Ros-dev

ros-dev@reactos.org

1 participants
5793 discussions

Cache / Memory Manager / FileSystemDrivers
by Aleksey Bragin 08 Jan '09

08 Jan '09

Hello, I recently spent some time experimenting with various Cc rewrites we have in our tree, and before that I extensively tested our fastfat driver in a real NT environment (MS Windows 2003). The conclusion of the above work is that our fastfat driver either needs serious bugfixing, or a rewrite is needed. It corrupts directory tables in a real NT, leads to unusual behaviour of cc- rewrite branch, and may have side effects on arty's newcc. Before this is done (bugfixing or rewriting, developing and testing against a Windows 2003 at least, not ReactOS), it's meaningless to continue any other work on Cc and related Mm parts. WBR, Aleksey Bragin.

9 14

[ros-diffs] [hyperion] 38621: PSEH can't make code exception-safe (yet). Add explicit "volatile"s where it might be necessary
by Dmitry Gorbachev 07 Jan '09

07 Jan '09

Maybe better to use a macro, something like _SEH2_VOLATILE, instead of a plain "volatile"?

1 0

[ros-diffs] [jimtabor] 38590: - Patch by Michael Martin: Fix for Get/Set CallProc.
by James Tabor 06 Jan '09

06 Jan '09

Hi all! Here is the test program provided by BugBoy. It also shows another bug, see bug 3971. Please test before and after the patch. Thanks, James PS Good Job, Michael!

1 0

questions about mmc?
by Jack Yi 05 Jan '09

05 Jan '09

Hi,guys I'm just a fan of ReactOS,now I have two questions. 1, Is(or will be) there an disk management utility for ReactOS, or in another way, will we make an implementation of Microsoft diskmgmt.msc? 2, Will ReactOS group make an implementation of Microsoft Management Console for ReactOS? Thank you!

1 0

Re: [ros-dev] [ros-diffs] [fireball] 38552: - Revert changes to NtSetVolumeInformationFile made in 38550, it should indeed use IoGetRelatedDeviceObject, not target device.
by Alex Ionescu 04 Jan '09

04 Jan '09

Yeah, sorry, the target device should be used to send a TARGET_DEVICE_CUSTOM_NOTIFICAITON of type GUID_IO_VOLUME_CHANGE at the end of the function (IoReportTargetDeviceChange). FYI, Pierre, when you implement the notification system. On 4-Jan-09, at 5:12 AM, fireball(a)svn.reactos.org wrote: > Author: fireball > Date: Sun Jan 4 04:12:38 2009 > New Revision: 38552 > > URL: http://svn.reactos.org/svn/reactos?rev=38552&view=rev > Log: > - Revert changes to NtSetVolumeInformationFile made in 38550, it > should indeed use IoGetRelatedDeviceObject, not target device. > > Modified: > trunk/reactos/ntoskrnl/io/iomgr/iofunc.c > > Modified: trunk/reactos/ntoskrnl/io/iomgr/iofunc.c > URL: http://svn.reactos.org/svn/reactos/trunk/reactos/ntoskrnl/io/iomgr/iofunc.c… > = > = > = > = > = > = > = > = > ====================================================================== > --- trunk/reactos/ntoskrnl/io/iomgr/iofunc.c [iso-8859-1] (original) > +++ trunk/reactos/ntoskrnl/io/iomgr/iofunc.c [iso-8859-1] Sun Jan 4 > 04:12:38 2009 > @@ -3307,8 +3307,7 @@ > } > > /* Get the device object */ > - Status = IoGetRelatedTargetDevice(FileObject, &DeviceObject); > - if (!NT_SUCCESS(Status)) return Status; > + DeviceObject = IoGetRelatedDeviceObject(FileObject); > > /* Clear File Object event */ > KeClearEvent(&FileObject->Event); > Best regards, Alex Ionescu

1 0

Re: [ros-dev] [ros-diffs] [jimtabor] 38518: - Removed SEH abuse and add notes for the hook code, in NtUserDefSetText. - Update NtUserCallHwndLock subfunctions.
by Timo Kreuzer 03 Jan '09

