mirrors/Bochs
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
23d4785e8e
Bochs/bochs/main.cc

2335 lines
78 KiB
C++
Raw Normal View History

- add rcs header
2001-10-03 16:56:51 +04:00
/////////////////////////////////////////////////////////////////////////
applied a temporary fix so that it can compile
2002-03-30 02:37:07 +03:00
// $Id: main.cc,v 1.96 2002-03-29 23:37:07 instinc Exp $
- add rcs header
2001-10-03 16:56:51 +04:00
/////////////////////////////////////////////////////////////////////////
//
- update copyright dates to 2001 for all mandrake headers - for bochs files with other header, replaced with current mandrake header
2001-04-10 06:20:02 +04:00
// Copyright (C) 2001 MandrakeSoft S.A.
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
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
#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
#include <assert.h>
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
#include "state_file.h"
Fixes for bug [ #467530 ] cmd line overrides config interface - rework the order of initialization with and without the control panel. The thing that was bothering me most was the command line options were being processed after the user had set everything in the control panel. This is clearly not what's expected--the command line options should affect the startup defaults of the control panel, but whatever the user chooses in the cpanel menus is the final choice. - if the control panel (config interface) is not wanted, the user can put "-nocp" or "-nocontrolpanel" as the FIRST argument on the command line. Also, the "-psn" option which is automatically passed in by MacOS X when you doubleclick the application causes the control panel to be disabled. In this case, the order of operations is: 1. read bochsrc 2. parse command line options. - if the control panel is enabled (default), the command line options are parsed to provide the startup defaults for the control panel, but the control panel settings are the final answer. So the order is: 1. parse command line options 2. run control panel (if user chooses, he can read bochsrc from menus) - I haven't tested command line options with the debugger yet.
2001-10-07 02:31:31 +04:00
int enable_control_panel = 1;
- attempt to clean up the behavior when you run bochs from the command line or by double clicking. Running from the command line doesn't really work because of a NULL pointer problem in gui/carbon.cc.
2001-09-27 06:54:06 +04:00
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
extern "C" {
#include <signal.h>
}
#ifdef __MINGW32__
void alarm(int);
#endif
#if BX_PROVIDE_DEVICE_MODELS==1
// some prototypes from iodev/
// I want to stay away from including iodev/iodev.h here
Bit32u bx_unmapped_io_read_handler(Bit32u address, unsigned io_len);
void bx_unmapped_io_write_handler(Bit32u address, Bit32u value,
unsigned io_len);
void bx_close_harddrive(void);
#endif
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
void bx_init_bx_dbg (void);
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
void bx_emulate_hga_dumps_timer(void);
- change name of action "crash" to "fatal" - change Makefile.in to "cvs-snapshot" information rather than printing the release number, since in fact it's not the same as the release anymore - use #defines to give the actions names (ACT_IGNORE, ACT_REPORT, ACT_FATAL) - register all logfunctions as they call setio, and keep a list so that we can change their settings more easily. I admit I used a static-sized array to store them. Sorry. - instead of printing a prefix [GEN ] for generic messages, just leave the prefix blank [ ]. Otherwise you start wondering what does gen stand for. - avoid reentry into fatal, since fatal can call BX_INFO and friends. This avoids a potential infinite recursion. - now that files other than .bochsrc can be the rc file, print the actual name of the file in error messages.
2001-05-23 06:37:52 +04:00
static char *divider = "========================================================================";
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
/* typedefs */
split log stuff into a separate file
2001-06-19 20:25:41 +04:00
#define LOG_THIS genlog->
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
#if ( BX_PROVIDE_DEVICE_MODELS==1 )
bx_pc_system_c bx_pc_system;
class state_file state_stuff("state_file.out", "options");
#endif
bx_debug_t bx_dbg;
bx_options_t bx_options = {
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
{ NULL, NULL, NULL }, // floppya
{ NULL, NULL, NULL }, // floppyb
{ 0, NULL, 0, 0, 0 }, // diskc
{ 0, NULL, 0, 0, 0 }, // diskd
- commit serial port patch from Stu Grossman <grossman@juniper.net> Here are the patches (without the serial name fix). Let me know if you want me to change that. Oh, FYI, my changes also open up the com1 terminal in raw mode, which makes the emulation much more accurate. You'll also notice that I added infrastructure for com2->4 in the option parsing. I didn't add it to the serial code, as I think a bunch of things need to be untangled before that can work.
2002-03-03 09:03:29 +03:00
{ 0, NULL}, // com1
{ 0, NULL}, // com2
{ 0, NULL}, // com3
{ 0, NULL}, // com4
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
{ 0, NULL, 0 }, // cdromd
- turned rom address and vga rom path into bx_param_c parameters, so that now the whole "Bochs Memory Options" menu uses new style parameters. The next step is to remove the hardcoded stuff that generates and runs this menu, and replace it with general menu building code. All you should need to create this menu is the string "Bochs Memory Options", and the IDs of the bx_param_c options that should appear on the menu. The bx_param_c structure for each parameter tell what type it is, how to display it, constraints on the value, what to do when the parameter changes.
2001-06-17 17:50:52 +04:00
{ NULL, NULL }, // rom
- merged BRANCH-smp-bochs into main branch. For details see comments in BRANCH-smp-bochs revisions. - The general task was to make multiple CPU's which communicate through their APICs. So instead of BX_CPU and BX_MEM, we now have BX_CPU(x) and BX_MEM(y). For an SMP simulation you have several processors in a shared memory space, so there might be processors BX_CPU(0..3) but only one memory space BX_MEM(0). For cosimulation, you could have BX_CPU(0) with BX_MEM(0), then BX_CPU(1) with BX_MEM(1). WARNING: Cosimulation is almost certainly broken by the SMP changes. - to simulate multiple CPUs, you have to give each CPU time to execute in turn. This is currently implemented using debugger guards. The cpu loop steps one CPU for a few instructions, then steps the next CPU for a few instructions, etc. - there is some limited support in the debugger for two CPUs, for example printing information from each CPU when single stepping.
2001-05-23 12:16:07 +04:00
{ NULL }, // vgarom
- now ips, mouse_enabled, vga_update_interval, and rom_path are declared as bx_param_c * types in the bx_options structure. They are initialized in main.cc (bx_init_options) with default values. Access to parameters of this type should always be like this: bx_options.mouse_enabled->get (); bx_options.mouse_enabled->set (newval); Eventually I will be transferring all options to this format.
2001-06-17 03:08:32 +04:00
{ NULL }, // memory
- allow configuration of the parallel port in .bochsrc and the config interface menus. Parallel port #1 is implemented, and I left stubs for parallel port #2 in case we want to ever add it. If the parallel port is enabled, the init method of parallel.cc does an fopen() on the output file. If disabled, or if the fopen fails, the file handler remains NULL and no characters are printed. There is no attempt to enable/disable the operation of the parallel port, only the output to a file.
2001-11-12 05:35:09 +03:00
{ 0, NULL }, // parallel port 1
{ 0, NULL }, // parallel port 2
- implemented bx_write_configuration which, for the first time ever, writes a bochsrc for you. - since there were two options related to logging, I moved them both into a new struct called bx_log_options. This follows the pattern used by other devices. - in control.cc: removed option 1 from main menu, the one that said "Read options from bochsrc.txt." This was identical to choosing the next option, "Read options from..." and only saved you one keystroke, so I removed it.
2001-06-11 10:35:18 +04:00
{ 0, NULL, NULL, NULL, 0, 0, 0, 0 }, // SB16
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
NULL, // boot drive
- now ips, mouse_enabled, vga_update_interval, and rom_path are declared as bx_param_c * types in the bx_options structure. They are initialized in main.cc (bx_init_options) with default values. Access to parameters of this type should always be like this: bx_options.mouse_enabled->get (); bx_options.mouse_enabled->set (newval); Eventually I will be transferring all options to this format.
2001-06-17 03:08:32 +04:00
NULL, // vga update interval
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
NULL, // default keyboard serial path delay (usec)
- apply Christophe's patch.paste-for-dos. I will make minor tweaks as separate revisions.
2002-03-26 16:51:48 +03:00
NULL, // default keyboard paste delay (usec)
. Added a keyboard_type option in configuration file . Modified answer of keyboard controler to a keyboard-identify command depending on keyboard_type . Added the keyboard_type option in bochsrc man page
2001-12-12 13:43:36 +03:00
NULL, // default keyboard type
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
NULL, // default floppy command delay (usec)
- now ips, mouse_enabled, vga_update_interval, and rom_path are declared as bx_param_c * types in the bx_options structure. They are initialized in main.cc (bx_init_options) with default values. Access to parameters of this type should always be like this: bx_options.mouse_enabled->get (); bx_options.mouse_enabled->set (newval); Eventually I will be transferring all options to this format.
2001-06-17 03:08:32 +04:00
NULL, // ips
NULL, // default mouse_enabled
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
NULL, // default private_colormap
ok, now the amiga target on cvs is complete
2001-08-16 16:35:52 +04:00
#if BX_WITH_AMIGAOS
- add option called fullscreen, for Nicholai Benalal's amiga port
2001-08-16 00:17:19 +04:00
NULL, // full screen mode
- add string parameter called screenmode - put #if BX_USE_AMIGAOS...#endif around fullscreen and screenmode, since they only apply to amiga
2001-08-16 06:00:31 +04:00
NULL, // screen mode
#endif
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
NULL, // default i440FXSupport
{NULL, NULL}, // cmos path, cmos image boolean
- add options to the string parameter class that allow it to deal with raw hex bytes, for the MAC address.
2001-06-21 23:27:05 +04:00
{ NULL, NULL, NULL, NULL, NULL, NULL }, // ne2k
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
NULL, // newHardDriveSupport
- new, partially implemented, lines in .bochsrc allow run-time choice of what action to take on panic, error, info, and debug. The lines don't have any effect quite yet because of an initialization order problem with the logfunctions.
2001-05-23 00:01:40 +04:00
{ 0, NULL, NULL, NULL }, // load32bitOSImage hack stuff
- implemented bx_write_configuration which, for the first time ever, writes a bochsrc for you. - since there were two options related to logging, I moved them both into a new struct called bx_log_options. This follows the pattern used by other devices. - in control.cc: removed option 1 from main menu, the one that said "Read options from bochsrc.txt." This was identical to choosing the next option, "Read options from..." and only saved you one keystroke, so I removed it.
2001-06-11 10:35:18 +04:00
// log options: ignore debug, report info and error, crash on panic.
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
{ NULL, { ACT_IGNORE, ACT_REPORT, ACT_REPORT, ACT_ASK } },
- Added XFree86 keyboard mapping (us, fr, de)
2001-12-14 20:56:37 +03:00
{ NULL, NULL }, // KeyboardMapping
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
};
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
static void parse_line_unformatted(char *context, char *line);
static void parse_line_formatted(char *context, int num_params, char *params[]);
static int parse_bochsrc(char *rcfile);
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
- now ips, mouse_enabled, vga_update_interval, and rom_path are declared as bx_param_c * types in the bx_options structure. They are initialized in main.cc (bx_init_options) with default values. Access to parameters of this type should always be like this: bx_options.mouse_enabled->get (); bx_options.mouse_enabled->set (newval); Eventually I will be transferring all options to this format.
2001-06-17 03:08:32 +04:00
static Bit32s
bx_param_handler (bx_param_c *param, int set, Bit32s val)
{
- converted ne2k and sb16 to new style parameters and control panel, except for MAC address of the ne2k, which is next.
2001-06-21 22:34:50 +04:00
bx_id id = param->get_id ();
switch (id) {
- now ips, mouse_enabled, vga_update_interval, and rom_path are declared as bx_param_c * types in the bx_options structure. They are initialized in main.cc (bx_init_options) with default values. Access to parameters of this type should always be like this: bx_options.mouse_enabled->get (); bx_options.mouse_enabled->set (newval); Eventually I will be transferring all options to this format.
2001-06-17 03:08:32 +04:00
case BXP_VGA_UPDATE_INTERVAL:
// if after init, notify the vga device to change its timer.
if (set && SIM->get_init_done ())
bx_vga.set_update_interval (val);
break;
case BXP_MOUSE_ENABLED:
// if after init, notify the GUI
Minor updates to the mouse code that may help with enable/disable problems. Also fixed a warning in pc_system.h Modified Files: main.cc pc_system.h iodev/keyboard.cc iodev/keyboard.h
2001-09-21 06:46:17 +04:00
if (set && SIM->get_init_done ()) {
- now ips, mouse_enabled, vga_update_interval, and rom_path are declared as bx_param_c * types in the bx_options structure. They are initialized in main.cc (bx_init_options) with default values. Access to parameters of this type should always be like this: bx_options.mouse_enabled->get (); bx_options.mouse_enabled->set (newval); Eventually I will be transferring all options to this format.
2001-06-17 03:08:32 +04:00
bx_gui.mouse_enabled_changed (val!=0);
Minor updates to the mouse code that may help with enable/disable problems. Also fixed a warning in pc_system.h Modified Files: main.cc pc_system.h iodev/keyboard.cc iodev/keyboard.h
2001-09-21 06:46:17 +04:00
bx_keyboard.mouse_enabled_changed (val!=0);
}
- now ips, mouse_enabled, vga_update_interval, and rom_path are declared as bx_param_c * types in the bx_options structure. They are initialized in main.cc (bx_init_options) with default values. Access to parameters of this type should always be like this: bx_options.mouse_enabled->get (); bx_options.mouse_enabled->set (newval); Eventually I will be transferring all options to this format.
2001-06-17 03:08:32 +04:00
break;
- allow configuration of the parallel port in .bochsrc and the config interface menus. Parallel port #1 is implemented, and I left stubs for parallel port #2 in case we want to ever add it. If the parallel port is enabled, the init method of parallel.cc does an fopen() on the output file. If disabled, or if the fopen fails, the file handler remains NULL and no characters are printed. There is no attempt to enable/disable the operation of the parallel port, only the output to a file.
2001-11-12 05:35:09 +03:00
case BXP_PARPORT1_ENABLE:
SIM->get_param (BXP_PARPORT1_OUTFILE)->set_enabled (val!=0);
break;
case BXP_PARPORT2_ENABLE:
SIM->get_param (BXP_PARPORT2_OUTFILE)->set_enabled (val!=0);
break;
- add serial port options to the configuration interface, in the same menu as the parallel port options.
2002-03-03 09:10:04 +03:00
case BXP_COM1_PRESENT:
SIM->get_param (BXP_COM1_PATH)->set_enabled (val!=0);
break;
case BXP_COM2_PRESENT:
SIM->get_param (BXP_COM2_PATH)->set_enabled (val!=0);
break;
case BXP_COM3_PRESENT:
SIM->get_param (BXP_COM3_PATH)->set_enabled (val!=0);
break;
case BXP_COM4_PRESENT:
SIM->get_param (BXP_COM4_PATH)->set_enabled (val!=0);
break;
- converted ne2k and sb16 to new style parameters and control panel, except for MAC address of the ne2k, which is next.
2001-06-21 22:34:50 +04:00
case BXP_SB16_PRESENT:
if (set) {
int enable = (val != 0);
SIM->get_param (BXP_SB16_MIDIFILE)->set_enabled (enable);
SIM->get_param (BXP_SB16_WAVEFILE)->set_enabled (enable);
SIM->get_param (BXP_SB16_LOGFILE)->set_enabled (enable);
SIM->get_param (BXP_SB16_MIDIMODE)->set_enabled (enable);
SIM->get_param (BXP_SB16_WAVEMODE)->set_enabled (enable);
SIM->get_param (BXP_SB16_LOGLEVEL)->set_enabled (enable);
SIM->get_param (BXP_SB16_DMATIMER)->set_enabled (enable);
}
break;
case BXP_NE2K_VALID:
if (set) {
int enable = (val != 0);
SIM->get_param (BXP_NE2K_IOADDR)->set_enabled (enable);
SIM->get_param (BXP_NE2K_IRQ)->set_enabled (enable);
- also disable macaddress when the NE2K is disabled
2001-06-21 23:30:14 +04:00
SIM->get_param (BXP_NE2K_MACADDR)->set_enabled (enable);
- converted ne2k and sb16 to new style parameters and control panel, except for MAC address of the ne2k, which is next.
2001-06-21 22:34:50 +04:00
SIM->get_param (BXP_NE2K_ETHMOD)->set_enabled (enable);
SIM->get_param (BXP_NE2K_ETHDEV)->set_enabled (enable);
}
break;
- turn the last few options into parameters (cmos, loader, random stuff)
2001-06-21 23:57:21 +04:00
case BXP_LOAD32BITOS_WHICH:
if (set) {
int enable = (val != 0);
SIM->get_param (BXP_LOAD32BITOS_PATH)->set_enabled (enable);
SIM->get_param (BXP_LOAD32BITOS_IOLOG)->set_enabled (enable);
SIM->get_param (BXP_LOAD32BITOS_INITRD)->set_enabled (enable);
}
break;
case BXP_CMOS_IMAGE:
SIM->get_param (BXP_CMOS_PATH)->set_enabled (val);
break;
- cdrom and floppy options in configuration interface partially fixed * don't change the device state to "not present" at runtime if the path is changed to "none" - simply eject the media if inserted * don't change the device state to "present" at runtime if the previous state is "not present" * eject and re-insert the media if the current media state is "inserted" * changing the media state in the config interface works now * TODO: don't do anything if the path is unchanged - bx_write_cdrom_options(): writing media state fixed
2002-03-14 23:14:47 +03:00
case BXP_CDROM_INSERTED:
if ((set) && (SIM->get_init_done ())) {
bx_devices.hard_drive->set_cd_media_status(val == BX_INSERTED);
- renamed update_floppy_status_buttons() to update_drive_status_buttons() since it handles the cdrom status too
2002-03-17 23:57:54 +03:00
bx_gui.update_drive_status_buttons ();
- cdrom and floppy options in configuration interface partially fixed * don't change the device state to "not present" at runtime if the path is changed to "none" - simply eject the media if inserted * don't change the device state to "present" at runtime if the previous state is "not present" * eject and re-insert the media if the current media state is "inserted" * changing the media state in the config interface works now * TODO: don't do anything if the path is unchanged - bx_write_cdrom_options(): writing media state fixed
2002-03-14 23:14:47 +03:00
}
break;
case BXP_FLOPPYA_STATUS:
if ((set) && (SIM->get_init_done ())) {
bx_devices.floppy->set_media_status(0, val == BX_INSERTED);
- renamed update_floppy_status_buttons() to update_drive_status_buttons() since it handles the cdrom status too
2002-03-17 23:57:54 +03:00
bx_gui.update_drive_status_buttons ();
- cdrom and floppy options in configuration interface partially fixed * don't change the device state to "not present" at runtime if the path is changed to "none" - simply eject the media if inserted * don't change the device state to "present" at runtime if the previous state is "not present" * eject and re-insert the media if the current media state is "inserted" * changing the media state in the config interface works now * TODO: don't do anything if the path is unchanged - bx_write_cdrom_options(): writing media state fixed
2002-03-14 23:14:47 +03:00
}
break;
case BXP_FLOPPYB_STATUS:
if ((set) && (SIM->get_init_done ())) {
bx_devices.floppy->set_media_status(1, val == BX_INSERTED);
- renamed update_floppy_status_buttons() to update_drive_status_buttons() since it handles the cdrom status too
2002-03-17 23:57:54 +03:00
bx_gui.update_drive_status_buttons ();
- cdrom and floppy options in configuration interface partially fixed * don't change the device state to "not present" at runtime if the path is changed to "none" - simply eject the media if inserted * don't change the device state to "present" at runtime if the previous state is "not present" * eject and re-insert the media if the current media state is "inserted" * changing the media state in the config interface works now * TODO: don't do anything if the path is unchanged - bx_write_cdrom_options(): writing media state fixed
2002-03-14 23:14:47 +03:00
}
break;
- made keyboard_paste_delay a runtime configurable option - clean up comments & debug msgs in paste code
2002-03-26 17:46:03 +03:00
case BXP_KBD_PASTE_DELAY:
if (set) bx_keyboard.paste_delay_changed ();
break;
- now ips, mouse_enabled, vga_update_interval, and rom_path are declared as bx_param_c * types in the bx_options structure. They are initialized in main.cc (bx_init_options) with default values. Access to parameters of this type should always be like this: bx_options.mouse_enabled->get (); bx_options.mouse_enabled->set (newval); Eventually I will be transferring all options to this format.
2001-06-17 03:08:32 +04:00
default:
- converted ne2k and sb16 to new style parameters and control panel, except for MAC address of the ne2k, which is next.
2001-06-21 22:34:50 +04:00
BX_PANIC (("bx_param_handler called with unknown id %d", id));
- now ips, mouse_enabled, vga_update_interval, and rom_path are declared as bx_param_c * types in the bx_options structure. They are initialized in main.cc (bx_init_options) with default values. Access to parameters of this type should always be like this: bx_options.mouse_enabled->get (); bx_options.mouse_enabled->set (newval); Eventually I will be transferring all options to this format.
2001-06-17 03:08:32 +04:00
return -1;
}
return val;
}
- now the memory options menu, interface menu, and disk menu in the control panel are all implemented as bx_list_c, and look much like they did before. - removed many hardcoded UI functions from control.cc, since the much more general "text_ask" methods have replaced them. - add range checking on integer parameter values. This exposed several cases where my initial value for an integer parameter was not in range. - cleaned up behavior of get/set methods. The get/set methods allow the handler to override the value that is returned/set, or perform side effects. - the title parameter of a bx_list_c now defaults to the name. - now bx_param_c fields that used to be private are protected instead - removed references to bx_any - moved definition of set_handler from siminterface.h to siminterface.cc. I was considering with doing a "set" of the old value when the handler is first installed, but that remains commented out. - BX_BOOT_DISKA and BX_BOOT_DISKC are now 0,1 so that they can correspond with the values of a bx_param_enum_c. For a while they were 0x00 and 0x80 corresponding to the numbering convention of the bios, but it didn't really matter.
2001-06-21 18:37:55 +04:00
char *bx_param_string_handler (bx_param_string_c *param, int set, char *val, int maxlen)
{
int empty = 0;
if ((strlen(val) < 1) || !strcmp ("none", val)) {
empty = 1;
val = "none";
}
switch (param->get_id ()) {
case BXP_FLOPPYA_PATH:
if (set==1) {
- cdrom and floppy options in configuration interface partially fixed * don't change the device state to "not present" at runtime if the path is changed to "none" - simply eject the media if inserted * don't change the device state to "present" at runtime if the previous state is "not present" * eject and re-insert the media if the current media state is "inserted" * changing the media state in the config interface works now * TODO: don't do anything if the path is unchanged - bx_write_cdrom_options(): writing media state fixed
2002-03-14 23:14:47 +03:00
if (SIM->get_init_done ()) {
if (empty) {
bx_devices.floppy->set_media_status(0, 0);
- renamed update_floppy_status_buttons() to update_drive_status_buttons() since it handles the cdrom status too
2002-03-17 23:57:54 +03:00
bx_gui.update_drive_status_buttons ();
- cdrom and floppy options in configuration interface partially fixed * don't change the device state to "not present" at runtime if the path is changed to "none" - simply eject the media if inserted * don't change the device state to "present" at runtime if the previous state is "not present" * eject and re-insert the media if the current media state is "inserted" * changing the media state in the config interface works now * TODO: don't do anything if the path is unchanged - bx_write_cdrom_options(): writing media state fixed
2002-03-14 23:14:47 +03:00
} else {
if (!SIM->get_param_num(BXP_FLOPPYA_TYPE)->get_enabled()) {
BX_ERROR(("Cannot add a floppy drive at runtime"));
bx_options.floppya.Opath->set ("none");
}
}
if ((bx_devices.floppy) &&
(SIM->get_param_num(BXP_FLOPPYA_STATUS)->get () == BX_INSERTED)) {
// tell the device model that we removed, then inserted the disk
bx_devices.floppy->set_media_status(0, 0);
bx_devices.floppy->set_media_status(0, 1);
}
} else {
SIM->get_param_num(BXP_FLOPPYA_TYPE)->set_enabled (!empty);
SIM->get_param_num(BXP_FLOPPYA_STATUS)->set_enabled (!empty);
}
- now the memory options menu, interface menu, and disk menu in the control panel are all implemented as bx_list_c, and look much like they did before. - removed many hardcoded UI functions from control.cc, since the much more general "text_ask" methods have replaced them. - add range checking on integer parameter values. This exposed several cases where my initial value for an integer parameter was not in range. - cleaned up behavior of get/set methods. The get/set methods allow the handler to override the value that is returned/set, or perform side effects. - the title parameter of a bx_list_c now defaults to the name. - now bx_param_c fields that used to be private are protected instead - removed references to bx_any - moved definition of set_handler from siminterface.h to siminterface.cc. I was considering with doing a "set" of the old value when the handler is first installed, but that remains commented out. - BX_BOOT_DISKA and BX_BOOT_DISKC are now 0,1 so that they can correspond with the values of a bx_param_enum_c. For a while they were 0x00 and 0x80 corresponding to the numbering convention of the bios, but it didn't really matter.
2001-06-21 18:37:55 +04:00
}
break;
case BXP_FLOPPYB_PATH:
if (set==1) {
- cdrom and floppy options in configuration interface partially fixed * don't change the device state to "not present" at runtime if the path is changed to "none" - simply eject the media if inserted * don't change the device state to "present" at runtime if the previous state is "not present" * eject and re-insert the media if the current media state is "inserted" * changing the media state in the config interface works now * TODO: don't do anything if the path is unchanged - bx_write_cdrom_options(): writing media state fixed
2002-03-14 23:14:47 +03:00
if (SIM->get_init_done ()) {
if (empty) {
bx_devices.floppy->set_media_status(1, 0);
- renamed update_floppy_status_buttons() to update_drive_status_buttons() since it handles the cdrom status too
2002-03-17 23:57:54 +03:00
bx_gui.update_drive_status_buttons ();
- cdrom and floppy options in configuration interface partially fixed * don't change the device state to "not present" at runtime if the path is changed to "none" - simply eject the media if inserted * don't change the device state to "present" at runtime if the previous state is "not present" * eject and re-insert the media if the current media state is "inserted" * changing the media state in the config interface works now * TODO: don't do anything if the path is unchanged - bx_write_cdrom_options(): writing media state fixed
2002-03-14 23:14:47 +03:00
} else {
if (!SIM->get_param_num(BXP_FLOPPYB_TYPE)->get_enabled ()) {
BX_ERROR(("Cannot add a floppy drive at runtime"));
bx_options.floppyb.Opath->set ("none");
}
}
if ((bx_devices.floppy) &&
(SIM->get_param_num(BXP_FLOPPYB_STATUS)->get () == BX_INSERTED)) {
// tell the device model that we removed, then inserted the disk
bx_devices.floppy->set_media_status(1, 0);
bx_devices.floppy->set_media_status(1, 1);
}
} else {
SIM->get_param_num(BXP_FLOPPYB_TYPE)->set_enabled (!empty);
SIM->get_param_num(BXP_FLOPPYB_STATUS)->set_enabled (!empty);
}
- now the memory options menu, interface menu, and disk menu in the control panel are all implemented as bx_list_c, and look much like they did before. - removed many hardcoded UI functions from control.cc, since the much more general "text_ask" methods have replaced them. - add range checking on integer parameter values. This exposed several cases where my initial value for an integer parameter was not in range. - cleaned up behavior of get/set methods. The get/set methods allow the handler to override the value that is returned/set, or perform side effects. - the title parameter of a bx_list_c now defaults to the name. - now bx_param_c fields that used to be private are protected instead - removed references to bx_any - moved definition of set_handler from siminterface.h to siminterface.cc. I was considering with doing a "set" of the old value when the handler is first installed, but that remains commented out. - BX_BOOT_DISKA and BX_BOOT_DISKC are now 0,1 so that they can correspond with the values of a bx_param_enum_c. For a while they were 0x00 and 0x80 corresponding to the numbering convention of the bios, but it didn't really matter.
2001-06-21 18:37:55 +04:00
}
break;
case BXP_DISKC_PATH:
if (set==1) {
SIM->get_param_num(BXP_DISKC_PRESENT)->set (!empty);
SIM->get_param_num(BXP_DISKC_CYLINDERS)->set_enabled (!empty);
SIM->get_param_num(BXP_DISKC_HEADS)->set_enabled (!empty);
SIM->get_param_num(BXP_DISKC_SPT)->set_enabled (!empty);
}
break;
case BXP_DISKD_PATH:
if (set==1) {
SIM->get_param_num(BXP_DISKD_PRESENT)->set (!empty);
SIM->get_param_num(BXP_DISKD_CYLINDERS)->set_enabled (!empty);
SIM->get_param_num(BXP_DISKD_HEADS)->set_enabled (!empty);
SIM->get_param_num(BXP_DISKD_SPT)->set_enabled (!empty);
}
break;
case BXP_CDROM_PATH:
if (set==1) {
- cdrom and floppy options in configuration interface partially fixed * don't change the device state to "not present" at runtime if the path is changed to "none" - simply eject the media if inserted * don't change the device state to "present" at runtime if the previous state is "not present" * eject and re-insert the media if the current media state is "inserted" * changing the media state in the config interface works now * TODO: don't do anything if the path is unchanged - bx_write_cdrom_options(): writing media state fixed
2002-03-14 23:14:47 +03:00
if (SIM->get_init_done ()) {
if (empty) {
bx_devices.hard_drive->set_cd_media_status(0);
- renamed update_floppy_status_buttons() to update_drive_status_buttons() since it handles the cdrom status too
2002-03-17 23:57:54 +03:00
bx_gui.update_drive_status_buttons ();
- cdrom and floppy options in configuration interface partially fixed * don't change the device state to "not present" at runtime if the path is changed to "none" - simply eject the media if inserted * don't change the device state to "present" at runtime if the previous state is "not present" * eject and re-insert the media if the current media state is "inserted" * changing the media state in the config interface works now * TODO: don't do anything if the path is unchanged - bx_write_cdrom_options(): writing media state fixed
2002-03-14 23:14:47 +03:00
} else {
if (!SIM->get_param_num(BXP_CDROM_PRESENT)->get ()) {
BX_ERROR(("Cannot add a cdrom drive at runtime"));
bx_options.cdromd.Opath->set ("none");
}
}
if ((bx_devices.hard_drive) &&
(SIM->get_param_num(BXP_CDROM_INSERTED)->get () == BX_INSERTED)) {
// tell the device model that we removed, then inserted the cd
bx_devices.hard_drive->set_cd_media_status(0);
bx_devices.hard_drive->set_cd_media_status(1);
}
} else {
SIM->get_param_num(BXP_CDROM_PRESENT)->set (!empty);
SIM->get_param_num(BXP_CDROM_INSERTED)->set_enabled (!empty);
}
- now the memory options menu, interface menu, and disk menu in the control panel are all implemented as bx_list_c, and look much like they did before. - removed many hardcoded UI functions from control.cc, since the much more general "text_ask" methods have replaced them. - add range checking on integer parameter values. This exposed several cases where my initial value for an integer parameter was not in range. - cleaned up behavior of get/set methods. The get/set methods allow the handler to override the value that is returned/set, or perform side effects. - the title parameter of a bx_list_c now defaults to the name. - now bx_param_c fields that used to be private are protected instead - removed references to bx_any - moved definition of set_handler from siminterface.h to siminterface.cc. I was considering with doing a "set" of the old value when the handler is first installed, but that remains commented out. - BX_BOOT_DISKA and BX_BOOT_DISKC are now 0,1 so that they can correspond with the values of a bx_param_enum_c. For a while they were 0x00 and 0x80 corresponding to the numbering convention of the bios, but it didn't really matter.
