Ros-dev September 2015

ros-dev@reactos.org

10 participants
12 discussions

Re: [ros-dev] [ros-diffs] [sginsberg] 69393: [NTOS] Fix the Ob wait system calls to only catch the exceptions that are expected to be raised by the Ke wait functions (and not potentially silently catching *any* exception an...

by Pierre Schweitzer

On 28/09/2015 11:01, sginsberg(a)svn.reactos.org wrote: > Author: sginsberg > Date: Mon Sep 28 09:01:11 2015 > New Revision: 69393 > > URL: http://svn.reactos.org/svn/reactos?rev=69393&view=rev > Log: > [NTOS] Fix the Ob wait system calls to only catch the exceptions that are expected to be raised by the Ke wait functions (and not potentially silently catching *any* exception and corrupting everything in the process). Also fixup some code logic. SEH Mega Fixup 1/??? > > Modified: > trunk/reactos/ntoskrnl/ob/obwait.c > > Modified: trunk/reactos/ntoskrnl/ob/obwait.c > URL: http://svn.reactos.org/svn/reactos/trunk/reactos/ntoskrnl/ob/obwait.c?rev=6… > ============================================================================== > --- trunk/reactos/ntoskrnl/ob/obwait.c [iso-8859-1] (original) > +++ trunk/reactos/ntoskrnl/ob/obwait.c [iso-8859-1] Mon Sep 28 09:01:11 2015 > @@ -49,12 +49,12 @@ > IN BOOLEAN Alertable, > IN PLARGE_INTEGER TimeOut OPTIONAL) > { > - PKWAIT_BLOCK WaitBlockArray = NULL; > + PKWAIT_BLOCK WaitBlockArray; > HANDLE Handles[MAXIMUM_WAIT_OBJECTS], KernelHandle; > PVOID Objects[MAXIMUM_WAIT_OBJECTS]; > PVOID WaitObjects[MAXIMUM_WAIT_OBJECTS]; > - ULONG i = 0, ReferencedObjects = 0, j; > - KPROCESSOR_MODE PreviousMode = ExGetPreviousMode(); > + ULONG i, ReferencedObjects, j; > + KPROCESSOR_MODE PreviousMode; > LARGE_INTEGER SafeTimeOut; > BOOLEAN LockInUse; > PHANDLE_TABLE_ENTRY HandleEntry; > @@ -65,31 +65,26 @@ > NTSTATUS Status; > PAGED_CODE(); > > - /* Enter a critical region since we'll play with handles */ > - LockInUse = TRUE; > - KeEnterCriticalRegion(); > - > /* Check for valid Object Count */ > if ((ObjectCount > MAXIMUM_WAIT_OBJECTS) || !(ObjectCount)) > { > /* Fail */ > - Status = STATUS_INVALID_PARAMETER_1; > - goto Quickie; > + return STATUS_INVALID_PARAMETER_1; > } > > /* Check for valid Wait Type */ > if ((WaitType != WaitAll) && (WaitType != WaitAny)) > { > /* Fail */ > - Status = STATUS_INVALID_PARAMETER_3; > - goto Quickie; > - } > - > - /* Enter SEH */ > - _SEH2_TRY > - { > - /* Check if the call came from user mode */ > - if (PreviousMode != KernelMode) > + return STATUS_INVALID_PARAMETER_3; > + } > + > + /* Enter SEH for user mode */ > + PreviousMode = ExGetPreviousMode(); > + if (PreviousMode != KernelMode) > + { > + /* Enter SEH */ > + _SEH2_TRY No, this is plain wrong. This is not because you're in kernel mode that the world is marvelous and callers trustable. A caller can pass you buggy address and you HAVE to wrap the RtlCopyMemory in SEH to make sure that if a buggy address is passed, the whole system isn't brought down (that's the whole purpose of the copy after all!). In case you have a doubt, just put some random: Status = ZwWaitForMultipleObjects(2, (void **)0x42424242, WaitAll, FALSE, NULL); In a kernel component. In w2k3, you'll get Status = STATUS_ACCESS_VIOLATION In ReactOS, with your changes: BSOD. Please before doing random changes that you believe are right: do testing. Alex already told you that. Cheers, -- Pierre Schweitzer <pierre at reactos.org> System & Network Administrator Senior Kernel Developer ReactOS Deutschland e.V.

9 years, 2 months

Status Meeting (September 2015)

by Aleksey Bragin

Hello, Let me invite you to the monthly status meeting taking place last Thursday of a month, 24th of September, 19:00 UTC, as usual. IRC service will only be started shortly before the meeting. Your participation passwords and server address will be emailed to you shortly before the meeting starts, and they are going to be different once again as they are not stored in any database. Hopefully it's not much of inconvenience. Please send agenda proposals to me before the meeting. Regards, Aleksey Bragin

9 years, 3 months

Re: [ros-dev] [ros-diffs] [tfaber] 69296: [KMTESTS] - Add missing OBJ_KERNEL_HANDLE - OBJ_KERNEL_HANDLE in user mode has no effect... no reason to make a define around it CORE-10207

