\marginparsep 0 in
\sloppy
+\usepackage{epsfig}
+
\newcommand{\note}[1]{{\em Note: #1}}
% DIMENSION OF TEXT:
\textheight 8.5 in
\tableofcontents
\newpage
-\section{Introduction}
+\part{Introduction}
+\section{Overview}
This document describes the GHC/Hugs run-time system. It serves as
a Glasgow/Yale/Nottingham ``contract'' about what the RTS does.
\end{itemize}
+%-----------------------------------------------------------------------------
+\part{Evaluation Model}
+\section{Compiled Execution}
-\section{The Scheduler}
-
-The Scheduler is the heart of the run-time system. A running program
-consists of a single running thread, and a list of runnable and
-blocked threads. The running thread returns to the scheduler when any
-of the following conditions arises:
-
-\begin{itemize}
-\item A heap check fails, and a garbage collection is required
-\item Compiled code needs to switch to interpreted code, and vice
-versa.
-\item The thread becomes blocked.
-\item The thread is preempted.
-\end{itemize}
-
-A world-switch (i.e. when compiled code encounters interpreted code,
-and vice-versa) can happen in six ways:
-
-\begin{itemize}
-\item A GHC thread enters a Hugs-built thunk.
-\item A GHC thread calls a Hugs-compiled function.
-\item A GHC thread returns to a Hugs-compiled return address.
-\item A Hugs thread enters a GHC-built thunk.
-\item A Hugs thread calls a GHC-compiled function.
-\item A Hugs thread returns to a Hugs-compiled return address.
-\end{itemize}
-
-A running system has a global state, consisting of
-
-\begin{itemize}
-\item @Hp@, the current heap pointer, which points to the next
-available address in the Heap.
-\item @HpLim@, the heap limit pointer, which points to the end of the
-heap.
-\item The Thread Preemption Flag, which is set whenever the currently
-running thread should be preempted at the next opportunity.
-\end{itemize}
-
-Each thread has a thread-local state, which consists of
-
-\begin{itemize}
-\item @TSO@, the Thread State Object for this thread. This is a heap
-object that is used to store the current thread state when the thread
-is blocked or sleeping.
-\item @Sp@, the current stack pointer.
-\item @Su@, the current stack update frame pointer. This register
-points to the most recent update frame on the stack, and is used to
-calculate the number of arguments available when entering a function.
-\item @SpLim@, the stack limit pointer. This points to the end of the
-current stack chunk.
-\item Several general purpose registers, used for passing arguments to
-functions.
-\end{itemize}
-
-\noindent and various other bits of information used in specialised
-circumstances, such as profiling and parallel execution. These are
-described in the appropriate sections.
-
-The following is pseudo-code for the inner loop of the scheduler
-itself.
-
-@
-while (threads_exist) {
- // handle global problems: GC, parallelism, etc
- if (need_gc) gc();
- if (external_message) service_message();
- // deal with other urgent stuff
-
- pick a runnable thread;
- do {
- switch (thread->whatNext) {
- case (RunGHC pc): status=runGHC(pc); break;
- case (RunHugs bc): status=runHugs(bc); break;
- }
- switch (status) { // handle local problems
- case (StackOverflow): enlargeStack; break;
- case (Error e) : error(thread,e); break;
- case (ExitWith e) : exit(e); break;
- case (Yield) : break;
- }
- } while (thread_runnable);
-}
-@
-
-Optimisations to avoid excess trampolining from Hugs into itself.
-How do we invoke GC, ccalls, etc.
-General ccall (@ccall-GC@) and optimised ccall.
-
-\section{Evaluation}
+This section describes the framework in which compiled code evaluates
+expressions. Only at certain points will compiled code need to be
+able to talk to the interpreted world; these are discussed in Section
+\ref{sect:switching-worlds}.
\subsection{Calling conventions}
@
case x of (a,b) -> E
@
-In a stack-based evaluator such as the STG machine,
-a @case@ expression is evaluated by pushing a {\em return address} on the stack
-before evaluating the scrutinee (@x@ in this case). Once evaluation of the
-scrutinee is complete, execution resumes at the return address, which
-points to the code for the expression @E@.
+
+The code for a @case@ expression looks like this:
+
+\begin{itemize}
+\item Push the free variables of the branches on the stack (fv(@E@) in
+this case).
