Subversion Repositories shark

Rev

Blame | Last modification | View Log | RSS feed

%----------------------------------------------------------------------------
\chapter{Kernel Internals}
%----------------------------------------------------------------------------

Since the Generic Kernel does not implement any memory protection,
all the Modules have access to the internal data structures used
for scheduling. Although the Modules could modify these data
structures, typically they do not need to do so.

%----------------------------------------------------------------------------
\section{Kernel types}
%----------------------------------------------------------------------------

The Generic Kernel defines a set of primitive data types, that are
briefly described below:

\begin{description}
\item [PID]This type is used to contain a task index. It is an
integer pointing to an entry of the task descriptor table. This
type can have values in the range {[}0...MAXPROC-1{]}, plus an
invalid value, NIL (that is, -1).

\item [IQUEUE]This type is used to implement the task queues typically used in
scheduling operations. For more information see Section \ref{KernSupport_Code}.

\item [TASK]This type is simply a
redefinition of the \texttt{void {*}} type and it can be used for
readiness ad the returned value in the task function declaration.

\item [LEVEL,~RLEVEL]They are used to index the Module descriptor
Tables. They are integers and they point to a particular entry.
The values of these types are
{[}0...\texttt{MAX\_SCHED\_LEVEL-1}{]} for the \texttt{LEVEL} type
and {[}0...\texttt{MAX\_RES\_LEVEL}{]} for the \texttt{RLEVEL}
type.

\item [bandwidth\_t]This type is used to store a real number
in the range {[}0,1{]}. It is a 32 bit unsigned integer and its
value is interpreted as $\frac{x}{MAX\_ BANDWIDTH}$.

\item
[task\_key\_t]This is an integer type and it is used as an index
for task specific data (similar to POSIX's Thread Specific Data).
\end{description
}

%----------------------------------------------------------------------------
\section{Descriptors}
%----------------------------------------------------------------------------

In this section we present all the descriptors defined in the
Generic Kernel. If not specified, these data structures are
defined in the \texttt{include/kernel/descr.h} file and they are
used (usually as arrays) in the \texttt{kernel/kern.c
} file.

%----------------------------------------------------------------------------
\subsection{Task Descriptor}
\label{KernSupport_Task_descriptor
}
%----------------------------------------------------------------------------

The \texttt{proc\_des} structures used to define task control blocks
\footnote{The name of this structure is derived from the previous versions of
Hartik. However, it is a \textit{task} descriptor and not a \textit{process}
descriptor.}. In the following paragraphs the fields of the task descriptor are
described \footnote{The following fields have been removed from the structure in
the last releases: request\_time, priority, timespec\_priority, prev, next.}.

\begin{description}
\item [DWORD~task\_ID]Progressive number assigned to a task at its
creation.

\item [LEVEL~task\_level]This field points to the Module
that owns the task. The Generic kernel uses this field to redirect
the calls to the Module owning the task.

\item
[CONTEXT~context]This field contains an index in the OS Lib
context array that handles the context of a task.

\item
[BYTE~{*}stack]This field is a pointer to the memory used as a
task stack.

\item [TASK~(*body)()]This field is the pointer to
the first instruction of a task body, and it is used at
initialization time.

\item [char~name[MAX\_TASKNAME]]This is a
symbolic name, whose length is defined by the
\texttt{MAX\_TASKNAME
} constant.

\item [WORD~status]This is the
task status.

\item [WORD~pclass]This field is the Class code of
the Task Model used during task creation. It can be used by
Modules to know the typology of the task (useful if the task
supports many types of Models; see
%
% Tool: such section does not exists.
%
%Section
%\ref{SchedModules_LevelCalls},
\texttt{level\_accept\_task\_model}
%function)
.

\item [WORD~group]This is the task group. The value 0
is used to index a single task. This field is not used as a GID in
classical Unix systems but is used to kill or activate a group of
tasks in an atomic way.

\item [WORD~stacksize]Stack dimension (in
bytes).

\item [DWORD~control]Task Status flags. A description of
the bits in this field is reported in
Section\ref{Modelli_TASK_MODEL}.

\item
[int~frozen\_activations]Number of frozen activations. Useful only
if the \texttt{FREEZE\_ACTIVATION} flag of the \texttt{control}
field is active (see Section \ref{Kernel_Attivazione}).