by Alex Ionescu

Actually, in XP, OBJ_KERNEL_HANDLE Did work, and gave you a kernel handle that you could never use. In Windows Vista, it returns a failure code. Not sure about 2003. Best regards, Alex Ionescu On Sat, Sep 19, 2015 at 2:46 PM, <tfaber(a)svn.reactos.org> wrote: > Author: tfaber > Date: Sat Sep 19 21:46:40 2015 > New Revision: 69296 > > URL: http://svn.reactos.org/svn/reactos?rev=69296&view=rev > Log: > [KMTESTS] > - Add missing OBJ_KERNEL_HANDLE > - OBJ_KERNEL_HANDLE in user mode has no effect... no reason to make a > define around it > CORE-10207 > > Modified: > trunk/rostests/kmtests/ntos_io/IoCreateFile.c > trunk/rostests/kmtests/ntos_ke/KeEvent.c > trunk/rostests/kmtests/rtl/RtlRegistry.c > > Modified: trunk/rostests/kmtests/ntos_io/IoCreateFile.c > URL: > http://svn.reactos.org/svn/reactos/trunk/rostests/kmtests/ntos_io/IoCreateF… > > ============================================================================== > --- trunk/rostests/kmtests/ntos_io/IoCreateFile.c [iso-8859-1] > (original) > +++ trunk/rostests/kmtests/ntos_io/IoCreateFile.c [iso-8859-1] Sat > Sep 19 21:46:40 2015 > @@ -454,7 +454,7 @@ > ok_eq_hex(IoStatusBlock.Status, 0xFFFFFFFF); > if (Status == STATUS_SUCCESS) > { > - ObCloseHandle(ParentHandle, KernelMode); > + ObCloseHandle(ParentHandle, UserMode); > } > > /* Attempt relative target open */ > @@ -462,7 +462,7 @@ > IoStatusBlock.Information = 0xFFFFFFFF; > InitializeObjectAttributes(&ObjectAttributes, > &SystemRoot, > - OBJ_CASE_INSENSITIVE, > + OBJ_CASE_INSENSITIVE | OBJ_KERNEL_HANDLE, > NULL, NULL); > Status = ZwOpenFile(&SystemRootHandle, > GENERIC_WRITE | GENERIC_READ | SYNCHRONIZE, > @@ -478,7 +478,7 @@ > IoStatusBlock.Information = 0xFFFFFFFF; > InitializeObjectAttributes(&ObjectAttributes, > &Regedit, > - OBJ_CASE_INSENSITIVE, > + OBJ_CASE_INSENSITIVE | > OBJ_KERNEL_HANDLE, > SystemRootHandle, > NULL); > Status = IoCreateFile(&ParentHandle, > > Modified: trunk/rostests/kmtests/ntos_ke/KeEvent.c > URL: > http://svn.reactos.org/svn/reactos/trunk/rostests/kmtests/ntos_ke/KeEvent.c… > > ============================================================================== > --- trunk/rostests/kmtests/ntos_ke/KeEvent.c [iso-8859-1] (original) > +++ trunk/rostests/kmtests/ntos_ke/KeEvent.c [iso-8859-1] Sat Sep 19 > 21:46:40 2015 > @@ -155,6 +155,7 @@ > PVOID ThreadObjects[RTL_NUMBER_OF(Threads)]; > LONG State; > PKTHREAD Thread = KeGetCurrentThread(); > + OBJECT_ATTRIBUTES ObjectAttributes; > > LongTimeout.QuadPart = -100 * MILLISECOND; > ShortTimeout.QuadPart = -1 * MILLISECOND; > @@ -165,7 +166,12 @@ > { > Threads[i].Event = Event; > Threads[i].Signal = FALSE; > - Status = PsCreateSystemThread(&Threads[i].Handle, GENERIC_ALL, > NULL, NULL, NULL, WaitForEventThread, &Threads[i]); > + InitializeObjectAttributes(&ObjectAttributes, > + NULL, > + OBJ_KERNEL_HANDLE, > + NULL, > + NULL); > + Status = PsCreateSystemThread(&Threads[i].Handle, GENERIC_ALL, > &ObjectAttributes, NULL, NULL, WaitForEventThread, &Threads[i]); > ok_eq_hex(Status, STATUS_SUCCESS); > Status = ObReferenceObjectByHandle(Threads[i].Handle, > SYNCHRONIZE, *PsThreadType, KernelMode, (PVOID *)&Threads[i].Thread, NULL); > ok_eq_hex(Status, STATUS_SUCCESS); > > Modified: trunk/rostests/kmtests/rtl/RtlRegistry.c > URL: > http://svn.reactos.org/svn/reactos/trunk/rostests/kmtests/rtl/RtlRegistry.c… > > ============================================================================== > --- trunk/rostests/kmtests/rtl/RtlRegistry.c [iso-8859-1] (original) > +++ trunk/rostests/kmtests/rtl/RtlRegistry.c [iso-8859-1] Sat Sep 19 > 21:46:40 2015 > @@ -7,12 +7,6 @@ > > #define KMT_EMULATE_KERNEL > #include <kmt_test.h> > - > -#ifdef KMT_KERNEL_MODE > -#define KMT_KERNEL_HANDLE OBJ_KERNEL_HANDLE > -#else > -#define KMT_KERNEL_HANDLE 0 > -#endif > > #ifndef RTL_NUMBER_OF > #define RTL_NUMBER_OF(x) (sizeof(x) / sizeof(x[0])) > @@ -212,7 +206,7 @@ > RtlInitUnicodeString(&KeyName, L"\\Registry\\MACHINE\\Software"); > InitializeObjectAttributes(&ObjectAttributes, > &KeyName, > - OBJ_CASE_INSENSITIVE | KMT_KERNEL_HANDLE, > + OBJ_CASE_INSENSITIVE | OBJ_KERNEL_HANDLE, > NULL, > NULL); > Status = ZwOpenKey(&SoftwareHandle, > @@ -225,7 +219,7 @@ > RtlInitUnicodeString(&KeyName, L"RtlRegistryKmtestKey"); > InitializeObjectAttributes(&ObjectAttributes, > &KeyName, > - OBJ_CASE_INSENSITIVE | KMT_KERNEL_HANDLE, > + OBJ_CASE_INSENSITIVE | OBJ_KERNEL_HANDLE, > SoftwareHandle, > NULL); > Status = ZwCreateKey(&KeyHandle, > > >

9 years, 3 months

Re: [ros-dev] [ros-diffs] [tfaber] 69236: [MSFS] - Use a NULL timeout for infinite waits instead of waiting for 100 ns. CORE-10188 #resolve - Wait for available read data in user mode to handle thread termination - Return S...

