Ros-dev

ros-dev@reactos.org

2 participants
5791 discussions

msvcrt/crtdll "merge"
by Gunnar Dalsnes 27 Jan '05

27 Jan '05

Hi, I'm currently "merging" msvcrt and crtdll (again). I'll move msvcrt into a library lib\crt and have msvcrt and crtdll link against it. Only dllmain.c will be left in msvcrt/crtdll. Most of (99.9%) crtdll will be dropped. It all seems to work nicely. One problem thou: I ran into some header problems, where i relied on some stuff in include\msvcrt\string.h but no matter what, mingw\include\string.h were included instead (and they both defined _STRING_H_). After looking at the headers in inlcude\msvcrt they all seem to be ripped from mingw. Does anyone know why they were put in include\msvcrt when mingw has headers for all of this stuff? Can i just remove them (include\msvcrt)? I looked at many depends files and saw most files depend on many mingw headers. Is this correct? Should we depend on mingw headers at all? Gunnar

1 0

videoprt.nostrip.sys
by Ge van Geldorp 27 Jan '05

27 Jan '05

It seems videoprt.nostrip.sys is somewhat messed up. I've attached the output of "objdump -p videoprt.nostrip.sys". Data Directory Entry 1 says there's an import directory at 0x0000b000. Then down under "PE File Base Relocations (interpreted .reloc section contents)" we find: Virtual Address: 0000b000 Chunk size 12 (0xc) Number of fixups 2 reloc 0 offset 14 [b014] HIGHLOW reloc 1 offset 20 [b020] HIGHLOW When the module is loaded, the relocations are applied. This messes up the import table, with the result that the module fails to load (copy videoprt.nostrip.sys to \reactos\system32\drivers\videoprt.sys and boot ReactOS). The 2 relocations given above are not present in the "normal" videoprt.sys module. I'm inclined to blame gcc (I'm using 3.4.1) or more likely binutils (2.15.90 20040222) for this, but before I make Filip unhappy, does anyone know if we're doing something wrong during the build process? Gé van Geldorp.

2 1

Re: [ros-svn] [ion] 13305: Dynamic 3GB support, part 1. Only multiboot.S remains to be changed, all other parts of the Kernel now used KERNEL_BASE based on the command-line (Please don't use /3gb if you don't compile 3GB)
by Alex Ionescu 26 Jan '05

26 Jan '05

ion(a)svn.reactos.com wrote: >Dynamic 3GB support, part 1. Only multiboot.S remains to be changed, all other parts of the Kernel now used KERNEL_BASE based on the command-line (Please don't use /3gb if you don't compile 3GB) > > >Updated files: >trunk/reactos/config >trunk/reactos/ntoskrnl/include/internal/i386/mm.h >trunk/reactos/ntoskrnl/ke/main.c >trunk/reactos/ntoskrnl/mm/mminit.c > >_______________________________________________ > > Royce and I are currently working on making this fully dynamic. We've got the plan set-up in our minds and it should be committed in less then a week, unless unexpected things happen. Speaking of which, is anyone against not making ACPI a compile-time instruction? It just makes debugging harder, one more thing to worry about, and I don't see the point either...I think it makes much more sense to have some kind of ACPISupported global variable that we either set at boot-up with proper table queries, and/or that the user can specify with an /ACPI command-line flag. Best regards, Alex Ionescu

2 1

new binutils snapshot
by Jonathan Wilson 26 Jan '05

26 Jan '05

http://sourceforge.net/project/showfiles.php?group_id=2435 Does this make those issues (dlltool I think) with using the previous snapshot go away?

