Patch by Jose Catena:
allmul produced wrong results when the higher 32 bits of any of the 64 bit operands are not zero. As the comment correctly said the definition of param locations must change
since esp was altered, but the definitions were not actually changed
See issue #5121 for more details.
svn path=/trunk/; revision=45088
- Switch to the new user handle manager. Added more handle types to win32k. One or two step to WND, desktop heap supported.
- Test with every application in my test suite. Needs more testing and clean up. Bleeding edge!
svn path=/trunk/; revision=45085
34969 Some ULONG / SIZE_T fixes, handle amd64 in PspCreateThread
34986 Fix several 64 bit issues
35968 Fix Size parameter of CmpAllpcate (ULONG->SIZE_T).
35970 Fix pointer <-> ULONG cast
35971 Fix pointer <-> ULONG cast fixes
36332 fix 64 bit warnings
37435 Fix several ULONG/ULONG_PTR issues. Add a parameter to DefaultQueryInfoBufferCheck for ULONG_PTR/SIZE_T ResultLength
39522 ExInitializeSystemlookasideList: Use InitializeSListHead instead of directly accessing the members. ExpAllocateLocallyUniqueId: use InterlockedCompareExchange64 instead of non-portable ExfInterlockedCompareExchange64
37433 fix InterlockedExchangePointer usage
38105 Use KeGetCurrentPrcb() instead of KeGetPcr()->Prcb / KeGetPcr()->CurrentPrcb
39160 Use wbinvd instruction for amd64, too.
svn path=/trunk/; revision=45066
[NTOS]: Optimize trap entry/exit by manually copying registers to the trap frame the correct way from the first time. Avoids conversion between PUSHA and KTRAP_FRAMEs and makes trap frames compatible the whole time (other than being slightly faster).
[NTOS]: Provide compiler with hints on likely code paths during trap entry and exit, which makes the code more linear and improves performance. The following assumptions (known to be true) are made: (1) Interrupts happen more often than system calls (per unit of time), so prioritize paths we take during interrupts. (2) The CPU spends most of its time in Ring 3, so prioritize traps from user-mode. (3) V8086 mode, debugging, 16-bit stacks, are uncommon, so de-prioritize them.
[NTOS]: Use KTRAP_FRAME offset names recommended by Timo instead of substraction which was confusing some people (still seems clearer to me).
svn path=/trunk/; revision=45064
KiExitV86Mode is supposed to return a pointer to the protected mode KTRAP_FRAME. Instead it was returning the value of TrapFrame->Edi, which would be identical to PmTrapFrame->Ebp, which doesn't have any particular meaning. Fix it by returning PmTrapFrame directly. Fixes 2nd stage boot.
svn path=/trunk/; revision=45062
[NTOS]: Add VDM debug spew to see why there's now an invalid opcode on Windows builds of VMWare and certain QEmu combinations. (Note: the double fault issue is fixed, this is a new issue).
svn path=/trunk/; revision=45057
[NTOS]: A trap can get us into a state where DS/ES are invalid, making any pointer dereference (on DS/ES segmented memory, not SS, the stack) crash (and probably double-fault). Therefore, we have to be careful to switch to a good DS/ES before touching the TrapFrame pointer, which we don't have in ESP like the ASM code, but in a DS/ES-segmented register. For V8086 traps we can switch to the good DS/ES immediately, but for other kinds of traps, we actually need to save the current (bad) segments first. So we save them on the stack now, then switch to the good ones, then store the stack values into the trap frame. This is what happens on a non-optimized (-O0) build. On an optimized build, the segments will end up in registers instead, which is fine too (they'll be direct values). The order of instructions is guaranteed since the segment macros are volatile.
[NTOS]: The GPF and Invalid Opcode handlers are performance critical when talking about V8086 traps, because they control the main flow of execution during that mode (GPFs will be issued for any privileged instruction we need to emulate, and invalid opcode might be generated for BOPs). Because of this, we employ a fast entry/exit macro into V8086 mode since we can make certain assumptions. We detect, and use, such scenarios when the V8086 flag is enabled in EFLAGS. However, because we can land in a GPF handler with an invalid DS/ES, as some V8086 code could trample this during BIOS calls for example, we must make sure that we are on a valid DS/ES before dereferencing any pointer. We fixup DS/ES either in KiEnterTrap (for normal entry/exit) or, for V86, in KiEnterV86Trap. Notice the problem: we need to detect which of these to use early on but we can't touch the EFLAGS in the frame because DS/ES could be invalid. Thankfully SS is always guaranteed valid, so stack dereferences are game! We therefore read the EFLAGS here, in assembly, where we can touch ESP as we please. We save this in EDX, which will be used as the second argument for the FASTCALL C trap entry. When we make the fast V86 check, we use the parameter instead of the trap frame, leading us to using the correct trap entry function, which fixes up DS/ES and lets us go on our merry way...
[NTOS]: Make appropriate changes to GENERATE_TRAP_HANDLERS macro.
[NTOS]: Switch to using well-known NT trap handler names (hex-based, double-zeroed) instead of decimal-based trap handler names which are confusing.
[NTOS]: Clean up some debug spew.
svn path=/trunk/; revision=45052
[NTOS]: Fix a logic bug in KiExitTrap: Always only restore segments if we came from user-mode (since they might be bogus on a kernel transition as they're not always saved), even if the caller wants segment restore.
[NTOS]: Small perf boot: do a JMP, not a CALL into C handling code.
[NTOS]: Make KiGetTickCount/KiCallbackReturn handled in C (as stubs).
[NTOS]: Implement KeSynchronizeExecution in C. Move Kei386SpinOnSpinLock to C stub.
[NTOS]: Implement overall architecture for handling hardware interrupts in C. Not used yet, since it needs C code in HAL.
svn path=/trunk/; revision=45045
- EhciDefferedRoutine: Fix late night brain fart by filling the members of the WorkItemData before queuing the work item.
svn path=/trunk/; revision=45043
- More movement to WND with updates to related files. Added new scroll and get/post message routines. In the mix we did moved to the MONITOR structure.
svn path=/trunk/; revision=45042
[NTOS]: Implement KiCoprocessorError in C.
[NTOS]: Make NMI handler fully C, now that the other parts are C too.
[NTOS]: Delete more ASM macros and code that are now unused.
svn path=/trunk/; revision=45040