- Ros-dev - ReactOS.org

Have anyone tried, stress prime on ReactOS?
by Ashuaria Lee 16 Dec '08

16 Dec '08

Hi, Guys(and Girls) did you heard about a program called "stress prime"? It was developed to calculate prime numbers. And it is very sensitive to small errors. Since it has some kinda detecting errors, many overclockers are using it to confirm overclocked H/W stability. Yes, I tried to run this program, just for fun, and surprise! The system is not overclocked, nor has problem when running stressprime on WindowsXP. The rounding error amount was huge. I saw the source code of the stress prime and I'm sorry it was assembly language, so I couldn't check what could make such prolem under ReactOS. It was all about the math, and I'm not good at both(asm,math) But, one thing sure, is something is not working properly(maybe kernel?) and it is affecting ReactOS stability. What is your opinion? any comments? -- Regards, Ashuaria Lee

2 3

[ros-diffs] [dgorbachev] 38083: - Use HEAP_ZERO_MEMORY instead of memset(). - Do not crash if no adapter found.
by Dmitry Gorbachev 15 Dec '08

15 Dec '08

Hi, Alex Ionescu wrote: > You should always use memset and void HEAP_ZERO_MEMORY. Could you please tell why HEAP_ZERO_MEMORY should not be used? It is used in many places, so if something wrong with RtlAllocateHeap, maybe better to fix it? Thanks, Dmitry

2 1

Re: [ros-dev] [ros-diffs] [dgorbachev] 38083: - Use HEAP_ZERO_MEMORY instead of memset(). - Do not crash if no adapter found.
by Alex Ionescu 14 Dec '08

14 Dec '08

You should always use memset and void HEAP_ZERO_MEMORY. On 14-Dec-08, at 3:45 PM, dgorbachev(a)svn.reactos.org wrote: > Author: dgorbachev > Date: Sun Dec 14 14:45:16 2008 > New Revision: 38083 > > URL: http://svn.reactos.org/svn/reactos?rev=38083&view=rev > Log: > - Use HEAP_ZERO_MEMORY instead of memset(). > - Do not crash if no adapter found. > > Modified: > trunk/reactos/dll/cpl/desk/desk.c > trunk/reactos/dll/cpl/desk/settings.c > > Modified: trunk/reactos/dll/cpl/desk/desk.c > URL: http://svn.reactos.org/svn/reactos/trunk/reactos/dll/cpl/desk/desk.c?rev=38… > = > = > = > = > = > = > = > = > ====================================================================== > --- trunk/reactos/dll/cpl/desk/desk.c [iso-8859-1] (original) > +++ trunk/reactos/dll/cpl/desk/desk.c [iso-8859-1] Sun Dec 14 > 14:45:16 2008 > @@ -162,7 +162,7 @@ > InitPropSheetPage(&psh, PropPages[i].idDlg, > PropPages[i].DlgProc, PropPages[i].Callback); > } > > - /* NOTE: Don;t call SHAddFromPropSheetExtArray here because > this applet only allows > + /* NOTE: Don't call SHAddFromPropSheetExtArray here because > this applet only allows > replacing the background page but not extending the > applet by more pages */ > > ret = (LONG)(PropertySheet(&psh) != -1); > > Modified: trunk/reactos/dll/cpl/desk/settings.c > URL: http://svn.reactos.org/svn/reactos/trunk/reactos/dll/cpl/desk/settings.c?re… > = > = > = > = > = > = > = > = > ====================================================================== > --- trunk/reactos/dll/cpl/desk/settings.c [iso-8859-1] (original) > +++ trunk/reactos/dll/cpl/desk/settings.c [iso-8859-1] Sun Dec 14 > 14:45:16 2008 > @@ -151,8 +151,7 @@ > DWORD ResolutionsCount = 1; > DWORD i; > > - newEntry = HeapAlloc(GetProcessHeap(), 0, > sizeof(DISPLAY_DEVICE_ENTRY)); > - memset(newEntry, 0, sizeof(DISPLAY_DEVICE_ENTRY)); > + newEntry = HeapAlloc(GetProcessHeap(), HEAP_ZERO_MEMORY, > sizeof(DISPLAY_DEVICE_ENTRY)); > if (!newEntry) goto ByeBye; > > newEntry->Settings = GetPossibleSettings(DisplayDevice->DeviceName, > &newEntry->SettingsCount, &newEntry->CurrentSettings); > @@ -311,6 +310,7 @@ > EnableWindow(GetDlgItem(hwndDlg, IDC_SETTINGS_RESOLUTION), FALSE); > EnableWindow(GetDlgItem(hwndDlg, IDC_SETTINGS_RESOLUTION_TEXT), > FALSE); > EnableWindow(GetDlgItem(hwndDlg, IDC_SETTINGS_ADVANCED), FALSE); > + ShowWindow(GetDlgItem(hwndDlg, IDC_SETTINGS_SPECTRUM), SW_HIDE); > } > else if (Result == 1) > { > Best regards, Alex Ionescu