\item
[int~sigmask]This is the blocked signal task mask (see
\texttt{kernel/signal.c} and also \cite{POSIX1003.1}).

\item
[int~sigpending]This is the pending signal task mask (see
\texttt{kernel/signal.c} and also \cite{POSIX1003.1}).

\item
[int~sigwaiting]This is the mask of the signals on which a task is
waiting, blocked on a \texttt{sigwait} primitive (or similar).

\item [int~avail\_time]This field contains the remaining
computation time of a task (see Section
\ref{KernSupport_Capacita}).

\item [PID~shadow]This is the shadow
pointer (see Section \ref{oview_Shadows}).

\item
[struct~\_task\_handler\_rec~{*}cleanup\_stack]Pointer to the
first element of the cleanup stack.

\item [int~errnumber]This is
an error number local to the running task. The symbol
\texttt{errno} defined into the C standard is implemented as a
macro that refers to this field of the running task.

\item
[TIME~jet\_table{[}JET\_TABLE\_DIM{]}]This table contains the
computation times consumed by the last JET\_TABLE\_DIM instances
of the task. The table is handled as a circular buffer.

\item
[int~jet\_tvalid]This is the number of valid items in the
\texttt{jet\_table} table.

\item [int~jet\_curr]This is the
current element in the \texttt{jet\_table} table.

\item
[TIME~jet\_max]This field contains the maximum time (in
microseconds) consumed by a task, among all its instances.

\item
[TIME~jet\_sum]This field accumulates the execution times (in
microseconds) of all task instances from the start of the task (or
from the last call to \texttt{jet\_delstat}). This field, together
with the \texttt{jet\_n} field, is used to compute the mean
execution time of a task instance.

\item [TIME~jet\_n]This field contains the number of instances involved in the
computation of \texttt{jet\_sum}.

\item [PID~waiting\_for\_me]This field contains
the identifier of the task which is currently blocked for a
\texttt{task\_join} on the current task. The field value is NIL if
no task is blocked for a join on it.

\item
[void~{*}return\_value]This is the value returned by a task when
it dies. This value is memorized if the task is joinable, waiting
for someone to synchronize on it with a join.

\item
[void~{*}keys{[}PTHREAD\_KEYS\_MAX{]}]This array contains the task
specific data.

\item
[struct~condition\_struct~{*}cond\_waiting]This field is NULL or
it points to the condition variable on which the task is
blocked \footnote{This field is present because if a task is killed during
a block on a condition wait, the POSIX standard requires that the
task reacquires the mutex linked to that condition before it
dies.}.

\item [int~delay\_timer]This field is used in the
implementation of the blocking primitives with timeout. Usually
this field contains the OS Lib index of the event created for the
wake-up.

\item [int~wcet]This field can be used by Modules to
store some temporal information. It is generally used in
conjunction with the \texttt{avail\_time} field.
\end{description
}

%----------------------------------------------------------------------------
\subsection{Level descriptor}
\label{KernSupport_Level_Descriptor
}
%----------------------------------------------------------------------------

In this Section we describe the fields contained in the
\texttt{level\_des} structure, that implements the Scheduling
Module Descriptor. These Modules will fill the proposed interface,
defining also all the interface functions. If a Module does not
implement an interface function, it redefines that function in a
way that it will raise an exception if called. The fields defined
for the \texttt{level\_des} structure are the following:

\begin{description}
\item [Private~Functions,~Public~Functions]The rest of the interface is made of
a set of function pointers that implements the \emph{virtual} functions of the
Scheduling Modules. That functions are described in the Chapter
\ref{CapSchedulingModules}. All that functions have as first parameter a LEVEL
field that can be used to find the Module Descriptor and then the private
structures of that Modules \footnote{The behaviour of the LEVEL parameter passed
to that function is the same of the hidden paramether \emph{this} of many Object
Oriented languages.}.
\end{description
}

%----------------------------------------------------------------------------
\subsection{Resource module descriptor}
\label{KernSupport_Descrittore_GestioneRisorse
}
%----------------------------------------------------------------------------

