Ros-dev June 2014

ros-dev@reactos.org

15 participants
15 discussions

by Colin Finck

Hi all, Fezile, one of our servers in Sweden, unexpectedly disappeared after a reboot today. Unfortunately, this is one of our older servers without an IPMI module, so we have to wait till it's rebooted manually. I'll keep you informed when this is done. The server outage affects the following services: * iso.reactos.org * doxygen.reactos.org * cppcheck.reactos.org * VMware Player Test slave * VMware Player Patchbot We apologize for the inconveniences! If you have any idea where we can still get a suitable IPMI module for these servers, we would be highly interested! The needed models are ASUS ASMB3 and Tyan M3291. Best regards, Colin Finck

10 years, 3 months

Status Meeting (June/July 2014)

by Aleksey Bragin

Hello, Let me invite you to the monthly status meeting taking place next Thursday, 3rd of July, 19:00 UTC. The meeting is one week off from the classic scheme due to various reasons (previous meeting was too close, and, you know, Germany - USA on Thursday). 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. If someone still is not getting passwords sent before a meeting - please email Pierre before the meeting started to get one. Please send agenda proposals to me before the meeting, so that we can start with a proposed agenda. Also all developers, please, may I so kindly ask you to prepare a very brief one-two lines message containing two things: 1. What you did since the previous meeting 2. What you are going to do in the coming month This is needed to understand who is working on what, to plan things, etc. Thanks, I hope it's not that much to ask for. Prepare that in advance! Everyone will have max 1 minute! Regards, Aleksey Bragin

10 years, 4 months

Re: [ros-dev] [ros-diffs] [hbelusca] 63610: [WIN32K] There is a bug in win32k (who would have thought that?) that consists in holding a winstation spinlock while running PAGED_CODE MmCopyToCaller function, when building the...

by Alex Ionescu

Win32k.sys is supposed to use MmPageEntireDriver, as such, it should run 100% at passive only. Best regards, Alex Ionescu On Tue, Jun 17, 2014 at 10:01 PM, <hbelusca(a)svn.reactos.org> wrote: > Author: hbelusca > Date: Tue Jun 17 20:01:23 2014 > New Revision: 63610 > > URL: http://svn.reactos.org/svn/reactos?rev=63610&view=rev > Log: > [WIN32K] > There is a bug in win32k (who would have thought that?) that consists in > holding a winstation spinlock while running PAGED_CODE MmCopyToCaller > function, when building the list of desktops of a given window station (the > bug is easily triggerable when calling EnumDesktopsW). Since this lock is > never used in anyplace but in this function, which, by the way, is just a > reader function that fills user buffer, I consider that it is safe to > remove this lock. However I want approval from win32k specialists. Hence I > just disable the code with a define USE_WINSTA_LOCK. If this lock is really > needed, please rewrite the BuildDesktopNameList function !! Otherwise > remove this lock and the associated code !! > This is a blocker for the shutdown code. > > Modified: > trunk/reactos/win32ss/user/ntuser/winsta.c > trunk/reactos/win32ss/user/ntuser/winsta.h > > Modified: trunk/reactos/win32ss/user/ntuser/winsta.c > URL: > http://svn.reactos.org/svn/reactos/trunk/reactos/win32ss/user/ntuser/winsta… > > ============================================================================== > --- trunk/reactos/win32ss/user/ntuser/winsta.c [iso-8859-1] (original) > +++ trunk/reactos/win32ss/user/ntuser/winsta.c [iso-8859-1] Tue Jun 17 > 20:01:23 2014 > @@ -451,7 +451,9 @@ > /* Initialize the window station */ > RtlZeroMemory(WindowStationObject, sizeof(WINSTATION_OBJECT)); > > +#ifdef USE_WINSTA_LOCK > KeInitializeSpinLock(&WindowStationObject->Lock); > +#endif > InitializeListHead(&WindowStationObject->DesktopListHead); > Status = RtlCreateAtomTable(37, &WindowStationObject->AtomTable); > WindowStationObject->Name = WindowStationName; > @@ -1203,7 +1205,9 @@ > { > NTSTATUS Status; > PWINSTATION_OBJECT WindowStation; > +#ifdef USE_WINSTA_LOCK > KIRQL OldLevel; > +#endif > PLIST_ENTRY DesktopEntry; > PDESKTOP DesktopObject; > DWORD EntryCount; > @@ -1220,7 +1224,9 @@ > return Status; > } > > +#ifdef USE_WINSTA_LOCK > KeAcquireSpinLock(&WindowStation->Lock, &OldLevel); > +#endif > > /* > * Count the required size of buffer. > @@ -1242,7 +1248,9 @@ > Status = MmCopyToCaller(pRequiredSize, &ReturnLength, > sizeof(ULONG)); > if (! NT_SUCCESS(Status)) > { > +#ifdef USE_WINSTA_LOCK > KeReleaseSpinLock(&WindowStation->Lock, OldLevel); > +#endif > ObDereferenceObject(WindowStation); > return STATUS_BUFFER_TOO_SMALL; > } > @@ -1253,7 +1261,9 @@ > */ > if (dwSize < ReturnLength) > { > +#ifdef USE_WINSTA_LOCK > KeReleaseSpinLock(&WindowStation->Lock, OldLevel); > +#endif > ObDereferenceObject(WindowStation); > return STATUS_BUFFER_TOO_SMALL; > } > @@ -1264,7 +1274,9 @@ > Status = MmCopyToCaller(lpBuffer, &EntryCount, sizeof(DWORD)); > if (! NT_SUCCESS(Status)) > { > +#ifdef USE_WINSTA_LOCK > KeReleaseSpinLock(&WindowStation->Lock, OldLevel); > +#endif > ObDereferenceObject(WindowStation); > return Status; > } > @@ -1280,7 +1292,9 @@ > Status = MmCopyToCaller(lpBuffer, DesktopName.Buffer, > DesktopName.Length); > if (! NT_SUCCESS(Status)) > { > +#ifdef USE_WINSTA_LOCK > KeReleaseSpinLock(&WindowStation->Lock, OldLevel); > +#endif > ObDereferenceObject(WindowStation); > return Status; > } > @@ -1288,7 +1302,9 @@ > Status = MmCopyToCaller(lpBuffer, &NullWchar, sizeof(WCHAR)); > if (! NT_SUCCESS(Status)) > { > +#ifdef USE_WINSTA_LOCK > KeReleaseSpinLock(&WindowStation->Lock, OldLevel); > +#endif > ObDereferenceObject(WindowStation); > return Status; > } > @@ -1298,7 +1314,9 @@ > /* > * Clean up > */ > +#ifdef USE_WINSTA_LOCK > KeReleaseSpinLock(&WindowStation->Lock, OldLevel); > +#endif > ObDereferenceObject(WindowStation); > > return STATUS_SUCCESS; > > Modified: trunk/reactos/win32ss/user/ntuser/winsta.h > URL: > http://svn.reactos.org/svn/reactos/trunk/reactos/win32ss/user/ntuser/winsta… > > ============================================================================== > --- trunk/reactos/win32ss/user/ntuser/winsta.h [iso-8859-1] (original) > +++ trunk/reactos/win32ss/user/ntuser/winsta.h [iso-8859-1] Tue Jun 17 > 20:01:23 2014 > @@ -7,11 +7,16 @@ > #define WSS_LOCKED (1) > #define WSS_NOINTERACTIVE (2) > > +// Uncomment for using WinSta spinlock > +// #define USE_WINSTA_LOCK > + > typedef struct _WINSTATION_OBJECT > { > DWORD dwSessionId; > > +#ifdef USE_WINSTA_LOCK > KSPIN_LOCK Lock; > +#endif > UNICODE_STRING Name; > LIST_ENTRY DesktopListHead; > PRTL_ATOM_TABLE AtomTable; > > >

