reactos/ntoskrnl/ps/query.c

2814 lines
91 KiB
C
Raw Permalink Normal View History

Thread/Process Termination/Repeaing Rewrite + Fixes --------------------------------------------------- - ps/cid.c: * Moved CID Lookup functions here - ps/security.c: * Moved all security related functions here. Makes other files neater and security functions easier to locate. - ps/thread.c: * Moved most of the Thread Scheduling/Dispatching code that belongs in the Kernel to /ke and renamed functions from Ps to Ki. * Implemented PsIsSystemThread. * Removed Reaper Thread Init (now obsolete). * Renamed PiDeleteThread to PspDeleteThread. * Moved Thread State functions from tinfo.c to here. - ps/process.c: * Removed Query/Set Process functions and moved to ps/query.c * Renamed PiDeletePRocess to PspDeleteProcess * Removed obsoleted Process Termination functions, moved persistent one to kill.c - ps/create.c: * Moved the security APIs to security.c * Correctly implemented PsCreateSystemThread to actually create system threads. - ps/suspend.c * Rewrote Nt Executive functions to use Kernel functions. * Moved Ps* Routines into ke/kthread.c and fixed them. The implementation was wrong in some aspects, especially the issue of the APC looping around the KeWaitXxx call and the fact that the routines excluded/ignored the FreezeCount. - ps/debug.c * Fixed completely broken implementation of Get/SetThreadContext. The old version crashed when called and did not work at all. Suspend Regression test now works. * Moved Context<->TrapFrame functions to ke/i386/ * Combined Set/GetThreadContext APCs into a single one, and used special context structure. - ps/query.c: * Moved Thread/Process Query/Set Routines here. - ps/tinfo.c: * Removed. - ps/kill.c * Removed complicated Process Termination semantics and useless Attach/Detach in favor for a much more lightweight function which performs the same tasks as before and actually works. TaskManager can now terminate foreign processes. * Rewrote Thread Reaping to use the HyperCritical Work Queue instead of manually controlled thread. This results in much less code as well as an increase in speed and less micro management. The reaper is PspReapRoutine. Closing CMD.EXE now works properly without requiring masks that were added as hacks to allow it. * Renamed PiTerminateProcessThreads to PspTerminateProcessThreads. Fixed it to work with new termination code. * Added PspDeleteProcess to handle Process Object deletion. Kills the CID Handle here as done by Hartmut. * Added PspDeletethread here. * Renamed and rewrote PsTerminateCurrentThread to PspExitThread. Used NT Implementation out- lined in Windows Internals, Chapter 13. Uses less locks, a more concise order of actions, actually parses the Termination Ports, handles Dbgk notification. Timers are now rundown, and Mutex rundown is in a dedicated Kernel function. Final termination handled by KeTerminate Thread as documented. * Renamed PsTerminateOtherThread to PspTerminateThreadByPointer and modified implementation to be compatible with the changes above. * Renamed and regrouped Process Termination into PspExitProcess. Also implemented as described above, and moved each subsystem specific termination helper into its own subsytem. * Improved NtTerminateProcess and added more debugging messages. * Improved NtTerminateThread and added check against System Thread and made it compatible with new implementation. * Corrected PsTerminateSystemThread now that we support System Threads. * Corrected NtRegisterThreadTerminatePort to use same structure name as on windows for the port, and added tag to pool allocation (documented in pooltag.txt) include/internal/*.h: * Defined Scheduler Functions and misc new functions or renamed functions. ke/apc.c: * Fixed critical bug where APCs were not delivered at all if the thread wastion and cancels any timers that are associated to a thread, as well as their APCs and DPCs. REGRESSIONS FOUND: NONE BUGS/REGRESSIOSN FIXED: * Thread/Get Set Context now works. * Suspend Regression test now works. * Task manager can now kill foreign processes, even hung ones (like it should). * ExitProcess/closing cmd.exe with the 'x' button now works correctly without hacks. KNOWN ISSUES: I left a bit of a mess in the headers and some stuff still needs to be moved into the right places. I just wanted to have this first part ready first, so that it won't get too big. svn path=/trunk/; revision=14174
2005-03-18 05:53:04 +00:00
/*
* PROJECT: ReactOS Kernel
* LICENSE: GPL - See COPYING in the top level directory
Thread/Process Termination/Repeaing Rewrite + Fixes --------------------------------------------------- - ps/cid.c: * Moved CID Lookup functions here - ps/security.c: * Moved all security related functions here. Makes other files neater and security functions easier to locate. - ps/thread.c: * Moved most of the Thread Scheduling/Dispatching code that belongs in the Kernel to /ke and renamed functions from Ps to Ki. * Implemented PsIsSystemThread. * Removed Reaper Thread Init (now obsolete). * Renamed PiDeleteThread to PspDeleteThread. * Moved Thread State functions from tinfo.c to here. - ps/process.c: * Removed Query/Set Process functions and moved to ps/query.c * Renamed PiDeletePRocess to PspDeleteProcess * Removed obsoleted Process Termination functions, moved persistent one to kill.c - ps/create.c: * Moved the security APIs to security.c * Correctly implemented PsCreateSystemThread to actually create system threads. - ps/suspend.c * Rewrote Nt Executive functions to use Kernel functions. * Moved Ps* Routines into ke/kthread.c and fixed them. The implementation was wrong in some aspects, especially the issue of the APC looping around the KeWaitXxx call and the fact that the routines excluded/ignored the FreezeCount. - ps/debug.c * Fixed completely broken implementation of Get/SetThreadContext. The old version crashed when called and did not work at all. Suspend Regression test now works. * Moved Context<->TrapFrame functions to ke/i386/ * Combined Set/GetThreadContext APCs into a single one, and used special context structure. - ps/query.c: * Moved Thread/Process Query/Set Routines here. - ps/tinfo.c: * Removed. - ps/kill.c * Removed complicated Process Termination semantics and useless Attach/Detach in favor for a much more lightweight function which performs the same tasks as before and actually works. TaskManager can now terminate foreign processes. * Rewrote Thread Reaping to use the HyperCritical Work Queue instead of manually controlled thread. This results in much less code as well as an increase in speed and less micro management. The reaper is PspReapRoutine. Closing CMD.EXE now works properly without requiring masks that were added as hacks to allow it. * Renamed PiTerminateProcessThreads to PspTerminateProcessThreads. Fixed it to work with new termination code. * Added PspDeleteProcess to handle Process Object deletion. Kills the CID Handle here as done by Hartmut. * Added PspDeletethread here. * Renamed and rewrote PsTerminateCurrentThread to PspExitThread. Used NT Implementation out- lined in Windows Internals, Chapter 13. Uses less locks, a more concise order of actions, actually parses the Termination Ports, handles Dbgk notification. Timers are now rundown, and Mutex rundown is in a dedicated Kernel function. Final termination handled by KeTerminate Thread as documented. * Renamed PsTerminateOtherThread to PspTerminateThreadByPointer and modified implementation to be compatible with the changes above. * Renamed and regrouped Process Termination into PspExitProcess. Also implemented as described above, and moved each subsystem specific termination helper into its own subsytem. * Improved NtTerminateProcess and added more debugging messages. * Improved NtTerminateThread and added check against System Thread and made it compatible with new implementation. * Corrected PsTerminateSystemThread now that we support System Threads. * Corrected NtRegisterThreadTerminatePort to use same structure name as on windows for the port, and added tag to pool allocation (documented in pooltag.txt) include/internal/*.h: * Defined Scheduler Functions and misc new functions or renamed functions. ke/apc.c: * Fixed critical bug where APCs were not delivered at all if the thread wastion and cancels any timers that are associated to a thread, as well as their APCs and DPCs. REGRESSIONS FOUND: NONE BUGS/REGRESSIOSN FIXED: * Thread/Get Set Context now works. * Suspend Regression test now works. * Task manager can now kill foreign processes, even hung ones (like it should). * ExitProcess/closing cmd.exe with the 'x' button now works correctly without hacks. KNOWN ISSUES: I left a bit of a mess in the headers and some stuff still needs to be moved into the right places. I just wanted to have this first part ready first, so that it won't get too big. svn path=/trunk/; revision=14174
2005-03-18 05:53:04 +00:00
* FILE: ntoskrnl/ps/query.c
* PURPOSE: Process Manager: Thread/Process Query/Set Information
* PROGRAMMERS: Alex Ionescu (alex.ionescu@reactos.org)
* Thomas Weidenmueller (w3seek@reactos.org)
* Eric Kohl
Thread/Process Termination/Repeaing Rewrite + Fixes --------------------------------------------------- - ps/cid.c: * Moved CID Lookup functions here - ps/security.c: * Moved all security related functions here. Makes other files neater and security functions easier to locate. - ps/thread.c: * Moved most of the Thread Scheduling/Dispatching code that belongs in the Kernel to /ke and renamed functions from Ps to Ki. * Implemented PsIsSystemThread. * Removed Reaper Thread Init (now obsolete). * Renamed PiDeleteThread to PspDeleteThread. * Moved Thread State functions from tinfo.c to here. - ps/process.c: * Removed Query/Set Process functions and moved to ps/query.c * Renamed PiDeletePRocess to PspDeleteProcess * Removed obsoleted Process Termination functions, moved persistent one to kill.c - ps/create.c: * Moved the security APIs to security.c * Correctly implemented PsCreateSystemThread to actually create system threads. - ps/suspend.c * Rewrote Nt Executive functions to use Kernel functions. * Moved Ps* Routines into ke/kthread.c and fixed them. The implementation was wrong in some aspects, especially the issue of the APC looping around the KeWaitXxx call and the fact that the routines excluded/ignored the FreezeCount. - ps/debug.c * Fixed completely broken implementation of Get/SetThreadContext. The old version crashed when called and did not work at all. Suspend Regression test now works. * Moved Context<->TrapFrame functions to ke/i386/ * Combined Set/GetThreadContext APCs into a single one, and used special context structure. - ps/query.c: * Moved Thread/Process Query/Set Routines here. - ps/tinfo.c: * Removed. - ps/kill.c * Removed complicated Process Termination semantics and useless Attach/Detach in favor for a much more lightweight function which performs the same tasks as before and actually works. TaskManager can now terminate foreign processes. * Rewrote Thread Reaping to use the HyperCritical Work Queue instead of manually controlled thread. This results in much less code as well as an increase in speed and less micro management. The reaper is PspReapRoutine. Closing CMD.EXE now works properly without requiring masks that were added as hacks to allow it. * Renamed PiTerminateProcessThreads to PspTerminateProcessThreads. Fixed it to work with new termination code. * Added PspDeleteProcess to handle Process Object deletion. Kills the CID Handle here as done by Hartmut. * Added PspDeletethread here. * Renamed and rewrote PsTerminateCurrentThread to PspExitThread. Used NT Implementation out- lined in Windows Internals, Chapter 13. Uses less locks, a more concise order of actions, actually parses the Termination Ports, handles Dbgk notification. Timers are now rundown, and Mutex rundown is in a dedicated Kernel function. Final termination handled by KeTerminate Thread as documented. * Renamed PsTerminateOtherThread to PspTerminateThreadByPointer and modified implementation to be compatible with the changes above. * Renamed and regrouped Process Termination into PspExitProcess. Also implemented as described above, and moved each subsystem specific termination helper into its own subsytem. * Improved NtTerminateProcess and added more debugging messages. * Improved NtTerminateThread and added check against System Thread and made it compatible with new implementation. * Corrected PsTerminateSystemThread now that we support System Threads. * Corrected NtRegisterThreadTerminatePort to use same structure name as on windows for the port, and added tag to pool allocation (documented in pooltag.txt) include/internal/*.h: * Defined Scheduler Functions and misc new functions or renamed functions. ke/apc.c: * Fixed critical bug where APCs were not delivered at all if the thread wastion and cancels any timers that are associated to a thread, as well as their APCs and DPCs. REGRESSIONS FOUND: NONE BUGS/REGRESSIOSN FIXED: * Thread/Get Set Context now works. * Suspend Regression test now works. * Task manager can now kill foreign processes, even hung ones (like it should). * ExitProcess/closing cmd.exe with the 'x' button now works correctly without hacks. KNOWN ISSUES: I left a bit of a mess in the headers and some stuff still needs to be moved into the right places. I just wanted to have this first part ready first, so that it won't get too big. svn path=/trunk/; revision=14174
2005-03-18 05:53:04 +00:00
*/
/* INCLUDES ******************************************************************/
#include <ntoskrnl.h>
#define NDEBUG
#include <debug.h>
Thread/Process Termination/Repeaing Rewrite + Fixes --------------------------------------------------- - ps/cid.c: * Moved CID Lookup functions here - ps/security.c: * Moved all security related functions here. Makes other files neater and security functions easier to locate. - ps/thread.c: * Moved most of the Thread Scheduling/Dispatching code that belongs in the Kernel to /ke and renamed functions from Ps to Ki. * Implemented PsIsSystemThread. * Removed Reaper Thread Init (now obsolete). * Renamed PiDeleteThread to PspDeleteThread. * Moved Thread State functions from tinfo.c to here. - ps/process.c: * Removed Query/Set Process functions and moved to ps/query.c * Renamed PiDeletePRocess to PspDeleteProcess * Removed obsoleted Process Termination functions, moved persistent one to kill.c - ps/create.c: * Moved the security APIs to security.c * Correctly implemented PsCreateSystemThread to actually create system threads. - ps/suspend.c * Rewrote Nt Executive functions to use Kernel functions. * Moved Ps* Routines into ke/kthread.c and fixed them. The implementation was wrong in some aspects, especially the issue of the APC looping around the KeWaitXxx call and the fact that the routines excluded/ignored the FreezeCount. - ps/debug.c * Fixed completely broken implementation of Get/SetThreadContext. The old version crashed when called and did not work at all. Suspend Regression test now works. * Moved Context<->TrapFrame functions to ke/i386/ * Combined Set/GetThreadContext APCs into a single one, and used special context structure. - ps/query.c: * Moved Thread/Process Query/Set Routines here. - ps/tinfo.c: * Removed. - ps/kill.c * Removed complicated Process Termination semantics and useless Attach/Detach in favor for a much more lightweight function which performs the same tasks as before and actually works. TaskManager can now terminate foreign processes. * Rewrote Thread Reaping to use the HyperCritical Work Queue instead of manually controlled thread. This results in much less code as well as an increase in speed and less micro management. The reaper is PspReapRoutine. Closing CMD.EXE now works properly without requiring masks that were added as hacks to allow it. * Renamed PiTerminateProcessThreads to PspTerminateProcessThreads. Fixed it to work with new termination code. * Added PspDeleteProcess to handle Process Object deletion. Kills the CID Handle here as done by Hartmut. * Added PspDeletethread here. * Renamed and rewrote PsTerminateCurrentThread to PspExitThread. Used NT Implementation out- lined in Windows Internals, Chapter 13. Uses less locks, a more concise order of actions, actually parses the Termination Ports, handles Dbgk notification. Timers are now rundown, and Mutex rundown is in a dedicated Kernel function. Final termination handled by KeTerminate Thread as documented. * Renamed PsTerminateOtherThread to PspTerminateThreadByPointer and modified implementation to be compatible with the changes above. * Renamed and regrouped Process Termination into PspExitProcess. Also implemented as described above, and moved each subsystem specific termination helper into its own subsytem. * Improved NtTerminateProcess and added more debugging messages. * Improved NtTerminateThread and added check against System Thread and made it compatible with new implementation. * Corrected PsTerminateSystemThread now that we support System Threads. * Corrected NtRegisterThreadTerminatePort to use same structure name as on windows for the port, and added tag to pool allocation (documented in pooltag.txt) include/internal/*.h: * Defined Scheduler Functions and misc new functions or renamed functions. ke/apc.c: * Fixed critical bug where APCs were not delivered at all if the thread wastion and cancels any timers that are associated to a thread, as well as their APCs and DPCs. REGRESSIONS FOUND: NONE BUGS/REGRESSIOSN FIXED: * Thread/Get Set Context now works. * Suspend Regression test now works. * Task manager can now kill foreign processes, even hung ones (like it should). * ExitProcess/closing cmd.exe with the 'x' button now works correctly without hacks. KNOWN ISSUES: I left a bit of a mess in the headers and some stuff still needs to be moved into the right places. I just wanted to have this first part ready first, so that it won't get too big. svn path=/trunk/; revision=14174
2005-03-18 05:53:04 +00:00
/* Debugging Level */
ULONG PspTraceLevel = 0;
/* PRIVATE FUNCTIONS *********************************************************/
NTSTATUS
NTAPI
PsReferenceProcessFilePointer(IN PEPROCESS Process,
OUT PFILE_OBJECT *FileObject)
{
PSECTION Section;
PAGED_CODE();
/* Lock the process */
if (!ExAcquireRundownProtection(&Process->RundownProtect))
{
return STATUS_PROCESS_IS_TERMINATING;
}
/* Get the section */
Section = Process->SectionObject;
if (Section)
{
/* Get the file object and reference it */
*FileObject = MmGetFileObjectForSection((PVOID)Section);
ObReferenceObject(*FileObject);
}
/* Release the protection */
ExReleaseRundownProtection(&Process->RundownProtect);
/* Return status */
return Section ? STATUS_SUCCESS : STATUS_UNSUCCESSFUL;
}
/* PUBLIC FUNCTIONS **********************************************************/
Thread/Process Termination/Repeaing Rewrite + Fixes --------------------------------------------------- - ps/cid.c: * Moved CID Lookup functions here - ps/security.c: * Moved all security related functions here. Makes other files neater and security functions easier to locate. - ps/thread.c: * Moved most of the Thread Scheduling/Dispatching code that belongs in the Kernel to /ke and renamed functions from Ps to Ki. * Implemented PsIsSystemThread. * Removed Reaper Thread Init (now obsolete). * Renamed PiDeleteThread to PspDeleteThread. * Moved Thread State functions from tinfo.c to here. - ps/process.c: * Removed Query/Set Process functions and moved to ps/query.c * Renamed PiDeletePRocess to PspDeleteProcess * Removed obsoleted Process Termination functions, moved persistent one to kill.c - ps/create.c: * Moved the security APIs to security.c * Correctly implemented PsCreateSystemThread to actually create system threads. - ps/suspend.c * Rewrote Nt Executive functions to use Kernel functions. * Moved Ps* Routines into ke/kthread.c and fixed them. The implementation was wrong in some aspects, especially the issue of the APC looping around the KeWaitXxx call and the fact that the routines excluded/ignored the FreezeCount. - ps/debug.c * Fixed completely broken implementation of Get/SetThreadContext. The old version crashed when called and did not work at all. Suspend Regression test now works. * Moved Context<->TrapFrame functions to ke/i386/ * Combined Set/GetThreadContext APCs into a single one, and used special context structure. - ps/query.c: * Moved Thread/Process Query/Set Routines here. - ps/tinfo.c: * Removed. - ps/kill.c * Removed complicated Process Termination semantics and useless Attach/Detach in favor for a much more lightweight function which performs the same tasks as before and actually works. TaskManager can now terminate foreign processes. * Rewrote Thread Reaping to use the HyperCritical Work Queue instead of manually controlled thread. This results in much less code as well as an increase in speed and less micro management. The reaper is PspReapRoutine. Closing CMD.EXE now works properly without requiring masks that were added as hacks to allow it. * Renamed PiTerminateProcessThreads to PspTerminateProcessThreads. Fixed it to work with new termination code. * Added PspDeleteProcess to handle Process Object deletion. Kills the CID Handle here as done by Hartmut. * Added PspDeletethread here. * Renamed and rewrote PsTerminateCurrentThread to PspExitThread. Used NT Implementation out- lined in Windows Internals, Chapter 13. Uses less locks, a more concise order of actions, actually parses the Termination Ports, handles Dbgk notification. Timers are now rundown, and Mutex rundown is in a dedicated Kernel function. Final termination handled by KeTerminate Thread as documented. * Renamed PsTerminateOtherThread to PspTerminateThreadByPointer and modified implementation to be compatible with the changes above. * Renamed and regrouped Process Termination into PspExitProcess. Also implemented as described above, and moved each subsystem specific termination helper into its own subsytem. * Improved NtTerminateProcess and added more debugging messages. * Improved NtTerminateThread and added check against System Thread and made it compatible with new implementation. * Corrected PsTerminateSystemThread now that we support System Threads. * Corrected NtRegisterThreadTerminatePort to use same structure name as on windows for the port, and added tag to pool allocation (documented in pooltag.txt) include/internal/*.h: * Defined Scheduler Functions and misc new functions or renamed functions. ke/apc.c: * Fixed critical bug where APCs were not delivered at all if the thread wastion and cancels any timers that are associated to a thread, as well as their APCs and DPCs. REGRESSIONS FOUND: NONE BUGS/REGRESSIOSN FIXED: * Thread/Get Set Context now works. * Suspend Regression test now works. * Task manager can now kill foreign processes, even hung ones (like it should). * ExitProcess/closing cmd.exe with the 'x' button now works correctly without hacks. KNOWN ISSUES: I left a bit of a mess in the headers and some stuff still needs to be moved into the right places. I just wanted to have this first part ready first, so that it won't get too big. svn path=/trunk/; revision=14174
2005-03-18 05:53:04 +00:00
/*
* @implemented