2001-06-21 18:37:55 +04:00
}
break;
- add string parameter called screenmode - put #if BX_USE_AMIGAOS...#endif around fullscreen and screenmode, since they only apply to amiga
2001-08-16 06:00:31 +04:00
case BXP_SCREENMODE:
if (set==1) {
BX_INFO (("Screen mode changed to %s", val));
}
break;
- tweaks to get rid of compile warnings
2001-09-11 20:56:48 +04:00
default:
BX_PANIC (("bx_string_handler called with unexpected parameter %d", param->get_id()));
- now the memory options menu, interface menu, and disk menu in the control panel are all implemented as bx_list_c, and look much like they did before. - removed many hardcoded UI functions from control.cc, since the much more general "text_ask" methods have replaced them. - add range checking on integer parameter values. This exposed several cases where my initial value for an integer parameter was not in range. - cleaned up behavior of get/set methods. The get/set methods allow the handler to override the value that is returned/set, or perform side effects. - the title parameter of a bx_list_c now defaults to the name. - now bx_param_c fields that used to be private are protected instead - removed references to bx_any - moved definition of set_handler from siminterface.h to siminterface.cc. I was considering with doing a "set" of the old value when the handler is first installed, but that remains commented out. - BX_BOOT_DISKA and BX_BOOT_DISKC are now 0,1 so that they can correspond with the values of a bx_param_enum_c. For a while they were 0x00 and 0x80 corresponding to the numbering convention of the bios, but it didn't really matter.
2001-06-21 18:37:55 +04:00
}
return val;
}
- now megs, ips, mouse_enabled, vga_update_interval, and rom_path are using Bryce's experimental parameter structures.
2001-06-16 23:29:59 +04:00
void bx_init_options ()
{
- now the memory options menu, interface menu, and disk menu in the control panel are all implemented as bx_list_c, and look much like they did before. - removed many hardcoded UI functions from control.cc, since the much more general "text_ask" methods have replaced them. - add range checking on integer parameter values. This exposed several cases where my initial value for an integer parameter was not in range. - cleaned up behavior of get/set methods. The get/set methods allow the handler to override the value that is returned/set, or perform side effects. - the title parameter of a bx_list_c now defaults to the name. - now bx_param_c fields that used to be private are protected instead - removed references to bx_any - moved definition of set_handler from siminterface.h to siminterface.cc. I was considering with doing a "set" of the old value when the handler is first installed, but that remains commented out. - BX_BOOT_DISKA and BX_BOOT_DISKC are now 0,1 so that they can correspond with the values of a bx_param_enum_c. For a while they were 0x00 and 0x80 corresponding to the numbering convention of the bios, but it didn't really matter.
2001-06-21 18:37:55 +04:00
bx_list_c *menu;
// floppya
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.floppya.Opath = new bx_param_string_c (BXP_FLOPPYA_PATH,
- now the memory options menu, interface menu, and disk menu in the control panel are all implemented as bx_list_c, and look much like they did before. - removed many hardcoded UI functions from control.cc, since the much more general "text_ask" methods have replaced them. - add range checking on integer parameter values. This exposed several cases where my initial value for an integer parameter was not in range. - cleaned up behavior of get/set methods. The get/set methods allow the handler to override the value that is returned/set, or perform side effects. - the title parameter of a bx_list_c now defaults to the name. - now bx_param_c fields that used to be private are protected instead - removed references to bx_any - moved definition of set_handler from siminterface.h to siminterface.cc. I was considering with doing a "set" of the old value when the handler is first installed, but that remains commented out. - BX_BOOT_DISKA and BX_BOOT_DISKC are now 0,1 so that they can correspond with the values of a bx_param_enum_c. For a while they were 0x00 and 0x80 corresponding to the numbering convention of the bios, but it didn't really matter.
2001-06-21 18:37:55 +04:00
"Floppy A image",
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
"Pathname of first floppy image file or device. If you're booting from floppy, this should be a bootable floppy.",
"", BX_PATHNAME_LEN);
- now the memory options menu, interface menu, and disk menu in the control panel are all implemented as bx_list_c, and look much like they did before. - removed many hardcoded UI functions from control.cc, since the much more general "text_ask" methods have replaced them. - add range checking on integer parameter values. This exposed several cases where my initial value for an integer parameter was not in range. - cleaned up behavior of get/set methods. The get/set methods allow the handler to override the value that is returned/set, or perform side effects. - the title parameter of a bx_list_c now defaults to the name. - now bx_param_c fields that used to be private are protected instead - removed references to bx_any - moved definition of set_handler from siminterface.h to siminterface.cc. I was considering with doing a "set" of the old value when the handler is first installed, but that remains commented out. - BX_BOOT_DISKA and BX_BOOT_DISKC are now 0,1 so that they can correspond with the values of a bx_param_enum_c. For a while they were 0x00 and 0x80 corresponding to the numbering convention of the bios, but it didn't really matter.
2001-06-21 18:37:55 +04:00
bx_options.floppya.Opath->set_ask_format ("Enter new filename, or 'none' for no disk: [%s] ");
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.floppya.Otype = new bx_param_enum_c (BXP_FLOPPYA_TYPE,
"floppya:type",
"Type of floppy disk",
- now the memory options menu, interface menu, and disk menu in the control panel are all implemented as bx_list_c, and look much like they did before. - removed many hardcoded UI functions from control.cc, since the much more general "text_ask" methods have replaced them. - add range checking on integer parameter values. This exposed several cases where my initial value for an integer parameter was not in range. - cleaned up behavior of get/set methods. The get/set methods allow the handler to override the value that is returned/set, or perform side effects. - the title parameter of a bx_list_c now defaults to the name. - now bx_param_c fields that used to be private are protected instead - removed references to bx_any - moved definition of set_handler from siminterface.h to siminterface.cc. I was considering with doing a "set" of the old value when the handler is first installed, but that remains commented out. - BX_BOOT_DISKA and BX_BOOT_DISKC are now 0,1 so that they can correspond with the values of a bx_param_enum_c. For a while they were 0x00 and 0x80 corresponding to the numbering convention of the bios, but it didn't really matter.
2001-06-21 18:37:55 +04:00
floppy_type_names,
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
BX_FLOPPY_NONE,
BX_FLOPPY_NONE);
- now the memory options menu, interface menu, and disk menu in the control panel are all implemented as bx_list_c, and look much like they did before. - removed many hardcoded UI functions from control.cc, since the much more general "text_ask" methods have replaced them. - add range checking on integer parameter values. This exposed several cases where my initial value for an integer parameter was not in range. - cleaned up behavior of get/set methods. The get/set methods allow the handler to override the value that is returned/set, or perform side effects. - the title parameter of a bx_list_c now defaults to the name. - now bx_param_c fields that used to be private are protected instead - removed references to bx_any - moved definition of set_handler from siminterface.h to siminterface.cc. I was considering with doing a "set" of the old value when the handler is first installed, but that remains commented out. - BX_BOOT_DISKA and BX_BOOT_DISKC are now 0,1 so that they can correspond with the values of a bx_param_enum_c. For a while they were 0x00 and 0x80 corresponding to the numbering convention of the bios, but it didn't really matter.
2001-06-21 18:37:55 +04:00
bx_options.floppya.Otype->set_ask_format ("What type of floppy disk? [%s] ");
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.floppya.Oinitial_status = new bx_param_enum_c (BXP_FLOPPYA_STATUS,
- now the memory options menu, interface menu, and disk menu in the control panel are all implemented as bx_list_c, and look much like they did before. - removed many hardcoded UI functions from control.cc, since the much more general "text_ask" methods have replaced them. - add range checking on integer parameter values. This exposed several cases where my initial value for an integer parameter was not in range. - cleaned up behavior of get/set methods. The get/set methods allow the handler to override the value that is returned/set, or perform side effects. - the title parameter of a bx_list_c now defaults to the name. - now bx_param_c fields that used to be private are protected instead - removed references to bx_any - moved definition of set_handler from siminterface.h to siminterface.cc. I was considering with doing a "set" of the old value when the handler is first installed, but that remains commented out. - BX_BOOT_DISKA and BX_BOOT_DISKC are now 0,1 so that they can correspond with the values of a bx_param_enum_c. For a while they were 0x00 and 0x80 corresponding to the numbering convention of the bios, but it didn't really matter.
2001-06-21 18:37:55 +04:00
"Is floppya inserted",
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
"Inserted or ejected",
- now the memory options menu, interface menu, and disk menu in the control panel are all implemented as bx_list_c, and look much like they did before. - removed many hardcoded UI functions from control.cc, since the much more general "text_ask" methods have replaced them. - add range checking on integer parameter values. This exposed several cases where my initial value for an integer parameter was not in range. - cleaned up behavior of get/set methods. The get/set methods allow the handler to override the value that is returned/set, or perform side effects. - the title parameter of a bx_list_c now defaults to the name. - now bx_param_c fields that used to be private are protected instead - removed references to bx_any - moved definition of set_handler from siminterface.h to siminterface.cc. I was considering with doing a "set" of the old value when the handler is first installed, but that remains commented out. - BX_BOOT_DISKA and BX_BOOT_DISKC are now 0,1 so that they can correspond with the values of a bx_param_enum_c. For a while they were 0x00 and 0x80 corresponding to the numbering convention of the bios, but it didn't really matter.
2001-06-21 18:37:55 +04:00
floppy_status_names,
BX_INSERTED,
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
BX_EJECTED);
- now the memory options menu, interface menu, and disk menu in the control panel are all implemented as bx_list_c, and look much like they did before. - removed many hardcoded UI functions from control.cc, since the much more general "text_ask" methods have replaced them. - add range checking on integer parameter values. This exposed several cases where my initial value for an integer parameter was not in range. - cleaned up behavior of get/set methods. The get/set methods allow the handler to override the value that is returned/set, or perform side effects. - the title parameter of a bx_list_c now defaults to the name. - now bx_param_c fields that used to be private are protected instead - removed references to bx_any - moved definition of set_handler from siminterface.h to siminterface.cc. I was considering with doing a "set" of the old value when the handler is first installed, but that remains commented out. - BX_BOOT_DISKA and BX_BOOT_DISKC are now 0,1 so that they can correspond with the values of a bx_param_enum_c. For a while they were 0x00 and 0x80 corresponding to the numbering convention of the bios, but it didn't really matter.
2001-06-21 18:37:55 +04:00
bx_options.floppya.Oinitial_status->set_ask_format ("Is the floppy inserted or ejected? [%s] ");
bx_options.floppya.Opath->set_format ("%s");
bx_options.floppya.Otype->set_format (", size=%s, ");
bx_options.floppya.Oinitial_status->set_format ("%s");
bx_param_c *floppya_init_list[] = {
bx_options.floppya.Opath,
bx_options.floppya.Otype,
bx_options.floppya.Oinitial_status,
NULL
};
menu = new bx_list_c (BXP_FLOPPYA, "Floppy Disk 0", "All options for first floppy disk", floppya_init_list);
menu->get_options ()->set (menu->BX_SERIES_ASK);
bx_options.floppya.Opath->set_handler (bx_param_string_handler);
bx_options.floppya.Opath->set ("none");
- cdrom and floppy options in configuration interface partially fixed * don't change the device state to "not present" at runtime if the path is changed to "none" - simply eject the media if inserted * don't change the device state to "present" at runtime if the previous state is "not present" * eject and re-insert the media if the current media state is "inserted" * changing the media state in the config interface works now * TODO: don't do anything if the path is unchanged - bx_write_cdrom_options(): writing media state fixed
2002-03-14 23:14:47 +03:00
bx_options.floppya.Oinitial_status->set_handler (bx_param_handler);
- now the memory options menu, interface menu, and disk menu in the control panel are all implemented as bx_list_c, and look much like they did before. - removed many hardcoded UI functions from control.cc, since the much more general "text_ask" methods have replaced them. - add range checking on integer parameter values. This exposed several cases where my initial value for an integer parameter was not in range. - cleaned up behavior of get/set methods. The get/set methods allow the handler to override the value that is returned/set, or perform side effects. - the title parameter of a bx_list_c now defaults to the name. - now bx_param_c fields that used to be private are protected instead - removed references to bx_any - moved definition of set_handler from siminterface.h to siminterface.cc. I was considering with doing a "set" of the old value when the handler is first installed, but that remains commented out. - BX_BOOT_DISKA and BX_BOOT_DISKC are now 0,1 so that they can correspond with the values of a bx_param_enum_c. For a while they were 0x00 and 0x80 corresponding to the numbering convention of the bios, but it didn't really matter.
2001-06-21 18:37:55 +04:00
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.floppyb.Opath = new bx_param_string_c (BXP_FLOPPYB_PATH,
"floppyb:path",
"Pathname of second floppy image file or device.",
"", BX_PATHNAME_LEN);
- now the memory options menu, interface menu, and disk menu in the control panel are all implemented as bx_list_c, and look much like they did before. - removed many hardcoded UI functions from control.cc, since the much more general "text_ask" methods have replaced them. - add range checking on integer parameter values. This exposed several cases where my initial value for an integer parameter was not in range. - cleaned up behavior of get/set methods. The get/set methods allow the handler to override the value that is returned/set, or perform side effects. - the title parameter of a bx_list_c now defaults to the name. - now bx_param_c fields that used to be private are protected instead - removed references to bx_any - moved definition of set_handler from siminterface.h to siminterface.cc. I was considering with doing a "set" of the old value when the handler is first installed, but that remains commented out. - BX_BOOT_DISKA and BX_BOOT_DISKC are now 0,1 so that they can correspond with the values of a bx_param_enum_c. For a while they were 0x00 and 0x80 corresponding to the numbering convention of the bios, but it didn't really matter.
2001-06-21 18:37:55 +04:00
bx_options.floppyb.Opath->set_ask_format ("Enter new filename, or 'none' for no disk: [%s] ");
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.floppyb.Otype = new bx_param_enum_c (BXP_FLOPPYB_TYPE,
"floppyb:type",
"Type of floppy disk",
floppy_type_names,
BX_FLOPPY_NONE,
BX_FLOPPY_NONE);
- now the memory options menu, interface menu, and disk menu in the control panel are all implemented as bx_list_c, and look much like they did before. - removed many hardcoded UI functions from control.cc, since the much more general "text_ask" methods have replaced them. - add range checking on integer parameter values. This exposed several cases where my initial value for an integer parameter was not in range. - cleaned up behavior of get/set methods. The get/set methods allow the handler to override the value that is returned/set, or perform side effects. - the title parameter of a bx_list_c now defaults to the name. - now bx_param_c fields that used to be private are protected instead - removed references to bx_any - moved definition of set_handler from siminterface.h to siminterface.cc. I was considering with doing a "set" of the old value when the handler is first installed, but that remains commented out. - BX_BOOT_DISKA and BX_BOOT_DISKC are now 0,1 so that they can correspond with the values of a bx_param_enum_c. For a while they were 0x00 and 0x80 corresponding to the numbering convention of the bios, but it didn't really matter.
2001-06-21 18:37:55 +04:00
bx_options.floppyb.Otype->set_ask_format ("What type of floppy disk? [%s] ");
bx_options.floppyb.Oinitial_status = new bx_param_enum_c (BXP_FLOPPYB_STATUS,
"Is floppyb inserted",
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
"Inserted or ejected",
- now the memory options menu, interface menu, and disk menu in the control panel are all implemented as bx_list_c, and look much like they did before. - removed many hardcoded UI functions from control.cc, since the much more general "text_ask" methods have replaced them. - add range checking on integer parameter values. This exposed several cases where my initial value for an integer parameter was not in range. - cleaned up behavior of get/set methods. The get/set methods allow the handler to override the value that is returned/set, or perform side effects. - the title parameter of a bx_list_c now defaults to the name. - now bx_param_c fields that used to be private are protected instead - removed references to bx_any - moved definition of set_handler from siminterface.h to siminterface.cc. I was considering with doing a "set" of the old value when the handler is first installed, but that remains commented out. - BX_BOOT_DISKA and BX_BOOT_DISKC are now 0,1 so that they can correspond with the values of a bx_param_enum_c. For a while they were 0x00 and 0x80 corresponding to the numbering convention of the bios, but it didn't really matter.
2001-06-21 18:37:55 +04:00
floppy_status_names,
BX_INSERTED,
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
BX_EJECTED);
- now the memory options menu, interface menu, and disk menu in the control panel are all implemented as bx_list_c, and look much like they did before. - removed many hardcoded UI functions from control.cc, since the much more general "text_ask" methods have replaced them. - add range checking on integer parameter values. This exposed several cases where my initial value for an integer parameter was not in range. - cleaned up behavior of get/set methods. The get/set methods allow the handler to override the value that is returned/set, or perform side effects. - the title parameter of a bx_list_c now defaults to the name. - now bx_param_c fields that used to be private are protected instead - removed references to bx_any - moved definition of set_handler from siminterface.h to siminterface.cc. I was considering with doing a "set" of the old value when the handler is first installed, but that remains commented out. - BX_BOOT_DISKA and BX_BOOT_DISKC are now 0,1 so that they can correspond with the values of a bx_param_enum_c. For a while they were 0x00 and 0x80 corresponding to the numbering convention of the bios, but it didn't really matter.
2001-06-21 18:37:55 +04:00
bx_options.floppyb.Oinitial_status->set_ask_format ("Is the floppy inserted or ejected? [%s] ");
bx_options.floppyb.Oinitial_status->set_format ("%s");
bx_options.floppyb.Opath->set_format ("%s");
bx_options.floppyb.Otype->set_format (", size=%s, ");
bx_options.floppyb.Oinitial_status->set_format ("%s");
bx_param_c *floppyb_init_list[] = {
bx_options.floppyb.Opath,
bx_options.floppyb.Otype,
bx_options.floppyb.Oinitial_status,
NULL
};
menu = new bx_list_c (BXP_FLOPPYB, "Floppy Disk 1", "All options for second floppy disk", floppyb_init_list);
menu->get_options ()->set (menu->BX_SERIES_ASK);
bx_options.floppyb.Opath->set_handler (bx_param_string_handler);
bx_options.floppyb.Opath->set ("none");
- cdrom and floppy options in configuration interface partially fixed * don't change the device state to "not present" at runtime if the path is changed to "none" - simply eject the media if inserted * don't change the device state to "present" at runtime if the previous state is "not present" * eject and re-insert the media if the current media state is "inserted" * changing the media state in the config interface works now * TODO: don't do anything if the path is unchanged - bx_write_cdrom_options(): writing media state fixed
2002-03-14 23:14:47 +03:00
bx_options.floppyb.Oinitial_status->set_handler (bx_param_handler);
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
// diskc options
bx_options.diskc.Opresent = new bx_param_bool_c (BXP_DISKC_PRESENT,
"diskc:present",
"Controls whether diskc is installed or not",
0);
bx_options.diskc.Opath = new bx_param_string_c (BXP_DISKC_PATH,
- now the memory options menu, interface menu, and disk menu in the control panel are all implemented as bx_list_c, and look much like they did before. - removed many hardcoded UI functions from control.cc, since the much more general "text_ask" methods have replaced them. - add range checking on integer parameter values. This exposed several cases where my initial value for an integer parameter was not in range. - cleaned up behavior of get/set methods. The get/set methods allow the handler to override the value that is returned/set, or perform side effects. - the title parameter of a bx_list_c now defaults to the name. - now bx_param_c fields that used to be private are protected instead - removed references to bx_any - moved definition of set_handler from siminterface.h to siminterface.cc. I was considering with doing a "set" of the old value when the handler is first installed, but that remains commented out. - BX_BOOT_DISKA and BX_BOOT_DISKC are now 0,1 so that they can correspond with the values of a bx_param_enum_c. For a while they were 0x00 and 0x80 corresponding to the numbering convention of the bios, but it didn't really matter.
2001-06-21 18:37:55 +04:00
"",
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
"Pathname of the hard drive image",
"", BX_PATHNAME_LEN);
bx_options.diskc.Ocylinders = new bx_param_num_c (BXP_DISKC_CYLINDERS,
"diskc:cylinders",
"Number of cylinders",
- fix bugs introduced in revision 1.48 which caused the default number of cylinders, heads, and spt of each hard disk to be 1 instead of 0. This caused the BIOS to report the existence of the hard disks, even when they were never configured. The most obvious symptom of this problem was "read multiple issued to non-disk" when a cdrom was enabled, which happened because the BIOS tried to send hard disk commands to the CDROM.
2001-06-23 22:03:55 +04:00
0, 65535,
0);
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.diskc.Oheads = new bx_param_num_c (BXP_DISKC_HEADS,
"diskc:heads",
"Number of heads",
- fix bugs introduced in revision 1.48 which caused the default number of cylinders, heads, and spt of each hard disk to be 1 instead of 0. This caused the BIOS to report the existence of the hard disks, even when they were never configured. The most obvious symptom of this problem was "read multiple issued to non-disk" when a cdrom was enabled, which happened because the BIOS tried to send hard disk commands to the CDROM.
2001-06-23 22:03:55 +04:00
0, 65535,
0);
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.diskc.Ospt = new bx_param_num_c (BXP_DISKC_SPT,
"diskc:spt",
"Number of sectors per track",
- fix bugs introduced in revision 1.48 which caused the default number of cylinders, heads, and spt of each hard disk to be 1 instead of 0. This caused the BIOS to report the existence of the hard disks, even when they were never configured. The most obvious symptom of this problem was "read multiple issued to non-disk" when a cdrom was enabled, which happened because the BIOS tried to send hard disk commands to the CDROM.
2001-06-23 22:03:55 +04:00
0, 65535,
0);
- now the memory options menu, interface menu, and disk menu in the control panel are all implemented as bx_list_c, and look much like they did before. - removed many hardcoded UI functions from control.cc, since the much more general "text_ask" methods have replaced them. - add range checking on integer parameter values. This exposed several cases where my initial value for an integer parameter was not in range. - cleaned up behavior of get/set methods. The get/set methods allow the handler to override the value that is returned/set, or perform side effects. - the title parameter of a bx_list_c now defaults to the name. - now bx_param_c fields that used to be private are protected instead - removed references to bx_any - moved definition of set_handler from siminterface.h to siminterface.cc. I was considering with doing a "set" of the old value when the handler is first installed, but that remains commented out. - BX_BOOT_DISKA and BX_BOOT_DISKC are now 0,1 so that they can correspond with the values of a bx_param_enum_c. For a while they were 0x00 and 0x80 corresponding to the numbering convention of the bios, but it didn't really matter.
2001-06-21 18:37:55 +04:00
bx_options.diskc.Opath->set_ask_format ("Enter new filename, or 'none' for no disk: [%s] ");
bx_options.diskc.Ocylinders->set_ask_format ("Enter number of cylinders: [%d] ");
bx_options.diskc.Oheads->set_ask_format ("Enter number of heads: [%d] ");
bx_options.diskc.Ospt->set_ask_format ("Enter number of sectors per track: [%d] ");
bx_options.diskc.Opath->set_format ("%s");
bx_options.diskc.Ocylinders->set_format (", %d cylinders, ");
bx_options.diskc.Oheads->set_format ("%d heads, ");
bx_options.diskc.Ospt->set_format ("%d sectors/track");
bx_param_c *diskc_init_list[] = {
bx_options.diskc.Opath,
bx_options.diskc.Ocylinders,
bx_options.diskc.Oheads,
bx_options.diskc.Ospt,
NULL
};
menu = new bx_list_c (BXP_DISKC, "Hard disk 0", "All options for hard disk 0", diskc_init_list);
menu->get_options ()->set (menu->BX_SERIES_ASK);
// if path is the word "none", then do not ask the other options and
// set present=0.
bx_options.diskc.Opath->set_handler (bx_param_string_handler);
bx_options.diskc.Opath->set ("none");
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
// diskd options
bx_options.diskd.Opresent = new bx_param_bool_c (BXP_DISKD_PRESENT,
"diskd:present",
"Controls whether diskd is installed or not",
0);
bx_options.diskd.Opath = new bx_param_string_c (BXP_DISKD_PATH,
"diskd:path",
"Pathname of the hard drive image",
"", BX_PATHNAME_LEN);
bx_options.diskd.Ocylinders = new bx_param_num_c (BXP_DISKD_CYLINDERS,
"diskd:cylinders",
"Number of cylinders",
- fix bugs introduced in revision 1.48 which caused the default number of cylinders, heads, and spt of each hard disk to be 1 instead of 0. This caused the BIOS to report the existence of the hard disks, even when they were never configured. The most obvious symptom of this problem was "read multiple issued to non-disk" when a cdrom was enabled, which happened because the BIOS tried to send hard disk commands to the CDROM.
2001-06-23 22:03:55 +04:00
0, 65535,
0);
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.diskd.Oheads = new bx_param_num_c (BXP_DISKD_HEADS,
"diskd:heads",
"Number of heads",
- fix bugs introduced in revision 1.48 which caused the default number of cylinders, heads, and spt of each hard disk to be 1 instead of 0. This caused the BIOS to report the existence of the hard disks, even when they were never configured. The most obvious symptom of this problem was "read multiple issued to non-disk" when a cdrom was enabled, which happened because the BIOS tried to send hard disk commands to the CDROM.
2001-06-23 22:03:55 +04:00
0, 65535,
0);
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.diskd.Ospt = new bx_param_num_c (BXP_DISKD_SPT,
"diskd:spt",
"Number of sectors per track",
- fix bugs introduced in revision 1.48 which caused the default number of cylinders, heads, and spt of each hard disk to be 1 instead of 0. This caused the BIOS to report the existence of the hard disks, even when they were never configured. The most obvious symptom of this problem was "read multiple issued to non-disk" when a cdrom was enabled, which happened because the BIOS tried to send hard disk commands to the CDROM.
2001-06-23 22:03:55 +04:00
0, 65535,
0);
- now the memory options menu, interface menu, and disk menu in the control panel are all implemented as bx_list_c, and look much like they did before. - removed many hardcoded UI functions from control.cc, since the much more general "text_ask" methods have replaced them. - add range checking on integer parameter values. This exposed several cases where my initial value for an integer parameter was not in range. - cleaned up behavior of get/set methods. The get/set methods allow the handler to override the value that is returned/set, or perform side effects. - the title parameter of a bx_list_c now defaults to the name. - now bx_param_c fields that used to be private are protected instead - removed references to bx_any - moved definition of set_handler from siminterface.h to siminterface.cc. I was considering with doing a "set" of the old value when the handler is first installed, but that remains commented out. - BX_BOOT_DISKA and BX_BOOT_DISKC are now 0,1 so that they can correspond with the values of a bx_param_enum_c. For a while they were 0x00 and 0x80 corresponding to the numbering convention of the bios, but it didn't really matter.
2001-06-21 18:37:55 +04:00
bx_options.diskd.Opath->set_ask_format ("Enter new filename, or none for no disk: [%s] ");
bx_options.diskd.Ocylinders->set_ask_format ("Enter number of cylinders: [%d] ");
bx_options.diskd.Oheads->set_ask_format ("Enter number of heads: [%d] ");
bx_options.diskd.Ospt->set_ask_format ("Enter number of sectors per track: [%d] ");
bx_options.diskd.Opath->set_format ("%s");
bx_options.diskd.Ocylinders->set_format (", %d cylinders, ");
bx_options.diskd.Oheads->set_format ("%d heads, ");
bx_options.diskd.Ospt->set_format ("%d sectors/track");
bx_param_c *diskd_init_list[] = {
bx_options.diskd.Opath,
bx_options.diskd.Ocylinders,
bx_options.diskd.Oheads,
bx_options.diskd.Ospt,
NULL
};
menu = new bx_list_c (BXP_DISKD, "Hard disk 1", "All options for hard disk 1", diskd_init_list);
menu->get_options ()->set (menu->BX_SERIES_ASK);
bx_options.diskd.Opath->set_handler (bx_param_string_handler);
bx_options.diskd.Opath->set ("none");
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
- commit serial port patch from Stu Grossman <grossman@juniper.net> Here are the patches (without the serial name fix). Let me know if you want me to change that. Oh, FYI, my changes also open up the com1 terminal in raw mode, which makes the emulation much more accurate. You'll also notice that I added infrastructure for com2->4 in the option parsing. I didn't add it to the serial code, as I think a bunch of things need to be untangled before that can work.
2002-03-03 09:03:29 +03:00
// com1 options
bx_options.com1.Opresent = new bx_param_bool_c (BXP_COM1_PRESENT,
- add serial port options to the configuration interface, in the same menu as the parallel port options.