+\item Push a \emph{return address} on the stack.
+\item Evaluate the scrutinee (@x@ in this case).
+\end{itemize}
+
+Once evaluation of the scrutinee is complete, execution resumes at the
+return address, which points to the code for the expression @E@.
When execution resumes at the return point, there must be some {\em
return convention} that defines where the components of the pair, @a@
heap.
When passing an unboxed tuple to a function, the components are
-flattened out and passed on the stack/in registers as usual.
+flattened out and passed in \Arg{1} \ldots \Arg{n} as usual.
\end{itemize}
@
data Maybe a = Nothing | Just a
@
-How does the return convention encode which of the two constructors is being returned?
-A @case@ expression scrutinising a value of @Maybe@ type would look like this:
+How does the return convention encode which of the two constructors is
+being returned? A @case@ expression scrutinising a value of @Maybe@
+type would look like this:
@
case E of
Nothing -> ...
The code at the return address will test the tag and jump to the
appropriate code for the case branch.
-\ToDo{Decide whether it's better to load the tag into \Arg{2} or not.
-May be affected by whether \Arg{2} is a real register.}
-
The choice of whether to use a vectored return or a direct return is
made on a type-by-type basis --- up to a certain maximum number of
constructors imposed by the update mechanism
(section~\ref{sect:data-updates}).
+Single-constructor data types also use direct returns, although in
+that case there is no need to return a tag in \Arg{2}.
+
\ToDo{Say whether we pop the return address before returning}
+\ToDo{Stack stubbing?}
+
\subsection{Updates}
\label{sect:data-updates}
\item The update frame is still on the stack.
\end{itemize}
+We can safely share a single statically-compiled update function
+between all types. However, the code must be able to handle both
+vectored and direct-return datatypes. This is done by arranging that
+the update code looks like this:
+
+@
+ | ^ |
+ | return vector |
+ |---------------|
+ | fixed-size |
+ | info table |
+ |---------------| <- update code pointer
+ | update code |
+ | v |
+@
+
+Each entry in the return vector (which is large enough to cover the
+largest vectored-return type) points to the update code.
+
The update code:
\begin{itemize}
\item overwrites the {\em updatee} with an indirection to \Arg{1};
\item enters \Arg{1}.
\end{itemize}
-This update code is the same for all data types, and can therefore be
-compiled statically in the runtime system.
-
-Since Haskell is polymorphic, we sometimes have to compile code for
-updatable thunks without knowing the type that will be returned. In
-this case, the update frame must work for both direct and vectored
-returns. This requires that we generate an infotable containing both
-a valid direct return address (which will perform the update and then
-perform a direct return) and a valid return vector (each entry of
-which will perform the update and then perform a vectored return).
-
+We enter \Arg{1} again, having probably just come from there, because
+it knows whether to perform a direct or vectored return. This could
+be optimised by compiling special update code for each slot in the
+return vector, which performs the correct return.
\subsection{Semi-tagging}
\label{sect:semi-tagging}
May have to revert black holes - ouch!
@
+\section{Interpreted Execution}
+
+This section describes how the Hugs interpreter interprets code in the
+same environment as compiled code executes. Both evaluation models
+use a common garbage collector, so they must agree on the form of
+objects in the heap.
+
+Hugs interprets code by converting it to byte-code and applying a
+byte-code interpreter to it. Wherever possible, we try to ensure that
+the byte-code is all that is required to interpret a section of code.
+This means not dynamically generating info tables, and hence we can
+only have a small number of possible heap objects each with a staticly
+compiled info table. Similarly for stack objects: in fact we only
+have one Hugs stack object, in which all information is tagged for the
+garbage collector.
+
+There is, however, one exception to this rule. Hugs must generate
+info tables for any constructors it is asked to compile, since the
+alternative is to force a context-switch each time compiled code
+enters a Hugs-built constructor, which would be prohibitively
+expensive.
+
+\subsection{Hugs Heap Objects}
+\label{sect:hugs-heap-objects}
+
+\subsubsection{Byte-Code Objects}
+
+Compiled byte code lives on the global heap, in objects called
+Byte-Code Objects (or BCOs). The layout of BCOs is described in
+detail in Section \ref{sect:BCO}, in this section we will describe
+their semantics.