Thread/Process Termination/Repeaing Rewrite + Fixes --------------------------------------------------- - ps/cid.c: * Moved CID Lookup functions here - ps/security.c: * Moved all security related functions here. Makes other files neater and security functions easier to locate. - ps/thread.c: * Moved most of the Thread Scheduling/Dispatching code that belongs in the Kernel to /ke and renamed functions from Ps to Ki. * Implemented PsIsSystemThread. * Removed Reaper Thread Init (now obsolete). * Renamed PiDeleteThread to PspDeleteThread. * Moved Thread State functions from tinfo.c to here. - ps/process.c: * Removed Query/Set Process functions and moved to ps/query.c * Renamed PiDeletePRocess to PspDeleteProcess * Removed obsoleted Process Termination functions, moved persistent one to kill.c - ps/create.c: * Moved the security APIs to security.c * Correctly implemented PsCreateSystemThread to actually create system threads. - ps/suspend.c * Rewrote Nt Executive functions to use Kernel functions. * Moved Ps* Routines into ke/kthread.c and fixed them. The implementation was wrong in some aspects, especially the issue of the APC looping around the KeWaitXxx call and the fact that the routines excluded/ignored the FreezeCount. - ps/debug.c * Fixed completely broken implementation of Get/SetThreadContext. The old version crashed when called and did not work at all. Suspend Regression test now works. * Moved Context<->TrapFrame functions to ke/i386/ * Combined Set/GetThreadContext APCs into a single one, and used special context structure. - ps/query.c: * Moved Thread/Process Query/Set Routines here. - ps/tinfo.c: * Removed. - ps/kill.c * Removed complicated Process Termination semantics and useless Attach/Detach in favor for a much more lightweight function which performs the same tasks as before and actually works. TaskManager can now terminate foreign processes. * Rewrote Thread Reaping to use the HyperCritical Work Queue instead of manually controlled thread. This results in much less code as well as an increase in speed and less micro management. The reaper is PspReapRoutine. Closing CMD.EXE now works properly without requiring masks that were added as hacks to allow it. * Renamed PiTerminateProcessThreads to PspTerminateProcessThreads. Fixed it to work with new termination code. * Added PspDeleteProcess to handle Process Object deletion. Kills the CID Handle here as done by Hartmut. * Added PspDeletethread here. * Renamed and rewrote PsTerminateCurrentThread to PspExitThread. Used NT Implementation out- lined in Windows Internals, Chapter 13. Uses less locks, a more concise order of actions, actually parses the Termination Ports, handles Dbgk notification. Timers are now rundown, and Mutex rundown is in a dedicated Kernel function. Final termination handled by KeTerminate Thread as documented. * Renamed PsTerminateOtherThread to PspTerminateThreadByPointer and modified implementation to be compatible with the changes above. * Renamed and regrouped Process Termination into PspExitProcess. Also implemented as described above, and moved each subsystem specific termination helper into its own subsytem. * Improved NtTerminateProcess and added more debugging messages. * Improved NtTerminateThread and added check against System Thread and made it compatible with new implementation. * Corrected PsTerminateSystemThread now that we support System Threads. * Corrected NtRegisterThreadTerminatePort to use same structure name as on windows for the port, and added tag to pool allocation (documented in pooltag.txt) include/internal/*.h: * Defined Scheduler Functions and misc new functions or renamed functions. ke/apc.c: * Fixed critical bug where APCs were not delivered at all if the thread wastion and cancels any timers that are associated to a thread, as well as their APCs and DPCs. REGRESSIONS FOUND: NONE BUGS/REGRESSIOSN FIXED: * Thread/Get Set Context now works. * Suspend Regression test now works. * Task manager can now kill foreign processes, even hung ones (like it should). * ExitProcess/closing cmd.exe with the 'x' button now works correctly without hacks. KNOWN ISSUES: I left a bit of a mess in the headers and some stuff still needs to be moved into the right places. I just wanted to have this first part ready first, so that it won't get too big. svn path=/trunk/; revision=14174
2005-03-18 05:53:04 +00:00
*/
NTSTATUS
NTAPI
NtQueryInformationProcess(IN HANDLE ProcessHandle,
IN PROCESSINFOCLASS ProcessInformationClass,
OUT PVOID ProcessInformation,
IN ULONG ProcessInformationLength,
OUT PULONG ReturnLength OPTIONAL)
Thread/Process Termination/Repeaing Rewrite + Fixes --------------------------------------------------- - ps/cid.c: * Moved CID Lookup functions here - ps/security.c: * Moved all security related functions here. Makes other files neater and security functions easier to locate. - ps/thread.c: * Moved most of the Thread Scheduling/Dispatching code that belongs in the Kernel to /ke and renamed functions from Ps to Ki. * Implemented PsIsSystemThread. * Removed Reaper Thread Init (now obsolete). * Renamed PiDeleteThread to PspDeleteThread. * Moved Thread State functions from tinfo.c to here. - ps/process.c: * Removed Query/Set Process functions and moved to ps/query.c * Renamed PiDeletePRocess to PspDeleteProcess * Removed obsoleted Process Termination functions, moved persistent one to kill.c - ps/create.c: * Moved the security APIs to security.c * Correctly implemented PsCreateSystemThread to actually create system threads. - ps/suspend.c * Rewrote Nt Executive functions to use Kernel functions. * Moved Ps* Routines into ke/kthread.c and fixed them. The implementation was wrong in some aspects, especially the issue of the APC looping around the KeWaitXxx call and the fact that the routines excluded/ignored the FreezeCount. - ps/debug.c * Fixed completely broken implementation of Get/SetThreadContext. The old version crashed when called and did not work at all. Suspend Regression test now works. * Moved Context<->TrapFrame functions to ke/i386/ * Combined Set/GetThreadContext APCs into a single one, and used special context structure. - ps/query.c: * Moved Thread/Process Query/Set Routines here. - ps/tinfo.c: * Removed. - ps/kill.c * Removed complicated Process Termination semantics and useless Attach/Detach in favor for a much more lightweight function which performs the same tasks as before and actually works. TaskManager can now terminate foreign processes. * Rewrote Thread Reaping to use the HyperCritical Work Queue instead of manually controlled thread. This results in much less code as well as an increase in speed and less micro management. The reaper is PspReapRoutine. Closing CMD.EXE now works properly without requiring masks that were added as hacks to allow it. * Renamed PiTerminateProcessThreads to PspTerminateProcessThreads. Fixed it to work with new termination code. * Added PspDeleteProcess to handle Process Object deletion. Kills the CID Handle here as done by Hartmut. * Added PspDeletethread here. * Renamed and rewrote PsTerminateCurrentThread to PspExitThread. Used NT Implementation out- lined in Windows Internals, Chapter 13. Uses less locks, a more concise order of actions, actually parses the Termination Ports, handles Dbgk notification. Timers are now rundown, and Mutex rundown is in a dedicated Kernel function. Final termination handled by KeTerminate Thread as documented. * Renamed PsTerminateOtherThread to PspTerminateThreadByPointer and modified implementation to be compatible with the changes above. * Renamed and regrouped Process Termination into PspExitProcess. Also implemented as described above, and moved each subsystem specific termination helper into its own subsytem. * Improved NtTerminateProcess and added more debugging messages. * Improved NtTerminateThread and added check against System Thread and made it compatible with new implementation. * Corrected PsTerminateSystemThread now that we support System Threads. * Corrected NtRegisterThreadTerminatePort to use same structure name as on windows for the port, and added tag to pool allocation (documented in pooltag.txt) include/internal/*.h: * Defined Scheduler Functions and misc new functions or renamed functions. ke/apc.c: * Fixed critical bug where APCs were not delivered at all if the thread wastion and cancels any timers that are associated to a thread, as well as their APCs and DPCs. REGRESSIONS FOUND: NONE BUGS/REGRESSIOSN FIXED: * Thread/Get Set Context now works. * Suspend Regression test now works. * Task manager can now kill foreign processes, even hung ones (like it should). * ExitProcess/closing cmd.exe with the 'x' button now works correctly without hacks. KNOWN ISSUES: I left a bit of a mess in the headers and some stuff still needs to be moved into the right places. I just wanted to have this first part ready first, so that it won't get too big. svn path=/trunk/; revision=14174
2005-03-18 05:53:04 +00:00
{
PEPROCESS Process;
KPROCESSOR_MODE PreviousMode = ExGetPreviousMode();
NTSTATUS Status;
ULONG Length = 0;
HANDLE DebugPort = 0;
PPROCESS_BASIC_INFORMATION ProcessBasicInfo =
(PPROCESS_BASIC_INFORMATION)ProcessInformation;
PKERNEL_USER_TIMES ProcessTime = (PKERNEL_USER_TIMES)ProcessInformation;
ULONG UserTime, KernelTime;
PPROCESS_PRIORITY_CLASS PsPriorityClass = (PPROCESS_PRIORITY_CLASS)ProcessInformation;
ULONG HandleCount;
PPROCESS_SESSION_INFORMATION SessionInfo =
(PPROCESS_SESSION_INFORMATION)ProcessInformation;
PVM_COUNTERS VmCounters = (PVM_COUNTERS)ProcessInformation;
PIO_COUNTERS IoCounters = (PIO_COUNTERS)ProcessInformation;
PQUOTA_LIMITS QuotaLimits = (PQUOTA_LIMITS)ProcessInformation;
PUNICODE_STRING ImageName;
ULONG Cookie, ExecuteOptions = 0;
ULONG_PTR Wow64 = 0;
PROCESS_VALUES ProcessValues;
ULONG Flags;
PAGED_CODE();
/* Check for user-mode caller */
if (PreviousMode != KernelMode)
{
/* Prepare to probe parameters */
_SEH2_TRY
{
/* Probe the buffer */
ProbeForRead(ProcessInformation,
ProcessInformationLength,
sizeof(ULONG));
/* Probe the return length if required */
if (ReturnLength) ProbeForWriteUlong(ReturnLength);
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Return the exception code */
_SEH2_YIELD(return _SEH2_GetExceptionCode());
}
_SEH2_END;
}
if (((ProcessInformationClass == ProcessCookie) ||
(ProcessInformationClass == ProcessImageInformation)) &&
(ProcessHandle != NtCurrentProcess()))
{
/*
* Retrieving the process cookie is only allowed for the calling process
* itself! XP only allows NtCurrentProcess() as process handles even if
* a real handle actually represents the current process.
*/
return STATUS_INVALID_PARAMETER;
}
Thread/Process Termination/Repeaing Rewrite + Fixes --------------------------------------------------- - ps/cid.c: * Moved CID Lookup functions here - ps/security.c: * Moved all security related functions here. Makes other files neater and security functions easier to locate. - ps/thread.c: * Moved most of the Thread Scheduling/Dispatching code that belongs in the Kernel to /ke and renamed functions from Ps to Ki. * Implemented PsIsSystemThread. * Removed Reaper Thread Init (now obsolete). * Renamed PiDeleteThread to PspDeleteThread. * Moved Thread State functions from tinfo.c to here. - ps/process.c: * Removed Query/Set Process functions and moved to ps/query.c * Renamed PiDeletePRocess to PspDeleteProcess * Removed obsoleted Process Termination functions, moved persistent one to kill.c - ps/create.c: * Moved the security APIs to security.c * Correctly implemented PsCreateSystemThread to actually create system threads. - ps/suspend.c * Rewrote Nt Executive functions to use Kernel functions. * Moved Ps* Routines into ke/kthread.c and fixed them. The implementation was wrong in some aspects, especially the issue of the APC looping around the KeWaitXxx call and the fact that the routines excluded/ignored the FreezeCount. - ps/debug.c * Fixed completely broken implementation of Get/SetThreadContext. The old version crashed when called and did not work at all. Suspend Regression test now works. * Moved Context<->TrapFrame functions to ke/i386/ * Combined Set/GetThreadContext APCs into a single one, and used special context structure. - ps/query.c: * Moved Thread/Process Query/Set Routines here. - ps/tinfo.c: * Removed. - ps/kill.c * Removed complicated Process Termination semantics and useless Attach/Detach in favor for a much more lightweight function which performs the same tasks as before and actually works. TaskManager can now terminate foreign processes. * Rewrote Thread Reaping to use the HyperCritical Work Queue instead of manually controlled thread. This results in much less code as well as an increase in speed and less micro management. The reaper is PspReapRoutine. Closing CMD.EXE now works properly without requiring masks that were added as hacks to allow it. * Renamed PiTerminateProcessThreads to PspTerminateProcessThreads. Fixed it to work with new termination code. * Added PspDeleteProcess to handle Process Object deletion. Kills the CID Handle here as done by Hartmut. * Added PspDeletethread here. * Renamed and rewrote PsTerminateCurrentThread to PspExitThread. Used NT Implementation out- lined in Windows Internals, Chapter 13. Uses less locks, a more concise order of actions, actually parses the Termination Ports, handles Dbgk notification. Timers are now rundown, and Mutex rundown is in a dedicated Kernel function. Final termination handled by KeTerminate Thread as documented. * Renamed PsTerminateOtherThread to PspTerminateThreadByPointer and modified implementation to be compatible with the changes above. * Renamed and regrouped Process Termination into PspExitProcess. Also implemented as described above, and moved each subsystem specific termination helper into its own subsytem. * Improved NtTerminateProcess and added more debugging messages. * Improved NtTerminateThread and added check against System Thread and made it compatible with new implementation. * Corrected PsTerminateSystemThread now that we support System Threads. * Corrected NtRegisterThreadTerminatePort to use same structure name as on windows for the port, and added tag to pool allocation (documented in pooltag.txt) include/internal/*.h: * Defined Scheduler Functions and misc new functions or renamed functions. ke/apc.c: * Fixed critical bug where APCs were not delivered at all if the thread wastion and cancels any timers that are associated to a thread, as well as their APCs and DPCs. REGRESSIONS FOUND: NONE BUGS/REGRESSIOSN FIXED: * Thread/Get Set Context now works. * Suspend Regression test now works. * Task manager can now kill foreign processes, even hung ones (like it should). * ExitProcess/closing cmd.exe with the 'x' button now works correctly without hacks. KNOWN ISSUES: I left a bit of a mess in the headers and some stuff still needs to be moved into the right places. I just wanted to have this first part ready first, so that it won't get too big. svn path=/trunk/; revision=14174
2005-03-18 05:53:04 +00:00
/* Check the information class */
switch (ProcessInformationClass)
{
/* Basic process information */
case ProcessBasicInformation:
if (ProcessInformationLength != sizeof(PROCESS_BASIC_INFORMATION))
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
/* Set return length */
Length = sizeof(PROCESS_BASIC_INFORMATION);
/* Reference the process */
Status = ObReferenceObjectByHandle(ProcessHandle,
PROCESS_QUERY_INFORMATION,
PsProcessType,
PreviousMode,
(PVOID*)&Process,
NULL);
if (!NT_SUCCESS(Status)) break;
/* Protect writes with SEH */
_SEH2_TRY
{
/* Write all the information from the EPROCESS/KPROCESS */
ProcessBasicInfo->ExitStatus = Process->ExitStatus;
ProcessBasicInfo->PebBaseAddress = Process->Peb;
ProcessBasicInfo->AffinityMask = Process->Pcb.Affinity;
ProcessBasicInfo->UniqueProcessId = (ULONG_PTR)Process->
UniqueProcessId;
ProcessBasicInfo->InheritedFromUniqueProcessId =
(ULONG_PTR)Process->InheritedFromUniqueProcessId;
ProcessBasicInfo->BasePriority = Process->Pcb.BasePriority;
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get exception code */
Status = _SEH2_GetExceptionCode();
}
_SEH2_END;
/* Dereference the process */
ObDereferenceObject(Process);
break;
/* Process quota limits */
case ProcessQuotaLimits:
if (ProcessInformationLength != sizeof(QUOTA_LIMITS))
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
Length = sizeof(QUOTA_LIMITS);
/* Reference the process */
Status = ObReferenceObjectByHandle(ProcessHandle,
PROCESS_QUERY_INFORMATION,
PsProcessType,
PreviousMode,
(PVOID*)&Process,
NULL);
if (!NT_SUCCESS(Status)) break;
/* Indicate success */
Status = STATUS_SUCCESS;
_SEH2_TRY
{
/* Set max/min working set sizes */
QuotaLimits->MaximumWorkingSetSize =
Process->Vm.MaximumWorkingSetSize << PAGE_SHIFT;
QuotaLimits->MinimumWorkingSetSize =
Process->Vm.MinimumWorkingSetSize << PAGE_SHIFT;
/* Set default time limits */
QuotaLimits->TimeLimit.LowPart = MAXULONG;
QuotaLimits->TimeLimit.HighPart = MAXULONG;
/* Is quota block a default one? */
if (Process->QuotaBlock == &PspDefaultQuotaBlock)
{
/* Set default pools and pagefile limits */
QuotaLimits->PagedPoolLimit = (SIZE_T)-1;
QuotaLimits->NonPagedPoolLimit = (SIZE_T)-1;
QuotaLimits->PagefileLimit = (SIZE_T)-1;
}
else
{
/* Get limits from non-default quota block */
QuotaLimits->PagedPoolLimit =
Process->QuotaBlock->QuotaEntry[PagedPool].Limit;
QuotaLimits->NonPagedPoolLimit =
Process->QuotaBlock->QuotaEntry[NonPagedPool].Limit;
QuotaLimits->PagefileLimit =
Process->QuotaBlock->QuotaEntry[2].Limit;
}
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get exception code */
Status = _SEH2_GetExceptionCode();
}
_SEH2_END;
/* Dereference the process */
ObDereferenceObject(Process);
break;
case ProcessIoCounters:
if (ProcessInformationLength != sizeof(IO_COUNTERS))
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
Length = sizeof(IO_COUNTERS);
/* Reference the process */
Status = ObReferenceObjectByHandle(ProcessHandle,
PROCESS_QUERY_INFORMATION,
PsProcessType,
PreviousMode,
(PVOID*)&Process,
NULL);
if (!NT_SUCCESS(Status)) break;
/* Query IO counters from the process */
KeQueryValuesProcess(&Process->Pcb, &ProcessValues);
_SEH2_TRY
{
RtlCopyMemory(IoCounters, &ProcessValues.IoInfo, sizeof(IO_COUNTERS));
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Ignore exception */
}
_SEH2_END;
/* Set status to success in any case */
Status = STATUS_SUCCESS;
/* Dereference the process */
ObDereferenceObject(Process);
break;
Thread/Process Termination/Repeaing Rewrite + Fixes --------------------------------------------------- - ps/cid.c: * Moved CID Lookup functions here - ps/security.c: * Moved all security related functions here. Makes other files neater and security functions easier to locate. - ps/thread.c: * Moved most of the Thread Scheduling/Dispatching code that belongs in the Kernel to /ke and renamed functions from Ps to Ki. * Implemented PsIsSystemThread. * Removed Reaper Thread Init (now obsolete). * Renamed PiDeleteThread to PspDeleteThread. * Moved Thread State functions from tinfo.c to here. - ps/process.c: * Removed Query/Set Process functions and moved to ps/query.c * Renamed PiDeletePRocess to PspDeleteProcess * Removed obsoleted Process Termination functions, moved persistent one to kill.c - ps/create.c: * Moved the security APIs to security.c * Correctly implemented PsCreateSystemThread to actually create system threads. - ps/suspend.c * Rewrote Nt Executive functions to use Kernel functions. * Moved Ps* Routines into ke/kthread.c and fixed them. The implementation was wrong in some aspects, especially the issue of the APC looping around the KeWaitXxx call and the fact that the routines excluded/ignored the FreezeCount. - ps/debug.c * Fixed completely broken implementation of Get/SetThreadContext. The old version crashed when called and did not work at all. Suspend Regression test now works. * Moved Context<->TrapFrame functions to ke/i386/ * Combined Set/GetThreadContext APCs into a single one, and used special context structure. - ps/query.c: * Moved Thread/Process Query/Set Routines here. - ps/tinfo.c: * Removed. - ps/kill.c * Removed complicated Process Termination semantics and useless Attach/Detach in favor for a much more lightweight function which performs the same tasks as before and actually works. TaskManager can now terminate foreign processes. * Rewrote Thread Reaping to use the HyperCritical Work Queue instead of manually controlled thread. This results in much less code as well as an increase in speed and less micro management. The reaper is PspReapRoutine. Closing CMD.EXE now works properly without requiring masks that were added as hacks to allow it. * Renamed PiTerminateProcessThreads to PspTerminateProcessThreads. Fixed it to work with new termination code. * Added PspDeleteProcess to handle Process Object deletion. Kills the CID Handle here as done by Hartmut. * Added PspDeletethread here. * Renamed and rewrote PsTerminateCurrentThread to PspExitThread. Used NT Implementation out- lined in Windows Internals, Chapter 13. Uses less locks, a more concise order of actions, actually parses the Termination Ports, handles Dbgk notification. Timers are now rundown, and Mutex rundown is in a dedicated Kernel function. Final termination handled by KeTerminate Thread as documented. * Renamed PsTerminateOtherThread to PspTerminateThreadByPointer and modified implementation to be compatible with the changes above. * Renamed and regrouped Process Termination into PspExitProcess. Also implemented as described above, and moved each subsystem specific termination helper into its own subsytem. * Improved NtTerminateProcess and added more debugging messages. * Improved NtTerminateThread and added check against System Thread and made it compatible with new implementation. * Corrected PsTerminateSystemThread now that we support System Threads. * Corrected NtRegisterThreadTerminatePort to use same structure name as on windows for the port, and added tag to pool allocation (documented in pooltag.txt) include/internal/*.h: * Defined Scheduler Functions and misc new functions or renamed functions. ke/apc.c: * Fixed critical bug where APCs were not delivered at all if the thread wastion and cancels any timers that are associated to a thread, as well as their APCs and DPCs. REGRESSIONS FOUND: NONE BUGS/REGRESSIOSN FIXED: * Thread/Get Set Context now works. * Suspend Regression test now works. * Task manager can now kill foreign processes, even hung ones (like it should). * ExitProcess/closing cmd.exe with the 'x' button now works correctly without hacks. KNOWN ISSUES: I left a bit of a mess in the headers and some stuff still needs to be moved into the right places. I just wanted to have this first part ready first, so that it won't get too big. svn path=/trunk/; revision=14174
2005-03-18 05:53:04 +00:00
/* Timing */
case ProcessTimes:
/* Set the return length */
if (ProcessInformationLength != sizeof(KERNEL_USER_TIMES))
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
Length = sizeof(KERNEL_USER_TIMES);
/* Reference the process */
Status = ObReferenceObjectByHandle(ProcessHandle,
PROCESS_QUERY_INFORMATION,
PsProcessType,
PreviousMode,
(PVOID*)&Process,
NULL);
if (!NT_SUCCESS(Status)) break;
/* Protect writes with SEH */
_SEH2_TRY
{
/* Copy time information from EPROCESS/KPROCESS */
KernelTime = KeQueryRuntimeProcess(&Process->Pcb, &UserTime);
ProcessTime->CreateTime = Process->CreateTime;
ProcessTime->UserTime.QuadPart = (LONGLONG)UserTime * KeMaximumIncrement;
ProcessTime->KernelTime.QuadPart = (LONGLONG)KernelTime * KeMaximumIncrement;
ProcessTime->ExitTime = Process->ExitTime;
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get exception code */
Status = _SEH2_GetExceptionCode();
}
_SEH2_END;
/* Dereference the process */
ObDereferenceObject(Process);
break;
/* Process Debug Port */
case ProcessDebugPort:
if (ProcessInformationLength != sizeof(HANDLE))
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
/* Set return length */
Length = sizeof(HANDLE);
/* Reference the process */
Status = ObReferenceObjectByHandle(ProcessHandle,
PROCESS_QUERY_INFORMATION,
PsProcessType,
PreviousMode,
(PVOID*)&Process,
NULL);
if (!NT_SUCCESS(Status)) break;
/* Protect write with SEH */
_SEH2_TRY
{
/* Return whether or not we have a debug port */
*(PHANDLE)ProcessInformation = (Process->DebugPort ?