2002-03-03 09:10:04 +03:00
"Enable serial port #1 (COM1)",
"",
- commit serial port patch from Stu Grossman <grossman@juniper.net> Here are the patches (without the serial name fix). Let me know if you want me to change that. Oh, FYI, my changes also open up the com1 terminal in raw mode, which makes the emulation much more accurate. You'll also notice that I added infrastructure for com2->4 in the option parsing. I didn't add it to the serial code, as I think a bunch of things need to be untangled before that can work.
2002-03-03 09:03:29 +03:00
0);
bx_options.com1.Odev = new bx_param_string_c (BXP_COM1_PATH,
- add serial port options to the configuration interface, in the same menu as the parallel port options.
2002-03-03 09:10:04 +03:00
"Pathname of the serial device for COM1",
- commit serial port patch from Stu Grossman <grossman@juniper.net> Here are the patches (without the serial name fix). Let me know if you want me to change that. Oh, FYI, my changes also open up the com1 terminal in raw mode, which makes the emulation much more accurate. You'll also notice that I added infrastructure for com2->4 in the option parsing. I didn't add it to the serial code, as I think a bunch of things need to be untangled before that can work.
2002-03-03 09:03:29 +03:00
"",
"", BX_PATHNAME_LEN);
- add serial port options to the configuration interface, in the same menu as the parallel port options.
2002-03-03 09:10:04 +03:00
bx_options.com1.Opresent->set_handler (bx_param_handler);
#if 0
- commit serial port patch from Stu Grossman <grossman@juniper.net> Here are the patches (without the serial name fix). Let me know if you want me to change that. Oh, FYI, my changes also open up the com1 terminal in raw mode, which makes the emulation much more accurate. You'll also notice that I added infrastructure for com2->4 in the option parsing. I didn't add it to the serial code, as I think a bunch of things need to be untangled before that can work.
2002-03-03 09:03:29 +03:00
// com2 options
bx_options.com2.Opresent = new bx_param_bool_c (BXP_COM2_PRESENT,
- add serial port options to the configuration interface, in the same menu as the parallel port options.
2002-03-03 09:10:04 +03:00
"Enable serial port #2 (COM1)",
- commit serial port patch from Stu Grossman <grossman@juniper.net> Here are the patches (without the serial name fix). Let me know if you want me to change that. Oh, FYI, my changes also open up the com1 terminal in raw mode, which makes the emulation much more accurate. You'll also notice that I added infrastructure for com2->4 in the option parsing. I didn't add it to the serial code, as I think a bunch of things need to be untangled before that can work.
2002-03-03 09:03:29 +03:00
"Controls whether com2 is installed or not",
0);
bx_options.com2.Odev = new bx_param_string_c (BXP_COM2_PATH,
"",
- add serial port options to the configuration interface, in the same menu as the parallel port options.
2002-03-03 09:10:04 +03:00
"",
- commit serial port patch from Stu Grossman <grossman@juniper.net> Here are the patches (without the serial name fix). Let me know if you want me to change that. Oh, FYI, my changes also open up the com1 terminal in raw mode, which makes the emulation much more accurate. You'll also notice that I added infrastructure for com2->4 in the option parsing. I didn't add it to the serial code, as I think a bunch of things need to be untangled before that can work.
2002-03-03 09:03:29 +03:00
"", BX_PATHNAME_LEN);
- add serial port options to the configuration interface, in the same menu as the parallel port options.
2002-03-03 09:10:04 +03:00
bx_options.com2.Opresent->set_handler (bx_param_handler);
- commit serial port patch from Stu Grossman <grossman@juniper.net> Here are the patches (without the serial name fix). Let me know if you want me to change that. Oh, FYI, my changes also open up the com1 terminal in raw mode, which makes the emulation much more accurate. You'll also notice that I added infrastructure for com2->4 in the option parsing. I didn't add it to the serial code, as I think a bunch of things need to be untangled before that can work.
2002-03-03 09:03:29 +03:00
// com3 options
bx_options.com3.Opresent = new bx_param_bool_c (BXP_COM3_PRESENT,
- add serial port options to the configuration interface, in the same menu as the parallel port options.
2002-03-03 09:10:04 +03:00
"Enable serial port #3 (COM1)",
"",
- commit serial port patch from Stu Grossman <grossman@juniper.net> Here are the patches (without the serial name fix). Let me know if you want me to change that. Oh, FYI, my changes also open up the com1 terminal in raw mode, which makes the emulation much more accurate. You'll also notice that I added infrastructure for com2->4 in the option parsing. I didn't add it to the serial code, as I think a bunch of things need to be untangled before that can work.
2002-03-03 09:03:29 +03:00
0);
bx_options.com3.Odev = new bx_param_string_c (BXP_COM3_PATH,
- add serial port options to the configuration interface, in the same menu as the parallel port options.
2002-03-03 09:10:04 +03:00
"Pathname of the serial device for COM3",
- commit serial port patch from Stu Grossman <grossman@juniper.net> Here are the patches (without the serial name fix). Let me know if you want me to change that. Oh, FYI, my changes also open up the com1 terminal in raw mode, which makes the emulation much more accurate. You'll also notice that I added infrastructure for com2->4 in the option parsing. I didn't add it to the serial code, as I think a bunch of things need to be untangled before that can work.
2002-03-03 09:03:29 +03:00
"",
"", BX_PATHNAME_LEN);
- add serial port options to the configuration interface, in the same menu as the parallel port options.
2002-03-03 09:10:04 +03:00
bx_options.com3.Opresent->set_handler (bx_param_handler);
- commit serial port patch from Stu Grossman <grossman@juniper.net> Here are the patches (without the serial name fix). Let me know if you want me to change that. Oh, FYI, my changes also open up the com1 terminal in raw mode, which makes the emulation much more accurate. You'll also notice that I added infrastructure for com2->4 in the option parsing. I didn't add it to the serial code, as I think a bunch of things need to be untangled before that can work.
2002-03-03 09:03:29 +03:00
// com4 options
bx_options.com4.Opresent = new bx_param_bool_c (BXP_COM4_PRESENT,
- add serial port options to the configuration interface, in the same menu as the parallel port options.
2002-03-03 09:10:04 +03:00
"Enable serial port #4 (COM1)",
"",
- commit serial port patch from Stu Grossman <grossman@juniper.net> Here are the patches (without the serial name fix). Let me know if you want me to change that. Oh, FYI, my changes also open up the com1 terminal in raw mode, which makes the emulation much more accurate. You'll also notice that I added infrastructure for com2->4 in the option parsing. I didn't add it to the serial code, as I think a bunch of things need to be untangled before that can work.
2002-03-03 09:03:29 +03:00
0);
bx_options.com4.Odev = new bx_param_string_c (BXP_COM4_PATH,
- add serial port options to the configuration interface, in the same menu as the parallel port options.
2002-03-03 09:10:04 +03:00
"Pathname of the serial device for COM4",
- commit serial port patch from Stu Grossman <grossman@juniper.net> Here are the patches (without the serial name fix). Let me know if you want me to change that. Oh, FYI, my changes also open up the com1 terminal in raw mode, which makes the emulation much more accurate. You'll also notice that I added infrastructure for com2->4 in the option parsing. I didn't add it to the serial code, as I think a bunch of things need to be untangled before that can work.
2002-03-03 09:03:29 +03:00
"",
"", BX_PATHNAME_LEN);
- add serial port options to the configuration interface, in the same menu as the parallel port options.
2002-03-03 09:10:04 +03:00
bx_options.com4.Opresent->set_handler (bx_param_handler);
#endif
- commit serial port patch from Stu Grossman <grossman@juniper.net> Here are the patches (without the serial name fix). Let me know if you want me to change that. Oh, FYI, my changes also open up the com1 terminal in raw mode, which makes the emulation much more accurate. You'll also notice that I added infrastructure for com2->4 in the option parsing. I didn't add it to the serial code, as I think a bunch of things need to be untangled before that can work.
2002-03-03 09:03:29 +03:00
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
// cdrom options
bx_options.cdromd.Opresent = new bx_param_bool_c (BXP_CDROM_PRESENT,
- now the memory options menu, interface menu, and disk menu in the control panel are all implemented as bx_list_c, and look much like they did before. - removed many hardcoded UI functions from control.cc, since the much more general "text_ask" methods have replaced them. - add range checking on integer parameter values. This exposed several cases where my initial value for an integer parameter was not in range. - cleaned up behavior of get/set methods. The get/set methods allow the handler to override the value that is returned/set, or perform side effects. - the title parameter of a bx_list_c now defaults to the name. - now bx_param_c fields that used to be private are protected instead - removed references to bx_any - moved definition of set_handler from siminterface.h to siminterface.cc. I was considering with doing a "set" of the old value when the handler is first installed, but that remains commented out. - BX_BOOT_DISKA and BX_BOOT_DISKC are now 0,1 so that they can correspond with the values of a bx_param_enum_c. For a while they were 0x00 and 0x80 corresponding to the numbering convention of the bios, but it didn't really matter.
2001-06-21 18:37:55 +04:00
"CDROM is present",
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
"Controls whether cdromd is installed or not",
0);
bx_options.cdromd.Opath = new bx_param_string_c (BXP_CDROM_PATH,
- now the memory options menu, interface menu, and disk menu in the control panel are all implemented as bx_list_c, and look much like they did before. - removed many hardcoded UI functions from control.cc, since the much more general "text_ask" methods have replaced them. - add range checking on integer parameter values. This exposed several cases where my initial value for an integer parameter was not in range. - cleaned up behavior of get/set methods. The get/set methods allow the handler to override the value that is returned/set, or perform side effects. - the title parameter of a bx_list_c now defaults to the name. - now bx_param_c fields that used to be private are protected instead - removed references to bx_any - moved definition of set_handler from siminterface.h to siminterface.cc. I was considering with doing a "set" of the old value when the handler is first installed, but that remains commented out. - BX_BOOT_DISKA and BX_BOOT_DISKC are now 0,1 so that they can correspond with the values of a bx_param_enum_c. For a while they were 0x00 and 0x80 corresponding to the numbering convention of the bios, but it didn't really matter.
2001-06-21 18:37:55 +04:00
"CDROM image filename",
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
"Pathname of the cdrom device or image",
"", BX_PATHNAME_LEN);
- now the memory options menu, interface menu, and disk menu in the control panel are all implemented as bx_list_c, and look much like they did before. - removed many hardcoded UI functions from control.cc, since the much more general "text_ask" methods have replaced them. - add range checking on integer parameter values. This exposed several cases where my initial value for an integer parameter was not in range. - cleaned up behavior of get/set methods. The get/set methods allow the handler to override the value that is returned/set, or perform side effects. - the title parameter of a bx_list_c now defaults to the name. - now bx_param_c fields that used to be private are protected instead - removed references to bx_any - moved definition of set_handler from siminterface.h to siminterface.cc. I was considering with doing a "set" of the old value when the handler is first installed, but that remains commented out. - BX_BOOT_DISKA and BX_BOOT_DISKC are now 0,1 so that they can correspond with the values of a bx_param_enum_c. For a while they were 0x00 and 0x80 corresponding to the numbering convention of the bios, but it didn't really matter.
2001-06-21 18:37:55 +04:00
bx_options.cdromd.Opath->set_format ("%s");
bx_options.cdromd.Opath->set_ask_format ("Enter new filename, or 'none' for no CDROM: [%s] ");
bx_options.cdromd.Oinserted = new bx_param_enum_c (BXP_CDROM_INSERTED,
"Is the CDROM inserted or ejected",
"Inserted or ejected",
floppy_status_names,
BX_INSERTED,
BX_EJECTED);
bx_options.cdromd.Oinserted->set_format (", %s");
bx_options.cdromd.Oinserted->set_ask_format ("Is the CDROM inserted or ejected? [%s] ");
bx_param_c *cdromd_init_list[] = {
bx_options.cdromd.Opath,
bx_options.cdromd.Oinserted,
NULL
};
menu = new bx_list_c (BXP_CDROMD, "CDROM", "Options for the CDROM", cdromd_init_list);
menu->get_options ()->set (menu->BX_SERIES_ASK);
bx_options.cdromd.Opath->set_handler (bx_param_string_handler);
bx_options.cdromd.Opath->set ("none");
- cdrom and floppy options in configuration interface partially fixed * don't change the device state to "not present" at runtime if the path is changed to "none" - simply eject the media if inserted * don't change the device state to "present" at runtime if the previous state is "not present" * eject and re-insert the media if the current media state is "inserted" * changing the media state in the config interface works now * TODO: don't do anything if the path is unchanged - bx_write_cdrom_options(): writing media state fixed
2002-03-14 23:14:47 +03:00
bx_options.cdromd.Oinserted->set_handler (bx_param_handler);
- now the memory options menu, interface menu, and disk menu in the control panel are all implemented as bx_list_c, and look much like they did before. - removed many hardcoded UI functions from control.cc, since the much more general "text_ask" methods have replaced them. - add range checking on integer parameter values. This exposed several cases where my initial value for an integer parameter was not in range. - cleaned up behavior of get/set methods. The get/set methods allow the handler to override the value that is returned/set, or perform side effects. - the title parameter of a bx_list_c now defaults to the name. - now bx_param_c fields that used to be private are protected instead - removed references to bx_any - moved definition of set_handler from siminterface.h to siminterface.cc. I was considering with doing a "set" of the old value when the handler is first installed, but that remains commented out. - BX_BOOT_DISKA and BX_BOOT_DISKC are now 0,1 so that they can correspond with the values of a bx_param_enum_c. For a while they were 0x00 and 0x80 corresponding to the numbering convention of the bios, but it didn't really matter.
2001-06-21 18:37:55 +04:00
bx_options.OnewHardDriveSupport = new bx_param_bool_c (BXP_NEWHARDDRIVESUPPORT,
"New Hard Drive Support",
"Enables new features found on newer hard drives.",
1);
bx_options.Obootdrive = new bx_param_enum_c (BXP_BOOTDRIVE,
"bootdrive",
. changes to enable El-Torito boot. It is disabled by default in rombios.c
2002-01-30 13:30:52 +03:00
"Boot A, C or CD",
- now the memory options menu, interface menu, and disk menu in the control panel are all implemented as bx_list_c, and look much like they did before. - removed many hardcoded UI functions from control.cc, since the much more general "text_ask" methods have replaced them. - add range checking on integer parameter values. This exposed several cases where my initial value for an integer parameter was not in range. - cleaned up behavior of get/set methods. The get/set methods allow the handler to override the value that is returned/set, or perform side effects. - the title parameter of a bx_list_c now defaults to the name. - now bx_param_c fields that used to be private are protected instead - removed references to bx_any - moved definition of set_handler from siminterface.h to siminterface.cc. I was considering with doing a "set" of the old value when the handler is first installed, but that remains commented out. - BX_BOOT_DISKA and BX_BOOT_DISKC are now 0,1 so that they can correspond with the values of a bx_param_enum_c. For a while they were 0x00 and 0x80 corresponding to the numbering convention of the bios, but it didn't really matter.
2001-06-21 18:37:55 +04:00
floppy_bootdisk_names,
BX_BOOT_FLOPPYA,
BX_BOOT_FLOPPYA);
bx_options.Obootdrive->set_format ("Boot from: %s drive");
. changes to enable El-Torito boot. It is disabled by default in rombios.c
2002-01-30 13:30:52 +03:00
bx_options.Obootdrive->set_ask_format ("Boot from floppy drive, hard drive or cdrom ? [%s] ");
- now the memory options menu, interface menu, and disk menu in the control panel are all implemented as bx_list_c, and look much like they did before. - removed many hardcoded UI functions from control.cc, since the much more general "text_ask" methods have replaced them. - add range checking on integer parameter values. This exposed several cases where my initial value for an integer parameter was not in range. - cleaned up behavior of get/set methods. The get/set methods allow the handler to override the value that is returned/set, or perform side effects. - the title parameter of a bx_list_c now defaults to the name. - now bx_param_c fields that used to be private are protected instead - removed references to bx_any - moved definition of set_handler from siminterface.h to siminterface.cc. I was considering with doing a "set" of the old value when the handler is first installed, but that remains commented out. - BX_BOOT_DISKA and BX_BOOT_DISKC are now 0,1 so that they can correspond with the values of a bx_param_enum_c. For a while they were 0x00 and 0x80 corresponding to the numbering convention of the bios, but it didn't really matter.
2001-06-21 18:37:55 +04:00
// disk menu
bx_param_c *disk_menu_init_list[] = {
SIM->get_param (BXP_FLOPPYA),
SIM->get_param (BXP_FLOPPYB),
SIM->get_param (BXP_DISKC),
SIM->get_param (BXP_DISKD),
SIM->get_param (BXP_CDROMD),
SIM->get_param (BXP_NEWHARDDRIVESUPPORT),
SIM->get_param (BXP_BOOTDRIVE),
NULL
};
menu = new bx_list_c (BXP_MENU_DISK, "Bochs Disk Options", "diskmenu", disk_menu_init_list);
menu->get_options ()->set (menu->BX_SHOW_PARENT);
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
- now the "memory options" menu is entirely defined using parameter data structures, see bx_init_options in main.cc. The implementation of this menu and all its choices is 17 lines long, see do_mem_options_menu in gui/control.cc.
2001-06-18 18:11:55 +04:00
// memory options menu
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.memory.Osize = new bx_param_num_c (BXP_MEM_SIZE,
- now ips, mouse_enabled, vga_update_interval, and rom_path are declared as bx_param_c * types in the bx_options structure. They are initialized in main.cc (bx_init_options) with default values. Access to parameters of this type should always be like this: bx_options.mouse_enabled->get (); bx_options.mouse_enabled->set (newval); Eventually I will be transferring all options to this format.
2001-06-17 03:08:32 +04:00
"megs",
"Amount of RAM in megabytes",
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
1, BX_MAX_INT,
- now ips, mouse_enabled, vga_update_interval, and rom_path are declared as bx_param_c * types in the bx_options structure. They are initialized in main.cc (bx_init_options) with default values. Access to parameters of this type should always be like this: bx_options.mouse_enabled->get (); bx_options.mouse_enabled->set (newval); Eventually I will be transferring all options to this format.
2001-06-17 03:08:32 +04:00
BX_DEFAULT_MEM_MEGS);
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.memory.Osize->set_format ("Memory size in megabytes: %d");
bx_options.memory.Osize->set_ask_format ("Enter memory size (MB): [%d] ");
- allow configuration of the parallel port in .bochsrc and the config interface menus. Parallel port #1 is implemented, and I left stubs for parallel port #2 in case we want to ever add it. If the parallel port is enabled, the init method of parallel.cc does an fopen() on the output file. If disabled, or if the fopen fails, the file handler remains NULL and no characters are printed. There is no attempt to enable/disable the operation of the parallel port, only the output to a file.
2001-11-12 05:35:09 +03:00
bx_options.par1.Oenable = new bx_param_bool_c (BXP_PARPORT1_ENABLE,
"Enable parallel port #1",
"Enables the first parallel port",
0);
bx_options.par1.Oenable->set_handler (bx_param_handler);
bx_options.par1.Ooutfile = new bx_param_string_c (BXP_PARPORT1_OUTFILE,
"Parallel port #1 output file",
"Data written to parport#1 by the guest OS is written to this file",
"parport.out", BX_PATHNAME_LEN);
#if 0
bx_options.par2.Oenable = new bx_param_bool_c (BXP_PARPORT2_ENABLE,
"Enable parallel port #2",
"Enables the second parallel port",
0);
bx_options.par2.Oenable->set_handler (bx_param_handler);
bx_options.par2.Ooutfile = new bx_param_string_c (BXP_PARPORT2_OUTFILE,
"Parallel port #2 output file",
"Data written to parport#2 by the guest OS is written to this file",
"parport2.out", BX_PATHNAME_LEN);
#endif
bx_param_c *parport_init_list[] = {
bx_options.par1.Oenable,
bx_options.par1.Ooutfile,
//bx_options.par2.Oenable,
//bx_options.par2.Ooutfile,
- add serial port options to the configuration interface, in the same menu as the parallel port options.
2002-03-03 09:10:04 +03:00
bx_options.com1.Opresent,
bx_options.com1.Odev,
//bx_options.com2.Opresent,
//bx_options.com2.Odev,
//bx_options.com3.Opresent,
//bx_options.com3.Odev,
//bx_options.com4.Opresent,
//bx_options.com4.Odev,
- allow configuration of the parallel port in .bochsrc and the config interface menus. Parallel port #1 is implemented, and I left stubs for parallel port #2 in case we want to ever add it. If the parallel port is enabled, the init method of parallel.cc does an fopen() on the output file. If disabled, or if the fopen fails, the file handler remains NULL and no characters are printed. There is no attempt to enable/disable the operation of the parallel port, only the output to a file.
2001-11-12 05:35:09 +03:00
NULL
};
- add serial port options to the configuration interface, in the same menu as the parallel port options.
2002-03-03 09:10:04 +03:00
menu = new bx_list_c (BXP_MENU_SERIAL_PARALLEL, "Serial and Parallel Port Options", "serial_parallel_menu", parport_init_list);
- allow configuration of the parallel port in .bochsrc and the config interface menus. Parallel port #1 is implemented, and I left stubs for parallel port #2 in case we want to ever add it. If the parallel port is enabled, the init method of parallel.cc does an fopen() on the output file. If disabled, or if the fopen fails, the file handler remains NULL and no characters are printed. There is no attempt to enable/disable the operation of the parallel port, only the output to a file.
2001-11-12 05:35:09 +03:00
menu->get_options ()->set (menu->BX_SHOW_PARENT);
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.rom.Opath = new bx_param_string_c (BXP_ROM_PATH,
- now the "memory options" menu is entirely defined using parameter data structures, see bx_init_options in main.cc. The implementation of this menu and all its choices is 17 lines long, see do_mem_options_menu in gui/control.cc.
2001-06-18 18:11:55 +04:00
"romimage",
"Pathname of ROM image to load",
"", BX_PATHNAME_LEN);
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.rom.Opath->set_format ("Name of ROM BIOS image: %s");
bx_options.rom.Oaddress = new bx_param_num_c (BXP_ROM_ADDRESS,
- now the "memory options" menu is entirely defined using parameter data structures, see bx_init_options in main.cc. The implementation of this menu and all its choices is 17 lines long, see do_mem_options_menu in gui/control.cc.
2001-06-18 18:11:55 +04:00
"romaddr",
"The address at which the ROM image should be loaded",
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
0, BX_MAX_INT,
- now the "memory options" menu is entirely defined using parameter data structures, see bx_init_options in main.cc. The implementation of this menu and all its choices is 17 lines long, see do_mem_options_menu in gui/control.cc.
2001-06-18 18:11:55 +04:00
0);
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.rom.Oaddress->set_format ("ROM BIOS address: 0x%05x");
bx_options.rom.Oaddress->set_base (16);
bx_options.vgarom.Opath = new bx_param_string_c (BXP_VGA_ROM_PATH,
- now the "memory options" menu is entirely defined using parameter data structures, see bx_init_options in main.cc. The implementation of this menu and all its choices is 17 lines long, see do_mem_options_menu in gui/control.cc.
2001-06-18 18:11:55 +04:00
"vgaromimage",
"Pathname of VGA ROM image to load",
"", BX_PATHNAME_LEN);
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.vgarom.Opath->set_format ("Name of VGA BIOS image: %s");
- now the memory options menu, interface menu, and disk menu in the control panel are all implemented as bx_list_c, and look much like they did before. - removed many hardcoded UI functions from control.cc, since the much more general "text_ask" methods have replaced them. - add range checking on integer parameter values. This exposed several cases where my initial value for an integer parameter was not in range. - cleaned up behavior of get/set methods. The get/set methods allow the handler to override the value that is returned/set, or perform side effects. - the title parameter of a bx_list_c now defaults to the name. - now bx_param_c fields that used to be private are protected instead - removed references to bx_any - moved definition of set_handler from siminterface.h to siminterface.cc. I was considering with doing a "set" of the old value when the handler is first installed, but that remains commented out. - BX_BOOT_DISKA and BX_BOOT_DISKC are now 0,1 so that they can correspond with the values of a bx_param_enum_c. For a while they were 0x00 and 0x80 corresponding to the numbering convention of the bios, but it didn't really matter.
2001-06-21 18:37:55 +04:00
bx_param_c *memory_init_list[] = {
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.memory.Osize,
bx_options.vgarom.Opath,
bx_options.rom.Opath,
bx_options.rom.Oaddress,
- now the "memory options" menu is entirely defined using parameter data structures, see bx_init_options in main.cc. The implementation of this menu and all its choices is 17 lines long, see do_mem_options_menu in gui/control.cc.
2001-06-18 18:11:55 +04:00
NULL
};
- now the memory options menu, interface menu, and disk menu in the control panel are all implemented as bx_list_c, and look much like they did before. - removed many hardcoded UI functions from control.cc, since the much more general "text_ask" methods have replaced them. - add range checking on integer parameter values. This exposed several cases where my initial value for an integer parameter was not in range. - cleaned up behavior of get/set methods. The get/set methods allow the handler to override the value that is returned/set, or perform side effects. - the title parameter of a bx_list_c now defaults to the name. - now bx_param_c fields that used to be private are protected instead - removed references to bx_any - moved definition of set_handler from siminterface.h to siminterface.cc. I was considering with doing a "set" of the old value when the handler is first installed, but that remains commented out. - BX_BOOT_DISKA and BX_BOOT_DISKC are now 0,1 so that they can correspond with the values of a bx_param_enum_c. For a while they were 0x00 and 0x80 corresponding to the numbering convention of the bios, but it didn't really matter.
2001-06-21 18:37:55 +04:00
menu = new bx_list_c (BXP_MENU_MEMORY, "Bochs Memory Options", "memmenu", memory_init_list);
- now the "memory options" menu is entirely defined using parameter data structures, see bx_init_options in main.cc. The implementation of this menu and all its choices is 17 lines long, see do_mem_options_menu in gui/control.cc.
2001-06-18 18:11:55 +04:00
menu->get_options ()->set (menu->BX_SHOW_PARENT);
- turn the last few options into parameters (cmos, loader, random stuff)
2001-06-21 23:57:21 +04:00
// interface
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.Ovga_update_interval = new bx_param_num_c (BXP_VGA_UPDATE_INTERVAL,
- now the memory options menu, interface menu, and disk menu in the control panel are all implemented as bx_list_c, and look much like they did before. - removed many hardcoded UI functions from control.cc, since the much more general "text_ask" methods have replaced them. - add range checking on integer parameter values. This exposed several cases where my initial value for an integer parameter was not in range. - cleaned up behavior of get/set methods. The get/set methods allow the handler to override the value that is returned/set, or perform side effects. - the title parameter of a bx_list_c now defaults to the name. - now bx_param_c fields that used to be private are protected instead - removed references to bx_any - moved definition of set_handler from siminterface.h to siminterface.cc. I was considering with doing a "set" of the old value when the handler is first installed, but that remains commented out. - BX_BOOT_DISKA and BX_BOOT_DISKC are now 0,1 so that they can correspond with the values of a bx_param_enum_c. For a while they were 0x00 and 0x80 corresponding to the numbering convention of the bios, but it didn't really matter.
2001-06-21 18:37:55 +04:00
"VGA Update Interval",
- now ips, mouse_enabled, vga_update_interval, and rom_path are declared as bx_param_c * types in the bx_options structure. They are initialized in main.cc (bx_init_options) with default values. Access to parameters of this type should always be like this: bx_options.mouse_enabled->get (); bx_options.mouse_enabled->set (newval); Eventually I will be transferring all options to this format.
2001-06-17 03:08:32 +04:00
"Number of microseconds between VGA updates",
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
1, BX_MAX_INT,
- now ips, mouse_enabled, vga_update_interval, and rom_path are declared as bx_param_c * types in the bx_options structure. They are initialized in main.cc (bx_init_options) with default values. Access to parameters of this type should always be like this: bx_options.mouse_enabled->get (); bx_options.mouse_enabled->set (newval); Eventually I will be transferring all options to this format.
2001-06-17 03:08:32 +04:00
30000);
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.Ovga_update_interval->set_handler (bx_param_handler);
- now the memory options menu, interface menu, and disk menu in the control panel are all implemented as bx_list_c, and look much like they did before. - removed many hardcoded UI functions from control.cc, since the much more general "text_ask" methods have replaced them. - add range checking on integer parameter values. This exposed several cases where my initial value for an integer parameter was not in range. - cleaned up behavior of get/set methods. The get/set methods allow the handler to override the value that is returned/set, or perform side effects. - the title parameter of a bx_list_c now defaults to the name. - now bx_param_c fields that used to be private are protected instead - removed references to bx_any - moved definition of set_handler from siminterface.h to siminterface.cc. I was considering with doing a "set" of the old value when the handler is first installed, but that remains commented out. - BX_BOOT_DISKA and BX_BOOT_DISKC are now 0,1 so that they can correspond with the values of a bx_param_enum_c. For a while they were 0x00 and 0x80 corresponding to the numbering convention of the bios, but it didn't really matter.