In this Section we describe the fields contained in the
\texttt{resource\_des} structure, which implements the Resource
Module Descriptor. These Modules will fill the proposed interface,
defining all the required functions. If a Module does not
implement an interface function, it redefines that function in a
way that it will raise an exception if it is called. The fields
defined for the \texttt{resource\_des} structure are the
following:

\begin{description}
\item [char~res\_name{[}MAX\_MODULENAME{]}]This filed contains a
symbolic Name of the Module, for statistical purposes. The file
\texttt{include/modules/codes.h} contains the name of the Modules
distributed with the Kernel. \item [WORD~res\_code]This field
contains a numeric identifier that identifies a Module in an
unique way. The codes of the Modules distributed with the kernel
are written in the file \texttt{include/modules/codes.h}. \item
[BYTE~res\_version]This is a version number for the Module. The
version numbers of the Modules distributed with the Kernel are
reported in the file \texttt{include/modules/codes.h}. \item
[int~rtype]This field is used to identify the extended interface
implemented by a Resource Module. This field is necessary to
implement some primitives that use a particular extended
interface, that need to know which of the resource modules
registered in the kernel have a particular extension. \item
[Resource~Calls]The rest of the interface is made of a set of
function pointers that implement the \emph{virtual} functions of
the Resource Modules. These functions are described in Chapter
\ref{CapModuliRisorsa}. All these functions have as first
parameter a RLEVEL field that can be used as the similar parameter
of the Scheduling Module virtual functions.
\end{description
}

%----------------------------------------------------------------------------
\section{System states}
%----------------------------------------------------------------------------

As mentioned in Section \ref{oview_Status} the concept of task
state has a local meaning, and it is stored in the \texttt{status}
field of its task descriptor.

The values that the status field can have are divided in four
intervals, described in Table \ref{KernSupport_Tab_StatusCodes}.

\begin{table}
\begin{center}
\begin{tabular}{|p{4cm}|p{9cm}|}
\hline
Interval reserved to & Codes\\ \hline \hline

Generic Kernel & {[}0...\texttt{MODULE\_STATUS\_BASE}-1{]} \\ \hline
Scheduling Algorithm \foreignlanguage{english}{} & {[}\texttt{MODULE\_STATUS\_BASE}
...\texttt{APER\_STATUS\_BASE}-1{
]}\\ \hline

Aperiodic Servers \foreignlanguage{english}{} & {[}\texttt{APER\_STATUS\_BASE}
...\texttt{LIB\_STATUS\_BASE}-1{
]}\\ \hline

Others (Resource Handling, libraries, etc.) \foreignlanguage{english}{} &
\texttt{>= LIB\_STATUS\_BASE}\\ \hline
\end{tabular}
\end{center}
\caption{Partitioning of the codes for the status field of the task descriptor.}
\label{KernSupport_Tab_StatusCodes}
\end{table
}

Each Module Should use the codes in the correct range of values.

This approach allows to handle many Module configurations. The
fact that two Scheduling Modules use the same status code for
different meanings is not a problem, because the codes are used
internally to the Modules.

%----------------------------------------------------------------------------
\section{Kernel Global Variables}
\label{KernSupport_Variabili
}
%----------------------------------------------------------------------------

The following global variables are defined in the Generic Kernel
and can be used by the Modules.

\begin{description}

\item [proc\_des~proc\_table{[}{]}]This variable is the task descriptor array.
Not all the entries of this array are used. The descriptor really allocated are
handled by the \texttt{task\_create} primitive.

\item [level\_des~{*}level\_table{[}{]}]This variable is the array that
memorizes a set of \emph{pointers} to the Scheduling Modules
Descriptors. When a Module registers itself into the Kernel, an
entry of this array is allocated, and it points to an
\emph{extension} of the \texttt{level\_des} descriptor type. In an
object oriented interpretation it can be thought as a polymorphic
Module array (see Section \ref{Modelli_MetodologiaOOP}).

\item [resource\_des~{*}resource\_table{[}{]}]This variable memorizes an
array of pointers to extension of the resource\_des structure. For
this variable the same comments for the level\_table apply.

\item [PID~exec]This variable contains the task returned by the
Scheduling Module scheduling operation. Typically this variable is
used only into the \texttt{scheduler()} function of the Generic
Kernel. This variable cannot be modified, it can only be read and
checked with the field \texttt{exec\_shadow}. This variable has
the value \texttt{NIL} (-1) when a primitive blocks the running
task, and the scheduler is not yet called. This variable
\emph{does not} point to the running task. Use
\texttt{exec\_shadow} instead!

