Ros-dev

ros-dev@reactos.org

1 participants
5793 discussions

start
by Imre Leber 30 Aug '06

30 Aug '06

So, if nobody objects I guess i'll start porting DOSemu to reactos. After all, we need to start somewhere. And when this is done, it would be easier to talk about why it is or it is not a good replacement for ntvdm. Imre

3 2

Re: [ros-dev] RE : Re: (Free)DOS subsystem
by Imre Leber 30 Aug '06

30 Aug '06

>-----Original Message----- >From: Sylvain Petreolle [mailto:spetreolle@yahoo.fr] >Sent: Wednesday, August 30, 2006 10:54 AM >To: 'ReactOS Development List' >Subject: [ros-dev] RE : Re: (Free)DOS subsystem > >Hi, >--- Imre Leber <imre.leber(a)telenet.be> a écrit : > >> >> DOSemu has a 80386 emulator on board. >> >> Or at least in the sources. It doesn't seem to be in the DOSemu documentation so it seems that >> it is non-functional. >> >> But it is always easier to fix something then to restart from scratch. >> >Like e.g. Windows ? Bad answer ;) > Windows??? Can you point me to the official download area? Imre >Kind regards, >Sylvain Petreolle (aka Usurp) >--- --- --- --- --- --- --- --- --- --- --- --- --- > > Run your favorite Windows apps with free ReactOS : http://www.reactos.org >Listen to non-DRMised Music: http://www.jamendo.com > > >Linux is not as well stable as it is told to. The proof is, mine has restarted two years ago, on the occasion of a power cut. >- H. Eychenne > >_______________________________________________ >Ros-dev mailing list >Ros-dev(a)reactos.org >http://www.reactos.org/mailman/listinfo/ros-dev > >

1 0

(Free)DOS subsystem
by Imre Leber 29 Aug '06

29 Aug '06

Hi, I guess i am new to this list. I have been a maintainer on the FreeDOS project for more then 7 years. I have also noticed that you are still looking into building a DOS subsystem in reactos. After the release of FreeDOS version 1, I would like to see wether I could not take up the task to having it running under reactos. My main idea involves using a port of DOSemu. Is there any documentation regarding the running of a DOS subsystem in windows that I should study beforehand? Imre

6 5

Re: [ros-dev] (Free)DOS subsystem
by Imre Leber 29 Aug '06

29 Aug '06

>-----Original Message----- >From: Imre Leber [mailto:imre.leber@telenet.be] >Sent: Tuesday, August 29, 2006 02:57 PM >To: 'ReactOS Development List' >Subject: Re: [ros-dev] (Free)DOS subsystem > > >>-----Original Message----- >>From: Peter Dolding [mailto:oiaohm@bluebottle.com] >>Sent: Tuesday, August 29, 2006 01:33 PM >>To: 'ReactOS Development List' >>Subject: Re: [ros-dev] (Free)DOS subsystem >> >>Myria wrote: >>> I've wanted to write an NTVDM for ReactOS, but I'm not sure when I'd ever >>> have time. >>> >>> Windows NT's DOS subsystem, NTVDM, is a user-mode program that runs on top >>> of Win32. It runs DOS programs as Windows processes, using v86 mode to run >>> the programs. DPMI programs are supported by asking the kernel to allocate >>> LDT segments and running directly. (If you write a DOS32 program and know >>> the correct addresses, you can call MessageBoxW in user32.dll and it will >>> work.) >>> >>> This differs significantly from DOSBox, which is effectively a machine >>> emulator rather than an API translator. DOSEmu, in comparison, is >>> essentially the same design as NTVDM but for Linux. >>> >>> FreeDOS in DOSBox works very well, because the FreeDOS kernel is its normal >>> self and is unaware that it's inside a VM. However, this doesn't work for >>> NTVDM/DOSEmu. In these, the FreeDOS kernel will need to be heavily >>> modified, particularly because the NT kernel handles file I/O. Programs >>> inside the virtual DOS environment use illegal opcodes to talk to NTVDM. >>> >>> Melissa >>> >>True Closest Freedos to Windows NT NTVDM is >>http://freedos-32.sourceforge.net/. Its was newer version under >>development. It could be stalled from 2005. >> > >Why? > >I never got it. It doesn't seem to add anything new. Nothing that can't be done in FreeDOS. > >FreeDOS is for all pratical purposes a capable 32 bit system. It can run 32 bit DOS and (some) windows programs and it has multithreading (eRTOS), TCP/IP (watt32), etc... . > Well. Maybe I should have said a 32 bit capable operating system. >I think it would not be very difficult to write a multitasking DOS extender either (maybe as an extension to cwsdpr0). > But who's ever going to care? >If it's because it is running in 32 bit, I am not all that convinced that DOS-C can not be ported to 32 bit. It was originally intended to run on apple macintosh. > >Imre Imre > >>Peter Dolding >> >>_______________________________________________ >>Ros-dev mailing list >>Ros-dev(a)reactos.org >>http://www.reactos.org/mailman/listinfo/ros-dev >> >> > > >_______________________________________________ >Ros-dev mailing list >Ros-dev(a)reactos.org >http://www.reactos.org/mailman/listinfo/ros-dev > >