1 0

Type definition in NDK
by Dmitry Gorbachev 12 Dec '08

12 Dec '08

May I know from where comes NLS_FILE_HEADER structure definition in include/ndk/rtltypes.h ? I think it is not correct.

1 0

why IOCTL_DISK_GET_DRIVE_GEOMETRY_EX failed in windows 2000?
by Jack Yi 10 Dec '08

10 Dec '08

hi,guys: I don't know if it is proper to ask such a question here. I'm sorry if I break some rules. why IOCTL_DISK_GET_DRIVE_GEOMETRY_EX fail in windows 2000 and succeed in windows xp and greater version? or how can I get a disk(whatever,ATA,SCSI,USB )'s real size(including additional sectors) in a easy way in windows 2000, just like there is such a function, __int 64GetDiskRealSize(HANDLE hDisk); Could anybody help me? Waiting for reply. Thank you!

4 6

Re: [ros-dev] [ros-diffs] [cwittich] 37955: SeUnsolicitedInputPrivilege is obsolete - the LUID is used by SE_MACHINE_ACCOUNT_PRIVILEGE now
by Alex Ionescu 10 Dec '08

10 Dec '08

Hm? On 9-Dec-08, at 3:33 AM, cwittich(a)svn.reactos.org wrote: > - Luid->LowPart = Priv + 1; > + Luid->LowPart = Priv + 2; Best regards, Alex Ionescu

1 0

Re: [ros-dev] [ros-diffs] [jimtabor] 37952: - Add CP_UNIXCP for CP_ACP, this will help cross tests.
by Steven Edwards 09 Dec '08

09 Dec '08

On Tue, Dec 9, 2008 at 12:58 AM, <jimtabor(a)svn.reactos.org> wrote: > Author: jimtabor > Date: Mon Dec 8 23:58:23 2008 > New Revision: 37952 > > URL: http://svn.reactos.org/svn/reactos?rev=37952&view=rev > Log: > - Add CP_UNIXCP for CP_ACP, this will help cross tests. > > Added: > trunk/reactos/include/reactos/wine/winnls.h (with props) > +#ifndef __WINE_WINNLS_H > +#define __WINE_WINNLS_H > + > +#define CP_UNIXCP CP_ACP Rather than keep the include_next brokeness, perhaps we should add the define to wine/port.h and include that in any offending source file. -- Steven Edwards "There is one thing stronger than all the armies in the world, and that is an idea whose time has come." - Victor Hugo

2 1

compatibility vs. correctness
by Matthias Kupfer 09 Dec '08

09 Dec '08

Hi folks, yesterday a user shows me a program what is working properly under windows, but not working under reactos. It was an open source program and it was easy to figure out the problem. It was an invalid win-api-call (wrong parameter value). Due to this situation I would like to ask one question. How compatible reactos should be? Even fault compatible? Matthias -- Matthias Kupfer Telefon +49 (0) 371 236 46 52 Wilhelm-Firl-Straße 21 Mobil +49 (0) 160 859 43 54 09122 Chemnitz

11 12

compatibility vs. correctness
by Dmitry Gorbachev 08 Dec '08

08 Dec '08

I believe that people wish "perfect world" API as an optional feature, not that they really want to broke compatibility.

