- Do not free resources at process termination.
- Free the provider catalog after the handle table, since the latter references the former.
svn path=/trunk/; revision=66337
- Do not merge memory blocks if they don't belong to the same VM "region" (ie not allocated from the same NtAllocateVirtualMemory call)
Fixes failures of ZwProtectVirtualMemory calls.
Bugs see DPH rollin', they hatin'
svn path=/trunk/; revision=66336
- MiIsEntireRangeCommitted: Ensure the PTE we are checking is really faulted in.
- Prefer MiPteToPde and MiPdeToPte (which should really be called MiFirstPteInPde) instead of MiAddressToPte and MiPteToAddress
Fixes weird failed ASSERT in page fault handler when using DPH.
svn path=/trunk/; revision=66334
- Load default application icon and display it in the taskbar for windows that have no icon. Fixes "GetIconInfo failed" debug prints all over the place (and looks prettier).
CORE-9211 #resolve
svn path=/trunk/; revision=66333
- Fix wine msg test_SetFocus tests. Use send message for WM_WINDOWPOSCHANGING. If nothing has been drawn and the no position changed bits set, set no Z order and no redraw. See CORE-7447.
svn path=/trunk/; revision=66332
- Add /n option to suppress console output. May or may not help to speed up testbot runs, but can't hurt to have in here.
svn path=/trunk/; revision=66318
- Abstract unidirectional anonymous pipes into a CPipe class
- Abstract a process with redirected output into a CPipedProcess class
- Use these abstractions to avoid polling for output from test processes. Instead, use blocking read operations to yield the CPU while waiting for data.
ROSTESTS-144 #resolve
svn path=/trunk/; revision=66316
- Sync to Wine 1.7.36
- Properly disable test_dib_formats. ROSTESTS-152, CORE-5922
- Fix skip message in test_mono_bitmap not to point to a resolved bug. ROSTESTS-153, CORE-5922
- Re-enable font tests that were left skipped even though the associated bug is correctly marked as resolved. ROSTESTS-8
svn path=/trunk/; revision=66312
- Fix wine msg:test_winevents.
- Fix global hooks while sending to same process different thread. Yes, interesting.
- Fix event hooks, use posting (faster) instead of sending. This fixed a problem while sending to the same thread.
svn path=/trunk/; revision=66310
NtQueryInformationFile: Implement FileAccessInformation and FileAlignmentInformation cases according to 'File System Internals' page 485.
svn path=/trunk/; revision=66307
- Part 2/2 of SrvLogon (see r66303): load the per-user shutdown timeouts.
- Retrieve those per-user shutdown timeouts from the registry when the user logs on, adapted from Alex' shutdown patch plus existing code.
- Commit the main part of shutdown code (finally!), still unfinished yet:
* Need to switch to the desktop where the hanging GUI app's window is present,
* Need to deal with apps from other users
* What about SYSTEM processes?
* What about console processes?
- Reuse the old timeout code.
- Fix the sending of WM_QUERYENDSESSION and WM_ENDSESSION messages (this is done by win32k directly; for that winsrv needs just to send one WM_CLIENTSHUTDOWN message with the correct wParam parameter).
Part 13/X
CORE-8322 #comment Big commit in revision 66306!
svn path=/trunk/; revision=66306
- In ExpReleaseOrWaitForKeyedEvent, always restore previous values for ETHREAD::KeyedWaitValue and ETHREAD::KeyedWaitChain, as these fields have alternative meanings outside of keyed events
- Add missing list walk in ExpReleaseOrWaitForKeyedEvent
- In Nt(WaitFor|Release)KeyedEvent, refuse keys that are not two-byte aligned as shown by Wine tests
- Don't forget to set KTHREAD::Process on thread creation
Fixes hang when running ntdll_winetest:om as well as failing tests.
ROSTESTS-118 #resolve
svn path=/trunk/; revision=66304
- Skip over service timeout tests because they take too long. This is not a bug, the tests are completing successfully -- however they are designed to cause service manager timeouts and take two minutes in doing so. Testbot time is too valuable to do this on every commit.
ROSTESTS-151
svn path=/trunk/; revision=66302
- Request only the access rights that are actually required in openTcpFile. Fixes many should-be access denied errors originating from netshell. While this doesn't fix the root cause of CORE-9172, it is a required step if we ever want netshell to work for non-Admin users.
CORE-9172
svn path=/trunk/; revision=66300
Add a stub for the "Windows Management Infrastructure" service.
This helps the Intel PRO eth card installer to go a bit farther
svn path=/trunk/; revision=66295
