According to the C11 and C++11 memory models only a read-modify-write
operation guarantees that we read the last value written in modification
order. Avoid the sequential consistent thread fence and instead use the
inter-processor interrupt to set the thread dispatch necessary
indicator.
This mutex implementation uses a thread priority queue with a simple
priority inheritance mechanism (similar to the object based mutexes).
The storage space must be supplied by the user (16 bytes on 32-bit
targets).
Move the storage for the thread queue heads to the threads. Each thread
provides a set of thread queue heads allocated from a dedicated memory
pool. In case a thread blocks on a queue, then it lends its heads to
the queue. In case the thread unblocks, then it takes a free set of
threads from the queue. Since a thread can block on at most one queue
this works. This mechanism is used in FreeBSD. The motivation for this
change is to reduce the memory demands of the synchronization objects.
On a 32-bit uni-processor configuration the Thread_queue_Control size is
now 8 bytes, compared to 64 bytes in RTEMS 4.10 (other changes reduced
the size as well).
Separate the thread queue heads and lock from the operations. This
enables the support for light weight objects which only support one
queuing discipline.
Add rtems_interrupt_local_disable|enable() as suggested by Pavel Pisa to
emphasize that interrupts are only disabled on the current processor.
Do not define the rtems_interrupt_disable|enable|flash() macros and
functions on SMP configurations since they don't ensure system wide
mutual exclusion.
This was obsolete and broken based upon recent time keeping changes.
Thie build option was previously enabled by adding
USE_TICKS_FOR_STATISTICS=1 to the configure command line.
This propagated into the code as preprocessor conditionals
using the __RTEMS_USE_TICKS_FOR_STATISTICS__ conditional.
Add an assert to ensure that the watchdog is the proper state for a
_Watchdog_Initialize(). This helps to detect invalid initializations
which may lead to a corrupt watchdog chain.
It is wrong to restore the floating point context here. The
_Context_Switch() ends up in _Thread_Handler() which will call
_Thread_Restore_fp(). In _Thread_Do_dispatch() the FP restore is after
the context switch.
In case the zlib compression was used, then the librtemscpu.a depended
on libz.a. To avoid a GCC patch or complicated link flags move the
JFFS2 support into a separate library to use a simple "-ljffs2 -lz" to
link the executable.
