the extra memory. It is in no way part
of the XMS interface to coordinate PAE paging between HIMEM
and the software that uses XMS. Of course you are right that
PAE paging most likely isn't needed, the simpler approach 32-bit paging with
4 MB pages should do as well.
japheth
Hello,
Japheth, I tried what you said: ran hdpmi32 -r-m, then the exe, and it
worked;
then tried cwsdpmi -x then the exe, that also worked. Remaining question:
how do I build-in the '-x' switch at compile time? The sandman site
does not address this point, and the delorie search archives
Eric, the program I wrote, which drew the page fault, uses the int86 in
four of the eight functions. the program
is a mouse event handler, and micro-gui. I already had included
'dos.h', but I did not realize ther was a 'dpmi.h'!
I'll have to pry into that header to try to see what exactly
Internet Exporer 5 and earlier, maybe 6.
Hardest-core CRAP, please don't pull it out if it's grave.
Using IE 5 instead of Arachne is a good idea. It has support for the MS
version of JavaScript (JScript) and runs even on Windows 3.11 standard mode.
This makes it render a lot of web sites
Hi,
Blake Stone Aliens of Gold requires the MS-DOS 6.22 or the Windows 9x
version. It can run on top of Windows 9x.
Wolfenstein 3D has the same problem because it's probably based on the
same underlying engine. It starts under Freedos, but none of the keys
work.
I just tried BS and W3D
Hello
Ahhh, it's amazing; not one tutorial specifically for the dos version of
'make'; even what there is, not too
specific. To wit: indeed, some thing you are trying to install,say , HX,
GEM, etc., they do include makefiles,
[snip]
You mentioned HX so I'm forced to feel somewhat
Are the BLASTER settings described anywhere in detail? VPC says it
emulates the Creative Labs Sound Blaster 16 ISA Plug and Play. Maybe my
settings are wrong.
You can use the GETSB.EXE tool which is contained in HXGUI
(http://www.japheth.de/hx.html). It reads the SB16 mixer status
also many things are running in background :)
just ask Eric, Mateusz, Rugxulo, Blair, ...
That's probably one of FreeDOS's problems: there's ony guy - I don't want to
be more specific - who forces all communication into private mails. This makes
the current FD status absolutely intransparent
why do you think that the RxDOS kernel is superior - besides having
many more bugs ?
What about built-in support for LFN?
--
Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM)
software. With
Hope you don't mind me to ask, why didn't you rather contribute to FreeDOS?
But he DOES contribute to FreeDOS: once the few remaining bugs are fixed in
the RxDOS kernel, it will replace the inferior current FD kernel.
Would you want a compressed filesystem at all?
My discouraging answer: I just don't need it.
I think, what FreeDOS needs for daily use is a good graphical web
browser, a nice e-mailer, a word processor like Abiword, FOSS USB
drivers, ...
In what way would FreeDOS differ from Linux 2.4
Sorry for long responce. I was update Jemm, because previous version
have no xmsstat. The xmsstat output attached as memstat.zip.
the report for Jemm did display
no handle regionsize(kB) locks flags
1 00A8
2. use Jemm386 instead of JemmEx and continue to use your previous
XMS host.
3. JemmEx's source can be extended to be able to manage more than
one block. This means some work, because the current implementation
isn't very flexible.
Nothing wrong in my fdconfig.sys?
JEMMEX.EXE A20 ALTBOOT
Hi,
an update for JWasm was necessary, now v1.91 is current. The update is
strongly recommended, v1.9 contains a pretty nasty bug in 16bit code
generation, bad code might be generated without warning.
Besides bugfixes, there's one improvement: the /bin switch allows to create
raw binaries, no
Note that JWASMD, a WASM variant with MASM .asm support, needs
the HDPMI32 DOS extender in your PATH. See www.japheth.de ;-).
there's now JWasm v1.9 available, which contains a stand-alone DOS binary.
download: http://www.japheth.de/JWasm.html
changelog:
Maybe it is context sensitive in TASM, 0 in .data and nop in .code?
