We will continue with this stuff when we can understand the root issues better
* Revert "[NTDLL_APITEST] Relicense implicit tls tests as MIT"
This reverts commit ec6bae2b31.
* Revert "[NTDLL:LDR] Add Implicit Thread Local Storage Support (#7594)"
This reverts commit 23043ce0b8.
Before Windows Vista, TLS directories were ignored by Ldr* in any module loaded after process creation, which meant that the TLS index in any delay-loaded module remained 0. This means that when running an application targeting Vista and above and its delay-load modules have TLS directories, the following events can happen:
One or more delay-load modules claim ThreadLocalStoragePointer idx 0, resulting in data corruption and possible buffer overflows as data intended for one of the immediate-load modules is overwritten.
None of the immediate-load modules has a TLS directory, so ThreadLocalStoragePointer is not initialized. This means that the program crashes when one of the delay-load modules attempts to access its TLS data.
Adding implicit TLS support improves support for NT 6.x/10 user-mode applications, including all contemporary major web browsers.
JIRA issue: CORE-19413
---------
Co-authored-by: Justin Miller <[email protected]>
Implement DLL Load Notification, an NT6+ feature.
https://learn.microsoft.com/en-us/windows/win32/devnotes/dll-load-notification
- [RTL] Sync `RTL_STATIC_LIST_HEAD` and `RtlFailFast` from XDK to NDK.
- [NTDLL_VISTA] Introduce ntdll_vista_static static library and link both ntdll_vista and ntdll to it.
- [NDK][LDR] Add and fix DLL Notification definitions.
- [NTDLL_VISTA] Code improvements.
- [NTDLL_VISTA:LDR] Implement Dll Notification.
- [NTDLL][NTDLL_APITEST] Add Dll Notification API test.
This removes all fake apiset forwarders,
and handles apisets inside ntdll.
This is not 100% compatible with how windows does it, but it should be good enough for us.
Make LdrUnlockLoaderLock Cookie type consistent with LdrLockLoaderLock
and LdrpMakeCookie functions.
In addition:
* Adjustment of the formatting string for ULONG_PTR;
* ntdll.spec: specify ptr for 2nd parameter of LdrUnlockLoaderLock:
This is a parameter whose length is pointer-like, platform-dependent.
Co-authored-by: Hermès BÉLUSCA - MAÏTO <[email protected]>
- Uniform function declarations in .c and .h with fixes and SAL2 improvements.
- Use LDR_[UN]LOCK_LOADER_LOCK_FLAG_XXX correctly.
- Fix incorrect UsedSize calculation in LdrQueryProcessModuleInformationEx (caught by ThFabba).
- Whatever the callback request stop the enumeration or not in LdrEnumerateLoadedModules, the following operations should be the same.
- Fix 2 incorrect DPRINT1 that printed incorrect parameter.
- Return error if RtlAllocateHeap failed in LdrpGetProcedureAddress, and add comments about NT6.2 new changes.
- Convert some notations to SAL2
- "Name" parameter in LdrGetProcedureAddress/LdrpGetProcedureAddress should be optional
- "DllCharacteristics" parameter in LdrGetDllHandle should be optional
We found out what we wanted to find out:
the remaining TODOs for the Python script.
So we can silence the DPRINTs again for the time being
to reduce spam from 3rd party dlls.
Paying tribute to Mark Jansen ;)
This reverts
0.4.13-dev-284-g 74be88ac47
The logging is helpful to alert us when we should really
recalculate the base addresses for performance reasons.
Spams a bit sometimes, sure, but it's worth it!
Otherwise we'd call ZwQuerySection on an invalid handle and get:
(ntoskrnl/mm/section.c:4320) Failed to reference section: 0xc0000008
Also correctly check DllCharacteristics: If the argument is missing, the
file is _not_ a system file, i.e. we should check for restrictions.
Timo Kreuzer