- Ros-dev - ReactOS.org

console functions are broken
by Hartmut Birr 26 Aug '05

26 Aug '05

Hi, parts of the changes from r16213, 17417 and 17431 have broken most of the console functions like WriteConsoleOutputCharacter or WriteConsoleOutputAttribute. In many cases the length of the buffers in kernel32 and the check of the buffers in csrss are wrong. As a result of this, Midnight Commander shows only a black screen. - Hartmut

2 1

New Homepage Content
by Maarten Bosma 26 Aug '05

26 Aug '05

Hi, Frik85 asked me write a suggestion how the structure of the content for the new homepage should look like. So here it is: http://wiki.reactos.com/New_Homepage_Content Let me know what you think about it. When we agreed about the headlines we can start to write the content into wiki. Some of it can be taken from the current wiki-pages and also some text from the old homepage can be reused. Maarten Bosma

1 0

Re: [ros-diffs] [weiden] 17542: add stubs for CreateSymbolicLinkA/W
by Joseph Galbraith 25 Aug '05

25 Aug '05

weiden(a)svn.reactos.com wrote: > add stubs for CreateSymbolicLinkA/W > > Modified: trunk/reactos/lib/kernel32/file/create.c > Modified: trunk/reactos/lib/kernel32/kernel32.def > > ------------------------------------------------------------------------ > *Modified: trunk/reactos/lib/kernel32/file/create.c* > > --- trunk/reactos/lib/kernel32/file/create.c 2005-08-25 17:38:15 UTC (rev 17541) > +++ trunk/reactos/lib/kernel32/file/create.c 2005-08-25 18:21:50 UTC (rev 17542) > @@ -357,4 +357,51 @@ > > return FileHandle; > } > > > + > +/* > + * @unimplemented > + */ > +BOOL STDCALL > +CreateSymbolicLinkW(IN LPCWSTR lpSymlinkFileName, > + IN LPCWSTR lpTargetFileName, > + IN DWORD dwFlags) > +{ > + SetLastError(ERROR_CALL_NOT_IMPLEMENTED); > + return FALSE; > +} > + > + > +/* > + * @implemented > + */ > +BOOL STDCALL > +CreateSymbolicLinkA(IN LPCSTR lpSymlinkFileName, > + IN LPCSTR lpTargetFileName, > + IN DWORD dwFlags) > +{ > + PWCHAR SymlinkW, TargetW; > + BOOL Ret; > + > + if(!lpSymlinkFileName || !lpTargetFileName) > + { > + SetLastError(ERROR_INVALID_PARAMETER); > + return FALSE; > + } > + > + if (!(SymlinkW = FilenameA2W(lpSymlinkFileName, FALSE))) > + return FALSE; > + > + if (!(TargetW = FilenameA2W(lpTargetFileName, TRUE))) > + return FALSE; > + > + Ret = CreateSymbolicLinkW(SymlinkW, > + TargetW, > + dwFlags); > + > + RtlFreeHeap(RtlGetProcessHeap(), 0, TargetW); > + > + return Ret; > +} Does this leak SymlinkW ? Thanks, Joseph

2 2

Re: [ros-svn] [royce] 17517: invoke _generate_dsp() have it open the output file, and fix some path parsing and const issues.
by Royce Mitchell III 25 Aug '05

25 Aug '05

royce(a)svn.reactos.com wrote: >invoke _generate_dsp() have it open the output file, and fix some path parsing and const issues. > > >Updated files: >trunk/reactos/tools/rbuild/backend/msvc/msvc.cpp >trunk/reactos/tools/rbuild/backend/msvc/msvc.h >trunk/reactos/tools/rbuild/backend/msvc/msvcmaker.cpp > > "make msvc" now creates a DSW file and DSP files such that msvc6 can load them. I seriously doubt anything will build yet, tho.

3 5

Re: [ros-svn] [hbirr] 17538: Fixed a bug in RtlLeaveCriticalSection. We have only to signal the event if someone waits on it.
by Thomas Weidenmueller 25 Aug '05

25 Aug '05

hbirr(a)svn.reactos.com wrote: > Fixed a bug in RtlLeaveCriticalSection. We have only to signal the event if someone waits on it. I believe this change is incorrect. This change introduces the exact same bug I fixed in r14326. The reason why this change is incorrect is, that you can't rely on LockCount. It may be incremented anytime a thread enters a critical section and waits for it. That's why there's the RecursionCount field, it is the only information we can rely on to determine whether we're about to release the lock or not. In this code path we already know that we left the last recursion, so no matter what the LockCount is at the moment (depends on how many threads are waiting for the critical section), we always need to unwait one waiter! This change most likely will introduce a dead-lock of cygwin. Best Regards, Thomas

2 3

Re: Re: [ros-dev] reactos.com or reactos.org?
by yash 25 Aug '05

25 Aug '05

>major Operating System manufacturer uses .org as their domain suffix. >Microsoft.com, Apple.com, Redhat.com. Given .ORG should be considered ROS is not a manufacturer, not a company, actually ROS community is a part of ReactOS source code and must be located at sourceforge. Will they register "dot sf" as a first level domain?

2 1

Re: [ros-dev] reactos.com or reactos.org?
by yash 25 Aug '05

25 Aug '05

My humble point is like this. Once ROS get's to it's first production release, you'll get many commercial offers and may need .com site for doing some commerce, like selling CD's and rotating ads. So many end users will come for download and support. Is it FreeBSD or RedHat way? Turning into monster is no good, but ROS might get huge end (very end) user hit, it is not freebsd and not even linux, it positions as a replacement for windoze itself and windoze means so much commercial and end user.... Yet my personal vision is to have www.reactos.com displaying "Server not found". So you gogo google and type reactos and if you're lucky you get to reactos.org and think "oh dear it is so cool and open source". >I was wondering if, together with the launch of the new website, we should >switch our "primary" domain from reactos.com to reactos.org? We have control >of both domains. An .org domain seems better suited for a open source >project. What do you all think? > >Ge van Geldorp (aka gvg(a)reactos.org :-)) > > >_______________________________________________ >Ros-dev mailing list >Ros-dev(a)reactos.com >http://reactos.com:8080/mailman/listinfo/ros-dev

1 0

RE: [ros-diffs] [navaraf] 17511: Implement the most over-optimized lockin NT - the NDIS_RW_LOCK.
by Casper Hornstrup 24 Aug '05

