* doc/sdccman.lyx: corrected cpp info

[fw/sdcc] / doc / sdccman.lyx

diff --git a/doc/sdccman.lyx b/doc/sdccman.lyx
index 1fd4a6880719cd8fc0d372831676feb0c7d5082d..fabc8d8dd1cdcc6a04ea8989aaa4c5a4694579b6 100644 (file)
--- a/doc/sdccman.lyx
+++ b/doc/sdccman.lyx
@@ -114,11 +114,11 @@ SDCC Compiler User Guide
 \begin_layout Date
 
 \size normal
-SDCC 2.8.6
+SDCC 2.8.9
 \size footnotesize
 
 \newline
-$Date: 2008-11-16#$ 
+$Date::            $ 
 \newline
 $Revision: 5272$
 \end_layout
@@ -7082,6 +7082,84 @@ name "-\\/-debug"
  is embedded in the library file itself.
  Library files created using sdcclib are used as described in the preceding
  sections.
+\end_layout
+
+\begin_layout Subsection
+Using ar to Create and Manage Libraries
+\begin_inset LatexCommand index
+name "ar"
+
+\end_inset
+
+
+\end_layout
+
+\begin_layout Standard
+Support for ar format libraries was introduced in sdcc 2.9.0.
+ Ar is a standard archive managing utility on unices (Linux, Mac OS X, several
+ unix flavors) so it is not included in the sdcc package.
+\series bold
+
+\newline
+
+\newline
+
+\series default
+For Windows platform you can find ar utility in GNU binutils package included
+ in several projects: Cygwin at 
+\begin_inset LatexCommand url
+target "http://www.cygwin.com/"
+
+\end_inset
+
+, MinGW at 
+\begin_inset LatexCommand url
+target "http://www.mingw.org/"
+
+\end_inset
+
+.
+\series bold
+
+\newline
+
+\newline
+
+\series default
+Both the GNU and BSD ar format variants are suppurated by asxxxx linkers.
+ Ar doesn't understand the asxxxx object file format, so there is a special
+ version of ranlib distributed with sdcc, called asranlib, which produces
+ the ar symbol lookup table.
+\series bold
+
+\newline
+
+\newline
+
+\series default
+To create a library containing asxxxx object files, you should use the following
+ sequence:
+\series bold
+
+\newline
+
+\end_layout
+
+\begin_layout Standard
+
+\family sans
+\series bold
+ar -Sq <library name>.lib <list of .rel files>
+\end_layout
+
+\begin_layout Standard
+
+\family sans
+\series bold
+asranlib <library name>.lib
+\end_layout
+
+\begin_layout Standard
 \begin_inset VSpace bigskip
 \end_inset
 
@@ -7689,16 +7767,16 @@ name "sdcpp (preprocessor)"
 \end_inset
 
 .
- SDCC uses an adapted version of the preprocessor 
-\emph on
-cpp
-\emph default
- of the GNU Compiler Collection
+ SDCC uses an adapted version of the GNU Compiler Collection
 \begin_inset LatexCommand index
 name "gcc (GNU Compiler Collection)"
 
 \end_inset
 
+ preprocessor 
+\emph on
+cpp
+\emph default
  (
 \emph on
 gcc
@@ -7709,11 +7787,11 @@ target "http://gcc.gnu.org/"
 
 \end_inset
 
-), if you need more dedicated options please refer to the GCC\InsetSpace ~
-4.1.1\InsetSpace ~
+).
+ If you need more dedicated options please refer to the GCC\InsetSpace ~
 CPP\InsetSpace ~
-Manual
- at 
+Manual at
+ 
 \begin_inset LatexCommand htmlurl
 target "http://www.gnu.org/software/gcc/onlinedocs/"
 
@@ -21001,7 +21079,7 @@ name "printf()"
 printf_large.c
 \family default
  does not support float (except on ds390).
- To enable this recompile it with the option 
+ To enable this, recompile it with the option 
 \emph on
 -
 \begin_inset ERT
@@ -21049,6 +21127,13 @@ name "-\\/-model-large"
 
 \emph default
  for the mcs51 port, since this uses a lot of memory.
+ To enable float support for the pic16 targets, see 
+\begin_inset LatexCommand ref
+reference "sub:pic16Libraries"
+
+\end_inset
+
+.
 \end_layout
 
 \begin_layout Standard
@@ -27254,9 +27339,19 @@ Building the libraries
 Before using SDCC/pic16 there are some libraries that need to be compiled.
  This process is done automatically if gputils are found at SDCC's compile
  time.
- Should you require to rebuild the pic16 libraries manually, these are the
- steps required to do so under Linux or Mac OS X (cygwin might work as well,
- but is untested):
+ Should you require to rebuild the pic16 libraries manually (e.g.
+ in order to enable output of float values
+\begin_inset LatexCommand index
+name "Floating point support"
+
+\end_inset
+
+ via 
+\family typewriter
+printf()
+\family default
+, see below), these are the steps required to do so under Linux or Mac OS
+ X (cygwin might work as well, but is untested):
 \end_layout
 
 \begin_layout LyX-Code
@@ -27295,6 +27390,54 @@ cd device/include
 su -c 'make install'     # install the headers, you need the root password
 \end_layout
 
+\begin_layout Subsubsection*
+Output of float values via printf()
+\end_layout
+
+\begin_layout Standard
+The library is normally built without support for displaying float values,
+ only <NO FLOAT> will appear instead of the value.
+ To change this, rebuild the library as stated above, but call 
+\family typewriter
+./configure.gnu --enable-floats 
+\family default
+instead of just 
+\family typewriter
+./configure.gnu
+\family default
+.
+ Also make sure that at least 
+\family typewriter
+libc/stdio/vfprintf.c
+\family default
+ is actually recompiled, e.g.
+ by 
+\family typewriter
+touch
+\family default
+ing it after the 
+\family typewriter
+configure
+\family default
+ run or deleting its 
+\family typewriter
+.o
+\family default
+ file.
+\end_layout
+
+\begin_layout Standard
+The more common appraoch of compiling 
+\family typewriter
+vfprintf.c
+\family default
+ manually with 
+\family typewriter
+-DUSE_FLOATS=1
+\family default
+ should also work, but is untested.
+\end_layout
+
 \begin_layout Subsection
 Adding New Devices to the Port
 \end_layout