3 2

compatibility vs. correctness
by Dmitry Gorbachev 08 Dec '08

08 Dec '08

Steven Edwards wrote: > We discussed something like this years ago. It would be nice if every sort > of legacy API that has any sort of hacks for brokeness as part of its > implementation, could be isolated and a simple recompile with something > like a #define NO_LEGACY_WIN32 so you could recompile without the hacks. > Not really that useful for the real world as you need the backward > compatibility but it is a nice idea for embedded or non-normal windows > targets like ARM. I think this is like compiler warnings. Some checks are disabled by default as they would cause more harm than good. But there is an option to enable them. It will help to catch bugs in software.

1 0

Please dont break functionality due to a compiler problem, just fix it.
by Dmitry Gorbachev 08 Dec '08

08 Dec '08

>> It was already broken in r37496. > Gdi is right Win32k is wrong. I mean that GCC (by default) silently discards code like if (i < 0) { ... } where i is unsigned. So there was a bug in GDI...

1 0

Please dont break functionality due to a compiler problem, just fix it.
by Dmitry Gorbachev 08 Dec '08

08 Dec '08

It was already broken in r37496.

2 1

[ros-diffs] [ekohl] 37891: Replace call to ExtractIconExW (shell32.dll) by a call to PrivateExtractIconExw (user32.dll) and remove shell32.dll from the list of linked libraries because shell32.dll requires a valid user account upon initialization. Pl
by James Tabor 07 Dec '08

07 Dec '08

Eric Kohl is an Ace Programmer! Thanks Eric! James

1 0

Re: [ros-dev] [ros-diffs] [dchapyshev] 37876: - Implement SetupInstallFileA/W - Stub implement SetupPromptForDiskA/W All from Wine.
by teknopup 07 Dec '08

07 Dec '08

1 0

Re: [ros-dev] [ros-diffs] [dchapyshev] 37876: - Implement SetupInstallFileA/W - Stub implement SetupPromptForDiskA/W All from Wine.
by Ged 06 Dec '08

06 Dec '08

dchapyshev(a)svn.reactos.org wrote: > + if (!(inf_source = HeapAlloc( GetProcessHeap(), 0, len * sizeof(WCHAR) ))) > + { > + SetLastError( ERROR_NOT_ENOUGH_MEMORY ); > + return FALSE; > + } > + if (!SetupGetStringFieldW( inf_context, 1, inf_source, len, NULL )) return FALSE; > + source = inf_source; There's a memory leak here. It's best to read Wine code when throughly, they have bugs too ;) Ged.

1 0

ddk header cleanup
by Timo Kreuzer 05 Dec '08

05 Dec '08

Now that Stefan has cleaned up some stuff in the ddk headers, maybe we should think of some additional design fixes here. We currently have ntddk.h include "per version" headers: winxp.h, win2k.h, winnt4.h. These contain very few stuff. My suggestion is to merge them into winddk.h. We could also merge winddk.h and ntddk.h It might make sense to split wdm.h and ntddk.h into architecture specific files Layout suggestion: - ntddk.h: includes wdm.h and contains arch-independent definitions currently in winddk.h/win2k.h/winnt4.h - wdm.h: includes arch-dependent headers, contains arch-independend wdm definitions - ddk_x86.h / ddk_x64.h, ddk_arm.h, ddk_ppc.h / ddk_mips.h: arch-dependent definitions for both wdm.h and ntddk.h definitions (#ifdef _NTDDK_ ...) Comments appreciated. Timo

