
|
if (!i->modC0()) {
|
2002-09-15 03:17:55 +04:00
|
|
|
BX_PANIC(("MOV_RqC(): rm field not a register!"));
|
|
|
|
}
|
|
|
|
|
|
|
|
if (protected_mode() && CPL!=0) {
|
|
|
|
BX_PANIC(("MOV_RqCq: CPL!=0"));
|
|
|
|
/* #GP(0) if CPL is not 0 */
|
|
|
|
exception(BX_GP_EXCEPTION, 0, 0);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
2002-09-18 02:50:53 +04:00
|
|
|
switch (i->nnn()) {
|
2002-09-15 03:17:55 +04:00
|
|
|
case 0: // CR0 (MSW)
|
|
|
|
val_64 = BX_CPU_THIS_PTR cr0.val32;
|
|
|
|
#if 0
|
|
|
|
BX_INFO(("MOV_RqCq:CR0: R64 = %08x @CS:EIP %04x:%04x",
|
|
|
|
(unsigned) val_64,
|
|
|
|
(unsigned) BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].selector.value,
|
|
|
|
(unsigned) EIP));
|
|
|
|
#endif
|
|
|
|
break;
|
|
|
|
case 1: /* CR1 */
|
|
|
|
BX_PANIC(("MOV_RqCq: CR1 not implemented yet"));
|
|
|
|
val_64 = 0;
|
|
|
|
break;
|
|
|
|
case 2: /* CR2 */
|
|
|
|
if (bx_dbg.creg)
|
|
|
|
BX_INFO(("MOV_RqCq: CR2"));
|
|
|
|
val_64 = BX_CPU_THIS_PTR cr2;
|
|
|
|
break;
|
|
|
|
case 3: // CR3
|
|
|
|
if (bx_dbg.creg)
|
|
|
|
BX_INFO(("MOV_RqCq: reading CR3"));
|
|
|
|
val_64 = BX_CPU_THIS_PTR cr3;
|
|
|
|
break;
|
|
|
|
case 4: // CR4
|
|
|
|
#if BX_CPU_LEVEL == 3
|
|
|
|
val_64 = 0;
|
|
|
|
BX_INFO(("MOV_RqCq: read of CR4 causes #UD"));
|
|
|
|
UndefinedOpcode(i);
|
|
|
|
#else
|
|
|
|
BX_INFO(("MOV_RqCq: read of CR4"));
|
|
|
|
val_64 = BX_CPU_THIS_PTR cr4.getRegister();
|
|
|
|
#endif
|
|
|
|
break;
|
|
|
|
default:
|
|
|
|
BX_PANIC(("MOV_RqCq: control register index out of range"));
|
|
|
|
val_64 = 0;
|
|
|
|
}
|
2002-09-18 02:50:53 +04:00
|
|
|
BX_WRITE_64BIT_REG(i->rm(), val_64);
|
2001-04-10 05:04:59 +04:00
|
|
|
#endif
|
|
|
|
}
|
2002-09-15 03:17:55 +04:00
|
|
|
#endif // #if BX_SUPPORT_X86_64
|
2001-04-10 05:04:59 +04:00
|
|
|
|
|
|
|
void
|
2002-09-18 02:50:53 +04:00
|
|
|
BX_CPU_C::MOV_TdRd(bxInstruction_c *i)
|
2001-04-10 05:04:59 +04:00
|
|
|
{
|
|
|
|
#if BX_CPU_LEVEL < 3
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_PANIC(("MOV_TdRd:"));
|
2001-04-10 05:04:59 +04:00
|
|
|
#elif BX_CPU_LEVEL <= 4
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_PANIC(("MOV_TdRd:"));
|
2001-04-10 05:04:59 +04:00
|
|
|
#else
|
|
|
|
// Pentium+ does not have TRx. They were redesigned using the MSRs.
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_INFO(("MOV_TdRd: causes #UD"));
|
2001-04-10 05:04:59 +04:00
|
|
|
UndefinedOpcode(i);
|
|
|
|
#endif
|
|
|
|
}
|
|
|
|
|
|
|
|
void
|
2002-09-18 02:50:53 +04:00
|
|
|
BX_CPU_C::MOV_RdTd(bxInstruction_c *i)
|
2001-04-10 05:04:59 +04:00
|
|
|
{
|
|
|
|
#if BX_CPU_LEVEL < 3
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_PANIC(("MOV_RdTd:"));
|
2001-04-10 05:04:59 +04:00
|
|
|
#elif BX_CPU_LEVEL <= 4
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_PANIC(("MOV_RdTd:"));
|
2001-04-10 05:04:59 +04:00
|
|
|
#else
|
|
|
|
// Pentium+ does not have TRx. They were redesigned using the MSRs.
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_INFO(("MOV_RdTd: causes #UD"));
|
2001-04-10 05:04:59 +04:00
|
|
|
UndefinedOpcode(i);
|
|
|
|
#endif
|
|
|
|
}
|
|
|
|
|
|
|
|
void
|
2002-09-18 02:50:53 +04:00
|
|
|
BX_CPU_C::LOADALL(bxInstruction_c *i)
|
2001-04-10 05:04:59 +04:00
|
|
|
{
|
|
|
|
#if BX_CPU_LEVEL < 2
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_PANIC(("undocumented LOADALL instruction not supported on 8086"));
|
2001-04-10 05:04:59 +04:00
|
|
|
#else
|
|
|
|
Bit16u msw, tr, flags, ip, ldtr;
|
|
|
|
Bit16u ds_raw, ss_raw, cs_raw, es_raw;
|
|
|
|
Bit16u di, si, bp, sp, bx, dx, cx, ax;
|
|
|
|
Bit16u base_15_0, limit;
|
|
|
|
Bit8u base_23_16, access;
|
|
|
|
|
2002-08-10 16:06:26 +04:00
|
|
|
if (v8086_mode()) BX_PANIC(("proc_ctrl: LOADALL in v8086 mode unsupported"));
|
2001-04-10 05:04:59 +04:00
|
|
|
|
|
|
|
#if BX_CPU_LEVEL > 2
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_PANIC(("loadall: not implemented for 386"));
|
2001-04-10 05:04:59 +04:00
|
|
|
/* ??? need to set G and other bits, and compute .limit_scaled also */
|
|
|
|
/* for all segments CS,DS,SS,... */
|
|
|
|
#endif
|
|
|
|
|
|
|
|
if (BX_CPU_THIS_PTR cr0.pe) {
|
merge in BRANCH-io-cleanup.
To see the commit logs for this use either cvsweb or
cvs update -r BRANCH-io-cleanup and then 'cvs log' the various files.
In general this provides a generic interface for logging.
logfunctions:: is a class that is inherited by some classes, and also
. allocated as a standalone global called 'genlog'. All logging uses
. one of the ::info(), ::error(), ::ldebug(), ::panic() methods of this
. class through 'BX_INFO(), BX_ERROR(), BX_DEBUG(), BX_PANIC()' macros
. respectively.
.
. An example usage:
. BX_INFO(("Hello, World!\n"));
iofunctions:: is a class that is allocated once by default, and assigned
as the iofunction of each logfunctions instance. It is this class that
maintains the file descriptor and other output related code, at this
point using vfprintf(). At some future point, someone may choose to
write a gui 'console' for bochs to which messages would be redirected
simply by assigning a different iofunction class to the various logfunctions
objects.
More cleanup is coming, but this works for now. If you want to see alot
of debugging output, in main.cc, change onoff[LOGLEV_DEBUG]=0 to =1.
Comments, bugs, flames, to me: todd@fries.net
2001-05-15 18:49:57 +04:00
|
|
|
BX_PANIC((
|
2001-05-30 22:56:02 +04:00
|
|
|
"LOADALL not yet supported for protected mode"));
|
2001-04-10 05:04:59 +04:00
|
|
|
}
|
|
|
|
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_PANIC(("LOADALL: handle CR0.val32"));
|
2001-04-10 05:04:59 +04:00
|
|
|
/* MSW */
|
Now, when you compile with --enable-guest2host-tlb, non-paged
mode uses the notion of the guest-to-host TLB. This has the
benefit of allowing more uniform and streamlined acceleration
code in access.cc which does not have to check if CR0.PG
is set, eliminating a few instructions per guest access.
Shaved just a little off execution time, as expected.
Also, access_linear now breaks accesses which span two pages,
into two calls the the physical memory routines, when paging
is off, just like it always has for paging on. Besides
being more uniform, this allows the physical memory access
routines to known the complete data item is contained
within a single physical page, and stop reapplying the
A20ADDR() macro to pointers as it increments them.
Perhaps things can be optimized a little more now there too...
I renamed the routines to {read,write}PhysicalPage() as
a reminder that these routines now operate on data
solely within one page.
I also added a little code so that the paging module is
notified when the A20 line is tweaked, so it can dump
whatever mappings it wants to.
2002-09-05 06:31:24 +04:00
|
|
|
BX_CPU_THIS_PTR mem->readPhysicalPage(this, 0x806, 2, &msw);
|
2001-04-10 05:04:59 +04:00
|
|
|
BX_CPU_THIS_PTR cr0.pe = (msw & 0x01); msw >>= 1;
|
|
|
|
BX_CPU_THIS_PTR cr0.mp = (msw & 0x01); msw >>= 1;
|
|
|
|
BX_CPU_THIS_PTR cr0.em = (msw & 0x01); msw >>= 1;
|
|
|
|
BX_CPU_THIS_PTR cr0.ts = (msw & 0x01);
|
|
|
|
|
2001-05-30 22:56:02 +04:00
|
|
|
//BX_INFO(("LOADALL: pe=%u, mp=%u, em=%u, ts=%u",
|
2001-04-10 05:04:59 +04:00
|
|
|
// (unsigned) BX_CPU_THIS_PTR cr0.pe, (unsigned) BX_CPU_THIS_PTR cr0.mp,
|
merge in BRANCH-io-cleanup.
To see the commit logs for this use either cvsweb or
cvs update -r BRANCH-io-cleanup and then 'cvs log' the various files.
In general this provides a generic interface for logging.
logfunctions:: is a class that is inherited by some classes, and also
. allocated as a standalone global called 'genlog'. All logging uses
. one of the ::info(), ::error(), ::ldebug(), ::panic() methods of this
. class through 'BX_INFO(), BX_ERROR(), BX_DEBUG(), BX_PANIC()' macros
. respectively.
.
. An example usage:
. BX_INFO(("Hello, World!\n"));
iofunctions:: is a class that is allocated once by default, and assigned
as the iofunction of each logfunctions instance. It is this class that
maintains the file descriptor and other output related code, at this
point using vfprintf(). At some future point, someone may choose to
write a gui 'console' for bochs to which messages would be redirected
simply by assigning a different iofunction class to the various logfunctions
objects.
More cleanup is coming, but this works for now. If you want to see alot
of debugging output, in main.cc, change onoff[LOGLEV_DEBUG]=0 to =1.
Comments, bugs, flames, to me: todd@fries.net
2001-05-15 18:49:57 +04:00
|
|
|
// (unsigned) BX_CPU_THIS_PTR cr0.em, (unsigned) BX_CPU_THIS_PTR cr0.ts));
|
2001-04-10 05:04:59 +04:00
|
|
|
|
|
|
|
if (BX_CPU_THIS_PTR cr0.pe || BX_CPU_THIS_PTR cr0.mp || BX_CPU_THIS_PTR cr0.em || BX_CPU_THIS_PTR cr0.ts)
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_PANIC(("LOADALL set PE, MP, EM or TS bits in MSW!"));
|
2001-04-10 05:04:59 +04:00
|
|
|
|
|
|
|
/* TR */
|
Now, when you compile with --enable-guest2host-tlb, non-paged
mode uses the notion of the guest-to-host TLB. This has the
benefit of allowing more uniform and streamlined acceleration
code in access.cc which does not have to check if CR0.PG
is set, eliminating a few instructions per guest access.
Shaved just a little off execution time, as expected.
Also, access_linear now breaks accesses which span two pages,
into two calls the the physical memory routines, when paging
is off, just like it always has for paging on. Besides
being more uniform, this allows the physical memory access
routines to known the complete data item is contained
within a single physical page, and stop reapplying the
A20ADDR() macro to pointers as it increments them.
Perhaps things can be optimized a little more now there too...
I renamed the routines to {read,write}PhysicalPage() as
a reminder that these routines now operate on data
solely within one page.
I also added a little code so that the paging module is
notified when the A20 line is tweaked, so it can dump
whatever mappings it wants to.
2002-09-05 06:31:24 +04:00
|
|
|
BX_CPU_THIS_PTR mem->readPhysicalPage(this, 0x816, 2, &tr);
|
2001-04-10 05:04:59 +04:00
|
|
|
BX_CPU_THIS_PTR tr.selector.value = tr;
|
|
|
|
BX_CPU_THIS_PTR tr.selector.rpl = (tr & 0x03); tr >>= 2;
|
|
|
|
BX_CPU_THIS_PTR tr.selector.ti = (tr & 0x01); tr >>= 1;
|
|
|
|
BX_CPU_THIS_PTR tr.selector.index = tr;
|
Now, when you compile with --enable-guest2host-tlb, non-paged
mode uses the notion of the guest-to-host TLB. This has the
benefit of allowing more uniform and streamlined acceleration
code in access.cc which does not have to check if CR0.PG
is set, eliminating a few instructions per guest access.
Shaved just a little off execution time, as expected.
Also, access_linear now breaks accesses which span two pages,
into two calls the the physical memory routines, when paging
is off, just like it always has for paging on. Besides
being more uniform, this allows the physical memory access
routines to known the complete data item is contained
within a single physical page, and stop reapplying the
A20ADDR() macro to pointers as it increments them.
Perhaps things can be optimized a little more now there too...
I renamed the routines to {read,write}PhysicalPage() as
a reminder that these routines now operate on data
solely within one page.
I also added a little code so that the paging module is
notified when the A20 line is tweaked, so it can dump
whatever mappings it wants to.
2002-09-05 06:31:24 +04:00
|
|
|
BX_CPU_THIS_PTR mem->readPhysicalPage(this, 0x860, 2, &base_15_0);
|
|
|
|
BX_CPU_THIS_PTR mem->readPhysicalPage(this, 0x862, 1, &base_23_16);
|
|
|
|
BX_CPU_THIS_PTR mem->readPhysicalPage(this, 0x863, 1, &access);
|
|
|
|
BX_CPU_THIS_PTR mem->readPhysicalPage(this, 0x864, 2, &limit);
|
2001-04-10 05:04:59 +04:00
|
|
|
|
|
|
|
|
|
|
|
BX_CPU_THIS_PTR tr.cache.valid =
|
|
|
|
BX_CPU_THIS_PTR tr.cache.p = (access & 0x80) >> 7;
|
|
|
|
BX_CPU_THIS_PTR tr.cache.dpl = (access & 0x60) >> 5;
|
|
|
|
BX_CPU_THIS_PTR tr.cache.segment = (access & 0x10) >> 4;
|
2001-10-10 01:15:14 +04:00
|
|
|
// don't allow busy bit in tr.cache.type, so bit 2 is masked away too.
|
|
|
|
BX_CPU_THIS_PTR tr.cache.type = (access & 0x0d);
|
2001-04-10 05:04:59 +04:00
|
|
|
BX_CPU_THIS_PTR tr.cache.u.tss286.base = (base_23_16 << 16) | base_15_0;
|
|
|
|
BX_CPU_THIS_PTR tr.cache.u.tss286.limit = limit;
|
|
|
|
|
|
|
|
if ( (BX_CPU_THIS_PTR tr.selector.value & 0xfffc) == 0 ) {
|
|
|
|
BX_CPU_THIS_PTR tr.cache.valid = 0;
|
|
|
|
}
|
|
|
|
if ( BX_CPU_THIS_PTR tr.cache.valid == 0 ) {
|
|
|
|
}
|
|
|
|
if ( BX_CPU_THIS_PTR tr.cache.u.tss286.limit < 43 ) {
|
|
|
|
BX_CPU_THIS_PTR tr.cache.valid = 0;
|
|
|
|
}
|
|
|
|
if ( BX_CPU_THIS_PTR tr.cache.type != 1 ) {
|
|
|
|
BX_CPU_THIS_PTR tr.cache.valid = 0;
|
|
|
|
}
|
|
|
|
if ( BX_CPU_THIS_PTR tr.cache.segment ) {
|
|
|
|
BX_CPU_THIS_PTR tr.cache.valid = 0;
|
|
|
|
}
|
|
|
|
if (BX_CPU_THIS_PTR tr.cache.valid==0) {
|
|
|
|
BX_CPU_THIS_PTR tr.cache.u.tss286.base = 0;
|
|
|
|
BX_CPU_THIS_PTR tr.cache.u.tss286.limit = 0;
|
|
|
|
BX_CPU_THIS_PTR tr.cache.p = 0;
|
|
|
|
BX_CPU_THIS_PTR tr.selector.value = 0;
|
|
|
|
BX_CPU_THIS_PTR tr.selector.index = 0;
|
|
|
|
BX_CPU_THIS_PTR tr.selector.ti = 0;
|
|
|
|
BX_CPU_THIS_PTR tr.selector.rpl = 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
/* FLAGS */
|
Now, when you compile with --enable-guest2host-tlb, non-paged
mode uses the notion of the guest-to-host TLB. This has the
benefit of allowing more uniform and streamlined acceleration
code in access.cc which does not have to check if CR0.PG
is set, eliminating a few instructions per guest access.