(HANDLE)-1 : NULL);
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get exception code */
Status = _SEH2_GetExceptionCode();
}
_SEH2_END;
/* Dereference the process */
ObDereferenceObject(Process);
break;
case ProcessHandleCount:
if (ProcessInformationLength != sizeof(ULONG))
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
/* Set the return length*/
Length = sizeof(ULONG);
/* Reference the process */
Status = ObReferenceObjectByHandle(ProcessHandle,
PROCESS_QUERY_INFORMATION,
PsProcessType,
PreviousMode,
(PVOID*)&Process,
NULL);
if (!NT_SUCCESS(Status)) break;
/* Count the number of handles this process has */
HandleCount = ObGetProcessHandleCount(Process);
/* Protect write in SEH */
_SEH2_TRY
{
/* Return the count of handles */
*(PULONG)ProcessInformation = HandleCount;
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get the exception code */
Status = _SEH2_GetExceptionCode();
}
_SEH2_END;
/* Dereference the process */
ObDereferenceObject(Process);
break;
Thread/Process Termination/Repeaing Rewrite + Fixes --------------------------------------------------- - ps/cid.c: * Moved CID Lookup functions here - ps/security.c: * Moved all security related functions here. Makes other files neater and security functions easier to locate. - ps/thread.c: * Moved most of the Thread Scheduling/Dispatching code that belongs in the Kernel to /ke and renamed functions from Ps to Ki. * Implemented PsIsSystemThread. * Removed Reaper Thread Init (now obsolete). * Renamed PiDeleteThread to PspDeleteThread. * Moved Thread State functions from tinfo.c to here. - ps/process.c: * Removed Query/Set Process functions and moved to ps/query.c * Renamed PiDeletePRocess to PspDeleteProcess * Removed obsoleted Process Termination functions, moved persistent one to kill.c - ps/create.c: * Moved the security APIs to security.c * Correctly implemented PsCreateSystemThread to actually create system threads. - ps/suspend.c * Rewrote Nt Executive functions to use Kernel functions. * Moved Ps* Routines into ke/kthread.c and fixed them. The implementation was wrong in some aspects, especially the issue of the APC looping around the KeWaitXxx call and the fact that the routines excluded/ignored the FreezeCount. - ps/debug.c * Fixed completely broken implementation of Get/SetThreadContext. The old version crashed when called and did not work at all. Suspend Regression test now works. * Moved Context<->TrapFrame functions to ke/i386/ * Combined Set/GetThreadContext APCs into a single one, and used special context structure. - ps/query.c: * Moved Thread/Process Query/Set Routines here. - ps/tinfo.c: * Removed. - ps/kill.c * Removed complicated Process Termination semantics and useless Attach/Detach in favor for a much more lightweight function which performs the same tasks as before and actually works. TaskManager can now terminate foreign processes. * Rewrote Thread Reaping to use the HyperCritical Work Queue instead of manually controlled thread. This results in much less code as well as an increase in speed and less micro management. The reaper is PspReapRoutine. Closing CMD.EXE now works properly without requiring masks that were added as hacks to allow it. * Renamed PiTerminateProcessThreads to PspTerminateProcessThreads. Fixed it to work with new termination code. * Added PspDeleteProcess to handle Process Object deletion. Kills the CID Handle here as done by Hartmut. * Added PspDeletethread here. * Renamed and rewrote PsTerminateCurrentThread to PspExitThread. Used NT Implementation out- lined in Windows Internals, Chapter 13. Uses less locks, a more concise order of actions, actually parses the Termination Ports, handles Dbgk notification. Timers are now rundown, and Mutex rundown is in a dedicated Kernel function. Final termination handled by KeTerminate Thread as documented. * Renamed PsTerminateOtherThread to PspTerminateThreadByPointer and modified implementation to be compatible with the changes above. * Renamed and regrouped Process Termination into PspExitProcess. Also implemented as described above, and moved each subsystem specific termination helper into its own subsytem. * Improved NtTerminateProcess and added more debugging messages. * Improved NtTerminateThread and added check against System Thread and made it compatible with new implementation. * Corrected PsTerminateSystemThread now that we support System Threads. * Corrected NtRegisterThreadTerminatePort to use same structure name as on windows for the port, and added tag to pool allocation (documented in pooltag.txt) include/internal/*.h: * Defined Scheduler Functions and misc new functions or renamed functions. ke/apc.c: * Fixed critical bug where APCs were not delivered at all if the thread wastion and cancels any timers that are associated to a thread, as well as their APCs and DPCs. REGRESSIONS FOUND: NONE BUGS/REGRESSIOSN FIXED: * Thread/Get Set Context now works. * Suspend Regression test now works. * Task manager can now kill foreign processes, even hung ones (like it should). * ExitProcess/closing cmd.exe with the 'x' button now works correctly without hacks. KNOWN ISSUES: I left a bit of a mess in the headers and some stuff still needs to be moved into the right places. I just wanted to have this first part ready first, so that it won't get too big. svn path=/trunk/; revision=14174
2005-03-18 05:53:04 +00:00
/* Session ID for the process */
case ProcessSessionInformation:
if (ProcessInformationLength != sizeof(PROCESS_SESSION_INFORMATION))
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
/* Set the return length*/
Length = sizeof(PROCESS_SESSION_INFORMATION);
/* Reference the process */
Status = ObReferenceObjectByHandle(ProcessHandle,
PROCESS_QUERY_INFORMATION,
PsProcessType,
PreviousMode,
(PVOID*)&Process,
NULL);
if (!NT_SUCCESS(Status)) break;
/* Enter SEH for write safety */
_SEH2_TRY
{
/* Write back the Session ID */
SessionInfo->SessionId = PsGetProcessSessionId(Process);
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get the exception code */
Status = _SEH2_GetExceptionCode();
}
_SEH2_END;
/* Dereference the process */
ObDereferenceObject(Process);
break;
/* Virtual Memory Statistics */
case ProcessVmCounters:
/* Validate the input length */
if ((ProcessInformationLength != sizeof(VM_COUNTERS)) &&
(ProcessInformationLength != sizeof(VM_COUNTERS_EX)))
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
/* Reference the process */
Status = ObReferenceObjectByHandle(ProcessHandle,
PROCESS_QUERY_INFORMATION,
PsProcessType,
PreviousMode,
(PVOID*)&Process,
NULL);
if (!NT_SUCCESS(Status)) break;
/* Enter SEH for write safety */
_SEH2_TRY
{
/* Return data from EPROCESS */
VmCounters->PeakVirtualSize = Process->PeakVirtualSize;
VmCounters->VirtualSize = Process->VirtualSize;
VmCounters->PageFaultCount = Process->Vm.PageFaultCount;
VmCounters->PeakWorkingSetSize = Process->Vm.PeakWorkingSetSize;
VmCounters->WorkingSetSize = Process->Vm.WorkingSetSize;
VmCounters->QuotaPeakPagedPoolUsage = Process->QuotaPeak[0];
VmCounters->QuotaPagedPoolUsage = Process->QuotaUsage[0];
VmCounters->QuotaPeakNonPagedPoolUsage = Process->QuotaPeak[1];
VmCounters->QuotaNonPagedPoolUsage = Process->QuotaUsage[1];
VmCounters->PagefileUsage = Process->QuotaUsage[2] << PAGE_SHIFT;
VmCounters->PeakPagefileUsage = Process->QuotaPeak[2] << PAGE_SHIFT;
//VmCounters->PrivateUsage = Process->CommitCharge << PAGE_SHIFT;
//
/* Set the return length */
Length = ProcessInformationLength;
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get the exception code */
Status = _SEH2_GetExceptionCode();
}
_SEH2_END;
/* Dereference the process */
ObDereferenceObject(Process);
break;
Thread/Process Termination/Repeaing Rewrite + Fixes --------------------------------------------------- - ps/cid.c: * Moved CID Lookup functions here - ps/security.c: * Moved all security related functions here. Makes other files neater and security functions easier to locate. - ps/thread.c: * Moved most of the Thread Scheduling/Dispatching code that belongs in the Kernel to /ke and renamed functions from Ps to Ki. * Implemented PsIsSystemThread. * Removed Reaper Thread Init (now obsolete). * Renamed PiDeleteThread to PspDeleteThread. * Moved Thread State functions from tinfo.c to here. - ps/process.c: * Removed Query/Set Process functions and moved to ps/query.c * Renamed PiDeletePRocess to PspDeleteProcess * Removed obsoleted Process Termination functions, moved persistent one to kill.c - ps/create.c: * Moved the security APIs to security.c * Correctly implemented PsCreateSystemThread to actually create system threads. - ps/suspend.c * Rewrote Nt Executive functions to use Kernel functions. * Moved Ps* Routines into ke/kthread.c and fixed them. The implementation was wrong in some aspects, especially the issue of the APC looping around the KeWaitXxx call and the fact that the routines excluded/ignored the FreezeCount. - ps/debug.c * Fixed completely broken implementation of Get/SetThreadContext. The old version crashed when called and did not work at all. Suspend Regression test now works. * Moved Context<->TrapFrame functions to ke/i386/ * Combined Set/GetThreadContext APCs into a single one, and used special context structure. - ps/query.c: * Moved Thread/Process Query/Set Routines here. - ps/tinfo.c: * Removed. - ps/kill.c * Removed complicated Process Termination semantics and useless Attach/Detach in favor for a much more lightweight function which performs the same tasks as before and actually works. TaskManager can now terminate foreign processes. * Rewrote Thread Reaping to use the HyperCritical Work Queue instead of manually controlled thread. This results in much less code as well as an increase in speed and less micro management. The reaper is PspReapRoutine. Closing CMD.EXE now works properly without requiring masks that were added as hacks to allow it. * Renamed PiTerminateProcessThreads to PspTerminateProcessThreads. Fixed it to work with new termination code. * Added PspDeleteProcess to handle Process Object deletion. Kills the CID Handle here as done by Hartmut. * Added PspDeletethread here. * Renamed and rewrote PsTerminateCurrentThread to PspExitThread. Used NT Implementation out- lined in Windows Internals, Chapter 13. Uses less locks, a more concise order of actions, actually parses the Termination Ports, handles Dbgk notification. Timers are now rundown, and Mutex rundown is in a dedicated Kernel function. Final termination handled by KeTerminate Thread as documented. * Renamed PsTerminateOtherThread to PspTerminateThreadByPointer and modified implementation to be compatible with the changes above. * Renamed and regrouped Process Termination into PspExitProcess. Also implemented as described above, and moved each subsystem specific termination helper into its own subsytem. * Improved NtTerminateProcess and added more debugging messages. * Improved NtTerminateThread and added check against System Thread and made it compatible with new implementation. * Corrected PsTerminateSystemThread now that we support System Threads. * Corrected NtRegisterThreadTerminatePort to use same structure name as on windows for the port, and added tag to pool allocation (documented in pooltag.txt) include/internal/*.h: * Defined Scheduler Functions and misc new functions or renamed functions. ke/apc.c: * Fixed critical bug where APCs were not delivered at all if the thread wastion and cancels any timers that are associated to a thread, as well as their APCs and DPCs. REGRESSIONS FOUND: NONE BUGS/REGRESSIOSN FIXED: * Thread/Get Set Context now works. * Suspend Regression test now works. * Task manager can now kill foreign processes, even hung ones (like it should). * ExitProcess/closing cmd.exe with the 'x' button now works correctly without hacks. KNOWN ISSUES: I left a bit of a mess in the headers and some stuff still needs to be moved into the right places. I just wanted to have this first part ready first, so that it won't get too big. svn path=/trunk/; revision=14174
2005-03-18 05:53:04 +00:00
/* Hard Error Processing Mode */
case ProcessDefaultHardErrorMode:
if (ProcessInformationLength != sizeof(ULONG))
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
/* Set the return length*/
Length = sizeof(ULONG);
/* Reference the process */
Status = ObReferenceObjectByHandle(ProcessHandle,
PROCESS_QUERY_INFORMATION,
PsProcessType,
PreviousMode,
(PVOID*)&Process,
NULL);
if (!NT_SUCCESS(Status)) break;
/* Enter SEH for writing back data */
_SEH2_TRY
{
/* Write the current processing mode */
*(PULONG)ProcessInformation = Process->
DefaultHardErrorProcessing;
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get the exception code */
Status = _SEH2_GetExceptionCode();
}
_SEH2_END;
/* Dereference the process */
ObDereferenceObject(Process);
break;
/* Priority Boosting status */
case ProcessPriorityBoost:
if (ProcessInformationLength != sizeof(ULONG))
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
/* Set the return length */
Length = sizeof(ULONG);
/* Reference the process */
Status = ObReferenceObjectByHandle(ProcessHandle,
PROCESS_QUERY_INFORMATION,
PsProcessType,
PreviousMode,
(PVOID*)&Process,
NULL);
if (!NT_SUCCESS(Status)) break;
/* Enter SEH for writing back data */
_SEH2_TRY
{
/* Return boost status */
*(PULONG)ProcessInformation = Process->Pcb.DisableBoost ?
TRUE : FALSE;
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
Thread/Process Termination/Repeaing Rewrite + Fixes --------------------------------------------------- - ps/cid.c: * Moved CID Lookup functions here - ps/security.c: * Moved all security related functions here. Makes other files neater and security functions easier to locate. - ps/thread.c: * Moved most of the Thread Scheduling/Dispatching code that belongs in the Kernel to /ke and renamed functions from Ps to Ki. * Implemented PsIsSystemThread. * Removed Reaper Thread Init (now obsolete). * Renamed PiDeleteThread to PspDeleteThread. * Moved Thread State functions from tinfo.c to here. - ps/process.c: * Removed Query/Set Process functions and moved to ps/query.c * Renamed PiDeletePRocess to PspDeleteProcess * Removed obsoleted Process Termination functions, moved persistent one to kill.c - ps/create.c: * Moved the security APIs to security.c * Correctly implemented PsCreateSystemThread to actually create system threads. - ps/suspend.c * Rewrote Nt Executive functions to use Kernel functions. * Moved Ps* Routines into ke/kthread.c and fixed them. The implementation was wrong in some aspects, especially the issue of the APC looping around the KeWaitXxx call and the fact that the routines excluded/ignored the FreezeCount. - ps/debug.c * Fixed completely broken implementation of Get/SetThreadContext. The old version crashed when called and did not work at all. Suspend Regression test now works. * Moved Context<->TrapFrame functions to ke/i386/ * Combined Set/GetThreadContext APCs into a single one, and used special context structure. - ps/query.c: * Moved Thread/Process Query/Set Routines here. - ps/tinfo.c: * Removed. - ps/kill.c * Removed complicated Process Termination semantics and useless Attach/Detach in favor for a much more lightweight function which performs the same tasks as before and actually works. TaskManager can now terminate foreign processes. * Rewrote Thread Reaping to use the HyperCritical Work Queue instead of manually controlled thread. This results in much less code as well as an increase in speed and less micro management. The reaper is PspReapRoutine. Closing CMD.EXE now works properly without requiring masks that were added as hacks to allow it. * Renamed PiTerminateProcessThreads to PspTerminateProcessThreads. Fixed it to work with new termination code. * Added PspDeleteProcess to handle Process Object deletion. Kills the CID Handle here as done by Hartmut. * Added PspDeletethread here. * Renamed and rewrote PsTerminateCurrentThread to PspExitThread. Used NT Implementation out- lined in Windows Internals, Chapter 13. Uses less locks, a more concise order of actions, actually parses the Termination Ports, handles Dbgk notification. Timers are now rundown, and Mutex rundown is in a dedicated Kernel function. Final termination handled by KeTerminate Thread as documented. * Renamed PsTerminateOtherThread to PspTerminateThreadByPointer and modified implementation to be compatible with the changes above. * Renamed and regrouped Process Termination into PspExitProcess. Also implemented as described above, and moved each subsystem specific termination helper into its own subsytem. * Improved NtTerminateProcess and added more debugging messages. * Improved NtTerminateThread and added check against System Thread and made it compatible with new implementation. * Corrected PsTerminateSystemThread now that we support System Threads. * Corrected NtRegisterThreadTerminatePort to use same structure name as on windows for the port, and added tag to pool allocation (documented in pooltag.txt) include/internal/*.h: * Defined Scheduler Functions and misc new functions or renamed functions. ke/apc.c: * Fixed critical bug where APCs were not delivered at all if the thread wastion and cancels any timers that are associated to a thread, as well as their APCs and DPCs. REGRESSIONS FOUND: NONE BUGS/REGRESSIOSN FIXED: * Thread/Get Set Context now works. * Suspend Regression test now works. * Task manager can now kill foreign processes, even hung ones (like it should). * ExitProcess/closing cmd.exe with the 'x' button now works correctly without hacks. KNOWN ISSUES: I left a bit of a mess in the headers and some stuff still needs to be moved into the right places. I just wanted to have this first part ready first, so that it won't get too big. svn path=/trunk/; revision=14174
2005-03-18 05:53:04 +00:00
{
/* Get the exception code */
Status = _SEH2_GetExceptionCode();
Thread/Process Termination/Repeaing Rewrite + Fixes --------------------------------------------------- - ps/cid.c: * Moved CID Lookup functions here - ps/security.c: * Moved all security related functions here. Makes other files neater and security functions easier to locate. - ps/thread.c: * Moved most of the Thread Scheduling/Dispatching code that belongs in the Kernel to /ke and renamed functions from Ps to Ki. * Implemented PsIsSystemThread. * Removed Reaper Thread Init (now obsolete). * Renamed PiDeleteThread to PspDeleteThread. * Moved Thread State functions from tinfo.c to here. - ps/process.c: * Removed Query/Set Process functions and moved to ps/query.c * Renamed PiDeletePRocess to PspDeleteProcess * Removed obsoleted Process Termination functions, moved persistent one to kill.c - ps/create.c: * Moved the security APIs to security.c * Correctly implemented PsCreateSystemThread to actually create system threads. - ps/suspend.c * Rewrote Nt Executive functions to use Kernel functions. * Moved Ps* Routines into ke/kthread.c and fixed them. The implementation was wrong in some aspects, especially the issue of the APC looping around the KeWaitXxx call and the fact that the routines excluded/ignored the FreezeCount. - ps/debug.c * Fixed completely broken implementation of Get/SetThreadContext. The old version crashed when called and did not work at all. Suspend Regression test now works. * Moved Context<->TrapFrame functions to ke/i386/ * Combined Set/GetThreadContext APCs into a single one, and used special context structure. - ps/query.c: * Moved Thread/Process Query/Set Routines here. - ps/tinfo.c: * Removed. - ps/kill.c * Removed complicated Process Termination semantics and useless Attach/Detach in favor for a much more lightweight function which performs the same tasks as before and actually works. TaskManager can now terminate foreign processes. * Rewrote Thread Reaping to use the HyperCritical Work Queue instead of manually controlled thread. This results in much less code as well as an increase in speed and less micro management. The reaper is PspReapRoutine. Closing CMD.EXE now works properly without requiring masks that were added as hacks to allow it. * Renamed PiTerminateProcessThreads to PspTerminateProcessThreads. Fixed it to work with new termination code. * Added PspDeleteProcess to handle Process Object deletion. Kills the CID Handle here as done by Hartmut. * Added PspDeletethread here. * Renamed and rewrote PsTerminateCurrentThread to PspExitThread. Used NT Implementation out- lined in Windows Internals, Chapter 13. Uses less locks, a more concise order of actions, actually parses the Termination Ports, handles Dbgk notification. Timers are now rundown, and Mutex rundown is in a dedicated Kernel function. Final termination handled by KeTerminate Thread as documented. * Renamed PsTerminateOtherThread to PspTerminateThreadByPointer and modified implementation to be compatible with the changes above. * Renamed and regrouped Process Termination into PspExitProcess. Also implemented as described above, and moved each subsystem specific termination helper into its own subsytem. * Improved NtTerminateProcess and added more debugging messages. * Improved NtTerminateThread and added check against System Thread and made it compatible with new implementation. * Corrected PsTerminateSystemThread now that we support System Threads. * Corrected NtRegisterThreadTerminatePort to use same structure name as on windows for the port, and added tag to pool allocation (documented in pooltag.txt) include/internal/*.h: * Defined Scheduler Functions and misc new functions or renamed functions. ke/apc.c: * Fixed critical bug where APCs were not delivered at all if the thread wastion and cancels any timers that are associated to a thread, as well as their APCs and DPCs. REGRESSIONS FOUND: NONE BUGS/REGRESSIOSN FIXED: * Thread/Get Set Context now works. * Suspend Regression test now works. * Task manager can now kill foreign processes, even hung ones (like it should). * ExitProcess/closing cmd.exe with the 'x' button now works correctly without hacks. KNOWN ISSUES: I left a bit of a mess in the headers and some stuff still needs to be moved into the right places. I just wanted to have this first part ready first, so that it won't get too big. svn path=/trunk/; revision=14174
2005-03-18 05:53:04 +00:00
}
_SEH2_END;
/* Dereference the process */
ObDereferenceObject(Process);
break;
/* DOS Device Map */
case ProcessDeviceMap:
if (ProcessInformationLength < sizeof(PROCESS_DEVICEMAP_INFORMATION))
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
if (ProcessInformationLength == sizeof(PROCESS_DEVICEMAP_INFORMATION_EX))
{
/* Protect read in SEH */
_SEH2_TRY
{
PPROCESS_DEVICEMAP_INFORMATION_EX DeviceMapEx = ProcessInformation;
Flags = DeviceMapEx->Flags;
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get the exception code */
Status = _SEH2_GetExceptionCode();
}
_SEH2_END;
if (!NT_SUCCESS(Status))
{
break;
}
/* Only one flag is supported and it needs LUID mappings */
if ((Flags & ~PROCESS_LUID_DOSDEVICES_ONLY) != 0 ||
!ObIsLUIDDeviceMapsEnabled())
{
Status = STATUS_INVALID_PARAMETER;
break;
}
}
else
{