1 0

Re: [ros-dev] (Free)DOS subsystem
by Imre Leber 29 Aug '06

29 Aug '06

>-----Original Message----- >From: Peter Dolding [mailto:oiaohm@bluebottle.com] >Sent: Tuesday, August 29, 2006 01:33 PM >To: 'ReactOS Development List' >Subject: Re: [ros-dev] (Free)DOS subsystem > >Myria wrote: >> I've wanted to write an NTVDM for ReactOS, but I'm not sure when I'd ever >> have time. >> >> Windows NT's DOS subsystem, NTVDM, is a user-mode program that runs on top >> of Win32. It runs DOS programs as Windows processes, using v86 mode to run >> the programs. DPMI programs are supported by asking the kernel to allocate >> LDT segments and running directly. (If you write a DOS32 program and know >> the correct addresses, you can call MessageBoxW in user32.dll and it will >> work.) >> >> This differs significantly from DOSBox, which is effectively a machine >> emulator rather than an API translator. DOSEmu, in comparison, is >> essentially the same design as NTVDM but for Linux. >> >> FreeDOS in DOSBox works very well, because the FreeDOS kernel is its normal >> self and is unaware that it's inside a VM. However, this doesn't work for >> NTVDM/DOSEmu. In these, the FreeDOS kernel will need to be heavily >> modified, particularly because the NT kernel handles file I/O. Programs >> inside the virtual DOS environment use illegal opcodes to talk to NTVDM. >> >> Melissa >> >True Closest Freedos to Windows NT NTVDM is >http://freedos-32.sourceforge.net/. Its was newer version under >development. It could be stalled from 2005. > Why? I never got it. It doesn't seem to add anything new. Nothing that can't be done in FreeDOS. FreeDOS is for all pratical purposes a capable 32 bit system. It can run 32 bit DOS and (some) windows programs and it has multithreading (eRTOS), TCP/IP (watt32), etc... . I think it would not be very difficult to write a multitasking DOS extender either (maybe as an extension to cwsdpr0). If it's because it is running in 32 bit, I am not all that convinced that DOS-C can not be ported to 32 bit. It was originally intended to run on apple macintosh. Imre >Peter Dolding > >_______________________________________________ >Ros-dev mailing list >Ros-dev(a)reactos.org >http://www.reactos.org/mailman/listinfo/ros-dev > >

1 0

Re: [ros-dev] [ros-diffs] [arty] 23765: Menu displays. Need stdin/hid objects.
by Thomas Weidenmueller 29 Aug '06

29 Aug '06

http://svn.reactos.org/svn/reactos/branches/powerpc/reactos/boot/freeldr/fr… > ============================================================================== --- branches/powerpc/reactos/boot/freeldr/freeldr/arch/powerpc/mach.c (original) > +++ branches/powerpc/reactos/boot/freeldr/freeldr/arch/powerpc/mach.c Tue > Aug 29 11:16:16 2006 > @@ -30,6 +30,7 @@ > BOOLEAN AcpiPresent = FALSE; > char BootPath[0x100] = { 0 }, BootPart[0x100] = { 0 }, CmdLine[0x100] = { > 0 }; > jmp_buf jmp; > +char *video_mem = 0; > > void le_swap( void *start_addr_v, > void *end_addr_v, To avoid incorrect code generation with enabled optimizations, the variable video_mem needs to be defined as: volatile char *video_mem = NULL; - Thomas

1 0

Frist test of clipboard patch from SoC project
by Magnus Olsen 28 Aug '06

28 Aug '06

Hi I I got the clipboard patch from thomas and waiting on the final dectsions from him. Thomas need some time to review it. for it is around 121Kbytes big I did test it in ReactOS and I alot happy 1. Copy text and picture betwin program is working with this big patch 2. Copy file from one place to anyther working. But I wisch it show file copy when it doing it 3. copy file to the desktop does not working nug in explore ?? 4. move icon on the desktop is a new bug in win32k. 5. mark the text is white background with white text instead for invert background with white text 6. cut/copy/paste is working 7. keys like cut/copy/paste (ctrl+x,ctrl+c,ctrl+v) is working in all program I tested. say this patch is really good sussess and adding a clipboard to reactos. that working betwin program and u can copy/paste files and more I say it it is up to thomas to aprove the code. I have not looked at the code at, more that a fast preview.