Shaved just a little off execution time, as expected.
Also, access_linear now breaks accesses which span two pages,
into two calls the the physical memory routines, when paging
is off, just like it always has for paging on. Besides
being more uniform, this allows the physical memory access
routines to known the complete data item is contained
within a single physical page, and stop reapplying the
A20ADDR() macro to pointers as it increments them.
Perhaps things can be optimized a little more now there too...
I renamed the routines to {read,write}PhysicalPage() as
a reminder that these routines now operate on data
solely within one page.
I also added a little code so that the paging module is
notified when the A20 line is tweaked, so it can dump
whatever mappings it wants to.
2002-09-05 06:31:24 +04:00
|
|
|
BX_CPU_THIS_PTR mem->readPhysicalPage(this, 0x818, 2, &flags);
|
2001-04-10 05:04:59 +04:00
|
|
|
write_flags(flags, 1, 1);
|
|
|
|
|
|
|
|
/* IP */
|
Now, when you compile with --enable-guest2host-tlb, non-paged
mode uses the notion of the guest-to-host TLB. This has the
benefit of allowing more uniform and streamlined acceleration
code in access.cc which does not have to check if CR0.PG
is set, eliminating a few instructions per guest access.
Shaved just a little off execution time, as expected.
Also, access_linear now breaks accesses which span two pages,
into two calls the the physical memory routines, when paging
is off, just like it always has for paging on. Besides
being more uniform, this allows the physical memory access
routines to known the complete data item is contained
within a single physical page, and stop reapplying the
A20ADDR() macro to pointers as it increments them.
Perhaps things can be optimized a little more now there too...
I renamed the routines to {read,write}PhysicalPage() as
a reminder that these routines now operate on data
solely within one page.
I also added a little code so that the paging module is
notified when the A20 line is tweaked, so it can dump
whatever mappings it wants to.
2002-09-05 06:31:24 +04:00
|
|
|
BX_CPU_THIS_PTR mem->readPhysicalPage(this, 0x81a, 2, &ip);
|
2001-04-10 05:04:59 +04:00
|
|
|
IP = ip;
|
|
|
|
|
|
|
|
/* LDTR */
|
Now, when you compile with --enable-guest2host-tlb, non-paged
mode uses the notion of the guest-to-host TLB. This has the
benefit of allowing more uniform and streamlined acceleration
code in access.cc which does not have to check if CR0.PG
is set, eliminating a few instructions per guest access.
Shaved just a little off execution time, as expected.
Also, access_linear now breaks accesses which span two pages,
into two calls the the physical memory routines, when paging
is off, just like it always has for paging on. Besides
being more uniform, this allows the physical memory access
routines to known the complete data item is contained
within a single physical page, and stop reapplying the
A20ADDR() macro to pointers as it increments them.
Perhaps things can be optimized a little more now there too...
I renamed the routines to {read,write}PhysicalPage() as
a reminder that these routines now operate on data
solely within one page.
I also added a little code so that the paging module is
notified when the A20 line is tweaked, so it can dump
whatever mappings it wants to.
2002-09-05 06:31:24 +04:00
|
|
|
BX_CPU_THIS_PTR mem->readPhysicalPage(this, 0x81c, 2, &ldtr);
|
2001-04-10 05:04:59 +04:00
|
|
|
BX_CPU_THIS_PTR ldtr.selector.value = ldtr;
|
|
|
|
BX_CPU_THIS_PTR ldtr.selector.rpl = (ldtr & 0x03); ldtr >>= 2;
|
|
|
|
BX_CPU_THIS_PTR ldtr.selector.ti = (ldtr & 0x01); ldtr >>= 1;
|
|
|
|
BX_CPU_THIS_PTR ldtr.selector.index = ldtr;
|
|
|
|
if ( (BX_CPU_THIS_PTR ldtr.selector.value & 0xfffc) == 0 ) {
|
|
|
|
BX_CPU_THIS_PTR ldtr.cache.valid = 0;
|
|
|
|
BX_CPU_THIS_PTR ldtr.cache.p = 0;
|
|
|
|
BX_CPU_THIS_PTR ldtr.cache.segment = 0;
|
|
|
|
BX_CPU_THIS_PTR ldtr.cache.type = 0;
|
|
|
|
BX_CPU_THIS_PTR ldtr.cache.u.ldt.base = 0;
|
|
|
|
BX_CPU_THIS_PTR ldtr.cache.u.ldt.limit = 0;
|
|
|
|
BX_CPU_THIS_PTR ldtr.selector.value = 0;
|
|
|
|
BX_CPU_THIS_PTR ldtr.selector.index = 0;
|
|
|
|
BX_CPU_THIS_PTR ldtr.selector.ti = 0;
|
|
|
|
}
|
|
|
|
else {
|
Now, when you compile with --enable-guest2host-tlb, non-paged
mode uses the notion of the guest-to-host TLB. This has the
benefit of allowing more uniform and streamlined acceleration
code in access.cc which does not have to check if CR0.PG
is set, eliminating a few instructions per guest access.
Shaved just a little off execution time, as expected.
Also, access_linear now breaks accesses which span two pages,
into two calls the the physical memory routines, when paging
is off, just like it always has for paging on. Besides
being more uniform, this allows the physical memory access
routines to known the complete data item is contained
within a single physical page, and stop reapplying the
A20ADDR() macro to pointers as it increments them.
Perhaps things can be optimized a little more now there too...
I renamed the routines to {read,write}PhysicalPage() as
a reminder that these routines now operate on data
solely within one page.
I also added a little code so that the paging module is
notified when the A20 line is tweaked, so it can dump
whatever mappings it wants to.
2002-09-05 06:31:24 +04:00
|
|
|
BX_CPU_THIS_PTR mem->readPhysicalPage(this, 0x854, 2, &base_15_0);
|
|
|
|
BX_CPU_THIS_PTR mem->readPhysicalPage(this, 0x856, 1, &base_23_16);
|
|
|
|
BX_CPU_THIS_PTR mem->readPhysicalPage(this, 0x857, 1, &access);
|
|
|
|
BX_CPU_THIS_PTR mem->readPhysicalPage(this, 0x858, 2, &limit);
|
2001-04-10 05:04:59 +04:00
|
|
|
BX_CPU_THIS_PTR ldtr.cache.valid =
|
|
|
|
BX_CPU_THIS_PTR ldtr.cache.p = access >> 7;
|
|
|
|
BX_CPU_THIS_PTR ldtr.cache.dpl = (access >> 5) & 0x03;
|
|
|
|
BX_CPU_THIS_PTR ldtr.cache.segment = (access >> 4) & 0x01;
|
|
|
|
BX_CPU_THIS_PTR ldtr.cache.type = (access & 0x0f);
|
|
|
|
BX_CPU_THIS_PTR ldtr.cache.u.ldt.base = (base_23_16 << 16) | base_15_0;
|
|
|
|
BX_CPU_THIS_PTR ldtr.cache.u.ldt.limit = limit;
|
|
|
|
|
|
|
|
if (access == 0) {
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_PANIC(("loadall: LDTR case access byte=0."));
|
2001-04-10 05:04:59 +04:00
|
|
|
}
|
|
|
|
if ( BX_CPU_THIS_PTR ldtr.cache.valid==0 ) {
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_PANIC(("loadall: ldtr.valid=0"));
|
2001-04-10 05:04:59 +04:00
|
|
|
}
|
|
|
|
if (BX_CPU_THIS_PTR ldtr.cache.segment) { /* not a system segment */
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_INFO((" AR byte = %02x", (unsigned) access));
|
|
|
|
BX_PANIC(("loadall: LDTR descriptor cache loaded with non system segment"));
|
2001-04-10 05:04:59 +04:00
|
|
|
}
|
|
|
|
if ( BX_CPU_THIS_PTR ldtr.cache.type != 2 ) {
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_PANIC(("loadall: LDTR.type(%u) != 2", (unsigned) (access & 0x0f)));
|
2001-04-10 05:04:59 +04:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/* DS */
|
Now, when you compile with --enable-guest2host-tlb, non-paged
mode uses the notion of the guest-to-host TLB. This has the
benefit of allowing more uniform and streamlined acceleration
code in access.cc which does not have to check if CR0.PG
is set, eliminating a few instructions per guest access.
Shaved just a little off execution time, as expected.
Also, access_linear now breaks accesses which span two pages,
into two calls the the physical memory routines, when paging
is off, just like it always has for paging on. Besides
being more uniform, this allows the physical memory access
routines to known the complete data item is contained
within a single physical page, and stop reapplying the
A20ADDR() macro to pointers as it increments them.
Perhaps things can be optimized a little more now there too...
I renamed the routines to {read,write}PhysicalPage() as
a reminder that these routines now operate on data
solely within one page.
I also added a little code so that the paging module is
notified when the A20 line is tweaked, so it can dump
whatever mappings it wants to.
2002-09-05 06:31:24 +04:00
|
|
|
BX_CPU_THIS_PTR mem->readPhysicalPage(this, 0x81e, 2, &ds_raw);
|
2001-04-10 05:04:59 +04:00
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_DS].selector.value = ds_raw;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_DS].selector.rpl = (ds_raw & 0x03); ds_raw >>= 2;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_DS].selector.ti = (ds_raw & 0x01); ds_raw >>= 1;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_DS].selector.index = ds_raw;
|
Now, when you compile with --enable-guest2host-tlb, non-paged
mode uses the notion of the guest-to-host TLB. This has the
benefit of allowing more uniform and streamlined acceleration
code in access.cc which does not have to check if CR0.PG
is set, eliminating a few instructions per guest access.
Shaved just a little off execution time, as expected.
Also, access_linear now breaks accesses which span two pages,
into two calls the the physical memory routines, when paging
is off, just like it always has for paging on. Besides
being more uniform, this allows the physical memory access
routines to known the complete data item is contained
within a single physical page, and stop reapplying the
A20ADDR() macro to pointers as it increments them.
Perhaps things can be optimized a little more now there too...
I renamed the routines to {read,write}PhysicalPage() as
a reminder that these routines now operate on data
solely within one page.
I also added a little code so that the paging module is
notified when the A20 line is tweaked, so it can dump
whatever mappings it wants to.
2002-09-05 06:31:24 +04:00
|
|
|
BX_CPU_THIS_PTR mem->readPhysicalPage(this, 0x848, 2, &base_15_0);
|
|
|
|
BX_CPU_THIS_PTR mem->readPhysicalPage(this, 0x84a, 1, &base_23_16);
|
|
|
|
BX_CPU_THIS_PTR mem->readPhysicalPage(this, 0x84b, 1, &access);
|
|
|
|
BX_CPU_THIS_PTR mem->readPhysicalPage(this, 0x84c, 2, &limit);
|
2001-04-10 05:04:59 +04:00
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_DS].cache.u.segment.base = (base_23_16 << 16) | base_15_0;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_DS].cache.u.segment.limit = limit;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_DS].cache.u.segment.a = (access & 0x01); access >>= 1;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_DS].cache.u.segment.r_w = (access & 0x01); access >>= 1;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_DS].cache.u.segment.c_ed = (access & 0x01); access >>= 1;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_DS].cache.u.segment.executable = (access & 0x01); access >>= 1;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_DS].cache.segment = (access & 0x01); access >>= 1;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_DS].cache.dpl = (access & 0x03); access >>= 2;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_DS].cache.valid =
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_DS].cache.p = (access & 0x01);
|
|
|
|
|
|
|
|
if ( (BX_CPU_THIS_PTR sregs[BX_SEG_REG_DS].selector.value & 0xfffc) == 0 ) {
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_DS].cache.valid = 0;
|
|
|
|
}
|
|
|
|
if (BX_CPU_THIS_PTR sregs[BX_SEG_REG_DS].cache.valid==0 ||
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_DS].cache.segment==0) {
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_PANIC(("loadall: DS invalid"));
|
2001-04-10 05:04:59 +04:00
|
|
|
}
|
|
|
|
|
|
|
|
/* SS */
|
Now, when you compile with --enable-guest2host-tlb, non-paged
mode uses the notion of the guest-to-host TLB. This has the
benefit of allowing more uniform and streamlined acceleration
code in access.cc which does not have to check if CR0.PG
is set, eliminating a few instructions per guest access.
Shaved just a little off execution time, as expected.
Also, access_linear now breaks accesses which span two pages,
into two calls the the physical memory routines, when paging
is off, just like it always has for paging on. Besides
being more uniform, this allows the physical memory access
routines to known the complete data item is contained
within a single physical page, and stop reapplying the
A20ADDR() macro to pointers as it increments them.
Perhaps things can be optimized a little more now there too...
I renamed the routines to {read,write}PhysicalPage() as
a reminder that these routines now operate on data
solely within one page.
I also added a little code so that the paging module is
notified when the A20 line is tweaked, so it can dump
whatever mappings it wants to.
2002-09-05 06:31:24 +04:00
|
|
|
BX_CPU_THIS_PTR mem->readPhysicalPage(this, 0x820, 2, &ss_raw);
|
2001-04-10 05:04:59 +04:00
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_SS].selector.value = ss_raw;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_SS].selector.rpl = (ss_raw & 0x03); ss_raw >>= 2;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_SS].selector.ti = (ss_raw & 0x01); ss_raw >>= 1;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_SS].selector.index = ss_raw;
|
Now, when you compile with --enable-guest2host-tlb, non-paged
mode uses the notion of the guest-to-host TLB. This has the
benefit of allowing more uniform and streamlined acceleration
code in access.cc which does not have to check if CR0.PG
is set, eliminating a few instructions per guest access.
Shaved just a little off execution time, as expected.
Also, access_linear now breaks accesses which span two pages,
into two calls the the physical memory routines, when paging
is off, just like it always has for paging on. Besides
being more uniform, this allows the physical memory access
routines to known the complete data item is contained
within a single physical page, and stop reapplying the
A20ADDR() macro to pointers as it increments them.
Perhaps things can be optimized a little more now there too...
I renamed the routines to {read,write}PhysicalPage() as
a reminder that these routines now operate on data
solely within one page.
I also added a little code so that the paging module is
notified when the A20 line is tweaked, so it can dump
whatever mappings it wants to.