if (ProcessInformationLength != sizeof(PROCESS_DEVICEMAP_INFORMATION))
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
/* No flags for standard call */
Flags = 0;
}
/* Set the return length */
Length = ProcessInformationLength;
/* Reference the process */
Status = ObReferenceObjectByHandle(ProcessHandle,
PROCESS_QUERY_INFORMATION,
PsProcessType,
PreviousMode,
(PVOID*)&Process,
NULL);
if (!NT_SUCCESS(Status)) break;
/* Query the device map information */
Status = ObQueryDeviceMapInformation(Process,
ProcessInformation,
Flags);
/* Dereference the process */
ObDereferenceObject(Process);
break;
Thread/Process Termination/Repeaing Rewrite + Fixes --------------------------------------------------- - ps/cid.c: * Moved CID Lookup functions here - ps/security.c: * Moved all security related functions here. Makes other files neater and security functions easier to locate. - ps/thread.c: * Moved most of the Thread Scheduling/Dispatching code that belongs in the Kernel to /ke and renamed functions from Ps to Ki. * Implemented PsIsSystemThread. * Removed Reaper Thread Init (now obsolete). * Renamed PiDeleteThread to PspDeleteThread. * Moved Thread State functions from tinfo.c to here. - ps/process.c: * Removed Query/Set Process functions and moved to ps/query.c * Renamed PiDeletePRocess to PspDeleteProcess * Removed obsoleted Process Termination functions, moved persistent one to kill.c - ps/create.c: * Moved the security APIs to security.c * Correctly implemented PsCreateSystemThread to actually create system threads. - ps/suspend.c * Rewrote Nt Executive functions to use Kernel functions. * Moved Ps* Routines into ke/kthread.c and fixed them. The implementation was wrong in some aspects, especially the issue of the APC looping around the KeWaitXxx call and the fact that the routines excluded/ignored the FreezeCount. - ps/debug.c * Fixed completely broken implementation of Get/SetThreadContext. The old version crashed when called and did not work at all. Suspend Regression test now works. * Moved Context<->TrapFrame functions to ke/i386/ * Combined Set/GetThreadContext APCs into a single one, and used special context structure. - ps/query.c: * Moved Thread/Process Query/Set Routines here. - ps/tinfo.c: * Removed. - ps/kill.c * Removed complicated Process Termination semantics and useless Attach/Detach in favor for a much more lightweight function which performs the same tasks as before and actually works. TaskManager can now terminate foreign processes. * Rewrote Thread Reaping to use the HyperCritical Work Queue instead of manually controlled thread. This results in much less code as well as an increase in speed and less micro management. The reaper is PspReapRoutine. Closing CMD.EXE now works properly without requiring masks that were added as hacks to allow it. * Renamed PiTerminateProcessThreads to PspTerminateProcessThreads. Fixed it to work with new termination code. * Added PspDeleteProcess to handle Process Object deletion. Kills the CID Handle here as done by Hartmut. * Added PspDeletethread here. * Renamed and rewrote PsTerminateCurrentThread to PspExitThread. Used NT Implementation out- lined in Windows Internals, Chapter 13. Uses less locks, a more concise order of actions, actually parses the Termination Ports, handles Dbgk notification. Timers are now rundown, and Mutex rundown is in a dedicated Kernel function. Final termination handled by KeTerminate Thread as documented. * Renamed PsTerminateOtherThread to PspTerminateThreadByPointer and modified implementation to be compatible with the changes above. * Renamed and regrouped Process Termination into PspExitProcess. Also implemented as described above, and moved each subsystem specific termination helper into its own subsytem. * Improved NtTerminateProcess and added more debugging messages. * Improved NtTerminateThread and added check against System Thread and made it compatible with new implementation. * Corrected PsTerminateSystemThread now that we support System Threads. * Corrected NtRegisterThreadTerminatePort to use same structure name as on windows for the port, and added tag to pool allocation (documented in pooltag.txt) include/internal/*.h: * Defined Scheduler Functions and misc new functions or renamed functions. ke/apc.c: * Fixed critical bug where APCs were not delivered at all if the thread wastion and cancels any timers that are associated to a thread, as well as their APCs and DPCs. REGRESSIONS FOUND: NONE BUGS/REGRESSIOSN FIXED: * Thread/Get Set Context now works. * Suspend Regression test now works. * Task manager can now kill foreign processes, even hung ones (like it should). * ExitProcess/closing cmd.exe with the 'x' button now works correctly without hacks. KNOWN ISSUES: I left a bit of a mess in the headers and some stuff still needs to be moved into the right places. I just wanted to have this first part ready first, so that it won't get too big. svn path=/trunk/; revision=14174
2005-03-18 05:53:04 +00:00
/* Priority class */
case ProcessPriorityClass:
if (ProcessInformationLength != sizeof(PROCESS_PRIORITY_CLASS))
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
/* Set the return length*/
Length = sizeof(PROCESS_PRIORITY_CLASS);
/* Reference the process */
Status = ObReferenceObjectByHandle(ProcessHandle,
PROCESS_QUERY_INFORMATION,
PsProcessType,
PreviousMode,
(PVOID*)&Process,
NULL);
if (!NT_SUCCESS(Status)) break;
/* Enter SEH for writing back data */
_SEH2_TRY
{
/* Return current priority class */
PsPriorityClass->PriorityClass = Process->PriorityClass;
PsPriorityClass->Foreground = FALSE;
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get the exception code */
Status = _SEH2_GetExceptionCode();
}
_SEH2_END;
/* Dereference the process */
ObDereferenceObject(Process);
break;
Thread/Process Termination/Repeaing Rewrite + Fixes --------------------------------------------------- - ps/cid.c: * Moved CID Lookup functions here - ps/security.c: * Moved all security related functions here. Makes other files neater and security functions easier to locate. - ps/thread.c: * Moved most of the Thread Scheduling/Dispatching code that belongs in the Kernel to /ke and renamed functions from Ps to Ki. * Implemented PsIsSystemThread. * Removed Reaper Thread Init (now obsolete). * Renamed PiDeleteThread to PspDeleteThread. * Moved Thread State functions from tinfo.c to here. - ps/process.c: * Removed Query/Set Process functions and moved to ps/query.c * Renamed PiDeletePRocess to PspDeleteProcess * Removed obsoleted Process Termination functions, moved persistent one to kill.c - ps/create.c: * Moved the security APIs to security.c * Correctly implemented PsCreateSystemThread to actually create system threads. - ps/suspend.c * Rewrote Nt Executive functions to use Kernel functions. * Moved Ps* Routines into ke/kthread.c and fixed them. The implementation was wrong in some aspects, especially the issue of the APC looping around the KeWaitXxx call and the fact that the routines excluded/ignored the FreezeCount. - ps/debug.c * Fixed completely broken implementation of Get/SetThreadContext. The old version crashed when called and did not work at all. Suspend Regression test now works. * Moved Context<->TrapFrame functions to ke/i386/ * Combined Set/GetThreadContext APCs into a single one, and used special context structure. - ps/query.c: * Moved Thread/Process Query/Set Routines here. - ps/tinfo.c: * Removed. - ps/kill.c * Removed complicated Process Termination semantics and useless Attach/Detach in favor for a much more lightweight function which performs the same tasks as before and actually works. TaskManager can now terminate foreign processes. * Rewrote Thread Reaping to use the HyperCritical Work Queue instead of manually controlled thread. This results in much less code as well as an increase in speed and less micro management. The reaper is PspReapRoutine. Closing CMD.EXE now works properly without requiring masks that were added as hacks to allow it. * Renamed PiTerminateProcessThreads to PspTerminateProcessThreads. Fixed it to work with new termination code. * Added PspDeleteProcess to handle Process Object deletion. Kills the CID Handle here as done by Hartmut. * Added PspDeletethread here. * Renamed and rewrote PsTerminateCurrentThread to PspExitThread. Used NT Implementation out- lined in Windows Internals, Chapter 13. Uses less locks, a more concise order of actions, actually parses the Termination Ports, handles Dbgk notification. Timers are now rundown, and Mutex rundown is in a dedicated Kernel function. Final termination handled by KeTerminate Thread as documented. * Renamed PsTerminateOtherThread to PspTerminateThreadByPointer and modified implementation to be compatible with the changes above. * Renamed and regrouped Process Termination into PspExitProcess. Also implemented as described above, and moved each subsystem specific termination helper into its own subsytem. * Improved NtTerminateProcess and added more debugging messages. * Improved NtTerminateThread and added check against System Thread and made it compatible with new implementation. * Corrected PsTerminateSystemThread now that we support System Threads. * Corrected NtRegisterThreadTerminatePort to use same structure name as on windows for the port, and added tag to pool allocation (documented in pooltag.txt) include/internal/*.h: * Defined Scheduler Functions and misc new functions or renamed functions. ke/apc.c: * Fixed critical bug where APCs were not delivered at all if the thread wastion and cancels any timers that are associated to a thread, as well as their APCs and DPCs. REGRESSIONS FOUND: NONE BUGS/REGRESSIOSN FIXED: * Thread/Get Set Context now works. * Suspend Regression test now works. * Task manager can now kill foreign processes, even hung ones (like it should). * ExitProcess/closing cmd.exe with the 'x' button now works correctly without hacks. KNOWN ISSUES: I left a bit of a mess in the headers and some stuff still needs to be moved into the right places. I just wanted to have this first part ready first, so that it won't get too big. svn path=/trunk/; revision=14174
2005-03-18 05:53:04 +00:00
case ProcessImageFileName:
/* Reference the process */
Status = ObReferenceObjectByHandle(ProcessHandle,
PROCESS_QUERY_INFORMATION,
PsProcessType,
PreviousMode,
(PVOID*)&Process,
NULL);
if (!NT_SUCCESS(Status)) break;
/* Get the image path */
Status = SeLocateProcessImageName(Process, &ImageName);
if (NT_SUCCESS(Status))
{
/* Set return length */
Length = ImageName->MaximumLength +
sizeof(OBJECT_NAME_INFORMATION);
/* Make sure it's large enough */
if (Length <= ProcessInformationLength)
{
/* Enter SEH to protect write */
_SEH2_TRY
{
/* Copy it */
RtlCopyMemory(ProcessInformation,
ImageName,
Length);
/* Update pointer */
((PUNICODE_STRING)ProcessInformation)->Buffer =
(PWSTR)((PUNICODE_STRING)ProcessInformation + 1);
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get the exception code */
Status = _SEH2_GetExceptionCode();
}
_SEH2_END;
}
else
{
/* Buffer too small */
Status = STATUS_INFO_LENGTH_MISMATCH;
}
/* Free the image path */
ExFreePoolWithTag(ImageName, TAG_SEPA);
}
/* Dereference the process */
ObDereferenceObject(Process);
break;
case ProcessDebugFlags:
if (ProcessInformationLength != sizeof(ULONG))
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
/* Set the return length*/
Length = sizeof(ULONG);
/* Reference the process */
Status = ObReferenceObjectByHandle(ProcessHandle,
PROCESS_QUERY_INFORMATION,
PsProcessType,
PreviousMode,
(PVOID*)&Process,
NULL);
if (!NT_SUCCESS(Status)) break;
/* Enter SEH for writing back data */
_SEH2_TRY
{
/* Return the debug flag state */
*(PULONG)ProcessInformation = Process->NoDebugInherit ? 0 : 1;
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get the exception code */
Status = _SEH2_GetExceptionCode();
}
_SEH2_END;
/* Dereference the process */
ObDereferenceObject(Process);
break;
case ProcessBreakOnTermination:
if (ProcessInformationLength != sizeof(ULONG))
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
/* Set the return length */
Length = sizeof(ULONG);
/* Reference the process */
Status = ObReferenceObjectByHandle(ProcessHandle,
PROCESS_QUERY_INFORMATION,
PsProcessType,
PreviousMode,
(PVOID*)&Process,
NULL);
if (!NT_SUCCESS(Status)) break;
/* Enter SEH for writing back data */
_SEH2_TRY
{
/* Return the BreakOnTermination state */
*(PULONG)ProcessInformation = Process->BreakOnTermination;
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get the exception code */
Status = _SEH2_GetExceptionCode();
}
_SEH2_END;
/* Dereference the process */
ObDereferenceObject(Process);
break;
/* Per-process security cookie */
case ProcessCookie:
/* Get the current process and cookie */
Process = PsGetCurrentProcess();
Cookie = Process->Cookie;
if (!Cookie)
{
LARGE_INTEGER SystemTime;
ULONG NewCookie;
PKPRCB Prcb;
/* Generate a new cookie */
KeQuerySystemTime(&SystemTime);
Prcb = KeGetCurrentPrcb();
NewCookie = Prcb->KeSystemCalls ^ Prcb->InterruptTime ^
SystemTime.u.LowPart ^ SystemTime.u.HighPart;
/* Set the new cookie or return the current one */
Cookie = InterlockedCompareExchange((LONG*)&Process->Cookie,
NewCookie,
Cookie);
if (!Cookie) Cookie = NewCookie;
/* Set return length */
Length = sizeof(ULONG);
}
Thread/Process Termination/Repeaing Rewrite + Fixes --------------------------------------------------- - ps/cid.c: * Moved CID Lookup functions here - ps/security.c: * Moved all security related functions here. Makes other files neater and security functions easier to locate. - ps/thread.c: * Moved most of the Thread Scheduling/Dispatching code that belongs in the Kernel to /ke and renamed functions from Ps to Ki. * Implemented PsIsSystemThread. * Removed Reaper Thread Init (now obsolete). * Renamed PiDeleteThread to PspDeleteThread. * Moved Thread State functions from tinfo.c to here. - ps/process.c: * Removed Query/Set Process functions and moved to ps/query.c * Renamed PiDeletePRocess to PspDeleteProcess * Removed obsoleted Process Termination functions, moved persistent one to kill.c - ps/create.c: * Moved the security APIs to security.c * Correctly implemented PsCreateSystemThread to actually create system threads. - ps/suspend.c * Rewrote Nt Executive functions to use Kernel functions. * Moved Ps* Routines into ke/kthread.c and fixed them. The implementation was wrong in some aspects, especially the issue of the APC looping around the KeWaitXxx call and the fact that the routines excluded/ignored the FreezeCount. - ps/debug.c * Fixed completely broken implementation of Get/SetThreadContext. The old version crashed when called and did not work at all. Suspend Regression test now works. * Moved Context<->TrapFrame functions to ke/i386/ * Combined Set/GetThreadContext APCs into a single one, and used special context structure. - ps/query.c: * Moved Thread/Process Query/Set Routines here. - ps/tinfo.c: * Removed. - ps/kill.c * Removed complicated Process Termination semantics and useless Attach/Detach in favor for a much more lightweight function which performs the same tasks as before and actually works. TaskManager can now terminate foreign processes. * Rewrote Thread Reaping to use the HyperCritical Work Queue instead of manually controlled thread. This results in much less code as well as an increase in speed and less micro management. The reaper is PspReapRoutine. Closing CMD.EXE now works properly without requiring masks that were added as hacks to allow it. * Renamed PiTerminateProcessThreads to PspTerminateProcessThreads. Fixed it to work with new termination code. * Added PspDeleteProcess to handle Process Object deletion. Kills the CID Handle here as done by Hartmut. * Added PspDeletethread here. * Renamed and rewrote PsTerminateCurrentThread to PspExitThread. Used NT Implementation out- lined in Windows Internals, Chapter 13. Uses less locks, a more concise order of actions, actually parses the Termination Ports, handles Dbgk notification. Timers are now rundown, and Mutex rundown is in a dedicated Kernel function. Final termination handled by KeTerminate Thread as documented. * Renamed PsTerminateOtherThread to PspTerminateThreadByPointer and modified implementation to be compatible with the changes above. * Renamed and regrouped Process Termination into PspExitProcess. Also implemented as described above, and moved each subsystem specific termination helper into its own subsytem. * Improved NtTerminateProcess and added more debugging messages. * Improved NtTerminateThread and added check against System Thread and made it compatible with new implementation. * Corrected PsTerminateSystemThread now that we support System Threads. * Corrected NtRegisterThreadTerminatePort to use same structure name as on windows for the port, and added tag to pool allocation (documented in pooltag.txt) include/internal/*.h: * Defined Scheduler Functions and misc new functions or renamed functions. ke/apc.c: * Fixed critical bug where APCs were not delivered at all if the thread wastion and cancels any timers that are associated to a thread, as well as their APCs and DPCs. REGRESSIONS FOUND: NONE BUGS/REGRESSIOSN FIXED: * Thread/Get Set Context now works. * Suspend Regression test now works. * Task manager can now kill foreign processes, even hung ones (like it should). * ExitProcess/closing cmd.exe with the 'x' button now works correctly without hacks. KNOWN ISSUES: I left a bit of a mess in the headers and some stuff still needs to be moved into the right places. I just wanted to have this first part ready first, so that it won't get too big. svn path=/trunk/; revision=14174
2005-03-18 05:53:04 +00:00
/* Indicate success */
Status = STATUS_SUCCESS;
/* Enter SEH to protect write */
_SEH2_TRY
{
/* Write back the cookie */
*(PULONG)ProcessInformation = Cookie;
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get the exception code */
Status = _SEH2_GetExceptionCode();
}
_SEH2_END;
break;
case ProcessImageInformation:
if (ProcessInformationLength != sizeof(SECTION_IMAGE_INFORMATION))
{
/* Break out */
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
/* Set the length required and validate it */
Length = sizeof(SECTION_IMAGE_INFORMATION);
/* Enter SEH to protect write */
_SEH2_TRY
{
MmGetImageInformation((PSECTION_IMAGE_INFORMATION)ProcessInformation);
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get the exception code */
Status = _SEH2_GetExceptionCode();
}
_SEH2_END;
/* Indicate success */
Status = STATUS_SUCCESS;
break;
case ProcessDebugObjectHandle:
if (ProcessInformationLength != sizeof(HANDLE))
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
/* Set the return length */
Length = sizeof(HANDLE);
/* Reference the process */
Status = ObReferenceObjectByHandle(ProcessHandle,
PROCESS_QUERY_INFORMATION,
PsProcessType,
PreviousMode,
(PVOID*)&Process,
NULL);
if (!NT_SUCCESS(Status)) break;
/* Get the debug port */
Status = DbgkOpenProcessDebugPort(Process, PreviousMode, &DebugPort);
/* Let go of the process */
ObDereferenceObject(Process);
/* Protect write in SEH */
_SEH2_TRY
{
/* Return debug port's handle */
*(PHANDLE)ProcessInformation = DebugPort;
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get the exception code */
Status = _SEH2_GetExceptionCode();
}
_SEH2_END;
break;
case ProcessHandleTracing:
DPRINT1("Handle tracing Not implemented: %lx\n", ProcessInformationClass);
Status = STATUS_NOT_IMPLEMENTED;
break;
case ProcessLUIDDeviceMapsEnabled:
if (ProcessInformationLength != sizeof(ULONG))
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
/* Set the return length */
Length = sizeof(ULONG);
/* Indicate success */
Status = STATUS_SUCCESS;
/* Protect write in SEH */
_SEH2_TRY
{
/* Query Ob */
*(PULONG)ProcessInformation = ObIsLUIDDeviceMapsEnabled();
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get the exception code */
Status = _SEH2_GetExceptionCode();
}
_SEH2_END;
break;
case ProcessWx86Information:
if (ProcessInformationLength != sizeof(ULONG))
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
/* Set the return length */
Length = sizeof(ULONG);
/* Reference the process */
Status = ObReferenceObjectByHandle(ProcessHandle,
PROCESS_QUERY_INFORMATION,
PsProcessType,
PreviousMode,
(PVOID*)&Process,
NULL);
if (!NT_SUCCESS(Status)) break;
/* Protect write in SEH */
_SEH2_TRY
{
/* Return if the flag is set */
*(PULONG)ProcessInformation = (ULONG)Process->VdmAllowed;
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get the exception code */
Status = _SEH2_GetExceptionCode();
}
_SEH2_END;
/* Dereference the process */
ObDereferenceObject(Process);
break;
case ProcessWow64Information:
if (ProcessInformationLength != sizeof(ULONG_PTR))
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
/* Set return length */
Length = sizeof(ULONG_PTR);
/* Reference the process */
Status = ObReferenceObjectByHandle(ProcessHandle,
PROCESS_QUERY_INFORMATION,
PsProcessType,
PreviousMode,
(PVOID*)&Process,
NULL);
if (!NT_SUCCESS(Status)) break;
/* Make sure the process isn't dying */
if (ExAcquireRundownProtection(&Process->RundownProtect))
{
/* Get the WOW64 process structure */
#ifdef _WIN64
Wow64 = (ULONG_PTR)Process->Wow64Process;
#else
Wow64 = 0;
#endif
/* Release the lock */
ExReleaseRundownProtection(&Process->RundownProtect);
}
/* Protect write with SEH */
_SEH2_TRY
{
/* Return whether or not we have a debug port */
*(PULONG_PTR)ProcessInformation = Wow64;
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get exception code */
Status = _SEH2_GetExceptionCode();
}
_SEH2_END;
/* Dereference the process */
ObDereferenceObject(Process);
break;
case ProcessExecuteFlags:
if (ProcessInformationLength != sizeof(ULONG))
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
/* Set return length */
Length = sizeof(ULONG);
if (ProcessHandle != NtCurrentProcess())
{
return STATUS_INVALID_PARAMETER;
}
/* Get the options */
Status = MmGetExecuteOptions(&ExecuteOptions);
if (NT_SUCCESS(Status))
{
/* Protect write with SEH */
_SEH2_TRY
{
/* Return them */
*(PULONG)ProcessInformation = ExecuteOptions;
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get exception code */
Status = _SEH2_GetExceptionCode();
}
_SEH2_END;
}
break;
case ProcessLdtInformation:
DPRINT1("VDM/16-bit not implemented: %lx\n", ProcessInformationClass);
Status = STATUS_NOT_IMPLEMENTED;
break;