2001-06-21 18:37:55 +04:00
bx_options.Ovga_update_interval->set_ask_format ("Type a new value for VGA update interval: [%d] ");
bx_options.Omouse_enabled = new bx_param_bool_c (BXP_MOUSE_ENABLED,
"Enable the mouse",
"Controls whether the mouse sends events to bochs",
0);
bx_options.Omouse_enabled->set_handler (bx_param_handler);
bx_options.Oips = new bx_param_num_c (BXP_IPS,
"Emulated instructions per second (IPS)",
"Emulated instructions per second, used to calibrate bochs emulated\ntime with wall clock time.",
1, BX_MAX_INT,
500000);
bx_options.Oprivate_colormap = new bx_param_bool_c (BXP_PRIVATE_COLORMAP,
"Use a private colormap",
"Request that the GUI create and use it's own non-shared colormap. This colormap will be used when in the bochs window. If not enabled, a shared colormap scheme may be used. Not implemented on all GUI's.",
0);
ok, now the amiga target on cvs is complete
2001-08-16 16:35:52 +04:00
#if BX_WITH_AMIGAOS
- add string parameter called screenmode - put #if BX_USE_AMIGAOS...#endif around fullscreen and screenmode, since they only apply to amiga
2001-08-16 06:00:31 +04:00
bx_options.Ofullscreen = new bx_param_bool_c (BXP_FULLSCREEN,
"Use full screen mode",
"When enabled, bochs occupies the whole screen instead of just a window.",
0);
bx_options.Oscreenmode = new bx_param_string_c (BXP_SCREENMODE,
"Screen mode name",
"Screen mode name",
"", BX_PATHNAME_LEN);
bx_options.Oscreenmode->set_handler (bx_param_string_handler);
#endif
- now the memory options menu, interface menu, and disk menu in the control panel are all implemented as bx_list_c, and look much like they did before. - removed many hardcoded UI functions from control.cc, since the much more general "text_ask" methods have replaced them. - add range checking on integer parameter values. This exposed several cases where my initial value for an integer parameter was not in range. - cleaned up behavior of get/set methods. The get/set methods allow the handler to override the value that is returned/set, or perform side effects. - the title parameter of a bx_list_c now defaults to the name. - now bx_param_c fields that used to be private are protected instead - removed references to bx_any - moved definition of set_handler from siminterface.h to siminterface.cc. I was considering with doing a "set" of the old value when the handler is first installed, but that remains commented out. - BX_BOOT_DISKA and BX_BOOT_DISKC are now 0,1 so that they can correspond with the values of a bx_param_enum_c. For a while they were 0x00 and 0x80 corresponding to the numbering convention of the bios, but it didn't really matter.
2001-06-21 18:37:55 +04:00
bx_param_c *interface_init_list[] = {
bx_options.Ovga_update_interval,
bx_options.Omouse_enabled,
bx_options.Oips,
bx_options.Oprivate_colormap,
ok, now the amiga target on cvs is complete
2001-08-16 16:35:52 +04:00
#if BX_WITH_AMIGAOS
- add option called fullscreen, for Nicholai Benalal's amiga port
2001-08-16 00:17:19 +04:00
bx_options.Ofullscreen,
- add string parameter called screenmode - put #if BX_USE_AMIGAOS...#endif around fullscreen and screenmode, since they only apply to amiga
2001-08-16 06:00:31 +04:00
bx_options.Oscreenmode,
#endif
- now the memory options menu, interface menu, and disk menu in the control panel are all implemented as bx_list_c, and look much like they did before. - removed many hardcoded UI functions from control.cc, since the much more general "text_ask" methods have replaced them. - add range checking on integer parameter values. This exposed several cases where my initial value for an integer parameter was not in range. - cleaned up behavior of get/set methods. The get/set methods allow the handler to override the value that is returned/set, or perform side effects. - the title parameter of a bx_list_c now defaults to the name. - now bx_param_c fields that used to be private are protected instead - removed references to bx_any - moved definition of set_handler from siminterface.h to siminterface.cc. I was considering with doing a "set" of the old value when the handler is first installed, but that remains commented out. - BX_BOOT_DISKA and BX_BOOT_DISKC are now 0,1 so that they can correspond with the values of a bx_param_enum_c. For a while they were 0x00 and 0x80 corresponding to the numbering convention of the bios, but it didn't really matter.
2001-06-21 18:37:55 +04:00
NULL
};
menu = new bx_list_c (BXP_MENU_INTERFACE, "Bochs Interface Menu", "intfmenu", interface_init_list);
menu->get_options ()->set (menu->BX_SHOW_PARENT);
- converted ne2k and sb16 to new style parameters and control panel, except for MAC address of the ne2k, which is next.
2001-06-21 22:34:50 +04:00
// NE2K options
bx_options.ne2k.Ovalid = new bx_param_bool_c (BXP_NE2K_VALID,
"NE2K is present",
"to be written",
0);
bx_options.ne2k.Oioaddr = new bx_param_num_c (BXP_NE2K_IOADDR,
"NE2K I/O Address",
"to be written",
0, 0xffff,
0);
bx_options.ne2k.Oioaddr->set_base (16);
bx_options.ne2k.Oirq = new bx_param_num_c (BXP_NE2K_IRQ,
"NE2K Interrupt",
"to be written",
0, 15,
0);
- add options to the string parameter class that allow it to deal with raw hex bytes, for the MAC address.
2001-06-21 23:27:05 +04:00
bx_options.ne2k.Omacaddr = new bx_param_string_c (BXP_NE2K_MACADDR,
"MAC Address",
"to be written",
"", 6);
bx_options.ne2k.Omacaddr->get_options ()->set (bx_options.ne2k.Omacaddr->BX_RAW_BYTES);
bx_options.ne2k.Omacaddr->set_separator (':');
- converted ne2k and sb16 to new style parameters and control panel, except for MAC address of the ne2k, which is next.
2001-06-21 22:34:50 +04:00
bx_options.ne2k.Oethmod = new bx_param_string_c (BXP_NE2K_ETHMOD,
"Ethernet module",
"to be written",
"null", 16);
bx_options.ne2k.Oethdev = new bx_param_string_c (BXP_NE2K_ETHDEV,
"Ethernet device",
"to be written",
"xl0", BX_PATHNAME_LEN);
bx_param_c *ne2k_init_list[] = {
bx_options.ne2k.Ovalid,
bx_options.ne2k.Oioaddr,
bx_options.ne2k.Oirq,
- add options to the string parameter class that allow it to deal with raw hex bytes, for the MAC address.
2001-06-21 23:27:05 +04:00
bx_options.ne2k.Omacaddr,
- converted ne2k and sb16 to new style parameters and control panel, except for MAC address of the ne2k, which is next.
2001-06-21 22:34:50 +04:00
bx_options.ne2k.Oethmod,
bx_options.ne2k.Oethdev,
NULL
};
menu = new bx_list_c (BXP_NE2K, "NE2K Configuration", "", ne2k_init_list);
menu->get_options ()->set (menu->BX_SHOW_PARENT);
bx_options.ne2k.Ovalid->set_handler (bx_param_handler);
bx_options.ne2k.Ovalid->set (0);
// SB16 options
bx_options.sb16.Opresent = new bx_param_bool_c (BXP_SB16_PRESENT,
"SB16 is present",
"to be written",
0);
bx_options.sb16.Omidifile = new bx_param_string_c (BXP_SB16_MIDIFILE,
"Midi file",
"to be written",
"", BX_PATHNAME_LEN);
bx_options.sb16.Owavefile = new bx_param_string_c (BXP_SB16_WAVEFILE,
"Wave file",
"to be written",
"", BX_PATHNAME_LEN);
bx_options.sb16.Ologfile = new bx_param_string_c (BXP_SB16_LOGFILE,
"Log file",
"to be written",
"", BX_PATHNAME_LEN);
bx_options.sb16.Omidimode = new bx_param_num_c (BXP_SB16_MIDIMODE,
"Midi mode",
"to be written",
0, BX_MAX_INT,
0);
bx_options.sb16.Owavemode = new bx_param_num_c (BXP_SB16_WAVEMODE,
"Wave mode",
"to be written",
0, BX_MAX_INT,
0);
bx_options.sb16.Ologlevel = new bx_param_num_c (BXP_SB16_LOGLEVEL,
"Log mode",
"to be written",
0, BX_MAX_INT,
0);
bx_options.sb16.Odmatimer = new bx_param_num_c (BXP_SB16_DMATIMER,
"DMA timer",
"to be written",
0, BX_MAX_INT,
0);
bx_param_c *sb16_init_list[] = {
bx_options.sb16.Opresent,
bx_options.sb16.Omidifile,
bx_options.sb16.Owavefile,
bx_options.sb16.Ologfile,
bx_options.sb16.Omidimode,
bx_options.sb16.Owavemode,
bx_options.sb16.Ologlevel,
bx_options.sb16.Odmatimer,
NULL
};
menu = new bx_list_c (BXP_SB16, "SB16 Configuration", "", sb16_init_list);
menu->get_options ()->set (menu->BX_SHOW_PARENT);
bx_options.sb16.Opresent->set_handler (bx_param_handler);
bx_options.sb16.Opresent->set (0);
- turn the last few options into parameters (cmos, loader, random stuff)
2001-06-21 23:57:21 +04:00
bx_options.log.Ofilename = new bx_param_string_c (BXP_LOG_FILENAME,
"log:filename",
"Pathname of bochs log file",
"-", BX_PATHNAME_LEN);
bx_options.log.Ofilename->set_ask_format ("Enter log filename: [%s] ");
// loader
bx_options.load32bitOSImage.OwhichOS = new bx_param_enum_c (BXP_LOAD32BITOS_WHICH,
"Which operating system?",
"Which OS to boot",
loader_os_names,
Load32bitOSNone,
Load32bitOSNone);
bx_options.load32bitOSImage.Opath = new bx_param_string_c (BXP_LOAD32BITOS_PATH,
"Pathname of OS to load",
NULL,
"", BX_PATHNAME_LEN);
bx_options.load32bitOSImage.Oiolog = new bx_param_string_c (BXP_LOAD32BITOS_IOLOG,
"Pathname of I/O log file",
NULL,
"", BX_PATHNAME_LEN);
bx_options.load32bitOSImage.Oinitrd = new bx_param_string_c (BXP_LOAD32BITOS_INITRD,
"Pathname of initrd",
NULL,
"", BX_PATHNAME_LEN);
bx_param_c *loader_init_list[] = {
bx_options.load32bitOSImage.OwhichOS,
bx_options.load32bitOSImage.Opath,
bx_options.load32bitOSImage.Oiolog,
bx_options.load32bitOSImage.Oinitrd,
NULL
};
bx_options.load32bitOSImage.OwhichOS->set_format ("os=%s");
bx_options.load32bitOSImage.Opath->set_format (", path=%s");
bx_options.load32bitOSImage.Oiolog->set_format (", iolog=%s");
bx_options.load32bitOSImage.Oinitrd->set_format (", initrd=%s");
bx_options.load32bitOSImage.OwhichOS->set_ask_format ("Enter OS to load: [%s] ");
bx_options.load32bitOSImage.Opath->set_ask_format ("Enter pathname of OS: [%s]");
bx_options.load32bitOSImage.Oiolog->set_ask_format ("Enter pathname of I/O log: [%s] ");
bx_options.load32bitOSImage.Oinitrd->set_ask_format ("Enter pathname of initrd: [%s] ");
menu = new bx_list_c (BXP_LOAD32BITOS, "32-bit OS Loader", "", loader_init_list);
menu->get_options ()->set (menu->BX_SERIES_ASK);
bx_options.load32bitOSImage.OwhichOS->set_handler (bx_param_handler);
bx_options.load32bitOSImage.OwhichOS->set (Load32bitOSNone);
// other
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.Okeyboard_serial_delay = new bx_param_num_c (BXP_KBD_SERIAL_DELAY,
"keyboard_serial_delay",
"Approximate time in microseconds that it takes one character to be transfered from the keyboard to controller over the serial path.",
1, BX_MAX_INT,
20000);
- apply Christophe's patch.paste-for-dos. I will make minor tweaks as separate revisions.
2002-03-26 16:51:48 +03:00
bx_options.Okeyboard_paste_delay = new bx_param_num_c (BXP_KBD_PASTE_DELAY,
"keyboard_paste_delay",
"Approximate time in microseconds between attemps to paste characters to the keyboard controller.",
- restore Christophe's minimum paste delay of 1ms. Linux works even if you turn the delay down to 0, but it doesn't actually paste any faster so there's no point. I will try to be satisfied with only 1000 chars per second. :)
2002-03-26 17:28:31 +03:00
1000, BX_MAX_INT,
- apply Christophe's patch.paste-for-dos. I will make minor tweaks as separate revisions.
2002-03-26 16:51:48 +03:00
100000);
- made keyboard_paste_delay a runtime configurable option - clean up comments & debug msgs in paste code
2002-03-26 17:46:03 +03:00
bx_options.Okeyboard_paste_delay->set_handler (bx_param_handler);
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.Ofloppy_command_delay = new bx_param_num_c (BXP_FLOPPY_CMD_DELAY,
"floppy_command_delay",
"Time in microseconds to wait before completing some floppy commands such as read/write/seek/etc, which normally have a delay associated. This used to be hardwired to 50,000 before.",
1, BX_MAX_INT,
50000);
bx_options.Oi440FXSupport = new bx_param_bool_c (BXP_I440FX_SUPPORT,
"i440FXSupport",
"Controls whether to emulate PCI I440FX",
0);
- turn the last few options into parameters (cmos, loader, random stuff)
2001-06-21 23:57:21 +04:00
bx_options.cmos.OcmosImage = new bx_param_bool_c (BXP_CMOS_IMAGE,
"Use a CMOS image",
NULL,
0);
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.cmos.Opath = new bx_param_string_c (BXP_CMOS_PATH,
"Pathname of CMOS image",
- turn the last few options into parameters (cmos, loader, random stuff)
2001-06-21 23:57:21 +04:00
NULL,
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
"", BX_PATHNAME_LEN);
bx_options.cmos.Otime0 = new bx_param_num_c (BXP_CMOS_TIME0,
- turn the last few options into parameters (cmos, loader, random stuff)
2001-06-21 23:57:21 +04:00
"Initial CMOS time for Bochs",
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
"Start time for Bochs CMOS clock, used if you really want two runs to be identical (cosimulation)",
- now the memory options menu, interface menu, and disk menu in the control panel are all implemented as bx_list_c, and look much like they did before. - removed many hardcoded UI functions from control.cc, since the much more general "text_ask" methods have replaced them. - add range checking on integer parameter values. This exposed several cases where my initial value for an integer parameter was not in range. - cleaned up behavior of get/set methods. The get/set methods allow the handler to override the value that is returned/set, or perform side effects. - the title parameter of a bx_list_c now defaults to the name. - now bx_param_c fields that used to be private are protected instead - removed references to bx_any - moved definition of set_handler from siminterface.h to siminterface.cc. I was considering with doing a "set" of the old value when the handler is first installed, but that remains commented out. - BX_BOOT_DISKA and BX_BOOT_DISKC are now 0,1 so that they can correspond with the values of a bx_param_enum_c. For a while they were 0x00 and 0x80 corresponding to the numbering convention of the bios, but it didn't really matter.
2001-06-21 18:37:55 +04:00
0, BX_MAX_INT,
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
0);
. Added a keyboard_type option in configuration file . Modified answer of keyboard controler to a keyboard-identify command depending on keyboard_type . Added the keyboard_type option in bochsrc man page
2001-12-12 13:43:36 +03:00
- Added XFree86 keyboard mapping (us, fr, de)
2001-12-14 20:56:37 +03:00
// Keyboard mapping
bx_options.keyboard.OuseMapping = new bx_param_bool_c(BXP_KEYBOARD_USEMAPPING,
"Use keyboard mapping",
NULL,
0);
bx_options.keyboard.Okeymap = new bx_param_string_c (BXP_KEYBOARD_MAP,
- New keymap architecture. The localisation is now read from keymap files The keymap files are in gui/keymaps directory
2002-03-06 12:31:55 +03:00
"Keymap filename",
- Added XFree86 keyboard mapping (us, fr, de)
2001-12-14 20:56:37 +03:00
NULL,
- New keymap architecture. The localisation is now read from keymap files The keymap files are in gui/keymaps directory
2002-03-06 12:31:55 +03:00
"", BX_PATHNAME_LEN);
- Added XFree86 keyboard mapping (us, fr, de)
2001-12-14 20:56:37 +03:00
// Keyboard type
. Added a keyboard_type option in configuration file . Modified answer of keyboard controler to a keyboard-identify command depending on keyboard_type . Added the keyboard_type option in bochsrc man page
2001-12-12 13:43:36 +03:00
bx_options.Okeyboard_type = new bx_param_enum_c (BXP_KBD_TYPE,
"Keyboard type",
"Keyboard type",
keyboard_type_names,
BX_KBD_MF_TYPE,
BX_KBD_XT_TYPE);
bx_options.Okeyboard_type->set_format ("Keyboard type: %s");
bx_options.Okeyboard_type->set_ask_format ("Enter keyboard type: [%s] ");
- turn the last few options into parameters (cmos, loader, random stuff)
2001-06-21 23:57:21 +04:00
bx_param_c *other_init_list[] = {
- apply Christophe's patch.paste-for-dos. I will make minor tweaks as separate revisions.
2002-03-26 16:51:48 +03:00
bx_options.Okeyboard_serial_delay,
bx_options.Okeyboard_paste_delay,
- turn the last few options into parameters (cmos, loader, random stuff)
2001-06-21 23:57:21 +04:00
bx_options.Ofloppy_command_delay,
bx_options.Oi440FXSupport,
bx_options.cmos.OcmosImage,
bx_options.cmos.Opath,
bx_options.cmos.Otime0,
SIM->get_param (BXP_LOAD32BITOS),
- Added XFree86 keyboard mapping (us, fr, de)
2001-12-14 20:56:37 +03:00
bx_options.keyboard.OuseMapping,
bx_options.keyboard.Okeymap,
. Added a keyboard_type option in configuration file . Modified answer of keyboard controler to a keyboard-identify command depending on keyboard_type . Added the keyboard_type option in bochsrc man page
2001-12-12 13:43:36 +03:00
bx_options.Okeyboard_type,
- turn the last few options into parameters (cmos, loader, random stuff)
2001-06-21 23:57:21 +04:00
NULL
};
menu = new bx_list_c (BXP_MENU_MISC, "Configure Everything Else", "", other_init_list);
menu->get_options ()->set (menu->BX_SHOW_PARENT);
bx_options.cmos.OcmosImage->set_handler (bx_param_handler);
bx_options.cmos.OcmosImage->set (0);
- converted ne2k and sb16 to new style parameters and control panel, except for MAC address of the ne2k, which is next.
2001-06-21 22:34:50 +04:00
- now megs, ips, mouse_enabled, vga_update_interval, and rom_path are using Bryce's experimental parameter structures.
2001-06-16 23:29:59 +04:00
}
split log stuff into a separate file
2001-06-19 20:25:41 +04:00
void bx_print_header(void);
void bx_print_header ()
{
fprintf (stderr, "%s\n", divider);
char buffer[128];
sprintf (buffer, "Bochs x86 Emulator %s\n", VER_STRING);
bx_center_print (stderr, buffer, 72);
if (REL_STRING[0]) {
sprintf (buffer, "%s\n", REL_STRING);
bx_center_print (stderr, buffer, 72);
}
fprintf (stderr, "%s\n", divider);
}
- fixed broken #ifdef...should have been an #if
2001-09-26 07:19:25 +04:00
#if BX_WITH_CARBON
- add code for MacOS X, contributed by emmanuel maillard <e.rsz@libertysurf.fr>
2001-09-26 04:12:12 +04:00
/* Original code by Darrell Walisser - dwaliss1@purdue.edu */
static void setupWorkingDirectory (char *path)
{
char parentdir[MAXPATHLEN];
char *c;
strncpy ( parentdir, path, MAXPATHLEN );
c = (char*) parentdir;
while (*c != '\0') /* go to end */
c++;
while (*c != '/') /* back up to parent */
c--;
*c = '\0'; /* cut off last part (binary name) */
- fixed a minor problem I introduced in macos X specific code
2001-10-09 07:21:48 +04:00
/* chdir to the binary app's parent */
int n;
n = chdir (parentdir);
if (n) BX_PANIC (("failed to change dir to parent"));
/* chdir to the .app's parent */
n = chdir ("../../../");
- on win32 in some configurations, code within an assert is completely removed! I used this trick sometimes to check that a function returned what I expected, like assert (func () == 0), but this caused the func() to never get called. Oops.
2001-10-07 22:16:01 +04:00
if (n) BX_PANIC (("failed to change to ../../.."));
- add code for MacOS X, contributed by emmanuel maillard <e.rsz@libertysurf.fr>
2001-09-26 04:12:12 +04:00
}
#endif
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
int
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
main(int argc, char *argv[])
{
Fixes for bug [ #467530 ] cmd line overrides config interface - rework the order of initialization with and without the control panel. The thing that was bothering me most was the command line options were being processed after the user had set everything in the control panel. This is clearly not what's expected--the command line options should affect the startup defaults of the control panel, but whatever the user chooses in the cpanel menus is the final choice. - if the control panel (config interface) is not wanted, the user can put "-nocp" or "-nocontrolpanel" as the FIRST argument on the command line. Also, the "-psn" option which is automatically passed in by MacOS X when you doubleclick the application causes the control panel to be disabled. In this case, the order of operations is: 1. read bochsrc 2. parse command line options. - if the control panel is enabled (default), the command line options are parsed to provide the startup defaults for the control panel, but the control panel settings are the final answer. So the order is: 1. parse command line options 2. run control panel (if user chooses, he can read bochsrc from menus) - I haven't tested command line options with the debugger yet.
2001-10-07 02:31:31 +04:00
int arg = 1;
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
// To deal with initialization order problems inherent in C++, use the macros
// SAFE_GET_IOFUNC and SAFE_GET_GENLOG to retrieve "io" and "genlog" in all
// constructors or functions called by constructors. The macros test for
// NULL and create the object if necessary, then return it. Ensure that io
// and genlog get created, by making one reference to each macro right here.
// All other code can reference io and genlog directly.
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
SAFE_GET_IOFUNC();
SAFE_GET_GENLOG();
- clean up stderr output a lot - give hints if they type "bochs" with no args and no .bochsrc is found
2001-05-22 22:47:30 +04:00
bx_print_header ();
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
bx_init_bx_dbg ();
- fixed broken #ifdef...should have been an #if
2001-09-26 07:19:25 +04:00
#if BX_WITH_CARBON
Fixes for bug [ #467530 ] cmd line overrides config interface - rework the order of initialization with and without the control panel. The thing that was bothering me most was the command line options were being processed after the user had set everything in the control panel. This is clearly not what's expected--the command line options should affect the startup defaults of the control panel, but whatever the user chooses in the cpanel menus is the final choice. - if the control panel (config interface) is not wanted, the user can put "-nocp" or "-nocontrolpanel" as the FIRST argument on the command line. Also, the "-psn" option which is automatically passed in by MacOS X when you doubleclick the application causes the control panel to be disabled. In this case, the order of operations is: 1. read bochsrc 2. parse command line options. - if the control panel is enabled (default), the command line options are parsed to provide the startup defaults for the control panel, but the control panel settings are the final answer. So the order is: 1. parse command line options 2. run control panel (if user chooses, he can read bochsrc from menus) - I haven't tested command line options with the debugger yet.
2001-10-07 02:31:31 +04:00
/* "-psn" is passed if we are launched by double-clicking */
- attempt to clean up the behavior when you run bochs from the command line or by double clicking. Running from the command line doesn't really work because of a NULL pointer problem in gui/carbon.cc.
2001-09-27 06:54:06 +04:00
if ( argc >= 2 && strncmp (argv[1], "-psn", 4) == 0 ) {
Fixes for bug [ #467530 ] cmd line overrides config interface - rework the order of initialization with and without the control panel. The thing that was bothering me most was the command line options were being processed after the user had set everything in the control panel. This is clearly not what's expected--the command line options should affect the startup defaults of the control panel, but whatever the user chooses in the cpanel menus is the final choice. - if the control panel (config interface) is not wanted, the user can put "-nocp" or "-nocontrolpanel" as the FIRST argument on the command line. Also, the "-psn" option which is automatically passed in by MacOS X when you doubleclick the application causes the control panel to be disabled. In this case, the order of operations is: 1. read bochsrc 2. parse command line options. - if the control panel is enabled (default), the command line options are parsed to provide the startup defaults for the control panel, but the control panel settings are the final answer. So the order is: 1. parse command line options 2. run control panel (if user chooses, he can read bochsrc from menus) - I haven't tested command line options with the debugger yet.
2001-10-07 02:31:31 +04:00
// ugly hack. I don't know how to open a window to print messages in,
// so put them in /tmp/early-bochs-out.txt. Sorry. -bbd
- attempt to clean up the behavior when you run bochs from the command line or by double clicking. Running from the command line doesn't really work because of a NULL pointer problem in gui/carbon.cc.
2001-09-27 06:54:06 +04:00
io->init_log("/tmp/early-bochs-out.txt");
BX_INFO (("I was launched by double clicking. Fixing home directory."));
Fixes for bug [ #467530 ] cmd line overrides config interface - rework the order of initialization with and without the control panel. The thing that was bothering me most was the command line options were being processed after the user had set everything in the control panel. This is clearly not what's expected--the command line options should affect the startup defaults of the control panel, but whatever the user chooses in the cpanel menus is the final choice. - if the control panel (config interface) is not wanted, the user can put "-nocp" or "-nocontrolpanel" as the FIRST argument on the command line. Also, the "-psn" option which is automatically passed in by MacOS X when you doubleclick the application causes the control panel to be disabled. In this case, the order of operations is: 1. read bochsrc 2. parse command line options. - if the control panel is enabled (default), the command line options are parsed to provide the startup defaults for the control panel, but the control panel settings are the final answer. So the order is: 1. parse command line options 2. run control panel (if user chooses, he can read bochsrc from menus) - I haven't tested command line options with the debugger yet.
2001-10-07 02:31:31 +04:00
argc = 1; // ignore all other args.
- attempt to clean up the behavior when you run bochs from the command line or by double clicking. Running from the command line doesn't really work because of a NULL pointer problem in gui/carbon.cc.
2001-09-27 06:54:06 +04:00
setupWorkingDirectory (argv[0]);
Fixes for bug [ #467530 ] cmd line overrides config interface - rework the order of initialization with and without the control panel. The thing that was bothering me most was the command line options were being processed after the user had set everything in the control panel. This is clearly not what's expected--the command line options should affect the startup defaults of the control panel, but whatever the user chooses in the cpanel menus is the final choice. - if the control panel (config interface) is not wanted, the user can put "-nocp" or "-nocontrolpanel" as the FIRST argument on the command line. Also, the "-psn" option which is automatically passed in by MacOS X when you doubleclick the application causes the control panel to be disabled. In this case, the order of operations is: 1. read bochsrc 2. parse command line options. - if the control panel is enabled (default), the command line options are parsed to provide the startup defaults for the control panel, but the control panel settings are the final answer. So the order is: 1. parse command line options 2. run control panel (if user chooses, he can read bochsrc from menus) - I haven't tested command line options with the debugger yet.
2001-10-07 02:31:31 +04:00
// there is no stdin/stdout so disable the text-based config interface.
enable_control_panel = 0;
- attempt to clean up the behavior when you run bochs from the command line or by double clicking. Running from the command line doesn't really work because of a NULL pointer problem in gui/carbon.cc.
2001-09-27 06:54:06 +04:00
}
Fixes for bug [ #467530 ] cmd line overrides config interface - rework the order of initialization with and without the control panel. The thing that was bothering me most was the command line options were being processed after the user had set everything in the control panel. This is clearly not what's expected--the command line options should affect the startup defaults of the control panel, but whatever the user chooses in the cpanel menus is the final choice. - if the control panel (config interface) is not wanted, the user can put "-nocp" or "-nocontrolpanel" as the FIRST argument on the command line. Also, the "-psn" option which is automatically passed in by MacOS X when you doubleclick the application causes the control panel to be disabled. In this case, the order of operations is: 1. read bochsrc 2. parse command line options. - if the control panel is enabled (default), the command line options are parsed to provide the startup defaults for the control panel, but the control panel settings are the final answer. So the order is: 1. parse command line options 2. run control panel (if user chooses, he can read bochsrc from menus) - I haven't tested command line options with the debugger yet.
2001-10-07 02:31:31 +04:00
// if it was started from command line, there could be some args still.
- attempt to clean up the behavior when you run bochs from the command line or by double clicking. Running from the command line doesn't really work because of a NULL pointer problem in gui/carbon.cc.
2001-09-27 06:54:06 +04:00
for (int a=0; a<argc; a++) {
BX_INFO (("argument %d is %s", a, argv[a]));
}
- add code for MacOS X, contributed by emmanuel maillard <e.rsz@libertysurf.fr>
2001-09-26 04:12:12 +04:00
- attempt to clean up the behavior when you run bochs from the command line or by double clicking. Running from the command line doesn't really work because of a NULL pointer problem in gui/carbon.cc.
2001-09-27 06:54:06 +04:00
char cwd[MAXPATHLEN];
getwd (cwd);
BX_INFO (("Now my working directory is %s", cwd));
- add code for MacOS X, contributed by emmanuel maillard <e.rsz@libertysurf.fr>
2001-09-26 04:12:12 +04:00
#endif
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
init_siminterface ();
Fixes for bug [ #467530 ] cmd line overrides config interface - rework the order of initialization with and without the control panel. The thing that was bothering me most was the command line options were being processed after the user had set everything in the control panel. This is clearly not what's expected--the command line options should affect the startup defaults of the control panel, but whatever the user chooses in the cpanel menus is the final choice. - if the control panel (config interface) is not wanted, the user can put "-nocp" or "-nocontrolpanel" as the FIRST argument on the command line. Also, the "-psn" option which is automatically passed in by MacOS X when you doubleclick the application causes the control panel to be disabled. In this case, the order of operations is: 1. read bochsrc 2. parse command line options. - if the control panel is enabled (default), the command line options are parsed to provide the startup defaults for the control panel, but the control panel settings are the final answer. So the order is: 1. parse command line options 2. run control panel (if user chooses, he can read bochsrc from menus) - I haven't tested command line options with the debugger yet.
2001-10-07 02:31:31 +04:00
bx_control_panel (BX_CPANEL_INIT);
- now the "memory options" menu is entirely defined using parameter data structures, see bx_init_options in main.cc. The implementation of this menu and all its choices is 17 lines long, see do_mem_options_menu in gui/control.cc.
2001-06-18 18:11:55 +04:00
bx_init_options ();
- add "-nocontrolpanel" command line option.