2002-09-05 06:31:24 +04:00
|
|
|
BX_CPU_THIS_PTR mem->readPhysicalPage(this, 0x842, 2, &base_15_0);
|
|
|
|
BX_CPU_THIS_PTR mem->readPhysicalPage(this, 0x844, 1, &base_23_16);
|
|
|
|
BX_CPU_THIS_PTR mem->readPhysicalPage(this, 0x845, 1, &access);
|
|
|
|
BX_CPU_THIS_PTR mem->readPhysicalPage(this, 0x846, 2, &limit);
|
2001-04-10 05:04:59 +04:00
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_SS].cache.u.segment.base = (base_23_16 << 16) | base_15_0;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_SS].cache.u.segment.limit = limit;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_SS].cache.u.segment.a = (access & 0x01); access >>= 1;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_SS].cache.u.segment.r_w = (access & 0x01); access >>= 1;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_SS].cache.u.segment.c_ed = (access & 0x01); access >>= 1;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_SS].cache.u.segment.executable = (access & 0x01); access >>= 1;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_SS].cache.segment = (access & 0x01); access >>= 1;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_SS].cache.dpl = (access & 0x03); access >>= 2;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_SS].cache.p = (access & 0x01);
|
|
|
|
|
|
|
|
if ( (BX_CPU_THIS_PTR sregs[BX_SEG_REG_SS].selector.value & 0xfffc) == 0 ) {
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_SS].cache.valid = 0;
|
|
|
|
}
|
|
|
|
if (BX_CPU_THIS_PTR sregs[BX_SEG_REG_SS].cache.valid==0 ||
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_SS].cache.segment==0) {
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_PANIC(("loadall: SS invalid"));
|
2001-04-10 05:04:59 +04:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
/* CS */
|
Now, when you compile with --enable-guest2host-tlb, non-paged
mode uses the notion of the guest-to-host TLB. This has the
benefit of allowing more uniform and streamlined acceleration
code in access.cc which does not have to check if CR0.PG
is set, eliminating a few instructions per guest access.
Shaved just a little off execution time, as expected.
Also, access_linear now breaks accesses which span two pages,
into two calls the the physical memory routines, when paging
is off, just like it always has for paging on. Besides
being more uniform, this allows the physical memory access
routines to known the complete data item is contained
within a single physical page, and stop reapplying the
A20ADDR() macro to pointers as it increments them.
Perhaps things can be optimized a little more now there too...
I renamed the routines to {read,write}PhysicalPage() as
a reminder that these routines now operate on data
solely within one page.
I also added a little code so that the paging module is
notified when the A20 line is tweaked, so it can dump
whatever mappings it wants to.
2002-09-05 06:31:24 +04:00
|
|
|
BX_CPU_THIS_PTR mem->readPhysicalPage(this, 0x822, 2, &cs_raw);
|
2001-04-10 05:04:59 +04:00
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].selector.value = cs_raw;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].selector.rpl = (cs_raw & 0x03); cs_raw >>= 2;
|
|
|
|
|
2001-05-30 22:56:02 +04:00
|
|
|
//BX_INFO(("LOADALL: setting cs.selector.rpl to %u",
|
merge in BRANCH-io-cleanup.
To see the commit logs for this use either cvsweb or
cvs update -r BRANCH-io-cleanup and then 'cvs log' the various files.
In general this provides a generic interface for logging.
logfunctions:: is a class that is inherited by some classes, and also
. allocated as a standalone global called 'genlog'. All logging uses
. one of the ::info(), ::error(), ::ldebug(), ::panic() methods of this
. class through 'BX_INFO(), BX_ERROR(), BX_DEBUG(), BX_PANIC()' macros
. respectively.
.
. An example usage:
. BX_INFO(("Hello, World!\n"));
iofunctions:: is a class that is allocated once by default, and assigned
as the iofunction of each logfunctions instance. It is this class that
maintains the file descriptor and other output related code, at this
point using vfprintf(). At some future point, someone may choose to
write a gui 'console' for bochs to which messages would be redirected
simply by assigning a different iofunction class to the various logfunctions
objects.
More cleanup is coming, but this works for now. If you want to see alot
of debugging output, in main.cc, change onoff[LOGLEV_DEBUG]=0 to =1.
Comments, bugs, flames, to me: todd@fries.net
2001-05-15 18:49:57 +04:00
|
|
|
// (unsigned) BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].selector.rpl));
|
2001-04-10 05:04:59 +04:00
|
|
|
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].selector.ti = (cs_raw & 0x01); cs_raw >>= 1;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].selector.index = cs_raw;
|
Now, when you compile with --enable-guest2host-tlb, non-paged
mode uses the notion of the guest-to-host TLB. This has the
benefit of allowing more uniform and streamlined acceleration
code in access.cc which does not have to check if CR0.PG
is set, eliminating a few instructions per guest access.
Shaved just a little off execution time, as expected.
Also, access_linear now breaks accesses which span two pages,
into two calls the the physical memory routines, when paging
is off, just like it always has for paging on. Besides
being more uniform, this allows the physical memory access
routines to known the complete data item is contained
within a single physical page, and stop reapplying the
A20ADDR() macro to pointers as it increments them.
Perhaps things can be optimized a little more now there too...
I renamed the routines to {read,write}PhysicalPage() as
a reminder that these routines now operate on data
solely within one page.
I also added a little code so that the paging module is
notified when the A20 line is tweaked, so it can dump
whatever mappings it wants to.
2002-09-05 06:31:24 +04:00
|
|
|
BX_CPU_THIS_PTR mem->readPhysicalPage(this, 0x83c, 2, &base_15_0);
|
|
|
|
BX_CPU_THIS_PTR mem->readPhysicalPage(this, 0x83e, 1, &base_23_16);
|
|
|
|
BX_CPU_THIS_PTR mem->readPhysicalPage(this, 0x83f, 1, &access);
|
|
|
|
BX_CPU_THIS_PTR mem->readPhysicalPage(this, 0x840, 2, &limit);
|
2001-04-10 05:04:59 +04:00
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].cache.u.segment.base = (base_23_16 << 16) | base_15_0;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].cache.u.segment.limit = limit;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].cache.u.segment.a = (access & 0x01); access >>= 1;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].cache.u.segment.r_w = (access & 0x01); access >>= 1;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].cache.u.segment.c_ed = (access & 0x01); access >>= 1;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].cache.u.segment.executable = (access & 0x01); access >>= 1;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].cache.segment = (access & 0x01); access >>= 1;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].cache.dpl = (access & 0x03); access >>= 2;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].cache.p = (access & 0x01);
|
|
|
|
|
|
|
|
if ( (BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].selector.value & 0xfffc) == 0 ) {
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].cache.valid = 0;
|
|
|
|
}
|
|
|
|
if (BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].cache.valid==0 ||
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].cache.segment==0) {
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_PANIC(("loadall: CS invalid"));
|
2001-04-10 05:04:59 +04:00
|
|
|
}
|
|
|
|
|
|
|
|
/* ES */
|
Now, when you compile with --enable-guest2host-tlb, non-paged
mode uses the notion of the guest-to-host TLB. This has the
benefit of allowing more uniform and streamlined acceleration
code in access.cc which does not have to check if CR0.PG
is set, eliminating a few instructions per guest access.
Shaved just a little off execution time, as expected.
Also, access_linear now breaks accesses which span two pages,
into two calls the the physical memory routines, when paging
is off, just like it always has for paging on. Besides
being more uniform, this allows the physical memory access
routines to known the complete data item is contained
within a single physical page, and stop reapplying the
A20ADDR() macro to pointers as it increments them.
Perhaps things can be optimized a little more now there too...
I renamed the routines to {read,write}PhysicalPage() as
a reminder that these routines now operate on data
solely within one page.
I also added a little code so that the paging module is
notified when the A20 line is tweaked, so it can dump
whatever mappings it wants to.
2002-09-05 06:31:24 +04:00
|
|
|
BX_CPU_THIS_PTR mem->readPhysicalPage(this, 0x824, 2, &es_raw);
|
2001-04-10 05:04:59 +04:00
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_ES].selector.value = es_raw;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_ES].selector.rpl = (es_raw & 0x03); es_raw >>= 2;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_ES].selector.ti = (es_raw & 0x01); es_raw >>= 1;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_ES].selector.index = es_raw;
|
Now, when you compile with --enable-guest2host-tlb, non-paged
mode uses the notion of the guest-to-host TLB. This has the
benefit of allowing more uniform and streamlined acceleration
code in access.cc which does not have to check if CR0.PG
is set, eliminating a few instructions per guest access.
Shaved just a little off execution time, as expected.
Also, access_linear now breaks accesses which span two pages,
into two calls the the physical memory routines, when paging
is off, just like it always has for paging on. Besides
being more uniform, this allows the physical memory access
routines to known the complete data item is contained
within a single physical page, and stop reapplying the
A20ADDR() macro to pointers as it increments them.
Perhaps things can be optimized a little more now there too...
I renamed the routines to {read,write}PhysicalPage() as
a reminder that these routines now operate on data
solely within one page.
I also added a little code so that the paging module is
notified when the A20 line is tweaked, so it can dump
whatever mappings it wants to.
2002-09-05 06:31:24 +04:00
|
|
|
BX_CPU_THIS_PTR mem->readPhysicalPage(this, 0x836, 2, &base_15_0);
|
|
|
|
BX_CPU_THIS_PTR mem->readPhysicalPage(this, 0x838, 1, &base_23_16);
|
|
|
|
BX_CPU_THIS_PTR mem->readPhysicalPage(this, 0x839, 1, &access);
|
|
|
|
BX_CPU_THIS_PTR mem->readPhysicalPage(this, 0x83a, 2, &limit);
|
2001-04-10 05:04:59 +04:00
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_ES].cache.u.segment.base = (base_23_16 << 16) | base_15_0;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_ES].cache.u.segment.limit = limit;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_ES].cache.u.segment.a = (access & 0x01); access >>= 1;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_ES].cache.u.segment.r_w = (access & 0x01); access >>= 1;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_ES].cache.u.segment.c_ed = (access & 0x01); access >>= 1;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_ES].cache.u.segment.executable = (access & 0x01); access >>= 1;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_ES].cache.segment = (access & 0x01); access >>= 1;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_ES].cache.dpl = (access & 0x03); access >>= 2;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_ES].cache.p = (access & 0x01);
|
|
|
|
|
|
|
|
#if 0
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_INFO(("cs.dpl = %02x", (unsigned) BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].cache.dpl));
|
|
|
|
BX_INFO(("ss.dpl = %02x", (unsigned) BX_CPU_THIS_PTR sregs[BX_SEG_REG_SS].cache.dpl));
|
|
|
|
BX_INFO(("BX_CPU_THIS_PTR sregs[BX_SEG_REG_DS].dpl = %02x", (unsigned) BX_CPU_THIS_PTR ds.cache.dpl));
|
|
|
|
BX_INFO(("BX_CPU_THIS_PTR sregs[BX_SEG_REG_ES].dpl = %02x", (unsigned) BX_CPU_THIS_PTR es.cache.dpl));
|
|
|
|
BX_INFO(("LOADALL: setting cs.selector.rpl to %u",
|
merge in BRANCH-io-cleanup.
To see the commit logs for this use either cvsweb or
cvs update -r BRANCH-io-cleanup and then 'cvs log' the various files.
In general this provides a generic interface for logging.
logfunctions:: is a class that is inherited by some classes, and also
. allocated as a standalone global called 'genlog'. All logging uses
. one of the ::info(), ::error(), ::ldebug(), ::panic() methods of this
. class through 'BX_INFO(), BX_ERROR(), BX_DEBUG(), BX_PANIC()' macros
. respectively.
.
. An example usage:
. BX_INFO(("Hello, World!\n"));
iofunctions:: is a class that is allocated once by default, and assigned
as the iofunction of each logfunctions instance. It is this class that
maintains the file descriptor and other output related code, at this
point using vfprintf(). At some future point, someone may choose to
write a gui 'console' for bochs to which messages would be redirected
simply by assigning a different iofunction class to the various logfunctions
objects.
More cleanup is coming, but this works for now. If you want to see alot
of debugging output, in main.cc, change onoff[LOGLEV_DEBUG]=0 to =1.
Comments, bugs, flames, to me: todd@fries.net
2001-05-15 18:49:57 +04:00
|
|
|
(unsigned) BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].selector.rpl));
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_INFO(("LOADALL: setting ss.selector.rpl to %u",
|
merge in BRANCH-io-cleanup.
To see the commit logs for this use either cvsweb or
cvs update -r BRANCH-io-cleanup and then 'cvs log' the various files.
In general this provides a generic interface for logging.
logfunctions:: is a class that is inherited by some classes, and also
. allocated as a standalone global called 'genlog'. All logging uses
. one of the ::info(), ::error(), ::ldebug(), ::panic() methods of this
. class through 'BX_INFO(), BX_ERROR(), BX_DEBUG(), BX_PANIC()' macros
. respectively.
.
. An example usage:
. BX_INFO(("Hello, World!\n"));
iofunctions:: is a class that is allocated once by default, and assigned
as the iofunction of each logfunctions instance. It is this class that
maintains the file descriptor and other output related code, at this
point using vfprintf(). At some future point, someone may choose to
write a gui 'console' for bochs to which messages would be redirected
simply by assigning a different iofunction class to the various logfunctions
objects.
More cleanup is coming, but this works for now. If you want to see alot
of debugging output, in main.cc, change onoff[LOGLEV_DEBUG]=0 to =1.
Comments, bugs, flames, to me: todd@fries.net
2001-05-15 18:49:57 +04:00
|
|
|
(unsigned) BX_CPU_THIS_PTR sregs[BX_SEG_REG_SS].selector.rpl));
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_INFO(("LOADALL: setting ds.selector.rpl to %u",
|
merge in BRANCH-io-cleanup.
To see the commit logs for this use either cvsweb or
cvs update -r BRANCH-io-cleanup and then 'cvs log' the various files.
In general this provides a generic interface for logging.
logfunctions:: is a class that is inherited by some classes, and also
. allocated as a standalone global called 'genlog'. All logging uses
. one of the ::info(), ::error(), ::ldebug(), ::panic() methods of this
. class through 'BX_INFO(), BX_ERROR(), BX_DEBUG(), BX_PANIC()' macros
. respectively.
.
. An example usage:
. BX_INFO(("Hello, World!\n"));
iofunctions:: is a class that is allocated once by default, and assigned
as the iofunction of each logfunctions instance. It is this class that
maintains the file descriptor and other output related code, at this
point using vfprintf(). At some future point, someone may choose to
write a gui 'console' for bochs to which messages would be redirected
simply by assigning a different iofunction class to the various logfunctions
objects.
More cleanup is coming, but this works for now. If you want to see alot
of debugging output, in main.cc, change onoff[LOGLEV_DEBUG]=0 to =1.
Comments, bugs, flames, to me: todd@fries.net
2001-05-15 18:49:57 +04:00
|
|
|
(unsigned) BX_CPU_THIS_PTR sregs[BX_SEG_REG_DS].selector.rpl));
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_INFO(("LOADALL: setting es.selector.rpl to %u",
|
merge in BRANCH-io-cleanup.
To see the commit logs for this use either cvsweb or
cvs update -r BRANCH-io-cleanup and then 'cvs log' the various files.
In general this provides a generic interface for logging.
logfunctions:: is a class that is inherited by some classes, and also
. allocated as a standalone global called 'genlog'. All logging uses
. one of the ::info(), ::error(), ::ldebug(), ::panic() methods of this
. class through 'BX_INFO(), BX_ERROR(), BX_DEBUG(), BX_PANIC()' macros
. respectively.
.
. An example usage:
. BX_INFO(("Hello, World!\n"));
iofunctions:: is a class that is allocated once by default, and assigned
as the iofunction of each logfunctions instance. It is this class that
maintains the file descriptor and other output related code, at this
point using vfprintf(). At some future point, someone may choose to
write a gui 'console' for bochs to which messages would be redirected
simply by assigning a different iofunction class to the various logfunctions
objects.