24 Aug '05

Some constants for the magic numbers would be nice. ________________________________________ From: ros-diffs-bounces(a)reactos.com [mailto:ros-diffs-bounces@reactos.com] On Behalf Of navaraf(a)svn.reactos.com Sent: 24. august 2005 14:21 To: ros-diffs(a)reactos.com Subject: [ros-diffs] [navaraf] 17511: Implement the most over-optimized lockin NT - the NDIS_RW_LOCK. Implement the most over-optimized lock in NT - the NDIS_RW_LOCK. Modified: trunk/reactos/drivers/net/ndis/ndis/control.c Modified: trunk/reactos/drivers/net/ndis/ndis/main.c Modified: trunk/reactos/drivers/net/ndis/ndis/stubs.c +/* + * @implemented + */ +VOID +EXPORT NdisAcquireReadWriteLock( IN PNDIS_RW_LOCK Lock, IN BOOLEAN fWrite, @@ -29,13 +48,88 @@ * NDIS 5.0 */ { - UNIMPLEMENTED + ULONG RefCount; + UCHAR ProcessorNumber; + UCHAR BusyLoop; + + ASSERT_IRQL(DISPATCH_LEVEL); + + if (fWrite) { + if (Lock->Context == PsGetCurrentThread()) { + LockState->LockState = 2; + } else { + KeAcquireSpinLock(&Lock->SpinLock, &LockState->OldIrql); + /* Check if any other processor helds a shared lock. */ + for (ProcessorNumber = KeNumberProcessors; ProcessorNumber--; ) { + if (ProcessorNumber != KeGetCurrentProcessorNumber()) { + /* Wait till the shared lock is released. */ + while (Lock->RefCount[ProcessorNumber].RefCount != 0) { + for (BusyLoop = 32; BusyLoop--; ) + ; + } + } + } + Lock->Context = PsGetCurrentThread(); + LockState->LockState = 4; + } + } else { + KeRaiseIrql(DISPATCH_LEVEL, &LockState->OldIrql); + RefCount = InterlockedIncrement((PLONG)&Lock->RefCount[KeGetCurrentProcessorNumber()].RefCount); + /* Racing with a exclusive write lock case. */ + if (Lock->SpinLock != 0) { + if (RefCount == 1) { + if (Lock->Context != PsGetCurrentThread()) { + /* Wait for the exclusive lock to be released. */ + Lock->RefCount[KeGetCurrentProcessorNumber()].RefCount--; + KefAcquireSpinLockAtDpcLevel(&Lock->SpinLock); + Lock->RefCount[KeGetCurrentProcessorNumber()].RefCount++; + KefReleaseSpinLockFromDpcLevel(&Lock->SpinLock); + } + } + } + LockState->LockState = 3; + } } /* * @implemented */ +VOID +EXPORT +NdisReleaseReadWriteLock( + IN PNDIS_RW_LOCK Lock, + IN PLOCK_STATE LockState) +/* + * FUNCTION: + * ARGUMENTS: + * NOTES: + * NDIS 5.0 + */ +{ + switch (LockState->LockState) { + case 2: /* Exclusive write lock, recursive */ + return; + + case 3: /* Shared read lock */ + Lock->RefCount[KeGetCurrentProcessorNumber()].RefCount--; + LockState->LockState = -1; + if (LockState->OldIrql < DISPATCH_LEVEL) + KeLowerIrql(LockState->OldIrql); + return; + + case 4: /* Exclusive write lock */ + Lock->Context = NULL; + LockState->LockState = -1; + KfReleaseSpinLock(&Lock->SpinLock, LockState->OldIrql); + return; + } +} + +

1 0

ATTN frik85: Website related bugs in Bugzilla
by WaxDragon 24 Aug '05

24 Aug '05

Here is the list of website related bugs currently in Bugzilla. Please keep these in mind when implementing the new site. When the new site is in place, please resolve the applicable bugs. "Bugzilla username/password weirdness" http://www.reactos.com/bugzilla/show_bug.cgi?id=22 "Roles on reactos.com" http://www.reactos.com/bugzilla/show_bug.cgi?id=68 "Various minor reactos.com issues" http://www.reactos.com/bugzilla/show_bug.cgi?id=76 "Taking a bug in bugzilla doesn't change the assignee" http://www.reactos.com/bugzilla/show_bug.cgi?id=196 "Hardware Compatiblity List (HCL)" http://www.reactos.com/bugzilla/show_bug.cgi?id=296 "rss feeds for reactos would be nice" http://www.reactos.com/bugzilla/show_bug.cgi?id=383 "users.reactos.com" http://www.reactos.com/bugzilla/show_bug.cgi?id=438 "Account Creation and Bugzilla is flaky" http://www.reactos.com/bugzilla/show_bug.cgi?id=459 "invisible graphic on the website seems not to exist" http://www.reactos.com/bugzilla/show_bug.cgi?id=612 "News Section on User Site is screwed up" http://www.reactos.com/bugzilla/show_bug.cgi?id=613 Thanks! WD -- #irc.freenode.net #reactos 01:03PM <filip2307> i don't know about any bug 01:04PM <filip2307> none exist 01:04PM <filip2307> ReactOS is prefect

2 2

We are linked from microsoft.com!!!
by reactos 24 Aug '05

24 Aug '05

Hello guys, Just want to encourage you about that we are linked from Microsoft.com!!! Nothing really nice, but we are... ...that's what matters, right? http://msdn.microsoft.com/library/default.asp?url=/library/en-us/dnsoftware… Yours sincerely, Jaix Bly

3 2

RE: Ros-dev Digest, Vol 12, Issue 38
by Lucio Diaz 24 Aug '05

24 Aug '05

The bug appears only in the instalation. I tried the livecd and it works fine. Best regards, Lucio Diaz. --------------------------------- Correo Yahoo! Comprueba qué es nuevo, aquí http://correo.yahoo.es

1 0

Re: ReactOS
by Jason Filby 24 Aug '05

24 Aug '05

Thanks Jim, I passed your regards on to the development team. I think sometimes we forget the great promise that ReactOS holds, but then someone like yourself reminds us. Cheers Jason On 8/23/05, Jim101 <jim101(a)bellsouth.net> wrote: > > I had hoped beyond hopes that someone was working on a truly open source > alternative to windows that could easily run any program ever made for > it, and I stumble across this project. > > All I can say is THANK YOU, and I wish your whole group the most > heartfelt best of luck. It would be the one thing that squared a wrong > done many years back by that little thief. To have an OS that would be > completely compatible with them all is a dream that far to many of us > share, and for it to be open source well. I just don't know what to say > I am ecstatic about it. Once again the very very best to each and > everyone of you that make it happen. > > For people like myself being disabled and on a fixed income the prices > that MS puts on their programs just sickens me. I have never intended to > be a thief in my life, I served my County and worked up until the day I > fell to the ground, never to recover completely. Then fought nearly 3 > years to get the measly little they give me back of my own money. And it > just irks me to see then do as they do, I mean I am all for a person > getting paid for their time and effort, it is only fair. but to charge > prices that most people can not archive, that is not how it was meant to > be. Anyway I am rambling, I just wanted to be sure you all knew that > there are many of us that would love to see your dream "and ours" come > true! > Best of luck! and God Bless! > >