Thread/Process Termination/Repeaing Rewrite + Fixes --------------------------------------------------- - ps/cid.c: * Moved CID Lookup functions here - ps/security.c: * Moved all security related functions here. Makes other files neater and security functions easier to locate. - ps/thread.c: * Moved most of the Thread Scheduling/Dispatching code that belongs in the Kernel to /ke and renamed functions from Ps to Ki. * Implemented PsIsSystemThread. * Removed Reaper Thread Init (now obsolete). * Renamed PiDeleteThread to PspDeleteThread. * Moved Thread State functions from tinfo.c to here. - ps/process.c: * Removed Query/Set Process functions and moved to ps/query.c * Renamed PiDeletePRocess to PspDeleteProcess * Removed obsoleted Process Termination functions, moved persistent one to kill.c - ps/create.c: * Moved the security APIs to security.c * Correctly implemented PsCreateSystemThread to actually create system threads. - ps/suspend.c * Rewrote Nt Executive functions to use Kernel functions. * Moved Ps* Routines into ke/kthread.c and fixed them. The implementation was wrong in some aspects, especially the issue of the APC looping around the KeWaitXxx call and the fact that the routines excluded/ignored the FreezeCount. - ps/debug.c * Fixed completely broken implementation of Get/SetThreadContext. The old version crashed when called and did not work at all. Suspend Regression test now works. * Moved Context<->TrapFrame functions to ke/i386/ * Combined Set/GetThreadContext APCs into a single one, and used special context structure. - ps/query.c: * Moved Thread/Process Query/Set Routines here. - ps/tinfo.c: * Removed. - ps/kill.c * Removed complicated Process Termination semantics and useless Attach/Detach in favor for a much more lightweight function which performs the same tasks as before and actually works. TaskManager can now terminate foreign processes. * Rewrote Thread Reaping to use the HyperCritical Work Queue instead of manually controlled thread. This results in much less code as well as an increase in speed and less micro management. The reaper is PspReapRoutine. Closing CMD.EXE now works properly without requiring masks that were added as hacks to allow it. * Renamed PiTerminateProcessThreads to PspTerminateProcessThreads. Fixed it to work with new termination code. * Added PspDeleteProcess to handle Process Object deletion. Kills the CID Handle here as done by Hartmut. * Added PspDeletethread here. * Renamed and rewrote PsTerminateCurrentThread to PspExitThread. Used NT Implementation out- lined in Windows Internals, Chapter 13. Uses less locks, a more concise order of actions, actually parses the Termination Ports, handles Dbgk notification. Timers are now rundown, and Mutex rundown is in a dedicated Kernel function. Final termination handled by KeTerminate Thread as documented. * Renamed PsTerminateOtherThread to PspTerminateThreadByPointer and modified implementation to be compatible with the changes above. * Renamed and regrouped Process Termination into PspExitProcess. Also implemented as described above, and moved each subsystem specific termination helper into its own subsytem. * Improved NtTerminateProcess and added more debugging messages. * Improved NtTerminateThread and added check against System Thread and made it compatible with new implementation. * Corrected PsTerminateSystemThread now that we support System Threads. * Corrected NtRegisterThreadTerminatePort to use same structure name as on windows for the port, and added tag to pool allocation (documented in pooltag.txt) include/internal/*.h: * Defined Scheduler Functions and misc new functions or renamed functions. ke/apc.c: * Fixed critical bug where APCs were not delivered at all if the thread wastion and cancels any timers that are associated to a thread, as well as their APCs and DPCs. REGRESSIONS FOUND: NONE BUGS/REGRESSIOSN FIXED: * Thread/Get Set Context now works. * Suspend Regression test now works. * Task manager can now kill foreign processes, even hung ones (like it should). * ExitProcess/closing cmd.exe with the 'x' button now works correctly without hacks. KNOWN ISSUES: I left a bit of a mess in the headers and some stuff still needs to be moved into the right places. I just wanted to have this first part ready first, so that it won't get too big. svn path=/trunk/; revision=14174
2005-03-18 05:53:04 +00:00
case ProcessWorkingSetWatch:
DPRINT1("WS Watch Not implemented: %lx\n", ProcessInformationClass);
Status = STATUS_NOT_IMPLEMENTED;
break;
case ProcessPooledUsageAndLimits:
DPRINT1("Pool limits Not implemented: %lx\n", ProcessInformationClass);
Status = STATUS_NOT_IMPLEMENTED;
break;
/* Not supported by Server 2003 */
default:
DPRINT1("Unsupported info class: %lx\n", ProcessInformationClass);
Status = STATUS_INVALID_INFO_CLASS;
}
/* Protect write with SEH */
_SEH2_TRY
{
/* Check if caller wanted return length */
if ((ReturnLength) && (Length)) *ReturnLength = Length;
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get exception code */
Status = _SEH2_GetExceptionCode();
}
_SEH2_END;
return Status;
}
/*
* @implemented
*/
NTSTATUS
NTAPI
NtSetInformationProcess(IN HANDLE ProcessHandle,
IN PROCESSINFOCLASS ProcessInformationClass,
IN PVOID ProcessInformation,
IN ULONG ProcessInformationLength)
{
PEPROCESS Process;
KPROCESSOR_MODE PreviousMode = ExGetPreviousMode();
ACCESS_MASK Access;
NTSTATUS Status;
HANDLE PortHandle = NULL;
HANDLE TokenHandle = NULL;
HANDLE DirectoryHandle = NULL;
PROCESS_SESSION_INFORMATION SessionInfo = {0};
PROCESS_PRIORITY_CLASS PriorityClass = {0};
PROCESS_FOREGROUND_BACKGROUND Foreground = {0};
PVOID ExceptionPort;
ULONG Break;
KAFFINITY ValidAffinity, Affinity = 0;
KPRIORITY BasePriority = 0;
UCHAR MemoryPriority = 0;
BOOLEAN DisableBoost = 0;
ULONG DefaultHardErrorMode = 0;
ULONG DebugFlags = 0, EnableFixup = 0, Boost = 0;
ULONG NoExecute = 0, VdmPower = 0;
BOOLEAN HasPrivilege;
PLIST_ENTRY Next;
PETHREAD Thread;
PAGED_CODE();
/* Verify Information Class validity */
#if 0
Status = DefaultSetInfoBufferCheck(ProcessInformationClass,
PsProcessInfoClass,
RTL_NUMBER_OF(PsProcessInfoClass),
ProcessInformation,
ProcessInformationLength,
PreviousMode);
if (!NT_SUCCESS(Status)) return Status;
#endif
/* Check what class this is */
Access = PROCESS_SET_INFORMATION;
if (ProcessInformationClass == ProcessSessionInformation)
{
/* Setting the Session ID needs a special mask */
Access |= PROCESS_SET_SESSIONID;
}
else if (ProcessInformationClass == ProcessExceptionPort)
{
/* Setting the exception port needs a special mask */
Access |= PROCESS_SUSPEND_RESUME;
}
/* Reference the process */
Status = ObReferenceObjectByHandle(ProcessHandle,
Access,
PsProcessType,
PreviousMode,
(PVOID*)&Process,
NULL);
if (!NT_SUCCESS(Status)) return Status;
Thread/Process Termination/Repeaing Rewrite + Fixes --------------------------------------------------- - ps/cid.c: * Moved CID Lookup functions here - ps/security.c: * Moved all security related functions here. Makes other files neater and security functions easier to locate. - ps/thread.c: * Moved most of the Thread Scheduling/Dispatching code that belongs in the Kernel to /ke and renamed functions from Ps to Ki. * Implemented PsIsSystemThread. * Removed Reaper Thread Init (now obsolete). * Renamed PiDeleteThread to PspDeleteThread. * Moved Thread State functions from tinfo.c to here. - ps/process.c: * Removed Query/Set Process functions and moved to ps/query.c * Renamed PiDeletePRocess to PspDeleteProcess * Removed obsoleted Process Termination functions, moved persistent one to kill.c - ps/create.c: * Moved the security APIs to security.c * Correctly implemented PsCreateSystemThread to actually create system threads. - ps/suspend.c * Rewrote Nt Executive functions to use Kernel functions. * Moved Ps* Routines into ke/kthread.c and fixed them. The implementation was wrong in some aspects, especially the issue of the APC looping around the KeWaitXxx call and the fact that the routines excluded/ignored the FreezeCount. - ps/debug.c * Fixed completely broken implementation of Get/SetThreadContext. The old version crashed when called and did not work at all. Suspend Regression test now works. * Moved Context<->TrapFrame functions to ke/i386/ * Combined Set/GetThreadContext APCs into a single one, and used special context structure. - ps/query.c: * Moved Thread/Process Query/Set Routines here. - ps/tinfo.c: * Removed. - ps/kill.c * Removed complicated Process Termination semantics and useless Attach/Detach in favor for a much more lightweight function which performs the same tasks as before and actually works. TaskManager can now terminate foreign processes. * Rewrote Thread Reaping to use the HyperCritical Work Queue instead of manually controlled thread. This results in much less code as well as an increase in speed and less micro management. The reaper is PspReapRoutine. Closing CMD.EXE now works properly without requiring masks that were added as hacks to allow it. * Renamed PiTerminateProcessThreads to PspTerminateProcessThreads. Fixed it to work with new termination code. * Added PspDeleteProcess to handle Process Object deletion. Kills the CID Handle here as done by Hartmut. * Added PspDeletethread here. * Renamed and rewrote PsTerminateCurrentThread to PspExitThread. Used NT Implementation out- lined in Windows Internals, Chapter 13. Uses less locks, a more concise order of actions, actually parses the Termination Ports, handles Dbgk notification. Timers are now rundown, and Mutex rundown is in a dedicated Kernel function. Final termination handled by KeTerminate Thread as documented. * Renamed PsTerminateOtherThread to PspTerminateThreadByPointer and modified implementation to be compatible with the changes above. * Renamed and regrouped Process Termination into PspExitProcess. Also implemented as described above, and moved each subsystem specific termination helper into its own subsytem. * Improved NtTerminateProcess and added more debugging messages. * Improved NtTerminateThread and added check against System Thread and made it compatible with new implementation. * Corrected PsTerminateSystemThread now that we support System Threads. * Corrected NtRegisterThreadTerminatePort to use same structure name as on windows for the port, and added tag to pool allocation (documented in pooltag.txt) include/internal/*.h: * Defined Scheduler Functions and misc new functions or renamed functions. ke/apc.c: * Fixed critical bug where APCs were not delivered at all if the thread wastion and cancels any timers that are associated to a thread, as well as their APCs and DPCs. REGRESSIONS FOUND: NONE BUGS/REGRESSIOSN FIXED: * Thread/Get Set Context now works. * Suspend Regression test now works. * Task manager can now kill foreign processes, even hung ones (like it should). * ExitProcess/closing cmd.exe with the 'x' button now works correctly without hacks. KNOWN ISSUES: I left a bit of a mess in the headers and some stuff still needs to be moved into the right places. I just wanted to have this first part ready first, so that it won't get too big. svn path=/trunk/; revision=14174
2005-03-18 05:53:04 +00:00
/* Check what kind of information class this is */
switch (ProcessInformationClass)
{
case ProcessWx86Information:
/* Check buffer length */
if (ProcessInformationLength != sizeof(HANDLE))
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
/* Use SEH for capture */
_SEH2_TRY
{
/* Capture the boolean */
VdmPower = *(PULONG)ProcessInformation;
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get the exception code */
Status = _SEH2_GetExceptionCode();
_SEH2_YIELD(break);
}
_SEH2_END;
/* Getting VDM powers requires the SeTcbPrivilege */
if (!SeSinglePrivilegeCheck(SeTcbPrivilege, PreviousMode))
{
/* We don't hold the privilege, bail out */
Status = STATUS_PRIVILEGE_NOT_HELD;
DPRINT1("Need TCB privilege\n");
break;
}
/* Set or clear the flag */
if (VdmPower)
{
PspSetProcessFlag(Process, PSF_VDM_ALLOWED_BIT);
}
else
{
PspClearProcessFlag(Process, PSF_VDM_ALLOWED_BIT);
}
break;
/* Error/Exception Port */
case ProcessExceptionPort:
/* Check buffer length */
if (ProcessInformationLength != sizeof(HANDLE))
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
/* Use SEH for capture */
_SEH2_TRY
{
/* Capture the handle */
PortHandle = *(PHANDLE)ProcessInformation;
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get the exception code */
Status = _SEH2_GetExceptionCode();
_SEH2_YIELD(break);
}
_SEH2_END;
/* Setting the error port requires the SeTcbPrivilege */
if (!SeSinglePrivilegeCheck(SeTcbPrivilege, PreviousMode))
{
/* We don't hold the privilege, bail out */
Status = STATUS_PRIVILEGE_NOT_HELD;
break;
}
/* Get the LPC Port */
Status = ObReferenceObjectByHandle(PortHandle,
0,
LpcPortObjectType,
PreviousMode,
(PVOID)&ExceptionPort,
NULL);
if (!NT_SUCCESS(Status)) break;
/* Change the pointer */
if (InterlockedCompareExchangePointer(&Process->ExceptionPort,
ExceptionPort,
NULL))
{
/* We already had one, fail */
ObDereferenceObject(ExceptionPort);
Status = STATUS_PORT_ALREADY_SET;
}
break;
/* Security Token */
case ProcessAccessToken:
/* Check buffer length */
if (ProcessInformationLength != sizeof(PROCESS_ACCESS_TOKEN))
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
/* Use SEH for capture */
_SEH2_TRY
{
/* Save the token handle */
TokenHandle = ((PPROCESS_ACCESS_TOKEN)ProcessInformation)->
Token;
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get the exception code */
Status = _SEH2_GetExceptionCode();
_SEH2_YIELD(break);
}
_SEH2_END;
Thread/Process Termination/Repeaing Rewrite + Fixes --------------------------------------------------- - ps/cid.c: * Moved CID Lookup functions here - ps/security.c: * Moved all security related functions here. Makes other files neater and security functions easier to locate. - ps/thread.c: * Moved most of the Thread Scheduling/Dispatching code that belongs in the Kernel to /ke and renamed functions from Ps to Ki. * Implemented PsIsSystemThread. * Removed Reaper Thread Init (now obsolete). * Renamed PiDeleteThread to PspDeleteThread. * Moved Thread State functions from tinfo.c to here. - ps/process.c: * Removed Query/Set Process functions and moved to ps/query.c * Renamed PiDeletePRocess to PspDeleteProcess * Removed obsoleted Process Termination functions, moved persistent one to kill.c - ps/create.c: * Moved the security APIs to security.c * Correctly implemented PsCreateSystemThread to actually create system threads. - ps/suspend.c * Rewrote Nt Executive functions to use Kernel functions. * Moved Ps* Routines into ke/kthread.c and fixed them. The implementation was wrong in some aspects, especially the issue of the APC looping around the KeWaitXxx call and the fact that the routines excluded/ignored the FreezeCount. - ps/debug.c * Fixed completely broken implementation of Get/SetThreadContext. The old version crashed when called and did not work at all. Suspend Regression test now works. * Moved Context<->TrapFrame functions to ke/i386/ * Combined Set/GetThreadContext APCs into a single one, and used special context structure. - ps/query.c: * Moved Thread/Process Query/Set Routines here. - ps/tinfo.c: * Removed. - ps/kill.c * Removed complicated Process Termination semantics and useless Attach/Detach in favor for a much more lightweight function which performs the same tasks as before and actually works. TaskManager can now terminate foreign processes. * Rewrote Thread Reaping to use the HyperCritical Work Queue instead of manually controlled thread. This results in much less code as well as an increase in speed and less micro management. The reaper is PspReapRoutine. Closing CMD.EXE now works properly without requiring masks that were added as hacks to allow it. * Renamed PiTerminateProcessThreads to PspTerminateProcessThreads. Fixed it to work with new termination code. * Added PspDeleteProcess to handle Process Object deletion. Kills the CID Handle here as done by Hartmut. * Added PspDeletethread here. * Renamed and rewrote PsTerminateCurrentThread to PspExitThread. Used NT Implementation out- lined in Windows Internals, Chapter 13. Uses less locks, a more concise order of actions, actually parses the Termination Ports, handles Dbgk notification. Timers are now rundown, and Mutex rundown is in a dedicated Kernel function. Final termination handled by KeTerminate Thread as documented. * Renamed PsTerminateOtherThread to PspTerminateThreadByPointer and modified implementation to be compatible with the changes above. * Renamed and regrouped Process Termination into PspExitProcess. Also implemented as described above, and moved each subsystem specific termination helper into its own subsytem. * Improved NtTerminateProcess and added more debugging messages. * Improved NtTerminateThread and added check against System Thread and made it compatible with new implementation. * Corrected PsTerminateSystemThread now that we support System Threads. * Corrected NtRegisterThreadTerminatePort to use same structure name as on windows for the port, and added tag to pool allocation (documented in pooltag.txt) include/internal/*.h: * Defined Scheduler Functions and misc new functions or renamed functions. ke/apc.c: * Fixed critical bug where APCs were not delivered at all if the thread wastion and cancels any timers that are associated to a thread, as well as their APCs and DPCs. REGRESSIONS FOUND: NONE BUGS/REGRESSIOSN FIXED: * Thread/Get Set Context now works. * Suspend Regression test now works. * Task manager can now kill foreign processes, even hung ones (like it should). * ExitProcess/closing cmd.exe with the 'x' button now works correctly without hacks. KNOWN ISSUES: I left a bit of a mess in the headers and some stuff still needs to be moved into the right places. I just wanted to have this first part ready first, so that it won't get too big. svn path=/trunk/; revision=14174
2005-03-18 05:53:04 +00:00
/* Assign the actual token */
Status = PspSetPrimaryToken(Process, TokenHandle, NULL);
break;
/* Hard error processing */
case ProcessDefaultHardErrorMode:
/* Check buffer length */
if (ProcessInformationLength != sizeof(ULONG))
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
/* Enter SEH for direct buffer read */
_SEH2_TRY
{
DefaultHardErrorMode = *(PULONG)ProcessInformation;
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get exception code */
Status = _SEH2_GetExceptionCode();
_SEH2_YIELD(break);
}
_SEH2_END;
/* Set the mode */
Process->DefaultHardErrorProcessing = DefaultHardErrorMode;
/* Call Ke for the update */
if (DefaultHardErrorMode & SEM_NOALIGNMENTFAULTEXCEPT)
{
KeSetAutoAlignmentProcess(&Process->Pcb, TRUE);
}
else
{
KeSetAutoAlignmentProcess(&Process->Pcb, FALSE);
}
Status = STATUS_SUCCESS;
break;
/* Session ID */
case ProcessSessionInformation:
/* Check buffer length */
if (ProcessInformationLength != sizeof(PROCESS_SESSION_INFORMATION))
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
/* Enter SEH for capture */
_SEH2_TRY
{
/* Capture the caller's buffer */
SessionInfo = *(PPROCESS_SESSION_INFORMATION)ProcessInformation;
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get the exception code */
Status = _SEH2_GetExceptionCode();
_SEH2_YIELD(break);
}
_SEH2_END;
/* Setting the session id requires the SeTcbPrivilege */
if (!SeSinglePrivilegeCheck(SeTcbPrivilege, PreviousMode))
{
/* We don't hold the privilege, bail out */
Status = STATUS_PRIVILEGE_NOT_HELD;
break;
}
Thread/Process Termination/Repeaing Rewrite + Fixes --------------------------------------------------- - ps/cid.c: * Moved CID Lookup functions here - ps/security.c: * Moved all security related functions here. Makes other files neater and security functions easier to locate. - ps/thread.c: * Moved most of the Thread Scheduling/Dispatching code that belongs in the Kernel to /ke and renamed functions from Ps to Ki. * Implemented PsIsSystemThread. * Removed Reaper Thread Init (now obsolete). * Renamed PiDeleteThread to PspDeleteThread. * Moved Thread State functions from tinfo.c to here. - ps/process.c: * Removed Query/Set Process functions and moved to ps/query.c * Renamed PiDeletePRocess to PspDeleteProcess * Removed obsoleted Process Termination functions, moved persistent one to kill.c - ps/create.c: * Moved the security APIs to security.c * Correctly implemented PsCreateSystemThread to actually create system threads. - ps/suspend.c * Rewrote Nt Executive functions to use Kernel functions. * Moved Ps* Routines into ke/kthread.c and fixed them. The implementation was wrong in some aspects, especially the issue of the APC looping around the KeWaitXxx call and the fact that the routines excluded/ignored the FreezeCount. - ps/debug.c * Fixed completely broken implementation of Get/SetThreadContext. The old version crashed when called and did not work at all. Suspend Regression test now works. * Moved Context<->TrapFrame functions to ke/i386/ * Combined Set/GetThreadContext APCs into a single one, and used special context structure. - ps/query.c: * Moved Thread/Process Query/Set Routines here. - ps/tinfo.c: * Removed. - ps/kill.c * Removed complicated Process Termination semantics and useless Attach/Detach in favor for a much more lightweight function which performs the same tasks as before and actually works. TaskManager can now terminate foreign processes. * Rewrote Thread Reaping to use the HyperCritical Work Queue instead of manually controlled thread. This results in much less code as well as an increase in speed and less micro management. The reaper is PspReapRoutine. Closing CMD.EXE now works properly without requiring masks that were added as hacks to allow it. * Renamed PiTerminateProcessThreads to PspTerminateProcessThreads. Fixed it to work with new termination code. * Added PspDeleteProcess to handle Process Object deletion. Kills the CID Handle here as done by Hartmut. * Added PspDeletethread here. * Renamed and rewrote PsTerminateCurrentThread to PspExitThread. Used NT Implementation out- lined in Windows Internals, Chapter 13. Uses less locks, a more concise order of actions, actually parses the Termination Ports, handles Dbgk notification. Timers are now rundown, and Mutex rundown is in a dedicated Kernel function. Final termination handled by KeTerminate Thread as documented. * Renamed PsTerminateOtherThread to PspTerminateThreadByPointer and modified implementation to be compatible with the changes above. * Renamed and regrouped Process Termination into PspExitProcess. Also implemented as described above, and moved each subsystem specific termination helper into its own subsytem. * Improved NtTerminateProcess and added more debugging messages. * Improved NtTerminateThread and added check against System Thread and made it compatible with new implementation. * Corrected PsTerminateSystemThread now that we support System Threads. * Corrected NtRegisterThreadTerminatePort to use same structure name as on windows for the port, and added tag to pool allocation (documented in pooltag.txt) include/internal/*.h: * Defined Scheduler Functions and misc new functions or renamed functions. ke/apc.c: * Fixed critical bug where APCs were not delivered at all if the thread wastion and cancels any timers that are associated to a thread, as well as their APCs and DPCs. REGRESSIONS FOUND: NONE BUGS/REGRESSIOSN FIXED: * Thread/Get Set Context now works. * Suspend Regression test now works. * Task manager can now kill foreign processes, even hung ones (like it should). * ExitProcess/closing cmd.exe with the 'x' button now works correctly without hacks. KNOWN ISSUES: I left a bit of a mess in the headers and some stuff still needs to be moved into the right places. I just wanted to have this first part ready first, so that it won't get too big. svn path=/trunk/; revision=14174
2005-03-18 05:53:04 +00:00
#if 0 // OLD AND DEPRECATED CODE!!!!