1 0

Slab cache
by Hartmut Birr 25 Jan '05

25 Jan '05

Hi, do we need the slab cache in ntoskrnl\mm\slab.c? I've replaced the kmap functions (ntoskrnl\mm\kmap.c) by a process local mapping (hyperspace). The slab cache needs the mapping for a longer time and cannot use the hyperspace mapping. I would remove the slab cache, because it is not used by any of the kernel functions. - Hartmut

4 3

user mode exceptions
by Hartmut Birr 25 Jan '05

25 Jan '05

Hi, on a user mode exception, I get hundreds of lines like this: (i386/exception.c:96) RtlpDispatchException() Dumping exception registrations: (0x0065FFE0) HANDLER (0x7C82CEAF) (0x7FFDF000) HANDLER (0x00000000) (i386/exception.c:96) RtlpDispatchException() Dumping exception registrations: (0x0065FFE0) HANDLER (0x7C82CEAF) (0x7FFDF000) HANDLER (0x00000000) (i386/exception.c:96) RtlpDispatchException() Dumping exception registrations: It needs a very large time, if ros does response again. It occurs on real hardware. Debugprints are going over a serial line. - Hartmut

1 0

Re: [ros-diffs] [hbirr] 13277: - Fixed the directory index of a file for FATX.
by Hartmut Birr 25 Jan '05

25 Jan '05

Hervé Poussineau schrieb: > hbirr(a)svn.reactos.com a écrit : > >> - Fixed the directory index of a file for FATX. >> + ULONG startIndex = rcFcb->startIndex; >> + if ((rcFcb->Flags & FCB_IS_FATX_ENTRY) && !vfatFCBIsRoot(Parent)) >> >> { >> >> + startIndex += 2; >> + } >> + if(startIndex >= DirContext->DirIndex) >> + { > > > This change looks suspicious to me. > "." and ".." don't exist on a FATX volume. They are FAT12/16/32 specific. > Please revert related changes in create.c, dirw.c, direntry.c, fcb.c > > Hervé > > PS: > Comment in direntry.c ("need to add . and .. entries".) is not very > clear. It may be modified to "Need to add . and .. entries, because > they don't exist in a FATX subdir" > _______________________________________________ > Ros-diffs mailing list > Ros-diffs(a)reactos.com > http://reactos.com:8080/mailman/listinfo/ros-diffs > > FATX doesn't have real '.' and '..' directories. But we use them on each directory except the root. The index from fcb is also used to update the entry of a file in its parent directory. For looking for a directory, '.' has index 0, '..' has index 1 and 'foo' has index 2 and so on. The real index of 'foo' is 0. We must make a difference between the real and the search index. I've modified some parts of freeldr and usetup. I can create and install ros on a fatx partition on a normal pc. I've also load vfatfs on W2K. At the beginning, the wrong index has scrambled the fatx partition. - Hartmut

2 1

We have a web browser] : Input ?
by Gge 25 Jan '05

25 Jan '05

With the Mozilla ActiveX Control , I can browse "www.reactos.com" from the Reactos Explorer as indicated before. The link to www.reactos.com is made by explorer at start but I do not know how to type an URL (www.google.com for example) in the Explorer window ? Any idea ? Regards Gerard

5 10

WineConf 2005
by Steven Edwards 25 Jan '05

25 Jan '05

Hello, If you are a ReactOS developer thats planning on coming to WineConf 2005 please fill out the RSVP form at http://www.winehq.org/site/wineconf We need to get a estimated head count so even if your not 100% sure that you will make it you can RSVP and note that on the webstie. So far GvG and myself are planning on attending and Thomas,KJK and fireball have said they will try to attend as well. Thanks Steven __________________________________ Do you Yahoo!? The all-new My Yahoo! - What will yours do? http://my.yahoo.com

1 0

Re: [ros-dev] Re: We have a web browser : Input ?
by barry day 25 Jan '05

25 Jan '05

On Mon, Jan 24, 2005 at 05:58:27PM -0600, Royce Mitchell III wrote: > > > > If we even tried, the only thing we would do is alienate potential users > of our OS, because their favorite killer app doesn't work on it. > Clone Microsoft's mistakes and you won't have potential users, there will be no reason to swap to Reactos.

4 4

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