2001-06-22 01:24:05 +04:00
Fixes for bug [ #467530 ] cmd line overrides config interface - rework the order of initialization with and without the control panel. The thing that was bothering me most was the command line options were being processed after the user had set everything in the control panel. This is clearly not what's expected--the command line options should affect the startup defaults of the control panel, but whatever the user chooses in the cpanel menus is the final choice. - if the control panel (config interface) is not wanted, the user can put "-nocp" or "-nocontrolpanel" as the FIRST argument on the command line. Also, the "-psn" option which is automatically passed in by MacOS X when you doubleclick the application causes the control panel to be disabled. In this case, the order of operations is: 1. read bochsrc 2. parse command line options. - if the control panel is enabled (default), the command line options are parsed to provide the startup defaults for the control panel, but the control panel settings are the final answer. So the order is: 1. parse command line options 2. run control panel (if user chooses, he can read bochsrc from menus) - I haven't tested command line options with the debugger yet.
2001-10-07 02:31:31 +04:00
// detect -nocontrolpanel or -nocp argument before anything else
if (argc > 1 &&
((!strcmp ("-nocontrolpanel", argv[1]))
|| (!strcmp ("-nocp", argv[1])))) {
- add "-nocontrolpanel" command line option.
2001-06-22 01:24:05 +04:00
// skip the control panel
Fixes for bug [ #467530 ] cmd line overrides config interface - rework the order of initialization with and without the control panel. The thing that was bothering me most was the command line options were being processed after the user had set everything in the control panel. This is clearly not what's expected--the command line options should affect the startup defaults of the control panel, but whatever the user chooses in the cpanel menus is the final choice. - if the control panel (config interface) is not wanted, the user can put "-nocp" or "-nocontrolpanel" as the FIRST argument on the command line. Also, the "-psn" option which is automatically passed in by MacOS X when you doubleclick the application causes the control panel to be disabled. In this case, the order of operations is: 1. read bochsrc 2. parse command line options. - if the control panel is enabled (default), the command line options are parsed to provide the startup defaults for the control panel, but the control panel settings are the final answer. So the order is: 1. parse command line options 2. run control panel (if user chooses, he can read bochsrc from menus) - I haven't tested command line options with the debugger yet.
2001-10-07 02:31:31 +04:00
arg++;
- add "-nocontrolpanel" command line option.
2001-06-22 01:24:05 +04:00
SIM->set_enabled (0);
Fixes for bug [ #467530 ] cmd line overrides config interface - rework the order of initialization with and without the control panel. The thing that was bothering me most was the command line options were being processed after the user had set everything in the control panel. This is clearly not what's expected--the command line options should affect the startup defaults of the control panel, but whatever the user chooses in the cpanel menus is the final choice. - if the control panel (config interface) is not wanted, the user can put "-nocp" or "-nocontrolpanel" as the FIRST argument on the command line. Also, the "-psn" option which is automatically passed in by MacOS X when you doubleclick the application causes the control panel to be disabled. In this case, the order of operations is: 1. read bochsrc 2. parse command line options. - if the control panel is enabled (default), the command line options are parsed to provide the startup defaults for the control panel, but the control panel settings are the final answer. So the order is: 1. parse command line options 2. run control panel (if user chooses, he can read bochsrc from menus) - I haven't tested command line options with the debugger yet.
2001-10-07 02:31:31 +04:00
enable_control_panel = 0;
- add "-nocontrolpanel" command line option.
2001-06-22 01:24:05 +04:00
}
Fixes for bug [ #467530 ] cmd line overrides config interface - rework the order of initialization with and without the control panel. The thing that was bothering me most was the command line options were being processed after the user had set everything in the control panel. This is clearly not what's expected--the command line options should affect the startup defaults of the control panel, but whatever the user chooses in the cpanel menus is the final choice. - if the control panel (config interface) is not wanted, the user can put "-nocp" or "-nocontrolpanel" as the FIRST argument on the command line. Also, the "-psn" option which is automatically passed in by MacOS X when you doubleclick the application causes the control panel to be disabled. In this case, the order of operations is: 1. read bochsrc 2. parse command line options. - if the control panel is enabled (default), the command line options are parsed to provide the startup defaults for the control panel, but the control panel settings are the final answer. So the order is: 1. parse command line options 2. run control panel (if user chooses, he can read bochsrc from menus) - I haven't tested command line options with the debugger yet.
2001-10-07 02:31:31 +04:00
#if !BX_USE_CONTROL_PANEL
enable_control_panel = 0;
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
#endif
- add "-nocontrolpanel" command line option.
2001-06-22 01:24:05 +04:00
Fixes for bug [ #467530 ] cmd line overrides config interface - rework the order of initialization with and without the control panel. The thing that was bothering me most was the command line options were being processed after the user had set everything in the control panel. This is clearly not what's expected--the command line options should affect the startup defaults of the control panel, but whatever the user chooses in the cpanel menus is the final choice. - if the control panel (config interface) is not wanted, the user can put "-nocp" or "-nocontrolpanel" as the FIRST argument on the command line. Also, the "-psn" option which is automatically passed in by MacOS X when you doubleclick the application causes the control panel to be disabled. In this case, the order of operations is: 1. read bochsrc 2. parse command line options. - if the control panel is enabled (default), the command line options are parsed to provide the startup defaults for the control panel, but the control panel settings are the final answer. So the order is: 1. parse command line options 2. run control panel (if user chooses, he can read bochsrc from menus) - I haven't tested command line options with the debugger yet.
2001-10-07 02:31:31 +04:00
if (!enable_control_panel) {
// if we don't intend to run the control panel, parse the .bochsrc now.
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
char *bochsrc = bx_find_bochsrc ();
if (bochsrc)
- compile with debugger has failed because it continued to call bx_bochs_init when main.cc no longer had one. Now compiling with debugger is working with the control panel. To get the control panel, you have to click the snapshot button, and to get the debugger, you have to press ^C. These should be better integrated (maybe a control panel menu choice that jumps into the debugger and a debugger command that starts the runtime control panel...)
2001-06-13 17:36:12 +04:00
bx_read_configuration (bochsrc);
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
Fixes for bug [ #467530 ] cmd line overrides config interface - rework the order of initialization with and without the control panel. The thing that was bothering me most was the command line options were being processed after the user had set everything in the control panel. This is clearly not what's expected--the command line options should affect the startup defaults of the control panel, but whatever the user chooses in the cpanel menus is the final choice. - if the control panel (config interface) is not wanted, the user can put "-nocp" or "-nocontrolpanel" as the FIRST argument on the command line. Also, the "-psn" option which is automatically passed in by MacOS X when you doubleclick the application causes the control panel to be disabled. In this case, the order of operations is: 1. read bochsrc 2. parse command line options. - if the control panel is enabled (default), the command line options are parsed to provide the startup defaults for the control panel, but the control panel settings are the final answer. So the order is: 1. parse command line options 2. run control panel (if user chooses, he can read bochsrc from menus) - I haven't tested command line options with the debugger yet.
2001-10-07 02:31:31 +04:00
if (bochsrc == NULL && arg>=argc) {
// no bochsrc used. This is legal since they may have everything on the
// command line. However if they have no arguments then give them some
// friendly advice.
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
fprintf (stderr, "%s\n", divider);
fprintf (stderr, "Before running Bochs, you should cd to a directory which contains\n");
fprintf (stderr, "a .bochsrc file and a disk image. If you downloaded a binary package,\n");
fprintf (stderr, "all the necessary files are already on your disk.\n");
#if defined(WIN32)
fprintf (stderr, "\nFor Windows installations, go to the dlxlinux direectory and\n");
fprintf (stderr, "double-click on the start.bat script.\n");
#elif !defined(macintosh)
fprintf (stderr, "\nFor UNIX installations, try running \"bochs-dlx\" for a demo. This script\n");
fprintf (stderr, "is basically equivalent to typing:\n");
fprintf (stderr, " cd /usr/local/bochs/dlxlinux\n");
fprintf (stderr, " bochs\n");
#endif
exit(1);
}
}
- clean up stderr output a lot - give hints if they type "bochs" with no args and no .bochsrc is found
2001-05-22 22:47:30 +04:00
Fixes for bug [ #467530 ] cmd line overrides config interface - rework the order of initialization with and without the control panel. The thing that was bothering me most was the command line options were being processed after the user had set everything in the control panel. This is clearly not what's expected--the command line options should affect the startup defaults of the control panel, but whatever the user chooses in the cpanel menus is the final choice. - if the control panel (config interface) is not wanted, the user can put "-nocp" or "-nocontrolpanel" as the FIRST argument on the command line. Also, the "-psn" option which is automatically passed in by MacOS X when you doubleclick the application causes the control panel to be disabled. In this case, the order of operations is: 1. read bochsrc 2. parse command line options. - if the control panel is enabled (default), the command line options are parsed to provide the startup defaults for the control panel, but the control panel settings are the final answer. So the order is: 1. parse command line options 2. run control panel (if user chooses, he can read bochsrc from menus) - I haven't tested command line options with the debugger yet.
2001-10-07 02:31:31 +04:00
// parse the rest of the command line.
if (bx_parse_cmdline (arg, argc, argv)) {
fprintf (stderr, "There were errors while parsing the command line.\n");
fprintf (stderr, "Bochs is exiting.\n");
exit (1);
}
if (enable_control_panel) {
- update log actions after reading command line or bochsrc
2001-12-29 14:40:37 +03:00
// update log actions before starting control panel
for (int level=0; level<N_LOGLEV; level++) {
int action = bx_options.log.actions[level];
io->set_log_action (level, action);
}
Fixes for bug [ #467530 ] cmd line overrides config interface - rework the order of initialization with and without the control panel. The thing that was bothering me most was the command line options were being processed after the user had set everything in the control panel. This is clearly not what's expected--the command line options should affect the startup defaults of the control panel, but whatever the user chooses in the cpanel menus is the final choice. - if the control panel (config interface) is not wanted, the user can put "-nocp" or "-nocontrolpanel" as the FIRST argument on the command line. Also, the "-psn" option which is automatically passed in by MacOS X when you doubleclick the application causes the control panel to be disabled. In this case, the order of operations is: 1. read bochsrc 2. parse command line options. - if the control panel is enabled (default), the command line options are parsed to provide the startup defaults for the control panel, but the control panel settings are the final answer. So the order is: 1. parse command line options 2. run control panel (if user chooses, he can read bochsrc from menus) - I haven't tested command line options with the debugger yet.
2001-10-07 02:31:31 +04:00
// Display the pre-simulation control panel.
SIM->set_enabled (1);
- update log actions after reading command line or bochsrc
2001-12-29 14:40:37 +03:00
bx_control_panel (BX_CPANEL_START_MENU);
Fixes for bug [ #467530 ] cmd line overrides config interface - rework the order of initialization with and without the control panel. The thing that was bothering me most was the command line options were being processed after the user had set everything in the control panel. This is clearly not what's expected--the command line options should affect the startup defaults of the control panel, but whatever the user chooses in the cpanel menus is the final choice. - if the control panel (config interface) is not wanted, the user can put "-nocp" or "-nocontrolpanel" as the FIRST argument on the command line. Also, the "-psn" option which is automatically passed in by MacOS X when you doubleclick the application causes the control panel to be disabled. In this case, the order of operations is: 1. read bochsrc 2. parse command line options. - if the control panel is enabled (default), the command line options are parsed to provide the startup defaults for the control panel, but the control panel settings are the final answer. So the order is: 1. parse command line options 2. run control panel (if user chooses, he can read bochsrc from menus) - I haven't tested command line options with the debugger yet.
2001-10-07 02:31:31 +04:00
}
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
#if BX_DEBUGGER
// If using the debugger, it will take control and call
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
// bx_init_hardware() and cpu_loop()
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
bx_dbg_main(argc, argv);
#else
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
bx_init_hardware();
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
if (bx_options.load32bitOSImage.OwhichOS->get ()) {
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
void bx_load32bitOSimagehack(void);
bx_load32bitOSimagehack();
}
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
SIM->set_init_done (1);
- added a new method to all GUIs called mouse_enabled_changed_specific, which notifies them that the mouse_enabled bit has changed. Now that mouse_enabled can be initialized or modified by parameter events in addition to GUI events, the guis must be prepared for it. I have pasted empty method definitions into mouse_enabled_changed_specific for all guis except for X11, which I did the right way. The implementation of this function must use the argument "val" rather than reading the parameter.
2001-06-23 07:23:19 +04:00
// The set handler for mouse_enabled does not actually update the gui
// until init_done is set. This forces the set handler to be called,
// which sets up the mouse enabled GUI-specific stuff correctly.
// Not a great solution but it works. BBD
bx_options.Omouse_enabled->set (bx_options.Omouse_enabled->get ());
- the SMP merge has reduced performance of even one processor, so this is the first attempt to regain the performance of pre-SMP bochs (1.1.2). When simulating only one processor, stay in cpu_loop forever as pre-SMP versions did. The overhead of returning from cpu_loop over and over was slowing us down.
2001-06-05 19:56:19 +04:00
if (BX_SMP_PROCESSORS == 1) {
// only one processor, run as fast as possible by not messing with
// quantums and loops.
BX_CPU(0)->cpu_loop(1);
- compile with debugger has failed because it continued to call bx_bochs_init when main.cc no longer had one. Now compiling with debugger is working with the control panel. To get the control panel, you have to click the snapshot button, and to get the debugger, you have to press ^C. These should be better integrated (maybe a control panel menu choice that jumps into the debugger and a debugger command that starts the runtime control panel...)
2001-06-13 17:36:12 +04:00
BX_PANIC (("cpu_loop should never return in a single-processor simulation"));
- the SMP merge has reduced performance of even one processor, so this is the first attempt to regain the performance of pre-SMP bochs (1.1.2). When simulating only one processor, stay in cpu_loop forever as pre-SMP versions did. The overhead of returning from cpu_loop over and over was slowing us down.
2001-06-05 19:56:19 +04:00
} else {
- compile with debugger has failed because it continued to call bx_bochs_init when main.cc no longer had one. Now compiling with debugger is working with the control panel. To get the control panel, you have to click the snapshot button, and to get the debugger, you have to press ^C. These should be better integrated (maybe a control panel menu choice that jumps into the debugger and a debugger command that starts the runtime control panel...)
2001-06-13 17:36:12 +04:00
// SMP simulation: do a few instructions on each processor, then switch
// to another. Increasing quantum speeds up overall performance, but
// reduces granularity of synchronization between processors.
- the SMP merge has reduced performance of even one processor, so this is the first attempt to regain the performance of pre-SMP bochs (1.1.2). When simulating only one processor, stay in cpu_loop forever as pre-SMP versions did. The overhead of returning from cpu_loop over and over was slowing us down.
2001-06-05 19:56:19 +04:00
int processor = 0;
int quantum = 5;
while (1) {
// do some instructions in each processor
BX_CPU(processor)->cpu_loop(quantum);
processor = (processor+1) % BX_SMP_PROCESSORS;
if (processor == 0)
BX_TICKN(quantum);
}
- merged BRANCH-smp-bochs into main branch. For details see comments in BRANCH-smp-bochs revisions. - The general task was to make multiple CPU's which communicate through their APICs. So instead of BX_CPU and BX_MEM, we now have BX_CPU(x) and BX_MEM(y). For an SMP simulation you have several processors in a shared memory space, so there might be processors BX_CPU(0..3) but only one memory space BX_MEM(0). For cosimulation, you could have BX_CPU(0) with BX_MEM(0), then BX_CPU(1) with BX_MEM(1). WARNING: Cosimulation is almost certainly broken by the SMP changes. - to simulate multiple CPUs, you have to give each CPU time to execute in turn. This is currently implemented using debugger guards. The cpu loop steps one CPU for a few instructions, then steps the next CPU for a few instructions, etc. - there is some limited support in the debugger for two CPUs, for example printing information from each CPU when single stepping.
2001-05-23 12:16:07 +04:00
}
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
#endif
return(0);
}
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
int
- compile with debugger has failed because it continued to call bx_bochs_init when main.cc no longer had one. Now compiling with debugger is working with the control panel. To get the control panel, you have to click the snapshot button, and to get the debugger, you have to press ^C. These should be better integrated (maybe a control panel menu choice that jumps into the debugger and a debugger command that starts the runtime control panel...)
2001-06-13 17:36:12 +04:00
bx_read_configuration (char *rcfile)
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
{
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
// parse rcfile first, then parse arguments in order.
BX_INFO (("reading configuration from %s", rcfile));
if (parse_bochsrc(rcfile) < 0) {
BX_ERROR (("reading from %s failed", rcfile));
return -1;
}
- update log actions after reading command line or bochsrc
2001-12-29 14:40:37 +03:00
// update log actions if control panel is enabled
if (SIM->get_enabled ()) {
for (int level=0; level<N_LOGLEV; level++) {
int action = bx_options.log.actions[level];
io->set_log_action (level, action);
}
}
- compile with debugger has failed because it continued to call bx_bochs_init when main.cc no longer had one. Now compiling with debugger is working with the control panel. To get the control panel, you have to click the snapshot button, and to get the debugger, you have to press ^C. These should be better integrated (maybe a control panel menu choice that jumps into the debugger and a debugger command that starts the runtime control panel...)
2001-06-13 17:36:12 +04:00
return 0;
}
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
Fixes for bug [ #467530 ] cmd line overrides config interface - rework the order of initialization with and without the control panel. The thing that was bothering me most was the command line options were being processed after the user had set everything in the control panel. This is clearly not what's expected--the command line options should affect the startup defaults of the control panel, but whatever the user chooses in the cpanel menus is the final choice. - if the control panel (config interface) is not wanted, the user can put "-nocp" or "-nocontrolpanel" as the FIRST argument on the command line. Also, the "-psn" option which is automatically passed in by MacOS X when you doubleclick the application causes the control panel to be disabled. In this case, the order of operations is: 1. read bochsrc 2. parse command line options. - if the control panel is enabled (default), the command line options are parsed to provide the startup defaults for the control panel, but the control panel settings are the final answer. So the order is: 1. parse command line options 2. run control panel (if user chooses, he can read bochsrc from menus) - I haven't tested command line options with the debugger yet.
2001-10-07 02:31:31 +04:00
int bx_parse_cmdline (int arg, int argc, char *argv[])
- compile with debugger has failed because it continued to call bx_bochs_init when main.cc no longer had one. Now compiling with debugger is working with the control panel. To get the control panel, you have to click the snapshot button, and to get the debugger, you have to press ^C. These should be better integrated (maybe a control panel menu choice that jumps into the debugger and a debugger command that starts the runtime control panel...)
2001-06-13 17:36:12 +04:00
{
Fixes for bug [ #467530 ] cmd line overrides config interface - rework the order of initialization with and without the control panel. The thing that was bothering me most was the command line options were being processed after the user had set everything in the control panel. This is clearly not what's expected--the command line options should affect the startup defaults of the control panel, but whatever the user chooses in the cpanel menus is the final choice. - if the control panel (config interface) is not wanted, the user can put "-nocp" or "-nocontrolpanel" as the FIRST argument on the command line. Also, the "-psn" option which is automatically passed in by MacOS X when you doubleclick the application causes the control panel to be disabled. In this case, the order of operations is: 1. read bochsrc 2. parse command line options. - if the control panel is enabled (default), the command line options are parsed to provide the startup defaults for the control panel, but the control panel settings are the final answer. So the order is: 1. parse command line options 2. run control panel (if user chooses, he can read bochsrc from menus) - I haven't tested command line options with the debugger yet.
2001-10-07 02:31:31 +04:00
//if (arg < argc) BX_INFO (("parsing command line arguments"));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
Fixes for bug [ #467530 ] cmd line overrides config interface - rework the order of initialization with and without the control panel. The thing that was bothering me most was the command line options were being processed after the user had set everything in the control panel. This is clearly not what's expected--the command line options should affect the startup defaults of the control panel, but whatever the user chooses in the cpanel menus is the final choice. - if the control panel (config interface) is not wanted, the user can put "-nocp" or "-nocontrolpanel" as the FIRST argument on the command line. Also, the "-psn" option which is automatically passed in by MacOS X when you doubleclick the application causes the control panel to be disabled. In this case, the order of operations is: 1. read bochsrc 2. parse command line options. - if the control panel is enabled (default), the command line options are parsed to provide the startup defaults for the control panel, but the control panel settings are the final answer. So the order is: 1. parse command line options 2. run control panel (if user chooses, he can read bochsrc from menus) - I haven't tested command line options with the debugger yet.
2001-10-07 02:31:31 +04:00
while (arg < argc) {
BX_INFO (("parsing arg %d, %s", arg, argv[arg]));
parse_line_unformatted("cmdline args", argv[arg]);
arg++;
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
}
return 0;
}
int
bx_init_hardware()
{
// all configuration has been read, now initialize everything.
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
- bx_init_hardware(): don't update the log actions if control panel is enabled
2002-01-02 12:57:58 +03:00
if (!SIM->get_enabled ()) {
for (int level=0; level<N_LOGLEV; level++) {
int action = bx_options.log.actions[level];
if (action == ACT_ASK) action = ACT_FATAL;
io->set_log_action (level, action);
}
- add "-nocontrolpanel" command line option.
2001-06-22 01:24:05 +04:00
}
- change name of action "crash" to "fatal" - change Makefile.in to "cvs-snapshot" information rather than printing the release number, since in fact it's not the same as the release anymore - use #defines to give the actions names (ACT_IGNORE, ACT_REPORT, ACT_FATAL) - register all logfunctions as they call setio, and keep a list so that we can change their settings more easily. I admit I used a static-sized array to store them. Sorry. - instead of printing a prefix [GEN ] for generic messages, just leave the prefix blank [ ]. Otherwise you start wondering what does gen stand for. - avoid reentry into fatal, since fatal can call BX_INFO and friends. This avoids a potential infinite recursion. - now that files other than .bochsrc can be the rc file, print the actual name of the file in error messages.
2001-05-23 06:37:52 +04:00
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_pc_system.init_ips(bx_options.Oips->get ());
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
if(bx_options.log.Ofilename->getptr()[0]!='-') {
BX_INFO (("using log file %s", bx_options.log.Ofilename->getptr ()));
io->init_log(bx_options.log.Ofilename->getptr ());
- more stderr cleanup
2001-05-22 23:07:53 +04:00
}
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
- merged BRANCH-smp-bochs into main branch. For details see comments in BRANCH-smp-bochs revisions. - The general task was to make multiple CPU's which communicate through their APICs. So instead of BX_CPU and BX_MEM, we now have BX_CPU(x) and BX_MEM(y). For an SMP simulation you have several processors in a shared memory space, so there might be processors BX_CPU(0..3) but only one memory space BX_MEM(0). For cosimulation, you could have BX_CPU(0) with BX_MEM(0), then BX_CPU(1) with BX_MEM(1). WARNING: Cosimulation is almost certainly broken by the SMP changes. - to simulate multiple CPUs, you have to give each CPU time to execute in turn. This is currently implemented using debugger guards. The cpu loop steps one CPU for a few instructions, then steps the next CPU for a few instructions, etc. - there is some limited support in the debugger for two CPUs, for example printing information from each CPU when single stepping.
2001-05-23 12:16:07 +04:00
// set up memory and CPU objects
- there were cases where BX_APIC_SUPPORT were used and others where BX_SUPPORT_APIC were used. To follow the pattern used by other names like this, I changed them all to BX_SUPPORT_APIC. Thanks to Tom Lindström for chasing this down!
2001-06-12 17:07:43 +04:00
#if BX_SUPPORT_APIC
- merged BRANCH-smp-bochs into main branch. For details see comments in BRANCH-smp-bochs revisions. - The general task was to make multiple CPU's which communicate through their APICs. So instead of BX_CPU and BX_MEM, we now have BX_CPU(x) and BX_MEM(y). For an SMP simulation you have several processors in a shared memory space, so there might be processors BX_CPU(0..3) but only one memory space BX_MEM(0). For cosimulation, you could have BX_CPU(0) with BX_MEM(0), then BX_CPU(1) with BX_MEM(1). WARNING: Cosimulation is almost certainly broken by the SMP changes. - to simulate multiple CPUs, you have to give each CPU time to execute in turn. This is currently implemented using debugger guards. The cpu loop steps one CPU for a few instructions, then steps the next CPU for a few instructions, etc. - there is some limited support in the debugger for two CPUs, for example printing information from each CPU when single stepping.
2001-05-23 12:16:07 +04:00
memset(apic_index, 0, sizeof(apic_index[0]) * APIC_MAX_ID);
#endif
- another speed boost: when not using SMP, use BX_CPU_C bx_cpu; BX_MEM_C bx_mem; and when more than one processor, use BX_CPU_C *bx_cpu_array[BX_SMP_PROCESSORS]; BX_MEM_C *bx_mem_array[BX_ADDRESS_SPACES]; The changeover is controlled by BX_SMP_PROCESSORS, but there are only a few code changes since nearly all code uses the BX_CPU(n) and BX_MEM(n) macros. - This turns out to make a 10% speed difference! With this revision, the CVS version now gets 95% of the performance of the 3/25/2000 snapshot, which I've been using as my baseline.
2001-06-05 21:35:08 +04:00
#if BX_SMP_PROCESSORS==1
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
BX_MEM(0)->init_memory(bx_options.memory.Osize->get () * 1024*1024);
BX_MEM(0)->load_ROM(bx_options.rom.Opath->getptr (), bx_options.rom.Oaddress->get ());
BX_MEM(0)->load_ROM(bx_options.vgarom.Opath->getptr (), 0xc0000);
- another speed boost: when not using SMP, use BX_CPU_C bx_cpu; BX_MEM_C bx_mem; and when more than one processor, use BX_CPU_C *bx_cpu_array[BX_SMP_PROCESSORS]; BX_MEM_C *bx_mem_array[BX_ADDRESS_SPACES]; The changeover is controlled by BX_SMP_PROCESSORS, but there are only a few code changes since nearly all code uses the BX_CPU(n) and BX_MEM(n) macros. - This turns out to make a 10% speed difference! With this revision, the CVS version now gets 95% of the performance of the 3/25/2000 snapshot, which I've been using as my baseline.
2001-06-05 21:35:08 +04:00
BX_CPU(0)->init (BX_MEM(0));
- fix segfault that happened if APIC was enabled with one processor
2002-03-27 19:05:13 +03:00
#if BX_SUPPORT_APIC
applied a temporary fix so that it can compile
2002-03-30 02:37:07 +03:00
BX_CPU(0)->local_apic.set_id (0);
- fix segfault that happened if APIC was enabled with one processor
2002-03-27 19:05:13 +03:00
#endif
- another speed boost: when not using SMP, use BX_CPU_C bx_cpu; BX_MEM_C bx_mem; and when more than one processor, use BX_CPU_C *bx_cpu_array[BX_SMP_PROCESSORS]; BX_MEM_C *bx_mem_array[BX_ADDRESS_SPACES]; The changeover is controlled by BX_SMP_PROCESSORS, but there are only a few code changes since nearly all code uses the BX_CPU(n) and BX_MEM(n) macros. - This turns out to make a 10% speed difference! With this revision, the CVS version now gets 95% of the performance of the 3/25/2000 snapshot, which I've been using as my baseline.
2001-06-05 21:35:08 +04:00
BX_CPU(0)->reset(BX_RESET_HARDWARE);
#else
// SMP initialization
- merged BRANCH-smp-bochs into main branch. For details see comments in BRANCH-smp-bochs revisions. - The general task was to make multiple CPU's which communicate through their APICs. So instead of BX_CPU and BX_MEM, we now have BX_CPU(x) and BX_MEM(y). For an SMP simulation you have several processors in a shared memory space, so there might be processors BX_CPU(0..3) but only one memory space BX_MEM(0). For cosimulation, you could have BX_CPU(0) with BX_MEM(0), then BX_CPU(1) with BX_MEM(1). WARNING: Cosimulation is almost certainly broken by the SMP changes. - to simulate multiple CPUs, you have to give each CPU time to execute in turn. This is currently implemented using debugger guards. The cpu loop steps one CPU for a few instructions, then steps the next CPU for a few instructions, etc. - there is some limited support in the debugger for two CPUs, for example printing information from each CPU when single stepping.
2001-05-23 12:16:07 +04:00
bx_mem_array[0] = new BX_MEM_C ();
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_mem_array[0]->init_memory(bx_options.memory.Osize->get () * 1024*1024);
allow to compile with SMP
2001-11-06 18:41:34 +03:00
bx_mem_array[0]->load_ROM(bx_options.rom.Opath->getptr (), bx_options.rom.Oaddress->get ());
bx_mem_array[0]->load_ROM(bx_options.vgarom.Opath->getptr (), 0xc0000);
- merged BRANCH-smp-bochs into main branch. For details see comments in BRANCH-smp-bochs revisions. - The general task was to make multiple CPU's which communicate through their APICs. So instead of BX_CPU and BX_MEM, we now have BX_CPU(x) and BX_MEM(y). For an SMP simulation you have several processors in a shared memory space, so there might be processors BX_CPU(0..3) but only one memory space BX_MEM(0). For cosimulation, you could have BX_CPU(0) with BX_MEM(0), then BX_CPU(1) with BX_MEM(1). WARNING: Cosimulation is almost certainly broken by the SMP changes. - to simulate multiple CPUs, you have to give each CPU time to execute in turn. This is currently implemented using debugger guards. The cpu loop steps one CPU for a few instructions, then steps the next CPU for a few instructions, etc. - there is some limited support in the debugger for two CPUs, for example printing information from each CPU when single stepping.
2001-05-23 12:16:07 +04:00
for (int i=0; i<BX_SMP_PROCESSORS; i++) {
BX_CPU(i) = new BX_CPU_C ();
BX_CPU(i)->init (bx_mem_array[0]);
// assign apic ID from the index of this loop
- there were cases where BX_APIC_SUPPORT were used and others where BX_SUPPORT_APIC were used. To follow the pattern used by other names like this, I changed them all to BX_SUPPORT_APIC. Thanks to Tom Lindström for chasing this down!
2001-06-12 17:07:43 +04:00
// if !BX_SUPPORT_APIC, this will not compile.