More cleanup is coming, but this works for now. If you want to see alot
of debugging output, in main.cc, change onoff[LOGLEV_DEBUG]=0 to =1.
Comments, bugs, flames, to me: todd@fries.net
2001-05-15 18:49:57 +04:00
|
|
|
(unsigned) BX_CPU_THIS_PTR sregs[BX_SEG_REG_ES].selector.rpl));
|
2001-04-10 05:04:59 +04:00
|
|
|
#endif
|
|
|
|
|
|
|
|
if ( (BX_CPU_THIS_PTR sregs[BX_SEG_REG_ES].selector.value & 0xfffc) == 0 ) {
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_ES].cache.valid = 0;
|
|
|
|
}
|
|
|
|
if (BX_CPU_THIS_PTR sregs[BX_SEG_REG_ES].cache.valid==0 ||
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_ES].cache.segment==0) {
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_PANIC(("loadall: ES invalid"));
|
2001-04-10 05:04:59 +04:00
|
|
|
}
|
|
|
|
|
|
|
|
/* DI */
|
Now, when you compile with --enable-guest2host-tlb, non-paged
mode uses the notion of the guest-to-host TLB. This has the
benefit of allowing more uniform and streamlined acceleration
code in access.cc which does not have to check if CR0.PG
is set, eliminating a few instructions per guest access.
Shaved just a little off execution time, as expected.
Also, access_linear now breaks accesses which span two pages,
into two calls the the physical memory routines, when paging
is off, just like it always has for paging on. Besides
being more uniform, this allows the physical memory access
routines to known the complete data item is contained
within a single physical page, and stop reapplying the
A20ADDR() macro to pointers as it increments them.
Perhaps things can be optimized a little more now there too...
I renamed the routines to {read,write}PhysicalPage() as
a reminder that these routines now operate on data
solely within one page.
I also added a little code so that the paging module is
notified when the A20 line is tweaked, so it can dump
whatever mappings it wants to.
2002-09-05 06:31:24 +04:00
|
|
|
BX_CPU_THIS_PTR mem->readPhysicalPage(this, 0x826, 2, &di);
|
2001-04-10 05:04:59 +04:00
|
|
|
DI = di;
|
|
|
|
|
|
|
|
/* SI */
|
Now, when you compile with --enable-guest2host-tlb, non-paged
mode uses the notion of the guest-to-host TLB. This has the
benefit of allowing more uniform and streamlined acceleration
code in access.cc which does not have to check if CR0.PG
is set, eliminating a few instructions per guest access.
Shaved just a little off execution time, as expected.
Also, access_linear now breaks accesses which span two pages,
into two calls the the physical memory routines, when paging
is off, just like it always has for paging on. Besides
being more uniform, this allows the physical memory access
routines to known the complete data item is contained
within a single physical page, and stop reapplying the
A20ADDR() macro to pointers as it increments them.
Perhaps things can be optimized a little more now there too...
I renamed the routines to {read,write}PhysicalPage() as
a reminder that these routines now operate on data
solely within one page.
I also added a little code so that the paging module is
notified when the A20 line is tweaked, so it can dump
whatever mappings it wants to.
2002-09-05 06:31:24 +04:00
|
|
|
BX_CPU_THIS_PTR mem->readPhysicalPage(this, 0x828, 2, &si);
|
2001-04-10 05:04:59 +04:00
|
|
|
SI = si;
|
|
|
|
|
|
|
|
/* BP */
|
Now, when you compile with --enable-guest2host-tlb, non-paged
mode uses the notion of the guest-to-host TLB. This has the
benefit of allowing more uniform and streamlined acceleration
code in access.cc which does not have to check if CR0.PG
is set, eliminating a few instructions per guest access.
Shaved just a little off execution time, as expected.
Also, access_linear now breaks accesses which span two pages,
into two calls the the physical memory routines, when paging
is off, just like it always has for paging on. Besides
being more uniform, this allows the physical memory access
routines to known the complete data item is contained
within a single physical page, and stop reapplying the
A20ADDR() macro to pointers as it increments them.
Perhaps things can be optimized a little more now there too...
I renamed the routines to {read,write}PhysicalPage() as
a reminder that these routines now operate on data
solely within one page.
I also added a little code so that the paging module is
notified when the A20 line is tweaked, so it can dump
whatever mappings it wants to.
2002-09-05 06:31:24 +04:00
|
|
|
BX_CPU_THIS_PTR mem->readPhysicalPage(this, 0x82a, 2, &bp);
|
2001-04-10 05:04:59 +04:00
|
|
|
BP = bp;
|
|
|
|
|
|
|
|
/* SP */
|
Now, when you compile with --enable-guest2host-tlb, non-paged
mode uses the notion of the guest-to-host TLB. This has the
benefit of allowing more uniform and streamlined acceleration
code in access.cc which does not have to check if CR0.PG
is set, eliminating a few instructions per guest access.
Shaved just a little off execution time, as expected.
Also, access_linear now breaks accesses which span two pages,
into two calls the the physical memory routines, when paging
is off, just like it always has for paging on. Besides
being more uniform, this allows the physical memory access
routines to known the complete data item is contained
within a single physical page, and stop reapplying the
A20ADDR() macro to pointers as it increments them.
Perhaps things can be optimized a little more now there too...
I renamed the routines to {read,write}PhysicalPage() as
a reminder that these routines now operate on data
solely within one page.
I also added a little code so that the paging module is
notified when the A20 line is tweaked, so it can dump
whatever mappings it wants to.
2002-09-05 06:31:24 +04:00
|
|
|
BX_CPU_THIS_PTR mem->readPhysicalPage(this, 0x82c, 2, &sp);
|
2001-04-10 05:04:59 +04:00
|
|
|
SP = sp;
|
|
|
|
|
|
|
|
/* BX */
|
Now, when you compile with --enable-guest2host-tlb, non-paged
mode uses the notion of the guest-to-host TLB. This has the
benefit of allowing more uniform and streamlined acceleration
code in access.cc which does not have to check if CR0.PG
is set, eliminating a few instructions per guest access.
Shaved just a little off execution time, as expected.
Also, access_linear now breaks accesses which span two pages,
into two calls the the physical memory routines, when paging
is off, just like it always has for paging on. Besides
being more uniform, this allows the physical memory access
routines to known the complete data item is contained
within a single physical page, and stop reapplying the
A20ADDR() macro to pointers as it increments them.
Perhaps things can be optimized a little more now there too...
I renamed the routines to {read,write}PhysicalPage() as
a reminder that these routines now operate on data
solely within one page.
I also added a little code so that the paging module is
notified when the A20 line is tweaked, so it can dump
whatever mappings it wants to.
2002-09-05 06:31:24 +04:00
|
|
|
BX_CPU_THIS_PTR mem->readPhysicalPage(this, 0x82e, 2, &bx);
|
2001-04-10 05:04:59 +04:00
|
|
|
BX = bx;
|
|
|
|
|
|
|
|
/* DX */
|
Now, when you compile with --enable-guest2host-tlb, non-paged
mode uses the notion of the guest-to-host TLB. This has the
benefit of allowing more uniform and streamlined acceleration
code in access.cc which does not have to check if CR0.PG
is set, eliminating a few instructions per guest access.
Shaved just a little off execution time, as expected.
Also, access_linear now breaks accesses which span two pages,
into two calls the the physical memory routines, when paging
is off, just like it always has for paging on. Besides
being more uniform, this allows the physical memory access
routines to known the complete data item is contained
within a single physical page, and stop reapplying the
A20ADDR() macro to pointers as it increments them.
Perhaps things can be optimized a little more now there too...
I renamed the routines to {read,write}PhysicalPage() as
a reminder that these routines now operate on data
solely within one page.
I also added a little code so that the paging module is
notified when the A20 line is tweaked, so it can dump
whatever mappings it wants to.
2002-09-05 06:31:24 +04:00
|
|
|
BX_CPU_THIS_PTR mem->readPhysicalPage(this, 0x830, 2, &dx);
|
2001-04-10 05:04:59 +04:00
|
|
|
DX = dx;
|
|
|
|
|
|
|
|
/* CX */
|
Now, when you compile with --enable-guest2host-tlb, non-paged
mode uses the notion of the guest-to-host TLB. This has the
benefit of allowing more uniform and streamlined acceleration
code in access.cc which does not have to check if CR0.PG
is set, eliminating a few instructions per guest access.
Shaved just a little off execution time, as expected.
Also, access_linear now breaks accesses which span two pages,
into two calls the the physical memory routines, when paging
is off, just like it always has for paging on. Besides
being more uniform, this allows the physical memory access
routines to known the complete data item is contained
within a single physical page, and stop reapplying the
A20ADDR() macro to pointers as it increments them.
Perhaps things can be optimized a little more now there too...
I renamed the routines to {read,write}PhysicalPage() as
a reminder that these routines now operate on data
solely within one page.
I also added a little code so that the paging module is
notified when the A20 line is tweaked, so it can dump
whatever mappings it wants to.
2002-09-05 06:31:24 +04:00
|
|
|
BX_CPU_THIS_PTR mem->readPhysicalPage(this, 0x832, 2, &cx);
|
2001-04-10 05:04:59 +04:00
|
|
|
CX = cx;
|
|
|
|
|
|
|
|
/* AX */
|
Now, when you compile with --enable-guest2host-tlb, non-paged
mode uses the notion of the guest-to-host TLB. This has the
benefit of allowing more uniform and streamlined acceleration
code in access.cc which does not have to check if CR0.PG
is set, eliminating a few instructions per guest access.
Shaved just a little off execution time, as expected.
Also, access_linear now breaks accesses which span two pages,
into two calls the the physical memory routines, when paging
is off, just like it always has for paging on. Besides
being more uniform, this allows the physical memory access
routines to known the complete data item is contained
within a single physical page, and stop reapplying the
A20ADDR() macro to pointers as it increments them.
Perhaps things can be optimized a little more now there too...
I renamed the routines to {read,write}PhysicalPage() as
a reminder that these routines now operate on data
solely within one page.
I also added a little code so that the paging module is
notified when the A20 line is tweaked, so it can dump
whatever mappings it wants to.
2002-09-05 06:31:24 +04:00
|
|
|
BX_CPU_THIS_PTR mem->readPhysicalPage(this, 0x834, 2, &ax);
|
2001-04-10 05:04:59 +04:00
|
|
|
AX = ax;
|
|
|
|
|
|
|
|
/* GDTR */
|
Now, when you compile with --enable-guest2host-tlb, non-paged
mode uses the notion of the guest-to-host TLB. This has the
benefit of allowing more uniform and streamlined acceleration
code in access.cc which does not have to check if CR0.PG
is set, eliminating a few instructions per guest access.
Shaved just a little off execution time, as expected.
Also, access_linear now breaks accesses which span two pages,
into two calls the the physical memory routines, when paging
is off, just like it always has for paging on. Besides
being more uniform, this allows the physical memory access
routines to known the complete data item is contained
within a single physical page, and stop reapplying the
A20ADDR() macro to pointers as it increments them.
Perhaps things can be optimized a little more now there too...
I renamed the routines to {read,write}PhysicalPage() as
a reminder that these routines now operate on data
solely within one page.
I also added a little code so that the paging module is
notified when the A20 line is tweaked, so it can dump
whatever mappings it wants to.
2002-09-05 06:31:24 +04:00
|
|
|
BX_CPU_THIS_PTR mem->readPhysicalPage(this, 0x84e, 2, &base_15_0);
|
|
|
|
BX_CPU_THIS_PTR mem->readPhysicalPage(this, 0x850, 1, &base_23_16);
|
|
|
|
BX_CPU_THIS_PTR mem->readPhysicalPage(this, 0x851, 1, &access);
|
|
|
|
BX_CPU_THIS_PTR mem->readPhysicalPage(this, 0x852, 2, &limit);
|
2001-04-10 05:04:59 +04:00
|
|
|
BX_CPU_THIS_PTR gdtr.base = (base_23_16 << 16) | base_15_0;
|
|
|
|
BX_CPU_THIS_PTR gdtr.limit = limit;
|
|
|
|
|
|
|
|
#if 0
|
|
|
|
if (access)
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_INFO(("LOADALL: GDTR access bits not 0 (%02x).",
|
merge in BRANCH-io-cleanup.
To see the commit logs for this use either cvsweb or
cvs update -r BRANCH-io-cleanup and then 'cvs log' the various files.
In general this provides a generic interface for logging.
logfunctions:: is a class that is inherited by some classes, and also
. allocated as a standalone global called 'genlog'. All logging uses
. one of the ::info(), ::error(), ::ldebug(), ::panic() methods of this
. class through 'BX_INFO(), BX_ERROR(), BX_DEBUG(), BX_PANIC()' macros
. respectively.
.
. An example usage:
. BX_INFO(("Hello, World!\n"));
iofunctions:: is a class that is allocated once by default, and assigned
as the iofunction of each logfunctions instance. It is this class that
maintains the file descriptor and other output related code, at this
point using vfprintf(). At some future point, someone may choose to
write a gui 'console' for bochs to which messages would be redirected
simply by assigning a different iofunction class to the various logfunctions
objects.
More cleanup is coming, but this works for now. If you want to see alot
of debugging output, in main.cc, change onoff[LOGLEV_DEBUG]=0 to =1.
Comments, bugs, flames, to me: todd@fries.net
2001-05-15 18:49:57 +04:00
|
|
|
(unsigned) access));
|
2001-04-10 05:04:59 +04:00
|
|
|
#endif
|
|
|
|
|
|
|
|
/* IDTR */
|
Now, when you compile with --enable-guest2host-tlb, non-paged
mode uses the notion of the guest-to-host TLB. This has the
benefit of allowing more uniform and streamlined acceleration
code in access.cc which does not have to check if CR0.PG
is set, eliminating a few instructions per guest access.
Shaved just a little off execution time, as expected.
Also, access_linear now breaks accesses which span two pages,
into two calls the the physical memory routines, when paging
is off, just like it always has for paging on. Besides
being more uniform, this allows the physical memory access
routines to known the complete data item is contained
within a single physical page, and stop reapplying the
A20ADDR() macro to pointers as it increments them.
Perhaps things can be optimized a little more now there too...
I renamed the routines to {read,write}PhysicalPage() as
a reminder that these routines now operate on data
solely within one page.
I also added a little code so that the paging module is
notified when the A20 line is tweaked, so it can dump
whatever mappings it wants to.