by Alex Ionescu

I'm not convinced on the user->kernel changes Thomas. Best regards, Alex Ionescu On Tue, Sep 15, 2015 at 5:40 AM, <tfaber(a)svn.reactos.org> wrote: > Author: tfaber > Date: Tue Sep 15 09:40:30 2015 > New Revision: 69236 > > URL: http://svn.reactos.org/svn/reactos?rev=69236&view=rev > Log: > [MSFS] > - Use a NULL timeout for infinite waits instead of waiting for 100 ns. > CORE-10188 #resolve > - Wait for available read data in user mode to handle thread termination > - Return STATUS_IO_TIMEOUT also for a zero-length timeout. Fixes Wine tests > - Avoid MmGetSystemAddressForMdl > - Acquiring a mutex is not a UserRequest > > Modified: > trunk/reactos/drivers/filesystems/msfs/msfs.h > trunk/reactos/drivers/filesystems/msfs/rw.c > > Modified: trunk/reactos/drivers/filesystems/msfs/msfs.h > URL: > http://svn.reactos.org/svn/reactos/trunk/reactos/drivers/filesystems/msfs/m… > > ============================================================================== > --- trunk/reactos/drivers/filesystems/msfs/msfs.h [iso-8859-1] > (original) > +++ trunk/reactos/drivers/filesystems/msfs/msfs.h [iso-8859-1] Tue > Sep 15 09:40:30 2015 > @@ -54,7 +54,7 @@ > > > #define KeLockMutex(x) KeWaitForSingleObject(x, \ > - UserRequest, \ > + Executive, \ > KernelMode, \ > FALSE, \ > NULL); > > Modified: trunk/reactos/drivers/filesystems/msfs/rw.c > URL: > http://svn.reactos.org/svn/reactos/trunk/reactos/drivers/filesystems/msfs/r… > > ============================================================================== > --- trunk/reactos/drivers/filesystems/msfs/rw.c [iso-8859-1] (original) > +++ trunk/reactos/drivers/filesystems/msfs/rw.c [iso-8859-1] Tue Sep 15 > 09:40:30 2015 > @@ -29,6 +29,7 @@ > ULONG LengthRead = 0; > PVOID Buffer; > NTSTATUS Status; > + PLARGE_INTEGER Timeout; > > DPRINT("MsfsRead(DeviceObject %p Irp %p)\n", DeviceObject, Irp); > > @@ -52,16 +53,21 @@ > > Length = IoStack->Parameters.Read.Length; > if (Irp->MdlAddress) > - Buffer = MmGetSystemAddressForMdl (Irp->MdlAddress); > + Buffer = MmGetSystemAddressForMdlSafe(Irp->MdlAddress, > NormalPagePriority); > else > Buffer = Irp->UserBuffer; > > + if (Fcb->TimeOut.QuadPart == -1LL) > + Timeout = NULL; > + else > + Timeout = &Fcb->TimeOut; > + > Status = KeWaitForSingleObject(&Fcb->MessageEvent, > UserRequest, > - KernelMode, > + UserMode, > FALSE, > - &Fcb->TimeOut); > - if (NT_SUCCESS(Status)) > + Timeout); > + if (Status != STATUS_USER_APC) > { > if (Fcb->MessageCount > 0) > { > @@ -84,7 +90,7 @@ > KeClearEvent(&Fcb->MessageEvent); > } > } > - else if (Fcb->TimeOut.QuadPart != 0LL) > + else > { > /* No message found after waiting */ > Status = STATUS_IO_TIMEOUT; > @@ -135,7 +141,7 @@ > > Length = IoStack->Parameters.Write.Length; > if (Irp->MdlAddress) > - Buffer = MmGetSystemAddressForMdl (Irp->MdlAddress); > + Buffer = MmGetSystemAddressForMdlSafe(Irp->MdlAddress, > NormalPagePriority); > else > Buffer = Irp->UserBuffer; > > > >

9 years, 3 months

Re: [ros-dev] [ros-diffs] [spetreolle] 69186: [TRANSLATIONS] Translate remaining French strings in msgina.dll

