PR#4880: document how to get backtraces when embedding Caml native code inside C.

Updates concerning the various ports (no more Alpha, etc).


git-svn-id: http://caml.inria.fr/svn/ocamldoc/trunk@11892 f963ae5c-01c2-4b8c-9fe0-0dff7051ff02

manual/manual/cmds/intf-c.etex

@@ -1455,15 +1455,23 @@ compilation of OCaml, as the variables "BYTECCLINKOPTS"
 -output-obj").  Currently, the only ports that require special
 attention are:
 \begin{itemize}
-\item Alpha under Digital Unix / Tru64 Unix with "gcc":
-object files produced by "ocamlc -output-obj" must be linked with the
-"gcc" options "-Wl,-T,12000000 -Wl,-D,14000000".
-This is not necessary for object files produced by "ocamlopt -output-obj".
-\item Windows NT: the object file produced by OCaml have been
-compiled with the "/MD" flag, and therefore all other object files
-linked with it should also be compiled with "/MD".
+\item Windows with the MSVC compiler: the object file produced by
+Objective Caml have been compiled with the "/MD" flag, and therefore
+all other object files linked with it should also be compiled with
+"/MD".
 \end{itemize}
 
+\paragraph{Stack backtraces.}  When OCaml bytecode produced by
+"ocamlc -g" is embedded in a C program, no debugging information is
+included, and therefore it is impossible to print stack backtraces on
+uncaught exceptions.  This is not the case when native code produced
+by "ocamlopt -g" is embedded in a C program: stack backtrace
+information is available, but the backtrace mechanism needs to be
+turned on programmatically.   This can be achieved from the OCaml side
+by calling "Printexc.record_backtrace true" in the initialization of
+one of the Caml modules.  This can also be achieved from the C side
+by calling "caml_record_backtrace(Val_int(1));" in the Caml-C glue code.
+
 \section{Advanced example with callbacks}
 \pdfsection{Advanced example with callbacks}
 

manual/manual/cmds/native.etex

@@ -258,7 +258,7 @@ compile-time and at link-time.  Linking object files not compiled with
 information about the profiles.
 
 Full support for "gprof" is only available for certain platforms
-(currently: Intel x86/Linux and Alpha/Digital Unix).
+(currently: Intel x86 32 and 64 bits under Linux, BSD and MacOS X).
 On other platforms, the "-p" option will result in a less precise
 profile (no call graph information, only a time profile).
 \end{unix}
@@ -539,17 +539,11 @@ platform used:
 (MS Windows 64 bits).
 \end{itemize}
 
-\item On IA32 processors only (Intel Pentium, AMD Athlon, etc, in
-32-bit mode), some intermediate results in floating-point computations
-are kept in extended precision rather than being rounded to double
+\item On IA32 processors only (Intel and AMD x86 processors in 32-bit
+mode), some intermediate results in floating-point computations are
+kept in extended precision rather than being rounded to double
 precision like the bytecode compiler always does.  Floating-point
-results can therefore differ between bytecode and native code; in
-general, the results obtained with native code are ``more exact''
-(less affected by rounding errors and loss of precision).
-
-\item On the Alpha processor only, floating-point operations involving
-infinite or denormalized numbers can abort the program on a
-``floating-point exception'' signal.
+results can therefore differ slightly between bytecode and native code.
 
 \end{itemize}
 

manual/manual/cmds/profil.etex

@@ -146,8 +146,8 @@ The output of "gprof" is described in the Unix manual page for
 "gprof(1)".  It generally consists of two parts: a ``flat'' profile
 showing the time spent in each function and the number of invocation
 of each function, and a ``hierarchical'' profile based on the call
-graph.  Currently, only the Intel x86/Linux and Alpha/Digital Unix
-ports of "ocamlopt" support the two profiles.  On other platforms,
+graph.  Currently, only the Intel x86 ports of "ocamlopt" under
+Linux, BSD and MacOS X support the two profiles.  On other platforms,
 "gprof" will report only the ``flat'' profile with just time
 information.  When reading the output of "gprof", keep in mind that
 the accumulated times computed by "gprof" are based on heuristics and