+
+Since byte-code lives on the heap, it can be garbage collected just
+like any other heap-resident data. Hugs maintains a table of
+currently live BCOs, which is treated as a table of live pointers by
+the garbage collector. When a module is unloaded, the pointers to its
+BCOs are removed from the table, and the code will be garbage
+collected some time later.
+
+A BCO represents a basic block of code - all entry points are at the
+beginning of a BCO, and it is impossible to jump into the middle of
+one. A BCO represents not only the code for a function, but also its
+closure; a BCO can be entered just like any other closure. Hugs
+performs lambda-lifting during compilation to byte-code, and each
+top-level combinator becomes a BCO in the heap.
+
+\subsubsection{Thunks}
+
+A thunk consists of a code pointer, and values for the free variables
+of that code. Since Hugs byte-code is lambda-lifted, free variables
+become arguments and are expected to be on the stack by the called
+function.
+
+Hugs represents thunks with an AP object. The AP object contains one
+or more pointers to other heap objects. When it is entered, it pushes
+an update frame followed by its payload on the stack, and enters the
+first word (which will be a pointer to a BCO). The layout of AP
+objects is described in more detail in Section \ref{sect:AP}.
+
+\subsubsection{Partial Applications}
+
+Partial applications are represented by PAP objects. A PAP object is
+exactly the same as an AP object, except that it is non-updatable.
+The layout of PAP objects is described in Section \ref{sect:PAP}.
+
+\subsection{Calling conventions}
+\label{sect:hugs-calling-conventions}
+
+The calling convention for any byte-code function is straightforward:
+
+\begin{itemize}
+\item Push any arguments on the stack.
+\item Push a pointer to the BCO.
+\item Begin interpreting the byte code.
+\end{itemize}
+
+The @ENTER@ byte-code instruction decides how to enter its argument.
+The object being entered must be either
+
+\begin{itemize}
+\item A BCO,
+\item An AP,
+\item A PAP,
+\item A constructor,
+\item A GHC-built closure, or
+\item An indirection.
+\end{itemize}
+
+If @ENTER@ is applied to a BCO, we just begin interpreting the
+byte-code contained therein. If the object is an AP, we push an
+update frame, push the values from the AP on the stack, and enter its
+associated object. If the object is a PAP, we push its values on the
+stack and enter the first one. If the object is a constructor, we
+simply return (see Section \ref{sect:hugs-return-convention}). The
+fourth case is convered in Section \ref{sect:hugs-to-ghc-closure}. If
+the object is an indirection, we simply enter the object it points to.
+
+\subsection{Return convention}
+\label{sect:hugs-return-convention}
+When Hugs pushes a return address, it pushes both a pointer to the BCO
+to return to, and a pointer to a static code fragment @HUGS_RET@ (this
+will be described in Section \ref{sect:ghc-to-hugs-return}). The
+stack layout is shown in Figure \ref{fig:hugs-return-fig}.
+
+\begin{figure}
+\begin{center}
+\input{hugs_ret.pstex_t}
+\end{center}
+\caption{Stack layout for a hugs return address}
+\label{fig:hugs-return-stack}
+\end{figure}
+
+\begin{figure}
+\begin{center}
+\input{hugs_ret2.pstex_t}
+\end{center}
+\caption{Stack layout on enterings a hugs return address}
+\label{fig:hugs-return2}
+\end{figure}
+
+When a Hugs byte-code sequence is returning, it first places the
+return value on the stack. It then examines the return address (now
+the second word on the stack):
+
+\begin{itemize}
+
+\item If the return address is @HUGS_RET@, rearrange the stack so that
+it has the returned object followed by the pointer to the BCO at the
+top, then enter the BCO (Figure \ref{fig:hugsreturn2}).
+
+\item If the top of the stack is not @HUGS_RET@, we need to do a world
+switch as described in Section \ref{sect:hugs-to-ghc-return}.
+
+\end{itemize}
+
+
+\section{The Scheduler}
+
+The Scheduler is the heart of the run-time system. A running program
+consists of a single running thread, and a list of runnable and
+blocked threads. The running thread returns to the scheduler when any
+of the following conditions arises:
+
+\begin{itemize}
+\item A heap check fails, and a garbage collection is required
+\item Compiled code needs to switch to interpreted code, and vice
+versa.
