STRIDE 3.0.01xx: Difference between revisions

From STRIDE Wiki
Jump to navigation Jump to search
No edit summary
(SCR 8891 initial)
Line 36: Line 36:
* '''Memory Management'''
* '''Memory Management'''
In case of multi-process target, recommended approach is to use <u>Memory-Mapped Files</u> to implement memory segments.
In case of multi-process target, recommended approach is to use <u>Memory-Mapped Files</u> to implement memory segments.
  palMemSegmentOpen - Open/create memory segment  
  palMemSegmentOpen() - Open/create memory segment  
  palMemSegmentClose - Close memory segment
  palMemSegmentClose() - Close memory segment


* '''Named Mutexes'''
* '''Named Mutexes'''
  palMutexInit - Initialize a mutex object
  palMutexInit() - Initialize a mutex object
  palMutexDestroy - Destroy a mutex object
  palMutexDestroy() - Destroy a mutex object
  palMutexLock - Lock a mutex object
  palMutexLock() - Lock a mutex object
  palMutexUnlock - Unlock a mutex object
  palMutexUnlock() - Unlock a mutex object


* '''Task Synchronization'''
* '''Task Synchronization'''
  palGetThreadId - Get current thread Id
  palGetThreadId() - Get current thread Id
  palGetProcessId - Get current process Id
  palGetProcessId() - Get current process Id
  palSleep - Suspend the execution of the current thread
  palSleep() - Suspend the execution of the current thread


* '''Logging'''
* '''Logging'''
This is optional.
This is optional.
  palLog - Logging utility
  palLog() - Logging utility


===Updated Routines===
===Updated Routines===
Function signatures of palMemAlloc() & palMemFree() have been updated.
* Function signatures and functionalities of palWait() & palNotify() have been updated.
  palMemAlloc - now returns a void pointer
 
  palMemFree - now takes in a void pointer as input
  palWait() - now takes in an in/out pointer for events notification
  palNotify() - mail box Id has been updated to notify any event including mail box Ids.
 
* Function signatures of palMemAlloc() & palMemFree() have been updated.


===Removed / Old Routines===
===Removed / Old Routines===
Line 64: Line 67:
  palCriticalErr
  palCriticalErr


===Removed Module===
===Removed Modules===
  S2Mem - Memory module provided as part of PAL has been removed. If native OS does not support dynamic memory, Runtime's memory management module ''srMem'' can be used instead of ''S2Mem''.
  S2Mem - Memory module provided as part of PAL has been removed. If native OS does not support dynamic memory, Runtime's memory management module ''srMem'' can be used instead of ''S2Mem''.


Line 121: Line 124:
* Implement new routine palSleep().
* Implement new routine palSleep().


* Update synchronization routines, palCreateNID(), palDeleteNID(), palWait() & palNotify(), to support sharing of the synchronization objects among multiple processors if shared memory management is enabled.
* Update synchronization routines, palCreateNID(), palDeleteNID(), palWait() & palNotify(), to support event notifications and in case of multi-process target is enabled, sharing of the synchronization objects among multiple processors.


* If your PAL was using ''S2Mem'' routines for dynamic memory, you should use new Runtime module ''srMem''. Update palMemAlloc() to call _srMem_Allocate() and palMemFree() to call _srMem_Free().
* If your PAL uses ''S2Mem'' routines for dynamic memory, you should use new Runtime module ''srMem''. Update palMemAlloc() to call _srMem_Allocate() and palMemFree() to call _srMem_Free().


* Optionally, implement new routine palLog() as a logging utility for PAL and Runtime.
* Optionally, implement new routine palLog() as a logging utility for PAL and Runtime.
Line 129: Line 132:
==Runtime==
==Runtime==
* Edit '''srcfg.h''' file to configure the new memory management (srCFG_MEMORY_MANAGEMENT) and multi-process target (srCFG_MULTI_PROC_TARGET) settings. By default, memory management and multi-process target are disabled. If you enable multi-process target you also need to enable memory management and configure the sizes and the maximum number of memory blocks. New Runtime module ''srMem'' uses these configurations to allocate and manage memory segments.
* Edit '''srcfg.h''' file to configure the new memory management (srCFG_MEMORY_MANAGEMENT) and multi-process target (srCFG_MULTI_PROC_TARGET) settings. By default, memory management and multi-process target are disabled. If you enable multi-process target you also need to enable memory management and configure the sizes and the maximum number of memory blocks. New Runtime module ''srMem'' uses these configurations to allocate and manage memory segments.
* Edit '''srcfg.h''' file to configure the new target connection timeout setting (srCFG_CONNECTION_TIMEOUT).


==Host Apps using Target Runtime lib s2shostapptrt==
==Host Apps using Target Runtime lib s2shostapptrt==
* A new routine, '''HostAppSetMaster()''', has been added to s2shostapptrt library. This routine needs to be called from the application that should be starting the runtime thread, srThread, before calling HostAppRun(). Usually within a multi-process target, the ''master'' process or the ''daemon'' calls this routine.
* A new routine, '''HostAppSetMaster()''', has been added to s2shostapptrt library. This routine needs to be called from the application that should be starting the runtime thread, srThread, before calling HostAppRun(). Usually within a multi-process target, the ''master'' process or the ''daemon'' calls this routine.

Revision as of 23:38, 12 June 2008

Release Notes

This page documents the changes in STRIDE version 2.1.0301 (code name StoneSteps).

Please review this information before upgrading from an earlier version.

What's New

Build Tools

A set of new command line build tools has been implemented:

  • Stride compiler
  • Stride database binder
  • Stride instrumantation generator

These utilities has been ported for Windows and Linux. Using them would allow seamless integration with "make" like build environment.