2 1

ROS-release prepared
by Robert Köpferl 24 Aug '05

24 Aug '05

Hi, I prepared a distribution of ros 0.2.7 release. However, I have no good feeling. Not very reproducable (at least I didn't get the rule, yet) it makes problems in qemu. -Sometimes the setup without format doesn't work (crash) -Then the creation of some hive didn't work -> freeldr can't boot -Install works, boot fails At least I have an archive with qemu and working 0.2.7 Maybe that's all related. Possibly am I operating at some limit. Moving the mouse fills up the last bit of memory. Actually, however, if this is the same on our customers pcs, I don't like the idea of delivering that... Opinions

8 19

ReactOS 0.2.7 setup bug
by Lucio Diaz 23 Aug '05

23 Aug '05

Tested in real software. Sintoms: Setup start loading files, after the load a blue screen with the sentence "reactos setup" and two red lines appears in the top and the system lights become busy. After ten minutes of business nothing have happened and the bloq num keys of the keyboard are blocked (no light) so a hang or eternal cicle must be suspected. This computer used to load 0.2.5 (no hardware changes that i know, only software ones) Hardware: Semperon 2800+ 2 Ide HDD -Linux in the second (master boot record of the computer is here) with Grub loading both Suse and Yoper.(Filesystems a bit complex layout: fat, ext2, ReiserF, swap) -Windows 98 with a screwed MBR cause on installing grub i selected the wrong HDD once and my reactos loader was overwriten with grub.(all fat) A PS2 mouse and a PS2 keyboard. Gygabyte 2004 RZ mainboard with integrated ethernet and sound. Winfast graphic card. Realtek netcard (dont remember the model). DVD writemaster CD burner polaroid Any idea? Best Regards, Lucio Diaz. ______________________________________________ Renovamos el Correo Yahoo! Nuevos servicios, más seguridad http://correo.yahoo.es

