1 #LyX 1.3 created this file. For more info see http://www.lyx.org/
6 pdftitle={SDCC Compiler User Guide},
7 pdfauthor={SDCC development team},
8 pdfsubject={installation, user manual},
9 pdfkeywords={8032, 8051, ansi, c, compiler, CPU, DS390,
10 embedded, GPL, HC08, manual, mcs51, PIC, Z80},
12 linkcolor=blue] {hyperref}
16 \emergencystretch=30pt
21 \inputencoding default
24 \paperfontsize default
26 \papersize letterpaper
31 \use_numerical_citations 0
32 \paperorientation portrait
39 \paragraph_separation indent
41 \quotes_language swedish
49 Please note: double dashed longoptions (e.g.
50 --version) are written this way: -
64 three consecutive dashes simply result in a long resp.
68 Architecture specific stuff (like memory models, code examples) should maybe
72 into seperate sections/chapters/appendices (it is hard to document PIC or
76 a 8051 centered document) - for now simply add.
79 SDCC Compiler User Guide
93 The above strings enclosed in $ are automatically updated by cvs
97 \begin_inset LatexCommand \tableofcontents{}
130 ompiler) is a Freeware, retargettable, optimizing ANSI-C compiler by
134 designed for 8 bit Microprocessors.
135 The current version targets Intel MCS51 based Microprocessors (8031, 8032,
137 \begin_inset LatexCommand \index{8031, 8032, 8051, 8052, mcs51 CPU}
141 , etc.), Dallas DS80C390 variants, Motorola HC08 and Zilog Z80 based MCUs.
142 It can be retargetted for other microprocessors, support for Microchip
143 PIC, Atmel AVR is under development.
144 The entire source code for the compiler is distributed under GPL.
145 SDCC uses ASXXXX & ASLINK, a Freeware, retargettable assembler & linker.
146 SDCC has extensive language extensions suitable for utilizing various microcont
147 rollers and underlying hardware effectively.
152 In addition to the MCU specific optimizations SDCC also does a host of standard
156 global sub expression elimination,
159 loop optimizations (loop invariant, strength reduction of induction variables
163 constant folding & propagation,
169 dead code elimination
179 For the back-end SDCC uses a global register allocation scheme which should
180 be well suited for other 8 bit MCUs.
185 The peep hole optimizer uses a rule based substitution mechanism which is
191 Supported data-types are:
194 char (8 bits, 1 byte),
197 short and int (16 bits, 2 bytes),
200 long (32 bit, 4 bytes)
207 The compiler also allows
209 inline assembler code
211 to be embedded anywhere in a function.
212 In addition, routines developed in assembly can also be called.
216 SDCC also provides an option (-
226 -cyclomatic) to report the relative complexity of a function.
227 These functions can then be further optimized, or hand coded in assembly
233 SDCC also comes with a companion source level debugger SDCDB, the debugger
234 currently uses ucSim a freeware simulator for 8051 and other micro-controllers.
239 The latest version can be downloaded from
240 \begin_inset LatexCommand \url{http://sdcc.sourceforge.net/snap.php}
250 Please note: the compiler will probably always be some steps ahead of this
255 \begin_inset LatexCommand \index{Status of documentation}
265 Obviously this has pros and cons
274 All packages used in this compiler system are
282 ; source code for all the sub-packages (pre-processor, assemblers, linkers
283 etc) is distributed with the package.
284 This documentation is maintained using a freeware word processor (LyX).
286 This program is free software; you can redistribute it and/or modify it
287 under the terms of the GNU General Public License
288 \begin_inset LatexCommand \index{GNU General Public License, GPL}
292 as published by the Free Software Foundation; either version 2, or (at
293 your option) any later version.
