From the author:
I've implemented functions ALOb(), ObAL(), YbAL(),
ALXb() and eAXXv() from dis_groups.cc.
I've also changed dis_decode.cc to print cmpsd if code
is 32 bits and cmpsw if code is 16 bits. The same with
stosd, lodsd and scasd.
appears with the right dimensions. Font sizes other than 16 are working now.
The maximum font height in text mode is 32.
- the text mode cursor is now drawn with the right colors
The 64 bit variant of MOVNTI was not decoded. The proper fix for this is to work on
fetchdecode64.cc to call a 64 bit variant of SSE instructions or fail it with a
invalid op. A careful check needs to be done with the AMD manuals to determine if
there are any other SSE instructions that have a special 64 bit decoding.
into the host-specific files and wrapped access to them
with atomic operations since that's a structure global to
all the VMs. I think all the other globals are SMP clean
since they are only written once during module init time,
and read thereafter my all VMs.
Renamed all host OS specific functions to hostOS*(). All host
independent functions to host*().
I'd like to rename all monitor space functions to mon*() next.
bablokb says in [ 664926 ]
please remove the patch patches/patch.bochs.sh. This
patch was developed for 1.4.x and does not work
anymore with 2.0.x.
It has been superseded by bxtstart in Bochs-Tools (see
http://www.bablokb.de/bochs-tools/).
all the memory it needs, and the plex86 kernel module uses
get_user_pages() from the Linux kernel to get at them and
pin the few that are needed statically (and later up to
a watermark of pages that are needed dynamically).
Guest physical memory pages are now dynamically pinned/unpinned.
For now, I use a hard limit of 4Megs of pinned pages and
a really primitive algorithm to decide which one to unpin
when the limit is reached and one needs to be bumped. Seems
to work. Though I haven't run into the limit yet since I'm using
just a small test program.
and pinning user pages from the plex86 kernel module. The
guest physical memory array is no longer mmap()'d from kernel
space where it used to be allocated.
in the plex86 module and mmap()'ing it into user space (bochs),
to letting bochs malloc() it normally and using the
Linux kernel facility get_user_pages() to get the associated
physical pages and pin them in memory. I only have code for
Linux kernel 2.4.20 and up, as that's the first version to
export the get_user_pages() symbol so modules can use it.
- add int75_handler, for msdos compatbility fpu exception
This handler calls int 2 (nmi). The NMI handler should really be aware
of this and should iret in case of FPU exception (to be implemented)
a region of virtual memory. Now the same one works for getting
the pages of the kernel driver and memory objects allocated via
vmalloc().
Converted to using Linux interfaces to walk the page tables to
get at the physical memory addresses above. The old code was
digging up this info starting with looking at CR3. Linux has
functions/macros to do this, which can handle 2/3-level cases.
Wrapped the page table walk with proper locks. A spin lock
for new Linuxes, a big kernel lock for old ones.
- implementation of the UIP bit using a new timer handler. The one_second_timer()
function only sets the UIP bit and starts the UIP timer. The uip_timer()
function handles the date / time update, the alarm check and finally clears
the UIP bit.
- writing to control register A doesn't change the UIP bit
Here are comments from the anonymous patch author :
I found bug when using the FPU in 16 bit protected
mode. When using the fsave command the 32bit
structures were used instead of the 16bit ones.
It now distinguishes between 16 and 32 bit protected mode
which wasn't the case before.
Now Ultima8 runs fine:)
new experimental stripped-down version of plex86, which is now
a user-code-only VM. I ripped out all the fancy stuff in plex86,
such that under that right conditions, user-code (protection level 3)
can run at near native speeds inside the plex86 VM.
The general idea is that bochs emulates all the initial real-mode code,
and guest kernel code (protection level 0). When it senses the
right conditions (like the context switches to user-code), a shim
is called to execute the guest inside the plex86 VM. All guest-generated
faults/exceptions are then forwarded back to bochs to be handled in
the emulator.