2002-09-05 06:31:24 +04:00
|
|
|
BX_CPU_THIS_PTR mem->readPhysicalPage(this, 0x85a, 2, &base_15_0);
|
|
|
|
BX_CPU_THIS_PTR mem->readPhysicalPage(this, 0x85c, 1, &base_23_16);
|
|
|
|
BX_CPU_THIS_PTR mem->readPhysicalPage(this, 0x85d, 1, &access);
|
|
|
|
BX_CPU_THIS_PTR mem->readPhysicalPage(this, 0x85e, 2, &limit);
|
2001-04-10 05:04:59 +04:00
|
|
|
BX_CPU_THIS_PTR idtr.base = (base_23_16 << 16) | base_15_0;
|
|
|
|
BX_CPU_THIS_PTR idtr.limit = limit;
|
|
|
|
#endif
|
|
|
|
}
|
|
|
|
|
2003-01-14 10:46:05 +03:00
|
|
|
/* Get CPU feature flags. Returned by CPUID functions 1 and 80000001. */
|
|
|
|
static inline Bit32u
|
|
|
|
get_std_cpuid_features()
|
2001-04-10 05:04:59 +04:00
|
|
|
{
|
2003-01-14 10:46:05 +03:00
|
|
|
unsigned features;
|
2001-04-10 05:04:59 +04:00
|
|
|
|
|
|
|
// EAX[3:0] Stepping ID
|
|
|
|
// EAX[7:4] Model: starts at 1
|
|
|
|
// EAX[11:8] Family: 4=486, 5=Pentium, 6=PPro
|
|
|
|
// EAX[13:12] Type: 0=OEM,1=overdrive,2=dual cpu,3=reserved
|
|
|
|
// EAX[31:14] Reserved
|
|
|
|
// EBX: Reserved (0)
|
|
|
|
// ECX: Reserved (0)
|
|
|
|
// EDX: Feature Flags
|
|
|
|
// [0:0] FPU on chip
|
|
|
|
// [1:1] VME: Virtual-8086 Mode enhancements
|
|
|
|
// [2:2] DE: Debug Extensions (I/O breakpoints)
|
|
|
|
// [3:3] PSE: Page Size Extensions
|
|
|
|
// [4:4] TSC: Time Stamp Counter
|
|
|
|
// [5:5] MSR: RDMSR and WRMSR support
|
|
|
|
// [6:6] PAE: Physical Address Extensions
|
|
|
|
// [7:7] MCE: Machine Check Exception
|
|
|
|
// [8:8] CXS: CMPXCHG8B instruction
|
|
|
|
// [9:9] APIC: APIC on Chip
|
2003-01-20 23:10:31 +03:00
|
|
|
// [10:10] Reserved
|
|
|
|
// [11:11] SYSENTER/SYSEXIT support
|
2001-04-10 05:04:59 +04:00
|
|
|
// [12:12] MTRR: Memory Type Range Reg
|
|
|
|
// [13:13] PGE/PTE Global Bit
|
|
|
|
// [14:14] MCA: Machine Check Architecture
|
|
|
|
// [15:15] CMOV: Cond Mov/Cmp Instructions
|
|
|
|
// [22:16] Reserved
|
|
|
|
// [23:23] MMX Technology
|
2002-09-23 18:45:44 +04:00
|
|
|
// [24] FXSR: FSAVE/FXRSTOR (also indicates CR4.OSFXSR is available)
|
2002-09-23 18:31:21 +04:00
|
|
|
// [25] SSE: SSE Extensions
|
|
|
|
// [26] SSE2: SSE2 Extensions
|
2002-09-23 18:38:14 +04:00
|
|
|
// [28] Reserved
|
2002-09-23 18:31:21 +04:00
|
|
|
// [29] TM: Therm Monitor
|
2002-09-23 18:38:14 +04:00
|
|
|
// [31:30] Reserved
|
2001-04-10 05:04:59 +04:00
|
|
|
|
|
|
|
features = 0; // start with none
|
|
|
|
|
|
|
|
# if BX_SUPPORT_FPU
|
|
|
|
features |= 0x01;
|
|
|
|
# endif
|
|
|
|
|
2003-01-14 10:46:05 +03:00
|
|
|
#if (BX_CPU_LEVEL >= 5)
|
|
|
|
features |= (1<<8); // Support CMPXCHG8B instruction
|
2001-05-23 12:16:07 +04:00
|
|
|
features |= (1<<4); // implement TSC
|
2002-09-09 20:11:25 +04:00
|
|
|
# if BX_SUPPORT_MMX
|
|
|
|
features |= (1<<23); // support MMX
|
|
|
|
# endif
|
2002-09-23 18:31:21 +04:00
|
|
|
#endif
|
2003-01-14 10:46:05 +03:00
|
|
|
|
|
|
|
#if BX_CPU_LEVEL == 6
|
2002-09-01 08:01:14 +04:00
|
|
|
features |= (1<<15); // Implement CMOV instructions.
|
2001-05-23 19:54:05 +04:00
|
|
|
# if BX_SUPPORT_APIC
|
2001-05-23 12:16:07 +04:00
|
|
|
features |= (1<<9); // APIC on chip
|
2001-05-23 19:54:05 +04:00
|
|
|
# endif
|
2002-11-14 00:00:05 +03:00
|
|
|
# if BX_SUPPORT_SSE >= 1
|
2002-10-16 21:37:35 +04:00
|
|
|
features |= (1<<24); // support FSAVE/FXRSTOR
|
|
|
|
features |= (1<<25); // support SSE
|
|
|
|
# endif
|
2002-11-14 00:00:05 +03:00
|
|
|
# if BX_SUPPORT_SSE >= 2
|
2002-10-16 21:37:35 +04:00
|
|
|
features |= (1<<26); // support SSE2
|
|
|
|
# endif
|
2002-09-09 20:11:25 +04:00
|
|
|
#endif
|
Integrated patches for:
- Paging code rehash. You must now use --enable-4meg-pages to
use 4Meg pages, with the default of disabled, since we don't well
support 4Meg pages yet. Paging table walks model a real CPU
more closely now, and I fixed some bugs in the old logic.
- Segment check redundancy elimination. After a segment is loaded,
reads and writes are marked when a segment type check succeeds, and
they are skipped thereafter, when possible.
- Repeated IO and memory string copy acceleration. Only some variants
of instructions are available on all platforms, word and dword
variants only on x86 for the moment due to alignment and endian issues.
This is compiled in currently with no option - I should add a configure
option.
- Added a guest linear address to host TLB. Actually, I just stick
the host address (mem.vector[addr] address) in the upper 29 bits
of the field 'combined_access' since they are unused. Convenient
for now. I'm only storing page frame addresses. This was the
simplest for of such a TLB. We can likely enhance this. Also,
I only accelerated the normal read/write routines in access.cc.
Could also modify the read-modify-write versions too. You must
use --enable-guest2host-tlb, to try this out. Currently speeds
up Win95 boot time by about 3.5% for me. More ground to cover...
- Minor mods to CPUI/MOV_CdRd for CMOV.
- Integrated enhancements from Volker to getHostMemAddr() for PCI
being enabled.
2002-09-02 00:12:09 +04:00
|
|
|
|
|
|
|
#if BX_SUPPORT_4MEG_PAGES
|
2002-11-08 23:26:12 +03:00
|
|
|
features |= (1<<3); // Support Page-Size Extension (4M pages)
|
Integrated patches for:
- Paging code rehash. You must now use --enable-4meg-pages to
use 4Meg pages, with the default of disabled, since we don't well
support 4Meg pages yet. Paging table walks model a real CPU
more closely now, and I fixed some bugs in the old logic.
- Segment check redundancy elimination. After a segment is loaded,
reads and writes are marked when a segment type check succeeds, and
they are skipped thereafter, when possible.
- Repeated IO and memory string copy acceleration. Only some variants
of instructions are available on all platforms, word and dword
variants only on x86 for the moment due to alignment and endian issues.
This is compiled in currently with no option - I should add a configure
option.
- Added a guest linear address to host TLB. Actually, I just stick
the host address (mem.vector[addr] address) in the upper 29 bits
of the field 'combined_access' since they are unused. Convenient
for now. I'm only storing page frame addresses. This was the
simplest for of such a TLB. We can likely enhance this. Also,
I only accelerated the normal read/write routines in access.cc.
Could also modify the read-modify-write versions too. You must
use --enable-guest2host-tlb, to try this out. Currently speeds
up Win95 boot time by about 3.5% for me. More ground to cover...
- Minor mods to CPUI/MOV_CdRd for CMOV.
- Integrated enhancements from Volker to getHostMemAddr() for PCI
being enabled.
2002-09-02 00:12:09 +04:00
|
|
|
#endif
|
2001-04-10 05:04:59 +04:00
|
|
|
|
2002-09-10 07:52:32 +04:00
|
|
|
#if BX_SupportGlobalPages
|
2002-11-08 23:26:12 +03:00
|
|
|
features |= (1<<13); // Support Global pages.
|
2002-09-10 07:52:32 +04:00
|
|
|
#endif
|
|
|
|
|
2002-09-15 03:17:55 +04:00
|
|
|
#if BX_SupportPAE
|
2002-11-08 23:26:12 +03:00
|
|
|
features |= (1<<6); // Support PAE.
|
2002-09-15 03:17:55 +04:00
|
|
|
#endif
|
|
|
|
|
2003-01-14 10:46:05 +03:00
|
|
|
#if BX_SUPPORT_X86_64
|
2003-01-20 23:10:31 +03:00
|
|
|
features |= (1<<5); // AMD specific MSR's
|
|
|
|
#endif
|
|
|
|
|
|
|
|
#if BX_SUPPORT_SEP
|
|
|
|
features |= (1<<11); // SYSENTER/SYSEXIT
|
2003-01-14 10:46:05 +03:00
|
|
|
#endif
|
|
|
|
|
|
|
|
return features;
|
|
|
|
}
|
|
|
|
|
|
|
|
void
|
|
|
|
BX_CPU_C::CPUID(bxInstruction_c *i)
|
|
|
|
{
|
|
|
|
#if BX_CPU_LEVEL >= 4
|
2003-05-02 16:22:48 +04:00
|
|
|
unsigned family, model, stepping;
|
|
|
|
#if BX_SUPPORT_X86_64
|
|
|
|
unsigned features;
|
|
|
|
#endif
|
2002-09-29 20:23:03 +04:00
|
|
|
#endif
|
|
|
|
|
2003-01-14 10:46:05 +03:00
|
|
|
invalidate_prefetch_q();
|
|
|
|
|
|
|
|
#if BX_CPU_LEVEL >= 4
|
|
|
|
switch (EAX) {
|
|
|
|
case 0:
|
|
|
|
// EAX: highest input value understood by CPUID
|
|
|
|
// EBX: vendor ID string
|
|
|
|
// EDX: vendor ID string
|
|
|
|
// ECX: vendor ID string
|
|
|
|
RAX = 1; // 486 or pentium
|
|
|
|
#if BX_SUPPORT_X86_64
|
|
|
|
RBX = 0x68747541; // "Auth"
|
|
|
|
RDX = 0x69746e65; // "enti"
|
|
|
|
RCX = 0x444d4163; // "cAMD"
|
|
|
|
#else
|
|
|
|
RBX = 0x756e6547; // "Genu"
|
|
|
|
RDX = 0x49656e69; // "ineI"
|
|
|
|
RCX = 0x6c65746e; // "ntel"
|
|
|
|
#endif
|
|
|
|
break;
|
|
|
|
|
|
|
|
case 1:
|
|
|
|
|
|
|
|
#if BX_CPU_LEVEL == 4
|
|
|
|
family = 4;
|
|
|
|
# if BX_SUPPORT_FPU
|
|
|
|
// 486dx
|
|
|
|
model = 1;
|
|
|
|
stepping = 3;
|
|
|
|
# else
|
|
|
|
// 486sx
|
|
|
|
model = 2;
|
|
|
|
stepping = 3;
|
|
|
|
# endif
|
|
|
|
|
|
|
|
#elif BX_CPU_LEVEL == 5
|
|
|
|
family = 5;
|
|
|
|
model = 1; // Pentium (60,66)
|
|
|
|
stepping = 3; // ???
|
|
|
|
|
|
|
|
#elif BX_CPU_LEVEL == 6
|
|
|
|
family = 6;
|
|
|
|
#if BX_SUPPORT_X86_64
|
|
|
|
model = 2; // Hammer returns what?
|
|
|
|
#else
|
|
|
|
model = 1; // Pentium Pro
|
|
|
|
#endif
|
|
|
|
stepping = 3; // ???
|
|
|
|
#else
|
|
|
|
BX_PANIC(("CPUID: not implemented for > 6"));
|
|
|
|
#endif
|
2002-09-29 20:23:03 +04:00
|
|
|
|
2002-09-15 03:17:55 +04:00
|
|
|
RAX = (family <<8) | (model<<4) | stepping;
|
|
|
|
RBX = RCX = 0; // reserved
|
2003-01-14 10:46:05 +03:00
|
|
|
RDX = get_std_cpuid_features ();
|
2002-09-15 03:17:55 +04:00
|
|
|
break;
|
|
|
|
|
|
|
|
#if BX_SUPPORT_X86_64
|
|
|
|
// x86-64 functions
|
|
|
|
case 0x80000000:
|
|
|
|
// max function supported.
|
|
|
|
RAX = 0x80000008;
|
|
|
|
RBX = 0x68747541; // "Auth"
|
|
|
|
RDX = 0x69746e65; // "enti"
|
|
|
|
RCX = 0x444d4163; // "cAMD"
|
|
|
|
break;
|
|
|
|
|
|
|
|
case 0x80000001:
|
|
|
|
// long mode supported.
|
2003-01-14 10:46:05 +03:00
|
|
|
features = get_std_cpuid_features ();
|
|
|
|
RAX = features;
|
|
|
|
// Many of the bits in EDX are the same as EAX
|
|
|
|
// [10:10] Reserved
|
|
|
|
// [11:11] SYSCALL and SYSRET instructions
|
|
|
|
// [18:19] Reserved
|
|
|
|
// [20:20] No-Execute page protection
|
|
|
|
// [21:21] Reserved
|
|
|
|
// [22:22] AMD MMX Extensions
|
|
|
|
// [25:28] Reserved
|
|
|
|
// [29:29] Long Mode
|
|
|
|
// [30:30] AMD 3DNow Extensions
|
|
|
|
// [31:31] AMD 3DNow Intructions
|
|
|
|
features = features & 0x0183F3FF;
|
|
|
|
RDX = features | (1 << 29) | (1 << 11);
|
2002-09-15 03:17:55 +04:00
|
|
|
RAX = RBX = RCX = 0;
|
2001-04-10 05:04:59 +04:00
|
|
|
break;
|
|
|
|
|
2002-09-15 03:17:55 +04:00
|
|
|
case 0x80000008:
|
|
|
|
// virtual & phys address size in low 2 bytes.