03 Jan '09

SEH is still needed. SafeText doesn't really deserve it's name, as it's only a safe copy of the UNICODE_STRING structure returned by ProbeForReadUnicodeString(), but with the still unsafe string buffer. Also the Buffer was never probed (ProbeForReadUnicodeString only checks the UNICODE_STRING and copies it) IMO the function is dangerous, as it implies that the Buffer was probed, too. Timo jimtabor(a)svn.reactos.org schrieb: > Author: jimtabor > Date: Fri Jan 2 22:02:54 2009 > New Revision: 38518 > > URL: http://svn.reactos.org/svn/reactos?rev=38518&view=rev > Log: > - Removed SEH abuse and add notes for the hook code, in NtUserDefSetText. > - Update NtUserCallHwndLock subfunctions. > > Modified: > trunk/reactos/subsystems/win32/win32k/ntuser/simplecall.c > trunk/reactos/subsystems/win32/win32k/ntuser/window.c > > ... > > Modified: trunk/reactos/subsystems/win32/win32k/ntuser/window.c > URL: http://svn.reactos.org/svn/reactos/trunk/reactos/subsystems/win32/win32k/nt… > ============================================================================== > --- trunk/reactos/subsystems/win32/win32k/ntuser/window.c [iso-8859-1] (original) > +++ trunk/reactos/subsystems/win32/win32k/ntuser/window.c [iso-8859-1] Fri Jan 2 22:02:54 2009 > @@ -4569,57 +4569,47 @@ > } > Wnd = Window->Wnd; > > - if(SafeText.Length != 0) > - { > - _SEH2_TRY > - { > - if (Wnd->WindowName.MaximumLength > 0 && > - SafeText.Length <= Wnd->WindowName.MaximumLength - sizeof(UNICODE_NULL)) > - { > - ASSERT(Wnd->WindowName.Buffer != NULL); > - > - Wnd->WindowName.Length = SafeText.Length; > - Wnd->WindowName.Buffer[SafeText.Length / sizeof(WCHAR)] = L'\0'; > - RtlCopyMemory(Wnd->WindowName.Buffer, > - SafeText.Buffer, > - SafeText.Length); > - } > - else > - { > - PWCHAR buf; > - Wnd->WindowName.MaximumLength = Wnd->WindowName.Length = 0; > - buf = Wnd->WindowName.Buffer; > - Wnd->WindowName.Buffer = NULL; > - if (buf != NULL) > - { > - DesktopHeapFree(Wnd->pdesktop, > - buf); > - } > - > - Wnd->WindowName.Buffer = DesktopHeapAlloc(Wnd->pdesktop, > - SafeText.Length + sizeof(UNICODE_NULL)); > - if (Wnd->WindowName.Buffer != NULL) > - { > - Wnd->WindowName.Buffer[SafeText.Length / sizeof(WCHAR)] = L'\0'; > - RtlCopyMemory(Wnd->WindowName.Buffer, > - SafeText.Buffer, > - SafeText.Length); > - Wnd->WindowName.MaximumLength = SafeText.Length + sizeof(UNICODE_NULL); > - Wnd->WindowName.Length = SafeText.Length; > - } > - else > - { > - SetLastWin32Error(ERROR_NOT_ENOUGH_MEMORY); > - Ret = FALSE; > - } > - } > - } > - _SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER) > - { > - SetLastNtError(_SEH2_GetExceptionCode()); > - Ret = FALSE; > - } > - _SEH2_END; > + if (SafeText.Length != 0) > + { > + if (Wnd->WindowName.MaximumLength > 0 && > + SafeText.Length <= Wnd->WindowName.MaximumLength - sizeof(UNICODE_NULL)) > + { > + ASSERT(Wnd->WindowName.Buffer != NULL); > + > + Wnd->WindowName.Length = SafeText.Length; > + Wnd->WindowName.Buffer[SafeText.Length / sizeof(WCHAR)] = L'\0'; > + RtlCopyMemory(Wnd->WindowName.Buffer, > + SafeText.Buffer, > + SafeText.Length); > + } > + else > + { > + PWCHAR buf; > + Wnd->WindowName.MaximumLength = Wnd->WindowName.Length = 0; > + buf = Wnd->WindowName.Buffer; > + Wnd->WindowName.Buffer = NULL; > + if (buf != NULL) > + { > + DesktopHeapFree(Wnd->pdesktop, buf); > + } > + > + Wnd->WindowName.Buffer = DesktopHeapAlloc(Wnd->pdesktop, > + SafeText.Length + sizeof(UNICODE_NULL)); > + if (Wnd->WindowName.Buffer != NULL) > + { > + Wnd->WindowName.Buffer[SafeText.Length / sizeof(WCHAR)] = L'\0'; > + RtlCopyMemory(Wnd->WindowName.Buffer, > + SafeText.Buffer, > + SafeText.Length); > + Wnd->WindowName.MaximumLength = SafeText.Length + sizeof(UNICODE_NULL); > + Wnd->WindowName.Length = SafeText.Length; > + } > + else > + { > + SetLastWin32Error(ERROR_NOT_ENOUGH_MEMORY); > + Ret = FALSE; > + } > + } > } > else > { > @@ -4628,6 +4618,9 @@ > Wnd->WindowName.Buffer[0] = L'\0'; > } > > + // HAX! FIXME! Windows does not do this in here! > + // In User32, these are called after: NotifyWinEvent EVENT_OBJECT_NAMECHANGE than > + // RepaintButton, StaticRepaint, NtUserCallHwndLock HWNDLOCK_ROUTINE_REDRAWFRAMEANDHOOK, etc. > /* Send shell notifications */ > if (!IntGetOwner(Window) && !IntGetParent(Window)) > { > > >