by Pierre Schweitzer

Comments inline. On 11/09/2015 21:30, spetreolle(a)svn.reactos.org wrote: > URL: http://svn.reactos.org/svn/reactos/trunk/reactos/dll/win32/msgina/lang/fr-F… > ============================================================================== > --- trunk/reactos/dll/win32/msgina/lang/fr-FR.rc [iso-8859-1] (original) > +++ trunk/reactos/dll/win32/msgina/lang/fr-FR.rc [iso-8859-1] Fri Sep 11 19:30:29 2015 > @@ -25,11 +25,11 @@ > FONT 8, "MS Shell Dlg", 400, 0, 1 > BEGIN > CONTROL IDI_ROSLOGO, IDC_ROSLOGO, "Static", SS_BITMAP, 0, 0, 275, 54 > - LTEXT "Utilisateur :", IDC_STATIC, 6, 60, 52, 8 > + LTEXT "Utilisateur:", IDC_STATIC, 6, 60, 52, 8 This is a broken change. > EDITTEXT IDC_USERNAME, 60, 57, 155, 14, ES_AUTOHSCROLL > - LTEXT "Mot de passe :", IDC_STATIC, 6, 78, 52, 8 > + LTEXT "Mot de passe:", IDC_STATIC, 6, 78, 52, 8 This is a broken change. > @@ -68,50 +68,50 @@ > > IDD_UNLOCK_DLG DIALOGEX 0, 0, 275, 179 > STYLE NOT WS_VISIBLE | DS_CENTER | DS_MODALFRAME | DS_SHELLFONT | WS_BORDER | WS_CAPTION | WS_DLGFRAME | WS_POPUP > -CAPTION "Unlock Computer" > +CAPTION "DÃ©verrouiller l'ordinateur" > FONT 8, "MS Shell Dlg", 400, 0, 1 > BEGIN > CONTROL IDI_ROSLOGO, IDC_ROSLOGO, "Static", SS_BITMAP, 0, 0, 275, 54 > ICON IDI_LOCKICON, -1, 7, 59, 20, 20 > - LTEXT "This computer is in use and has been locked.", IDC_STATIC, 36, 61, 232, 8 > + LTEXT "Cet ordinateur est actuellement verrouillÃ©.", IDC_STATIC, 36, 61, 232, 8 > LTEXT "Message", IDC_LOCKMSG, 36, 75, 232, 26 > - LTEXT "User name:", IDC_STATIC, 36, 107, 40, 8 > + LTEXT "Utilisateur:", IDC_STATIC, 36, 107, 40, 8 French typography requires a space before and after ':' > EDITTEXT IDC_USERNAME, 84, 104, 119, 14, ES_AUTOHSCROLL > - LTEXT "Password:", IDC_STATIC, 36, 125, 42, 8 > + LTEXT "Mot de passe:", IDC_STATIC, 36, 125, 50, 8 See upper comment. > EDITTEXT IDC_PASSWORD, 84, 123, 119, 14, ES_AUTOHSCROLL | ES_PASSWORD > PUSHBUTTON "OK", IDOK, 80, 154, 50, 14, BS_DEFPUSHBUTTON > - PUSHBUTTON "Cancel", IDCANCEL, 144, 154, 50, 14 > + PUSHBUTTON "Annuler", IDCANCEL, 144, 154, 50, 14 > END > > IDD_CHANGE_PASSWORD DIALOGEX 0, 0, 275, 166 > STYLE NOT WS_VISIBLE | DS_CENTER | DS_MODALFRAME | DS_SHELLFONT | WS_BORDER | WS_CAPTION | WS_DLGFRAME | WS_POPUP > -CAPTION "Change Password" > +CAPTION "Changer le mot de passe" > FONT 8, "MS Shell Dlg", 400, 0, 1 > BEGIN > CONTROL IDI_ROSLOGO, IDC_ROSLOGO, "Static", SS_BITMAP, 0, 0, 275, 54 > - LTEXT "User name:", IDC_STATIC, 7, 61, 78, 8 > + LTEXT "Utilisateur:", IDC_STATIC, 7, 61, 78, 8 Same here. > EDITTEXT IDC_CHANGEPWD_USERNAME, 90, 59, 127, 12, ES_AUTOHSCROLL > - LTEXT "Log on to:", IDC_STATIC, 7, 78, 78, 8 > + LTEXT "Se connecter Ã :", IDC_STATIC, 7, 78, 78, 8 > COMBOBOX IDC_CHANGEPWD_DOMAIN, 90, 75, 127, 144, CBS_DROPDOWNLIST | CBS_SORT | WS_TABSTOP > - LTEXT "Old Password:", IDC_STATIC, 7, 95, 78, 8 > + LTEXT "Ancien mot de passe:", IDC_STATIC, 7, 95, 78, 8 See upper comment. > EDITTEXT IDC_CHANGEPWD_OLDPWD, 90, 92, 127, 12, ES_AUTOHSCROLL | ES_PASSWORD > - LTEXT "New Password:", IDC_STATIC, 7, 111, 78, 8 > + LTEXT "Nouveau mot de passe:", IDC_STATIC, 7, 111, 78, 8 Same here. > EDITTEXT IDC_CHANGEPWD_NEWPWD1, 90, 109, 127, 12, ES_AUTOHSCROLL | ES_PASSWORD > - LTEXT "Confirm new Password:", IDC_STATIC, 7, 127, 78, 8 > + LTEXT "Confirmez le nouveau mot de passe:", IDC_STATIC, 7, 127, 78, 16 See upper comment. > EDITTEXT IDC_CHANGEPWD_NEWPWD2, 90, 125, 127, 12, ES_AUTOHSCROLL | ES_PASSWORD > PUSHBUTTON "OK", IDOK, 164, 145, 50, 14, BS_DEFPUSHBUTTON > - PUSHBUTTON "Cancel", IDCANCEL, 218, 145, 50, 14 > + PUSHBUTTON "Annuler", IDCANCEL, 218, 145, 50, 14 > END > > IDD_LOGOFF_DLG DIALOGEX 0, 0, 188, 60 > STYLE NOT WS_VISIBLE | DS_CENTER | DS_MODALFRAME | DS_SHELLFONT | WS_BORDER | WS_CAPTION | WS_DLGFRAME | WS_SYSMENU | WS_POPUP > -CAPTION "Log Off ReactOS" > +CAPTION "DÃ©connexion de ReactOS" > FONT 8, "MS Shell Dlg", 400, 0, 1 > BEGIN > ICON IDI_LOCKICON, -1, 7, 7, 20, 20 > - LTEXT "Are you sure you want to log off?", IDC_STATIC, 35, 16, 146, 8 > - PUSHBUTTON "Yes", IDYES, 41, 39, 50, 14, BS_DEFPUSHBUTTON > - PUSHBUTTON "No", IDNO, 95, 39, 50, 14 > + LTEXT "Ãtes vous sÃ»r de vouloir vous dÃ©connecter?", IDC_STATIC, 35, 16, 146, 8 You need a dash here. > @@ -160,34 +160,34 @@ > IDS_ASKFORUSER "Utilisateur : " > IDS_ASKFORPASSWORD "Mot de passe : " > IDS_FORCELOGOFF "Cela dÃ©connectera l'utilisateur en cours et perdra les donnÃ©es non sauvÃ©es. Continuer ?" > - IDS_LOCKMSG "Seulement %s ou bien un Administrateur peut dÃ©verrouiller cet ordinateur." > - IDS_LOGONMSG "You are logged on as %s." > - IDS_LOGONDATE "Logon date: %s %s" > - IDS_COMPUTERLOCKED "Computer locked" > - IDS_LOCKEDWRONGPASSWORD "The password is wrong. Please enter your password again. Letters in passwords must be typed using the correct case." > - IDS_LOCKEDWRONGUSER "This computer is locked. Only %s\\%s or an Administrator can unlock this computer." > - IDS_CHANGEPWDTITLE "Change Password" > - IDS_NONMATCHINGPASSWORDS "The passwords you typed do not match. Type the same password in both text boxes." > - IDS_PASSWORDCHANGED "Your password has been changed." > - IDS_LOGONTITLE "Logon Message" > - IDS_LOGONWRONGUSERORPWD "The system could not log you on. Make sure your User name and domain are correct, then type your password again. Letters in passwords must be typed using the correct case." > - IDS_LOGONUSERDISABLED "Your account has been disabled. Please see your system administrator." > - IDS_PASSWORDMUSTCHANGE "You are required to change your password at first logon." > - IDS_PASSWORDEXPIRED "Your password has expired and must be changed." > + IDS_LOCKMSG "Seul %s ou un administrateur peut dÃ©verrouiller cet ordinateur." > + IDS_LOGONMSG "Vous Ãªtes connectÃ© en tant que %s." > + IDS_LOGONDATE "Date de connexion: %s %s" See comment about ':' > + IDS_COMPUTERLOCKED "Ordinateur verrouillÃ©" > + IDS_LOCKEDWRONGPASSWORD "Le mot de passe est incorrect. Entrez Ã nouveau votre mot de passe en respectant les majuscules." > + IDS_LOCKEDWRONGUSER "Cet ordinateur est actuellement verrouillÃ©. Seul %s\\%s ou un administrateur peut dÃ©verrouiller cet ordinateur." > + IDS_CHANGEPWDTITLE "Changer le mot de passe" > + IDS_NONMATCHINGPASSWORDS "Les mots de passe ne correspondent pas. Entrez le mÃªme mot de passe dans les deux zones de texte." > + IDS_PASSWORDCHANGED "Votre mot de passe a Ã©tÃ© changÃ©." > + IDS_LOGONTITLE "Connexion" > + IDS_LOGONWRONGUSERORPWD "Le systÃ¨me ne peut vous connecter. VÃ©rifiez vos noms d'utilisateur et de domaine. Entrez Ã nouveau votre mot de passe en respectant les majuscules." > + IDS_LOGONUSERDISABLED "Votre compte est dÃ©sactivÃ©. Voyez votre administrateur systÃ¨me." Consultez votre ? -- Pierre Schweitzer <pierre at reactos.org> System & Network Administrator Senior Kernel Developer ReactOS Deutschland e.V.

