* Work around a Clang bug w.r.t. dwarf emission. Brought to you by the mighty Arty.
CORE-8516 #comment Now ReactOS can compile a bootcd with Clang, and it boots to shell. Only some PSEH bug(s) remain for it to be fully functional and ready for testman (I'm looking at you, Timo ;) )
svn path=/trunk/; revision=64203
- Do not make callouts to user-mode while holding a lock on GDI objects.
Fixes crashes in user32_winetest:win and :msg
svn path=/trunk/; revision=64189
- Enable all debug channels if DEBUGCHANNEL is set to "+all".
- Fix GDI objects exclusive locks counting, fixing a memory corruption altogether.
- Add a missing lock release on error path.
- Add a debug print.
svn path=/trunk/; revision=64187
- Implement RtlpCloseRegistryHandle helper, which closes a registry key, if RTL_REGISTRY_HANDLE was not passed as RelativeTo
- Use RtlpCloseRegistryHandle whereever required, fixing some cases, where we unconditionally closed the handle
svn path=/trunk/; revision=64186
- Add dummy INT 18h (ROM Basic) and INT 19h (Bootstrap loader) interrupts. If INT 18h will never be implemented, INT 19h can be used one day...
- Add the standard bootstrap code snippets at F000:FFF0 and F000:E05B.
- For compatibility purposes, and nice info reporting via tools like Microsoft Diagnostics, or Symantec SysInfo, or whatever else..., register nice BIOS Copyright and Version strings. Also add a version string at the end of the BIOS.
- Fix our reported BIOS model.
- Add some temporary documentation (from Bochs BIOS) about the ROM BIOS compatibility entry-points that some apps may use (instead of passing through more documented & standard interfaces).
svn path=/trunk/; revision=64185
- Slap *a bit* of sense in the way we use region objects and handles, avoiding a massive orgy of handle allocations, locks, leaks and so on.
The motto here being "if you don't need a valid user-mode handle, then chances are that you don't need one at all."
CORE-8506 #resolve #comment Stale region handles should not be passed anymore to user-mode apps now.
svn path=/trunk/; revision=64177
- Fix up values got from win32k in GetOutlineTextMetrics and do not fail if the provided buffersize is only sizeof(OUTLINETEXTMETRICW).
CORE-8507 #resolve
svn path=/trunk/; revision=64176
- Add Etw trace api stubs (moved from avdapi32)
- Add exports for Etw* functions
[ADVAPI32]
- Add forwarders of trace api functions to ntdll Etw* functions
svn path=/trunk/; revision=64172
CORE-8537 #resolve #comment Demote UNIMPLEMENTED_DBGBREAK for UNIMPLEMENTED in the non-implemented case of PopFlushVolumes. Fixes the apparent BSOD.
svn path=/trunk/; revision=64170
SampRemoveUserFromAllGroups:
- Do not treat a STATUS_MEMBER_NOT_IN_GROUP, returned by SampRemoveMemberFromGroup, as error but as success.
- Clear the Groups attribute at the end.
svn path=/trunk/; revision=64169
- Move the mouse driver to DOS since it's our 32bit version of mouse.com, in some sense,
- Temporarily disable the mouse user callbacks because currently they went called in a parallel thread than the CPU one, that caused CPU corruption,
- In the same way we currently trigger IRQ1 interrupts, we do the same for IRQ12,
- Make the console input pump more modular (and mouse presence detection code is WIP),
- Put keyboard code into a dedicated file (as done for the mouse),
- Adapt the existing PS/2 emulation code to support two PS/2 ports controlled by 1 controller. Please note that some documentations precise that the response byte (answer to a PS/2 controller command) is written directly to the PS/2 output port.
svn path=/trunk/; revision=64168
- Code formatting and precise an info in a comment,
- Remove an unneeded commented call, and
- Mention that INT 15h functions C1h and C2h need to be implemented for PS/2 support.
svn path=/trunk/; revision=64167
- As pid and tid 0 have a special meaning in GDB, use off-by-one thread and process ID when communicating with it
- Properly read registers and memory from foreign thread and processes. (This time it was tested and proved to work reliably. __writecr3 ftw!)
- Loop the list of processes and threads when trying to find them from ID, as PsLookupProessByThreadId and friends can't be used since we can be at any IRQL.
- Add a few more debug prints to help diagnosing problems.
CORE-8531
svn path=/trunk/; revision=64166
- Read MaxSize and Retention from the Registry when a new log file is created.
- Use ntdll file functions instead of their kernel32 counterparts.
svn path=/trunk/; revision=64159
- Don't use svn:eol-style = native for PCI ids but then have the code expect CRLF. Should fix all devices being "Unknown device" on Unix builds.
svn path=/trunk/; revision=64158