I believe NASM has a whole range of ways to align things, including
pad with user-specified value and pick a multi byte noop machine
code instruction where appropriate but you would have to check the
docs which is selected
It's fixed in v1.8.
Odd, my report already was about 1.80 ...
Ok, then you forgot to tell us that you have access to a time machine. Because
the versions stored on my site are v1.7 and v1.8pre.
Then an error message would have been better than a 0 value...
Sure. Looking back, many things
Hi,
thanks for posting your results!
- TASM evendata pads with 0, JWASM even pads with 0fch
I do not know how to change this, I just used db 0.
this is indeed hard-coded in the source and will be used only if a 16-bit CODE
segment has to be padded by 1 byte. The reason why not just 0x00
Hello,
v1.13 of FD Debug is available: http://www.japheth.de/debxxf.html.
v5.71 of Jemm is available: http://www.japheth.de/jemm.html.
Both are just bugfix releases, and the bugs fixed are minor.
-
This SF.net email is
Hi :-)
v1.13 of FD Debug is available: http://www.japheth.de/debxxf.html.
v5.71 of Jemm is available: http://www.japheth.de/jemm.html.
Both are just bugfix releases, and the bugs fixed are minor.
Nice,
Thanks for your kind remark! :)
but could you update the two readme files on your
Hello,
Debug v1.11 is out. 7 bugs have been fixed (same as in Jemm v5.70 - seems to
become my new standard). Also, debugx can now display floating point
registers much nicer. So this update is highly recommended - at least for
those who still know what DEBUG is good for.
1.11 [06
The forum discussion suggests that most of the bugfixes
in jemm386 5.70 fix minor or rarely occurring bugs :-).
Yes. The one thing which I'm unable to decide whether it's relevant is the
lacking reentrancy of v5.69 converning XMS block moves. A TSR which hooks the
timer interrupt and then
... is out:
http://www.bttr-software.de/forum/forum_entry.php?id=3132page=0order=timedescasc=DESCcategory=all
-
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
is out. It can be found at http://www.japheth.de/jemm.html
changes for v5.69
- memory block copies are now done with interrupts enabled - if client's
interrupt flag is set. This avoids interrupt latency and makes the
previously implemented interrupt windows obsolete. It also avoids
problems
Actually: Why does hdpmi use a non-1:1 mapping linear/physical?
I mean it has every right to do so, but it turned out that this
can make DPMI usage a bit more foolproof ;-). I assume it saves
a bit of page table space? Being NOT foolproof can be good as
it can help to detect bugs, but otoh, a
I'm happy that this thing - which is old and hard to maintain...
But I was talking about hdpmi32...!?
Yes, me too. hdpmi32 v1.0 (10/1993) is older than FreeDOS v0.001
-
This SF.net email is sponsored by: Microsoft
Defy
You can download it as part of my NEWTRACK eye-tracking DOS
software, and of course you need no dual VGA for PCXTEST ;-).
The download is at www.coli.uni-saarland.de/~eric/newtrack/
I found out that:
- pcxtest always works with dr emm386
- pcxtest fails (while some selector points
Just tell me what the bug is :-).
in vesainit.c, search for /* update */
It happens to run with CWSDPMI, but I'm almost
sure this is the only host it really runs with.
What is special about CWSDPMI (and DR EMM386) then?
I don't know.
Even easier, you can tell DOSEMU to emulate a VGA
PS: The CTS Toasted 1996 ctstoast demo runs okay in LFB
mode if CWSDPMI is resident or if you use DOSEMU, but I
promise you that I did not write CTS Toasted and did not
ask the authors to give it the same bug as PCXTEXT ;-).
I don't understand what you are talking about. There was a bug in
No, the crashing program just uses a DOS extender for
direct framebuffer access.
The DOS extender uses its own memory context (CR3, page tables), GDT and IDT.
The EMM (Jemm) is responsible vor v86-mode only.
-
This
- the program uses dos4gw but using dos32a instead does not help
- the program crashes with fd emm386 and with jemm386...