Actually, I'm not yet adding the mods to the bochs code (other than
the shim code which is in a separate file), until I hear that we're
back in a more development mode with bochs after the 2.0 release.
The plex86 subdirectory is really a separate project. It's just more
convenient to co-develop it with bochs for now. Both projects are
currently LGPL, but each should be taken to be a separate project,
and have their own license file. Plex86 (it's only a kernel driver
now) could ultimately be used with other projects, as it's modular.
I talked with Bryce, and we both agreed it's OK to keep plex86 as
a subdir in bochs for now.
PSRAW_PqQq (MMX)
PSRAD_PqQq (MMX)
PSRAW_PqIb (MMX)
PSRAD_PqIb (MMX)
PSRAW_VdqWdq (SSE)
PSRAD_VdqWdq (SSE)
PSRAW_PdqIb (SSE)
PSRAD_PdqIb (SSE)
When register was shifted by 0 bits the result produced was incorrect.
Now Bochs fully passes MMX test provided by
Hentai Yagi [hentai_yagi@yahoo.com.au] !
the icon problem on Win2000. RegisterClassEx() would be required if we want
to specify a special mini icon.
- the win32 gui now uses the new function/macro DEV_vga_actl_pal_idx() to
determine the text mode colors.
palette - it's an index to the attribute controller palette index.
The new function get_actl_palette_idx() now returns the real DAC index.
The guis sdl, wx and x11 are using the new function. TODO: Fix this bug in
all other guis.
- FAQ - sound question: added FreeBSD to the list of supported platforms
- list of configure shortcut files updated
- compiling footnote: file locations updated
- replaced reference to sound.html (old docs) by a reference to the
"sb16-emulation" section
- compiling instructions for BeOS updated
- keyboard mapping: supported guis added
the current code page's mode. It was &'ing against fetchModeMask:
((pageWriteStamp & fetchModeMask) == fetchModeMask)
instead of against a constant mask which populates all the
relevant bits:
((pageWriteStamp & ICacheFetchModeMask) == fetchModeMask)
* Moved the check above to ::prefetch(), since nothing should change
between calls to that function. I added some code to functions
which load CS to make sure that is the case.
* Commented out several calls to invalidate_prefetch_q() in
ctrl_xfer{16,32,64}.cc. Changes to only EIP (i.e. CS is not
modified) don't matter any more. If EIP lands outside the
window of the current code page, ::prefetch() is called.
* These changes add a few percent of performance.
- move configure options section right after Standard Compile. It
was buried at the end of the section and hard to find before.
- fix section issue for sb16
- fix typo in pit description
- fix some FIXMEs
- move documentation section to the top
- move wxWindows after config interface and display library sections
- remove the massive list of SF bugs,features,etc. since they have
(mostly) been covered above
- erase to-do list since it's done
sash to run.
1) fixed fetchdecode64.cc to fix the operand size at 64 bits in long mode for moves
to/from CRx
2) minor patches to sse2.cc to fix unimplemented and 64 bit variants of sse2
instructions.
Because source files were added/removed it would require an update
of the windows and macos project files, so I want to wait until after 2.0.
M Makefile.in 1.51 back to 1.50
M cpu.h 1.121 back to 1.120
M fetchdecode.cc 1.37 back to 1.36
M fetchdecode64.cc 1.33 back to 1.32
M sse.cc 1.17 back to 1.16
A sse2.cc 1.27 back to 1.26 (added back)
R sse_move.cc removed
R sse_pfp.cc removed
- to bring these changes back again, all we have to do is
"cvs update -j tmp-before1 -j tmp-after1"
sse.cc -> general SSE stuff and SSE integer (MMX extensions)
sse_move.cc -> memory transfer and shuffle opcodes
sse_pfp.cc -> packed floating point operations
the script to behave. It checks the current directory and the directory one up
for bochs.app. If both those checks fails then it just outright asks the user
for what bochs to run. (And notes the path so it won't ask again unless it
doesn't find things where they last were)
Modified Files:
bochs.applescript script_compiled.rsrc
- add notes about dlcompat on macosx, libtool ltdl, DLL building on win32
- add x86-64 notes from Peter Tattam
- add PSE,PGE,PAE in cu section
- specify some of those carbon improvements based on reading logs
- update to-do list
value be the application's directory, no matter where it gets
installed. This is consistent with installing the bioses and
keymaps/* into the same directory as bochs.app.