9 years, 3 months

Re: [ros-dev] [ros-diffs] [sginsberg] 69248: [HAL] Get rid off REGISTERCALL ("regparm(3)" for GCC) and replace with good old NTAPI (stdcall). Such an obsession with avoiding stack usage is not healthy nor makes much sense t...

by Alex Ionescu

You seem to have a huge misunderstanding of modern processor architecture. Everything you said is false. In fact, this is why x64 uses register passing, as does ARM and every other major architecture except x86. Push/pop operations are extremely expensive and break pipelining. The compiler will then place the parameters in registers anyways after the fact. Suffice it to say though, the original reason these were regparm/fastcall is to easily allow them to be called from assembly at the time. Best regards, Alex Ionescu On Tue, Sep 15, 2015 at 7:03 PM, <sginsberg(a)svn.reactos.org> wrote: > Author: sginsberg > Date: Tue Sep 15 23:03:42 2015 > New Revision: 69248 > > URL: http://svn.reactos.org/svn/reactos?rev=69248&view=rev > Log: > [HAL] Get rid off REGISTERCALL ("regparm(3)" for GCC) and replace with > good old NTAPI (stdcall). Such an obsession with avoiding stack usage is > not healthy nor makes much sense today (or even a long time before today) > with processors that have a decent L1 cache, whose "cost of access" is > basically the same as to that of a register, and with processors being > capable of recognising basic access patterns to ensure frequently used > memory (read: stack) is in the cache. Slapping FASTCALL/regparm on > frequently used code does not ensure it actually operates faster. You want > to know what really hurts performance (and cache)? Slapping FORCEINLINE on > everything like if it was some kind of asm macro, making code needlessly > bloated. > > Modified: > trunk/reactos/hal/halx86/include/halp.h > trunk/reactos/hal/halx86/up/pic.c > > Modified: trunk/reactos/hal/halx86/include/halp.h > URL: > http://svn.reactos.org/svn/reactos/trunk/reactos/hal/halx86/include/halp.h?… > > ============================================================================== > --- trunk/reactos/hal/halx86/include/halp.h [iso-8859-1] (original) > +++ trunk/reactos/hal/halx86/include/halp.h [iso-8859-1] Tue Sep 15 > 23:03:42 2015 > @@ -10,12 +10,6 @@ > #define INIT_SECTION /* Done via alloc_text for MSC */ > #endif > > - > -#ifdef _MSC_VER > -#define REGISTERCALL FASTCALL > -#else > -#define REGISTERCALL __attribute__((regparm(3))) > -#endif > > #ifdef CONFIG_SMP > #define HAL_BUILD_TYPE (DBG ? PRCB_BUILD_DEBUG : 0) > @@ -430,14 +424,14 @@ > > typedef > BOOLEAN > -( REGISTERCALL *PHAL_DISMISS_INTERRUPT)( > +(NTAPI *PHAL_DISMISS_INTERRUPT)( > IN KIRQL Irql, > IN ULONG Irq, > OUT PKIRQL OldIrql > ); > > BOOLEAN > -REGISTERCALL > +NTAPI > HalpDismissIrqGeneric( > IN KIRQL Irql, > IN ULONG Irq, > @@ -445,7 +439,7 @@ > ); > > BOOLEAN > -REGISTERCALL > +NTAPI > HalpDismissIrq15( > IN KIRQL Irql, > IN ULONG Irq, > @@ -453,7 +447,7 @@ > ); > > BOOLEAN > -REGISTERCALL > +NTAPI > HalpDismissIrq13( > IN KIRQL Irql, > IN ULONG Irq, > @@ -461,7 +455,7 @@ > ); > > BOOLEAN > -REGISTERCALL > +NTAPI > HalpDismissIrq07( > IN KIRQL Irql, > IN ULONG Irq, > @@ -469,7 +463,7 @@ > ); > > BOOLEAN > -REGISTERCALL > +NTAPI > HalpDismissIrqLevel( > IN KIRQL Irql, > IN ULONG Irq, > @@ -477,7 +471,7 @@ > ); > > BOOLEAN > -REGISTERCALL > +NTAPI > HalpDismissIrq15Level( > IN KIRQL Irql, > IN ULONG Irq, > @@ -485,7 +479,7 @@ > ); > > BOOLEAN > -REGISTERCALL > +NTAPI > HalpDismissIrq13Level( > IN KIRQL Irql, > IN ULONG Irq, > @@ -493,7 +487,7 @@ > ); > > BOOLEAN > -REGISTERCALL > +NTAPI > HalpDismissIrq07Level( > IN KIRQL Irql, > IN ULONG Irq, > > Modified: trunk/reactos/hal/halx86/up/pic.c > URL: > http://svn.reactos.org/svn/reactos/trunk/reactos/hal/halx86/up/pic.c?rev=69… > > ============================================================================== > --- trunk/reactos/hal/halx86/up/pic.c [iso-8859-1] (original) > +++ trunk/reactos/hal/halx86/up/pic.c [iso-8859-1] Tue Sep 15 23:03:42 > 2015 > @@ -843,7 +843,7 @@ > } > > BOOLEAN > -REGISTERCALL > +NTAPI > HalpDismissIrqGeneric(IN KIRQL Irql, > IN ULONG Irq, > OUT PKIRQL OldIrql) > @@ -853,7 +853,7 @@ > } > > BOOLEAN > -REGISTERCALL > +NTAPI > HalpDismissIrq15(IN KIRQL Irql, > IN ULONG Irq, > OUT PKIRQL OldIrql) > @@ -889,7 +889,7 @@ > > > BOOLEAN > -REGISTERCALL > +NTAPI > HalpDismissIrq13(IN KIRQL Irql, > IN ULONG Irq, > OUT PKIRQL OldIrql) > @@ -902,7 +902,7 @@ > } > > BOOLEAN > -REGISTERCALL > +NTAPI > HalpDismissIrq07(IN KIRQL Irql, > IN ULONG Irq, > OUT PKIRQL OldIrql) > @@ -986,7 +986,7 @@ > } > > BOOLEAN > -REGISTERCALL > +NTAPI > HalpDismissIrqLevel(IN KIRQL Irql, > IN ULONG Irq, > OUT PKIRQL OldIrql) > @@ -996,7 +996,7 @@ > } > > BOOLEAN > -REGISTERCALL > +NTAPI > HalpDismissIrq15Level(IN KIRQL Irql, > IN ULONG Irq, > OUT PKIRQL OldIrql) > @@ -1031,7 +1031,7 @@ > } > > BOOLEAN > -REGISTERCALL > +NTAPI > HalpDismissIrq13Level(IN KIRQL Irql, > IN ULONG Irq, > OUT PKIRQL OldIrql) > @@ -1044,7 +1044,7 @@ > } > > BOOLEAN > -REGISTERCALL > +NTAPI > HalpDismissIrq07Level(IN KIRQL Irql, > IN ULONG Irq, > OUT PKIRQL OldIrql) > > >