\item [PID~exec\_shadow]This variable points to the running task. It can be
different from the value contained into \texttt{exec} because of the Shadow
mechanism (see Section \ref{oview_Shadows}). This variable cannot be modified,
it can only be read and check with the field \texttt{exec}. This variable has
the value \texttt{NIL} (-1) when a primitive blocks the running task, and the
scheduler is not yet called. When a reference to the running task is needed,
this variable has to be used.

\item [int~cap\_timer]This variable is different from -1 only if the Generic
Kernel or a Module created a capacity event. This variable can be used by the
Scheduling Modules that do not utilize the \texttt{CONTROL\_CAP flag} for their
tasks (see Section \ref{KernSupport_Capacita}).

\item [struct~timespec~schedule\_time]This variable contains the system time in
which the scheduler was called the last time. The first thing the Kernel
function \texttt{scheduler()} does is the setting of that variable. This
variable is also used as finish time for an execution time interval.

\item [struct~timespec~cap\_lasttime]This variable is a copy of the
\texttt{schedule\_time} variable, done at each call of the \texttt{scheduler()}
function. The value of that variable is used as the start time of the last
interval executed by a task. It is used to implement the time accounting.

\item [DWORD~res\_levels]This variable is the number of resource
module level descriptors allocated. It is modified only by the
\texttt{resource\_alloc\_descriptor
} function
%
% Tool: such section does not exists.
%
% (see Section \ref{SchedModules_RegFunctions})
.

\item [int~task\_counter]This variable is a variable that counts the number of
User tasks actually present into the system (see Section
\ref{Kernel_Inizializzazione}).

\item [int~system\_counter]This
variable is a variable that counts the number of System tasks
actually present into the system (see Section
\ref{Kernel_Inizializzazione
}).

%
% Tool: no more sys_end nor sys_abort. Check this in the code!
%
% \item [int~mustexit]This variable
% is used by the system primitives \texttt{sys\_end} and
% \texttt{sys\_abort} to block the context changes into the event
% handlers.

\item [int~calling\_runlevel\_func]This variable is a
flag. It is set to 1 when the system executes some system
functions (registered through the \texttt{sys\_atrunlevel()}
function). This flag influence the \texttt{task\_activate}
primitive, because that primitive has to do different stuffs
depending on the value of that variable (see the code of the
primitive in the file \texttt{kernel/activate.c}).

\item
[IQUEUE~freedesc]This variable is the free task descriptor queue.
It is handled by the Kernel and must be used by a Scheduling
Module into the call \texttt{task\_end} call to free a task
descriptor of a task terminated correctly.

\item
[TIME~sys\_tick]If the OS Lib is initialized with the \textit{one
shot timer} the variable contains the system tick (in
microseconds). Otherwise, the one-shot timer is used, and this
variable has a value of \texttt{0}.
\end{description
}

%----------------------------------------------------------------------------
\section{Temporal protection}
\label{KernSupport_Capacita
}
%----------------------------------------------------------------------------

The Generic Kernel supports task temporal protection through a set
of procedures and data structures that may be used by the
Scheduling Modules. The generic Kernel does force to use its own
functions, so a Module can define itself a policy to ensure
temporal protection. In this Section the proposed functions to
implement the temporal protection are described.

The Generic Kernel handles the time capacity of the tasks in the
system using the creation and the deletion of a specific OS Lib
event that simply reschedules the system. The end of the capacity
is seen by the Modules as a normal epilogue done in the case of a
preemption.

The timer event is created when a task is dispatched, and it is
deleted by all the primitives that may cause preemption. In an
instant there is at least one capacity event pending, that is the
event of the running task.

The Generic Kernel defines for each task the bit
\texttt{CONTROL\_CAP} that is memorized into the field
\texttt{control} of the task descriptor. That bit is set by the
Scheduling Modules to require for a task the creation of the
capacity event at dispatch time.

The Scheduling Modules have also to check the capacity exhaustion
each time a task is descheduled.

Capacities are handled using the following variables provided by
the Kernel:

\begin{itemize}

\item The \texttt{wcet} field of the task descriptor usually memorizes a
characteristic time for a task (i.e., it is used to store a worst-case execution
time or a mean execution time). It is included in the task descriptor to avoid
the allocation of such an information for all the modules. It is \emph{not} used
by the Generic Kernel.