2 1

ReactOS 0.2.7 Release is out
by Robert Köpferl 23 Aug '05

23 Aug '05

After a long time and some delays, the next version of Reactos is out. Reactos 0.2.7 is out on sourceforge.net, ready to download and explore. For your convenience, here's the official change log: ChangeLog-0.2.7 From ReactOS The changelog for 0.2.7 in terms meaningful to technical end-users. Table of contents 1 Generic 1.1 Overview 1.2 Main 0.2.7 Changes 1.3 Known Bugs 1.4 Translation 1 ReactOS Core 1.1 Kernel + Executive (NTOSKRNL) 1.2 Kernel Mode Drivers 1.2.1 Input Device 1.2.2 ACPI 1.2.3 PCI 1.2.4 SERIAL 1.2.5 USB 1.2.6 VFAT 1.2.7 VGA 2 Networking 3 Session Manager (SMSS) 4 Win32(tm) Personality 4.1 User mode subsystem server (CSRSS) 4.2 Kernel mode subsystem server (WIN32K) 4.3 Win32(tm) Libraries 4.3.1 ADVAPI32 4.3.2 KERNEL32 4.3.3 SETUPAPI 4.3.4 DirectX 4.3.5 Userenv 4.4 Libraries shared with Wine 4.4.1 USER32 4.5 Win32(tm) Applications 4.5.1 IPConfig 4.5.2 CMD 4.5.3 TASKMGR 4.5.4 EXPLORER 4.5.5 WINEFILE Generic Overview With the release of 0.2.7 we, the ReactOS Team, have once again gained further compatibility with Microsoft Windows. We have aligned our headers with those of the Windows SDK which gives us a greater range of driver compatibility. We have also kept ourselves compliant with the latest GCC compiler which with our new build system provides a faster and more efficient building processing. We have elected a new UI coordinator and work has begun to redo the interfaces seen by the user. Many of our advancements aren't obvious to the end user, but rest assured that the inner workings are coming together with work in USB, Plug and Play, and networking that will blow you away in our next release. Main 0.2.7 Changes * Header clean up to compatible Windows SDK headers. Implemented and used NDK. ReactOS is now built with our own headers. Alex Ionescu, Filip Navara, Steven Edwards * New My computer, Command Prompt, and Logo Icons. Mindflyer * New build system called rbuild. Based on xml, it allows for auto generated makefiles. Casper Hornstup * Improved appearance of the first stage installer. Alex Ionescu * GCC 4.x.x build fixes. Thomas Weidenmueller, Alex Ionescu Known Bugs ReactOS will fail to boot with a AllocConsole error when a ps/2 mouse is not attached on some hardware. Bug 688 (http://reactos.com/bugzilla/show_bug.cgi?id=688) This ReactOS release requires 64MB RAM to complete the first stage installer. Bug 703 (http://reactos.com/bugzilla/show_bug.cgi?id=703) Translation * Swedish translations by Andreas Bjerkeholt (harteex(a)gmail.com) * Swedish translations by Kris Engeman * Czech translations by Potapnik <jirka(a)studioprojekt.cz> * Swedish translations by David Nordenberg * French translation by Usurp * Polish translation by Sebastian Gasiorek <zebasoftis(a)gmail.com> * Belgian (point/Flemish) keyboard layout by i386DX <i386dx(a)hotmail.com> ReactOS Core Kernel + Executive (NTOSKRNL) (Hervé Poussineau) * Implemented: o IoRegisterDeviceInterface o IoSetDeviceInterfaceState o IoRegisterPlugPlayNotification o IoUnregisterPlugPlayNotification o PoRequestPowerIrp o IRP_MN_QUERY_RESOURCE_REQUIREMENTS for devices enumerated by Root bus (Eric_Kohl) * Implemented: o PLUGPLAY_GET_RELATED_DEVICE and PLUGPLAY_DEVICE_STATUS. o PlugPlayControlProperty and PlugPlayControlGetDeviceDepth o CM_Get_Global_State/CM_Get_Global_State_Ex (Alex Ionescu) * Kernel (Ke) o Implemented Guarded Mutex: + KeAcquireGuardedMutex, KeEnterGuardedRegion, KeLeaveGuardedRegion, KeInitializeGuardedMutex, KeAcquireGuardedMutexUnsafe, KeReleaseGuardedMutexUnsafe, KeAcquireGuardedMutex, KeReleaseGuardedMutex, KeTryToAcquireGuardedMutex o Fixed critical APC queuing and delivery bugs o Optimized Entering/Leaving critical sections and enabled APC delivery after each leave, if required. o Fix KeWaitForMultipleObjects if WaitAll was used. o Rewrote Context Switching to be faster. o Fixed KINTERRUPT structure and fixed KeDisconnectInterrupt o Fixed a bug where Kernel Queues were inserted on the Ready Thread list. * Input/Output (Io) o Implemented IoIsfileOriginRemote, IoGetLowerDeviceObject, IoGetDiskDeviceObject, IoGetRequestorSessionId, IoGetRequestorProcessId, IoRegisterBootDriverRetinialization, IoAttachDevicetoDeviceStackSafe, IoEnumerateDeviceObjectList, IoGetDeviceAttachmentBaseRef, IoDetachDevice, IoRaiseHardError o Fixed Controller Objects implementation. o Fixed Attaching to devices (IoAttachDevice) the driver will be notified with IO_ATTACH_DEVICE_API. o Fixed IoAttachDevicetoDeviceStack and IoAttachDevice to call the Safe cuntion. o Optimized IoGetRelatedDeviceObject o Removed IoOpenDeviceInstanceKey and IoQueryDeviceEnumInfo. o Cleaned up IopAllocateVpb o Optimized IoCreateDevice, add support for more flags and remove hard-coded sector size and incorrect sizes being set. o Fixed IRP Code not to zero out the IRP, free MDLs in failure cases, set the right IRP flags and set the I/O Object Type. o Reimplemtented 2nd-stage completion for IRPs to free ALL MDLs, free memory depending on the flags used, don't call I/O Completion if an APC is registered, don't set event/call APCs in some failure cases, don't use certain fields after the pointer can become invalid. o Use the right stack count in I/O Operations. o Fixed IopDeleteFile to fix a memory leak and dereference the completion port. o Fixed IopCloseFile, NtQueryInformationFile, NtFlushBuffersFile, NtQueryDirectoryFile, NtReadFile, NtWriteFile, NtSetInformationFile, IopSecurityFile, IopQueryFileName, NtDeviceIoControlFile, NtLockFile, NtUnlockFile which contained several bugs related to IRPs and completion, which were making assumptions or not supporting all the possibilities, as well as signlaing the wrong event or making the wrong kind of call, or using the wrong device object. o Implemented Lookaside lists for IRP packets to increase the allocation/deallocation speed by over 400%. o Optimized Completion Packets by piggybacking them on IRP packets if possible, and added the right memory flags to free them properly. o Share NtDeviceIoControlFile and NtFsIoControlFile o Fixed IRP cancelation. o Rewrote I/O Interrupt Functions to match new structure and optimized some code paths. * Process Manager (Ps) o Created Memory Manager (Mm) functions when toucing process/memory o Created Kernel (Ke) functins when touching kernel structures and semantics. o Cached and optimized System DLL (ntdll) Loading/Mapping, so it is done only at startup. o Implemented NtOpenProcess, PsRemoveLoadImageNotifyRoutine, PsGetCurrentProcessSessionId, PsSetLegoNotifyRoutine, PsRemoveCreateThreadNotifyroutine, PsGetVersion o Rewrote Process/Thread creation and exit functions. * Memory Manager (Mm) o PEB and TEB are now properly allocated in memory, allowing 4KB granularity instead of 64KB, removed all the hacks which allowed this earlier. o Implement MmCreateKernelStack and MmDeleteKernelStack o Took out many system structures from NonPaged Pool to Paged Pool to reduce physical memory consumption. o Removed pool debugging functions in retail builds to increase the speed. o Don't allorw NtQueryVirtualMemory for kernel-mode addresses. o Fixed bug in memory mapping which caused large applications to BSOD the system. o Made the PE Loader more lenient so it can load some other valid executables. * Executive (Ex) o Fixed the lookaside functions, their macros and the way the functions were being exported. o Moved Win32k Object Registration inside Win32k. The pointers and initlization is now done when Win32k loads. * Debugging Services (Dbgk/Kd) o Implemented some Dbgk code for user-mode bugging. o Implemented modular debugging services for Bochs, GDB, etc. * Object Manager (Ob) o Implemented Fast Referencing stubs. o Rewrote ObQueryNameString. o Rewrote ObjectType creation to match the structures, flags and semantics present in NT's Object Manager, from the caller's point of view. o Implemented Object Create Information structure and semantics when capturing data from user-mode, securing and removing a lot of kernel exloits. o Fixed ObCreateObject and ObInsertObject to work like on Windows. ObCreateObject only allocates the Object, while ObInsertObject does everything else. * File System Runtime (FsRtl) o FsRtlMdlRead, FsRtlMdlReadComplete, FsMdlReadCompleteDev, FsRtlMdlWRiteComplete, FsRtlMdlWriteCompleteDev, FsRtlPrepareMdlWrite, CcMdlReadCompleted, CcMdlWriteComplete, CcMdlReadCompleteDev, FsRtlAllocateResource, FsRtlIsPagingFile, FsRtlBalanceReads * Security Subsystem (Se) o Implemented SeCreateAccessState, SeDeleteAccessState, SeSetAccessStateGenericMapping based on a patch fy Javier M. Mellid. (Thomas_Weidenmueller) * Implemented RtlHashUnicodeString * Moved ntdll's atom table implementation to rtl, rewrote it to use proper structures and share the generic implementation between ntoskrnl and ntdll * Updated the rtl handle table implementation to use proper structures. Reserved handles are not yet supported correctly. Kernel Mode Drivers Input Device * i8042prt driver by tinus. ACPI * Implement IRP_MN_QUERY_RESOURCE_REQUIREMENTS (Hervé Poussineau) PCI * Report correct list in IRP_MN_QUERY_RESOURCE_REQUIREMENTS (Hervé Poussineau) SERIAL (Hervé Poussineau) * serial.sys driver completed except control handflow (XON/XOFF) * serenum.sys