3 3

Re: [ros-dev] [ros-diffs] [pschweitzer] 38523: - Implemeted Io(p)GetRelatedTargetDevice - Fixed FsRtlNotifyVolumeEvent
by Alex Ionescu 03 Jan '09

03 Jan '09

You can/should use IopGetDeviceNode instead of the ugly (but necessary) triple-casting you have. > + *DeviceNode = (PDEVICE_NODE) > ((PEXTENDED_DEVOBJ_EXTENSION)DeviceRelations->Objects[0]- > >DeviceObjectExtension)->DeviceNode; Also, I believe there is a bug in Nt/IoQuery/SetVolumeInformation -- we call IoGetRelatedDeviceObject instead of IoGetRelatedTargetDevice, because the function was missing. Good job! On 3-Jan-09, at 4:35 AM, pschweitzer(a)svn.reactos.org wrote: > Author: pschweitzer > Date: Sat Jan 3 03:35:10 2009 > New Revision: 38523 > > URL: http://svn.reactos.org/svn/reactos?rev=38523&view=rev > Log: > - Implemeted Io(p)GetRelatedTargetDevice > - Fixed FsRtlNotifyVolumeEvent > > Modified: > branches/pierre-fsd/ntoskrnl/fsrtl/pnp.c > branches/pierre-fsd/ntoskrnl/include/internal/io.h > branches/pierre-fsd/ntoskrnl/io/iomgr/device.c > > Modified: branches/pierre-fsd/ntoskrnl/fsrtl/pnp.c > URL: http://svn.reactos.org/svn/reactos/branches/pierre-fsd/ntoskrnl/fsrtl/pnp.c… > = > = > = > = > = > = > = > = > ====================================================================== > --- branches/pierre-fsd/ntoskrnl/fsrtl/pnp.c [iso-8859-1] (original) > +++ branches/pierre-fsd/ntoskrnl/fsrtl/pnp.c [iso-8859-1] Sat Jan 3 > 03:35:10 2009 > @@ -43,8 +43,7 @@ > PDEVICE_OBJECT DeviceObject = NULL; > TARGET_DEVICE_CUSTOM_NOTIFICATION Notification; > > - /* FIXME: We should call IoGetRelatedTargetDevice here */ > - DeviceObject = IoGetRelatedDeviceObject(FileObject); > + IoGetRelatedTargetDevice(FileObject, &DeviceObject); > if (DeviceObject) > { > Notification.Version = 1; > > Modified: branches/pierre-fsd/ntoskrnl/include/internal/io.h > URL: http://svn.reactos.org/svn/reactos/branches/pierre-fsd/ntoskrnl/include/int… > = > = > = > = > = > = > = > = > ====================================================================== > --- branches/pierre-fsd/ntoskrnl/include/internal/io.h [iso-8859-1] > (original) > +++ branches/pierre-fsd/ntoskrnl/include/internal/io.h [iso-8859-1] > Sat Jan 3 03:35:10 2009 > @@ -669,6 +669,12 @@ > IN BOOLEAN ForceUnload > ); > > +NTSTATUS > +NTAPI > +IoGetRelatedTargetDevice(IN PFILE_OBJECT FileObject, > + OUT PDEVICE_OBJECT *DeviceObject > +); > + > // > // IRP Routines > // > > Modified: branches/pierre-fsd/ntoskrnl/io/iomgr/device.c > URL: http://svn.reactos.org/svn/reactos/branches/pierre-fsd/ntoskrnl/io/iomgr/de… > = > = > = > = > = > = > = > = > ====================================================================== > --- branches/pierre-fsd/ntoskrnl/io/iomgr/device.c [iso-8859-1] > (original) > +++ branches/pierre-fsd/ntoskrnl/io/iomgr/device.c [iso-8859-1] Sat > Jan 3 03:35:10 2009 > @@ -552,6 +552,49 @@ > } > } > > +NTSTATUS > +NTAPI > +IopGetRelatedTargetDevice(IN PFILE_OBJECT FileObject, > + OUT PDEVICE_NODE *DeviceNode) > +{ > + NTSTATUS Status; > + IO_STACK_LOCATION Stack; > + IO_STATUS_BLOCK IoStatusBlock; > + PDEVICE_RELATIONS DeviceRelations; > + PDEVICE_OBJECT DeviceObject = NULL; > + > + ASSERT(FileObject); > + > + /* Get DeviceObject related to given FileObject */ > + DeviceObject = IoGetRelatedDeviceObject(FileObject); > + if (!DeviceObject) > + { > + return STATUS_NO_SUCH_DEVICE; > + } > + > + /* Call the driver to query all the relations (IRP_MJ_PNP) */ > + Status = IopInitiatePnpIrp(DeviceObject, &IoStatusBlock, > + IRP_MN_QUERY_DEVICE_RELATIONS, > &Stack); > + if (NT_SUCCESS(Status)) > + { > + DeviceRelations = > (PDEVICE_RELATIONS)IoStatusBlock.Information; > + ASSERT(DeviceRelations); > + ASSERT(DeviceRelations->Count == 1); > + > + /* We finally get the device node */ > + *DeviceNode = (PDEVICE_NODE) > ((PEXTENDED_DEVOBJ_EXTENSION)DeviceRelations->Objects[0]- > >DeviceObjectExtension)->DeviceNode; > + if (!*DeviceNode) > + { > + Status = STATUS_NO_SUCH_DEVICE; > + } > + > + /* Free the DEVICE_RELATIONS structure, we don't need it > anymore */ > + ExFreePool(DeviceRelations); > + } > + > + return Status; > +} > + > /* PUBLIC FUNCTIONS > ***********************************************************/ > > /* > @@ -1205,6 +1248,26 @@ > > /* Return the DO we found */ > return DeviceObject; > +} > + > +/* > + * @implemented > + */ > +NTSTATUS > +NTAPI > +IoGetRelatedTargetDevice(IN PFILE_OBJECT FileObject, > + OUT PDEVICE_OBJECT *DeviceObject) > +{ > + NTSTATUS Status; > + PDEVICE_NODE DeviceNode = NULL; > + > + /* We call the internal function to do all the work */ > + Status = IopGetRelatedTargetDevice(FileObject, &DeviceNode); > + if (NT_SUCCESS(Status) && DeviceNode) > + { > + *DeviceObject = DeviceNode->PhysicalDeviceObject; > + } > + return Status; > } > > /* > Best regards, Alex Ionescu