- the program works okay if only himem but no emm is loaded :-)
this is no proof at all that the error is in the EMM! :))
So maybe it uses the int 15 memory copy function
Remember that there is a parameter in HIMEM to set apart a little memory
for int15.. Did you try that?
I tried nothing because it is Eric who has a crashing program. :)
And from the error descriptions I got so far the problem is inside the
DOS-extended application, not in the EMM.
What
Hi, I just double-checked my own EMM problem:
With my PCIe GeForce 7600 VGA, both with the
classic FD EMM386 and with JEMM386, I get a
page fault when attempting to use VESA VBE
2.0/3.0 LFB framebuffer access to graphics.
Please explain in more detail how you come to think that this is a
Lelio wrote:
Thanks, now was with 613K of Convetional Memory Free... and with DOS
VPC's Guest Additions.
Good to know! The mentioned bug in Jemm which must be cirumvented currently by
setting option X=CC00-CCFF will be fixed in the next release.
I replaced HIMEM and EMM386 by HIMEMX and JEMM386 today (I am running FreeDOS
1.0). Unfortunately, it hangs my PC. The system is booting okay, but when I
use the PC, it displays an error message Jemm386: exception 08 occured
at... and various hex numbers.
It never happens (so far) when I
There it is: Jemm386: Exception 08 occured at: ... ERRC = 00110B40Col:1
Yes, the ERRC=0011 confirms that it is an IRQ 0, not an exception.
However, I have no clue what the Col:1 suffix is meant to be. If this isn't
a typo, there is a chance that Jemm's code in memory has been -
Hi,
You should have mentioned something about the context
instead of only throwing URLs at us ;-). The first URL
talks about making more low RAM free by loading things
high into UMB and by using fewer drivers. The second
URL talks about DR DOS. One thing that might be useful
for FreeDOS and
--
Creative SB16 Emulation Driver, Version 5.00
Copyright(c) 1996-2001, Creative Technology Ltd., All Rights Reserved.
Creative Audio @ Port d800, IRQ 7
error: Counld not allocate code/patch RAM below 4 Mbyte boundary. Try
loading SBEINIT.COM before SMARTDRV.EXE or
Hello,
when I run ctmouse v2.1beta in Qemu, it says Installed at COM1 (...) in Mouse
Systems mode and the mouse does NOT work inside the emulator.
If I try ctmouse /P, it displays Error: device not found
The old ctmouse v1.91a does not have such a problem, it detects a PS/2 mouse
and works in
I made tests running xdma32/xcdrom32 in Qemu, VirtualBox and VirtualPC. There
are significant speed increases to achieve, but they require to set additional
parameters. For details see:
Qemu: http://www.bttr-software.de/forum/forum_entry.php?id=2463
VBox VPC:
Jack, my fRiend, do you think it is a good idea to make a federal case of a
missing little r? Try to calm down!
Jim Hall isn't a FOOL by having ignored it, it's at best an indication that he
isn't paranoid.
The guys you mentioned to be your witnesses, Kuhnt and Lucho, they are
slightly
I was referring to one forum already banned some
of the people involved in this flamewar and another
forum was closed because of this flamewar. I am
not suggesting to follow those examples... ;-).
Eric, you are a wag. :))
btw, I was banned by the admin of the EDR forum (Evil DoctoR) because
Hello,
Jemm v5.68 is now available at http://www.japheth.de/Jemm.html
changelog for 5.68:
- bugfix: JLOAD's Lock DMA function ignored flag to check for 64 kB
border crossing.
- NMIs occuring inside the monitor are now silently routed to v86-mode.
Previous Jemm versions displayed an
Hello,
running Phar Lab TNT binaries didn't work with Jemm versions 5.60-5.65.
This is fixed in v5.66:
http://www.bttr-software.de/forum/forum_entry.php?id=342
-
This SF.net email is sponsored by: Splunk Inc.
Still
16 kB alignment.
Japheth
-
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems? Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE
Eric Auer wrote:
... [very detailed explanation] ...