\item The \texttt{avail\_time} field of the task descriptor is used to handle
the available execution time for a task. The capacity event generated by the
Generic Kernel at dispatch time has an activation time that is equal to
\texttt{schedule\_time + avail\_time}. The field is updated by the generic
kernel each time a task is descheduled, decrementing the slice just executed by
the task.

\item The \texttt{cap\_timer} field is used by the Generic Kernel to know
whether a capacity event is or not pending. If the field is equal to -1, there
are no events, otherwise the value is the index of the capacity event. If the
field is not equal to -1 the Generic kernel will delete automatically the event
each time the system is rescheduled. The Generic Kernel does not check who has
created the event pointed by \texttt{cap\_timer}, so a Scheduling Module that
not use the flag \texttt{CONTROL\_CAP} can create a capacity event and put his
index into \texttt{cap\_timer}, avoiding the removal of the event directly in
the code of the Module.

\item The \texttt{cap\_lasttime} field contains the time of the previous
reschedule of the system. The accounting of the computation times is done from
the difference between \texttt{schedule\_time} and \texttt{cap\_lasttime}. The
time used by the Kernel primitives is accounted in the running task.

\end{itemize}

The functions called by the capacity event have only to reschedule
the system. The Scheduling Modules that have to generate
themselves the capacity event can store their events in the
\texttt{cap\_timer} filed if they want an event to be removed at
each preemption. The events can call the Generic Kernel Function
called \texttt{capacity\_timer
} (it does not accept any
parameter).

%----------------------------------------------------------------------------
\subsection{Negative capacities}
%----------------------------------------------------------------------------

The contents of the \texttt{avail\_time} field can be negative.
This fact, theoretically impossible, can happen for two reasons:

\begin{itemize}
\item first, the OS LIb event handler can not guarantee the
delivery of a capacity event at exact time (a small delay of some
microseconds may occur). So, a capacity event may be delivered
after the capacity is exhausted, so that the task capacity becomes
negative. We can suppose that negatives values are smaller than
the usual time for the tasks; \item second, because the shadow
mechanism can execute a task that has finished his budget; that
task will not be descheduled until the shadow of the higher
priority task points to it. This behavior is correct because we
want that the blocking task will ends its critical section to
limit the blocking time of the higher priority tasks. This
approach is supported by some theoretical results (see
\cite{Gha95}).
\end{itemize}

Scheduling modules can use two strategies to cope with the problem
given by negative capacities.

The first solution is to recharges the capacity with an amount
equal to $Q_{s}$; if the capacity is still negative, the task will
{}``loose a turn''. This solution is good for algorithms that
provide a periodic replenishment (like for example Round Robin,
Polling Server or Deferrable Server) but not for algorithms like
Sporadic Server or Constant Bandwidth Server, where the
replenishment is due to the capacity usage (in these algorithms
the server would not become active).

The second solution is to replenish the budget up to $Q_{s}$; this
solution is good for all types of algorithms, but we have to
consider that the server utilization factor becomes
$\frac{Q_{s}+\Delta}{T_{s}}$
.

%----------------------------------------------------------------------------
\section{Utility functions}
%----------------------------------------------------------------------------

This section contains a set of utility functions that can be used
to simplify the writing of the Modules. Often these functions are
simple redefinitions of the OS Lib functions. Use these functions
instead of those provided by the OS Lib.

%----------------------------------------------------------------------------
\subsection{Event Handling}
\label{KernSupport_GestioneEventi
}
%----------------------------------------------------------------------------

The Generic Kernel uses the Generic kernel event handling provided
by the OS Lib. The following functions should be used (see
\texttt{include/kernel/func.h}):

\begin{itemize}
\item \texttt{kern\_event\_post} can be used to post an event. The
redefinition adds a check on the index, raising an exception if
the OS Lib event queue is full. \item \texttt{kern\_event\_delete}
can be used to remove an OSLIB event. The unique parameter of the
funcion is the eveni ID returned by the kern\_event\_post
function. \item \texttt{event\_need\_reschedule}, with no
parameters. It has to be called into the event handlers that have
to reschedule the system. The Modules never have to call the
schedule function into event handlers.
\end{itemize
}

%----------------------------------------------------------------------------
\subsection{Exception Handling}
%----------------------------------------------------------------------------

The Generic Kernel exception handling is based on the real-time
signal interface of the POSIX standard. In particular, the
exceptions are implemented with the signal number 9, SIGHEXC, that
typically writes a message to the console and ends the system.