294 This program is distributed in the hope that it will be useful, but WITHOUT
295 ANY WARRANTY; without even the implied warranty
296 \begin_inset LatexCommand \index{warranty}
300 of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
301 See the GNU General Public License for more details.
302 You should have received a copy of the GNU General Public License along
303 with this program; if not, write to the Free Software Foundation, 59 Temple
304 Place - Suite 330, Boston, MA 02111-1307, USA.
305 In other words, you are welcome to use, share and improve this program.
306 You are forbidden to forbid anyone else to use, share and improve what
308 Help stamp out software-hoarding!
311 Typographic conventions
312 \begin_inset LatexCommand \index{Typographic conventions}
319 Throughout this manual, we will use the following convention.
320 Commands you have to type in are printed in
328 Code samples are printed in
333 Interesting items and new terms are printed in
338 Compatibility with previous versions
341 This version has numerous bug fixes compared with the previous version.
342 But we also introduced some incompatibilities with older versions.
343 Not just for the fun of it, but to make the compiler more stable, efficient
345 \begin_inset LatexCommand \index{ANSI-compliance}
350 \begin_inset LatexCommand \ref{sub:ANSI-Compliance}
354 for ANSI-Compliance).
360 short is now equivalent to int (16 bits), it used to be equivalent to char
361 (8 bits) which is not ANSI compliant
364 the default directory for gcc-builds where include, library and documentation
365 files are stored is now in /usr/local/share
368 char type parameters to vararg functions are casted to int unless explicitly
385 will push a as an int and as a char resp.
398 -regextend has been removed
411 -noregparms has been removed
424 -stack-after-data has been removed
429 <pending: more incompatibilities?>
435 What do you need before you start installation of SDCC? A computer, and
437 The preferred method of installation is to compile SDCC from source using
439 For Windows some pre-compiled binary distributions are available for your
441 You should have some experience with command line tools and compiler use.
447 The SDCC home page at
448 \begin_inset LatexCommand \url{http://sdcc.sourceforge.net/}
452 is a great place to find distribution sets.
453 You can also find links to the user mailing lists that offer help or discuss
454 SDCC with other SDCC users.
455 Web links to other SDCC related sites can also be found here.
456 This document can be found in the DOC directory of the source package as
458 Some of the other tools (simulator and assembler) included with SDCC contain
459 their own documentation and can be found in the source distribution.
460 If you want the latest unreleased software, the complete source package
461 is available directly by anonymous CVS on cvs.sdcc.sourceforge.net.
464 Wishes for the future
467 There are (and always will be) some things that could be done.
468 Here are some I can think of:
475 char KernelFunction3(char p) at 0x340;
483 \begin_inset LatexCommand \index{code banking (not supported)}
493 If you can think of some more, please see the section
494 \begin_inset LatexCommand \ref{sub:Requesting-Features}
498 about filing feature requests
499 \begin_inset LatexCommand \index{Requesting features}
504 \begin_inset LatexCommand \index{Feature request}
514 \begin_inset LatexCommand \index{Installation}
521 For most users it is sufficient to skip to either section
522 \begin_inset LatexCommand \ref{sub:Building-SDCC-on-Linux}
527 \begin_inset LatexCommand \ref{sub:Windows-Install}
532 More detailled instructions follow below.
536 \begin_inset LatexCommand \index{Options SDCC configuration}
543 The install paths, search paths and other options are defined when running
545 The defaults can be overridden by:
547 \labelwidthstring 00.00.0000
559 -prefix see table below
561 \labelwidthstring 00.00.0000
573 -exec_prefix see table below
575 \labelwidthstring 00.00.0000
587 -bindir see table below
589 \labelwidthstring 00.00.0000
601 -datadir see table below
603 \labelwidthstring 00.00.0000
605 docdir environment variable, see table below
607 \labelwidthstring 00.00.0000
609 include_dir_suffix environment variable, see table below
611 \labelwidthstring 00.00.0000
613 lib_dir_suffix environment variable, see table below
615 \labelwidthstring 00.00.0000
617 sdccconf_h_dir_separator environment variable, either / or
622 This character will only be used in sdccconf.h; don't forget it's a C-header,
623 therefore a double-backslash is needed there.
