RTEMS 4.6.2 On-Line Library

Explicit Synchronization

RTEMS Remote Debugger Server Specifications

7.6.2: Explicit Synchronization

This chapter will describe the synchronization variables that need to be implemented in order to sequence debug events in a way that is compatible with what GDB code expects. The root of the problem is that GDB code mainly expects that once a debug event has occurred on the debuggee, the entire debuggee is frozen and no other event will occur before the CONTINUE command is issued. This behavior is hard to achieve in our case as once we hit a breakpoint, only the task that hits the breakpoint will be asleep on a synchronization object. Other tasks may hit other breakpoints while we are waiting commands from GDB generating potential unexpected events. There is a solutions if RGDBSD itself use RTEMS threads to fix this problem by creating a task that loops forever at a priority superior to any debugged task but below RGDBSD task priorities. Unfortunately this will not work for the case we use the nano-kernel implementation and we think it is better to study synchronization problems now. We also expects that multi-thread debug support hardening in GDB will remove some event serializations requirements. Here is the list of synchronization variables we plan to use and their usage. They are all regular semaphores. They are not binary semaphores because the task that does V is not the task that has done the P.

WakeUpEventTask : used by exception handler code to wake up the EVENT_MNGT task by doing a V operation on this object. When target code is running normally the EVENT_MNGT task sleeps due to a P operation on this semaphore,
SerializeDebugEvent : used to serialize events in a way compatible to what GDB expects. Before doing a V operation on WakeUpEventTask, the exception handler does a P on this semaphore to be sure processing of another exception is not in progress. Upon reception of a CONTINUE command, the COMMAND_MNGT task will issue a V operation so that the exception code can wake up EVENT_MNGT task using the mechanism described above,
RestartFromException : (in fact one semaphore per task) used by exception handling code to put a faulty task to sleep once it has generated an exception by doing a P operation on this semaphore. In the case the exception was generated due to a breakpoint, GDB command will modify back the BREAKPOINT opcode to the original value before doing the CONTINUE command. This command will perform a V on this semaphore. In the case it is a real non restartable exception (faulty memory reference via invalid pointer for example), GDB will not allow to restart the program avoiding any loop. So not special analysis of cause of exception is foreseen as far as RGDBSD code is concerned,

RTEMS Remote Debugger Server Specifications