+\item The thread becomes blocked.
+\item The thread is preempted.
+\end{itemize}
+
+A running system has a global state, consisting of
+
+\begin{itemize}
+\item @Hp@, the current heap pointer, which points to the next
+available address in the Heap.
+\item @HpLim@, the heap limit pointer, which points to the end of the
+heap.
+\item The Thread Preemption Flag, which is set whenever the currently
+running thread should be preempted at the next opportunity.
+\item A list of runnable threads.
+\item A list of blocked threads.
+\end{itemize}
+
+Each thread is represented by a Thread State Object (TSO), which is
+described in detail in Section \ref{sect:TSO}.
+
+The following is pseudo-code for the inner loop of the scheduler
+itself.
+
+@
+while (threads_exist) {
+ // handle global problems: GC, parallelism, etc
+ if (need_gc) gc();
+ if (external_message) service_message();
+ // deal with other urgent stuff
+
+ pick a runnable thread;
+ do {
+ switch (thread->whatNext) {
+ case (RunGHC pc): status=runGHC(pc); break;
+ case (RunHugs bc): status=runHugs(bc); break;
+ }
+ switch (status) { // handle local problems
+ case (StackOverflow): enlargeStack; break;
+ case (Error e) : error(thread,e); break;
+ case (ExitWith e) : exit(e); break;
+ case (Yield) : break;
+ }
+ } while (thread_runnable);
+}
+@
+
+Optimisations to avoid excess trampolining from Hugs into itself.
+How do we invoke GC, ccalls, etc.
+General ccall (@ccall-GC@) and optimised ccall.
+
+\section{Switching Worlds}
+
+\label{sect:switching-worlds}
+
+Because this is a combined compiled/interpreted system, the
+interpreter will sometimes encounter compiled code, and vice-versa.
+
+All world-switches go via the scheduler, ensuring that the world is in
+a known state ready to enter either compiled code or the interpreter.
+When a thread is run from the scheduler, the @whatNext@ field in the
+TSO (Section \ref{sect:TSO}) is checked to find out how to execute the
+thread.
+
+\begin{itemize}
+\item If @whatNext@ is set to @ReturnGHC@, we load up the required
+registers from the TSO and jump to the address at the top of the user
+stack.
+\item If @whatNext@ is set to @EnterGHC@, we load up the required
+registers from the TSO and enter the closure pointed to by the top
+word of the stack.
+\item If @whatNext@ is set to @EnterHugs@, we enter the top thing on
+the stack, using the interpreter.
+\end{itemize}
+
+There are four cases we need to consider:
+
+\begin{enumerate}
+\item A GHC thread enters a Hugs-built closure.
+\item A GHC thread returns to a Hugs-compiled return address.
+\item A Hugs thread enters a GHC-built closure.
+\item A Hugs thread returns to a Hugs-compiled return address.
+\end{enumerate}
+
+GHC-compiled modules cannot call functions in a Hugs-compiled module
+directly, because the compiler has no information about arities in the
+external module. Therefore it must assume any top-level objects are
+CAFs, and enter their closures.
+
+\ToDo{dynamic linking stuff}
+\ToDo{Hugs-built constructors?}
+
+We now examine the various cases one by one and describe how the
+switch happens in each situation.
+
+\subsection{A GHC thread enters a Hugs-built closure}
+\label{sect:ghc-to-hugs-closure}
+
+There are two possibilities: GHC has entered the BCO directly (for a
+top-level function closure), or it has entered an AP.
+
+The code for both objects is the same:
+
+\begin{itemize}
+\item Push the address of the BCO on the stack.
+\item Save the current state of the thread in its TSO.
+\item Return to the scheduler, setting @whatNext@ to @EnterHugs@.
+\end{itemize}
+
+\subsection{A GHC thread returns to a Hugs-compiled return address}
+\label{sect:ghc-to-hugs-return}
+
+Hugs return addresses are laid out as in Figure
+\ref{fig:hugs-return-stack}. If GHC is returning, it will return to
+the address at the top of the stack, namely @HUGS_RET@. The code at
+@HUGS_RET@ performs the following:
+
+\begin{itemize}
+\item pushes \Arg{1} (the return value) on the stack.
+\item saves the thread state in the TSO
+\item returns to the scheduler with @whatNext@ set to @EnterHugs@.