625 \labelwidthstring 00.00.0000
637 -disable-mcs51-port Excludes the Intel mcs51 port
639 \labelwidthstring 00.00.0000
651 -disable-gbz80-port Excludes the Gameboy gbz80 port
653 \labelwidthstring 00.00.0000
665 -disable-z80-port Excludes the z80 port
667 \labelwidthstring 00.00.0000
679 -disable-avr-port Excludes the AVR port
681 \labelwidthstring 00.00.0000
693 -disable-ds390-port Excludes the DS390 port
695 \labelwidthstring 00.00.0000
707 -disable-hc08-port Excludes the HC08 port
709 \labelwidthstring 00.00.0000
721 -disable-pic-port Excludes the PIC port
723 \labelwidthstring 00.00.0000
735 -disable-xa51-port Excludes the XA51 port
737 \labelwidthstring 00.00.0000
749 -disable-ucsim Disables configuring and building of ucsim
751 \labelwidthstring 00.00.0000
763 -disable-device-lib-build Disables automatically building device libraries
765 \labelwidthstring 00.00.0000
777 -disable-packihx Disables building packihx
779 \labelwidthstring 00.00.0000
791 -enable-libgc Use the Bohem memory allocator.
792 Lower runtime footprint.
795 Furthermore the environment variables CC, CFLAGS, ...
796 the tools and their arguments can be influenced.
797 Please see `configure -
807 -help` and the man/info pages of `configure` for details.
811 The names of the standard libraries STD_LIB, STD_INT_LIB, STD_LONG_LIB,
812 STD_FP_LIB, STD_DS390_LIB, STD_XA51_LIB and the environment variables SDCC_DIR_
813 NAME, SDCC_INCLUDE_NAME, SDCC_LIB_NAME are defined by `configure` too.
814 At the moment it's not possible to change the default settings (it was
815 simply never required).
819 These configure options are compiled into the binaries, and can only be
820 changed by rerunning 'configure' and recompiling SDCC.
821 The configure options are written in
825 to distinguish them from run time environment variables (see section search
831 \begin_inset Quotes sld
835 \begin_inset Quotes srd
838 are used by the SDCC team to build the official Win32 binaries.
839 The SDCC team uses Mingw32 to build the official Windows binaries, because
846 a gcc compiler and last but not least
849 the binaries can be built by cross compiling on Sourceforge's compile farm.
852 See the examples, how to pass the Win32 settings to 'configure'.
853 The other Win32 builds using Borland, VC or whatever don't use 'configure',
854 but a header file sdcc_vc_in.h is the same as sdccconf.h built by 'configure'
866 <lyxtabular version="3" rows="8" columns="3">
868 <column alignment="block" valignment="top" leftline="true" width="0in">
869 <column alignment="block" valignment="top" leftline="true" width="0in">
870 <column alignment="block" valignment="top" leftline="true" rightline="true" width="0in">
871 <row topline="true" bottomline="true">
872 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
880 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
888 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
898 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
908 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
916 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
928 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
938 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
948 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
960 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
970 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
982 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
998 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1008 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1020 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
1031 <row topline="true">
1032 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1042 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1054 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
1069 <row topline="true">
1070 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1080 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1088 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
1097 <row topline="true" bottomline="true">
1098 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1108 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1116 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
1134 'configure' also computes relative paths.