- merged BRANCH-smp-bochs into main branch. For details see comments in BRANCH-smp-bochs revisions. - The general task was to make multiple CPU's which communicate through their APICs. So instead of BX_CPU and BX_MEM, we now have BX_CPU(x) and BX_MEM(y). For an SMP simulation you have several processors in a shared memory space, so there might be processors BX_CPU(0..3) but only one memory space BX_MEM(0). For cosimulation, you could have BX_CPU(0) with BX_MEM(0), then BX_CPU(1) with BX_MEM(1). WARNING: Cosimulation is almost certainly broken by the SMP changes. - to simulate multiple CPUs, you have to give each CPU time to execute in turn. This is currently implemented using debugger guards. The cpu loop steps one CPU for a few instructions, then steps the next CPU for a few instructions, etc. - there is some limited support in the debugger for two CPUs, for example printing information from each CPU when single stepping.
2001-05-23 12:16:07 +04:00
BX_CPU(i)->local_apic.set_id (i);
BX_CPU(i)->reset(BX_RESET_HARDWARE);
}
- another speed boost: when not using SMP, use BX_CPU_C bx_cpu; BX_MEM_C bx_mem; and when more than one processor, use BX_CPU_C *bx_cpu_array[BX_SMP_PROCESSORS]; BX_MEM_C *bx_mem_array[BX_ADDRESS_SPACES]; The changeover is controlled by BX_SMP_PROCESSORS, but there are only a few code changes since nearly all code uses the BX_CPU(n) and BX_MEM(n) macros. - This turns out to make a 10% speed difference! With this revision, the CVS version now gets 95% of the performance of the 3/25/2000 snapshot, which I've been using as my baseline.
2001-06-05 21:35:08 +04:00
#endif
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
#if BX_DEBUGGER == 0
- merged BRANCH-smp-bochs into main branch. For details see comments in BRANCH-smp-bochs revisions. - The general task was to make multiple CPU's which communicate through their APICs. So instead of BX_CPU and BX_MEM, we now have BX_CPU(x) and BX_MEM(y). For an SMP simulation you have several processors in a shared memory space, so there might be processors BX_CPU(0..3) but only one memory space BX_MEM(0). For cosimulation, you could have BX_CPU(0) with BX_MEM(0), then BX_CPU(1) with BX_MEM(1). WARNING: Cosimulation is almost certainly broken by the SMP changes. - to simulate multiple CPUs, you have to give each CPU time to execute in turn. This is currently implemented using debugger guards. The cpu loop steps one CPU for a few instructions, then steps the next CPU for a few instructions, etc. - there is some limited support in the debugger for two CPUs, for example printing information from each CPU when single stepping.
2001-05-23 12:16:07 +04:00
bx_devices.init(BX_MEM(0));
- This implements a solution that lets any GUI take over handling of any signal. First, selection of the GUI should cause BX_GUI_SIGHANDLER to be defined in config.h.in. Then, the GUI should define member functions Bit32u get_sighandler_mask (); void sighandler (int sig); The mask function returns a bitfield where one bit corresponds to each signal. For any signal whose bit is set to 1 in the return value of get_sighandler_mask, the gui will control that signal. When the signal arrives, bx_gui.sighandler(sig) will be called by bx_signal_handler, instead of the default behavior of that signal.
2001-05-09 00:18:04 +04:00
bx_gui.init_signal_handlers ();
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
bx_pc_system.start_timers();
#endif
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_DEBUG(("bx_init_hardware is setting signal handlers"));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
// if not using debugger, then we can take control of SIGINT.
// If using debugger, it needs control of this.
#if BX_DEBUGGER==0
signal(SIGINT, bx_signal_handler);
#endif
#if BX_SHOW_IPS
#ifndef __MINGW32__
signal(SIGALRM, bx_signal_handler);
#endif
alarm( 1 );
#endif
return(0);
}
void
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
bx_init_bx_dbg (void)
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
{
bx_dbg.floppy = 0;
bx_dbg.keyboard = 0;
bx_dbg.video = 0;
bx_dbg.disk = 0;
bx_dbg.pit = 0;
bx_dbg.pic = 0;
bx_dbg.bios = 0;
bx_dbg.cmos = 0;
bx_dbg.a20 = 0;
bx_dbg.interrupts = 0;
bx_dbg.exceptions = 0;
bx_dbg.unsupported = 0;
bx_dbg.temp = 0;
bx_dbg.reset = 0;
bx_dbg.mouse = 0;
bx_dbg.io = 0;
bx_dbg.debugger = 0;
bx_dbg.xms = 0;
bx_dbg.v8086 = 0;
bx_dbg.paging = 0;
bx_dbg.creg = 0;
bx_dbg.dreg = 0;
bx_dbg.dma = 0;
bx_dbg.unsupported_io = 0;
bx_dbg.record_io = 0;
bx_dbg.serial = 0;
bx_dbg.cdrom = 0;
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
#ifdef MAGIC_BREAKPOINT
bx_dbg.magic_break_enabled = 0;
#endif
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
void
bx_atexit(void)
{
static Boolean been_here = 0;
#if BX_PROVIDE_DEVICE_MODELS==1
if (been_here == 0) {
bx_pc_system.exit();
}
#endif
#if BX_DEBUGGER == 0
- merged BRANCH-smp-bochs into main branch. For details see comments in BRANCH-smp-bochs revisions. - The general task was to make multiple CPU's which communicate through their APICs. So instead of BX_CPU and BX_MEM, we now have BX_CPU(x) and BX_MEM(y). For an SMP simulation you have several processors in a shared memory space, so there might be processors BX_CPU(0..3) but only one memory space BX_MEM(0). For cosimulation, you could have BX_CPU(0) with BX_MEM(0), then BX_CPU(1) with BX_MEM(1). WARNING: Cosimulation is almost certainly broken by the SMP changes. - to simulate multiple CPUs, you have to give each CPU time to execute in turn. This is currently implemented using debugger guards. The cpu loop steps one CPU for a few instructions, then steps the next CPU for a few instructions, etc. - there is some limited support in the debugger for two CPUs, for example printing information from each CPU when single stepping.
2001-05-23 12:16:07 +04:00
for (int cpu=0; cpu<BX_SMP_PROCESSORS; cpu++)
if (BX_CPU(cpu)) BX_CPU(cpu)->atexit();
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
#endif
#if BX_PCI_SUPPORT
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
if (bx_options.Oi440FXSupport->get ()) {
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_devices.pci->print_i440fx_state();
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
#endif
}
#if (BX_PROVIDE_CPU_MEMORY==1) && (BX_EMULATE_HGA_DUMPS>0)
void
bx_emulate_hga_dumps_timer(void)
{
void bx_hga_set_video_memory(Bit8u *ptr);
bx_hga_set_video_memory(&bx_phy_memory[0xb0000]);
}
#endif
#if BX_PROVIDE_MAIN
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
char *
bx_find_bochsrc ()
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
{
- clean up compiler warnings
2001-06-12 00:39:05 +04:00
FILE *fd = NULL;
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
char rcfile[512];
- clean up stderr output a lot - give hints if they type "bochs" with no args and no .bochsrc is found
2001-05-22 22:47:30 +04:00
Bit32u retry = 0, found = 0;
// try several possibilities for the bochsrc before giving up
while (!found) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
rcfile[0] = 0;
- clean up stderr output a lot - give hints if they type "bochs" with no args and no .bochsrc is found
2001-05-22 22:47:30 +04:00
switch (retry++) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
case 0: strcpy (rcfile, ".bochsrc"); break;
case 1: strcpy (rcfile, "bochsrc"); break;
case 2: strcpy (rcfile, "bochsrc.txt"); break;
- clean up stderr output a lot - give hints if they type "bochs" with no args and no .bochsrc is found
2001-05-22 22:47:30 +04:00
case 3:
- correct stupid problem that I caused yesterday. Must use #if for BX_WITH_MACOS, not #ifdef!!!
2001-09-26 19:22:52 +04:00
#if (!defined(WIN32)) && !BX_WITH_MACOS
- clean up stderr output a lot - give hints if they type "bochs" with no args and no .bochsrc is found
2001-05-22 22:47:30 +04:00
// only try this on unix
{
char *ptr = getenv("HOME");
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
if (ptr) sprintf (rcfile, "%s/.bochsrc", ptr);
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
- clean up stderr output a lot - give hints if they type "bochs" with no args and no .bochsrc is found
2001-05-22 22:47:30 +04:00
#endif
break;
default:
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
return NULL;
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
if (rcfile[0]) {
BX_DEBUG (("looking for configuration in %s", rcfile));
fd = fopen(rcfile, "r");
- clean up stderr output a lot - give hints if they type "bochs" with no args and no .bochsrc is found
2001-05-22 22:47:30 +04:00
if (fd) found = 1;
}
}
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
assert (fd != NULL && rcfile[0] != 0);
fclose (fd);
return strdup (rcfile);
}
static int
parse_bochsrc(char *rcfile)
{
FILE *fd = NULL;
char *ret;
char line[512];
// try several possibilities for the bochsrc before giving up
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
fd = fopen (rcfile, "r");
if (!fd) return -1;
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
do {
ret = fgets(line, sizeof(line)-1, fd);
line[sizeof(line) - 1] = '\0';
- bug fix: zero length line could cause bochs to segfault. (Maybe it only happened at the end.)
2001-05-23 19:00:10 +04:00
int len = strlen(line);
if (len>0)
line[len-1] = '\0';
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
if ((ret != NULL) && strlen(line)) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
parse_line_unformatted(rcfile, line);
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
} while (!feof(fd));
- clean up compiler warnings
2001-06-12 00:39:05 +04:00
return 0;
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
static void
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
parse_line_unformatted(char *context, char *line)
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
{
- apply patch.memleaks-tominac from Darko Tominac <darko.tominac@zg.tel.hr> which fixes some potential memory leaks
2001-12-21 22:33:18 +03:00
#define MAX_PARAMS_LEN 40
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
char *ptr;
unsigned i, string_i;
char string[512];
- apply patch.memleaks-tominac from Darko Tominac <darko.tominac@zg.tel.hr> which fixes some potential memory leaks
2001-12-21 22:33:18 +03:00
char *params[MAX_PARAMS_LEN];
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
int num_params;
Boolean inquotes = 0;
- apply patch.memleaks-tominac from Darko Tominac <darko.tominac@zg.tel.hr> which fixes some potential memory leaks
2001-12-21 22:33:18 +03:00
memset(params, 0, sizeof(params));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
if (line == NULL) return;
// if passed nothing but whitespace, just return
for (i=0; i<strlen(line); i++) {
if (!isspace(line[i])) break;
}
if (i>=strlen(line))
return;
num_params = 0;
ptr = strtok(line, ":");
while (ptr) {
string_i = 0;
for (i=0; i<strlen(ptr); i++) {
if (ptr[i] == '"')
inquotes = !inquotes;
- do environment variable substitutions on any strings in the .bochsrc
2001-11-11 08:44:40 +03:00
else {
#if BX_HAVE_GETENV
// substitute environment variables.
if (ptr[i] == '$') {
char varname[512];
char *pv = varname;
char *value;
*pv = 0;
i++;
while (isalpha(ptr[i]) || ptr[i]=='_') {
*pv = ptr[i]; pv++; i++;
}
*pv = 0;
if (strlen(varname)<1 || !(value = getenv(varname))) {
BX_PANIC (("could not look up environment variable '%s'\n", varname));
} else {
// append value to the string
for (pv=value; *pv; pv++)
string[string_i++] = *pv;
}
}
#endif
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
if (!isspace(ptr[i]) || inquotes) {
- do environment variable substitutions on any strings in the .bochsrc
2001-11-11 08:44:40 +03:00
string[string_i++] = ptr[i];
}
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
- do environment variable substitutions on any strings in the .bochsrc
2001-11-11 08:44:40 +03:00
}
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
string[string_i] = '\0';
- apply patch.memleaks-tominac from Darko Tominac <darko.tominac@zg.tel.hr> which fixes some potential memory leaks
2001-12-21 22:33:18 +03:00
if ( params[num_params] != NULL )
{
free(params[num_params]);
params[num_params] = NULL;
}
if ( num_params < MAX_PARAMS_LEN )
{
- do environment variable substitutions on any strings in the .bochsrc
2001-11-11 08:44:40 +03:00
params[num_params++] = strdup (string);
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
ptr = strtok(NULL, ",");
- apply patch.memleaks-tominac from Darko Tominac <darko.tominac@zg.tel.hr> which fixes some potential memory leaks
2001-12-21 22:33:18 +03:00
}
else
{
BX_PANIC (("too many parameters, max is %d\n", MAX_PARAMS_LEN));
}
- do environment variable substitutions on any strings in the .bochsrc
2001-11-11 08:44:40 +03:00
}
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
parse_line_formatted(context, num_params, &params[0]);
- apply patch.memleaks-tominac from Darko Tominac <darko.tominac@zg.tel.hr> which fixes some potential memory leaks
2001-12-21 22:33:18 +03:00
for (i=0; i < MAX_PARAMS_LEN; i++)
{
if ( params[i] != NULL )
{
free(params[i]);
params[i] = NULL;
}
}
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
static void
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
parse_line_formatted(char *context, int num_params, char *params[])
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
{
int i;
if (num_params < 1) return;
if (params[0][0] == '#') return; /* comment */
else if (!strcmp(params[0], "floppya")) {
for (i=1; i<num_params; i++) {
- 2.88M floppy patch by Todd Fries
2001-05-08 21:45:44 +04:00
if (!strncmp(params[i], "2_88=", 5)) {
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.floppya.Opath->set (&params[i][5]);
bx_options.floppya.Otype->set (BX_FLOPPY_2_88);
- 2.88M floppy patch by Todd Fries
2001-05-08 21:45:44 +04:00
}
else if (!strncmp(params[i], "1_44=", 5)) {
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.floppya.Opath->set (&params[i][5]);
bx_options.floppya.Otype->set (BX_FLOPPY_1_44);
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
else if (!strncmp(params[i], "1_2=", 4)) {
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.floppya.Opath->set (&params[i][4]);
bx_options.floppya.Otype->set (BX_FLOPPY_1_2);
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
else if (!strncmp(params[i], "720k=", 5)) {
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.floppya.Opath->set (&params[i][5]);
bx_options.floppya.Otype->set (BX_FLOPPY_720K);
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
else if (!strncmp(params[i], "status=ejected", 14)) {
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.floppya.Oinitial_status->set (BX_EJECTED);
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
else if (!strncmp(params[i], "status=inserted", 15)) {
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.floppya.Oinitial_status->set (BX_INSERTED);
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
else {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: floppya attribute '%s' not understood.", context,
- replaced all the "fprintf(stderr...); exit(1);" lines with BX_PANIC and BX_INFO calls instead.
2001-05-23 19:13:05 +04:00
params[i]));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
}
}
else if (!strcmp(params[0], "floppyb")) {
for (i=1; i<num_params; i++) {
- 2.88M floppy patch by Todd Fries
2001-05-08 21:45:44 +04:00
if (!strncmp(params[i], "2_88=", 5)) {
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.floppyb.Opath->set (&params[i][5]);
bx_options.floppyb.Otype->set (BX_FLOPPY_2_88);
- 2.88M floppy patch by Todd Fries
2001-05-08 21:45:44 +04:00
}
else if (!strncmp(params[i], "1_44=", 5)) {
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.floppyb.Opath->set (&params[i][5]);
bx_options.floppyb.Otype->set (BX_FLOPPY_1_44);
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
else if (!strncmp(params[i], "1_2=", 4)) {
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.floppyb.Opath->set (&params[i][4]);
bx_options.floppyb.Otype->set (BX_FLOPPY_1_2);
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
else if (!strncmp(params[i], "720k=", 5)) {
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.floppyb.Opath->set (&params[i][5]);
bx_options.floppyb.Otype->set (BX_FLOPPY_720K);
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
else if (!strncmp(params[i], "status=ejected", 14)) {
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.floppyb.Oinitial_status->set (BX_EJECTED);
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
else if (!strncmp(params[i], "status=inserted", 15)) {
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.floppyb.Oinitial_status->set (BX_INSERTED);
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
else {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: floppyb attribute '%s' not understood.", context,
- replaced all the "fprintf(stderr...); exit(1);" lines with BX_PANIC and BX_INFO calls instead.
2001-05-23 19:13:05 +04:00
params[i]));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
}
}
else if (!strcmp(params[0], "diskc")) {
if (num_params != 5) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: diskc directive malformed.", context));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
if (strncmp(params[1], "file=", 5) ||
strncmp(params[2], "cyl=", 4) ||
strncmp(params[3], "heads=", 6) ||
strncmp(params[4], "spt=", 4)) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: diskc directive malformed.", context));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.diskc.Opath->set (&params[1][5]);
bx_options.diskc.Ocylinders->set (atol(&params[2][4]));
bx_options.diskc.Oheads->set (atol(&params[3][6]));
bx_options.diskc.Ospt->set (atol(&params[4][4]));
bx_options.diskc.Opresent->set (1);
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
else if (!strcmp(params[0], "diskd")) {
if (num_params != 5) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: diskd directive malformed.", context));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
if (strncmp(params[1], "file=", 5) ||
strncmp(params[2], "cyl=", 4) ||
strncmp(params[3], "heads=", 6) ||
strncmp(params[4], "spt=", 4)) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: diskd directive malformed.", context));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.diskd.Opath->set (&params[1][5]);
bx_options.diskd.Ocylinders->set (atol( &params[2][4]));
bx_options.diskd.Oheads->set (atol( &params[3][6] ));
bx_options.diskd.Ospt->set (atol( &params[4][4] ));
bx_options.diskd.Opresent->set (1);
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
- commit serial port patch from Stu Grossman <grossman@juniper.net> Here are the patches (without the serial name fix). Let me know if you want me to change that. Oh, FYI, my changes also open up the com1 terminal in raw mode, which makes the emulation much more accurate. You'll also notice that I added infrastructure for com2->4 in the option parsing. I didn't add it to the serial code, as I think a bunch of things need to be untangled before that can work.
2002-03-03 09:03:29 +03:00
else if (!strcmp(params[0], "com1")) {
if (num_params != 2) {
BX_PANIC(("%s: com1 directive malformed.", context));
}
if (strncmp(params[1], "dev=", 4)) {
BX_PANIC(("%s: com1 directive malformed.", context));
}
bx_options.com1.Odev->set (&params[1][4]);
bx_options.com1.Opresent->set (1);
}
else if (!strcmp(params[0], "com2")) {
if (num_params != 2) {
BX_PANIC(("%s: com2 directive malformed.", context));
}
if (strncmp(params[1], "dev=", 4)) {
BX_PANIC(("%s: com2 directive malformed.", context));
}
bx_options.com2.Odev->set (&params[1][4]);
bx_options.com2.Opresent->set (1);
}
else if (!strcmp(params[0], "com3")) {
if (num_params != 2) {
BX_PANIC(("%s: com3 directive malformed.", context));
}
if (strncmp(params[1], "dev=", 4)) {
BX_PANIC(("%s: com3 directive malformed.", context));
}
bx_options.com3.Odev->set (&params[1][4]);
bx_options.com3.Opresent->set (1);
}
else if (!strcmp(params[0], "com4")) {
if (num_params != 2) {
BX_PANIC(("%s: com4 directive malformed.", context));
}
if (strncmp(params[1], "dev=", 4)) {
BX_PANIC(("%s: com4 directive malformed.", context));
}
bx_options.com4.Odev->set (&params[1][4]);
bx_options.com4.Opresent->set (1);
}
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
else if (!strcmp(params[0], "cdromd")) {
if (num_params != 3) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: cdromd directive malformed.", context));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
if (strncmp(params[1], "dev=", 4) || strncmp(params[2], "status=", 7)) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: cdromd directive malformed.", context));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.cdromd.Opath->set (&params[1][4]);
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
if (!strcmp(params[2], "status=inserted"))
- since "bx_options.cdromd.Oinserted" changed from bool to enum, I had to change the parsing code to set BX_INSERTED and BX_EJECTED instead of 0,1.
2001-06-21 19:22:26 +04:00
bx_options.cdromd.Oinserted->set (BX_INSERTED);
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
else if (!strcmp(params[2], "status=ejected"))
- since "bx_options.cdromd.Oinserted" changed from bool to enum, I had to change the parsing code to set BX_INSERTED and BX_EJECTED instead of 0,1.
2001-06-21 19:22:26 +04:00
bx_options.cdromd.Oinserted->set (BX_EJECTED);
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
else {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: cdromd directive malformed.", context));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.cdromd.Opresent->set (1);
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
else if (!strcmp(params[0], "boot")) {
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
if (!strcmp(params[1], "a")) {
bx_options.Obootdrive->set (BX_BOOT_FLOPPYA);
} else if (!strcmp(params[1], "c")) {
bx_options.Obootdrive->set (BX_BOOT_DISKC);
. changes to enable El-Torito boot. It is disabled by default in rombios.c
2002-01-30 13:30:52 +03:00
} else if (!strcmp(params[1], "cdrom")) {
bx_options.Obootdrive->set (BX_BOOT_CDROM);
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
} else {
. changes to enable El-Torito boot. It is disabled by default in rombios.c
2002-01-30 13:30:52 +03:00
BX_PANIC(("%s: boot directive with unknown boot device '%s'. use 'a', 'c' or 'cdrom'.", context, params[1]));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
}
else if (!strcmp(params[0], "log")) {
if (num_params != 2) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: log directive has wrong # args.", context));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.log.Ofilename->set (params[1]);
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
- new, partially implemented, lines in .bochsrc allow run-time choice of what action to take on panic, error, info, and debug. The lines don't have any effect quite yet because of an initialization order problem with the logfunctions.
2001-05-23 00:01:40 +04:00
else if (!strcmp(params[0], "panic")) {
if (num_params != 2) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: panic directive malformed.", context));
- new, partially implemented, lines in .bochsrc allow run-time choice of what action to take on panic, error, info, and debug. The lines don't have any effect quite yet because of an initialization order problem with the logfunctions.
2001-05-23 00:01:40 +04:00
}
if (strncmp(params[1], "action=", 7)) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: panic directive malformed.", context));
- new, partially implemented, lines in .bochsrc allow run-time choice of what action to take on panic, error, info, and debug. The lines don't have any effect quite yet because of an initialization order problem with the logfunctions.
2001-05-23 00:01:40 +04:00
}
char *action = 7 + params[1];
- change name of action "crash" to "fatal" - change Makefile.in to "cvs-snapshot" information rather than printing the release number, since in fact it's not the same as the release anymore - use #defines to give the actions names (ACT_IGNORE, ACT_REPORT, ACT_FATAL) - register all logfunctions as they call setio, and keep a list so that we can change their settings more easily. I admit I used a static-sized array to store them. Sorry. - instead of printing a prefix [GEN ] for generic messages, just leave the prefix blank [ ]. Otherwise you start wondering what does gen stand for. - avoid reentry into fatal, since fatal can call BX_INFO and friends. This avoids a potential infinite recursion. - now that files other than .bochsrc can be the rc file, print the actual name of the file in error messages.
2001-05-23 06:37:52 +04:00
if (!strcmp(action, "fatal"))
- implemented bx_write_configuration which, for the first time ever, writes a bochsrc for you. - since there were two options related to logging, I moved them both into a new struct called bx_log_options. This follows the pattern used by other devices. - in control.cc: removed option 1 from main menu, the one that said "Read options from bochsrc.txt." This was identical to choosing the next option, "Read options from..." and only saved you one keystroke, so I removed it.
2001-06-11 10:35:18 +04:00
bx_options.log.actions[LOGLEV_PANIC] = ACT_FATAL;
- new, partially implemented, lines in .bochsrc allow run-time choice of what action to take on panic, error, info, and debug. The lines don't have any effect quite yet because of an initialization order problem with the logfunctions.
2001-05-23 00:01:40 +04:00
else if (!strcmp (action, "report"))
- implemented bx_write_configuration which, for the first time ever, writes a bochsrc for you. - since there were two options related to logging, I moved them both into a new struct called bx_log_options. This follows the pattern used by other devices. - in control.cc: removed option 1 from main menu, the one that said "Read options from bochsrc.txt." This was identical to choosing the next option, "Read options from..." and only saved you one keystroke, so I removed it.
2001-06-11 10:35:18 +04:00
bx_options.log.actions[LOGLEV_PANIC] = ACT_REPORT;
- new, partially implemented, lines in .bochsrc allow run-time choice of what action to take on panic, error, info, and debug. The lines don't have any effect quite yet because of an initialization order problem with the logfunctions.
2001-05-23 00:01:40 +04:00
else if (!strcmp (action, "ignore"))
- implemented bx_write_configuration which, for the first time ever, writes a bochsrc for you. - since there were two options related to logging, I moved them both into a new struct called bx_log_options. This follows the pattern used by other devices. - in control.cc: removed option 1 from main menu, the one that said "Read options from bochsrc.txt." This was identical to choosing the next option, "Read options from..." and only saved you one keystroke, so I removed it.
2001-06-11 10:35:18 +04:00
bx_options.log.actions[LOGLEV_PANIC] = ACT_IGNORE;
- added a new log action "ask" which sends the log message to the control panel. It gives the user the option of continue this time, continue and never ask again for this type of log message, or die. - the communication between control.cc and siminterface.cc is still somewhat crude. I'm trying to implement it so that a network protocol based on this interface will be simple. I'll get this cleaned up soon.
2001-06-11 18:03:35 +04:00
else if (!strcmp (action, "ask"))
bx_options.log.actions[LOGLEV_PANIC] = ACT_ASK;
- new, partially implemented, lines in .bochsrc allow run-time choice of what action to take on panic, error, info, and debug. The lines don't have any effect quite yet because of an initialization order problem with the logfunctions.
2001-05-23 00:01:40 +04:00
else {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: panic directive malformed.", context));
- new, partially implemented, lines in .bochsrc allow run-time choice of what action to take on panic, error, info, and debug. The lines don't have any effect quite yet because of an initialization order problem with the logfunctions.
2001-05-23 00:01:40 +04:00
}
}
else if (!strcmp(params[0], "error")) {
if (num_params != 2) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: error directive malformed.", context));
- new, partially implemented, lines in .bochsrc allow run-time choice of what action to take on panic, error, info, and debug. The lines don't have any effect quite yet because of an initialization order problem with the logfunctions.
2001-05-23 00:01:40 +04:00
}
if (strncmp(params[1], "action=", 7)) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: error directive malformed.", context));
- new, partially implemented, lines in .bochsrc allow run-time choice of what action to take on panic, error, info, and debug. The lines don't have any effect quite yet because of an initialization order problem with the logfunctions.
2001-05-23 00:01:40 +04:00
}
char *action = 7 + params[1];
- change name of action "crash" to "fatal" - change Makefile.in to "cvs-snapshot" information rather than printing the release number, since in fact it's not the same as the release anymore - use #defines to give the actions names (ACT_IGNORE, ACT_REPORT, ACT_FATAL) - register all logfunctions as they call setio, and keep a list so that we can change their settings more easily. I admit I used a static-sized array to store them. Sorry. - instead of printing a prefix [GEN ] for generic messages, just leave the prefix blank [ ]. Otherwise you start wondering what does gen stand for. - avoid reentry into fatal, since fatal can call BX_INFO and friends. This avoids a potential infinite recursion. - now that files other than .bochsrc can be the rc file, print the actual name of the file in error messages.
2001-05-23 06:37:52 +04:00
if (!strcmp(action, "fatal"))
- implemented bx_write_configuration which, for the first time ever, writes a bochsrc for you. - since there were two options related to logging, I moved them both into a new struct called bx_log_options. This follows the pattern used by other devices. - in control.cc: removed option 1 from main menu, the one that said "Read options from bochsrc.txt." This was identical to choosing the next option, "Read options from..." and only saved you one keystroke, so I removed it.
2001-06-11 10:35:18 +04:00
bx_options.log.actions[LOGLEV_ERROR] = ACT_FATAL;
- new, partially implemented, lines in .bochsrc allow run-time choice of what action to take on panic, error, info, and debug. The lines don't have any effect quite yet because of an initialization order problem with the logfunctions.
2001-05-23 00:01:40 +04:00
else if (!strcmp (action, "report"))
- implemented bx_write_configuration which, for the first time ever, writes a bochsrc for you. - since there were two options related to logging, I moved them both into a new struct called bx_log_options. This follows the pattern used by other devices. - in control.cc: removed option 1 from main menu, the one that said "Read options from bochsrc.txt." This was identical to choosing the next option, "Read options from..." and only saved you one keystroke, so I removed it.
2001-06-11 10:35:18 +04:00
bx_options.log.actions[LOGLEV_ERROR] = ACT_REPORT;
- new, partially implemented, lines in .bochsrc allow run-time choice of what action to take on panic, error, info, and debug. The lines don't have any effect quite yet because of an initialization order problem with the logfunctions.
2001-05-23 00:01:40 +04:00
else if (!strcmp (action, "ignore"))
- implemented bx_write_configuration which, for the first time ever, writes a bochsrc for you. - since there were two options related to logging, I moved them both into a new struct called bx_log_options. This follows the pattern used by other devices. - in control.cc: removed option 1 from main menu, the one that said "Read options from bochsrc.txt." This was identical to choosing the next option, "Read options from..." and only saved you one keystroke, so I removed it.
2001-06-11 10:35:18 +04:00
bx_options.log.actions[LOGLEV_ERROR] = ACT_IGNORE;
- added a new log action "ask" which sends the log message to the control panel. It gives the user the option of continue this time, continue and never ask again for this type of log message, or die. - the communication between control.cc and siminterface.cc is still somewhat crude. I'm trying to implement it so that a network protocol based on this interface will be simple. I'll get this cleaned up soon.
2001-06-11 18:03:35 +04:00
else if (!strcmp (action, "ask"))
bx_options.log.actions[LOGLEV_PANIC] = ACT_ASK;
- new, partially implemented, lines in .bochsrc allow run-time choice of what action to take on panic, error, info, and debug. The lines don't have any effect quite yet because of an initialization order problem with the logfunctions.