2 1

Re: [ros-dev] [ros-diffs] [sginsberg] 38508: - Add Wine's GetPrivateObjectSecurity and use it instead of our that uses the unimplemented RtlQuerySecurityObject routine
by Alex Ionescu 03 Jan '09

03 Jan '09

Please don't do this -- now nobody will know that the Rtl function needs to be called! On 2-Jan-09, at 12:14 PM, sginsberg(a)svn.reactos.org wrote: > Author: sginsberg > Date: Fri Jan 2 11:14:42 2009 > New Revision: 38508 > > URL: http://svn.reactos.org/svn/reactos?rev=38508&view=rev > Log: > - Add Wine's GetPrivateObjectSecurity and use it instead of our that > uses the unimplemented RtlQuerySecurityObject routine > > Modified: > trunk/reactos/dll/win32/advapi32/sec/misc.c > > Modified: trunk/reactos/dll/win32/advapi32/sec/misc.c > URL: http://svn.reactos.org/svn/reactos/trunk/reactos/dll/win32/advapi32/sec/mis… > = > = > = > = > = > = > = > = > ====================================================================== > --- trunk/reactos/dll/win32/advapi32/sec/misc.c [iso-8859-1] > (original) > +++ trunk/reactos/dll/win32/advapi32/sec/misc.c [iso-8859-1] Fri > Jan 2 11:14:42 2009 > @@ -2250,19 +2250,25 @@ > } > > > +#if 0 > +// > +// Use when RtlQuerySecurityObject is implemented > +// > + > /* > * @implemented > */ > BOOL > WINAPI > -GetPrivateObjectSecurity(PSECURITY_DESCRIPTOR ObjectDescriptor, > - SECURITY_INFORMATION SecurityInformation, > - PSECURITY_DESCRIPTOR ResultantDescriptor, > - DWORD DescriptorLength, > - PDWORD ReturnLength) > +GetPrivateObjectSecurity(IN PSECURITY_DESCRIPTOR ObjectDescriptor, > + IN SECURITY_INFORMATION SecurityInformation, > + OUT PSECURITY_DESCRIPTOR > ResultantDescriptor OPTIONAL, > + IN DWORD DescriptorLength, > + OUT PDWORD ReturnLength) > { > NTSTATUS Status; > > + /* Call RTL */ > Status = RtlQuerySecurityObject(ObjectDescriptor, > SecurityInformation, > ResultantDescriptor, > @@ -2270,12 +2276,73 @@ > ReturnLength); > if (!NT_SUCCESS(Status)) > { > + /* Fail */ > SetLastError(RtlNtStatusToDosError(Status)); > return FALSE; > } > > + /* Success */ > return TRUE; > } > +#else > +// > +// Wine's implementation (as of December 30th 2008) > +// > + > +/* > + * @implemented > + */ > +BOOL > +WINAPI > +GetPrivateObjectSecurity(IN PSECURITY_DESCRIPTOR ObjectDescriptor, > + IN SECURITY_INFORMATION SecurityInformation, > + OUT PSECURITY_DESCRIPTOR > ResultantDescriptor OPTIONAL, > + IN DWORD DescriptorLength, > + OUT PDWORD ReturnLength) > +{ > + SECURITY_DESCRIPTOR desc; > + BOOL defaulted, present; > + PACL pacl; > + PSID psid; > + > + TRACE("(%p,0x%08x,%p,0x%08x,%p)\n", ObjectDescriptor, > SecurityInformation, > + ResultantDescriptor, DescriptorLength, ReturnLength); > + > + if (!InitializeSecurityDescriptor(&desc, > SECURITY_DESCRIPTOR_REVISION)) > + return FALSE; > + > + if (SecurityInformation & OWNER_SECURITY_INFORMATION) > + { > + if (!GetSecurityDescriptorOwner(ObjectDescriptor, &psid, > &defaulted)) > + return FALSE; > + SetSecurityDescriptorOwner(&desc, psid, defaulted); > + } > + > + if (SecurityInformation & GROUP_SECURITY_INFORMATION) > + { > + if (!GetSecurityDescriptorGroup(ObjectDescriptor, &psid, > &defaulted)) > + return FALSE; > + SetSecurityDescriptorGroup(&desc, psid, defaulted); > + } > + > + if (SecurityInformation & DACL_SECURITY_INFORMATION) > + { > + if (!GetSecurityDescriptorDacl(ObjectDescriptor, &present, > &pacl, &defaulted)) > + return FALSE; > + SetSecurityDescriptorDacl(&desc, present, pacl, defaulted); > + } > + > + if (SecurityInformation & SACL_SECURITY_INFORMATION) > + { > + if (!GetSecurityDescriptorSacl(ObjectDescriptor, &present, > &pacl, &defaulted)) > + return FALSE; > + SetSecurityDescriptorSacl(&desc, present, pacl, defaulted); > + } > + > + *ReturnLength = DescriptorLength; > + return MakeSelfRelativeSD(&desc, ResultantDescriptor, > ReturnLength); > +} > +#endif > > > /* > Best regards, Alex Ionescu