9 years, 3 months

Re: [ros-dev] [ros-diffs] [pschweitzer] 69221: [MOUNTMGR] Implement the IOCTL IOCTL_MOUNTMGR_VOLUME_MOUNT_POINT_CREATED: - Implement WriteRemoteDatabaseEntry() - Implement MountMgrVolumeMountPointCreated()

by Alex Ionescu

Lol, make sure not to implement the huge vulnerability Microsoft patched two months ago (win2k->xp-style database migration). Best regards, Alex Ionescu On Sun, Sep 13, 2015 at 6:52 PM, <pschweitzer(a)svn.reactos.org> wrote: > Author: pschweitzer > Date: Sun Sep 13 22:52:07 2015 > New Revision: 69221 > > URL: http://svn.reactos.org/svn/reactos?rev=69221&view=rev > Log: > [MOUNTMGR] > Implement the IOCTL IOCTL_MOUNTMGR_VOLUME_MOUNT_POINT_CREATED: > - Implement WriteRemoteDatabaseEntry() > - Implement MountMgrVolumeMountPointCreated() > > Modified: > trunk/reactos/drivers/filters/mountmgr/database.c > trunk/reactos/drivers/filters/mountmgr/device.c > trunk/reactos/drivers/filters/mountmgr/mntmgr.h > trunk/reactos/drivers/filters/mountmgr/mountmgr.c > > Modified: trunk/reactos/drivers/filters/mountmgr/database.c > URL: > http://svn.reactos.org/svn/reactos/trunk/reactos/drivers/filters/mountmgr/d… > > ============================================================================== > --- trunk/reactos/drivers/filters/mountmgr/database.c [iso-8859-1] > (original) > +++ trunk/reactos/drivers/filters/mountmgr/database.c [iso-8859-1] Sun > Sep 13 22:52:07 2015 > @@ -192,6 +192,39 @@ > } > > return Entry; > +} > + > +/* > + * @implemented > + */ > +NTSTATUS > +WriteRemoteDatabaseEntry(IN HANDLE Database, > + IN LONG Offset, > + IN PDATABASE_ENTRY Entry) > +{ > + NTSTATUS Status; > + LARGE_INTEGER ByteOffset; > + IO_STATUS_BLOCK IoStatusBlock; > + > + ByteOffset.QuadPart = Offset; > + Status = ZwWriteFile(Database, > + NULL, > + NULL, > + NULL, > + &IoStatusBlock, > + Entry, > + Entry->EntrySize, > + &ByteOffset, > + NULL); > + if (NT_SUCCESS(Status)) > + { > + if (IoStatusBlock.Information < Entry->EntrySize) > + { > + Status = STATUS_INSUFFICIENT_RESOURCES; > + } > + } > + > + return Status; > } > > /* > > Modified: trunk/reactos/drivers/filters/mountmgr/device.c > URL: > http://svn.reactos.org/svn/reactos/trunk/reactos/drivers/filters/mountmgr/d… > > ============================================================================== > --- trunk/reactos/drivers/filters/mountmgr/device.c [iso-8859-1] > (original) > +++ trunk/reactos/drivers/filters/mountmgr/device.c [iso-8859-1] Sun > Sep 13 22:52:07 2015 > @@ -1688,15 +1688,242 @@ > return Status; > } > > +/* > + * @implemented > + */ > NTSTATUS > MountMgrVolumeMountPointCreated(IN PDEVICE_EXTENSION DeviceExtension, > IN PIRP Irp, > IN NTSTATUS LockStatus) > { > - UNREFERENCED_PARAMETER(DeviceExtension); > - UNREFERENCED_PARAMETER(Irp); > - UNREFERENCED_PARAMETER(LockStatus); > - return STATUS_NOT_IMPLEMENTED; > + LONG Offset; > + BOOLEAN Found; > + NTSTATUS Status; > + HANDLE RemoteDatabase; > + PMOUNTDEV_UNIQUE_ID UniqueId; > + PDATABASE_ENTRY DatabaseEntry; > + PASSOCIATED_DEVICE_ENTRY AssociatedEntry; > + PDEVICE_INFORMATION DeviceInformation, TargetDeviceInformation; > + UNICODE_STRING LinkTarget, SourceDeviceName, SourceSymbolicName, > TargetVolumeName, VolumeName, DbName; > + > + /* Initialize string */ > + LinkTarget.Length = 0; > + LinkTarget.MaximumLength = 0xC8; > + LinkTarget.Buffer = AllocatePool(LinkTarget.MaximumLength); > + if (LinkTarget.Buffer == NULL) > + { > + return STATUS_INSUFFICIENT_RESOURCES; > + } > + > + /* If the mount point was created, then, it changed! > + * Also use it to query some information > + */ > + Status = MountMgrVolumeMountPointChanged(DeviceExtension, Irp, > LockStatus, &SourceDeviceName, &SourceSymbolicName, &TargetVolumeName); > + /* Pending means DB are under synchronization, bail out */ > + if (Status == STATUS_PENDING) > + { > + FreePool(LinkTarget.Buffer); > + FreePool(SourceDeviceName.Buffer); > + FreePool(SourceSymbolicName.Buffer); > + return STATUS_PENDING; > + } > + else if (!NT_SUCCESS(Status)) > + { > + FreePool(LinkTarget.Buffer); > + return Status; > + } > + > + /* Query the device information */ > + Status = FindDeviceInfo(DeviceExtension, &SourceDeviceName, FALSE, > &DeviceInformation); > + if (!NT_SUCCESS(Status)) > + { > + /* If it failed, first try to get volume name */ > + Status = QueryVolumeName(0, NULL, &SourceDeviceName, &LinkTarget, > &VolumeName); > + if (!NT_SUCCESS(Status)) > + { > + /* Then, try to read the symlink */ > + Status = MountMgrQuerySymbolicLink(&SourceDeviceName, > &LinkTarget); > + if (!NT_SUCCESS(Status)) > + { > + FreePool(LinkTarget.Buffer); > + FreePool(SourceDeviceName.Buffer); > + FreePool(SourceSymbolicName.Buffer); > + return Status; > + } > + } > + else > + { > + FreePool(VolumeName.Buffer); > + } > + > + FreePool(SourceDeviceName.Buffer); > + > + SourceDeviceName.Length = LinkTarget.Length; > + SourceDeviceName.MaximumLength = LinkTarget.MaximumLength; > + SourceDeviceName.Buffer = LinkTarget.Buffer; > + > + /* Now that we have the correct source, reattempt to query > information */ > + Status = FindDeviceInfo(DeviceExtension, &SourceDeviceName, > FALSE, &DeviceInformation); > + if (!NT_SUCCESS(Status)) > + { > + FreePool(SourceDeviceName.Buffer); > + FreePool(SourceSymbolicName.Buffer); > + return Status; > + } > + } > + > + FreePool(SourceDeviceName.Buffer); > + > + /* Get information about target device */ > + Status = FindDeviceInfo(DeviceExtension, &TargetVolumeName, FALSE, > &TargetDeviceInformation); > + if (!NT_SUCCESS(Status)) > + { > + FreePool(SourceSymbolicName.Buffer); > + return Status; > + } > + > + /* Notify if not disabled */ > + if (!TargetDeviceInformation->SkipNotifications) > + { > + PostOnlineNotification(DeviceExtension, > &TargetDeviceInformation->SymbolicName); > + } > + > + /* Open the remote database */ > + RemoteDatabase = OpenRemoteDatabase(DeviceInformation, TRUE); > + if (RemoteDatabase == 0) > + { > + FreePool(SourceSymbolicName.Buffer); > + return STATUS_INSUFFICIENT_RESOURCES; > + } > + > + /* Browse all the entries */ > + Offset = 0; > + Found = FALSE; > + for (;;) > + { > + DatabaseEntry = GetRemoteDatabaseEntry(RemoteDatabase, Offset); > + if (DatabaseEntry == NULL) > + { > + break; > + } > + > + /* Try to find ourselves */ > + DbName.MaximumLength = DatabaseEntry->SymbolicNameLength; > + DbName.Length = DbName.MaximumLength; > + DbName.Buffer = (PWSTR)((ULONG_PTR)DatabaseEntry + > DatabaseEntry->SymbolicNameOffset); > + if (RtlEqualUnicodeString(&TargetVolumeName, &DbName, TRUE)) > + { > + ++DatabaseEntry->DatabaseOffset; > + Status = WriteRemoteDatabaseEntry(RemoteDatabase, Offset, > DatabaseEntry); > + FreePool(DatabaseEntry); > + Found = TRUE; > + break; > + } > + > + Offset += DatabaseEntry->EntrySize; > + FreePool(DatabaseEntry); > + } > + > + /* We couldn't find ourselves, we'll have to add ourselves */ > + if (!Found) > + { > + ULONG EntrySize; > + PUNIQUE_ID_REPLICATE UniqueIdReplicate; > + > + /* Query the device unique ID */ > + Status = QueryDeviceInformation(&TargetVolumeName, NULL, > &UniqueId, NULL, NULL, NULL, NULL, NULL); > + if (!NT_SUCCESS(Status)) > + { > + FreePool(SourceSymbolicName.Buffer); > + CloseRemoteDatabase(RemoteDatabase); > + return Status; > + } > + > + /* Allocate a database entry */ > + EntrySize = UniqueId->UniqueIdLength + TargetVolumeName.Length + > sizeof(DATABASE_ENTRY); > + DatabaseEntry = AllocatePool(EntrySize); > + if (DatabaseEntry == NULL) > + { > + FreePool(UniqueId); > + FreePool(SourceSymbolicName.Buffer); > + CloseRemoteDatabase(RemoteDatabase); > + return STATUS_INSUFFICIENT_RESOURCES; > + } > + > + /* Fill it in */ > + DatabaseEntry->EntrySize = EntrySize; > + DatabaseEntry->DatabaseOffset = 1; > + DatabaseEntry->SymbolicNameOffset = sizeof(DATABASE_ENTRY); > + DatabaseEntry->SymbolicNameLength = TargetVolumeName.Length; > + DatabaseEntry->UniqueIdOffset = TargetVolumeName.Length + > sizeof(DATABASE_ENTRY); > + DatabaseEntry->UniqueIdLength = UniqueId->UniqueIdLength; > + RtlCopyMemory((PVOID)((ULONG_PTR)DatabaseEntry + > sizeof(DATABASE_ENTRY)), TargetVolumeName.Buffer, > DatabaseEntry->SymbolicNameLength); > + RtlCopyMemory((PVOID)((ULONG_PTR)DatabaseEntry + > DatabaseEntry->UniqueIdOffset), UniqueId->UniqueId, > UniqueId->UniqueIdLength); > + > + /* And write it down */ > + Status = AddRemoteDatabaseEntry(RemoteDatabase, DatabaseEntry); > + FreePool(DatabaseEntry); > + if (!NT_SUCCESS(Status)) > + { > + FreePool(UniqueId); > + FreePool(SourceSymbolicName.Buffer); > + CloseRemoteDatabase(RemoteDatabase); > + return Status; > + } > + > + /* And now, allocate an Unique ID item */ > + UniqueIdReplicate = AllocatePool(sizeof(UNIQUE_ID_REPLICATE)); > + if (UniqueIdReplicate == NULL) > + { > + FreePool(UniqueId); > + FreePool(SourceSymbolicName.Buffer); > + CloseRemoteDatabase(RemoteDatabase); > + return Status; > + } > + > + /* To associate it with the device */ > + UniqueIdReplicate->UniqueId = UniqueId; > + InsertTailList(&DeviceInformation->ReplicatedUniqueIdsListHead, > &UniqueIdReplicate->ReplicatedUniqueIdsListEntry); > + } > + > + /* We're done with the remote database */ > + CloseRemoteDatabase(RemoteDatabase); > + > + /* Check we were find writing the entry */ > + if (!NT_SUCCESS(Status)) > + { > + FreePool(SourceSymbolicName.Buffer); > + return Status; > + } > + > + /* This is the end, allocate an associated entry */ > + AssociatedEntry = AllocatePool(sizeof(ASSOCIATED_DEVICE_ENTRY)); > + if (AssociatedEntry == NULL) > + { > + FreePool(SourceSymbolicName.Buffer); > + return STATUS_INSUFFICIENT_RESOURCES; > + } > + > + /* Initialize its source name string */ > + AssociatedEntry->String.Length = SourceSymbolicName.Length; > + AssociatedEntry->String.MaximumLength = > AssociatedEntry->String.Length + sizeof(UNICODE_NULL); > + AssociatedEntry->String.Buffer = > AllocatePool(AssociatedEntry->String.MaximumLength); > + if (AssociatedEntry->String.Buffer == NULL) > + { > + FreePool(AssociatedEntry); > + FreePool(SourceSymbolicName.Buffer); > + return STATUS_INSUFFICIENT_RESOURCES; > + } > + > + /* Copy data & insert in list */ > + RtlCopyMemory(AssociatedEntry->String.Buffer, > SourceSymbolicName.Buffer, SourceSymbolicName.Length); > + AssociatedEntry->String.Buffer[SourceSymbolicName.Length / > sizeof(WCHAR)] = UNICODE_NULL; > + AssociatedEntry->DeviceInformation = DeviceInformation; > + InsertTailList(&TargetDeviceInformation->AssociatedDevicesHead, > &AssociatedEntry->AssociatedDevicesEntry); > + > + /* We're done! */ > + FreePool(SourceSymbolicName.Buffer); > + return STATUS_SUCCESS; > } > > NTSTATUS > > Modified: trunk/reactos/drivers/filters/mountmgr/mntmgr.h > URL: > http://svn.reactos.org/svn/reactos/trunk/reactos/drivers/filters/mountmgr/m… > > ============================================================================== > --- trunk/reactos/drivers/filters/mountmgr/mntmgr.h [iso-8859-1] > (original) > +++ trunk/reactos/drivers/filters/mountmgr/mntmgr.h [iso-8859-1] Sun > Sep 13 22:52:07 2015 > @@ -298,6 +298,36 @@ > OUT PUNICODE_STRING VolumeName > ); > > +HANDLE > +OpenRemoteDatabase( > + IN PDEVICE_INFORMATION DeviceInformation, > + IN BOOLEAN MigrateDatabase > +); > + > +PDATABASE_ENTRY > +GetRemoteDatabaseEntry( > + IN HANDLE Database, > + IN LONG StartingOffset > +); > + > +NTSTATUS > +WriteRemoteDatabaseEntry( > + IN HANDLE Database, > + IN LONG Offset, > + IN PDATABASE_ENTRY Entry > +); > + > +NTSTATUS > +CloseRemoteDatabase( > + IN HANDLE Database > +); > + > +NTSTATUS > +AddRemoteDatabaseEntry( > + IN HANDLE Database, > + IN PDATABASE_ENTRY Entry > +); > + > /* device.c */ > > DRIVER_DISPATCH MountMgrDeviceControl; > @@ -458,4 +488,10 @@ > IN BOOLEAN MarkOffline > ); > > +NTSTATUS > +MountMgrQuerySymbolicLink( > + IN PUNICODE_STRING SymbolicName, > + IN OUT PUNICODE_STRING LinkTarget > +); > + > #endif /* _MNTMGR_H_ */ > > Modified: trunk/reactos/drivers/filters/mountmgr/mountmgr.c > URL: > http://svn.reactos.org/svn/reactos/trunk/reactos/drivers/filters/mountmgr/m… > > ============================================================================== > --- trunk/reactos/drivers/filters/mountmgr/mountmgr.c [iso-8859-1] > (original) > +++ trunk/reactos/drivers/filters/mountmgr/mountmgr.c [iso-8859-1] Sun > Sep 13 22:52:07 2015 > @@ -48,7 +48,6 @@ > * - MountMgrQueryDosVolumePaths > * - MountMgrQueryVolumePaths > * - MountMgrValidateBackPointer > - * - MountMgrVolumeMountPointCreated > * - MountMgrVolumeMountPointDeleted > * - ReconcileThisDatabaseWithMasterWorker > */ > > >