Authors: Br'fin, Bryce
- for MacOSX/Carbon, if you start the app by double clicking in the finder,
there is no text console. So in logfunctions::fatal(), display the
final message in a Carbon dialog instead of using fprintf.
- in the text-mode config interface, return an error code if the
user tries to use the text menus when there's no text console.
Back in bxmain() this error code causes a panic
"Bochs needed the text console, but it was not usable"
- if a config file was not found, normally we turn off quick start
and drop into the text menus. However in the no-console case the
text menus won't work, so we added a panic: Unable to start Bochs
without a bochsrc.txt and without a text console. This is a much
more useful error than the generic one above.
- removed ensure_text_console() which was making matters worse instead
of better.
Modified Files:
logio.cc main.cc gui/siminterface.cc gui/siminterface.h gui/textconfig.cc
been closed since 1.4.1. This massive list is not going to stay in
here forever--it's just to help us remember, summarize, and cross
reference as we write the changelog. Note that some have been
closed but not resolved--we can't just assume that every single thing
has been marked "Fixed".
completely ignored and Bochs has always been returning 0. Now Bochs should
exit with the proper exit code.
- implementation: BX_EXIT(exitcode) calls SIM->quit_sim(exitcode).
quit_sim stores the exit code in the SIM object, then uses longjmp
to unwind the stack and return to main(). main() calls
SIM->get_exit_code() to retrieve the exit code and returns it.
- If you are wondering why we don't just call exit(exitcode), it is
because in the wxWindows interface all of the CPU simulation runs as one
thread while the user interface runs as another thread. With the longjmp
technique the simulation thread ends cleanly and it is possible to
start a new simulation thread without making a new process.
and miss index.html. Then they would only find the documentation about
the documentation--not so helpful. Therefore, on win32 package install
documentation into
- docs/index.html
- docs/user/*
- docs/development/*
- docs/documentation/*
be used for more things than just the BIOS. Also the keymaps need to
be located relative to the share path, and maybe other things like
graphics and icons, later.
with scp and then log in and gunzip it on the shell server. This speeds
up "make webinst" quite a bit.
- add "make dl_docs" which downloads a bochsdoc.tar.gz from the web site.
This will be used to grab the rendered documentation for platforms that
do not have docbook tools. (Most don't.)
- stop installing "README" because depending on where you put it, it can
overwrite the main README.
you have dlcompat installed. Without dlcompat, the configure/build works
fine but the plugins cannot be loaded. This new test prints a meaningful
error message before you waste any time trying to build it and figure
out why the plugins cannot be dlopened.
authors. If someone has contributed a section, we would give them credit in
that section. If they write enough sections, we can take out the
section-by-section credits and list them as an author.
make install it will build the docbook documentation and install it
into $(prefix)/share/doc/bochs. If it is not specified, then it will
be enabled only if the program "docbook2html" is found in your path.
- this completes [ 652957 ] install docbook docs instead of HTML
Modified Files:
Makefile.in configure configure.in doc/docbook/Makefile.in
- only rebuild docbook stuff when the source is newer
- add install target
- allow $SFUSER environment variable to define the sf username, instead of
hardcoding the user->sfuser translations in the makefile.
- now alldocs is called index.html
- add dist-clean target
- now configure script does substitutions in doc/docbook/Makefile.in
- add "make install_docbook" target in main makefile that runs make install
in doc/docbook. It will not be run unless you ask for it.
Modified Files:
configure.in Makefile.in configure doc/docbook/Makefile.in
things up for MacOSX, and I don't want to risk that right now just before
the release. I changed it to #if BX_WITH_MACOS to be safe.