/* FIXME - update the session id for the process token */
//Status = PsLockProcess(Process, FALSE);
if (!NT_SUCCESS(Status)) break;
/* Write the session ID in the EPROCESS */
Process->Session = UlongToPtr(SessionInfo.SessionId); // HACK!!!
/* Check if the process also has a PEB */
if (Process->Peb)
{
/*
* Attach to the process to make sure we're in the right
* context to access the PEB structure
*/
KeAttachProcess(&Process->Pcb);
/* Enter SEH for write to user-mode PEB */
_SEH2_TRY
{
/* Write the session ID */
Process->Peb->SessionId = SessionInfo.SessionId;
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get exception code */
Status = _SEH2_GetExceptionCode();
}
_SEH2_END;
/* Detach from the process */
KeDetachProcess();
}
Thread/Process Termination/Repeaing Rewrite + Fixes --------------------------------------------------- - ps/cid.c: * Moved CID Lookup functions here - ps/security.c: * Moved all security related functions here. Makes other files neater and security functions easier to locate. - ps/thread.c: * Moved most of the Thread Scheduling/Dispatching code that belongs in the Kernel to /ke and renamed functions from Ps to Ki. * Implemented PsIsSystemThread. * Removed Reaper Thread Init (now obsolete). * Renamed PiDeleteThread to PspDeleteThread. * Moved Thread State functions from tinfo.c to here. - ps/process.c: * Removed Query/Set Process functions and moved to ps/query.c * Renamed PiDeletePRocess to PspDeleteProcess * Removed obsoleted Process Termination functions, moved persistent one to kill.c - ps/create.c: * Moved the security APIs to security.c * Correctly implemented PsCreateSystemThread to actually create system threads. - ps/suspend.c * Rewrote Nt Executive functions to use Kernel functions. * Moved Ps* Routines into ke/kthread.c and fixed them. The implementation was wrong in some aspects, especially the issue of the APC looping around the KeWaitXxx call and the fact that the routines excluded/ignored the FreezeCount. - ps/debug.c * Fixed completely broken implementation of Get/SetThreadContext. The old version crashed when called and did not work at all. Suspend Regression test now works. * Moved Context<->TrapFrame functions to ke/i386/ * Combined Set/GetThreadContext APCs into a single one, and used special context structure. - ps/query.c: * Moved Thread/Process Query/Set Routines here. - ps/tinfo.c: * Removed. - ps/kill.c * Removed complicated Process Termination semantics and useless Attach/Detach in favor for a much more lightweight function which performs the same tasks as before and actually works. TaskManager can now terminate foreign processes. * Rewrote Thread Reaping to use the HyperCritical Work Queue instead of manually controlled thread. This results in much less code as well as an increase in speed and less micro management. The reaper is PspReapRoutine. Closing CMD.EXE now works properly without requiring masks that were added as hacks to allow it. * Renamed PiTerminateProcessThreads to PspTerminateProcessThreads. Fixed it to work with new termination code. * Added PspDeleteProcess to handle Process Object deletion. Kills the CID Handle here as done by Hartmut. * Added PspDeletethread here. * Renamed and rewrote PsTerminateCurrentThread to PspExitThread. Used NT Implementation out- lined in Windows Internals, Chapter 13. Uses less locks, a more concise order of actions, actually parses the Termination Ports, handles Dbgk notification. Timers are now rundown, and Mutex rundown is in a dedicated Kernel function. Final termination handled by KeTerminate Thread as documented. * Renamed PsTerminateOtherThread to PspTerminateThreadByPointer and modified implementation to be compatible with the changes above. * Renamed and regrouped Process Termination into PspExitProcess. Also implemented as described above, and moved each subsystem specific termination helper into its own subsytem. * Improved NtTerminateProcess and added more debugging messages. * Improved NtTerminateThread and added check against System Thread and made it compatible with new implementation. * Corrected PsTerminateSystemThread now that we support System Threads. * Corrected NtRegisterThreadTerminatePort to use same structure name as on windows for the port, and added tag to pool allocation (documented in pooltag.txt) include/internal/*.h: * Defined Scheduler Functions and misc new functions or renamed functions. ke/apc.c: * Fixed critical bug where APCs were not delivered at all if the thread wastion and cancels any timers that are associated to a thread, as well as their APCs and DPCs. REGRESSIONS FOUND: NONE BUGS/REGRESSIOSN FIXED: * Thread/Get Set Context now works. * Suspend Regression test now works. * Task manager can now kill foreign processes, even hung ones (like it should). * ExitProcess/closing cmd.exe with the 'x' button now works correctly without hacks. KNOWN ISSUES: I left a bit of a mess in the headers and some stuff still needs to be moved into the right places. I just wanted to have this first part ready first, so that it won't get too big. svn path=/trunk/; revision=14174
2005-03-18 05:53:04 +00:00
/* Unlock the process */
//PsUnlockProcess(Process);
#endif
/*
* Since we cannot change the session ID of the given
* process anymore because it is set once and for all
* at process creation time and because it is stored
* inside the Process->Session structure managed by MM,
* we fake changing it: we just return success if the
* user-defined value is the same as the session ID of
* the process, and otherwise we fail.
*/
if (SessionInfo.SessionId == PsGetProcessSessionId(Process))
{
Status = STATUS_SUCCESS;
}
else
{
Status = STATUS_ACCESS_DENIED;
}
break;
case ProcessPriorityClass:
/* Check buffer length */
if (ProcessInformationLength != sizeof(PROCESS_PRIORITY_CLASS))
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
/* Enter SEH for capture */
_SEH2_TRY
{
/* Capture the caller's buffer */
PriorityClass = *(PPROCESS_PRIORITY_CLASS)ProcessInformation;
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Return the exception code */
Status = _SEH2_GetExceptionCode();
_SEH2_YIELD(break);
}
_SEH2_END;
/* Check for invalid PriorityClass value */
if (PriorityClass.PriorityClass > PROCESS_PRIORITY_CLASS_ABOVE_NORMAL)
{
Status = STATUS_INVALID_PARAMETER;
break;
}
if ((PriorityClass.PriorityClass != Process->PriorityClass) &&
(PriorityClass.PriorityClass == PROCESS_PRIORITY_CLASS_REALTIME))
{
/* Check the privilege */
HasPrivilege = SeCheckPrivilegedObject(SeIncreaseBasePriorityPrivilege,
ProcessHandle,
PROCESS_SET_INFORMATION,
PreviousMode);
if (!HasPrivilege)
{
ObDereferenceObject(Process);
DPRINT1("Privilege to change priority to realtime lacking\n");
return STATUS_PRIVILEGE_NOT_HELD;
}
}
/* Check if we have a job */
if (Process->Job)
{
DPRINT1("Jobs not yet supported\n");
}
/* Set process priority class */
Process->PriorityClass = PriorityClass.PriorityClass;
/* Set process priority mode (foreground or background) */
PsSetProcessPriorityByClass(Process,
PriorityClass.Foreground ?
PsProcessPriorityForeground :
PsProcessPriorityBackground);
Status = STATUS_SUCCESS;
break;
case ProcessForegroundInformation:
/* Check buffer length */
if (ProcessInformationLength != sizeof(PROCESS_FOREGROUND_BACKGROUND))
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
/* Enter SEH for capture */
_SEH2_TRY
{
/* Capture the caller's buffer */
Foreground = *(PPROCESS_FOREGROUND_BACKGROUND)ProcessInformation;
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Return the exception code */
Status = _SEH2_GetExceptionCode();
_SEH2_YIELD(break);
}
_SEH2_END;
/* Set process priority mode (foreground or background) */
PsSetProcessPriorityByClass(Process,
Foreground.Foreground ?
PsProcessPriorityForeground :
PsProcessPriorityBackground);
Status = STATUS_SUCCESS;
break;
case ProcessBasePriority:
/* Validate input length */
if (ProcessInformationLength != sizeof(KPRIORITY))
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
/* Enter SEH for direct buffer read */
_SEH2_TRY
{
BasePriority = *(KPRIORITY*)ProcessInformation;
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get exception code */
Break = 0;
Status = _SEH2_GetExceptionCode();
_SEH2_YIELD(break);
}
_SEH2_END;
/* Extract the memory priority out of there */
if (BasePriority & 0x80000000)
{
MemoryPriority = MEMORY_PRIORITY_FOREGROUND;
BasePriority &= ~0x80000000;
}
else
{
MemoryPriority = MEMORY_PRIORITY_BACKGROUND;
}
/* Validate the number */
if ((BasePriority > HIGH_PRIORITY) || (BasePriority <= LOW_PRIORITY))
{
ObDereferenceObject(Process);
return STATUS_INVALID_PARAMETER;
}
/* Check if the new base is higher */
if (BasePriority > Process->Pcb.BasePriority)
{
HasPrivilege = SeCheckPrivilegedObject(SeIncreaseBasePriorityPrivilege,
ProcessHandle,
PROCESS_SET_INFORMATION,
PreviousMode);
if (!HasPrivilege)
{
ObDereferenceObject(Process);
DPRINT1("Privilege to change priority from %lx to %lx lacking\n", BasePriority, Process->Pcb.BasePriority);
return STATUS_PRIVILEGE_NOT_HELD;
}
}
/* Call Ke */
KeSetPriorityAndQuantumProcess(&Process->Pcb, BasePriority, 0);
/* Now set the memory priority */
MmSetMemoryPriorityProcess(Process, MemoryPriority);
Status = STATUS_SUCCESS;
break;
case ProcessRaisePriority:
/* Validate input length */
if (ProcessInformationLength != sizeof(ULONG))
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
/* Enter SEH for direct buffer read */
_SEH2_TRY
{
Boost = *(PULONG)ProcessInformation;
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get exception code */
Break = 0;
Status = _SEH2_GetExceptionCode();
_SEH2_YIELD(break);
}
_SEH2_END;
/* Make sure the process isn't dying */
if (ExAcquireRundownProtection(&Process->RundownProtect))
{
/* Lock it */
KeEnterCriticalRegion();
ExAcquirePushLockShared(&Process->ProcessLock);
/* Loop the threads */
for (Next = Process->ThreadListHead.Flink;
Next != &Process->ThreadListHead;
Next = Next->Flink)
{
/* Call Ke for the thread */
Thread = CONTAINING_RECORD(Next, ETHREAD, ThreadListEntry);
KeBoostPriorityThread(&Thread->Tcb, Boost);
}
/* Release the lock and rundown */
ExReleasePushLockShared(&Process->ProcessLock);
KeLeaveCriticalRegion();
ExReleaseRundownProtection(&Process->RundownProtect);
/* Set success code */
Status = STATUS_SUCCESS;
}
else
{
/* Avoid race conditions */
Status = STATUS_PROCESS_IS_TERMINATING;
}
break;
case ProcessBreakOnTermination:
/* Check buffer length */
if (ProcessInformationLength != sizeof(ULONG))
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
/* Enter SEH for direct buffer read */
_SEH2_TRY
{
Break = *(PULONG)ProcessInformation;
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get exception code */
Break = 0;
Status = _SEH2_GetExceptionCode();
_SEH2_YIELD(break);
}
_SEH2_END;
/* Setting 'break on termination' requires the SeDebugPrivilege */
if (!SeSinglePrivilegeCheck(SeDebugPrivilege, PreviousMode))
{
/* We don't hold the privilege, bail out */
Status = STATUS_PRIVILEGE_NOT_HELD;
break;
}
/* Set or clear the flag */
if (Break)
{
PspSetProcessFlag(Process, PSF_BREAK_ON_TERMINATION_BIT);
}
else
{
PspClearProcessFlag(Process, PSF_BREAK_ON_TERMINATION_BIT);
}
break;
case ProcessAffinityMask:
/* Check buffer length */
if (ProcessInformationLength != sizeof(KAFFINITY))
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
/* Enter SEH for direct buffer read */
_SEH2_TRY
{
Affinity = *(PKAFFINITY)ProcessInformation;
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get exception code */
Break = 0;
Status = _SEH2_GetExceptionCode();
_SEH2_YIELD(break);
}
_SEH2_END;
/* Make sure it's valid for the CPUs present */
ValidAffinity = Affinity & KeActiveProcessors;
if (!Affinity || (ValidAffinity != Affinity))
{
Status = STATUS_INVALID_PARAMETER;
break;
}
/* Check if it's within job affinity limits */
if (Process->Job)
{
/* Not yet implemented */
UNIMPLEMENTED;
Status = STATUS_NOT_IMPLEMENTED;
break;
}
/* Make sure the process isn't dying */
if (ExAcquireRundownProtection(&Process->RundownProtect))
{
/* Lock it */
KeEnterCriticalRegion();
ExAcquirePushLockShared(&Process->ProcessLock);
/* Call Ke to do the work */
KeSetAffinityProcess(&Process->Pcb, ValidAffinity);
/* Release the lock and rundown */
ExReleasePushLockShared(&Process->ProcessLock);
KeLeaveCriticalRegion();
ExReleaseRundownProtection(&Process->RundownProtect);
/* Set success code */
Status = STATUS_SUCCESS;
}
else
{
/* Avoid race conditions */
Status = STATUS_PROCESS_IS_TERMINATING;
}
break;
/* Priority Boosting status */
case ProcessPriorityBoost:
/* Validate input length */
if (ProcessInformationLength != sizeof(ULONG))
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
/* Enter SEH for direct buffer read */
_SEH2_TRY
{
DisableBoost = *(PBOOLEAN)ProcessInformation;
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get exception code */
Break = 0;
Status = _SEH2_GetExceptionCode();
_SEH2_YIELD(break);
}
_SEH2_END;
/* Make sure the process isn't dying */
if (ExAcquireRundownProtection(&Process->RundownProtect))
{
/* Lock it */
KeEnterCriticalRegion();
ExAcquirePushLockShared(&Process->ProcessLock);
/* Call Ke to do the work */
KeSetDisableBoostProcess(&Process->Pcb, DisableBoost);
/* Loop the threads too */
for (Next = Process->ThreadListHead.Flink;
Next != &Process->ThreadListHead;
Next = Next->Flink)
{
/* Call Ke for the thread */
Thread = CONTAINING_RECORD(Next, ETHREAD, ThreadListEntry);
KeSetDisableBoostThread(&Thread->Tcb, DisableBoost);
}
/* Release the lock and rundown */
ExReleasePushLockShared(&Process->ProcessLock);
KeLeaveCriticalRegion();
ExReleaseRundownProtection(&Process->RundownProtect);
/* Set success code */
Status = STATUS_SUCCESS;
}
else
{
/* Avoid race conditions */
Status = STATUS_PROCESS_IS_TERMINATING;
}
break;
case ProcessDebugFlags:
/* Check buffer length */
if (ProcessInformationLength != sizeof(ULONG))
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
/* Enter SEH for direct buffer read */
_SEH2_TRY
{
DebugFlags = *(PULONG)ProcessInformation;
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get exception code */
Status = _SEH2_GetExceptionCode();
_SEH2_YIELD(break);
}
_SEH2_END;
/* Set the mode */
if (DebugFlags & ~1)
{
Status = STATUS_INVALID_PARAMETER;
}
else
{
if (DebugFlags & 1)
{
PspClearProcessFlag(Process, PSF_NO_DEBUG_INHERIT_BIT);
}
else
{
PspSetProcessFlag(Process, PSF_NO_DEBUG_INHERIT_BIT);
}
}
/* Done */
Status = STATUS_SUCCESS;
break;
case ProcessEnableAlignmentFaultFixup:
/* Check buffer length */
if (ProcessInformationLength != sizeof(ULONG))
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
/* Enter SEH for direct buffer read */
_SEH2_TRY
{
EnableFixup = *(PULONG)ProcessInformation;
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get exception code */
Status = _SEH2_GetExceptionCode();
_SEH2_YIELD(break);
}
_SEH2_END;
/* Set the mode */
if (EnableFixup)
{
Process->DefaultHardErrorProcessing |= SEM_NOALIGNMENTFAULTEXCEPT;
}
else
{
Process->DefaultHardErrorProcessing &= ~SEM_NOALIGNMENTFAULTEXCEPT;
}
/* Call Ke for the update */
KeSetAutoAlignmentProcess(&Process->Pcb, FALSE);
Status = STATUS_SUCCESS;
break;
case ProcessUserModeIOPL:
/* Only TCB can do this */
if (!SeSinglePrivilegeCheck(SeTcbPrivilege, PreviousMode))
{
/* We don't hold the privilege, bail out */
DPRINT1("Need TCB to set IOPL\n");
Status = STATUS_PRIVILEGE_NOT_HELD;
break;
}
/* Only supported on x86 */
#if defined (_X86_)
Ke386SetIOPL();
#else
Status = STATUS_NOT_IMPLEMENTED;
#endif
/* Done */
break;
case ProcessExecuteFlags:
/* Check buffer length */
if (ProcessInformationLength != sizeof(ULONG))
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
if (ProcessHandle != NtCurrentProcess())
{
Status = STATUS_INVALID_PARAMETER;
break;
}
/* Enter SEH for direct buffer read */
_SEH2_TRY
{
NoExecute = *(PULONG)ProcessInformation;
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get exception code */
Status = _SEH2_GetExceptionCode();
_SEH2_YIELD(break);
}
_SEH2_END;
/* Call Mm for the update */
Status = MmSetExecuteOptions(NoExecute);
break;
case ProcessDeviceMap:
/* Check buffer length */
if (ProcessInformationLength != sizeof(HANDLE))
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
/* Use SEH for capture */
_SEH2_TRY
{
/* Capture the handle */
DirectoryHandle = *(PHANDLE)ProcessInformation;
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get the exception code */
Status = _SEH2_GetExceptionCode();
_SEH2_YIELD(break);
}
_SEH2_END;
/* Call Ob to set the device map */
Status = ObSetDeviceMap(Process, DirectoryHandle);
break;
/* We currently don't implement any of these */
case ProcessLdtInformation:
case ProcessLdtSize:
case ProcessIoPortHandlers:
DPRINT1("VDM/16-bit Request not implemented: %lx\n", ProcessInformationClass);
Status = STATUS_NOT_IMPLEMENTED;
break;
case ProcessQuotaLimits:
Status = PspSetQuotaLimits(Process,
1,
ProcessInformation,
ProcessInformationLength,
PreviousMode);
break;
case ProcessWorkingSetWatch:
DPRINT1("WS watch not implemented\n");
Status = STATUS_NOT_IMPLEMENTED;
break;
case ProcessHandleTracing:
DPRINT1("Handle tracing not implemented\n");
Status = STATUS_NOT_IMPLEMENTED;
break;
/* Anything else is invalid */
default:
DPRINT1("Invalid Server 2003 Info Class: %lx\n", ProcessInformationClass);
Status = STATUS_INVALID_INFO_CLASS;
}
/* Dereference and return status */
ObDereferenceObject(Process);
return Status;
Thread/Process Termination/Repeaing Rewrite + Fixes --------------------------------------------------- - ps/cid.c: * Moved CID Lookup functions here - ps/security.c: * Moved all security related functions here. Makes other files neater and security functions easier to locate. - ps/thread.c: * Moved most of the Thread Scheduling/Dispatching code that belongs in the Kernel to /ke and renamed functions from Ps to Ki. * Implemented PsIsSystemThread. * Removed Reaper Thread Init (now obsolete). * Renamed PiDeleteThread to PspDeleteThread. * Moved Thread State functions from tinfo.c to here. - ps/process.c: * Removed Query/Set Process functions and moved to ps/query.c * Renamed PiDeletePRocess to PspDeleteProcess * Removed obsoleted Process Termination functions, moved persistent one to kill.c - ps/create.c: * Moved the security APIs to security.c * Correctly implemented PsCreateSystemThread to actually create system threads. - ps/suspend.c * Rewrote Nt Executive functions to use Kernel functions. * Moved Ps* Routines into ke/kthread.c and fixed them. The implementation was wrong in some aspects, especially the issue of the APC looping around the KeWaitXxx call and the fact that the routines excluded/ignored the FreezeCount. - ps/debug.c * Fixed completely broken implementation of Get/SetThreadContext. The old version crashed when called and did not work at all. Suspend Regression test now works. * Moved Context<->TrapFrame functions to ke/i386/ * Combined Set/GetThreadContext APCs into a single one, and used special context structure. - ps/query.c: * Moved Thread/Process Query/Set Routines here. - ps/tinfo.c: * Removed. - ps/kill.c * Removed complicated Process Termination semantics and useless Attach/Detach in favor for a much more lightweight function which performs the same tasks as before and actually works. TaskManager can now terminate foreign processes. * Rewrote Thread Reaping to use the HyperCritical Work Queue instead of manually controlled thread. This results in much less code as well as an increase in speed and less micro management. The reaper is PspReapRoutine. Closing CMD.EXE now works properly without requiring masks that were added as hacks to allow it. * Renamed PiTerminateProcessThreads to PspTerminateProcessThreads. Fixed it to work with new termination code. * Added PspDeleteProcess to handle Process Object deletion. Kills the CID Handle here as done by Hartmut. * Added PspDeletethread here. * Renamed and rewrote PsTerminateCurrentThread to PspExitThread. Used NT Implementation out- lined in Windows Internals, Chapter 13. Uses less locks, a more concise order of actions, actually parses the Termination Ports, handles Dbgk notification. Timers are now rundown, and Mutex rundown is in a dedicated Kernel function. Final termination handled by KeTerminate Thread as documented. * Renamed PsTerminateOtherThread to PspTerminateThreadByPointer and modified implementation to be compatible with the changes above. * Renamed and regrouped Process Termination into PspExitProcess. Also implemented as described above, and moved each subsystem specific termination helper into its own subsytem. * Improved NtTerminateProcess and added more debugging messages. * Improved NtTerminateThread and added check against System Thread and made it compatible with new implementation. * Corrected PsTerminateSystemThread now that we support System Threads. * Corrected NtRegisterThreadTerminatePort to use same structure name as on windows for the port, and added tag to pool allocation (documented in pooltag.txt) include/internal/*.h: * Defined Scheduler Functions and misc new functions or renamed functions. ke/apc.c: * Fixed critical bug where APCs were not delivered at all if the thread wastion and cancels any timers that are associated to a thread, as well as their APCs and DPCs. REGRESSIONS FOUND: NONE BUGS/REGRESSIOSN FIXED: * Thread/Get Set Context now works. * Suspend Regression test now works. * Task manager can now kill foreign processes, even hung ones (like it should). * ExitProcess/closing cmd.exe with the 'x' button now works correctly without hacks. KNOWN ISSUES: I left a bit of a mess in the headers and some stuff still needs to be moved into the right places. I just wanted to have this first part ready first, so that it won't get too big. svn path=/trunk/; revision=14174