driver completed. It enumerates mice plugged on serial ports USB * UHCI HCD driver supports recognizing Memory type of resource (Aleksey_Bragin) (Hervé Poussineau) * UHCI controller driver, that uses Cromwell USB stack * Basic USB hub driver, that sometimes reports connected devices VFAT * Implement FSCTL_IS_VOLUME_DIRTY and FSCTL_MARK_VOLUME_DIRTY (Hervé Poussineau) VGA * Implement IOCTL_VIDEO_QUERY_AVAIL_MODES, IOCTL_VIDEO_QUERY_CURRENT_MODE, IOCTL_VIDEO_QUERY_NUM_AVAIL_MODES (Hervé Poussineau) * VMWare 5 video driver works (GvG) and (Filip_Navara) * Qemu video drive works (GvG), (Alex_Ionescu), and (Filip_Navara) Networking * Implemented WSAStringToAddressA and WSAStringToAddressW in ws2_32 (Magnus Olsen) * Added dhcp service and make it start. (Art_Yerkes) (WaxDragon) * Implement get* integer reading. * Properly implement ipv4addrs (validates a set of IPv4 addresses) * Limit returned DNS servers to 1 until we fix iphlpapi. Session Manager (SMSS) * Removed loading of the kernel mode part of Win32 emulator (win32k.sys). * Removed running winlogon.exe Win32(tm) Personality User mode subsystem server (CSRSS) * Added loading of the kernel mode part of the Win32 emulator (win32k.sys); * Run winlogon.exe * Implemented EnumSystemLocalesW (Aleksey_Bragin) Kernel mode subsystem server (WIN32K) (Magnus Olsen) * Implement NtGdiDdCanCreateSurface and NtGdiDdBlt for DirectX (untested) * StrechBitBlt for all Bpp (not full implemet use the code as ref) * partially implemented fullscreen in changedisplay setting I can play winquake in fullscreen now :) * Implement NtGdiGetSystemPaletteUse and NtGdiSetSystemPaletteUse this code have been taken from wine * fix winquake color glitc bug the text are now whie instead for black (Magnus Olsen) * Implement NtGdiAnimatePalette (partly ripped from Wine does not anime 100% of the palette) (Hervé Poussineau) * Speed optimze the bitblt By (Magnus Olsen), (Gregor_Anich), (Alex_Ionescu), (GvG) see svn log * Repair GDI handle debugging functionality. (Filip_Navara) * Implement NtGdiUnrealizedObject (James_Tabor) * Implement WH_KEYBOARD_LL hook (GvG) Win32(tm) Libraries ADVAPI32 (Eric_Kohl) * Implemented LockServiceDatabase, UnlockServiceDatabase, ControlService, QueryServiceStatus. * Implemented OpenSCManagerA, OpenServiceA, OpenServiceW and QueryServiceStatus (Thomas_Weidenmueller) * Ported BuildTrusteeWithObjectsAndName and BuildTrusteeWithObjectsAndSid from wine. * Implemented RegOpenCurrentUser * Implemented OpenAndMapFileForRead, RetreiveFileSecurity, StampFileSecurity, TakeOwnershipOfFile and UnmapAndCloseFile. * Implemented RegOpenUserClassesRoot * Implemented IsTokenRestricted(), inspired by a patch to winehq by James Hawkins * Implemented TokenRestrictedSids KERNEL32 * Implemented GetCommProperties, ClearCommError, CommConfigDialogA, CommConfigDialogW (Saveliy Tretiakov) * Implemented GetCommConfig, SetCommConfig, FindFirstFileExW (Dmitry Philippov) SETUPAPI (Hervé Poussineau) * Work on devices enumeration by implementing: o CM_Enumerate_Classes(_Ex) o SetupDiCreateDeviceInfoA o SetupDiCreateDeviceInfoListExW o SetupDiEnumDeviceInfo o SetupDiGetActualSectionToInstallA o SetupDiGetClassDescriptionExA o SetupDiGetClassDevs(Ex)A/W o SetupDiGetDeviceInterfaceDetailA/W o SetupDiGetDeviceRegistryPropertyA/W (Eric_Kohl) * Implemented: o ConcatenatePaths o MyGetFileTitle o GetVersionInfoFromImage o StringTableDuplicate DirectX Dorp last sync with wine and userhook Before we can sync with wine we need to rewrite the enum and reg of device so it working fine in reactos and windows. Add svn rev 15043 + only userhooks for the mouse at last the mouse are working in tribles windows mode but in full screen it is bit chopy. Not tested tribles in reactos with the new code. Userenv David Nordenberg: Swedish translation, proofread by Andreas Bjerkeholt Libraries shared with Wine (GvG) * Sync to Wine-20050419 * Sync to Wine-20040419 * Sync to Wine-20050524 * Sync to Wine-20050628 USER32 * Implementation of DragDetect. Based on Wine code (C) 1993, 1994 Alexandre Julliard. (Filip_Navara) (James_Tabor) * Implemented: o TrackMouseEvent. o NtUserGetAsyncKeyState and support for TrackMouseEvent. o DrawMenuBar. o CheckMenuRadioItem. Not fully tested. o GetMenuString A & W. Not fully tested. o ModifyMenu A & W. o MDICascade, MDITile and WIN_ListChildren Win32(tm) Applications IPConfig (Tim Jobling <tjob800(a)yahoo.co.uk>) * Relicense to GPL. * Display NodeType with meaningfull Human readable names. * Exclusively use TCHAR strings. * Display Physical Address, DHCP enabled state, IP Addresses/Netmasks, * Default Gateway, DHCP server and DHCP Lease times. * Parse command line options. * Default to only showing the IP/SM/DG is no options specified * Handel option: /All and /? * Display message about all unimplemented options. * Changed C++ style commenting to C style CMD (Magnus Olsen) * Remove all hardcode string to own rc file. * Cache codepage instead call on win32 api for it whole time. * Add *.* syntax to Dir * Bugfix Color now it work as it should * Bugfix CD "program files" so it works * Implement CD s* * translate %errorlevel% to a value when it pass at command line. Now is errorlevel implement as it should. left todo check all cmd command that they are setting right value * fix a small bug in choice.c so it does print out choice "sadsad" right * add %time% and %cd% internal value. Example echo %cd% or echo %time%. * add %DATE% example echo %date% are working now. Bugfix date so it print out the week days names. * adding %RANDOM% example how to use it echo %random% * adding %cmdcmdline%, example how to use it echo %CMDCMDLINE% * add %CMDEXTVERSION% example to use it echo %CMDEXTVERSION% the value are hardcode to 2. for it is that value ms win2k / winxp report back. thx arty to test it in win xp. * quick dirty hack getting our copy working with 1>null by me, thanks to Hartex and Brandon to hunt down where the bug was. (Frik85) * Add Help command (By Frik85) * Add german language resource to the ReactOS Command Processor (not completely finished, I will update it as soon as possible) Martin Rottensteiner (2005only(a)pianonote.at): * set errorlevel to 9009 if command not found * implemented exit /b # in batchfiles (Brandon_Turner) * Added exclusive deletion "del * -abc.txt -text*.txt" * Fixed bug to allow MS style wildcards + code clean up added /y and /-y in move.c * simple check to fix > and | bug with 'rem' * Implemented /A in delete, example "del /A:H /A:-R *.exe -ping.exe" * Bug fix color.c, now it is working as ms cmd color. * Fix bug "mv foo.txt c:\temp gives you c:\tempfoo.txt" TASKMGR * Remove some hardcode string tested by Harteex (Filip_Navara) EXPLORER * Stepwise taskbar resizing (charn <charn89(a)hotmail.com>) * Swedish translation ([[David Nordenberg (dnordenberg [at] users.sourceforge.net) ]]) * option to build Explorer as ROS Shell without integrated explorer part (Martin Fuchs) * display of custom folders in start menu root (Martin Fuchs) * fix of listbox insert algorithmus (Martin Fuchs) * Czech translation of Explorer (Luk _ "denzil" Frolka <d3nzil(a)gmail.com>) * Russian Translation (Done by Dmitry Philippov, checked by fireball, DarkHobbit and others) * French Translation (hpoussin) * split of big explorer resource file into smaller language specific rescource scripts (Martin Fuchs) * Implemented part of screensaver functions, get values from reg, and show screensaver (<sikker2004(a)yahoo.com>) WINEFILE * Czech translation (Denzil <d3nzil(a)gmail.com>) * Swedish translation (David Nordenberg, proofread by Andreas Bjerkeholt) * implementation of owner drawn context menus (Martin Fuchs) * from WINE: add czech and swedish resource files (Martin Fuchs) * from WINE: Change SUBLANG_DEFAULT to SUBLANG_NEUTRAL for LANG_SPANISH in all resources, so that Spanish locales other than Spain also use Spanish resources. (Alex Villacís Lasso <a_villacis(a)palosanto.com>) * from WINE: Spanish translation (José Manuel Ferrer Ortiz <jmfo1982(a)yahoo.es>) * from WINE: Update of Portuguese translation (Marcelo Duarte <wine-devel(a)bol.com.br>, Am‚rico Jos‚ Melo <mmodem00(a)netvisao.pt>, Francois Gouget <fgouget(a)codeweavers.com>) * from WINE: Update of German translation. (Henning Gerhardt <henning.gerhardt(a)web.de>) * display source path in "move file" dialog (Martin Fuchs) * network connect and disconnect dialogs (Martin Fuchs) * "format disk" dialog (Martin Fuchs) * display free and total disk space (Martin Fuchs) * switching of file sort order (Martin Fuchs) * from WINE: add polish resource file (Martin Fuchs) * window refresh in shell mode (Martin Fuchs) * implement file filtering (matching file name patterns and file types) (Martin Fuchs) * refresh display after executing a context menu command (Martin Fuchs) * file copy, move and delete commands (Martin Fuchs) * file properties dialog ([Martin Fuchs], partly based on Rob D.'s winfile code) * from WINE: Update German resource files (Henning Gerhardt <henning.gerhardt(a)web.de>) * switch to WIN32 API string functions (Martin Fuchs) * Updated winefile French resources (Jonathan Ernst <Jonathan(a)ErnstFamily.ch>) * Sync source code and resource scripts between WINE and ROS WINE and switch to UNICODE compilaton in Wine (Martin Fuchs)