Thanks for your thoughts and hints, Eric!
It is
interesting that the catch just puts an int 19 there,
is that a real reboot or just a fastboot in this case?
I've forgotten by now. But you can test with DEBUG:
a
:0100jmp
Hello,
There is a bugfix release available for HimemX at
http://www.japheth.de/Himem331.zip
The first bug caused v3.30 to fail on cpus pentium when running in
real-mode, the latter gave unexpected results when running on MS-DOS 6.x
(and most likely 5.x, 4.x and 3.x as well).
In case you
There is a bugfix release available for HimemX at
http://www.japheth.de/Himem331.zip
sorry, the link actually is
http://www.japheth.de/Download/Himem331.zip
-
This SF.net email is sponsored by: Splunk Inc.
Still
We (many people in this list) espent months testing himem and emm386
from Michael Devore. Are those two just a bugfix from *that* well tested
version or is it descendent from somewhere else?
As for HimemX: there are 12 bugfixes and some optimizations, all described in
the readme.txt.
As
Hi,
Hi Japheth,
Japheth escreveu:
Btw, in the name of how many people are you speaking exactly? And what do
you
regard as well tested. Because, as for me, I seriously doubt that a
program
is well tested just because some dozens (or some hundreds? I have no idea
how big the FD user
- FDHimem opens an interrupt window in v86-mode only. HimemX opens
an interrupt window in both v86- and real-mode.
that was probably the point where Jack started his own civil war ;)
At least it WAS a bug, because this issue is mentioned in the XMS docs.
- in FDHimem the routine which
Bart has added support for that to the CVS/SVN version of
our kernel. I hope Japheth noticed :-). The changes are:
On int19 (prepare for warm boot / warm boot / run boot
sector or however you want to call it), FreeDOS now un-
hooks all hooked BIOS interrupts. And on boot, FreeDOS
now stores
The hxrtd.zip file mentioned above also seems to have something wrong
with it, the 'ole32.dll' file has a file size of 0 bytes!
Thanks for the hint! It has been fixed. However, this file isn't relevant for
mplayer AFAICS.
I tried it too again with those files, but get lots of errors with
Then you should try www.japheth.de JEMM386 and maybe some manual
tuning for the options. Try for example X=e000-efff X=test and
maybe some options which are recommended in the JEMM386 readme.
The most important setting in Jemm386 concerning MS VPC might be NOINVLPG.
Indeed, this option was
May you show examples? Starting from minimal void main(){}, then with
printf(Hello, world!\n), then something more complex? With OW 1.6:
a dummy main and a hello world main are utterly irrelevant if you want to
compare size/speed optimisations (leaving CRT aside).
A complex sample is
How much?
MS VC is 6% better than OW, both optimizing for size. OW is not bad,
especially if comparing MSVC with TC, where MSVC is 25% better.
-
Take Surveys. Earn Cash. Influence the Future of IT
Join
Better than watcom? How much better and in which areas?
In Open Watcom just the 32bit C(++) compiler and WD are pretty good. The rest
is of mixed quality.
With optimisation on MSVC CL.EXE beats OW WCC.EXE significantly both in speed
and binary size (not counting the CRT part).
One
I own MS VC 1.5 which runs in DOS with the Phar Lab TNT DOS Extender.
IMO it is the best 16-bit C(++) optimising compiler :)
-
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and
Are the binaries unmodified?
I'm asking because there is a KNOWN bug in OW 1.6 with DOS4G applications
launching other applications. The enviroment of the launched application
contains garbage, which makes it pretty impossible to use BINW\WMAKE for
example.
of Linux or Windows. This means that for example you have
to wait until your app gets a new DHCP lease every time
when you start it, because no central networking driver
of the operating system would be around to cache that.
IIRC watt-32 (not sure about wattcp) caches this info in file
Lester wrote:
I'm trying to replace SRDISK with SHSURDRV (v1.11, 2005-12-21) to make a
ram drive E. Fdconfig and Fdauto load ok but when using Arachne
browser, after exiting (to dos shell or permanently), I get the following
error:
error reading from drive E: DOS area: write protection
Lester Vedrox wrote:
My connection: DSL modem + router + Realtek NIC.