1 0

Re: [ros-dev] [ros-diffs] [jimtabor] 37783: - Fixed hooking bug for Firefox, a plus, new SEH found a bug!
by Timo Kreuzer 04 Dec '08

04 Dec '08

This looks like you're creating a mem leak. > Author: jimtabor > Date: Sun Nov 30 17:39:20 2008 > New Revision: 37783 > > URL: http://svn.reactos.org/svn/reactos?rev=37783&view=rev <http://svn.reactos.org/svn/reactos?rev=37783&view=rev> > Log: > - Fixed hooking bug for Firefox, a plus, new SEH found a bug! > > Modified: > trunk/reactos/subsystems/win32/win32k/ntuser/callback.c > > Modified: trunk/reactos/subsystems/win32/win32k/ntuser/callback.c > URL: http://svn.reactos.org/svn/reactos/trunk/reactos/subsystems/win32/win32k/nt… <http://svn.reactos.org/svn/reactos/trunk/reactos/subsystems/win32/win32k/nt…> > ============================================================================== > --- trunk/reactos/subsystems/win32/win32k/ntuser/callback.c [iso-8859-1] (original) > +++ trunk/reactos/subsystems/win32/win32k/ntuser/callback.c [iso-8859-1] Sun Nov 30 17:39:20 2008 > @@ -507,12 +507,12 @@ > } > _SEH2_END; > > - IntCbFreeMemory(Argument); > - > if (!NT_SUCCESS(Status)) > { > return 0; > } > + > + if (Argument) IntCbFreeMemory(Argument); > > return Result; > } >

2 1

Re: [ros-dev] [ros-diffs] [sginsberg] 37792: - Get rid of DDK's dependency on windef.h and winnt.h - Clean up, fix, and add missing types to ntdef.h/ntifs.h/winddk.h/wdm.h/ntpoapi.h - Add ntimage.h and use it in rossym and NDK - Fix a gcc compatibility issue -- some internal NT structures have their pointer types defined in the DDK headers, but we define them along with the "struct" type when defining the full structure. MSVC is okay with this double definition, but gcc treats it as an error (hidden by syst
by Timo Kreuzer 04 Dec '08

04 Dec '08

What's the reason to copy & paste comments from ms headers? Please remember: Interfaces are not copyrighted, header files (including the comments) are! > +// > +// Types to use to contain PFNs and their counts. > +// > +typedef ULONG PFN_COUNT; > typedef ULONG PFN_NUMBER, *PPFN_NUMBER; > +typedef LONG SPFN_NUMBER, *PSPFN_NUMBER; >

1 0

Boot broken, trunk locked
by Dmitry Gorbachev 04 Dec '08

04 Dec '08

Hi! I'm sorry I broke the trunk. About "extra" comma within braces, it is not a mistake. It is allowed. There is a fragment of C grammar: initializer : assignment_expression | '{' initializer_list '}' | '{' initializer_list ',' '}' ;

1 0

Boot broken, trunk locked
by Aleksey Bragin 03 Dec '08

03 Dec '08

Hello, booting bootcd leads to a stackfault of VMWare, build server experiences similar problems. Trunk is locked with write access available to Dmitry Gorbachev to fix the issue, and to anyone by request who has the fix. WBR, Aleksey Bragin.

2 1

Re: [ros-dev] [ros-diffs] [dgorbachev] 37830: Silence compiler warnings (8/11).
by Aleksey Bragin 03 Dec '08

03 Dec '08

On Dec 3, 2008, at 8:34 PM, dgorbachev(a)svn.reactos.org wrote: > Author: dgorbachev > Date: Wed Dec 3 11:34:49 2008 > New Revision: 37830 > -static FREELDR_SETTINGS Settings = { 0, {0}, 0 }; > +static FREELDR_SETTINGS Settings = { 0, { 0, }, 0, 0, FALSE }; I suspect there is a typo in the above line ("extra" comma within braces, should've been something like 0, {0, 0}, 0, FALSE instead probably). WBR, Aleksey.