(Bochs makefiles currently define -Dmacintosh on OSX. Maybe they
shouldn't but I'm not going to change it right now.)
be zero. I changed it to #ifdef macintosh.
- removed extern "C" {} that only encompassed two #defines. If there
were any includes in there it would be important, but there weren't.
is important because many people don't have the tools to rebuild them.
They will only be removed if you type "make bios-clean", which people
are unlikely to do by accident.
- I had to hack around the disk image size measurement and force it to use
a 20meg disk image. Otherwise it would run out of space while copying
the files into the image. I don't know why this was necessary.
eject, insert, eject, etc. several extra times before allowing you to
actually change the CD. The extra insert/eject cycles were coming from the
param handlers for cdrom status and for the cdrom pathname. To fix this, I
changed wxdialog.cc and textconfig.cc to only call set() on parameters if
they are different from the original value. Now it generally does what you
would expect.
- to change physical CD: change the Bochs CD status to "ejected". On some OSes
that will actually eject it, but on others it will simply close the file so
that the eject button will actually work. Then put the new CD into the
drive, and change the Bochs CD status to "inserted".
Modified Files:
main.cc gui/siminterface.cc gui/siminterface.h
gui/textconfig.cc gui/wxdialog.cc
the first cdrom. Implemented by MyFrame::editFirstCdrom().
- eliminate OnOtherEvent and call OnEditATA directly from the event table.
Modified Files:
gui/wxmain.cc gui/wxmain.h
- move DOC-linux.html into the docbook in a section called "Quick Start for
Linux Users", and removed all references to the DOC-linux file that I could
find. Most of the text has been moved into the Linux RPM section. Parts
that did not fit into the docbook are now in doc/docbook/misc.txt.
- put vga.pcf into INSTALL_LIST_X11 so that it's only installed when
you configure with --with-x11.
- transfer section on VGA font installation from docs-html into user guide
- remove RCS id from screen shots within user guide, since they are changed
to the revision info of user.dbk.
Modified Files:
Makefile.in configure configure.in build/linux/.cvsignore
build/redhat/bochs.rpmspec.template doc/docbook/misc.txt
doc/docbook/user/user.dbk doc/man/bochs-dlx.1 doc/man/bochs.1
doc/man/bochsrc.5 doc/man/bximage.1 docs-html/install.html
Removed Files:
build/linux/DOC-linux.html.in build/linux/bochs-docs.in
* missing VBE support added, use xmax / ymax variables instead of the
values 640 / 480
* use the constant Y_TILESIZE when checking the y position
* variable vga_mem_updated was always set twice
- it used to list the options that were enabled in .conf.win32-vcpp, but I
removed it. It was out of date, and it's unlikely to stay in sync
with the file even if I fix it now.
reaches a certain maximum length. Otherwise the debug console grows and
grows and never frees any memory. The max length is defined as a constant
in wxdialog.h. The max length should be made into a variable some day.
- This fixes bug [ 624317 ] wx: debug console never discards lines.
Modified Files:
gui/wxdialog.h gui/wxdialog.cc
- add tip section about serial port (no content yet)
- add tip section on how to enter special key combination
- add diclaimer in Windows NT 4.0 section
- add comment about installing and running Windows NT 4.0
- add comment about installing Windows XP
the first hard drive or first cdrom drive no matter where they are found.
Now the wxWindows interface recognizes this fact. It allows you to
select HD boot if there is a hard drive in any ATA location, and
select CDROM boot if there is a cdrom in any ATA location.
- this fixes bug [ 616139 ] wx: boot hd/cd must be in ata0 interface
Modified Files:
gui/siminterface.cc gui/siminterface.h gui/wxmain.cc
This version is a bit of a hack that needs to be cleaned up before release.
There is also a glaring bug that lets things go WAY too fast when
we're trying to catch up to real time. This is my highest priority
right now.