2001-05-23 00:01:40 +04:00
else {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: error directive malformed.", context));
- new, partially implemented, lines in .bochsrc allow run-time choice of what action to take on panic, error, info, and debug. The lines don't have any effect quite yet because of an initialization order problem with the logfunctions.
2001-05-23 00:01:40 +04:00
}
}
else if (!strcmp(params[0], "info")) {
if (num_params != 2) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: info directive malformed.", context));
- new, partially implemented, lines in .bochsrc allow run-time choice of what action to take on panic, error, info, and debug. The lines don't have any effect quite yet because of an initialization order problem with the logfunctions.
2001-05-23 00:01:40 +04:00
}
if (strncmp(params[1], "action=", 7)) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: info directive malformed.", context));
- new, partially implemented, lines in .bochsrc allow run-time choice of what action to take on panic, error, info, and debug. The lines don't have any effect quite yet because of an initialization order problem with the logfunctions.
2001-05-23 00:01:40 +04:00
}
char *action = 7 + params[1];
- change name of action "crash" to "fatal" - change Makefile.in to "cvs-snapshot" information rather than printing the release number, since in fact it's not the same as the release anymore - use #defines to give the actions names (ACT_IGNORE, ACT_REPORT, ACT_FATAL) - register all logfunctions as they call setio, and keep a list so that we can change their settings more easily. I admit I used a static-sized array to store them. Sorry. - instead of printing a prefix [GEN ] for generic messages, just leave the prefix blank [ ]. Otherwise you start wondering what does gen stand for. - avoid reentry into fatal, since fatal can call BX_INFO and friends. This avoids a potential infinite recursion. - now that files other than .bochsrc can be the rc file, print the actual name of the file in error messages.
2001-05-23 06:37:52 +04:00
if (!strcmp(action, "fatal"))
- implemented bx_write_configuration which, for the first time ever, writes a bochsrc for you. - since there were two options related to logging, I moved them both into a new struct called bx_log_options. This follows the pattern used by other devices. - in control.cc: removed option 1 from main menu, the one that said "Read options from bochsrc.txt." This was identical to choosing the next option, "Read options from..." and only saved you one keystroke, so I removed it.
2001-06-11 10:35:18 +04:00
bx_options.log.actions[LOGLEV_INFO] = ACT_FATAL;
- new, partially implemented, lines in .bochsrc allow run-time choice of what action to take on panic, error, info, and debug. The lines don't have any effect quite yet because of an initialization order problem with the logfunctions.
2001-05-23 00:01:40 +04:00
else if (!strcmp (action, "report"))
- implemented bx_write_configuration which, for the first time ever, writes a bochsrc for you. - since there were two options related to logging, I moved them both into a new struct called bx_log_options. This follows the pattern used by other devices. - in control.cc: removed option 1 from main menu, the one that said "Read options from bochsrc.txt." This was identical to choosing the next option, "Read options from..." and only saved you one keystroke, so I removed it.
2001-06-11 10:35:18 +04:00
bx_options.log.actions[LOGLEV_INFO] = ACT_REPORT;
- new, partially implemented, lines in .bochsrc allow run-time choice of what action to take on panic, error, info, and debug. The lines don't have any effect quite yet because of an initialization order problem with the logfunctions.
2001-05-23 00:01:40 +04:00
else if (!strcmp (action, "ignore"))
- implemented bx_write_configuration which, for the first time ever, writes a bochsrc for you. - since there were two options related to logging, I moved them both into a new struct called bx_log_options. This follows the pattern used by other devices. - in control.cc: removed option 1 from main menu, the one that said "Read options from bochsrc.txt." This was identical to choosing the next option, "Read options from..." and only saved you one keystroke, so I removed it.
2001-06-11 10:35:18 +04:00
bx_options.log.actions[LOGLEV_INFO] = ACT_IGNORE;
- added a new log action "ask" which sends the log message to the control panel. It gives the user the option of continue this time, continue and never ask again for this type of log message, or die. - the communication between control.cc and siminterface.cc is still somewhat crude. I'm trying to implement it so that a network protocol based on this interface will be simple. I'll get this cleaned up soon.
2001-06-11 18:03:35 +04:00
else if (!strcmp (action, "ask"))
bx_options.log.actions[LOGLEV_PANIC] = ACT_ASK;
- new, partially implemented, lines in .bochsrc allow run-time choice of what action to take on panic, error, info, and debug. The lines don't have any effect quite yet because of an initialization order problem with the logfunctions.
2001-05-23 00:01:40 +04:00
else {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: info directive malformed.", context));
- new, partially implemented, lines in .bochsrc allow run-time choice of what action to take on panic, error, info, and debug. The lines don't have any effect quite yet because of an initialization order problem with the logfunctions.
2001-05-23 00:01:40 +04:00
}
}
else if (!strcmp(params[0], "debug")) {
if (num_params != 2) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: debug directive malformed.", context));
- new, partially implemented, lines in .bochsrc allow run-time choice of what action to take on panic, error, info, and debug. The lines don't have any effect quite yet because of an initialization order problem with the logfunctions.
2001-05-23 00:01:40 +04:00
}
if (strncmp(params[1], "action=", 7)) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: debug directive malformed.", context));
- new, partially implemented, lines in .bochsrc allow run-time choice of what action to take on panic, error, info, and debug. The lines don't have any effect quite yet because of an initialization order problem with the logfunctions.
2001-05-23 00:01:40 +04:00
}
char *action = 7 + params[1];
- change name of action "crash" to "fatal" - change Makefile.in to "cvs-snapshot" information rather than printing the release number, since in fact it's not the same as the release anymore - use #defines to give the actions names (ACT_IGNORE, ACT_REPORT, ACT_FATAL) - register all logfunctions as they call setio, and keep a list so that we can change their settings more easily. I admit I used a static-sized array to store them. Sorry. - instead of printing a prefix [GEN ] for generic messages, just leave the prefix blank [ ]. Otherwise you start wondering what does gen stand for. - avoid reentry into fatal, since fatal can call BX_INFO and friends. This avoids a potential infinite recursion. - now that files other than .bochsrc can be the rc file, print the actual name of the file in error messages.
2001-05-23 06:37:52 +04:00
if (!strcmp(action, "fatal"))
- implemented bx_write_configuration which, for the first time ever, writes a bochsrc for you. - since there were two options related to logging, I moved them both into a new struct called bx_log_options. This follows the pattern used by other devices. - in control.cc: removed option 1 from main menu, the one that said "Read options from bochsrc.txt." This was identical to choosing the next option, "Read options from..." and only saved you one keystroke, so I removed it.
2001-06-11 10:35:18 +04:00
bx_options.log.actions[LOGLEV_DEBUG] = ACT_FATAL;
- new, partially implemented, lines in .bochsrc allow run-time choice of what action to take on panic, error, info, and debug. The lines don't have any effect quite yet because of an initialization order problem with the logfunctions.
2001-05-23 00:01:40 +04:00
else if (!strcmp (action, "report"))
- implemented bx_write_configuration which, for the first time ever, writes a bochsrc for you. - since there were two options related to logging, I moved them both into a new struct called bx_log_options. This follows the pattern used by other devices. - in control.cc: removed option 1 from main menu, the one that said "Read options from bochsrc.txt." This was identical to choosing the next option, "Read options from..." and only saved you one keystroke, so I removed it.
2001-06-11 10:35:18 +04:00
bx_options.log.actions[LOGLEV_DEBUG] = ACT_REPORT;
- new, partially implemented, lines in .bochsrc allow run-time choice of what action to take on panic, error, info, and debug. The lines don't have any effect quite yet because of an initialization order problem with the logfunctions.
2001-05-23 00:01:40 +04:00
else if (!strcmp (action, "ignore"))
- implemented bx_write_configuration which, for the first time ever, writes a bochsrc for you. - since there were two options related to logging, I moved them both into a new struct called bx_log_options. This follows the pattern used by other devices. - in control.cc: removed option 1 from main menu, the one that said "Read options from bochsrc.txt." This was identical to choosing the next option, "Read options from..." and only saved you one keystroke, so I removed it.
2001-06-11 10:35:18 +04:00
bx_options.log.actions[LOGLEV_DEBUG] = ACT_IGNORE;
- added a new log action "ask" which sends the log message to the control panel. It gives the user the option of continue this time, continue and never ask again for this type of log message, or die. - the communication between control.cc and siminterface.cc is still somewhat crude. I'm trying to implement it so that a network protocol based on this interface will be simple. I'll get this cleaned up soon.
2001-06-11 18:03:35 +04:00
else if (!strcmp (action, "ask"))
bx_options.log.actions[LOGLEV_PANIC] = ACT_ASK;
- new, partially implemented, lines in .bochsrc allow run-time choice of what action to take on panic, error, info, and debug. The lines don't have any effect quite yet because of an initialization order problem with the logfunctions.
2001-05-23 00:01:40 +04:00
else {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: debug directive malformed.", context));
- new, partially implemented, lines in .bochsrc allow run-time choice of what action to take on panic, error, info, and debug. The lines don't have any effect quite yet because of an initialization order problem with the logfunctions.
2001-05-23 00:01:40 +04:00
}
}
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
else if (!strcmp(params[0], "romimage")) {
if (num_params != 3) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: romimage directive: wrong # args.", context));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
if (strncmp(params[1], "file=", 5)) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: romimage directive malformed.", context));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
if (strncmp(params[2], "address=", 8)) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: romimage directive malformed.", context));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.rom.Opath->set (&params[1][5]);
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
if ( (params[2][8] == '0') && (params[2][9] == 'x') )
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.rom.Oaddress->set (strtoul (&params[2][8], NULL, 16));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
else
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.rom.Oaddress->set (strtoul (&params[2][8], NULL, 10));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
else if (!strcmp(params[0], "vgaromimage")) {
if (num_params != 2) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: vgaromimage directive: wrong # args.", context));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.vgarom.Opath->set (params[1]);
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
else if (!strcmp(params[0], "vga_update_interval")) {
if (num_params != 2) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: vga_update_interval directive: wrong # args.", context));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.Ovga_update_interval->set (atol(params[1]));
if (bx_options.Ovga_update_interval->get () < 50000) {
- now ips, mouse_enabled, vga_update_interval, and rom_path are declared as bx_param_c * types in the bx_options structure. They are initialized in main.cc (bx_init_options) with default values. Access to parameters of this type should always be like this: bx_options.mouse_enabled->get (); bx_options.mouse_enabled->set (newval); Eventually I will be transferring all options to this format.
2001-06-17 03:08:32 +04:00
BX_INFO(("%s: vga_update_interval seems awfully small!", context));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
}
else if (!strcmp(params[0], "keyboard_serial_delay")) {
if (num_params != 2) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: keyboard_serial_delay directive: wrong # args.", context));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.Okeyboard_serial_delay->set (atol(params[1]));
if (bx_options.Okeyboard_serial_delay->get () < 5) {
BX_ERROR (("%s: keyboard_serial_delay not big enough!", context));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
}
- apply Christophe's patch.paste-for-dos. I will make minor tweaks as separate revisions.
2002-03-26 16:51:48 +03:00
else if (!strcmp(params[0], "keyboard_paste_delay")) {
if (num_params != 2) {
BX_PANIC(("%s: keyboard_paste_delay directive: wrong # args.", context));
}
bx_options.Okeyboard_paste_delay->set (atol(params[1]));
if (bx_options.Okeyboard_paste_delay->get () < 1000) {
BX_ERROR (("%s: keyboard_paste_delay not big enough!", context));
}
}
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
else if (!strcmp(params[0], "megs")) {
if (num_params != 2) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: megs directive: wrong # args.", context));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.memory.Osize->set (atol(params[1]));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
else if (!strcmp(params[0], "floppy_command_delay")) {
if (num_params != 2) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: floppy_command_delay directive: wrong # args.", context));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.Ofloppy_command_delay->set (atol(params[1]));
if (bx_options.Ofloppy_command_delay->get () < 100) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: floppy_command_delay not big enough!", context));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
}
else if (!strcmp(params[0], "ips")) {
if (num_params != 2) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: ips directive: wrong # args.", context));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.Oips->set (atol(params[1]));
if (bx_options.Oips->get () < 200000) {
BX_ERROR(("%s: WARNING: ips is AWFULLY low!", context));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
}
- added two new bochsrc and control panel options: max_ips and system_clock_sync
2001-06-21 20:59:56 +04:00
else if (!strcmp(params[0], "max_ips")) {
if (num_params != 2) {
BX_PANIC(("%s: max_ips directive: wrong # args.", context));
}
- remove references to max_ips and system_clock_sync. If you need to revert, find patches/patch.remove-max-ips.
2001-10-07 04:58:10 +04:00
BX_INFO(("WARNING: max_ips not implemented"));
- added two new bochsrc and control panel options: max_ips and system_clock_sync
2001-06-21 20:59:56 +04:00
}
else if (!strcmp(params[0], "system_clock_sync")) {
if (num_params != 2) {
BX_PANIC(("%s: system_clock_sync directive malformed.", context));
}
if (strncmp(params[1], "enabled=", 8)) {
BX_PANIC(("%s: system_clock_sync directive malformed.", context));
}
if (params[1][8] == '0' || params[1][8] == '1')
- remove references to max_ips and system_clock_sync. If you need to revert, find patches/patch.remove-max-ips.
2001-10-07 04:58:10 +04:00
BX_INFO (("WARNING: system_clock_sync not implemented"));
- added two new bochsrc and control panel options: max_ips and system_clock_sync
2001-06-21 20:59:56 +04:00
else
BX_PANIC(("%s: system_clock_sync directive malformed.", context));
}
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
else if (!strcmp(params[0], "mouse")) {
if (num_params != 2) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: mouse directive malformed.", context));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
if (strncmp(params[1], "enabled=", 8)) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: mouse directive malformed.", context));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
- now ips, mouse_enabled, vga_update_interval, and rom_path are declared as bx_param_c * types in the bx_options structure. They are initialized in main.cc (bx_init_options) with default values. Access to parameters of this type should always be like this: bx_options.mouse_enabled->get (); bx_options.mouse_enabled->set (newval); Eventually I will be transferring all options to this format.
2001-06-17 03:08:32 +04:00
if (params[1][8] == '0' || params[1][8] == '1')
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.Omouse_enabled->set (params[1][8] - '0');
- now ips, mouse_enabled, vga_update_interval, and rom_path are declared as bx_param_c * types in the bx_options structure. They are initialized in main.cc (bx_init_options) with default values. Access to parameters of this type should always be like this: bx_options.mouse_enabled->get (); bx_options.mouse_enabled->set (newval); Eventually I will be transferring all options to this format.
2001-06-17 03:08:32 +04:00
else
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: mouse directive malformed.", context));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
else if (!strcmp(params[0], "private_colormap")) {
if (num_params != 2) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: private_colormap directive malformed.", context));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
if (strncmp(params[1], "enabled=", 8)) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: private_colormap directive malformed.", context));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
if (params[1][8] == '0' || params[1][8] == '1')
bx_options.Oprivate_colormap->set (params[1][8] - '0');
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
else {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: private_colormap directive malformed.", context));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
}
- add option called fullscreen, for Nicholai Benalal's amiga port
2001-08-16 00:17:19 +04:00
else if (!strcmp(params[0], "fullscreen")) {
ok, now the amiga target on cvs is complete
2001-08-16 16:35:52 +04:00
#if BX_WITH_AMIGAOS
- add option called fullscreen, for Nicholai Benalal's amiga port
2001-08-16 00:17:19 +04:00
if (num_params != 2) {
BX_PANIC(("%s: fullscreen directive malformed.", context));
}
if (strncmp(params[1], "enabled=", 8)) {
BX_PANIC(("%s: fullscreen directive malformed.", context));
}
- add string parameter called screenmode - put #if BX_USE_AMIGAOS...#endif around fullscreen and screenmode, since they only apply to amiga
2001-08-16 06:00:31 +04:00
if (params[1][8] == '0' || params[1][8] == '1') {
- add option called fullscreen, for Nicholai Benalal's amiga port
2001-08-16 00:17:19 +04:00
bx_options.Ofullscreen->set (params[1][8] - '0');
- add string parameter called screenmode - put #if BX_USE_AMIGAOS...#endif around fullscreen and screenmode, since they only apply to amiga
2001-08-16 06:00:31 +04:00
} else {
- add option called fullscreen, for Nicholai Benalal's amiga port
2001-08-16 00:17:19 +04:00
BX_PANIC(("%s: fullscreen directive malformed.", context));
}
- add string parameter called screenmode - put #if BX_USE_AMIGAOS...#endif around fullscreen and screenmode, since they only apply to amiga
2001-08-16 06:00:31 +04:00
#endif
}
else if (!strcmp(params[0], "screenmode")) {
ok, now the amiga target on cvs is complete
2001-08-16 16:35:52 +04:00
#if BX_WITH_AMIGAOS
- add string parameter called screenmode - put #if BX_USE_AMIGAOS...#endif around fullscreen and screenmode, since they only apply to amiga
2001-08-16 06:00:31 +04:00
if (num_params != 2) {
BX_PANIC(("%s: screenmode directive malformed.", context));
}
if (strncmp(params[1], "name=", 5)) {
BX_PANIC(("%s: screenmode directive malformed.", context));
}
bx_options.Oscreenmode->set (strdup(&params[1][5]));
#endif
- add option called fullscreen, for Nicholai Benalal's amiga port
2001-08-16 00:17:19 +04:00
}
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
else if (!strcmp(params[0], "sb16")) {
for (i=1; i<num_params; i++) {
- converted ne2k and sb16 to new style parameters and control panel, except for MAC address of the ne2k, which is next.
2001-06-21 22:34:50 +04:00
bx_options.sb16.Opresent->set (1);
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
if (!strncmp(params[i], "midi=", 5)) {
- converted ne2k and sb16 to new style parameters and control panel, except for MAC address of the ne2k, which is next.
2001-06-21 22:34:50 +04:00
bx_options.sb16.Omidifile->set (strdup(&params[i][5]));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
else if (!strncmp(params[i], "midimode=", 9)) {
- converted ne2k and sb16 to new style parameters and control panel, except for MAC address of the ne2k, which is next.
2001-06-21 22:34:50 +04:00
bx_options.sb16.Omidimode->set (atol(&params[i][9]));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
else if (!strncmp(params[i], "wave=", 5)) {
- converted ne2k and sb16 to new style parameters and control panel, except for MAC address of the ne2k, which is next.
2001-06-21 22:34:50 +04:00
bx_options.sb16.Owavefile->set (strdup(&params[i][5]));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
else if (!strncmp(params[i], "wavemode=", 9)) {
- converted ne2k and sb16 to new style parameters and control panel, except for MAC address of the ne2k, which is next.
2001-06-21 22:34:50 +04:00
bx_options.sb16.Owavemode->set (atol(&params[i][9]));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
else if (!strncmp(params[i], "log=", 4)) {
- converted ne2k and sb16 to new style parameters and control panel, except for MAC address of the ne2k, which is next.
2001-06-21 22:34:50 +04:00
bx_options.sb16.Ologfile->set (strdup(&params[i][4]));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
else if (!strncmp(params[i], "loglevel=", 9)) {
- converted ne2k and sb16 to new style parameters and control panel, except for MAC address of the ne2k, which is next.
2001-06-21 22:34:50 +04:00
bx_options.sb16.Ologlevel->set (atol(&params[i][9]));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
else if (!strncmp(params[i], "dmatimer=", 9)) {
- converted ne2k and sb16 to new style parameters and control panel, except for MAC address of the ne2k, which is next.
2001-06-21 22:34:50 +04:00
bx_options.sb16.Odmatimer->set (atol(&params[i][9]));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
}
}
- allow configuration of the parallel port in .bochsrc and the config interface menus. Parallel port #1 is implemented, and I left stubs for parallel port #2 in case we want to ever add it. If the parallel port is enabled, the init method of parallel.cc does an fopen() on the output file. If disabled, or if the fopen fails, the file handler remains NULL and no characters are printed. There is no attempt to enable/disable the operation of the parallel port, only the output to a file.
2001-11-12 05:35:09 +03:00
else if (!strcmp(params[0], "parport1")) {
for (i=1; i<num_params; i++) {
if (!strncmp(params[i], "enable=", 7)) {
bx_options.par1.Oenable->set (atol(&params[i][7]));
}
else if (!strncmp(params[i], "file=", 5)) {
bx_options.par1.Ooutfile->set (strdup(&params[i][5]));
}
}
}
#if 0
else if (!strcmp(params[0], "parport2")) {
for (i=1; i<num_params; i++) {
if (!strncmp(params[i], "enable=", 7)) {
bx_options.par1.Oenable->set (atol(&params[i][7]));
}
else if (!strncmp(params[i], "file=", 5)) {
bx_options.par1.Ooutfile->set (strdup(&params[i][5]));
}
}
}
#endif
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
else if (!strcmp(params[0], "i440fxsupport")) {
if (num_params != 2) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: i440FXSupport directive malformed.", context));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
if (strncmp(params[1], "enabled=", 8)) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: i440FXSupport directive malformed.", context));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
if (params[1][8] == '0')
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.Oi440FXSupport->set (0);
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
else if (params[1][8] == '1')
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.Oi440FXSupport->set (1);
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
else {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: i440FXSupport directive malformed.", context));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
}
else if (!strcmp(params[0], "newharddrivesupport")) {
if (num_params != 2) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: newharddrivesupport directive malformed.", context));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
if (strncmp(params[1], "enabled=", 8)) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: newharddrivesupport directive malformed.", context));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
if (params[1][8] == '0')
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.OnewHardDriveSupport->set (0);
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
else if (params[1][8] == '1')
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.OnewHardDriveSupport->set (1);
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
else {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: newharddrivesupport directive malformed.", context));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
}
else if (!strcmp(params[0], "cmosimage")) {
if (num_params != 2) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: cmosimage directive: wrong # args.", context));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.cmos.Opath->set (strdup(params[1]));
bx_options.cmos.OcmosImage->set (1); // CMOS Image is true
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
else if (!strcmp(params[0], "time0")) {
if (num_params != 2) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: time0 directive: wrong # args.", context));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.cmos.Otime0->set (atoi(params[1]));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
#ifdef MAGIC_BREAKPOINT
else if (!strcmp(params[0], "magic_break")) {
if (num_params != 2) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: magic_break directive: wrong # args.", context));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
if (strncmp(params[1], "enabled=", 8)) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: magic_break directive malformed.", context));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
if (params[1][8] == '0') {
strip '\n' from BX_{INFO,DEBUG,ERROR,PANIC} don't need it, moved the output of it into the general io functions. saves space, as well as removes the confusing output if a '\n' is left off
2001-05-30 22:56:02 +04:00
BX_INFO(("Ignoring magic break points"));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
bx_dbg.magic_break_enabled = 0;
}
else if (params[1][8] == '1') {
strip '\n' from BX_{INFO,DEBUG,ERROR,PANIC} don't need it, moved the output of it into the general io functions. saves space, as well as removes the confusing output if a '\n' is left off
2001-05-30 22:56:02 +04:00
BX_INFO(("Stopping on magic break points"));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
bx_dbg.magic_break_enabled = 1;
}
else {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: magic_break directive malformed.", context));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
}
#endif
else if (!strcmp(params[0], "ne2k")) {
int tmp[6];
- add options to the string parameter class that allow it to deal with raw hex bytes, for the MAC address.
2001-06-21 23:27:05 +04:00
char tmpchar[6];
- converted ne2k and sb16 to new style parameters and control panel, except for MAC address of the ne2k, which is next.
2001-06-21 22:34:50 +04:00
bx_options.ne2k.Ovalid->set (0);
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
if ((num_params < 4) || (num_params > 6)) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: ne2k directive malformed.", context));
- implemented bx_write_configuration which, for the first time ever, writes a bochsrc for you. - since there were two options related to logging, I moved them both into a new struct called bx_log_options. This follows the pattern used by other devices. - in control.cc: removed option 1 from main menu, the one that said "Read options from bochsrc.txt." This was identical to choosing the next option, "Read options from..." and only saved you one keystroke, so I removed it.
2001-06-11 10:35:18 +04:00
return;
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
- converted ne2k and sb16 to new style parameters and control panel, except for MAC address of the ne2k, which is next.
2001-06-21 22:34:50 +04:00
bx_options.ne2k.Oethmod->set ("null");
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
if (strncmp(params[1], "ioaddr=", 7)) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: ne2k directive malformed.", context));
- implemented bx_write_configuration which, for the first time ever, writes a bochsrc for you. - since there were two options related to logging, I moved them both into a new struct called bx_log_options. This follows the pattern used by other devices. - in control.cc: removed option 1 from main menu, the one that said "Read options from bochsrc.txt." This was identical to choosing the next option, "Read options from..." and only saved you one keystroke, so I removed it.
2001-06-11 10:35:18 +04:00
return;
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
if (strncmp(params[2], "irq=", 4)) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: ne2k directive malformed.", context));
- implemented bx_write_configuration which, for the first time ever, writes a bochsrc for you. - since there were two options related to logging, I moved them both into a new struct called bx_log_options. This follows the pattern used by other devices. - in control.cc: removed option 1 from main menu, the one that said "Read options from bochsrc.txt." This was identical to choosing the next option, "Read options from..." and only saved you one keystroke, so I removed it.
2001-06-11 10:35:18 +04:00
return;
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
if (strncmp(params[3], "mac=", 4)) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: ne2k directive malformed.", context));
- implemented bx_write_configuration which, for the first time ever, writes a bochsrc for you. - since there were two options related to logging, I moved them both into a new struct called bx_log_options. This follows the pattern used by other devices. - in control.cc: removed option 1 from main menu, the one that said "Read options from bochsrc.txt." This was identical to choosing the next option, "Read options from..." and only saved you one keystroke, so I removed it.
2001-06-11 10:35:18 +04:00
return;
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
- converted ne2k and sb16 to new style parameters and control panel, except for MAC address of the ne2k, which is next.
2001-06-21 22:34:50 +04:00
bx_options.ne2k.Oioaddr->set (strtoul(&params[1][7], NULL, 16));
bx_options.ne2k.Oirq->set (atol(&params[2][4]));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
i = sscanf(&params[3][4], "%x:%x:%x:%x:%x:%x",
&tmp[0],&tmp[1],&tmp[2],&tmp[3],&tmp[4],&tmp[5]);
if (i != 6) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: ne2k mac address malformed.", context));
- implemented bx_write_configuration which, for the first time ever, writes a bochsrc for you. - since there were two options related to logging, I moved them both into a new struct called bx_log_options. This follows the pattern used by other devices. - in control.cc: removed option 1 from main menu, the one that said "Read options from bochsrc.txt." This was identical to choosing the next option, "Read options from..." and only saved you one keystroke, so I removed it.
2001-06-11 10:35:18 +04:00
return;
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
for (i=0;i<6;i++)
- add options to the string parameter class that allow it to deal with raw hex bytes, for the MAC address.
2001-06-21 23:27:05 +04:00
tmpchar[i] = (unsigned char)tmp[i];
bx_options.ne2k.Omacaddr->set (tmpchar);
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
if (num_params > 4) {
if (strncmp(params[4], "ethmod=", 7)) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: ne2k directive malformed.", context));
- implemented bx_write_configuration which, for the first time ever, writes a bochsrc for you. - since there were two options related to logging, I moved them both into a new struct called bx_log_options. This follows the pattern used by other devices. - in control.cc: removed option 1 from main menu, the one that said "Read options from bochsrc.txt." This was identical to choosing the next option, "Read options from..." and only saved you one keystroke, so I removed it.
2001-06-11 10:35:18 +04:00
return;
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
- converted ne2k and sb16 to new style parameters and control panel, except for MAC address of the ne2k, which is next.
2001-06-21 22:34:50 +04:00
bx_options.ne2k.Oethmod->set (strdup(&params[4][7]));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
if (num_params == 6) {
if (strncmp(params[5], "ethdev=", 7)) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: ne2k directive malformed.", context));
- implemented bx_write_configuration which, for the first time ever, writes a bochsrc for you. - since there were two options related to logging, I moved them both into a new struct called bx_log_options. This follows the pattern used by other devices. - in control.cc: removed option 1 from main menu, the one that said "Read options from bochsrc.txt." This was identical to choosing the next option, "Read options from..." and only saved you one keystroke, so I removed it.
2001-06-11 10:35:18 +04:00
return;
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
- converted ne2k and sb16 to new style parameters and control panel, except for MAC address of the ne2k, which is next.
2001-06-21 22:34:50 +04:00
bx_options.ne2k.Oethdev->set (strdup(&params[5][7]));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
}
- converted ne2k and sb16 to new style parameters and control panel, except for MAC address of the ne2k, which is next.
2001-06-21 22:34:50 +04:00
bx_options.ne2k.Ovalid->set (1);
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
else if (!strcmp(params[0], "load32bitOSImage")) {
if ( (num_params!=4) && (num_params!=5) ) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: load32bitOSImage directive: wrong # args.", context));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
if (strncmp(params[1], "os=", 3)) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: load32bitOSImage: directive malformed.", context));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
if (!strcmp(&params[1][3], "nullkernel")) {
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.load32bitOSImage.OwhichOS->set (Load32bitOSNullKernel);
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
else if (!strcmp(&params[1][3], "linux")) {
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.load32bitOSImage.OwhichOS->set (Load32bitOSLinux);
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
else {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: load32bitOSImage: unsupported OS.", context));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
if (strncmp(params[2], "path=", 5)) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: load32bitOSImage: directive malformed.", context));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
if (strncmp(params[3], "iolog=", 6)) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: load32bitOSImage: directive malformed.", context));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.load32bitOSImage.Opath->set (strdup(&params[2][5]));
bx_options.load32bitOSImage.Oiolog->set (strdup(&params[3][6]));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
if (num_params == 5) {
if (strncmp(params[4], "initrd=", 7)) {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(("%s: load32bitOSImage: directive malformed.", context));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
bx_options.load32bitOSImage.Oinitrd->set (strdup(&params[4][7]));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
}
. Added a keyboard_type option in configuration file . Modified answer of keyboard controler to a keyboard-identify command depending on keyboard_type . Added the keyboard_type option in bochsrc man page
2001-12-12 13:43:36 +03:00
else if (!strcmp(params[0], "keyboard_type")) {
if (num_params != 2) {
BX_PANIC(("%s: keyboard_type directive: wrong # args.", context));
}
if(strcmp(params[1],"xt")==0){
bx_options.Okeyboard_type->set (BX_KBD_XT_TYPE);
}
else if(strcmp(params[1],"at")==0){
bx_options.Okeyboard_type->set (BX_KBD_AT_TYPE);
}
else if(strcmp(params[1],"mf")==0){
bx_options.Okeyboard_type->set (BX_KBD_MF_TYPE);
}
else{
BX_PANIC(("%s: keyboard_type directive: wrong arg %s.", context,params[1]));
}
}
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
- Added XFree86 keyboard mapping (us, fr, de)
2001-12-14 20:56:37 +03:00
else if (!strcmp(params[0], "keyboard_mapping")
||!strcmp(params[0], "keyboardmapping")) {
for (i=1; i<num_params; i++) {
if (!strncmp(params[i], "enabled=", 8)) {
bx_options.keyboard.OuseMapping->set (atol(&params[i][8]));
}
else if (!strncmp(params[i], "map=", 4)) {
bx_options.keyboard.Okeymap->set (strdup(&params[i][4]));
}
}
}
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
else {
- more work on text-mode control panel. Now the control panel starts before the simulation begins, to give you a chance to choose between bochsrcs, choose the boot disk, etc.