4 3

Problems bochs and reactos
by Jeroen van Iddekinge 22 Aug '05

22 Aug '05

Hi all, I downloaded ReactOS 0.2.7 iso. THe installation whent fine, but after booting the system stops and bochs keeps outputting the following messages: 00311505974i[CPU0 ] allow_io(): IO addr 3c4 (len 1) outside TSS IO permission map (base=ffff, limit=2068) #GP(0) 00311506230i[CPU0 ] allow_io(): IO addr 3c5 (len 1) outside TSS IO permission map (base=ffff, limit=2068) #GP(0) 00311506532i[CPU0 ] allow_io(): IO addr cf8 (len 4) outside TSS IO permission map (base=ffff, limit=2068) #GP(0) 00311506815i[CPU0 ] allow_io(): IO addr cfc (len 4) outside TSS IO permission map (base=ffff, limit=2068) #GP(0) what whrong? The bochsrc.txt is: ata0:enabled=1, ioaddr1=0x1f0,ioaddr2=0x3f0,irq=14 ata1:enabled=1, ioaddr1=0x170,ioaddr2=0x370,irq=15 ata0-master: type=disk, path="c.img", mode=flat, cylinders=1040, heads=16, spt=63 ata1-master: type=cdrom,path=ReactOS.iso,status=inserted megs:64 romimage:file=/usr/share/bochs/BIOS-bochs-latest, address=0xf0000 vgaromimage:file=/usr/share/bochs/VGABIOS-lgpl-latest-cirrus vga: extension=cirrus boot:c config_interface:textconfig display_library:sdl log:bochs.log debug:action=ignore info:action=report error:action=report panic:action=ask vga_update_interval: 250000 keyboard_serial_delay: 200 keyboard_paste_delay: 100000 ips: 10000000 Regards, Jer

1 1

Re: [ros-diffs] [weiden] 17474: added a macro IsKernelPointer() to test whether a pointer value points to the kernel address space. This is needed because on IA-64 the MSB is not necessarily set for pointers to the kernel address space.
by Filip Navara 22 Aug '05