The function to use to raise an exception is: \begin{center}
\fbox{\tt{ void kern\_raise(int n, PID p); }} \end{center}

As a result of this function call, a real-time signal is enqueued
into the system and the n parameter is passed into the field
\texttt{si\_value} of the structure \texttt{siginfo\_t
} passed as
a parameter to the signal handler.

%----------------------------------------------------------------------------
\subsection{Memory Management}
\label{KernSupport_GestioneMemoria
}
%----------------------------------------------------------------------------

The Generic kernel provides a memory allocator based on Flux
OS-Kit LMM\cite{Bry97}. The LMM is in the \texttt{kernel/mem}
directory; the functions defined in this subsection are in the
file \texttt{kernel/mem.c}. These functions must be called with
interrupt disabled.

The allocator divides the memory into three regions that can be
specified with some \#defines into the flag parameter of some
functions:

\begin{itemize}
\item Memory addresses below 1 Mb (flag field set to
\texttt{MEMORY\_UNDER\_1M}) \item Memory addresses from 1 Mb to 16
Mb (flag field set to \texttt{MEMORY\_FROM\_1M\_TO\_16M}) \item
Memory addresses below 16 Mb (flag field set to
\texttt{MEMORY\_UNDER\_16M}) \item Memory addresses over 16 Mb
(flag field set to \texttt{0})
\end{itemize}

Memory management functions are described below (see also Figure
\ref{KernSupport_mem_functions}):

\begin{figure}
\begin{center}
\begin{minipage}{6cm}
\begin{verbatim}
void *kern_alloc(DWORD s);
void *kern_alloc_aligned(size_t size, lmm_flags_t flags,
      int align_bits, DWORD align_ofs);
void *kern_alloc_gen(size_t size, lmm_flags_t flags,
      int align_bits, DWORD align_ofs,
      DWORD bounds_min, DWORD bounds_max);
void kern_free(void *block, size_t size);
void *kern_alloc_page(lmm_flags_t flags);
void kern_free_page(void *block);
void *DOS_alloc(DWORD size);
void DOS_free(void *ptr, DWORD size);
\end{verbatim}
\end{minipage}
\end{center}
\label{KernSupport_mem_functions}
\caption{Memory allocation functions.}
\end{figure}

\begin{description}
\item [\texttt{kern\_alloc}]This function allocates a memory block
of s bytes returning a \texttt{void~{*}}. All the addresses given
are supposed to be good for the block. The function returns
\texttt{NULL} if there is not a free memory block with the
required size. \item [\texttt{kern\_alloc\_aligned}]This functions
allocates an aligned memory block. The first \texttt{align\_bits}
of the block must be equal to the lowest \texttt{align\_bits} of
\texttt{align\_ofs}. In other words, \texttt{align\_bits}
specifies an alignment as a power of 2, whereas
\texttt{align\_ofs} is a natural offset. The function returns
\texttt{NULL} if there is not a block with the required
characteristics. \item [\texttt{kern\_alloc\_gen}]This function
allocates an aligned memory block in a way similar to
\texttt{kern\_alloc\_aligned}, with the additional condition that
the block must be found within the addresses \texttt{bounds\_min}
and \texttt{bounds\_max}. \item [\texttt{kern\_free}]This function
will free a block allocated with one of the previous functions.
Note that this function requires the dimension of the block to
allocate. \item [\texttt{kern\_alloc\_page}]This function
allocates an aligned page of memory (4 Kb). \item
[\texttt{kern\_free\_page}]This function frees a page allocated
with the \texttt{kern\_alloc\_page} function. \item
[\texttt{DOS\_alloc,~DOS\_free}]These functions allocate memory
under the first Mb.
\end{description
}

%----------------------------------------------------------------------------
\subsection{Context switch}
%----------------------------------------------------------------------------

The creation and the deletion of a context in the Generic Kernel
is made using the OS Lib functions, properly renamed into
\texttt{kern\_context\_create} and \texttt{kern\_context\_delete}.