|
|
|
|
RAX = 0x00003028;
|
|
|
|
RBX = RCX = RDX = 0;
|
|
|
|
break;
|
|
|
|
#endif // #if BX_SUPPORT_X86_64
|
|
|
|
|
2001-04-10 05:04:59 +04:00
|
|
|
default:
|
2002-09-15 03:17:55 +04:00
|
|
|
RAX = RBX = RCX = RDX = 0; // Reserved, undefined
|
2001-04-10 05:04:59 +04:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
#else
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_PANIC(("CPUID: not available on < late 486"));
|
2001-04-10 05:04:59 +04:00
|
|
|
#endif
|
|
|
|
}
|
|
|
|
|
|
|
|
void
|
|
|
|
BX_CPU_C::SetCR0(Bit32u val_32)
|
|
|
|
{
|
|
|
|
// from either MOV_CdRd() or debug functions
|
|
|
|
// protection checks made already or forcing from debug
|
2002-10-25 15:44:41 +04:00
|
|
|
bx_bool prev_pe, prev_pg;
|
2002-09-07 09:21:28 +04:00
|
|
|
Bit32u oldCR0 = BX_CPU_THIS_PTR cr0.val32, newCR0;
|
2001-04-10 05:04:59 +04:00
|
|
|
|
|
|
|
prev_pe = BX_CPU_THIS_PTR cr0.pe;
|
|
|
|
prev_pg = BX_CPU_THIS_PTR cr0.pg;
|
|
|
|
|
|
|
|
BX_CPU_THIS_PTR cr0.pe = val_32 & 0x01;
|
|
|
|
BX_CPU_THIS_PTR cr0.mp = (val_32 >> 1) & 0x01;
|
|
|
|
BX_CPU_THIS_PTR cr0.em = (val_32 >> 2) & 0x01;
|
|
|
|
BX_CPU_THIS_PTR cr0.ts = (val_32 >> 3) & 0x01;
|
|
|
|
// cr0.et is hardwired to 1
|
|
|
|
#if BX_CPU_LEVEL >= 4
|
|
|
|
BX_CPU_THIS_PTR cr0.ne = (val_32 >> 5) & 0x01;
|
|
|
|
BX_CPU_THIS_PTR cr0.wp = (val_32 >> 16) & 0x01;
|
|
|
|
BX_CPU_THIS_PTR cr0.am = (val_32 >> 18) & 0x01;
|
|
|
|
BX_CPU_THIS_PTR cr0.nw = (val_32 >> 29) & 0x01;
|
|
|
|
BX_CPU_THIS_PTR cr0.cd = (val_32 >> 30) & 0x01;
|
|
|
|
#endif
|
|
|
|
BX_CPU_THIS_PTR cr0.pg = (val_32 >> 31) & 0x01;
|
|
|
|
|
|
|
|
// handle reserved bits behaviour
|
|
|
|
#if BX_CPU_LEVEL == 3
|
2002-09-07 09:21:28 +04:00
|
|
|
newCR0 = val_32 | 0x7ffffff0;
|
2001-04-10 05:04:59 +04:00
|
|
|
#elif BX_CPU_LEVEL == 4
|
2002-09-07 09:21:28 +04:00
|
|
|
newCR0 = (val_32 | 0x00000010) & 0xe005003f;
|
2001-04-10 05:04:59 +04:00
|
|
|
#elif BX_CPU_LEVEL == 5
|
2002-09-07 09:21:28 +04:00
|
|
|
newCR0 = val_32 | 0x00000010;
|
2001-04-10 05:04:59 +04:00
|
|
|
#elif BX_CPU_LEVEL == 6
|
2002-09-07 09:21:28 +04:00
|
|
|
newCR0 = (val_32 | 0x00000010) & 0xe005003f;
|
2001-04-10 05:04:59 +04:00
|
|
|
#else
|
|
|
|
#error "MOV_CdRd: implement reserved bits behaviour for this CPU_LEVEL"
|
|
|
|
#endif
|
2002-09-07 09:21:28 +04:00
|
|
|
BX_CPU_THIS_PTR cr0.val32 = newCR0;
|
2001-04-10 05:04:59 +04:00
|
|
|
|
|
|
|
//if (BX_CPU_THIS_PTR cr0.ts)
|
2001-05-30 22:56:02 +04:00
|
|
|
// BX_INFO(("MOV_CdRd:CR0.TS set 0x%x", (unsigned) val_32));
|
2001-04-10 05:04:59 +04:00
|
|
|
|
|
|
|
if (prev_pe==0 && BX_CPU_THIS_PTR cr0.pe) {
|
|
|
|
enter_protected_mode();
|
2002-09-30 02:38:18 +04:00
|
|
|
if (BX_CPU_THIS_PTR get_VM()) BX_PANIC(("EFLAGS.VM=1, enter_PM"));
|
|
|
|
BX_CPU_THIS_PTR protectedMode = 1;
|
|
|
|
BX_CPU_THIS_PTR v8086Mode = 0;
|
|
|
|
BX_CPU_THIS_PTR realMode = 0;
|
2001-04-10 05:04:59 +04:00
|
|
|
}
|
|
|
|
else if (prev_pe==1 && BX_CPU_THIS_PTR cr0.pe==0) {
|
|
|
|
enter_real_mode();
|
2002-09-30 02:38:18 +04:00
|
|
|
BX_CPU_THIS_PTR protectedMode = 0;
|
|
|
|
BX_CPU_THIS_PTR v8086Mode = 0;
|
|
|
|
BX_CPU_THIS_PTR realMode = 1;
|
2001-04-10 05:04:59 +04:00
|
|
|
}
|
|
|
|
|
2002-09-15 03:17:55 +04:00
|
|
|
#if BX_SUPPORT_X86_64
|
|
|
|
if (prev_pg==0 && BX_CPU_THIS_PTR cr0.pg) {
|
|
|
|
if (BX_CPU_THIS_PTR msr.lme) {
|
|
|
|
if (!BX_CPU_THIS_PTR cr4.get_PAE()) {
|
|
|
|
BX_PANIC(("SetCR0: attempt to enter x86-64 LONG mode without enabling CR4.PAE !!!"));
|
|
|
|
exception(BX_GP_EXCEPTION, 0, 0);
|
|
|
|
}
|
|
|
|
BX_CPU_THIS_PTR msr.lma = 1;
|
|
|
|
BX_CPU_THIS_PTR cpu_mode = BX_MODE_LONG_COMPAT;
|
|
|
|
#if BX_EXTERNAL_DEBUGGER
|
2003-01-14 10:40:21 +03:00
|
|
|
//trap_debugger(0);
|
2002-09-15 03:17:55 +04:00
|
|
|
#endif
|
|
|
|
}
|
|
|
|
}
|
|
|
|
else if (prev_pg==1 && BX_CPU_THIS_PTR cr0.pg==0) {
|
|
|
|
if (BX_CPU_THIS_PTR msr.lma) {
|
|
|
|
if (BX_CPU_THIS_PTR dword.rip_upper != 0) {
|
|
|
|
BX_PANIC(("SetCR0: attempt to leave x86-64 LONG mode with RIP upper != 0 !!!"));
|
|
|
|
}
|
|
|
|
BX_CPU_THIS_PTR msr.lma = 0;
|
|
|
|
BX_CPU_THIS_PTR cpu_mode = BX_MODE_IA32;
|
|
|
|
#if BX_EXTERNAL_DEBUGGER
|
2003-01-14 10:40:21 +03:00
|
|
|
//trap_debugger(0);
|
2002-09-15 03:17:55 +04:00
|
|
|
#endif
|
|
|
|
}
|
|
|
|
}
|
|
|
|
#endif // #if BX_SUPPORT_X86_64
|
|
|
|
|
2002-09-07 09:21:28 +04:00
|
|
|
// Give the paging unit a chance to look for changes in bits
|
|
|
|
// it cares about, like {PG,PE}, so it can flush cache entries etc.
|
|
|
|
pagingCR0Changed(oldCR0, newCR0);
|
2001-04-10 05:04:59 +04:00
|
|
|
}
|
|
|
|
|
2002-09-18 02:14:33 +04:00
|
|
|
#if BX_CPU_LEVEL >= 4
|
2002-09-15 03:17:55 +04:00
|
|
|
void
|
|
|
|
BX_CPU_C::SetCR4(Bit32u val_32)
|
|
|
|
{
|
2002-09-23 18:45:44 +04:00
|
|
|
|
|
|
|
// CR4 bit definitions from AMD Hammer manual:
|
|
|
|
// [63-11] Reserved, Must be Zero
|
|
|
|
// [10] OSXMMEXCPT: Operating System Unmasked Exception Support R/W
|
|
|
|
// [9] OSFXSR: Operating System FXSAVE/FXRSTOR Support R/W
|
|
|
|
// [8] PCE: Performance-Monitoring Counter Enable R/W
|
|
|
|
// [7] PGE: Page-Global Enable R/W
|
|
|
|
// [6] MCE: Machine Check Enable R/W
|
|
|
|
// [5] PAE: Physical-Address Extension R/W
|
|
|
|
// [4] PSE: Page Size Extensions R/W
|
|
|
|
// [3] DE: Debugging Extensions R/W
|
|
|
|
// [2] TSD: Time Stamp Disable R/W
|
|
|
|
// [1] PVI: Protected-Mode Virtual Interrupts R/W
|
|
|
|
// [0] VME: Virtual-8086 Mode Extensions R/W
|
|
|
|
|
2002-09-15 03:17:55 +04:00
|
|
|
Bit32u allowMask = 0;
|
|
|
|
Bit32u oldCR4 = BX_CPU_THIS_PTR cr4.getRegister();
|
|
|
|
|
|
|
|
#if BX_SUPPORT_4MEG_PAGES
|
|
|
|
allowMask |= (1<<4);
|
|
|
|
#endif
|
|
|
|
|
|
|
|
#if BX_SupportGlobalPages
|
|
|
|
allowMask |= (1<<7);
|
|
|
|
#endif
|
|
|
|
|
|
|
|
#if BX_SupportPAE
|
|
|
|
allowMask |= (1<<5);
|
|
|
|
#endif
|
|
|
|
|
2003-05-15 20:41:17 +04:00
|
|
|
#if BX_SUPPORT_SSE
|
2002-11-08 23:26:12 +03:00
|
|
|
allowMask |= (1<<9); /* OSFXSR */
|
2003-05-15 20:41:17 +04:00
|
|
|
allowMask |= (1<<10); /* OSXMMECPT */
|
2002-11-08 23:26:12 +03:00
|
|
|
#endif
|
2002-09-15 03:17:55 +04:00
|
|
|
|
|
|
|
#if BX_SUPPORT_X86_64
|
|
|
|
// need to GPF #0 if LME=1 and PAE=0
|
2003-05-15 20:41:17 +04:00
|
|
|
if ((BX_CPU_THIS_PTR msr.lme)
|
2002-09-15 03:17:55 +04:00
|
|
|
&& (!(val_32 >> 5) & 1)
|
2003-05-15 20:41:17 +04:00
|
|
|
&& (BX_CPU_THIS_PTR cr4.get_PAE()))
|
|
|
|
{
|
2002-09-15 03:17:55 +04:00
|
|
|
exception(BX_GP_EXCEPTION, 0, 0);
|
2003-05-15 20:41:17 +04:00
|
|
|
}
|
2002-09-15 03:17:55 +04:00
|
|
|
#endif
|
|
|
|
|
|
|
|
// Need to GPF if trying to set undefined bits.
|
|
|
|
if (val_32 & ~allowMask) {
|
2003-05-15 20:41:17 +04:00
|
|
|
BX_INFO(("#GP(0): SetCR4: Write of 0x%08x not supported (allowMask=0x%x)",
|
2002-09-15 03:17:55 +04:00
|
|
|
val_32, allowMask));
|
|
|
|
exception(BX_GP_EXCEPTION, 0, 0);
|
2003-05-15 20:41:17 +04:00
|
|
|
}
|
2002-09-15 03:17:55 +04:00
|
|
|
val_32 &= allowMask; // Screen out unsupported bits. (not needed, for good measure)
|
|
|
|
BX_CPU_THIS_PTR cr4.setRegister(val_32);
|
|
|
|
pagingCR4Changed(oldCR4, BX_CPU_THIS_PTR cr4.getRegister());
|
|
|
|
}
|
2002-09-18 02:14:33 +04:00
|
|
|
#endif
|
2002-09-15 03:17:55 +04:00
|
|
|
|
2001-04-10 05:04:59 +04:00
|
|
|
|
|
|
|
void
|
2002-09-18 02:50:53 +04:00
|
|
|
BX_CPU_C::RSM(bxInstruction_c *i)
|
2001-04-10 05:04:59 +04:00
|
|
|
{
|
|
|
|
#if BX_CPU_LEVEL >= 4
|
|
|
|
invalidate_prefetch_q();
|
|
|
|
|
2001-05-30 22:56:02 +04:00
|
|
|
BX_PANIC(("RSM: System Management Mode not implemented yet"));
|
2001-04-10 05:04:59 +04:00
|
|
|
#else
|
|
|
|
UndefinedOpcode(i);
|
|
|
|
#endif
|
|
|
|
}
|
|
|
|
|
|
|
|
void
|
2002-09-18 02:50:53 +04:00
|
|
|
BX_CPU_C::RDTSC(bxInstruction_c *i)
|
2001-04-10 05:04:59 +04:00
|
|
|
{
|
|
|
|
#if BX_CPU_LEVEL >= 5
|
2002-10-25 15:44:41 +04:00
|
|
|
bx_bool tsd = BX_CPU_THIS_PTR cr4.get_TSD();
|
|
|
|
bx_bool cpl = CPL;
|
2001-05-23 12:16:07 +04:00
|
|
|
if ((tsd==0) || (tsd==1 && cpl==0)) {
|
|
|
|
// return ticks
|
|
|
|
Bit64u ticks = bx_pc_system.time_ticks ();
|
2002-09-15 03:17:55 +04:00
|
|
|
RAX = (Bit32u) (ticks & 0xffffffff);
|
|
|
|
RDX = (Bit32u) ((ticks >> 32) & 0xffffffff);
|
2001-05-30 22:56:02 +04:00
|
|
|
//BX_INFO(("RDTSC: returning EDX:EAX = %08x:%08x", EDX, EAX));
|
2001-05-23 12:16:07 +04:00
|
|
|
} else {
|
|
|
|
// not allowed to use RDTSC!
