mirrors/Bochs
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
6,119 Commits 11 Branches 185 Tags 48 MiB
9a7930a23f
Commit Graph

1256 Commits

Author SHA1 Message Date
Volker Ruppert
9a7930a23f - destructor now deletes the device buffer 
- attempt to implement the "get max. LUN" command for mass storage devices
- removed unnecessary spaces
 2007-03-05 18:09:57 +00:00
Volker Ruppert
0be7d6e5e3 - when clearing the run/stop bit, the "host halted" bit in the status register 
must be set
- ignore unused bits on i/o writes
- fixed two warnings
 2007-03-01 18:29:36 +00:00
Volker Ruppert
a3b352d4fd - don't abort command when setting maximum logical head number to 0 (Linux 2.6.x 
kernels use this value when LBA48 support is available)
 2007-02-27 18:16:20 +00:00
Volker Ruppert
8beb290f67 - NACK for invalid PS/2 mouse command must be added to the mouse queue 2007-02-25 13:18:33 +00:00
Volker Ruppert
252946be66 - initial ACPI SMBus support added (ported from Qemu, TODO: register devices, 
bus transactions)
 2007-02-20 09:39:06 +00:00
Volker Ruppert
4a8996153f - partial support for non-DMA mode (initial patch #1641816 by John Comeau) 2007-02-09 14:23:50 +00:00
Stanislav Shwartsman
2f9e1bee68 fixed code duplication and comments 2007-02-03 17:56:35 +00:00
Volker Ruppert
ef9c6ff8b1 - CMOS RTC index port read message type changed to BX_DEBUG 2007-01-02 16:37:11 +00:00
Volker Ruppert
d7c627fdc1 - added support for serial 3-button mouse with Mousesystems protocol 2006-12-31 11:56:14 +00:00
Volker Ruppert
795d07e90f - hard drive id now reports UDMA supported and ATA version 6 
- setting transfer mode to UDMA now supported (no effect on the emulation speed)
- drive id arrays now initialized with memset()
 2006-12-29 11:57:04 +00:00
Volker Ruppert
7213eae1bb - fixed some gcc 4.x warnings 2006-12-27 15:21:03 +00:00
Volker Ruppert
ee6bd95cfc - ATA/ATAPI identify device methods must now be called only one time per device in 
a Bochs session unless the configuration (transfer mode) changes. Then the
  "drive id" data is simply copied to the output buffer.
- removed unused "drive id" example (Conner CFA540A)
- added stubs for the ATA flush cache commands
- converted some tabs to spaces
 2006-12-26 17:08:12 +00:00
Volker Ruppert
fa43431365 - LBA48 support implemented (tested with Linux - TODO: readback of hob* values) 2006-12-25 09:34:32 +00:00
Volker Ruppert
a6d772caa7 - fixed several MSVC warnings 2006-12-19 16:42:27 +00:00
Volker Ruppert
f1799211ae - identify_drive() returns max. 16383 cylinders (higher values are usually 
calculated using the total number of sectors, heads and sectors per track)
 2006-12-18 17:56:05 +00:00
Volker Ruppert
9c8b9eae6a - support for VMWare version 4 disk image format implemented (patch by Sharvil 
Nanavati)
- siminterface + wx code cleanup (removed unused stuff and wx log messages,
  floppy media type name handling simplified)
 2006-12-17 08:17:28 +00:00
Volker Ruppert
d363ceef3b - the ATA command "verify without retries" works the same was as the normal 
verify command (should fix SVR4/386 installation)
 2006-12-13 16:59:29 +00:00
Alexander Krisak
5f64f48e51 Additional validation of theCmosDevice is added 2006-12-06 14:27:02 +00:00
Volker Ruppert
4314b6a911 - networking modules eth_tap and eth_tuntap fixed for OpenBSD 
- removed unnecessary spaces
 2006-11-23 17:21:59 +00:00
Volker Ruppert
307ca616a2 - enable BPF immediate mode (patch from SF tracker item #1597451) 2006-11-19 16:16:35 +00:00
Volker Ruppert
bf0c76ee75 - fixed remaining type conflicts in vmware lseek() method and the bxcommit utility 
- the bximage and bxcommit utilities depend on hdimage.h
- removed unnecessary spaces
 2006-11-19 09:55:23 +00:00
Volker Ruppert
0a174516ff - fixed type conflicts in hdimage lseek() methods. Now using always Bit64s for 
the lseek() methods and off_t is only used for the lowlevel lseek() function.