2 1

About r38504
by Pierre SCHWEITZER 02 Jan '09

02 Jan '09

Hi, the declaration of FsRtlNotifyReportChange was correct, even if this function is now deprecated (that's why it doesn't need to be in DDK). WBR, P. Schweitzer _________________________________________________________________ Inédit ! Des Emoticônes Déjantées! Installez les dans votre Messenger ! http://www.ilovemessenger.fr/Emoticones/EmoticonesDejantees.aspx

1 0

Re: [ros-dev] Problems installing REACTOS
by Ged 02 Jan '09

02 Jan '09

Why do you need to start on real hardware? This is fairly unsupported at the moment, so you’re really in no-mans-land here. I’d suggest a virtual machine until you become more familiar with it. Anyway, looking at your hardware spec I don’t see any untoward. Continuing down the real hardware path, your next step is to obtain debug output. The only current method to do this is via a null-modem cable, as specified in the link I gave you. Ged. From: James Muthee [mailto:jsoftpacks@yahoo.co.uk] Sent: 31 December 2008 09:12 To: Ged Subject: RE: [ros-dev] Problems installing REACTOS Hello Ged. Thanks for the prompt reply. First, I need to test the software on a real machine. No Virtualization, no support form any other OS. As for the exact configuration of the system, I am forwarding the entire report from the windows system info. I know its lengthy but, then again you would be looking for someting specific. My disks are FAT32 Please help. Thanks James I. M. --- On Tue, 30/12/08, Ged <gedmurphy(a)gmail.com> wrote: From: Ged <gedmurphy(a)gmail.com> Subject: RE: [ros-dev] Problems installing REACTOS To: jsoftpacks(a)yahoo.co.uk, "'ReactOS Development List'" <ros-dev(a)reactos.org> Date: Tuesday, 30 December, 2008, 1:59 PM It could be a whole range of things. Without having your hardware spec and debug output, it’s impossible to tell. If you just want to test it, it’s better to run it in a virtual machine. If you’re insistent on running it on hardware, then you’ll need to start by supplying your hardware details. If that doesn’t show any obvious problems (e.g. SATA, NTFS, etc) then you will need to grab debug output via a null cable modem <http://www.reactos.org/wiki/index.php/Debugging#Real_computer:_Physical_ser…> http://www.reactos.org/wiki/index.php/Debugging#Real_computer:_Physical_ser… Good luck. Ged. From: ros-dev-bounces(a)reactos.org [mailto:ros-dev-bounces@reactos.org] On Behalf Of James Muthee Sent: 30 December 2008 11:21 To: ros-dev(a)reactos.org Subject: [ros-dev] Problems installing REACTOS Hello. Having downloaded the software (I was amazed it was less than 40MB), I wrote the ISO on a CD and started the installation. 0.3.7 failed to install completely, stopping dead just after loading the initial files. I downloaded the 0.3.6 and tried again. This time, I was able to install the first portion (copying the textmode installation to the hard disk and modifying the MBR). On restart, A GUI window flashed then a blue screen with a single line about the Build, and that was the end. What am I doing wrong? I would appreciate any help. I really want to test drive this software. James I. M.

2 2

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