Runtime/PAL

Based on customer requirements, in this release we have made significant changes to STRIDE Runtime and PAL.

Multi-Process Targets

New routines and support have been added to Runtime and PAL to support multi-process target. Changes include implementation of memory management and shared memory, protecting using named Mutex objects, synchronization of multiple processors and multiple threads, and usage of current process and thread IDs.

Logging (Optional)

Runtime and PAL use a logging routine that can optionally be implemented in PAL.

Change Details

PAL

Memory Management

To support multi-process target, memory management support for dynamic, configurable, and internal static memory has been implemented in Runtime. Runtime makes PAL calls to acquire memory segments and manages the memory according to runtime configurations set by user. In case of multi-process target is enabled, these memory will be shared among multiple applications, and PAL should implement shared memory.

Protection using Mutex

Runtime now uses named mutex objects to protect critical data from multiple and simultaneous accesses by multiple threads and, in case of multi-process target is enabled, by multiple applications. Runtime makes PAL calls to obtain and control mutex objects.

New Routines

  • Memory Management

In case of multi-process target, recommended approach is to use Memory-Mapped Files to implement memory segments.

palMemSegmentOpen() - Open/create memory segment 
palMemSegmentClose() - Close memory segment
  • Named Mutexes
palMutexInit() - Initialize a mutex object
palMutexDestroy() - Destroy a mutex object
palMutexLock() - Lock a mutex object
palMutexUnlock() - Unlock a mutex object
  • Task Synchronization
palGetThreadId() - Get current thread Id
palGetProcessId() - Get current process Id
palSleep() - Suspend the execution of the current thread
  • Logging

This is optional.

palLog() - Logging utility

Updated Routines

  • Function signatures and functionalities of palWait() & palNotify() have been updated.
palWait() - now takes in an in/out pointer for events notification
palNotify() - mail box Id has been updated to notify any event including mail box Ids.
  • Function signatures of palMemAlloc() & palMemFree() have been updated.

Removed / Old Routines

palProtect
palUnprotect
palCriticalErr

Removed Modules

S2Mem - Memory module provided as part of PAL has been removed. If native OS does not support dynamic memory, Runtime's memory management module srMem can be used instead of S2Mem.

Runtime

  • STRIDE Host Release 2.1.0301 is compatible with the Runtime Version 3.00.
  • To support multi-process target, significant changes have been made to runtime.
  • A new public routine, srUninit(), has been added to sr.h & srapi.c.
  • A new module, srMem, for Memory Management has been added. Its new routines, _srMem_Allocate() & _srMem_Free(), will be used in PAL's palMemAlloc() & palMemFree() in case of multi-process target is enabled.
  • Calls to palProtect() & palUnprotect() have been replaced by new routines _srProtect() & _srUnprotect(), in which the implementation uses its own mutex object that calls PAL's mutex routines.
  • The following Runtime files have been modified:
sr.h
srapi.c
srconn.c
srconn.h
srib.c
srib.h
sribtr.c
sribtr.h
srmsgptr.c
srmsgptr.h
srmsgque.c
srmsgque.h
srmsgrt.c
srmsgrt.h
srmsgsub.c
srmsgsub.h
srstid.c
srstid.h
srsuid.c
srsuid.h
srtime.c
srtime.h
  • The following Runtime files have been added:
srmem.c
srmem.h

Migration to 2.1.0301

Recommended steps for migration from a previous version:

PAL

Please refer to Platform Abstraction Layer (PAL) Specification for detailed information on new APIs. You may also refer to Linux implementation of PAL, which has been updated with new routines.

  • Implement new shared memory routines, palMemSegmentOpen() & palMemSegmentClose(). In case of multi-process target, these routines should implement shared memory using memory-mapped files.
  • Runtime now can manage dynamic, configurable, and internal static memory. In case of multi-process target, update palMemAlloc() to simply call _srMem_Allocate() and palMemFree() to simply call _srMem_Free(). In case of single-process target, you can simply allocate dynamic memory according to your Operating System. If you are upgrading, you can simply leave the current implementation as it is.
  • Implement new mutex routines, palMutexInit(), palMutexDestroy(), palMutexLock() & palMutexUnlock, to handle any named mutex objects. It is recommended to implement a separate code for shared memory case if it is necessary.
  • Remove old routines palProtect(), palUnprotect() and palCriticalErr().
  • Implement new routines palGetThreadId() and palGetProcessId().
  • Implement new routine palSleep().
  • Update synchronization routines, palCreateNID(), palDeleteNID(), palWait() & palNotify(), to support event notifications and in case of multi-process target is enabled, sharing of the synchronization objects among multiple processors.
  • If your PAL uses S2Mem routines for dynamic memory, you should use new Runtime module srMem. Update palMemAlloc() to call _srMem_Allocate() and palMemFree() to call _srMem_Free().
  • Optionally, implement new routine palLog() as a logging utility for PAL and Runtime.

Runtime

  • Edit srcfg.h file to configure the new memory management (srCFG_MEMORY_MANAGEMENT) and multi-process target (srCFG_MULTI_PROC_TARGET) settings. By default, memory management and multi-process target are disabled. If you enable multi-process target you also need to enable memory management and configure the sizes and the maximum number of memory blocks. New Runtime module srMem uses these configurations to allocate and manage memory segments.
  • Edit srcfg.h file to configure the new target connection timeout setting (srCFG_CONNECTION_TIMEOUT).

Host Apps using Target Runtime lib s2shostapptrt

  • A new routine, HostAppSetMaster(), has been added to s2shostapptrt library. This routine needs to be called from the application that should be starting the runtime thread, srThread, before calling HostAppRun(). Usually within a multi-process target, the master process or the daemon calls this routine.