|
2002-04-11 05:19:24 +04:00
|
|
|
exception (BX_GP_EXCEPTION, 0, 0);
|
2001-05-23 12:16:07 +04:00
|
|
|
}
|
2001-04-10 05:04:59 +04:00
|
|
|
#else
|
|
|
|
UndefinedOpcode(i);
|
|
|
|
#endif
|
|
|
|
}
|
|
|
|
|
|
|
|
void
|
2002-09-18 02:50:53 +04:00
|
|
|
BX_CPU_C::RDMSR(bxInstruction_c *i)
|
2001-04-10 05:04:59 +04:00
|
|
|
{
|
|
|
|
#if BX_CPU_LEVEL >= 5
|
2002-10-04 21:04:33 +04:00
|
|
|
invalidate_prefetch_q();
|
2002-09-15 03:17:55 +04:00
|
|
|
|
2002-10-04 21:04:33 +04:00
|
|
|
if (v8086_mode()) {
|
|
|
|
BX_INFO(("RDMSR: Invalid whilst in virtual 8086 mode"));
|
|
|
|
goto do_exception;
|
|
|
|
}
|
2002-03-27 19:04:05 +03:00
|
|
|
|
2002-10-04 21:04:33 +04:00
|
|
|
if (CPL!= 0) {
|
|
|
|
BX_INFO(("RDMSR: CPL!= 0"));
|
|
|
|
goto do_exception;
|
|
|
|
}
|
2002-03-27 19:04:05 +03:00
|
|
|
|
2002-10-04 21:04:33 +04:00
|
|
|
/* We have the requested MSR register in ECX */
|
|
|
|
switch(ECX) {
|
2003-01-20 23:10:31 +03:00
|
|
|
|
|
|
|
#if BX_SUPPORT_SEP
|
|
|
|
case BX_MSR_SYSENTER_CS: { EAX = BX_CPU_THIS_PTR sysenter_cs_msr; EDX = 0; return; }
|
|
|
|
case BX_MSR_SYSENTER_ESP: { EAX = BX_CPU_THIS_PTR sysenter_esp_msr; EDX = 0; return; }
|
|
|
|
case BX_MSR_SYSENTER_EIP: { EAX = BX_CPU_THIS_PTR sysenter_eip_msr; EDX = 0; return; }
|
|
|
|
#endif
|
|
|
|
|
2002-03-27 19:04:05 +03:00
|
|
|
#if BX_CPU_LEVEL == 5
|
2002-10-04 21:04:33 +04:00
|
|
|
/* The following registers are defined for Pentium only */
|
|
|
|
case BX_MSR_P5_MC_ADDR:
|
|
|
|
case BX_MSR_MC_TYPE:
|
|
|
|
/* TODO */
|
|
|
|
return;
|
|
|
|
|
|
|
|
case BX_MSR_TSC:
|
|
|
|
RDTSC(i);
|
|
|
|
return;
|
|
|
|
|
|
|
|
case BX_MSR_CESR:
|
|
|
|
/* TODO */
|
|
|
|
return;
|
2001-04-10 05:04:59 +04:00
|
|
|
#else
|
2002-10-04 21:04:33 +04:00
|
|
|
/* These are noops on i686... */
|
|
|
|
case BX_MSR_P5_MC_ADDR:
|
|
|
|
case BX_MSR_MC_TYPE:
|
|
|
|
/* do nothing */
|
|
|
|
return;
|
|
|
|
|
|
|
|
case BX_MSR_TSC:
|
|
|
|
RDTSC(i);
|
|
|
|
return;
|
|
|
|
|
|
|
|
/* ... And these cause an exception on i686 */
|
|
|
|
case BX_MSR_CESR:
|
|
|
|
case BX_MSR_CTR0:
|
|
|
|
case BX_MSR_CTR1:
|
|
|
|
goto do_exception;
|
|
|
|
#endif /* BX_CPU_LEVEL == 5 */
|
|
|
|
|
|
|
|
/* MSR_APICBASE
|
|
|
|
0:7 Reserved
|
|
|
|
8 This is set if its the BSP
|
|
|
|
9:10 Reserved
|
|
|
|
11 APIC Global Enable bit (1=enabled 0=disabled)
|
|
|
|
12:35 APIC Base Address
|
|
|
|
36:63 Reserved
|
|
|
|
*/
|
|
|
|
case BX_MSR_APICBASE:
|
|
|
|
/* we return low 32 bits in EAX, and high in EDX */
|
|
|
|
RAX = Bit32u(BX_CPU_THIS_PTR msr.apicbase & 0xffffffff);
|
|
|
|
RDX = Bit32u(BX_CPU_THIS_PTR msr.apicbase >> 32);
|
|
|
|
BX_INFO(("RDMSR: Read %08x:%08x from MSR_APICBASE", EDX, EAX));
|
|
|
|
return;
|
2002-09-15 03:17:55 +04:00
|
|
|
|
|
|
|
#if BX_SUPPORT_X86_64
|
|
|
|
case BX_MSR_EFER:
|
|
|
|
RAX = (BX_CPU_THIS_PTR msr.sce << 0)
|
|
|
|
| (BX_CPU_THIS_PTR msr.lme << 8)
|
|
|
|
| (BX_CPU_THIS_PTR msr.lma << 10);
|
|
|
|
RDX = 0;
|
|
|
|
return;
|
|
|
|
|
|
|
|
case BX_MSR_STAR:
|
2002-09-25 18:09:08 +04:00
|
|
|
RAX = MSR_STAR;
|
|
|
|
RDX = MSR_STAR >> 32;
|
2002-09-15 03:17:55 +04:00
|
|
|
return;
|
|
|
|
|
|
|
|
case BX_MSR_LSTAR:
|
2002-09-25 18:09:08 +04:00
|
|
|
RAX = MSR_LSTAR;
|
|
|
|
RDX = MSR_LSTAR >> 32;
|
2002-09-15 03:17:55 +04:00
|
|
|
return;
|
|
|
|
|
|
|
|
case BX_MSR_CSTAR:
|
2002-09-25 18:09:08 +04:00
|
|
|
RAX = MSR_CSTAR;
|
|
|
|
RDX = MSR_CSTAR >> 32;
|
2002-09-15 03:17:55 +04:00
|
|
|
return;
|
|
|
|
|
|
|
|
case BX_MSR_FMASK:
|
2002-09-25 18:09:08 +04:00
|
|
|
RAX = MSR_FMASK;
|
|
|
|
RDX = MSR_FMASK >> 32;
|
2002-09-15 03:17:55 +04:00
|
|
|
return;
|
|
|
|
|
|
|
|
case BX_MSR_FSBASE:
|
2002-09-25 18:09:08 +04:00
|
|
|
RAX = MSR_FSBASE;
|
|
|
|
RDX = MSR_FSBASE >> 32;
|
2002-09-15 03:17:55 +04:00
|
|
|
return;
|
|
|
|
|
|
|
|
case BX_MSR_GSBASE:
|
2002-09-25 18:09:08 +04:00
|
|
|
RAX = MSR_GSBASE;
|
|
|
|
RDX = MSR_GSBASE >> 32;
|
2002-09-15 03:17:55 +04:00
|
|
|
return;
|
|
|
|
|
|
|
|
case BX_MSR_KERNELGSBASE:
|
2002-09-25 18:09:08 +04:00
|
|
|
RAX = MSR_KERNELGSBASE;
|
|
|
|
RDX = MSR_KERNELGSBASE >> 32;
|
2002-09-15 03:17:55 +04:00
|
|
|
return;
|
|
|
|
#endif // #if BX_SUPPORT_X86_64
|
|
|
|
|
2002-10-04 21:04:33 +04:00
|
|
|
default:
|
2002-09-24 12:29:06 +04:00
|
|
|
#if BX_IGNORE_BAD_MSR
|
|
|
|
BX_ERROR(("RDMSR: Unknown register %#x", ECX));
|
|
|
|
return;
|
I integrated my hacks to get Linux/x86-64 booting. To keep
these from interfering from a normal compile here's what I did.
In config.h.in (which will generate config.h after a configure),
I added a #define called KPL64Hacks:
#define KPL64Hacks
*After* running configure, you must set this by hand. It will
default to off, so you won't get my hacks in a normal compile.
This will go away soon. There is also a macro just after that
called BailBigRSP(). You don't need to enabled that, but you
can. In many of the instructions which seemed like they could
be hit by the fetchdecode64() process, but which also touched
EIP/ESP, I inserted a macro. Usually this macro expands to nothing.
If you like, you can enabled it, and it will panic if it finds
the upper bits of RIP/RSP set. This helped me find bugs.
Also, I cleaned up the emulation in ctrl_xfer{8,16,32}.cc.
There were some really old legacy code snippets which directly
accessed operands on the stack with access_linear. Lots of
ugly code instead of just pop_32() etc. Cleaning those up,
minimized the number of instructions which directly manipulate
the stack pointer, which should help in refining 64-bit support.
2002-09-24 04:44:56 +04:00
|
|
|
#else
|
2002-10-04 21:04:33 +04:00
|
|
|
BX_PANIC(("RDMSR: Unknown register %#x", ECX));
|
I integrated my hacks to get Linux/x86-64 booting. To keep
these from interfering from a normal compile here's what I did.
In config.h.in (which will generate config.h after a configure),
I added a #define called KPL64Hacks:
#define KPL64Hacks
*After* running configure, you must set this by hand. It will
default to off, so you won't get my hacks in a normal compile.
This will go away soon. There is also a macro just after that
called BailBigRSP(). You don't need to enabled that, but you
can. In many of the instructions which seemed like they could
be hit by the fetchdecode64() process, but which also touched
EIP/ESP, I inserted a macro. Usually this macro expands to nothing.
If you like, you can enabled it, and it will panic if it finds
the upper bits of RIP/RSP set. This helped me find bugs.
Also, I cleaned up the emulation in ctrl_xfer{8,16,32}.cc.
There were some really old legacy code snippets which directly
accessed operands on the stack with access_linear. Lots of
ugly code instead of just pop_32() etc. Cleaning those up,
minimized the number of instructions which directly manipulate
the stack pointer, which should help in refining 64-bit support.
2002-09-24 04:44:56 +04:00
|
|
|
#endif
|
2002-10-04 21:04:33 +04:00
|
|
|
goto do_exception;
|
2002-09-15 03:17:55 +04:00
|
|
|
|
2002-10-04 21:04:33 +04:00
|
|
|
}
|
|
|
|
#endif /* BX_CPU_LEVEL >= 5 */
|
2002-03-27 19:04:05 +03:00
|
|
|
|
|
|
|
do_exception:
|
2002-10-04 21:04:33 +04:00
|
|
|
exception(BX_GP_EXCEPTION, 0, 0);
|
2001-04-10 05:04:59 +04:00
|
|
|
}
|
|
|
|
|
|
|
|
void
|
2002-09-18 02:50:53 +04:00
|
|
|
BX_CPU_C::WRMSR(bxInstruction_c *i)
|
2001-04-10 05:04:59 +04:00
|
|
|
{
|
|
|
|
#if BX_CPU_LEVEL >= 5
|
2002-10-04 21:04:33 +04:00
|
|
|
invalidate_prefetch_q();
|
2002-03-27 19:04:05 +03:00
|
|
|
|
2002-10-04 21:04:33 +04:00
|
|
|
if (v8086_mode()) {
|
|
|
|
BX_INFO(("WRMSR: Invalid whilst in virtual 8086 mode"));
|
|
|
|
goto do_exception;
|
|
|
|
}
|
2002-03-27 19:04:05 +03:00
|
|
|
|
2002-10-04 21:04:33 +04:00
|
|
|
if (CPL!= 0) {
|
|
|
|
BX_INFO(("WDMSR: CPL!= 0"));
|
|
|
|
goto do_exception;
|
|
|
|
}
|
2002-03-27 19:04:05 +03:00
|
|
|
|
2002-10-04 21:04:33 +04:00
|
|
|
/* ECX has the MSR to write to */
|
|
|
|
switch(ECX) {
|
2003-01-20 23:10:31 +03:00
|
|
|
|
|
|
|
#if BX_SUPPORT_SEP
|
|
|
|
case BX_MSR_SYSENTER_CS: {
|
|
|
|
if (EAX & 3) BX_PANIC (("writing sysenter_cs_msr with non-kernel mode selector %X", EAX)); // not a bug according to book
|
|
|
|
BX_CPU_THIS_PTR sysenter_cs_msr = EAX; // ... but very stOOpid
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
case BX_MSR_SYSENTER_ESP: { BX_CPU_THIS_PTR sysenter_esp_msr = EAX; return; }
|
|
|
|
case BX_MSR_SYSENTER_EIP: { BX_CPU_THIS_PTR sysenter_eip_msr = EAX; return; }
|
|
|
|
#endif
|
|
|
|
|
2002-03-27 19:04:05 +03:00
|
|
|
#if BX_CPU_LEVEL == 5
|
2002-10-04 21:04:33 +04:00
|
|
|
/* The following registers are defined for Pentium only */
|
|
|
|
case BX_MSR_P5_MC_ADDR:
|
|
|
|
case BX_MSR_MC_TYPE:
|
|
|
|
case BX_MSR_TSC:
|
|
|
|
case BX_MSR_CESR:
|
|
|
|
/* TODO */
|
|
|
|
return;
|
2001-04-10 05:04:59 +04:00
|
|
|
#else
|
2002-10-04 21:04:33 +04:00
|
|
|
/* These are noops on i686... */
|
|
|
|
case BX_MSR_P5_MC_ADDR:
|
|
|
|
case BX_MSR_MC_TYPE:
|
|
|
|
case BX_MSR_TSC:
|
|
|
|
/* do nothing */
|
|
|
|
return;
|
|
|
|
|
|
|
|
/* ... And these cause an exception on i686 */
|
|
|
|
case BX_MSR_CESR:
|
|
|
|
case BX_MSR_CTR0:
|
|
|
|
case BX_MSR_CTR1:
|
|
|
|
goto do_exception;
|
|
|
|
#endif /* BX_CPU_LEVEL == 5 */
|
|
|
|
|
|
|
|
/* MSR_APICBASE
|
|
|
|
0:7 Reserved
|
|
|
|
8 This is set if its the BSP
|
|
|
|
9:10 Reserved
|
|
|
|
11 APIC Global Enable bit (1=enabled 0=disabled)
|
|
|
|
12:35 APIC Base Address
|
|
|
|
36:63 Reserved
|
|
|
|
*/
|
|
|
|
|
|
|
|
case BX_MSR_APICBASE:
|
|
|
|
BX_CPU_THIS_PTR msr.apicbase = ((Bit64u) EDX << 32) + EAX;
|
|
|
|
BX_INFO(("WRMSR: wrote %08x:%08x to MSR_APICBASE", EDX, EAX));
|
|
|
|
return;
|
2002-09-15 03:17:55 +04:00
|
|
|
|
|
|
|
#if BX_SUPPORT_X86_64
|
|
|
|
case BX_MSR_EFER:
|
|
|
|
// GPF #0 if lme 0->1 and cr0.pg = 1
|
|
|
|
// GPF #0 if lme 1->0 and cr0.pg = 1
|
|
|
|
if ( (BX_CPU_THIS_PTR msr.lme != (EAX >> 8) & 1)
|
|
|
|
&& (BX_CPU_THIS_PTR cr0.pg == 1)) {
|
|
|
|
exception(BX_GP_EXCEPTION, 0, 0);
|
|
|
|
}
|
|
|
|
BX_CPU_THIS_PTR msr.sce = (EAX >> 0) & 1;
|
|
|
|
BX_CPU_THIS_PTR msr.lme = (EAX >> 8) & 1;
|
|
|
|
return;
|
|
|
|
case BX_MSR_STAR:
|
2002-09-25 18:09:08 +04:00
|
|
|
MSR_STAR = ((Bit64u) EDX << 32) + EAX;
|
2002-09-15 03:17:55 +04:00
|
|
|
return;
|
|
|
|
case BX_MSR_LSTAR:
|
2002-09-25 18:09:08 +04:00
|
|
|
MSR_LSTAR = ((Bit64u) EDX << 32) + EAX;
|
2002-09-15 03:17:55 +04:00
|
|
|
return;
|
|
|
|
case BX_MSR_CSTAR:
|
2002-09-25 18:09:08 +04:00
|
|
|
MSR_CSTAR = ((Bit64u) EDX << 32) + EAX;
|
2002-09-15 03:17:55 +04:00
|
|
|
return;
|
|
|
|
case BX_MSR_FMASK:
|
2002-09-25 18:09:08 +04:00
|
|
|
MSR_FMASK = ((Bit64u) EDX << 32) + EAX;
|
2002-09-15 03:17:55 +04:00
|
|
|
return;
|
|
|
|
case BX_MSR_FSBASE:
|
2002-09-25 18:09:08 +04:00
|
|
|
MSR_FSBASE = ((Bit64u) EDX << 32) + EAX;
|
2002-09-15 03:17:55 +04:00
|
|
|
return;
|
|
|
|
case BX_MSR_GSBASE:
|
2002-09-25 18:09:08 +04:00
|
|
|
MSR_GSBASE = ((Bit64u) EDX << 32) + EAX;
|
2002-09-15 03:17:55 +04:00
|
|
|
return;
|
|
|
|
case BX_MSR_KERNELGSBASE:
|
2002-09-25 18:09:08 +04:00
|
|
|
MSR_KERNELGSBASE = ((Bit64u) EDX << 32) + EAX;
|
2002-09-15 03:17:55 +04:00
|
|
|
return;
|
|
|
|
#endif // #if BX_SUPPORT_X86_64
|
|
|
|
|
2002-10-04 21:04:33 +04:00
|
|
|
default:
|
2002-09-24 12:29:06 +04:00
|
|
|
#if BX_IGNORE_BAD_MSR
|
|
|
|
BX_ERROR(("WRMSR: Unknown register %#x", ECX));
|
|
|
|
return;
|
I integrated my hacks to get Linux/x86-64 booting. To keep
these from interfering from a normal compile here's what I did.
In config.h.in (which will generate config.h after a configure),
I added a #define called KPL64Hacks:
#define KPL64Hacks
*After* running configure, you must set this by hand. It will
default to off, so you won't get my hacks in a normal compile.
This will go away soon. There is also a macro just after that
called BailBigRSP(). You don't need to enabled that, but you
can. In many of the instructions which seemed like they could
be hit by the fetchdecode64() process, but which also touched
EIP/ESP, I inserted a macro. Usually this macro expands to nothing.
If you like, you can enabled it, and it will panic if it finds
the upper bits of RIP/RSP set. This helped me find bugs.
Also, I cleaned up the emulation in ctrl_xfer{8,16,32}.cc.
There were some really old legacy code snippets which directly
accessed operands on the stack with access_linear. Lots of
ugly code instead of just pop_32() etc. Cleaning those up,
minimized the number of instructions which directly manipulate
the stack pointer, which should help in refining 64-bit support.
2002-09-24 04:44:56 +04:00
|
|
|
#else
|
2002-10-04 21:04:33 +04:00
|
|
|
BX_PANIC(("WRMSR: Unknown register %#x", ECX));
|
I integrated my hacks to get Linux/x86-64 booting. To keep
these from interfering from a normal compile here's what I did.
In config.h.in (which will generate config.h after a configure),
I added a #define called KPL64Hacks:
#define KPL64Hacks
*After* running configure, you must set this by hand. It will
default to off, so you won't get my hacks in a normal compile.