3 3

GetSiteSidFromToken in advapi32
by Grzegorz Jaskiewicz 28 Aug '06

28 Aug '06

Hi folks based on my knowledge and some docs I used to read ages ago. Plus some recent googling, I kinda recreated function GetSiteSidFromToken from advapi32.dll - as it was missing there, and I couldn't start w2k taskmgr because of that. Of course simple stub returning NULL should be sufficient, but I gave it a shot and tried to implement it fresh and fully functional. First, I am not sure of few things there. And if someone is more knowledgable than I am - please let me know your thoughts. As far as I know the function returns pointer to SID from token. But there's more than one token. So it returns token that has "SITE" SID. Now, there were quite few SIDs missing in ROS, I tried to add them - mainly guessing their names based on some googling. Some docs on msdn say that they are all defined in ntseapi.h - but I couldn't found that file anywhere. For more details lookup attachment 1043 and corresponding bug. Function is quite simple, it iterates through all sids, trying to match the one with "SITE" authority. Question now arises - what if there's more than one such SID ? Does ROS use SIDs 1-6 to 1-8 at all ? is the SECURITY_INTERNETSITE_AUTHORITY the SITE authority or perhaps SECURITY_SITESERVER_AUTHORITY ? I don't know. I will try to write simple code that would ask original function from original dll about it, and see what it comes up with. First part of the code comes as simple c&p from other functions. The iteration is quite obvious. I didn't knew what should I use to compare SIDs, so I used memcmp. But perhaps simple comparing values in the table one by one would be simple, or maybe there's some sort of ROSish/NTish specific function that is more "right" than memcmp. Another question I have - is it normal practice that you alloc extra ram and return pointer to copy (seems to be in other places in your code), or should I just simply return pointer to the SID ? I do appreciate any comment. Patch is attached. if someone has the ntseapi.h file anywhere on their discs, I would like you to verify SECURITY_*_AUTHORITY names as well as values. I spent whole day gathering info, and I used all info I could find. Sorry for such long letter. Much too long I suppose. this is my first attempt to hack anything for ros. I promise next time I would try to take care of something more obvious and better documented. till than, ta. -- GJ

1 0

Status of trunk
by Aleksey Bragin 27 Aug '06

27 Aug '06

Hello, after recent fixes by Alex Ionescu and Thomas Weidenmueller trunk works rather good again. What I tested: - bootcd: installs, 2nd stage install works, explorer shows up and is usable - livecd: doesn't boot due to broken mkhive (it should use cmlib and it doesn't do at the moment) - GDB support in ntoskrnl: I was able to connect to the GDB with a cmdline client, disassembly, however I wasn't able to test it more because our keyboard/mouse drivers don't work when vmware has no focus during booting. WBR, Aleksey Bragin.

1 0

Shutdown
by James Tabor 27 Aug '06

27 Aug '06

Hi, Instead of posting things on chat and coming across as flames, so here we go. I added excerpts and opt edited from "Inside Microsoft Windows 2000". http://www.reactos.org/wiki/index.php/ShutdownProcess "Csrss in turn impersonates the caller and sends a Window message to a hidden window owned by Winlogon telling it to perform a system shutdown." The best way I could do this at one time, take for example. Find the program ntpmon.exe run it and open a console window. Type ctrl-C or break at the same time watch ntpmon. You will see cmd clone itself. One step further in the console window run listdlls.exe and watch ntpmon. You will see "impersonation", threads created under processes to access each process context. Cool right! I could have written this for lib/rtl/dbgbuffer.c but I thought KJK_Hyperions way was cooler and used NtReadVirtualMemory to peek at process context instead. I did have a note in lib/rtl/dbgbuffer.c alluded to this process but one of "the coders that knows more" removed it. So, impersonates; user access rights, its a thread in Csrss created under the user program process context, it finds the parent and sends the shutdown message. I guess, James

2 1

Jump to page:

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

Ros-dev