To change context, the Generic Kernel provides two functions,
\texttt{kern\_context\_save} and \texttt{kern\_context\_load}.
First function can be used to start a kernel primitive (disabling
interrupts, whereas the second can be used to enable the
interrupts, change the context using the OS Lib functions,
dispatch pending signals and test for asynchronous cancellation.
Note that currently the kernel primitives run on the same stack of
the caller thread, and the primitives are simply called disabling
the interrupts.

These functions must never be called into the standard interface
of the scheduling modules. When a OS Lib event need to reschedule
the system because there is a preemption, the function
\texttt{event\_need\_reschedule} must be called instead.

The functions described in this paragraph are into the
\texttt{include/kernel/func.h
} file.

%----------------------------------------------------------------------------
\subsection{Queues, arrays and pointers}
\label{KernSupport_Code
}
%----------------------------------------------------------------------------

To simplify the use of the internal Kernel references, the
following approach is used:

\begin{itemize}
\item Whenever possible, the descriptor arrays are statically
allocated. \item Each descriptor is identified by an integer which
represents the index in the descriptor array (e.g. PID, LEVEL,
RLEVEL).
\end{itemize}

To handle task queues, the Generic Kernel provides some utility
functions which can speed up the Module writing; however a module
can use its own functions to enqueue tasks.

The implemented functions do not use any fields of the task
descriptor; they are implemented with a double linked list. The
prototypes of the functions for queue management are shown in
Figure \ref{KernSupport_queue_functions}. They are into the file
\texttt{kernel/iqueue.c}, and \texttt{include/kernel/iqueue.h},
described below.

\begin{figure}
\texttt{\#define IQUEUE\_NO\_PRIORITY 1}
\texttt{\#define IQUEUE\_NO\_TIMESPEC 2}
\texttt{struct IQUEUE\_shared \{ }
\texttt{~~PID prev{[}MAX\_PROC{]}; }
\texttt{~~PID next{[}MAX\_PROC{]}; }
\texttt{~~struct timespec {*}timespec\_priority; }
\texttt{~~DWORD {*}priority; }
\texttt{\};}
\texttt{typedef struct \{ }
\texttt{~~PID first; }
\texttt{~~PID last; }
\texttt{~~struct IQUEUE\_shared {*}s; }
\texttt{\} IQUEUE; }
\texttt{~}
\texttt{void iq\_init (IQUEUE {*}q, IQUEUE {*}share, int flags);}
\texttt{void iq\_priority\_insert (PID p, IQUEUE {*}q); }
\texttt{void iq\_timespec\_insert (PID p, IQUEUE {*}q); }
\texttt{void iq\_insertfirst (PID p, IQUEUE {*}q); }
\texttt{void iq\_insertlast (PID p, IQUEUE {*}q);}
\texttt{void iq\_extract (PID p, IQUEUE {*}q); }
\texttt{PID iq\_getfirst ( IQUEUE {*}q); }
\texttt{PID iq\_getlast ( IQUEUE {*}q); }
\texttt{PID iq\_query\_first(IQUEUE {*}q);}
\texttt{PID iq\_query\_last(IQUEUE {*}q);}
\texttt{struct timespec {*}iq\_query\_timespec(PID p, IQUEUE
{*}q);}
\texttt{DWORD {*}iq\_query\_priority (PID p, IQUEUE {*}q);}
\texttt{PID iq\_query\_next (PID p, IQUEUE {*}q);}
\texttt{PID iq\_query\_prev (PID p, IQUEUE {*}q);}
\texttt{int iq\_isempty (IQUEUE {*}q);}
\caption{Prototypes of the task queue handling functions.}
\label{KernSupport_queue_functions}
\end{figure}


Basically, an IQUEUE has an \char`\"{}I\char`\"{}nternal prev/next
structure, that may be shared between one or more queue. Of
course, the user MUST guarantee that the same task will not be
inserted in two IQUEUEs that share the same prev/next buffer.

Internal queue initialization:

\begin{description}
\item [share~=~\&x]the internal data structure of the IQUEUE x is
used to enqueue the tasks. \item [\texttt{share~=~NULL}]an
internal data structure to handle prev/next pairs is dynamically
allocated (The amount of memory that is allocated can be reduced
using the flags). \item [flags]can be used to reduce the memory
usage of an IQUEUE when \texttt{share=NULL}:

\begin{description}
\item [\texttt{IQUEUE\_NO\_PRIORITY}]the iqueue do not provide
internally a priority field \item
[\texttt{IQUEUE\_NO\_TIMESPEC}]the iqueue do not provide
internally a timespec field
\end{description}

note that, if these flags are used, the corresponding insert
functions will not work!. The default value for the flags is, of
course, 0.