The packet driver RTSPKT.com loads fine but Arachne won't open any web page.
_http://img138.imageshack.us/img138/9059/10112006035337it6.png_
Which version of Arachne.cfg shown below is correct?
Ver #1 (all tcp/ip
The host stack problem is solved. Fine :)
Not a severe problem (IMO it's more like a feature), but the A20 line
switching will only work if an UMB handler has been installed. There are some
constellation where this is not true, and then the A20 gate is unhandled by
Emm386.
There is also still
Also, your patch doesn't solves issue (which you
try to explain) in case, if extended key pressed together with Ctrl-Alt.
Ie., even if problem exist, then you fix for it incomplete.
Yes, since it is a bug in *qemu*, not Emm386. Emm386 can do its best to work
around this bug, but not fix it.
Barring the fileset I requested, or other testable indication that your
problem occurs with anyone else, I should not think that would be a
problem. In other words, no need to wait for EMM386 3.0 - codenamed The
Arkady Factor.
Since the error is evident in my eyes, I will not spent any
I have been made aware that there is no description of the error I was
refering previously here on the list. That's true, sorry, it was described in
BugZilla. But a brief overview is:
- it requires Himem + FD Emm386 to be loaded
- it occurs when FD Himem has been started with parameter
Are you willing to do that trivial task and
demonstrate an actual interest in helping to correct the issue, or will you
just preach and post here?
Neither nor. The issue is of minor general interest, since most FD users will
use FD Himem and FD Emm386 with the standard parameters. For me it
J (apparently gives problems with extended keys).
Which ones?
As you possibly know, by reading port 60h you send the keyboard controller an
interrupt acknowledged signal, so it can send the next byte. On real
hardware this takes some time (some microseconds at least), but qemu is
this is the c:\fdos\fdauto.bat which was generated. May give a hint where the
messages are coming from.
@echo off
SET dosdir=C:\FDOS
SET LFN=N
set PATH=%dosdir%\bin
set NLSPATH=%dosdir%\NLS
set HELPPATH=%dosdir%\HELP
set temp=%dosdir%\temp
set tmp=%dosdir%\temp
SET BLASTER=A220 I5 D1 H5 P330
REM
After the first installation I wanted to test if FD is now able to detect if
it has enough free space (or at least is able to abort the install process in
a friendly manner).
I renamed directory FDOS to FDOSALT. About 6,5 MB free space are left on C:
- I booted from CD and select Install, FD
Blair Campbell wrote:
I renamed directory FDOS to FDOSALT. About 6,5 MB free space are left on C:
- I booted from CD and select Install, FD does not complain!
I am so sorry. I forgot to include the updated install.exe ! I
have re-uploaded the ISO with the updated install.exe included.
On my machine A20 is always enabled. BIOS int 15h reports kbd and fast PS/2
methods being available, but none works to disable A20. IMO if A20 cannot be
disabled in real-mode, emm386 shouldn't try to emulate A20 disable in
v86-mode.
This is my experience so far:
First attempt
I tried to install on a pretty small partition (QEMU) where MS-DOS 5.0 was
already installed. The installation very soon stops due to insufficient
hard disk space (although nothing is told, you just see no more progress
on the screen). After that the
Without a true SB card, I can't duplicate the problem, but I would strongly
bet against the Windows-introduced and flavored GEMMIS being used in any
meaningful (or at least critical) manner by DOS4G -- or any other DOS
extender of notable popularity. Function 4a15h is much more likely,
But: DOOM does work with the current himem/emm386.
As for me this is true for the newest versions of himem/emm386 (July 2006)
only. In fact, any DOS4G application didn't work with FD-Emm386 previously
because my machine has 768 MB.
And do work doesn't mean works well. For example, the MIDI
Digital Mars
---
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache
76 matches
Mail list logo