1 0

Re: [ros-dev] [ros-diffs] [ekohl] 37801: Sync to wine-1.1.1 (Patch 4 of 10): ATTENTION: You MUST run "clean" before building ReactOS, otherwise WIDL will crash when it is run. - Dan Hipschman <dsh at linux.ucla.edu> Tue, 1 Jul 2008 widl: Implement type redefinition
by Colin Finck 01 Dec '08

01 Dec '08

Hi Eric, Just wanted to tell you that your patch is correct and the Release Buildslave is failing as you can see from its log. For some reason, it stopped cleaning properly today and also has weird file deletion issues sometimes. But as long as the patch builds on the Debug slave, it's correct in 99% of all cases. Wanted to finish this mail before you reverted in r37802, but I was too slow. :-( So, it's safe for you to reapply r37801 and it's here to stay! Best regards, Colin

1 0

Re: [ros-dev] [ros-diffs] [gschneider] 37774: Update C++ demangle code to recent wine, now all msvcrt cpp tests pass except one.
by Aleksey Bragin 01 Dec '08

01 Dec '08

On Nov 30, 2008, at 10:19 PM, gschneider(a)svn.reactos.org wrote: > Author: gschneider > Date: Sun Nov 30 13:19:45 2008 > New Revision: 37774 > - if (!(tmp = (char*)und_alloc(sym, len))) return NULL; > + if (!(tmp = und_alloc(sym, len))) return NULL; You might want to keep these when syncing, it reduces compiler warnings.

1 0

Question about east asian language support
by Ashuaria Lee 01 Dec '08

01 Dec '08

Hello, everyone. Since I'm new to this mailing list, I may ask stupid questions, so don't be mad at me. ;-) I was trying to get the status of east asian language support, but couldn't find it in the wiki. Could someone tell me some brief status about east asian support? -- Regards, Ashuaria Lee

2 2

Re: [ros-dev] [ros-diffs] [jimtabor] 37766: - Sync in rev 37720: Magnus Olsen, add GdiGetBitmapBitsSize is a new api from windows xp sp3, the code was tested in vs2003.
by Aleksey Bragin 01 Dec '08

01 Dec '08

Instead of copying the whole formula, it could be done as ULONG Height; if (lpbmi->bmiHeader.biHeight >=0 ) Height = lpbmi->bmiHeader.biHeight; else Height = -lpbmi->bmiHeader.biHeight; retSize = Height * ((lpbmi->bmiHeader.biWidth * lpbmi->bmiHeader.biPlanes * lpbmi- >bmiHeader.biBitCount + 31) & -32) / 8; Also comments contain typos /* Make height postiive if it negitve then calc the bits Size and align it*/ postiive, negitive WBR, Aleksey. On Nov 30, 2008, at 4:17 PM, jimtabor(a)svn.reactos.org wrote: > Author: jimtabor > Date: Sun Nov 30 07:17:19 2008 > New Revision: 37766 > + if (lpbmi->bmiHeader.biHeight >=0 ) > + { > + /* Calc the bits Size and align it*/ > + retSize = lpbmi->bmiHeader.biHeight * ((lpbmi- > >bmiHeader.biWidth * > + lpbmi->bmiHeader.biPlanes * lpbmi- > >bmiHeader.biBitCount + 31) & -32) / 8; > + } > + else > + { > + /* Make height postiive if it negitve then calc > the bits Size and align it*/ > + retSize = (-lpbmi->bmiHeader.biHeight) * ((lpbmi- > >bmiHeader.biWidth * > + lpbmi->bmiHeader.biPlanes * lpbmi- > >bmiHeader.biBitCount + 31) & -32) / 8; > + }

2 1

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

Ros-dev