10 years, 4 months

Jira comments

by Hermès BÉLUSCA - MAÏTO

Hey guys ! It seems that suddently Jira SVN comments work again. How is it possible? H.

10 years, 4 months

[CONSRV] Ensure szMenuStringBack is NULL terminated to prevent possible buffer overread and fix char to wchar_t assignment.

by Vyronas Tsingaras

Is this patch acceptable? Please kindly explain any possible mistakes.

10 years, 4 months

Re: [ros-dev] [ros-diffs] [jgardou] 63642: [SETUPAPI_APITEST] - Add a test that shows that SetupDiGetClassDevsExW doesn't fail on non-existing GUID

by Timo Kreuzer

10 years, 4 months

CRT changes

by Ged Murphy

Some interesting changes are coming to the CRT which the VS builds will need to prepare for. The obvious one is the %s/%S format changes http://blogs.msdn.com/b/vcblog/archive/2014/06/18/crt-features-fixes-and-bre aking-changes-in-visual-studio-14-ctp1.aspx Ged.

10 years, 5 months

by Timo Kreuzer

10 years, 5 months

Re: [ros-dev] [ros-diffs] [hbelusca] 63599: [TASKMGR]: Use the EndTask API to kill tasks.

by Pierre Schweitzer

On 15/06/2014 19:47, hbelusca(a)svn.reactos.org wrote: > + /* Trick: on Windows, pressing the CTRL key forces the task to be ended */ > + BOOL ForceEndTask = !!(GetKeyState(VK_CONTROL) & 0x8000); > + Was it really meant to be '!!'? -- Pierre Schweitzer<pierre at reactos.org> System Administrator ReactOS Foundation

10 years, 5 months

Re: [ros-dev] Upgrade complete

by Alexander Rechitskiy

All is down! > Fixed. > > On 15/06/2014 00:56, Alexander Rechitskiy wrote: > >> https://code.reactos.org/ does NOT work >> 15.06.2014, 02:40, "Pierre Schweitzer" <pierre(a)reactos.org>: >>> As a side note, upgrade to Fisheye 3.4.4 didn't fix ONLINE-461 (which was the >>> main reason behind the upgrade). >>> >>> On 15/06/2014 00:36, Zachary Gorden wrote: >>>> Fisheye was upgraded successfully, Jira went off the rails and took us >>>> about three hours to recover from. Please report any data loss as we had to >>>> restore from backups. >>>> >>>> >>>> >>>> _______________________________________________ >>>> Ros-dev mailing list >>>> Ros-dev(a)reactos.org <mailto:Ros-dev@reactos.org> >>>> http://www.reactos.org/mailman/listinfo/ros-dev >>> >>> >>> -- >>> Pierre Schweitzer<pierre at reactos.org> >>> System Administrator >>> ReactOS Foundation >>> , >>> >>> _______________________________________________ >>> Ros-dev mailing list >>> Ros-dev(a)reactos.org <mailto:Ros-dev@reactos.org> >>> http://www.reactos.org/mailman/listinfo/ros-dev >>> >> -- >> Best regards, >> >> _______________________________________________ >> Ros-dev mailing list >> Ros-dev(a)reactos.org >> http://www.reactos.org/mailman/listinfo/ros-dev > > -- > Pierre Schweitzer<pierre at reactos.org> > System Administrator > ReactOS Foundation -- -- Best regards,

10 years, 5 months

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

Ros-dev June 2014