+\end{itemize}
+
+\noindent When Hugs runs, it will enter the return value, which will
+return using the correct Hugs convention (Section
+\ref{sect:hugs-return-convention}) to the return address underneath it
+on the stack.
+
+\subsection{A Hugs thread enters a GHC-compiled closure}
+\label{sect:hugs-to-ghc-closure}
+
+Hugs can recognise a GHC-built closure as not being one of the
+following types of object:
+
+\begin{itemize}
+\item A BCO.
+\item An AP.
+\item A constructor.
+\end{itemize}
+
+When Hugs is called on to enter a GHC closure, it executes the
+following sequence of instructions:
+
+\begin{itemize}
+\item Push the address of the closure on the stack.
+\item Save the current state of the thread in the TSO.
+\item Return to the scheduler, with the @whatNext@ field set to
+@EnterGHC@.
+\end{itemize}
+
+\subsection{A Hugs thread returns to a GHC-compiled return address}
+\label{sect:hugs-to-ghc-return}
+
+When hugs encounters a return address on the stack that is not
+@HUGS_RET@, it knows that a world-switch is required. At this point
+the stack contains a pointer to the return value, followed by the GHC
+return address. The following sequence is then performed:
+
+\begin{itemize}
+\item save the state of the thread in the TSO.
+\item return to the scheduler, setting @whatNext@ to @EnterGHC@.
+\end{itemize}
+
+The first thing that GHC will do is enter the object on the top of the
+stack, which is a pointer to the return value. This value will then
+return itself to the return address using the GHC return convention.
+
+\part{Implementation}
\section{Heap objects}
\label{sect:fixed-header}
+\subsection{Hugs Objects}
+
+\subsubsection{Byte-Code Objects}
+\label{sect:BCO}
+
+A Byte-Code Object (BCO) is a container for a a chunk of byte-code,
+which can be executed by Hugs. For a top-level function, the BCO also
+serves as the closure for the function.
+
+The semantics of BCOs are described in Section
+\ref{sect:hugs-heap-objects}. A BCO has the following structure:
+
+\begin{center}
+\begin{tabular}{|l|l|l|l|l|l|}
+\hline
+\emph{BCO} & \emph{Layout} & \emph{Offset} & \emph{Size} &
+\emph{Literals} & \emph{Byte code} \\
+\hline
+\end{tabular}
+\end{center}
+
+\noindent where:
+\begin{itemize}
+\item \emph{BCO} is a pointer to a static code fragment/info table that
+returns to the scheduler to invoke Hugs (Section
+\ref{sect:ghc-to-hugs-closure}).
+\item \emph{Layout} contains the number of pointer literals in the
+\emph{Literals} field.
+\item \emph{Offset} is the offset to the byte code from the start of
+the object.
+\item \emph{Size} is the number of words of byte code in the object.
+\item \emph{Literals} contains any pointer and non-pointer literals used in
+the byte-codes (including jump addresses), pointers first.
+\item \emph{Byte code} contains \emph{Size} words of non-pointer byte
+code.
+\end{itemize}
+
+\subsubsection{AP objects}
+\label{sect:AP}
+
+Hugs uses a standard object called an AP for thunks and partial
+applications. The layout of an AP is
+
+\begin{center}
+\begin{tabular}{|l|l|l|l|}
+\hline
+\emph{AP} & \emph{BCO} & \emph{Layout} & \emph{Free Variables} \\
+\hline
+\end{tabular}
+\end{center}
+
+\noindent where:
+
+\begin{itemize}
+\item \emph{AP} is a pointer to a statically-compiled code
+fragment/info table that returns to the scheduler to invoke Hugs
+(Sections \ref{sect:ghc-to-hugs-closure}, \ref{sect:ghc-to-hugs-return}).
+\item \emph{BCO} is a pointer to the BCO for the thunk.
+\item \emph{Layout} contains the number of pointers and the size of
+the \emph{Free Variables} field.
+\item \emph{Free Variables} contains the free variables of the
+thunk/partial application/return address, pointers first.
+\end{itemize}
+
\subsection{Pointed Objects}
All pointed objects can be entered.
\subsection{Internal workings of the Generational Collector}
+\section{Dynamic Linking}
\section{Profiling}
projects / ghc-hetmet.git / blobdiff