- Removed all the code in userptrap.c and removed KiKernelTrapHandler.
- Set Traps 11, 12, 16 and 17 as unhandled for now(Segment fault, stack fault, fpu fault and alignment fault). We weren't really "handling" them in the past either.
- I probably need to implement GPF handler for non-V86 mode to support Lazy segment loading and fix the vmware syscall bug that has been haunting me for two years.
svn path=/trunk/; revision=23636
* font.c/layout.c/console.c/options.c: add UNREFERENCED_PARAMETER macros, fix sharing of global struct
* console.def: fix building with MSVC
svn path=/trunk/; revision=23629
- Handle FPU/SSE saving in the CONTEXT<->KTRAP_FRAME routines directly, and using KiFlushNPXState assembly function.
- Implement a naive Trap 7 (FPU Fault) handler in assembly that properly does most of the work required to handle an FPU fault. Not yet complete however since it almost ignores user-mode faults (just like trunk). FPU isn't done in ctx switching yet.
- Remove all the other .c code that dealt with FPU stuff and had become incompatible since more then a year ago.
svn path=/trunk/; revision=23628
modified porting-tools/rdesktop-core-tester/rdesktop-core-tester.cpp
And from the chrysalis out came... the caterpillar. MissTosca stops just short of being demonstrable. I'll keep working on it, but the Google Summer of Code is, for all intent and purposes, over. The matter is in the hands of my judges now. Wish me luck
modified rdesktop/iso.c
modified rdesktop/proto.h
modified rdesktop/rdesktop.h
modified rdesktop/rdp.c
Very minor fixes
modified rdesktop/tcp.c
Use asynchronous I/O and alertable waits to allow aborting the protocol thread ASAP
svn path=/trunk/; revision=23627
- Detect if kernel APCs are pending and request APC_LEVEL software interrupt from HAL if they are, returning with the right apc status.
- Also update INT21 VDM Descriptor handler when updating LDT descriptor.
svn path=/trunk/; revision=23624
- Detect and crash if context switch is done inside a DPC.
- Switch kernel stack a bit earlier
- Don't switch address space if we're still in the same process, and fix some duplicated code that was mixed up together.
- Move LDT setup out-of-line.
svn path=/trunk/; revision=23621
1. Bitstream Vera Sans to MS Sans Serif
2. Bitstream Vera Sans Mono to Courier New
This fixed some problem that bug 1191 mentor but not all.
I tested this change in ReactOS
1. using BootCD install whole process and run some ros own program
and tested cmd. No problem was found with this change.
svn path=/trunk/; revision=23614
modified porting-tools/rdesktop-core-tester/rdesktop-core-tester.vcproj
modified porting-tools/rdesktop-core-tester/stdafx.h
Support acting as both the stand-alone DLL (mstscax.dll) and the redistributable (msrdp.ocx)
Wrap the non-standard "sealed" keyword in a macro
Almost all state now correctly passed between rdesktop-core and the ActiveX
Some debugging aids
Prepared a functional DllMain
Pulled in Common Controls, needed for the full-screen titlebar
modified rdesktop/rdesktop.h
modified rdesktop/rdp.c
modified rdesktop/secure.c
Minor clean-up
svn path=/trunk/; revision=23613
- When compiled with gcc we assume that CPU can't do TSC (the function will be rewritten anyway since it's weird to use inline assembly in a driver)
- Changed to the correct .rc file, removed duplicate one (ehci.rc)
- usbdriver now compiles with reactos build system
svn path=/trunk/; revision=23610
1. No plans to support NT 3.x hives (thus no need for OldCell / NewCell)
2. Cmlib is a bit architecturally different from NT's implementation:
"... i don't need the UserData, because that is what i return ... i don't consider it part of the header anymore"
svn path=/trunk/; revision=23605
It supports UHCI/EHCI controllers.
Don't expect it to compile with ReactOS build environment. I intend to do some work on it, and use this place to keep my changes, before it's suitable for ReactOS.
"I would like to see it useful for ReactOS" (c) woodhead
svn path=/trunk/; revision=23602