This will go away soon. There is also a macro just after that
called BailBigRSP(). You don't need to enabled that, but you
can. In many of the instructions which seemed like they could
be hit by the fetchdecode64() process, but which also touched
EIP/ESP, I inserted a macro. Usually this macro expands to nothing.
If you like, you can enabled it, and it will panic if it finds
the upper bits of RIP/RSP set. This helped me find bugs.
Also, I cleaned up the emulation in ctrl_xfer{8,16,32}.cc.
There were some really old legacy code snippets which directly
accessed operands on the stack with access_linear. Lots of
ugly code instead of just pop_32() etc. Cleaning those up,
minimized the number of instructions which directly manipulate
the stack pointer, which should help in refining 64-bit support.
2002-09-24 04:44:56 +04:00
|
|
|
#endif
|
2002-10-04 21:04:33 +04:00
|
|
|
goto do_exception;
|
2002-09-15 03:17:55 +04:00
|
|
|
|
2002-10-04 21:04:33 +04:00
|
|
|
}
|
|
|
|
#endif /* BX_CPU_LEVEL >= 5 */
|
2002-03-27 19:04:05 +03:00
|
|
|
|
|
|
|
do_exception:
|
2002-10-04 21:04:33 +04:00
|
|
|
exception(BX_GP_EXCEPTION, 0, 0);
|
2002-03-27 19:04:05 +03:00
|
|
|
|
2003-01-20 23:10:31 +03:00
|
|
|
}
|
|
|
|
|
|
|
|
void
|
|
|
|
BX_CPU_C::SYSENTER (bxInstruction_c *i)
|
|
|
|
{
|
|
|
|
#if BX_SUPPORT_SEP
|
|
|
|
if (!protected_mode ()) {
|
|
|
|
BX_INFO (("sysenter not from protected mode"));
|
|
|
|
exception (BX_GP_EXCEPTION, 0, 0);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
if (BX_CPU_THIS_PTR sysenter_cs_msr == 0) {
|
|
|
|
BX_INFO (("sysenter with zero sysenter_cs_msr"));
|
|
|
|
exception (BX_GP_EXCEPTION, 0, 0);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
2003-05-11 02:25:55 +04:00
|
|
|
invalidate_prefetch_q();
|
2003-01-20 23:10:31 +03:00
|
|
|
|
|
|
|
BX_CPU_THIS_PTR set_VM(0); // do this just like the book says to do
|
|
|
|
BX_CPU_THIS_PTR set_IF(0);
|
|
|
|
BX_CPU_THIS_PTR set_RF(0);
|
|
|
|
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].selector.value = BX_CPU_THIS_PTR sysenter_cs_msr;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].selector.index = BX_CPU_THIS_PTR sysenter_cs_msr >> 3;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].selector.ti = (BX_CPU_THIS_PTR sysenter_cs_msr >> 2) & 1;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].selector.rpl = 0;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].cache.u.segment.executable = 1; // code segment
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].cache.u.segment.c_ed = 0; // non-conforming
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].cache.u.segment.r_w = 1; // readable
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].cache.u.segment.a = 1; // accessed
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].cache.u.segment.base = 0; // base address
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].cache.u.segment.limit = 0xFFFF; // segment limit
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].cache.u.segment.limit_scaled = 0xFFFFFFFF; // scaled segment limit
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].cache.u.segment.g = 1; // 4k granularity
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].cache.u.segment.d_b = 1; // 32-bit mode
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].cache.u.segment.avl = 0; // available for use by system
|
|
|
|
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_SS].selector.value = BX_CPU_THIS_PTR sysenter_cs_msr + 8;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_SS].selector.index = (BX_CPU_THIS_PTR sysenter_cs_msr + 8) >> 3;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_SS].selector.ti = (BX_CPU_THIS_PTR sysenter_cs_msr >> 2) & 1;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_SS].selector.rpl = 0;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_SS].cache.u.segment.executable = 0; // data segment
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_SS].cache.u.segment.c_ed = 0; // expand-up
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_SS].cache.u.segment.r_w = 1; // writeable
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_SS].cache.u.segment.a = 1; // accessed
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_SS].cache.u.segment.base = 0; // base address
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_SS].cache.u.segment.limit = 0xFFFF; // segment limit
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_SS].cache.u.segment.limit_scaled = 0xFFFFFFFF; // scaled segment limit
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_SS].cache.u.segment.g = 1; // 4k granularity
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_SS].cache.u.segment.d_b = 1; // 32-bit mode
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_SS].cache.u.segment.avl = 0; // available for use by system
|
|
|
|
|
|
|
|
// BX_INFO (("sysenter: old eip %X, esp %x, new eip %x, esp %X, edx %X", BX_CPU_THIS_PTR prev_eip, ESP, BX_CPU_THIS_PTR sysenter_eip_msr, BX_CPU_THIS_PTR sysenter_esp_msr, EDX));
|
|
|
|
|
|
|
|
ESP = BX_CPU_THIS_PTR sysenter_esp_msr;
|
|
|
|
EIP = BX_CPU_THIS_PTR sysenter_eip_msr;
|
|
|
|
#else
|
2003-06-20 12:58:12 +04:00
|
|
|
BX_INFO(("SYSENTER: use --enable-sep to enable SYSENTER/SYSEXIT support"));
|
2003-01-20 23:10:31 +03:00
|
|
|
UndefinedOpcode (i);
|
|
|
|
#endif
|
|
|
|
}
|
|
|
|
|
|
|
|
void
|
|
|
|
BX_CPU_C::SYSEXIT (bxInstruction_c *i)
|
|
|
|
{
|
|
|
|
#if BX_SUPPORT_SEP
|
|
|
|
if (!protected_mode ()) {
|
|
|
|
BX_INFO (("sysexit not from protected mode"));
|
|
|
|
exception (BX_GP_EXCEPTION, 0, 0);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
if (BX_CPU_THIS_PTR sysenter_cs_msr == 0) {
|
|
|
|
BX_INFO (("sysexit with zero sysenter_cs_msr"));
|
|
|
|
exception (BX_GP_EXCEPTION, 0, 0);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
if (CPL != 0) {
|
|
|
|
BX_INFO (("sysexit at non-zero cpl %u", CPL));
|
|
|
|
exception (BX_GP_EXCEPTION, 0, 0);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
2003-05-11 02:25:55 +04:00
|
|
|
invalidate_prefetch_q();
|
2003-01-20 23:10:31 +03:00
|
|
|
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].selector.value = (BX_CPU_THIS_PTR sysenter_cs_msr + 16) | 3;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].selector.index = (BX_CPU_THIS_PTR sysenter_cs_msr + 16) >> 3;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].selector.ti = (BX_CPU_THIS_PTR sysenter_cs_msr >> 2) & 1;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].selector.rpl = 3;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].cache.u.segment.executable = 1; // code segment
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].cache.u.segment.c_ed = 0; // non-conforming
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].cache.u.segment.r_w = 1; // readable
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].cache.u.segment.a = 1; // accessed
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].cache.u.segment.base = 0; // base address
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].cache.u.segment.limit = 0xFFFF; // segment limit
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].cache.u.segment.limit_scaled = 0xFFFFFFFF; // scaled segment limit
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].cache.u.segment.g = 1; // 4k granularity
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].cache.u.segment.d_b = 1; // 32-bit mode
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_CS].cache.u.segment.avl = 0; // available for use by system
|
|
|
|
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_SS].selector.value = (BX_CPU_THIS_PTR sysenter_cs_msr + 24) | 3;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_SS].selector.index = (BX_CPU_THIS_PTR sysenter_cs_msr + 24) >> 3;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_SS].selector.ti = (BX_CPU_THIS_PTR sysenter_cs_msr >> 2) & 1;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_SS].selector.rpl = 3;
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_SS].cache.u.segment.executable = 0; // data segment
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_SS].cache.u.segment.c_ed = 0; // expand-up
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_SS].cache.u.segment.r_w = 1; // writeable
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_SS].cache.u.segment.a = 1; // accessed
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_SS].cache.u.segment.base = 0; // base address
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_SS].cache.u.segment.limit = 0xFFFF; // segment limit
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_SS].cache.u.segment.limit_scaled = 0xFFFFFFFF; // scaled segment limit
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_SS].cache.u.segment.g = 1; // 4k granularity
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_SS].cache.u.segment.d_b = 1; // 32-bit mode
|
|
|
|
BX_CPU_THIS_PTR sregs[BX_SEG_REG_SS].cache.u.segment.avl = 0; // available for use by system
|
|
|
|
|
|
|
|
// BX_INFO (("sysexit: old eip %X, esp %x, new eip %x, esp %X, eax %X", BX_CPU_THIS_PTR prev_eip, ESP, EDX, ECX, EAX));
|
|
|
|
|
|
|
|
ESP = ECX;
|
|
|
|
EIP = EDX;
|
|
|
|
#else
|
2003-06-20 12:58:12 +04:00
|
|
|
BX_INFO(("SYSEXIT: use --enable-sep to enable SYSENTER/SYSEXIT support"));
|
2003-01-20 23:10:31 +03:00
|
|
|
UndefinedOpcode (i);
|
|
|
|
#endif
|
2001-04-10 05:04:59 +04:00
|
|
|
}
|
|
|
|
|
2002-09-25 18:09:08 +04:00
|
|
|
#if BX_SUPPORT_X86_64
|
|
|
|
void
|
|
|
|
BX_CPU_C::SWAPGS(bxInstruction_c *i)
|
|
|
|
{
|
|
|
|
Bit64u temp_GS_base;
|
|
|
|
|
|
|
|
if(CPL != 0) {
|
|
|
|
exception(BX_GP_EXCEPTION, 0, 0);
|
|
|
|
}
|
|
|
|
temp_GS_base = MSR_GSBASE;
|
|
|
|
MSR_GSBASE = MSR_KERNELGSBASE;
|
2002-10-08 18:43:18 +04:00
|
|
|
MSR_KERNELGSBASE = temp_GS_base;
|
2002-09-25 18:09:08 +04:00
|
|
|
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
|
2001-04-10 05:04:59 +04:00
|
|
|
#if BX_X86_DEBUGGER
|
|
|
|
Bit32u
|
|
|
|
BX_CPU_C::hwdebug_compare(Bit32u laddr_0, unsigned size,
|
|
|
|
unsigned opa, unsigned opb)
|
|
|
|
{
|
|
|
|
// Support x86 hardware debug facilities (DR0..DR7)
|
|
|
|
Bit32u dr7 = BX_CPU_THIS_PTR dr7;
|
|
|
|
|
2002-10-25 15:44:41 +04:00
|
|
|
bx_bool ibpoint_found = 0;
|
2001-04-10 05:04:59 +04:00
|
|
|
Bit32u laddr_n = laddr_0 + (size - 1);
|
|
|
|
Bit32u dr0, dr1, dr2, dr3;
|
|
|
|
Bit32u dr0_n, dr1_n, dr2_n, dr3_n;
|
|
|
|
Bit32u len0, len1, len2, len3;
|
|
|
|
static unsigned alignment_mask[4] =
|
|
|
|
// 00b=1 01b=2 10b=undef 11b=4
|
|
|
|
{ 0xffffffff, 0xfffffffe, 0xffffffff, 0xfffffffc };
|
|
|
|
Bit32u dr0_op, dr1_op, dr2_op, dr3_op;
|
|
|
|
|
|
|
|
len0 = (dr7>>18) & 3;
|
|
|
|
len1 = (dr7>>22) & 3;
|
|
|
|
len2 = (dr7>>26) & 3;
|
|
|
|
len3 = (dr7>>30) & 3;
|
|
|
|
|
|
|
|
dr0 = BX_CPU_THIS_PTR dr0 & alignment_mask[len0];
|
|
|
|
dr1 = BX_CPU_THIS_PTR dr1 & alignment_mask[len1];
|
|
|
|
dr2 = BX_CPU_THIS_PTR dr2 & alignment_mask[len2];
|
|
|
|
dr3 = BX_CPU_THIS_PTR dr3 & alignment_mask[len3];
|
|
|
|
|
|
|
|
dr0_n = dr0 + len0;
|
|
|
|
dr1_n = dr1 + len1;
|
|
|
|
dr2_n = dr2 + len2;
|
|
|
|
dr3_n = dr3 + len3;
|
|
|
|
|
|
|
|
dr0_op = (dr7>>16) & 3;
|
|
|
|
dr1_op = (dr7>>20) & 3;
|
|
|
|
dr2_op = (dr7>>24) & 3;
|
|
|
|
dr3_op = (dr7>>28) & 3;
|
|
|
|
|
|
|
|
// See if this instruction address matches any breakpoints
|
|
|
|
if ( (dr7 & 0x00000003) ) {
|
|
|
|
if ( (dr0_op==opa || dr0_op==opb) &&
|
|
|
|
(laddr_0 <= dr0_n) &&
|
|
|
|
(laddr_n >= dr0) )
|
|
|
|
ibpoint_found = 1;
|
|
|
|
}
|
|
|
|
if ( (dr7 & 0x0000000c) ) {
|
|
|
|
if ( (dr1_op==opa || dr1_op==opb) &&
|
|
|
|
(laddr_0 <= dr1_n) &&
|
|
|
|
(laddr_n >= dr1) )
|
|
|
|
ibpoint_found = 1;
|
|
|
|
}
|
|
|
|
if ( (dr7 & 0x00000030) ) {
|
|
|
|
if ( (dr2_op==opa || dr2_op==opb) &&
|
|
|
|
(laddr_0 <= dr2_n) &&
|
|
|
|
(laddr_n >= dr2) )
|
|
|
|
ibpoint_found = 1;
|
|
|
|
}
|
|
|
|
if ( (dr7 & 0x000000c0) ) {
|
|
|
|
if ( (dr3_op==opa || dr3_op==opb) &&
|
|
|
|
(laddr_0 <= dr3_n) &&
|
|
|
|
(laddr_n >= dr3) )
|
|
|
|
ibpoint_found = 1;
|
|
|
|
}
|
|
|
|
|
|
|
|
// If *any* enabled breakpoints matched, then we need to
|
|
|
|
// set status bits for *all* breakpoints, even disabled ones,
|
|
|
|
// as long as they meet the other breakpoint criteria.
|
|
|
|
// This code is similar to that above, only without the
|
|
|
|
// breakpoint enabled check. Seems weird to duplicate effort,
|
|
|
|
// but its more efficient to do it this way.
|
|
|
|
if (ibpoint_found) {
|
|
|
|
// dr6_mask is the return value. These bits represent the bits to
|
|
|
|
// be OR'd into DR6 as a result of the debug event.
|
|
|
|
Bit32u dr6_mask=0;
|
|
|
|
if ( (dr0_op==opa || dr0_op==opb) &&
|
|
|
|
(laddr_0 <= dr0_n) &&
|
|
|
|
(laddr_n >= dr0) )
|
|
|
|
dr6_mask |= 0x01;
|
|
|
|
if ( (dr1_op==opa || dr1_op==opb) &&
|
|
|
|
(laddr_0 <= dr1_n) &&
|
|
|
|
(laddr_n >= dr1) )
|
|
|
|
dr6_mask |= 0x02;
|
|
|
|
if ( (dr2_op==opa || dr2_op==opb) &&
|
|
|
|
(laddr_0 <= dr2_n) &&
|
|
|
|
(laddr_n >= dr2) )
|
|
|
|
dr6_mask |= 0x04;
|
|
|
|
if ( (dr3_op==opa || dr3_op==opb) &&
|
|
|
|
(laddr_0 <= dr3_n) &&
|
|
|
|
(laddr_n >= dr3) )
|
|
|
|
dr6_mask |= 0x08;
|
|
|
|
return(dr6_mask);
|
|
|
|
}
|
|
|
|
return(0);
|
|
|
|
}
|
|
|
|
#endif
|