- fixed indent mode and removed unnecessary spaces
 2006-11-18 11:51:07 +00:00
Volker Ruppert
3dd9df62c7 - don't modify display library options parameter when splitting it into a list 
of arguments
 2006-11-12 18:26:58 +00:00
Volker Ruppert
e1ed0cd384 - register iodebug ports for word access only and simplify code 2006-11-01 15:44:38 +00:00
Volker Ruppert
f3bee13cad - Bochs on Solaris always needs to be linked with the socket library 
- serial terminal support now enabled on Solaris
- only complain about the missing wxWidgets library if requested
 2006-10-12 17:58:09 +00:00
Volker Ruppert
0b2331c1fe - sequencer register 0x07 must be initialized with 0 if cirrus extension is not 
selected
- cirrus specific mem write methods don't need to be virtual
 2006-10-08 15:21:26 +00:00
Volker Ruppert
f8f01db46a - replaced call of obsolete bcopy() by memcpy() 
- the cfmakeraw() function is not available on all platforms (now using code
  that does the same job)
- removed useless strcpy() call
 2006-10-08 10:59:21 +00:00
Volker Ruppert
c9c01954b9 - PM/SM i/o space enable control bits implemented 
- ACPI debug port added
 2006-10-07 08:59:15 +00:00
Stanislav Shwartsman
1aa7ae677d Small cleanup in he APIC code, removed unused methods 2006-10-02 21:49:49 +00:00
Volker Ruppert
8dfd8eefa7 - ExtINT implemented (some progress with FreeBSD 5.3 boot cd) 
- fixed a warning
 2006-10-02 15:08:21 +00:00
Volker Ruppert
815f2a026d - fixed compilation with ACPI disabled 2006-10-02 10:14:07 +00:00
Volker Ruppert
5d1e848a6a - SMI generation implemented (needs to be checked for SMP) 2006-10-01 19:51:49 +00:00
Volker Ruppert
0223ffdc6e - ported most of the ACPI controller capabilities from Qemu (TODO: generate SMI 
on APM control port write)
 2006-10-01 13:47:26 +00:00
Volker Ruppert
fb762477fa - some more basic work on the acpi device 2006-09-26 18:43:42 +00:00
Stanislav Shwartsman
aab3feb2e0 Fixed Makefile to compile with ACPI stub disabled 
Remove bx_debug.h from bochs.h to reduce compilation dependencies
 2006-09-25 20:13:52 +00:00
Volker Ruppert
2cdd6596f1 - implementation of the PIIX4 ACPI controller started. The code is not yet 
enabled and the i/o handler only generate info messages.
 2006-09-24 16:58:13 +00:00
Volker Ruppert
bf6d61de5e - experimental USB device change support added (some more work in the emulation 
code may be required)
- set mouse/keypad connection status to 0 when removing device
- pciusb code now accepts "none" as the device name for an unused port
 2006-09-24 10:10:21 +00:00
Volker Ruppert
93534890b0 - the USB mouse connection status no longer depends on the mouse capture status. 
The connection status is now only controlled by the USB port options (TODO:
  device change support at runtime)
- a connected USB mouse now has the priority when sending mouse data from the
  gui to the emulated mouse device.
 2006-09-23 12:59:57 +00:00
Volker Ruppert
1f5d311ca1 - fixed memory leaks found with valgrind when trying to start a second 
simulation with wxBochs:
  * cpu: delete pageWriteStampTable array
  * devices: simplified setup of initial default i/o handlers
  * ne2k: delete ethernet module in destructor
 2006-09-20 20:52:23 +00:00
Volker Ruppert
72283f6760 - delete i/o handlers before unloading plugins 
- memory for i/o handler name is now dynamicly allocated (strdup's in harddrv
  code removed)
 2006-09-20 18:24:17 +00:00
Volker Ruppert
26f7f49b9a - virtual timer method 'reset' renamed to 'setup' (it's doing other things than 
the reset methods for devices)
- PCI register dump on exit now done in the destructor
 2006-09-18 18:10:49 +00:00
Volker Ruppert
c592b9527c - deactivate device parameter handlers in destructor 2006-09-17 20:39:37 +00:00
Volker Ruppert
b8c84baddb - new PIT: exit method added to reset the timer id 
- old PIT: exit method stub added and some code cleanup
- virtual timer: reset method added with basic init code from the constructor
 2006-09-17 19:19:15 +00:00
Volker Ruppert
fdefdc5b9f - remove logfunction from list when device is deleted 
- fixed typos in harddrv code
 2006-09-17 18:09:33 +00:00
Volker Ruppert
4c1e82f18c - moved some parts of the timer init from the constructor to initialize() 
- exit() method sets the number of registered timers back to 1 (or 2 if APIC
  is present)
- fixed some parameter types for save/restore
- debug message for timer registration added
- slowdown timer: exit method added to reset the timer id
 2006-09-16 19:30:56 +00:00
Volker Ruppert
647b0e3e6c - removed exit() method from the devmodel object, since all that stuff can be 
done by the object destructor (changes in cmos, hard drive and keyboard)
- bx_unload_plugins() now deletes the devmodel object in non-plugin mode
- CMOS device now prints the time on exit in human-readable format
 2006-09-16 14:47:40 +00:00
Volker Ruppert
38caf3bf19 - allocate memory for the IRQ handler name in register_irq() and free it in 
unregister_irq(). This is required since there are some devices that don't
  use a string constant for the handler name.
- replaced true / false by portable 0 / 1 values.
 2006-09-13 18:51:25 +00:00
Volker Ruppert
4cf2f745d1 - some fixes to avoid segfaults after early panics 
- fixed gameport log type
- wx: store mouse capture mode in the gui code (only access the mouse parameter
  when the capture mode changes)
- wx: don't initialize wxScreen if already done and clear screen on exit
 2006-09-12 13:05:07 +00:00
Volker Ruppert
190b9391cf - delete the device object in the plugin_fini() function of each plugin device 
(TODO: find a way to call plugin_fini() for non-core devices if plugins are
   disabled)
- set the device pointers back to stubs after unloading plugins
- added debug message "Exit" in all device plugin destructors
 2006-09-10 17:18:44 +00:00
Volker Ruppert
c07197758b - plugin unload mechanism implemented in the plugin and devices code 
- added exit() method to the devmodel object for devices that need to do special
  cleanups on exit. Device plugins should call this method from their plugin_fini()
  function. If Bochs is compiled without plugin support, for non-core plugins
  the exit() method is called directly from bx_unload_plugins().
- exit() method implemented in the keyboard device
- wx "Show Keyboard" feature now only appear when the wxdebug keyboard subtree
  is not empty
- cpu cleanup must be done before the devices cleanup
- moved PCI chipset dump call to the devices exit method
 2006-09-10 09:13:47 +00:00