2001-06-10 00:01:12 +04:00
BX_PANIC(( "%s: directive '%s' not understood", context, params[0]));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
- diskd implemented now. Only panic if diskd and cdrom are used together, since at present there's no way to actually manage this. To use both, we will need to put the cdrom (or diskd) onto second ide controller with a different IRQ.
2001-05-04 00:59:38 +04:00
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
if (bx_options.diskd.Opresent->get () && bx_options.cdromd.Opresent->get ())
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(("At present, using both diskd and cdromd at once is not supported."));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
- implemented bx_write_configuration which, for the first time ever, writes a bochsrc for you. - since there were two options related to logging, I moved them both into a new struct called bx_log_options. This follows the pattern used by other devices. - in control.cc: removed option 1 from main menu, the one that said "Read options from bochsrc.txt." This was identical to choosing the next option, "Read options from..." and only saved you one keystroke, so I removed it.
2001-06-11 10:35:18 +04:00
static char *fdtypes[] = {
"none", "1_2", "1_44", "2_88", "720k"
};
int
bx_write_floppy_options (FILE *fp, int drive, bx_floppy_options *opt)
{
BX_ASSERT (drive==0 || drive==1);
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
if (opt->Otype->get () == BX_FLOPPY_NONE) {
- implemented bx_write_configuration which, for the first time ever, writes a bochsrc for you. - since there were two options related to logging, I moved them both into a new struct called bx_log_options. This follows the pattern used by other devices. - in control.cc: removed option 1 from main menu, the one that said "Read options from bochsrc.txt." This was identical to choosing the next option, "Read options from..." and only saved you one keystroke, so I removed it.
2001-06-11 10:35:18 +04:00
fprintf (fp, "# no floppy%c\n", (char)'a'+drive);
return 0;
}
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
BX_ASSERT (opt->Otype->get () > BX_FLOPPY_NONE && opt->Otype->get () <= BX_FLOPPY_LAST);
fprintf (fp, "floppy%c: %s=\"%s\", status=%s\n",
- implemented bx_write_configuration which, for the first time ever, writes a bochsrc for you. - since there were two options related to logging, I moved them both into a new struct called bx_log_options. This follows the pattern used by other devices. - in control.cc: removed option 1 from main menu, the one that said "Read options from bochsrc.txt." This was identical to choosing the next option, "Read options from..." and only saved you one keystroke, so I removed it.
2001-06-11 10:35:18 +04:00
(char)'a'+drive,
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
fdtypes[opt->Otype->get () - BX_FLOPPY_NONE],
opt->Opath->getptr (),
opt->Oinitial_status->get ()==BX_EJECTED ? "ejected" : "inserted");
- implemented bx_write_configuration which, for the first time ever, writes a bochsrc for you. - since there were two options related to logging, I moved them both into a new struct called bx_log_options. This follows the pattern used by other devices. - in control.cc: removed option 1 from main menu, the one that said "Read options from bochsrc.txt." This was identical to choosing the next option, "Read options from..." and only saved you one keystroke, so I removed it.
2001-06-11 10:35:18 +04:00
return 0;
}
int
bx_write_disk_options (FILE *fp, int drive, bx_disk_options *opt)
{
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
if (!opt->Opresent->get ()) {
- implemented bx_write_configuration which, for the first time ever, writes a bochsrc for you. - since there were two options related to logging, I moved them both into a new struct called bx_log_options. This follows the pattern used by other devices. - in control.cc: removed option 1 from main menu, the one that said "Read options from bochsrc.txt." This was identical to choosing the next option, "Read options from..." and only saved you one keystroke, so I removed it.
2001-06-11 10:35:18 +04:00
fprintf (fp, "# no disk%c\n", (char)'c'+drive);
return 0;
}
fprintf (fp, "disk%c: file=\"%s\", cyl=%d, heads=%d, spt=%d\n",
(char)'c'+drive,
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
opt->Opath->getptr (),
opt->Ocylinders->get (),
opt->Oheads->get (),
opt->Ospt->get ());
- implemented bx_write_configuration which, for the first time ever, writes a bochsrc for you. - since there were two options related to logging, I moved them both into a new struct called bx_log_options. This follows the pattern used by other devices. - in control.cc: removed option 1 from main menu, the one that said "Read options from bochsrc.txt." This was identical to choosing the next option, "Read options from..." and only saved you one keystroke, so I removed it.
2001-06-11 10:35:18 +04:00
return 0;
}
int
bx_write_cdrom_options (FILE *fp, int drive, bx_cdrom_options *opt)
{
BX_ASSERT (drive == 0);
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
if (!opt->Opresent->get ()) {
- implemented bx_write_configuration which, for the first time ever, writes a bochsrc for you. - since there were two options related to logging, I moved them both into a new struct called bx_log_options. This follows the pattern used by other devices. - in control.cc: removed option 1 from main menu, the one that said "Read options from bochsrc.txt." This was identical to choosing the next option, "Read options from..." and only saved you one keystroke, so I removed it.
2001-06-11 10:35:18 +04:00
fprintf (fp, "# no cdromd\n");
return 0;
}
- cdrom and floppy options in configuration interface partially fixed * don't change the device state to "not present" at runtime if the path is changed to "none" - simply eject the media if inserted * don't change the device state to "present" at runtime if the previous state is "not present" * eject and re-insert the media if the current media state is "inserted" * changing the media state in the config interface works now * TODO: don't do anything if the path is unchanged - bx_write_cdrom_options(): writing media state fixed
2002-03-14 23:14:47 +03:00
fprintf (fp, "cdromd: dev=%s, status=%s\n",
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
opt->Opath->getptr (),
- cdrom and floppy options in configuration interface partially fixed * don't change the device state to "not present" at runtime if the path is changed to "none" - simply eject the media if inserted * don't change the device state to "present" at runtime if the previous state is "not present" * eject and re-insert the media if the current media state is "inserted" * changing the media state in the config interface works now * TODO: don't do anything if the path is unchanged - bx_write_cdrom_options(): writing media state fixed
2002-03-14 23:14:47 +03:00
opt->Oinserted->get ()==BX_INSERTED ? "inserted" : "ejected");
- implemented bx_write_configuration which, for the first time ever, writes a bochsrc for you. - since there were two options related to logging, I moved them both into a new struct called bx_log_options. This follows the pattern used by other devices. - in control.cc: removed option 1 from main menu, the one that said "Read options from bochsrc.txt." This was identical to choosing the next option, "Read options from..." and only saved you one keystroke, so I removed it.
2001-06-11 10:35:18 +04:00
return 0;
}
- allow configuration of the parallel port in .bochsrc and the config interface menus. Parallel port #1 is implemented, and I left stubs for parallel port #2 in case we want to ever add it. If the parallel port is enabled, the init method of parallel.cc does an fopen() on the output file. If disabled, or if the fopen fails, the file handler remains NULL and no characters are printed. There is no attempt to enable/disable the operation of the parallel port, only the output to a file.
2001-11-12 05:35:09 +03:00
int
bx_write_parport_options (FILE *fp, bx_parport_options *opt, int n)
{
if (!opt->Oenable->get ()) {
fprintf (fp, "# no parport #%d\n", n);
return 0;
}
fprintf (fp, "parport%d: enable=%d", n, opt->Oenable->get ());
if (opt->Oenable->get ()) {
fprintf (fp, ", file=\"%s\"", opt->Ooutfile->getptr ());
}
fprintf (fp, "\n");
return 0;
}
- implemented bx_write_configuration which, for the first time ever, writes a bochsrc for you. - since there were two options related to logging, I moved them both into a new struct called bx_log_options. This follows the pattern used by other devices. - in control.cc: removed option 1 from main menu, the one that said "Read options from bochsrc.txt." This was identical to choosing the next option, "Read options from..." and only saved you one keystroke, so I removed it.
2001-06-11 10:35:18 +04:00
int
bx_write_sb16_options (FILE *fp, bx_sb16_options *opt)
{
- converted ne2k and sb16 to new style parameters and control panel, except for MAC address of the ne2k, which is next.
2001-06-21 22:34:50 +04:00
if (!opt->Opresent->get ()) {
- implemented bx_write_configuration which, for the first time ever, writes a bochsrc for you. - since there were two options related to logging, I moved them both into a new struct called bx_log_options. This follows the pattern used by other devices. - in control.cc: removed option 1 from main menu, the one that said "Read options from bochsrc.txt." This was identical to choosing the next option, "Read options from..." and only saved you one keystroke, so I removed it.
2001-06-11 10:35:18 +04:00
fprintf (fp, "# no sb16\n");
return 0;
}
- converted ne2k and sb16 to new style parameters and control panel, except for MAC address of the ne2k, which is next.
2001-06-21 22:34:50 +04:00
fprintf (fp, "sb16: midimode=%d, midi=%s, wavemode=%d, wave=%s, loglevel=%d, log=%s, dmatimer=%d\n", opt->Omidimode->get (), opt->Omidifile->getptr (), opt->Owavemode->get (), opt->Owavefile->getptr (), opt->Ologlevel->get (), opt->Ologfile->getptr (), opt->Odmatimer->get ());
- implemented bx_write_configuration which, for the first time ever, writes a bochsrc for you. - since there were two options related to logging, I moved them both into a new struct called bx_log_options. This follows the pattern used by other devices. - in control.cc: removed option 1 from main menu, the one that said "Read options from bochsrc.txt." This was identical to choosing the next option, "Read options from..." and only saved you one keystroke, so I removed it.
2001-06-11 10:35:18 +04:00
return 0;
}
int
bx_write_ne2k_options (FILE *fp, bx_ne2k_options *opt)
{
- converted ne2k and sb16 to new style parameters and control panel, except for MAC address of the ne2k, which is next.
2001-06-21 22:34:50 +04:00
if (!opt->Ovalid->get ()) {
- implemented bx_write_configuration which, for the first time ever, writes a bochsrc for you. - since there were two options related to logging, I moved them both into a new struct called bx_log_options. This follows the pattern used by other devices. - in control.cc: removed option 1 from main menu, the one that said "Read options from bochsrc.txt." This was identical to choosing the next option, "Read options from..." and only saved you one keystroke, so I removed it.
2001-06-11 10:35:18 +04:00
fprintf (fp, "# no ne2k\n");
return 0;
}
- add options to the string parameter class that allow it to deal with raw hex bytes, for the MAC address.
2001-06-21 23:27:05 +04:00
char *ptr = opt->Omacaddr->getptr ();
- implemented bx_write_configuration which, for the first time ever, writes a bochsrc for you. - since there were two options related to logging, I moved them both into a new struct called bx_log_options. This follows the pattern used by other devices. - in control.cc: removed option 1 from main menu, the one that said "Read options from bochsrc.txt." This was identical to choosing the next option, "Read options from..." and only saved you one keystroke, so I removed it.
2001-06-11 10:35:18 +04:00
fprintf (fp, "ne2k: ioaddr=0x%x, irq=%d, mac=%02x:%02x:%02x:%02x:%02x:%02x, ethmod=%s, ethdev=%s\n",
- converted ne2k and sb16 to new style parameters and control panel, except for MAC address of the ne2k, which is next.
2001-06-21 22:34:50 +04:00
opt->Oioaddr->get (),
opt->Oirq->get (),
- fixed: mac addresses used to be printed as an 8-digit hex number. Now I AND each byte with 0xff before printing.
2001-06-25 11:12:45 +04:00
(unsigned int)(0xff & ptr[0]),
(unsigned int)(0xff & ptr[1]),
(unsigned int)(0xff & ptr[2]),
(unsigned int)(0xff & ptr[3]),
(unsigned int)(0xff & ptr[4]),
(unsigned int)(0xff & ptr[5]),
- converted ne2k and sb16 to new style parameters and control panel, except for MAC address of the ne2k, which is next.
2001-06-21 22:34:50 +04:00
opt->Oethmod->getptr (),
opt->Oethdev->getptr ());
- implemented bx_write_configuration which, for the first time ever, writes a bochsrc for you. - since there were two options related to logging, I moved them both into a new struct called bx_log_options. This follows the pattern used by other devices. - in control.cc: removed option 1 from main menu, the one that said "Read options from bochsrc.txt." This was identical to choosing the next option, "Read options from..." and only saved you one keystroke, so I removed it.
2001-06-11 10:35:18 +04:00
return 0;
}
int
bx_write_loader_options (FILE *fp, bx_load32bitOSImage_t *opt)
{
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
if (opt->OwhichOS->get () == 0) {
- implemented bx_write_configuration which, for the first time ever, writes a bochsrc for you. - since there were two options related to logging, I moved them both into a new struct called bx_log_options. This follows the pattern used by other devices. - in control.cc: removed option 1 from main menu, the one that said "Read options from bochsrc.txt." This was identical to choosing the next option, "Read options from..." and only saved you one keystroke, so I removed it.
2001-06-11 10:35:18 +04:00
fprintf (fp, "# no loader\n");
return 0;
}
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
BX_ASSERT(opt->OwhichOS->get () == Load32bitOSLinux || opt->OwhichOS->get () == Load32bitOSNullKernel);
- implemented bx_write_configuration which, for the first time ever, writes a bochsrc for you. - since there were two options related to logging, I moved them both into a new struct called bx_log_options. This follows the pattern used by other devices. - in control.cc: removed option 1 from main menu, the one that said "Read options from bochsrc.txt." This was identical to choosing the next option, "Read options from..." and only saved you one keystroke, so I removed it.
2001-06-11 10:35:18 +04:00
fprintf (fp, "load32bitOSImage: os=%s, path=%s, iolog=%s, initrd=%s\n",
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
(opt->OwhichOS->get () == Load32bitOSLinux) ? "linux" : "nullkernel",
opt->Opath->getptr (),
opt->Oiolog->getptr (),
opt->Oinitrd->getptr ());
- implemented bx_write_configuration which, for the first time ever, writes a bochsrc for you. - since there were two options related to logging, I moved them both into a new struct called bx_log_options. This follows the pattern used by other devices. - in control.cc: removed option 1 from main menu, the one that said "Read options from bochsrc.txt." This was identical to choosing the next option, "Read options from..." and only saved you one keystroke, so I removed it.
2001-06-11 10:35:18 +04:00
return 0;
}
int
bx_write_log_options (FILE *fp, bx_log_options *opt)
{
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
fprintf (fp, "log: %s\n", opt->Ofilename->getptr ());
- implemented bx_write_configuration which, for the first time ever, writes a bochsrc for you. - since there were two options related to logging, I moved them both into a new struct called bx_log_options. This follows the pattern used by other devices. - in control.cc: removed option 1 from main menu, the one that said "Read options from bochsrc.txt." This was identical to choosing the next option, "Read options from..." and only saved you one keystroke, so I removed it.
2001-06-11 10:35:18 +04:00
// no syntax to describe all the possible action settings for every
// device. Instead, take a vote and record the most popular action
// for each level of event.
int action_tally[N_ACT];
int most_popular_action[N_LOGLEV];
- clean up compiler warnings
2001-06-12 00:39:05 +04:00
int i,lev;
- implemented bx_write_configuration which, for the first time ever, writes a bochsrc for you. - since there were two options related to logging, I moved them both into a new struct called bx_log_options. This follows the pattern used by other devices. - in control.cc: removed option 1 from main menu, the one that said "Read options from bochsrc.txt." This was identical to choosing the next option, "Read options from..." and only saved you one keystroke, so I removed it.
2001-06-11 10:35:18 +04:00
for (lev = 0; lev < N_LOGLEV; lev++) {
// clear tally
for (i=0; i<N_ACT; i++)
action_tally[i] = 0;
// count how many devices use each action
for (i=0; i<io->get_n_logfns (); i++) {
logfunc_t *fn = io->get_logfn (i);
int action = fn->getonoff(lev);
BX_ASSERT (action >= 0 && action < N_ACT);
action_tally[action]++;
}
// count the votes
int best = 0, best_votes = action_tally[0];
for (i=1; i<N_ACT; i++) {
if (action_tally[i] > best_votes) {
best = i;
best_votes = action_tally[i];
}
}
most_popular_action[lev] = best;
}
fprintf (fp, "panic: action=%s\n",
io->getaction(most_popular_action[LOGLEV_PANIC]));
fprintf (fp, "error: action=%s\n",
io->getaction(most_popular_action[LOGLEV_ERROR]));
fprintf (fp, "info: action=%s\n",
io->getaction(most_popular_action[LOGLEV_INFO]));
fprintf (fp, "debug: action=%s\n",
io->getaction(most_popular_action[LOGLEV_DEBUG]));
- clean up compiler warnings
2001-06-12 00:39:05 +04:00
return 0;
- implemented bx_write_configuration which, for the first time ever, writes a bochsrc for you. - since there were two options related to logging, I moved them both into a new struct called bx_log_options. This follows the pattern used by other devices. - in control.cc: removed option 1 from main menu, the one that said "Read options from bochsrc.txt." This was identical to choosing the next option, "Read options from..." and only saved you one keystroke, so I removed it.
2001-06-11 10:35:18 +04:00
}
- Added XFree86 keyboard mapping (us, fr, de)
2001-12-14 20:56:37 +03:00
int
bx_write_keyboard_options (FILE *fp, bx_keyboard_options *opt)
{
fprintf (fp, "keyboard_mapping: enabled=%d, map=%s\n", opt->OuseMapping->get(), opt->Okeymap->getptr());
return 0;
}
- implemented bx_write_configuration which, for the first time ever, writes a bochsrc for you. - since there were two options related to logging, I moved them both into a new struct called bx_log_options. This follows the pattern used by other devices. - in control.cc: removed option 1 from main menu, the one that said "Read options from bochsrc.txt." This was identical to choosing the next option, "Read options from..." and only saved you one keystroke, so I removed it.
2001-06-11 10:35:18 +04:00
// return values:
// 0: written ok
// -1: failed
// -2: already exists, and overwrite was off
int
bx_write_configuration (char *rc, int overwrite)
{
BX_INFO (("write configuration to %s\n", rc));
// check if it exists. If so, only proceed if overwrite is set.
FILE *fp = fopen (rc, "r");
if (fp != NULL) {
fclose (fp);
if (!overwrite) return -2;
}
fp = fopen (rc, "w");
if (fp == NULL) return -1;
// finally it's open and we can start writing.
fprintf (fp, "# configuration file generated by Bochs\n");
// it would be nice to put this type of function as methods on
// the structs like bx_floppy_options::print or something.
bx_write_floppy_options (fp, 0, &bx_options.floppya);
bx_write_floppy_options (fp, 1, &bx_options.floppyb);
bx_write_disk_options (fp, 0, &bx_options.diskc);
bx_write_disk_options (fp, 1, &bx_options.diskd);
bx_write_cdrom_options (fp, 0, &bx_options.cdromd);
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
if (strlen (bx_options.rom.Opath->getptr ()) > 0)
fprintf (fp, "romimage: file=%s, address=0x%05x\n", bx_options.rom.Opath->getptr(), (unsigned int)bx_options.rom.Oaddress->get ());
- implemented bx_write_configuration which, for the first time ever, writes a bochsrc for you. - since there were two options related to logging, I moved them both into a new struct called bx_log_options. This follows the pattern used by other devices. - in control.cc: removed option 1 from main menu, the one that said "Read options from bochsrc.txt." This was identical to choosing the next option, "Read options from..." and only saved you one keystroke, so I removed it.
2001-06-11 10:35:18 +04:00
else
fprintf (fp, "# no romimage\n");
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
if (strlen (bx_options.vgarom.Opath->getptr ()) > 0)
fprintf (fp, "vgaromimage: %s\n", bx_options.vgarom.Opath->getptr ());
- implemented bx_write_configuration which, for the first time ever, writes a bochsrc for you. - since there were two options related to logging, I moved them both into a new struct called bx_log_options. This follows the pattern used by other devices. - in control.cc: removed option 1 from main menu, the one that said "Read options from bochsrc.txt." This was identical to choosing the next option, "Read options from..." and only saved you one keystroke, so I removed it.
2001-06-11 10:35:18 +04:00
else
fprintf (fp, "# no vgaromimage\n");
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
fprintf (fp, "megs: %d\n", bx_options.memory.Osize->get ());
- allow configuration of the parallel port in .bochsrc and the config interface menus. Parallel port #1 is implemented, and I left stubs for parallel port #2 in case we want to ever add it. If the parallel port is enabled, the init method of parallel.cc does an fopen() on the output file. If disabled, or if the fopen fails, the file handler remains NULL and no characters are printed. There is no attempt to enable/disable the operation of the parallel port, only the output to a file.
2001-11-12 05:35:09 +03:00
bx_write_parport_options (fp, &bx_options.par1, 1);
//bx_write_parport_options (fp, &bx_options.par2);
- implemented bx_write_configuration which, for the first time ever, writes a bochsrc for you. - since there were two options related to logging, I moved them both into a new struct called bx_log_options. This follows the pattern used by other devices. - in control.cc: removed option 1 from main menu, the one that said "Read options from bochsrc.txt." This was identical to choosing the next option, "Read options from..." and only saved you one keystroke, so I removed it.
2001-06-11 10:35:18 +04:00
bx_write_sb16_options (fp, &bx_options.sb16);
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
int bootdrive = bx_options.Obootdrive->get ();
- fix for boot drive option incorrectly saved to configuration file
2002-03-07 02:31:02 +03:00
fprintf (fp, "boot: %s\n", (bootdrive==BX_BOOT_FLOPPYA) ? "a" : (bootdrive==BX_BOOT_DISKC) ? "c" : "cdrom");
- tweaks to get rid of compile warnings
2001-09-11 20:56:48 +04:00
fprintf (fp, "vga_update_interval: %u\n", bx_options.Ovga_update_interval->get ());
fprintf (fp, "keyboard_serial_delay: %u\n", bx_options.Okeyboard_serial_delay->get ());
- apply Christophe's patch.paste-for-dos. I will make minor tweaks as separate revisions.
2002-03-26 16:51:48 +03:00
fprintf (fp, "keyboard_paste_delay: %u\n", bx_options.Okeyboard_paste_delay->get ());
- tweaks to get rid of compile warnings
2001-09-11 20:56:48 +04:00
fprintf (fp, "floppy_command_delay: %u\n", bx_options.Ofloppy_command_delay->get ());
- correct problem in the ips line when saving a bochsrc, which I clumsily added in 1.71 or so.
2001-10-08 06:02:47 +04:00
fprintf (fp, "ips: %u\n", bx_options.Oips->get ());
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
fprintf (fp, "mouse: enabled=%d\n", bx_options.Omouse_enabled->get ());
fprintf (fp, "private_colormap: enabled=%d\n", bx_options.Oprivate_colormap->get ());
ok, now the amiga target on cvs is complete
2001-08-16 16:35:52 +04:00
#if BX_WITH_AMIGAOS
- add string parameter called screenmode - put #if BX_USE_AMIGAOS...#endif around fullscreen and screenmode, since they only apply to amiga
2001-08-16 06:00:31 +04:00
fprintf (fp, "fullscreen: enabled=%d\n", bx_options.Ofullscreen->get ());
fprintf (fp, "screenmode: name=\"%s\"\n", bx_options.Oscreenmode->getptr ());
#endif
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
fprintf (fp, "i440fxsupport: enabled=%d\n", bx_options.Oi440FXSupport->get ());
fprintf (fp, "time0: %u\n", bx_options.cmos.Otime0->get ());
- implemented bx_write_configuration which, for the first time ever, writes a bochsrc for you. - since there were two options related to logging, I moved them both into a new struct called bx_log_options. This follows the pattern used by other devices. - in control.cc: removed option 1 from main menu, the one that said "Read options from bochsrc.txt." This was identical to choosing the next option, "Read options from..." and only saved you one keystroke, so I removed it.
2001-06-11 10:35:18 +04:00
bx_write_ne2k_options (fp, &bx_options.ne2k);
- convert nearly all bx_options to parameter form. All options that have been converted into parameters temporarily have the letter "O" appended to their name. I don't want to keep it this way, but it has helped in the conversion process because the compiler refuses to compile the old uses of the name. Before I started using the "O" trick, there were many bugs like this: if (bx_options.diskc.present) {...} This was legal with the new parameters, but it was testing whether the parameter structure had been created, instead of testing the value of the present parameter. Renaming present to Opresent turns this into a compile error, which points out the incorrect use of the param. - the "--disable-control-panel" no longer works, I'm afraid. I can no longer support this and continue progress.
2001-06-20 18:01:39 +04:00
fprintf (fp, "newharddrivesupport: enabled=%d\n", bx_options.OnewHardDriveSupport->get ());
- implemented bx_write_configuration which, for the first time ever, writes a bochsrc for you. - since there were two options related to logging, I moved them both into a new struct called bx_log_options. This follows the pattern used by other devices. - in control.cc: removed option 1 from main menu, the one that said "Read options from bochsrc.txt." This was identical to choosing the next option, "Read options from..." and only saved you one keystroke, so I removed it.
2001-06-11 10:35:18 +04:00
bx_write_loader_options (fp, &bx_options.load32bitOSImage);
bx_write_log_options (fp, &bx_options.log);
- Added XFree86 keyboard mapping (us, fr, de)
2001-12-14 20:56:37 +03:00
bx_write_keyboard_options (fp, &bx_options.keyboard);
. Added a keyboard_type option in configuration file . Modified answer of keyboard controler to a keyboard-identify command depending on keyboard_type . Added the keyboard_type option in bochsrc man page
2001-12-12 13:43:36 +03:00
fprintf (fp, "keyboard_type: %s\n", bx_options.Okeyboard_type->get ()==BX_KBD_XT_TYPE?"xt":
bx_options.Okeyboard_type->get ()==BX_KBD_AT_TYPE?"at":"mf");
- implemented bx_write_configuration which, for the first time ever, writes a bochsrc for you. - since there were two options related to logging, I moved them both into a new struct called bx_log_options. This follows the pattern used by other devices. - in control.cc: removed option 1 from main menu, the one that said "Read options from bochsrc.txt." This was identical to choosing the next option, "Read options from..." and only saved you one keystroke, so I removed it.
2001-06-11 10:35:18 +04:00
fclose (fp);
return 0;
}
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
#endif // #if BX_PROVIDE_MAIN
void
bx_signal_handler( int signum)
{
- give the GUI signal handler first priority (before, BX_SHOW_IPS took precedence). This allows GUIs which want to take over SIGALRM to do so. Presently, no GUI uses this.
2001-05-09 00:18:05 +04:00
#if BX_GUI_SIGHANDLER
// GUI signal handler gets first priority, if the mask says it's wanted
if ((1<<signum) & bx_gui.get_sighandler_mask ()) {
bx_gui.sighandler (signum);
return;
}
#endif
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
#if BX_SHOW_IPS
extern unsigned long ips_count;
if (signum == SIGALRM ) {
strip '\n' from BX_{INFO,DEBUG,ERROR,PANIC} don't need it, moved the output of it into the general io functions. saves space, as well as removes the confusing output if a '\n' is left off
2001-05-30 22:56:02 +04:00
BX_INFO(("ips = %lu", ips_count));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
ips_count = 0;
#ifndef __MINGW32__
signal(SIGALRM, bx_signal_handler);
alarm( 1 );
#endif
return;
}
#endif
- merged BRANCH-smp-bochs into main branch. For details see comments in BRANCH-smp-bochs revisions. - The general task was to make multiple CPU's which communicate through their APICs. So instead of BX_CPU and BX_MEM, we now have BX_CPU(x) and BX_MEM(y). For an SMP simulation you have several processors in a shared memory space, so there might be processors BX_CPU(0..3) but only one memory space BX_MEM(0). For cosimulation, you could have BX_CPU(0) with BX_MEM(0), then BX_CPU(1) with BX_MEM(1). WARNING: Cosimulation is almost certainly broken by the SMP changes. - to simulate multiple CPUs, you have to give each CPU time to execute in turn. This is currently implemented using debugger guards. The cpu loop steps one CPU for a few instructions, then steps the next CPU for a few instructions, etc. - there is some limited support in the debugger for two CPUs, for example printing information from each CPU when single stepping.
2001-05-23 12:16:07 +04:00
#if BX_GUI_SIGHANDLER
if ((1<<signum) & bx_gui.get_sighandler_mask ()) {
bx_gui.sighandler (signum);
return;
}
#endif
strip '\n' from BX_{INFO,DEBUG,ERROR,PANIC} don't need it, moved the output of it into the general io functions. saves space, as well as removes the confusing output if a '\n' is left off
2001-05-30 22:56:02 +04:00
BX_PANIC(("SIGNAL %u caught", signum));
- entered original Bochs snapshot bochs-2000_0325a.tar.gz from ftp.bochs.com
2001-04-10 05:04:59 +04:00
}
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
Reference in New Issue Copy Permalink