\end{description}
The queue insertion is made by the following functions:

\begin{description}
\item [iq\_insert]insertion based on the \texttt{priority} field.
\item [iq\_timespec\_insert]same as above but use the
\texttt{timespec\_priority} field \item [iq\_insertfirst]insert in
the first position of the queue \item [iq\_insertlast]insert in
the last position of the queue
\end{description}

The queue extraction functions: basically extracts a task p from
the queue q. There are three versions of the function:

\begin{description}
\item [iq\_extract]extracts given a task p (that must be inserted
in the queue); \item [iq\_getfirst]extracts the first task in the
queue, NIL if the queue is empty; \item [iq\_getlast]extracts the
last task in the queue, NIL if the queue is empty;
\end{description
}

Seven queue query functions are also provided. The first two
functions (query first/last) return the first and the last task in
the queue, NIL if the queue is empty. The second two functions
(query priority/timespec) can be used to get/set the priority or
the timespec field used when queuing. The third two functions
(query next/prev) can be used to scan the queue elements. The last
function can be used to test if a queue is empty.

%----------------------------------------------------------------------------
\subsection{Initialization functions}
%----------------------------------------------------------------------------

The Generic Kernel supports the specification of the functions to
be called at system initialization and termination (see Section
\ref{Kernel_Inizializzazione}). These functions can be registered
through the following system primitive:

\begin{center}
\fbox{\tt{
\texttt{int sys\_atrunlevel(void ({*}f)(void {*}),void {*}p, BYTE when);
}}}
\end{center}

The parameters for that function are:

\begin{description}

\item [\texttt{f}]the function to be registered;

\item [\texttt{p}]the parameter to be passed to function \texttt{f} when the
function will be called;

\item [\texttt{when}]is the situation in witch that function will be called. The
correct values are the following:

\begin{description}

\item [\texttt{RUNLEVEL\_INIT}]The function will be called after Module
registration, when the system is just entered in multitasking mode but no thread
executed yet;

\item [\texttt{RUNLEVEL\_SHUTDOWN}]The function will be called after a call to
\texttt{sys\_abort} or \texttt{sys\_end}; The system is still in multitasking
mode;

\item [\texttt{RUNLEVEL\_BEFORE\_EXIT}]The function will be called when the
Kernel exits from multitasking mode;

\item [\texttt{RUNLEVEL\_AFTER\_EXIT}]The function is called before the system
hangs (or returns to the host OS, if the proprietary extender is used).

\end{description}

It is also possible to specify with an OR operator a flag
\texttt{NO\_AT\_ABORT} that disables the call to the functions if
the system is exiting with a \texttt{sys\_abort} function.
\end{description}

You can post at most \texttt{MAX\_RUNLEVEL\_FUNC
} functions.

%----------------------------------------------------------------------------
\subsection{Interrupt disabling, printf and system termination}
%----------------------------------------------------------------------------

The Generic Kernel redefines the following OS Lib functions (see
the \texttt{include/kernel/func.h} file):

\begin{description}
\item [\texttt{cli}]renamed in \texttt{kern\_cli} \item
[\texttt{sti}]renamed in \texttt{kern\_sti} \item
[\texttt{ll\_fsave}]renamed in \texttt{kern\_fsave} \item
[\texttt{ll\_frestore}]renamed in \texttt{kern\_frestore}.
\end{description}

To display some messages to the console, the Generic Kernel
provides two functions called \texttt{printk} and
\texttt{kern\_printf}. These functions are similar to the standard
C \texttt{printf} function, except that they write directly to the
console. The \texttt{printk} function allows to specify a kind of
``importance'' for the message that the function displays. For
more information look at the file \texttt{include/kernel/func.h}
and \texttt{include/kernel/log.h}.

Finally, a clean system termination can be obtained through the \texttt{exit}
primitive. For more information look at the file \texttt{include/kernel/func.h}
and read Section \ref{Kernel_Inizializzazione
}.