22 Aug '05

weiden(a)svn.reactos.com wrote: >added a macro IsKernelPointer() to test whether a pointer value points to the kernel address space. This is needed because on IA-64 the MSB is not necessarily set for pointers to the kernel address space. > >Modified: trunk/reactos/ntoskrnl/include/internal/ntoskrnl.h >Modified: trunk/reactos/ntoskrnl/ob/wait.c > > > ------------------------------------------------------------------------ > *Modified: trunk/reactos/ntoskrnl/include/internal/ntoskrnl.h* > >--- trunk/reactos/ntoskrnl/include/internal/ntoskrnl.h 2005-08-22 10:51:05 UTC (rev 17473) >+++ trunk/reactos/ntoskrnl/include/internal/ntoskrnl.h 2005-08-22 13:38:30 UTC (rev 17474) >@@ -147,8 +147,26 @@ > > #define ProbeForReadLargeInteger(Ptr) ((LARGE_INTEGER)ProbeForReadGenericType(&(Ptr)->QuadPart, LONGLONG, 0)) > #define ProbeForReadUlargeInteger(Ptr) ((ULARGE_INTEGER)ProbeForReadGenericType(&(Ptr)->QuadPart, ULONGLONG, 0)) > > > >+/* >+ * Use IsKernelPointer to test whether a pointer points to the kernel address >+ * space >+ */ >+#if defined(_X86_) || defined(_M_AMD64) > > > > > >+/* for x86 and x86-64 the MSB is 1 so we can simply test on that */ >+#define IsKernelPointer(Ptr) ((LONG_PTR)(Ptr) < 0) >+ >+#elif defined(_IA64_) >+ >+/* on Itanium if the 24 most significant bits are set, we're not dealing with >+ user mode pointers. */ >+#define IsKernelPointer(Ptr) (((ULONG_PTR)(Ptr) & 0xFFFFFF0000000000ULL) != 0) >+ >+#else >+#error IsKernelPointer() needs to be defined for this architecture > > > #endif > > >+ >+#endif > > This looks really really wrong ... the macro will incorrectly evaluate to TRUE for user mode pointers on 3GB kernels. I don't understand the purpose of the code, but why don't you just check for Ptr >= MM_SYSTEM_RANGE_START ? - Filip

2 1

dma problem
by Hartmut Birr 22 Aug '05

22 Aug '05

Hi, it seems that the routines for allocating a dma adapter object are broken. For a pci cards, the channel is always -1. The adapter object is stored at HalpEisaAdapter[-1]. It does overwrite something. This adapter object is used for all pci cards (scsi controller, network cards, ...). Usually every pci card must have its own adapter object. - Hartmut

2 1

include file problem
by Hartmut Birr 20 Aug '05

20 Aug '05

Hi, I need for some modifications the definitions of the resource structures (CM_FULL_RESOURCE_DESCRIPTOR, ..) in usetup. The structures are defined in winddk.h. It seems that it isn't possible to include winndk.h if ntndk.h is already included. I get errors like this: ... include/ndk/extypes.h:37:1: warning: "SEMAPHORE_QUERY_STATE" redefined In file included from w32api/include/ddk/ntddk.h:67, from subsys\system\usetup\usetup.h:38: w32api/include/ddk/winddk.h:342:1: warning: this is the location of the previous definition In file included from include/ndk/ntndk.h:43, from subsys\system\usetup\usetup.h:41: include/ndk/cmtypes.h:23: error: redeclaration of `enum _KEY_INFORMATION_CLASS' include/ndk/cmtypes.h:24: error: conflicting types for 'KeyBasicInformation' ... Our headers are broken. - Hartmut

2 1

RE: [ros-dev] Website Design Voting
by Murphy, Ged (Bolton) 19 Aug '05

19 Aug '05

This has been 2 weeks now. Can we finalise it and start working towards getting the new site up please? Thanks, Ged. > -----Original Message----- > From: ros-dev-bounces(a)reactos.com [mailto:ros-dev-bounces@reactos.com] On Behalf Of Freeworld > Sent: 4. august 2005 23:28 > To: ros-dev(a)reactos.com; ros-general(a)reactos.com; ros-web(a)reactos.com > Subject: [ros-dev] Website Design Voting > > Hello folks! > > Things are going very well with the new site and it is time that the > community/developers choose a design for our now homepage. The choices: > (Thanks to Klemens Friedl for carrying the suggestions together and for > all his other work.) > > 1) Jason Filby: > http://reactos.com/newsite/reactos_index.html > <http://reactos.com/newsite/reactos_index.html> > > 2) Klemens Friedl: > http://frik85.fr.funpic.de/reactos/ <http://frik85.fr.funpic.de/reactos/> > > 3) Mikko Tikkanen (created the layaut, only a pic) > http://devnet.dnsalias.net/guidesign/web_mockup.jpg > <http://devnet.dnsalias.net/guidesign/web_mockup.jpg> (Mikko Tikkanen > design) > Brandon Turner (created a html version): > http://brandonturner.org/ROS/index.html > <http://brandonturner.org/ROS/index.html> (Brandon Turner work) > > 4) Mindflyer (based on mikko's design): > http://www.mufunyo.net/reactos/web/ > <http://www.mufunyo.net/reactos/web/> (currently not available!) > > 5) jh: > http://reactos.com:8080/pipermail/ros-web/attachments/20050614/624eb1e9/ros- 0001.png > <http://reactos.com:8080/pipermail/ros-web/attachments/20050614/624eb1e9/ros -0001.png> > > 6) Gennadiy Dolzhenko: > http://optim.sorb.com.ua/ros > http://optim.sorb.com.ua/ros2 > > Allowed votes are: > 1) Jason Filby > 2) Klemens Friedl > 3) Mikko Tikkanen > 4) Mindflyer > 5) jh > 6) Gennadiy Dolzhenko 1 > 7) Gennadiy Dolzhenko 1 > > If the designers name is not right or you want the respective > mailinglist-posting please send me an email. > > The voting will last for two weeks. > > Greetings > Michael Wirth ************************************************************************ The information contained in this message or any of its attachments is confidential and is intended for the exclusive use of the addressee. The information may also be legally privileged. The views expressed may not be company policy, but the personal views of the originator. If you are not the addressee, any disclosure, reproduction, distribution or other dissemination or use of this communication is strictly prohibited. If you have received this message in error, please contact postmaster(a)exideuk.co.uk <mailto:postmaster@exideuk.co.uk> and then delete this message. Exide Technologies is an industrial and transportation battery producer and recycler with operations in 89 countries. Further information can be found at www.exide.com

4 3

Re: [ros-svn] [gvg] 17434: Import VMware open source BusLogic driver
by Alex Ionescu 19 Aug '05

19 Aug '05

gvg(a)svn.reactos.com wrote: >Hartmut made it work in ReactOS. > > I hope Hartmut made ReactOS work with *it*. We don't have the liberty of patching every single NT driver that we will want to run on ROS :) Best regards, Alex Ionescu