2005-03-18 05:53:04 +00:00
}
/*
* @implemented
Thread/Process Termination/Repeaing Rewrite + Fixes --------------------------------------------------- - ps/cid.c: * Moved CID Lookup functions here - ps/security.c: * Moved all security related functions here. Makes other files neater and security functions easier to locate. - ps/thread.c: * Moved most of the Thread Scheduling/Dispatching code that belongs in the Kernel to /ke and renamed functions from Ps to Ki. * Implemented PsIsSystemThread. * Removed Reaper Thread Init (now obsolete). * Renamed PiDeleteThread to PspDeleteThread. * Moved Thread State functions from tinfo.c to here. - ps/process.c: * Removed Query/Set Process functions and moved to ps/query.c * Renamed PiDeletePRocess to PspDeleteProcess * Removed obsoleted Process Termination functions, moved persistent one to kill.c - ps/create.c: * Moved the security APIs to security.c * Correctly implemented PsCreateSystemThread to actually create system threads. - ps/suspend.c * Rewrote Nt Executive functions to use Kernel functions. * Moved Ps* Routines into ke/kthread.c and fixed them. The implementation was wrong in some aspects, especially the issue of the APC looping around the KeWaitXxx call and the fact that the routines excluded/ignored the FreezeCount. - ps/debug.c * Fixed completely broken implementation of Get/SetThreadContext. The old version crashed when called and did not work at all. Suspend Regression test now works. * Moved Context<->TrapFrame functions to ke/i386/ * Combined Set/GetThreadContext APCs into a single one, and used special context structure. - ps/query.c: * Moved Thread/Process Query/Set Routines here. - ps/tinfo.c: * Removed. - ps/kill.c * Removed complicated Process Termination semantics and useless Attach/Detach in favor for a much more lightweight function which performs the same tasks as before and actually works. TaskManager can now terminate foreign processes. * Rewrote Thread Reaping to use the HyperCritical Work Queue instead of manually controlled thread. This results in much less code as well as an increase in speed and less micro management. The reaper is PspReapRoutine. Closing CMD.EXE now works properly without requiring masks that were added as hacks to allow it. * Renamed PiTerminateProcessThreads to PspTerminateProcessThreads. Fixed it to work with new termination code. * Added PspDeleteProcess to handle Process Object deletion. Kills the CID Handle here as done by Hartmut. * Added PspDeletethread here. * Renamed and rewrote PsTerminateCurrentThread to PspExitThread. Used NT Implementation out- lined in Windows Internals, Chapter 13. Uses less locks, a more concise order of actions, actually parses the Termination Ports, handles Dbgk notification. Timers are now rundown, and Mutex rundown is in a dedicated Kernel function. Final termination handled by KeTerminate Thread as documented. * Renamed PsTerminateOtherThread to PspTerminateThreadByPointer and modified implementation to be compatible with the changes above. * Renamed and regrouped Process Termination into PspExitProcess. Also implemented as described above, and moved each subsystem specific termination helper into its own subsytem. * Improved NtTerminateProcess and added more debugging messages. * Improved NtTerminateThread and added check against System Thread and made it compatible with new implementation. * Corrected PsTerminateSystemThread now that we support System Threads. * Corrected NtRegisterThreadTerminatePort to use same structure name as on windows for the port, and added tag to pool allocation (documented in pooltag.txt) include/internal/*.h: * Defined Scheduler Functions and misc new functions or renamed functions. ke/apc.c: * Fixed critical bug where APCs were not delivered at all if the thread wastion and cancels any timers that are associated to a thread, as well as their APCs and DPCs. REGRESSIONS FOUND: NONE BUGS/REGRESSIOSN FIXED: * Thread/Get Set Context now works. * Suspend Regression test now works. * Task manager can now kill foreign processes, even hung ones (like it should). * ExitProcess/closing cmd.exe with the 'x' button now works correctly without hacks. KNOWN ISSUES: I left a bit of a mess in the headers and some stuff still needs to be moved into the right places. I just wanted to have this first part ready first, so that it won't get too big. svn path=/trunk/; revision=14174
2005-03-18 05:53:04 +00:00
*/
NTSTATUS
NTAPI
NtSetInformationThread(IN HANDLE ThreadHandle,
IN THREADINFOCLASS ThreadInformationClass,
IN PVOID ThreadInformation,
IN ULONG ThreadInformationLength)
Thread/Process Termination/Repeaing Rewrite + Fixes --------------------------------------------------- - ps/cid.c: * Moved CID Lookup functions here - ps/security.c: * Moved all security related functions here. Makes other files neater and security functions easier to locate. - ps/thread.c: * Moved most of the Thread Scheduling/Dispatching code that belongs in the Kernel to /ke and renamed functions from Ps to Ki. * Implemented PsIsSystemThread. * Removed Reaper Thread Init (now obsolete). * Renamed PiDeleteThread to PspDeleteThread. * Moved Thread State functions from tinfo.c to here. - ps/process.c: * Removed Query/Set Process functions and moved to ps/query.c * Renamed PiDeletePRocess to PspDeleteProcess * Removed obsoleted Process Termination functions, moved persistent one to kill.c - ps/create.c: * Moved the security APIs to security.c * Correctly implemented PsCreateSystemThread to actually create system threads. - ps/suspend.c * Rewrote Nt Executive functions to use Kernel functions. * Moved Ps* Routines into ke/kthread.c and fixed them. The implementation was wrong in some aspects, especially the issue of the APC looping around the KeWaitXxx call and the fact that the routines excluded/ignored the FreezeCount. - ps/debug.c * Fixed completely broken implementation of Get/SetThreadContext. The old version crashed when called and did not work at all. Suspend Regression test now works. * Moved Context<->TrapFrame functions to ke/i386/ * Combined Set/GetThreadContext APCs into a single one, and used special context structure. - ps/query.c: * Moved Thread/Process Query/Set Routines here. - ps/tinfo.c: * Removed. - ps/kill.c * Removed complicated Process Termination semantics and useless Attach/Detach in favor for a much more lightweight function which performs the same tasks as before and actually works. TaskManager can now terminate foreign processes. * Rewrote Thread Reaping to use the HyperCritical Work Queue instead of manually controlled thread. This results in much less code as well as an increase in speed and less micro management. The reaper is PspReapRoutine. Closing CMD.EXE now works properly without requiring masks that were added as hacks to allow it. * Renamed PiTerminateProcessThreads to PspTerminateProcessThreads. Fixed it to work with new termination code. * Added PspDeleteProcess to handle Process Object deletion. Kills the CID Handle here as done by Hartmut. * Added PspDeletethread here. * Renamed and rewrote PsTerminateCurrentThread to PspExitThread. Used NT Implementation out- lined in Windows Internals, Chapter 13. Uses less locks, a more concise order of actions, actually parses the Termination Ports, handles Dbgk notification. Timers are now rundown, and Mutex rundown is in a dedicated Kernel function. Final termination handled by KeTerminate Thread as documented. * Renamed PsTerminateOtherThread to PspTerminateThreadByPointer and modified implementation to be compatible with the changes above. * Renamed and regrouped Process Termination into PspExitProcess. Also implemented as described above, and moved each subsystem specific termination helper into its own subsytem. * Improved NtTerminateProcess and added more debugging messages. * Improved NtTerminateThread and added check against System Thread and made it compatible with new implementation. * Corrected PsTerminateSystemThread now that we support System Threads. * Corrected NtRegisterThreadTerminatePort to use same structure name as on windows for the port, and added tag to pool allocation (documented in pooltag.txt) include/internal/*.h: * Defined Scheduler Functions and misc new functions or renamed functions. ke/apc.c: * Fixed critical bug where APCs were not delivered at all if the thread wastion and cancels any timers that are associated to a thread, as well as their APCs and DPCs. REGRESSIONS FOUND: NONE BUGS/REGRESSIOSN FIXED: * Thread/Get Set Context now works. * Suspend Regression test now works. * Task manager can now kill foreign processes, even hung ones (like it should). * ExitProcess/closing cmd.exe with the 'x' button now works correctly without hacks. KNOWN ISSUES: I left a bit of a mess in the headers and some stuff still needs to be moved into the right places. I just wanted to have this first part ready first, so that it won't get too big. svn path=/trunk/; revision=14174
2005-03-18 05:53:04 +00:00
{
PETHREAD Thread;
ULONG Access;
KPROCESSOR_MODE PreviousMode = ExGetPreviousMode();
NTSTATUS Status;
HANDLE TokenHandle = NULL;
KPRIORITY Priority = 0;
KAFFINITY Affinity = 0, CombinedAffinity;
PVOID Address = NULL;
PEPROCESS Process;
ULONG_PTR DisableBoost = 0;
ULONG_PTR IdealProcessor = 0;
ULONG_PTR Break = 0;
PTEB Teb;
ULONG_PTR TlsIndex = 0;
PVOID *ExpansionSlots;
PETHREAD ProcThread;
PAGED_CODE();
/* Verify Information Class validity */
#if 0
Status = DefaultSetInfoBufferCheck(ThreadInformationClass,
PsThreadInfoClass,
RTL_NUMBER_OF(PsThreadInfoClass),
ThreadInformation,
ThreadInformationLength,
PreviousMode);
if (!NT_SUCCESS(Status)) return Status;
#endif
/* Check what class this is */
Access = THREAD_SET_INFORMATION;
if (ThreadInformationClass == ThreadImpersonationToken)
{
/* Setting the impersonation token needs a special mask */
Access = THREAD_SET_THREAD_TOKEN;
}
/* Reference the thread */
Status = ObReferenceObjectByHandle(ThreadHandle,
Access,
PsThreadType,
PreviousMode,
(PVOID*)&Thread,
NULL);
if (!NT_SUCCESS(Status)) return Status;
/* Check what kind of information class this is */
switch (ThreadInformationClass)
{
/* Thread priority */
case ThreadPriority:
/* Check buffer length */
if (ThreadInformationLength != sizeof(KPRIORITY))
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
/* Use SEH for capture */
_SEH2_TRY
{
/* Get the priority */
Priority = *(PLONG)ThreadInformation;
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get the exception code */
Status = _SEH2_GetExceptionCode();
_SEH2_YIELD(break);
}
_SEH2_END;
/* Validate it */
if ((Priority > HIGH_PRIORITY) ||
(Priority <= LOW_PRIORITY))
{
/* Fail */
Status = STATUS_INVALID_PARAMETER;
break;
}
/* Set the priority */
KeSetPriorityThread(&Thread->Tcb, Priority);
break;
case ThreadBasePriority:
/* Check buffer length */
if (ThreadInformationLength != sizeof(LONG))
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
/* Use SEH for capture */
_SEH2_TRY
{
/* Get the priority */
Priority = *(PLONG)ThreadInformation;
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get the exception code */
Status = _SEH2_GetExceptionCode();
_SEH2_YIELD(break);
}
_SEH2_END;
/* Validate it */
if ((Priority > THREAD_BASE_PRIORITY_MAX) ||
(Priority < THREAD_BASE_PRIORITY_MIN))
{
/* These ones are OK */
if ((Priority != THREAD_BASE_PRIORITY_LOWRT + 1) &&
(Priority != THREAD_BASE_PRIORITY_IDLE - 1))
{
/* Check if the process is real time */
if (PsGetCurrentProcess()->PriorityClass !=
PROCESS_PRIORITY_CLASS_REALTIME)
{
/* It isn't, fail */
Status = STATUS_INVALID_PARAMETER;
break;
}
}
}
/* Set the base priority */
KeSetBasePriorityThread(&Thread->Tcb, Priority);
break;
case ThreadAffinityMask:
/* Check buffer length */
if (ThreadInformationLength != sizeof(ULONG_PTR))
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
/* Use SEH for capture */
_SEH2_TRY
{
/* Get the priority */
Affinity = *(PULONG_PTR)ThreadInformation;
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get the exception code */
Status = _SEH2_GetExceptionCode();
_SEH2_YIELD(break);
}
_SEH2_END;
/* Validate it */
if (!Affinity)
{
/* Fail */
Status = STATUS_INVALID_PARAMETER;
break;
}
/* Get the process */
Process = Thread->ThreadsProcess;
/* Try to acquire rundown */
if (ExAcquireRundownProtection(&Process->RundownProtect))
{
/* Lock it */
KeEnterCriticalRegion();
ExAcquirePushLockShared(&Process->ProcessLock);
/* Combine masks */
CombinedAffinity = Affinity & Process->Pcb.Affinity;
if (CombinedAffinity != Affinity)
{
/* Fail */
Status = STATUS_INVALID_PARAMETER;
}
else
{
/* Set the affinity */
KeSetAffinityThread(&Thread->Tcb, CombinedAffinity);
}
/* Release the lock and rundown */
ExReleasePushLockShared(&Process->ProcessLock);
KeLeaveCriticalRegion();
ExReleaseRundownProtection(&Process->RundownProtect);
}
else
{
/* Too late */
Status = STATUS_PROCESS_IS_TERMINATING;
}
/* Return status */
break;
case ThreadImpersonationToken:
/* Check buffer length */
if (ThreadInformationLength != sizeof(HANDLE))
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
/* Use SEH for capture */
_SEH2_TRY
{
/* Save the token handle */
TokenHandle = *(PHANDLE)ThreadInformation;
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get the exception code */
Status = _SEH2_GetExceptionCode();
_SEH2_YIELD(break);
}
_SEH2_END;
/* Assign the actual token */
Status = PsAssignImpersonationToken(Thread, TokenHandle);
break;
case ThreadQuerySetWin32StartAddress:
/* Check buffer length */
if (ThreadInformationLength != sizeof(ULONG_PTR))
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
/* Use SEH for capture */
_SEH2_TRY
{
/* Get the priority */
Address = *(PVOID*)ThreadInformation;
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get the exception code */
Status = _SEH2_GetExceptionCode();
_SEH2_YIELD(break);
}
_SEH2_END;
/* Set the address */
Thread->Win32StartAddress = Address;
break;
case ThreadIdealProcessor:
/* Check buffer length */
if (ThreadInformationLength != sizeof(ULONG_PTR))
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
/* Use SEH for capture */
_SEH2_TRY
{
/* Get the priority */
IdealProcessor = *(PULONG_PTR)ThreadInformation;
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get the exception code */
Status = _SEH2_GetExceptionCode();
_SEH2_YIELD(break);
}
_SEH2_END;
/* Validate it */
if (IdealProcessor > MAXIMUM_PROCESSORS)
{
/* Fail */
Status = STATUS_INVALID_PARAMETER;
break;
}
/* Set the ideal */
Status = KeSetIdealProcessorThread(&Thread->Tcb,
(CCHAR)IdealProcessor);
/* Get the TEB and protect the thread */
Teb = Thread->Tcb.Teb;
if ((Teb) && (ExAcquireRundownProtection(&Thread->RundownProtect)))
{
/* Save the ideal processor */
Teb->IdealProcessor = Thread->Tcb.IdealProcessor;
/* Release rundown protection */
ExReleaseRundownProtection(&Thread->RundownProtect);
}
break;
case ThreadPriorityBoost:
/* Check buffer length */
if (ThreadInformationLength != sizeof(ULONG_PTR))
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
/* Use SEH for capture */
_SEH2_TRY
{
/* Get the priority */
DisableBoost = *(PULONG_PTR)ThreadInformation;
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get the exception code */
Status = _SEH2_GetExceptionCode();
_SEH2_YIELD(break);
}
_SEH2_END;
/* Call the kernel */
KeSetDisableBoostThread(&Thread->Tcb, (BOOLEAN)DisableBoost);
break;
case ThreadZeroTlsCell:
/* Check buffer length */
if (ThreadInformationLength != sizeof(ULONG_PTR))
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
/* Use SEH for capture */
_SEH2_TRY
{
/* Get the priority */
TlsIndex = *(PULONG_PTR)ThreadInformation;
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get the exception code */
Status = _SEH2_GetExceptionCode();
_SEH2_YIELD(break);
}
_SEH2_END;
/* This is only valid for the current thread */
if (Thread != PsGetCurrentThread())
{
/* Fail */
Status = STATUS_INVALID_PARAMETER;
break;
}
/* Get the process */
Process = Thread->ThreadsProcess;
/* Loop the threads */
ProcThread = PsGetNextProcessThread(Process, NULL);
while (ProcThread)
{
/* Acquire rundown */
if (ExAcquireRundownProtection(&ProcThread->RundownProtect))
{
/* Get the TEB */
Teb = ProcThread->Tcb.Teb;
if (Teb)
{
/* Check if we're in the expansion range */
if (TlsIndex > TLS_MINIMUM_AVAILABLE - 1)
{
if (TlsIndex < (TLS_MINIMUM_AVAILABLE +
TLS_EXPANSION_SLOTS) - 1)
{
/* Check if we have expansion slots */
ExpansionSlots = Teb->TlsExpansionSlots;
if (ExpansionSlots)
{
/* Clear the index */
ExpansionSlots[TlsIndex - TLS_MINIMUM_AVAILABLE] = 0;
}
}
}
else
{
/* Clear the index */
Teb->TlsSlots[TlsIndex] = NULL;
}
}
/* Release rundown */
ExReleaseRundownProtection(&ProcThread->RundownProtect);
}
/* Go to the next thread */
ProcThread = PsGetNextProcessThread(Process, ProcThread);
}
/* All done */
break;
case ThreadBreakOnTermination:
/* Check buffer length */
if (ThreadInformationLength != sizeof(ULONG))
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
/* Enter SEH for direct buffer read */
_SEH2_TRY
{
Break = *(PULONG)ThreadInformation;
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get exception code */
Break = 0;
Status = _SEH2_GetExceptionCode();
_SEH2_YIELD(break);
}
_SEH2_END;
/* Setting 'break on termination' requires the SeDebugPrivilege */
if (!SeSinglePrivilegeCheck(SeDebugPrivilege, PreviousMode))
{
/* We don't hold the privilege, bail out */
Status = STATUS_PRIVILEGE_NOT_HELD;
break;
}
/* Set or clear the flag */
if (Break)
{
PspSetCrossThreadFlag(Thread, CT_BREAK_ON_TERMINATION_BIT);
}
else
{
PspClearCrossThreadFlag(Thread, CT_BREAK_ON_TERMINATION_BIT);
}
break;
case ThreadHideFromDebugger:
/* Check buffer length */
if (ThreadInformationLength != 0)
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
/* Set the flag */
PspSetCrossThreadFlag(Thread, CT_HIDE_FROM_DEBUGGER_BIT);
break;
default:
/* We don't implement it yet */
DPRINT1("Not implemented: %d\n", ThreadInformationClass);
Status = STATUS_NOT_IMPLEMENTED;
}
/* Dereference and return status */
ObDereferenceObject(Thread);
return Status;
Thread/Process Termination/Repeaing Rewrite + Fixes --------------------------------------------------- - ps/cid.c: * Moved CID Lookup functions here - ps/security.c: * Moved all security related functions here. Makes other files neater and security functions easier to locate. - ps/thread.c: * Moved most of the Thread Scheduling/Dispatching code that belongs in the Kernel to /ke and renamed functions from Ps to Ki. * Implemented PsIsSystemThread. * Removed Reaper Thread Init (now obsolete). * Renamed PiDeleteThread to PspDeleteThread. * Moved Thread State functions from tinfo.c to here. - ps/process.c: * Removed Query/Set Process functions and moved to ps/query.c * Renamed PiDeletePRocess to PspDeleteProcess * Removed obsoleted Process Termination functions, moved persistent one to kill.c - ps/create.c: * Moved the security APIs to security.c * Correctly implemented PsCreateSystemThread to actually create system threads. - ps/suspend.c * Rewrote Nt Executive functions to use Kernel functions. * Moved Ps* Routines into ke/kthread.c and fixed them. The implementation was wrong in some aspects, especially the issue of the APC looping around the KeWaitXxx call and the fact that the routines excluded/ignored the FreezeCount. - ps/debug.c * Fixed completely broken implementation of Get/SetThreadContext. The old version crashed when called and did not work at all. Suspend Regression test now works. * Moved Context<->TrapFrame functions to ke/i386/ * Combined Set/GetThreadContext APCs into a single one, and used special context structure. - ps/query.c: * Moved Thread/Process Query/Set Routines here. - ps/tinfo.c: * Removed. - ps/kill.c * Removed complicated Process Termination semantics and useless Attach/Detach in favor for a much more lightweight function which performs the same tasks as before and actually works. TaskManager can now terminate foreign processes. * Rewrote Thread Reaping to use the HyperCritical Work Queue instead of manually controlled thread. This results in much less code as well as an increase in speed and less micro management. The reaper is PspReapRoutine. Closing CMD.EXE now works properly without requiring masks that were added as hacks to allow it. * Renamed PiTerminateProcessThreads to PspTerminateProcessThreads. Fixed it to work with new termination code. * Added PspDeleteProcess to handle Process Object deletion. Kills the CID Handle here as done by Hartmut. * Added PspDeletethread here. * Renamed and rewrote PsTerminateCurrentThread to PspExitThread. Used NT Implementation out- lined in Windows Internals, Chapter 13. Uses less locks, a more concise order of actions, actually parses the Termination Ports, handles Dbgk notification. Timers are now rundown, and Mutex rundown is in a dedicated Kernel function. Final termination handled by KeTerminate Thread as documented. * Renamed PsTerminateOtherThread to PspTerminateThreadByPointer and modified implementation to be compatible with the changes above. * Renamed and regrouped Process Termination into PspExitProcess. Also implemented as described above, and moved each subsystem specific termination helper into its own subsytem. * Improved NtTerminateProcess and added more debugging messages. * Improved NtTerminateThread and added check against System Thread and made it compatible with new implementation. * Corrected PsTerminateSystemThread now that we support System Threads. * Corrected NtRegisterThreadTerminatePort to use same structure name as on windows for the port, and added tag to pool allocation (documented in pooltag.txt) include/internal/*.h: * Defined Scheduler Functions and misc new functions or renamed functions. ke/apc.c: * Fixed critical bug where APCs were not delivered at all if the thread wastion and cancels any timers that are associated to a thread, as well as their APCs and DPCs. REGRESSIONS FOUND: NONE BUGS/REGRESSIOSN FIXED: * Thread/Get Set Context now works. * Suspend Regression test now works. * Task manager can now kill foreign processes, even hung ones (like it should). * ExitProcess/closing cmd.exe with the 'x' button now works correctly without hacks. KNOWN ISSUES: I left a bit of a mess in the headers and some stuff still needs to be moved into the right places. I just wanted to have this first part ready first, so that it won't get too big. svn path=/trunk/; revision=14174