9 years, 3 months

Re: [ros-dev] [ros-diffs] [ion] 69180: [CDFS_NEW] - Resurrect back 7 years later (revert commit 34622). - Add correct MS-PL (Public) License. Not added to build/used, but I need some of the headers/data structures in here. ...

by Colin Finck

ion(a)svn.reactos.org wrote: > - Add correct MS-PL (Public) License. Not added to build/used, but I > need some of the headers/data structures in here. > - Source: https://github.com/Microsoft/Windows-driver-samples/blob/master/filesys/cdfs Yay! No more messing around with DDK licenses! Someone should take the other samples there and replace our DDK-inspired components (e.g. classpnp and disk) by these ones licensed under MS-PL. Cheers, Colin

9 years, 3 months

Re: [ros-dev] [ros-diffs] [hbelusca] 69139: OMGWTFBBQ!!!! We can now boot our ISOs on UEFI systems!! Thanks gigaherz for the "fatten" utility!, and others for testing. [CDMAKE] - Add multi-boot CD support, following El-Tori...

by Colin Finck

hbelusca(a)svn.reactos.org wrote: > +// FIXME! FIXME! Do it in a portable way!! > typedef unsigned char BYTE; > typedef unsigned short WORD; > typedef unsigned long DWORD; Try using our include/host/typedefs.h, which provides the most popular Windows type definitions in a portable way. - Colin

9 years, 3 months

Re: [ros-dev] [ros-diffs] [cwittich] 69079: [MODE] use ARRAYSIZE, don't include windows.h, fix two possible out-of-bounds reads CID 1206872 CID 1206873

by Colin Finck

cwittich(a)svn.reactos.org wrote: > Modified: trunk/reactos/base/applications/cmdutils/mode/mode.c What is that tool for anyway? Windows has a 16-bit "mode.com" for compatibility and I believe the actual "mode" command is built-in into cmd.exe these days. Cheers, Colin

9 years, 3 months

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

Ros-dev September 2015