2 1

Re: [ros-dev] Website Design Voting
by reactos 18 Aug '05

18 Aug '05

I think: just go ahead and publish it as fast as you can, the world has been waiting long enough. By the way, really nice result, it's simple and nice! Yours sincerely, Jaix Bly ----Ursprungligt meddelande----- From: Freeworld michael(a)freeworld.net Date: Thu, 18 Aug 2005 15:47:11 +0200 To: ReactOS Development List ros-dev(a)reactos.com Subject: Re: [ros-dev] Website Design Voting > Yes you're right. > > The votin is closed now. And the winner is... > > Klemens Friedl > http://frik85.fr.funpic.de/reactos/ > > He's got far the most votes. We're heavily working on the new site and > hope to get it up soon. What do you think of releasing it concurrent to > ReactOS 0.3? > > Greetings > Michael > > Murphy, Ged (Bolton) wrote: > > >This has been 2 weeks now. > >Can we finalise it and start working towards getting the new site up please? > > > >Thanks, > >Ged. > > > > > > > > > >>-----Original Message----- > >>From: ros-dev-bounces(a)reactos.com [mailto:ros-dev-bounces@reactos.com] On > >> > >> > >Behalf Of Freeworld > > > > > >>Sent: 4. august 2005 23:28 > >>To: ros-dev(a)reactos.com; ros-general(a)reactos.com; ros-web(a)reactos.com > >>Subject: [ros-dev] Website Design Voting > >> > >>Hello folks! > >> > >>Things are going very well with the new site and it is time that the > >>community/developers choose a design for our now homepage. The choices: > >>(Thanks to Klemens Friedl for carrying the suggestions together and for > >>all his other work.) > >> > >>1) Jason Filby: > >> http://reactos.com/newsite/reactos_index.html > >><http://reactos.com/newsite/reactos_index.html> > >> > >>2) Klemens Friedl: > >> http://frik85.fr.funpic.de/reactos/ > >> > >> > ><http://frik85.fr.funpic.de/reactos/> > > > > > >>3) Mikko Tikkanen (created the layaut, only a pic) > >> http://devnet.dnsalias.net/guidesign/web_mockup.jpg > >><http://devnet.dnsalias.net/guidesign/web_mockup.jpg> (Mikko Tikkanen > >> design) > >> Brandon Turner (created a html version): > >> http://brandonturner.org/ROS/index.html > >><http://brandonturner.org/ROS/index.html> (Brandon Turner work) > >> > >>4) Mindflyer (based on mikko's design): > >> http://www.mufunyo.net/reactos/web/ > >><http://www.mufunyo.net/reactos/web/> (currently not available!) > >> > >>5) jh: > >> > >> > >> > >http://reactos.com:8080/pipermail/ros-web/attachments/20050614/624eb1e9/ros- > >0001.png > > > > > ><http://reactos.com:8080/pipermail/ros-web/attachments/20050614/624eb1e9/ros > >-0001.png> > > > > > >>6) Gennadiy Dolzhenko: > >> http://optim.sorb.com.ua/ros > >> http://optim.sorb.com.ua/ros2 > >> > >>Allowed votes are: > >>1) Jason Filby > >>2) Klemens Friedl > >>3) Mikko Tikkanen > >>4) Mindflyer > >>5) jh > >>6) Gennadiy Dolzhenko 1 > >>7) Gennadiy Dolzhenko 1 > >> > >>If the designers name is not right or you want the respective > >>mailinglist-posting please send me an email. > >> > >>The voting will last for two weeks. > >> > >>Greetings > >>Michael Wirth > >> > >> > > > >************************************************************************ > >The information contained in this message or any of its > >attachments is confidential and is intended for the exclusive > >use of the addressee. The information may also be legally > >privileged. The views expressed may not be company policy, > >but the personal views of the originator. If you are not the > >addressee, any disclosure, reproduction, distribution or other > >dissemination or use of this communication is strictly prohibited. > >If you have received this message in error, please contact > >postmaster(a)exideuk.co.uk > ><mailto:postmaster@exideuk.co.uk> and then delete this message. > > > >Exide Technologies is an industrial and transportation battery > >producer and recycler with operations in 89 countries. > >Further information can be found at www.exide.com > > > > > >_______________________________________________ > >Ros-dev mailing list > >Ros-dev(a)reactos.com > >http://reactos.com:8080/mailman/listinfo/ros-dev > > > > > > > > > _______________________________________________ > Ros-dev mailing list > Ros-dev(a)reactos.com > http://reactos.com:8080/mailman/listinfo/ros-dev

1 0

RE: [ros-dev] Website Design Voting
by Murphy, Ged (Bolton) 18 Aug '05

18 Aug '05

Freeworld wrote: > What do you think of releasing it > concurrent to ReactOS 0.3? Do you mean 0.2.7 ? 0.3 could be another year away ! Ged. ************************************************************************ The information contained in this message or any of its attachments is confidential and is intended for the exclusive use of the addressee. The information may also be legally privileged. The views expressed may not be company policy, but the personal views of the originator. If you are not the addressee, any disclosure, reproduction, distribution or other dissemination or use of this communication is strictly prohibited. If you have received this message in error, please contact postmaster(a)exideuk.co.uk <mailto:postmaster@exideuk.co.uk> and then delete this message. Exide Technologies is an industrial and transportation battery producer and recycler with operations in 89 countries. Further information can be found at www.exide.com

1 0

0.2.7 Status
by Brandon Turner 17 Aug '05

17 Aug '05

Hey Robert, How are things looking for 0.2.7? Should be expecting it soon? http://www.reactos.com/wiki/index.php/ChangeLog-0.2.7 is still looking kinda bare, we might want to make sure that gives a better repersentation of 0.2.7 when it comes out. Brandon

5 4

RE: [ros-svn] [ion] 17417: Remove all non-official LPCstructures/defines/hardcoded hacks, and use actual correct sizes and structures.
by Ge van Geldorp 17 Aug '05

17 Aug '05

> From: ion(a)svn.reactos.com > > Remove all non-official LPC structures/defines/hardcoded > hacks, and use actual correct sizes and structures. This breaks a simple boot. ntoskrnl/include/internal/port.h previously defined QUEUEDMESSAGE as: typedef struct _QUEUEDMESSAGE { PEPORT Sender; LIST_ENTRY QueueListEntry; PORT_MESSAGE Message; UCHAR MessageData [MAX_MESSAGE_DATA]; } QUEUEDMESSAGE, *PQUEUEDMESSAGE; r17417 removed the MessageData member. Now in ntoskrnl/lpc/reply.c function EiReplyOrRequestPort() line 52: memcpy(&MessageReply->Message, LpcReply, LpcReply->u1.s1.TotalLength); writes outside allocated memory (MessageReply is a PQUEUEDMESSAGE, LpcReply->u1.s1.TotalLength is 292). This causes a subsequent ExFreePool to generate a page fault. Gé van Geldorp.

1 0

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