2005-03-18 05:53:04 +00:00
}
/*
* @implemented
*/
NTSTATUS
NTAPI
NtQueryInformationThread(IN HANDLE ThreadHandle,
IN THREADINFOCLASS ThreadInformationClass,
OUT PVOID ThreadInformation,
IN ULONG ThreadInformationLength,
OUT PULONG ReturnLength OPTIONAL)
Thread/Process Termination/Repeaing Rewrite + Fixes --------------------------------------------------- - ps/cid.c: * Moved CID Lookup functions here - ps/security.c: * Moved all security related functions here. Makes other files neater and security functions easier to locate. - ps/thread.c: * Moved most of the Thread Scheduling/Dispatching code that belongs in the Kernel to /ke and renamed functions from Ps to Ki. * Implemented PsIsSystemThread. * Removed Reaper Thread Init (now obsolete). * Renamed PiDeleteThread to PspDeleteThread. * Moved Thread State functions from tinfo.c to here. - ps/process.c: * Removed Query/Set Process functions and moved to ps/query.c * Renamed PiDeletePRocess to PspDeleteProcess * Removed obsoleted Process Termination functions, moved persistent one to kill.c - ps/create.c: * Moved the security APIs to security.c * Correctly implemented PsCreateSystemThread to actually create system threads. - ps/suspend.c * Rewrote Nt Executive functions to use Kernel functions. * Moved Ps* Routines into ke/kthread.c and fixed them. The implementation was wrong in some aspects, especially the issue of the APC looping around the KeWaitXxx call and the fact that the routines excluded/ignored the FreezeCount. - ps/debug.c * Fixed completely broken implementation of Get/SetThreadContext. The old version crashed when called and did not work at all. Suspend Regression test now works. * Moved Context<->TrapFrame functions to ke/i386/ * Combined Set/GetThreadContext APCs into a single one, and used special context structure. - ps/query.c: * Moved Thread/Process Query/Set Routines here. - ps/tinfo.c: * Removed. - ps/kill.c * Removed complicated Process Termination semantics and useless Attach/Detach in favor for a much more lightweight function which performs the same tasks as before and actually works. TaskManager can now terminate foreign processes. * Rewrote Thread Reaping to use the HyperCritical Work Queue instead of manually controlled thread. This results in much less code as well as an increase in speed and less micro management. The reaper is PspReapRoutine. Closing CMD.EXE now works properly without requiring masks that were added as hacks to allow it. * Renamed PiTerminateProcessThreads to PspTerminateProcessThreads. Fixed it to work with new termination code. * Added PspDeleteProcess to handle Process Object deletion. Kills the CID Handle here as done by Hartmut. * Added PspDeletethread here. * Renamed and rewrote PsTerminateCurrentThread to PspExitThread. Used NT Implementation out- lined in Windows Internals, Chapter 13. Uses less locks, a more concise order of actions, actually parses the Termination Ports, handles Dbgk notification. Timers are now rundown, and Mutex rundown is in a dedicated Kernel function. Final termination handled by KeTerminate Thread as documented. * Renamed PsTerminateOtherThread to PspTerminateThreadByPointer and modified implementation to be compatible with the changes above. * Renamed and regrouped Process Termination into PspExitProcess. Also implemented as described above, and moved each subsystem specific termination helper into its own subsytem. * Improved NtTerminateProcess and added more debugging messages. * Improved NtTerminateThread and added check against System Thread and made it compatible with new implementation. * Corrected PsTerminateSystemThread now that we support System Threads. * Corrected NtRegisterThreadTerminatePort to use same structure name as on windows for the port, and added tag to pool allocation (documented in pooltag.txt) include/internal/*.h: * Defined Scheduler Functions and misc new functions or renamed functions. ke/apc.c: * Fixed critical bug where APCs were not delivered at all if the thread wastion and cancels any timers that are associated to a thread, as well as their APCs and DPCs. REGRESSIONS FOUND: NONE BUGS/REGRESSIOSN FIXED: * Thread/Get Set Context now works. * Suspend Regression test now works. * Task manager can now kill foreign processes, even hung ones (like it should). * ExitProcess/closing cmd.exe with the 'x' button now works correctly without hacks. KNOWN ISSUES: I left a bit of a mess in the headers and some stuff still needs to be moved into the right places. I just wanted to have this first part ready first, so that it won't get too big. svn path=/trunk/; revision=14174
2005-03-18 05:53:04 +00:00
{
PETHREAD Thread;
KPROCESSOR_MODE PreviousMode = ExGetPreviousMode();
NTSTATUS Status;
ULONG Access;
ULONG Length = 0;
PTHREAD_BASIC_INFORMATION ThreadBasicInfo =
(PTHREAD_BASIC_INFORMATION)ThreadInformation;
PKERNEL_USER_TIMES ThreadTime = (PKERNEL_USER_TIMES)ThreadInformation;
KIRQL OldIrql;
ULONG ThreadTerminated;
PAGED_CODE();
/* Check if we were called from user mode */
if (PreviousMode != KernelMode)
{
/* Enter SEH */
_SEH2_TRY
{
/* Probe the buffer */
ProbeForWrite(ThreadInformation,
ThreadInformationLength,
sizeof(ULONG));
/* Probe the return length if required */
if (ReturnLength) ProbeForWriteUlong(ReturnLength);
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Return the exception code */
_SEH2_YIELD(return _SEH2_GetExceptionCode());
}
_SEH2_END;
}
/* Check what class this is */
Access = THREAD_QUERY_INFORMATION;
/* Reference the process */
Status = ObReferenceObjectByHandle(ThreadHandle,
Access,
PsThreadType,
PreviousMode,
(PVOID*)&Thread,
NULL);
if (!NT_SUCCESS(Status)) return Status;
/* Check what kind of information class this is */
switch (ThreadInformationClass)
{
/* Basic thread information */
case ThreadBasicInformation:
/* Set return length */
Length = sizeof(THREAD_BASIC_INFORMATION);
if (ThreadInformationLength != Length)
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
/* Protect writes with SEH */
_SEH2_TRY
{
/* Write all the information from the ETHREAD/KTHREAD */
ThreadBasicInfo->ExitStatus = Thread->ExitStatus;
ThreadBasicInfo->TebBaseAddress = (PVOID)Thread->Tcb.Teb;
ThreadBasicInfo->ClientId = Thread->Cid;
ThreadBasicInfo->AffinityMask = Thread->Tcb.Affinity;
ThreadBasicInfo->Priority = Thread->Tcb.Priority;
ThreadBasicInfo->BasePriority = KeQueryBasePriorityThread(&Thread->Tcb);
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get exception code */
Status = _SEH2_GetExceptionCode();
}
_SEH2_END;
break;
/* Thread time information */
case ThreadTimes:
/* Set the return length */
Length = sizeof(KERNEL_USER_TIMES);
if (ThreadInformationLength != Length)
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
/* Protect writes with SEH */
_SEH2_TRY
{
/* Copy time information from ETHREAD/KTHREAD */
ThreadTime->KernelTime.QuadPart = Thread->Tcb.KernelTime * KeMaximumIncrement;
ThreadTime->UserTime.QuadPart = Thread->Tcb.UserTime * KeMaximumIncrement;
ThreadTime->CreateTime = Thread->CreateTime;
/* Exit time is in a union and only valid on actual exit! */
if (KeReadStateThread(&Thread->Tcb))
{
ThreadTime->ExitTime = Thread->ExitTime;
}
else
{
ThreadTime->ExitTime.QuadPart = 0;
}
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get exception code */
Status = _SEH2_GetExceptionCode();
}
_SEH2_END;
break;
case ThreadQuerySetWin32StartAddress:
/* Set the return length*/
Length = sizeof(PVOID);
if (ThreadInformationLength != Length)
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
/* Protect write with SEH */
_SEH2_TRY
{
/* Return the Win32 Start Address */
*(PVOID*)ThreadInformation = Thread->Win32StartAddress;
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get exception code */
Status = _SEH2_GetExceptionCode();
}
_SEH2_END;
break;
case ThreadPerformanceCount:
/* Set the return length*/
Length = sizeof(LARGE_INTEGER);
if (ThreadInformationLength != Length)
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
/* Protect write with SEH */
_SEH2_TRY
{
/* FIXME */
(*(PLARGE_INTEGER)ThreadInformation).QuadPart = 0;
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get exception code */
Status = _SEH2_GetExceptionCode();
}
_SEH2_END;
break;
case ThreadAmILastThread:
/* Set the return length*/
Length = sizeof(ULONG);
if (ThreadInformationLength != Length)
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
/* Protect write with SEH */
_SEH2_TRY
{
/* Return whether or not we are the last thread */
*(PULONG)ThreadInformation = ((Thread->ThreadsProcess->
ThreadListHead.Flink->Flink ==
&Thread->ThreadsProcess->
ThreadListHead) ?
TRUE : FALSE);
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get exception code */
Status = _SEH2_GetExceptionCode();
}
_SEH2_END;
break;
case ThreadIsIoPending:
/* Set the return length*/
Length = sizeof(ULONG);
if (ThreadInformationLength != Length)
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
/* Raise the IRQL to protect the IRP list */
KeRaiseIrql(APC_LEVEL, &OldIrql);
/* Protect write with SEH */
_SEH2_TRY
{
/* Check if the IRP list is empty or not */
*(PULONG)ThreadInformation = !IsListEmpty(&Thread->IrpList);
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get exception code */
Status = _SEH2_GetExceptionCode();
}
_SEH2_END;
/* Lower IRQL back */
KeLowerIrql(OldIrql);
break;
/* LDT and GDT information */
case ThreadDescriptorTableEntry:
#if defined(_X86_)
/* Call the worker routine */
Status = PspQueryDescriptorThread(Thread,
ThreadInformation,
ThreadInformationLength,
ReturnLength);
#else
/* Only implemented on x86 */
Status = STATUS_NOT_IMPLEMENTED;
#endif
break;
case ThreadPriorityBoost:
/* Set the return length*/
Length = sizeof(ULONG);
if (ThreadInformationLength != Length)
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
_SEH2_TRY
{
*(PULONG)ThreadInformation = Thread->Tcb.DisableBoost ? 1 : 0;
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
Status = _SEH2_GetExceptionCode();
}
_SEH2_END;
break;
case ThreadIsTerminated:
/* Set the return length*/
Length = sizeof(ThreadTerminated);
if (ThreadInformationLength != Length)
{
Status = STATUS_INFO_LENGTH_MISMATCH;
break;
}
ThreadTerminated = PsIsThreadTerminating(Thread);
_SEH2_TRY
{
*(PULONG)ThreadInformation = ThreadTerminated ? 1 : 0;
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
Status = _SEH2_GetExceptionCode();
}
_SEH2_END;
break;
/* Anything else */
default:
/* Not yet implemented */
DPRINT1("Not implemented: %lx\n", ThreadInformationClass);
Status = STATUS_NOT_IMPLEMENTED;
}
/* Protect write with SEH */
_SEH2_TRY
{
/* Check if caller wanted return length */
if (ReturnLength) *ReturnLength = Length;
}
_SEH2_EXCEPT(EXCEPTION_EXECUTE_HANDLER)
{
/* Get exception code */
Status = _SEH2_GetExceptionCode();
}
_SEH2_END;
/* Dereference the thread, and return */
ObDereferenceObject(Thread);
return Status;
Thread/Process Termination/Repeaing Rewrite + Fixes --------------------------------------------------- - ps/cid.c: * Moved CID Lookup functions here - ps/security.c: * Moved all security related functions here. Makes other files neater and security functions easier to locate. - ps/thread.c: * Moved most of the Thread Scheduling/Dispatching code that belongs in the Kernel to /ke and renamed functions from Ps to Ki. * Implemented PsIsSystemThread. * Removed Reaper Thread Init (now obsolete). * Renamed PiDeleteThread to PspDeleteThread. * Moved Thread State functions from tinfo.c to here. - ps/process.c: * Removed Query/Set Process functions and moved to ps/query.c * Renamed PiDeletePRocess to PspDeleteProcess * Removed obsoleted Process Termination functions, moved persistent one to kill.c - ps/create.c: * Moved the security APIs to security.c * Correctly implemented PsCreateSystemThread to actually create system threads. - ps/suspend.c * Rewrote Nt Executive functions to use Kernel functions. * Moved Ps* Routines into ke/kthread.c and fixed them. The implementation was wrong in some aspects, especially the issue of the APC looping around the KeWaitXxx call and the fact that the routines excluded/ignored the FreezeCount. - ps/debug.c * Fixed completely broken implementation of Get/SetThreadContext. The old version crashed when called and did not work at all. Suspend Regression test now works. * Moved Context<->TrapFrame functions to ke/i386/ * Combined Set/GetThreadContext APCs into a single one, and used special context structure. - ps/query.c: * Moved Thread/Process Query/Set Routines here. - ps/tinfo.c: * Removed. - ps/kill.c * Removed complicated Process Termination semantics and useless Attach/Detach in favor for a much more lightweight function which performs the same tasks as before and actually works. TaskManager can now terminate foreign processes. * Rewrote Thread Reaping to use the HyperCritical Work Queue instead of manually controlled thread. This results in much less code as well as an increase in speed and less micro management. The reaper is PspReapRoutine. Closing CMD.EXE now works properly without requiring masks that were added as hacks to allow it. * Renamed PiTerminateProcessThreads to PspTerminateProcessThreads. Fixed it to work with new termination code. * Added PspDeleteProcess to handle Process Object deletion. Kills the CID Handle here as done by Hartmut. * Added PspDeletethread here. * Renamed and rewrote PsTerminateCurrentThread to PspExitThread. Used NT Implementation out- lined in Windows Internals, Chapter 13. Uses less locks, a more concise order of actions, actually parses the Termination Ports, handles Dbgk notification. Timers are now rundown, and Mutex rundown is in a dedicated Kernel function. Final termination handled by KeTerminate Thread as documented. * Renamed PsTerminateOtherThread to PspTerminateThreadByPointer and modified implementation to be compatible with the changes above. * Renamed and regrouped Process Termination into PspExitProcess. Also implemented as described above, and moved each subsystem specific termination helper into its own subsytem. * Improved NtTerminateProcess and added more debugging messages. * Improved NtTerminateThread and added check against System Thread and made it compatible with new implementation. * Corrected PsTerminateSystemThread now that we support System Threads. * Corrected NtRegisterThreadTerminatePort to use same structure name as on windows for the port, and added tag to pool allocation (documented in pooltag.txt) include/internal/*.h: * Defined Scheduler Functions and misc new functions or renamed functions. ke/apc.c: * Fixed critical bug where APCs were not delivered at all if the thread wastion and cancels any timers that are associated to a thread, as well as their APCs and DPCs. REGRESSIONS FOUND: NONE BUGS/REGRESSIOSN FIXED: * Thread/Get Set Context now works. * Suspend Regression test now works. * Task manager can now kill foreign processes, even hung ones (like it should). * ExitProcess/closing cmd.exe with the 'x' button now works correctly without hacks. KNOWN ISSUES: I left a bit of a mess in the headers and some stuff still needs to be moved into the right places. I just wanted to have this first part ready first, so that it won't get too big. svn path=/trunk/; revision=14174
2005-03-18 05:53:04 +00:00
}
Thread/Process Termination/Repeaing Rewrite + Fixes --------------------------------------------------- - ps/cid.c: * Moved CID Lookup functions here - ps/security.c: * Moved all security related functions here. Makes other files neater and security functions easier to locate. - ps/thread.c: * Moved most of the Thread Scheduling/Dispatching code that belongs in the Kernel to /ke and renamed functions from Ps to Ki. * Implemented PsIsSystemThread. * Removed Reaper Thread Init (now obsolete). * Renamed PiDeleteThread to PspDeleteThread. * Moved Thread State functions from tinfo.c to here. - ps/process.c: * Removed Query/Set Process functions and moved to ps/query.c * Renamed PiDeletePRocess to PspDeleteProcess * Removed obsoleted Process Termination functions, moved persistent one to kill.c - ps/create.c: * Moved the security APIs to security.c * Correctly implemented PsCreateSystemThread to actually create system threads. - ps/suspend.c * Rewrote Nt Executive functions to use Kernel functions. * Moved Ps* Routines into ke/kthread.c and fixed them. The implementation was wrong in some aspects, especially the issue of the APC looping around the KeWaitXxx call and the fact that the routines excluded/ignored the FreezeCount. - ps/debug.c * Fixed completely broken implementation of Get/SetThreadContext. The old version crashed when called and did not work at all. Suspend Regression test now works. * Moved Context<->TrapFrame functions to ke/i386/ * Combined Set/GetThreadContext APCs into a single one, and used special context structure. - ps/query.c: * Moved Thread/Process Query/Set Routines here. - ps/tinfo.c: * Removed. - ps/kill.c * Removed complicated Process Termination semantics and useless Attach/Detach in favor for a much more lightweight function which performs the same tasks as before and actually works. TaskManager can now terminate foreign processes. * Rewrote Thread Reaping to use the HyperCritical Work Queue instead of manually controlled thread. This results in much less code as well as an increase in speed and less micro management. The reaper is PspReapRoutine. Closing CMD.EXE now works properly without requiring masks that were added as hacks to allow it. * Renamed PiTerminateProcessThreads to PspTerminateProcessThreads. Fixed it to work with new termination code. * Added PspDeleteProcess to handle Process Object deletion. Kills the CID Handle here as done by Hartmut. * Added PspDeletethread here. * Renamed and rewrote PsTerminateCurrentThread to PspExitThread. Used NT Implementation out- lined in Windows Internals, Chapter 13. Uses less locks, a more concise order of actions, actually parses the Termination Ports, handles Dbgk notification. Timers are now rundown, and Mutex rundown is in a dedicated Kernel function. Final termination handled by KeTerminate Thread as documented. * Renamed PsTerminateOtherThread to PspTerminateThreadByPointer and modified implementation to be compatible with the changes above. * Renamed and regrouped Process Termination into PspExitProcess. Also implemented as described above, and moved each subsystem specific termination helper into its own subsytem. * Improved NtTerminateProcess and added more debugging messages. * Improved NtTerminateThread and added check against System Thread and made it compatible with new implementation. * Corrected PsTerminateSystemThread now that we support System Threads. * Corrected NtRegisterThreadTerminatePort to use same structure name as on windows for the port, and added tag to pool allocation (documented in pooltag.txt) include/internal/*.h: * Defined Scheduler Functions and misc new functions or renamed functions. ke/apc.c: * Fixed critical bug where APCs were not delivered at all if the thread wastion and cancels any timers that are associated to a thread, as well as their APCs and DPCs. REGRESSIONS FOUND: NONE BUGS/REGRESSIOSN FIXED: * Thread/Get Set Context now works. * Suspend Regression test now works. * Task manager can now kill foreign processes, even hung ones (like it should). * ExitProcess/closing cmd.exe with the 'x' button now works correctly without hacks. KNOWN ISSUES: I left a bit of a mess in the headers and some stuff still needs to be moved into the right places. I just wanted to have this first part ready first, so that it won't get too big. svn path=/trunk/; revision=14174
2005-03-18 05:53:04 +00:00
/* EOF */