1135 This is needed for full relocatability of a binary package and to complete
1136 search paths (see section search paths below):
1142 \begin_inset Tabular
1143 <lyxtabular version="3" rows="4" columns="3">
1145 <column alignment="block" valignment="top" leftline="true" width="0in">
1146 <column alignment="block" valignment="top" leftline="true" width="0in">
1147 <column alignment="block" valignment="top" leftline="true" rightline="true" width="0in">
1148 <row topline="true" bottomline="true">
1149 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1157 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1165 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
1174 <row topline="true" bottomline="true">
1175 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1185 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1193 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
1202 <row bottomline="true">
1203 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1213 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1221 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
1230 <row bottomline="true">
1231 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1241 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1249 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
1282 \begin_inset Quotes srd
1286 \begin_inset Quotes srd
1300 \begin_inset Quotes srd
1304 \begin_inset Quotes srd
1332 To cross compile on linux for Mingw32 (see also 'sdcc/support/scripts/sdcc_mingw
1341 \begin_inset Quotes srd
1344 i586-mingw32msvc-gcc
1345 \begin_inset Quotes srd
1349 \begin_inset Quotes srd
1352 i586-mingw32msvc-g++
1353 \begin_inset Quotes srd
1361 \begin_inset Quotes srd
1364 i586-mingw32msvc-ranlib
1365 \begin_inset Quotes srd
1373 \begin_inset Quotes srd
1376 i586-mingw32msvc-strip
1377 \begin_inset Quotes srd
1395 \begin_inset Quotes srd
1399 \begin_inset Quotes srd
1417 \begin_inset Quotes srd
1421 \begin_inset Quotes srd
1429 \begin_inset Quotes srd
1433 \begin_inset Quotes srd
1441 \begin_inset Quotes srd
1445 \begin_inset Quotes srd
1453 \begin_inset Quotes srd
1457 \begin_inset Quotes srd
1464 sdccconf_h_dir_separator=
1465 \begin_inset Quotes srd
1477 \begin_inset Quotes srd
1494 -disable-device-lib-build
1522 -host=i586-mingw32msvc -
1532 -build=unknown-unknown-linux-gnu
1536 \begin_inset Quotes sld
1540 \begin_inset Quotes srd
1543 compile on Cygwin for Mingw32 (see also sdcc/support/scripts/sdcc_cygwin_mingw32
1552 \begin_inset Quotes srd
1556 \begin_inset Quotes srd
1564 \begin_inset Quotes srd
1568 \begin_inset Quotes srd
1586 \begin_inset Quotes srd
1590 \begin_inset Quotes srd
1608 \begin_inset Quotes srd
1612 \begin_inset Quotes srd
1620 \begin_inset Quotes srd
1624 \begin_inset Quotes srd
1632 \begin_inset Quotes srd
1636 \begin_inset Quotes srd
1644 \begin_inset Quotes srd
1648 \begin_inset Quotes srd
1655 sdccconf_h_dir_separator=
1656 \begin_inset Quotes srd
1668 \begin_inset Quotes srd
1688 'configure' is quite slow on Cygwin (at least on windows before Win2000/XP).
1699 -C' turns on caching, which gives a little bit extra speed.
1700 However if options are changed, it can be necessary to delete the config.cache
1705 \begin_inset LatexCommand \label{sub:Install-paths}
1710 \begin_inset LatexCommand \index{Install paths}
1716 \added_space_top medskip \align center
1718 \begin_inset Tabular
1719 <lyxtabular version="3" rows="5" columns="4">
1721 <column alignment="center" valignment="top" leftline="true" width="0">
1722 <column alignment="center" valignment="top" leftline="true" width="0">
1723 <column alignment="center" valignment="top" leftline="true" width="0">
1724 <column alignment="center" valignment="top" leftline="true" rightline="true" width="0">
1725 <row topline="true" bottomline="true">
1726 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1736 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1746 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1756 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
1767 <row topline="true">
1768 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1776 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1786 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1794 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
1807 <row topline="true">
1808 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1816 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1823 $DATADIR/ $INCLUDE_DIR_SUFFIX
1826 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1831 /usr/local/share/sdcc/include
1834 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
1847 <row topline="true">
1848 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1856 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1863 $DATADIR/$LIB_DIR_SUFFIX
1866 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1871 /usr/local/share/sdcc/lib
1874 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
1887 <row topline="true" bottomline="true">
1888 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1896 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1906 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1911 /usr/local/share/sdcc/doc
1914 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
1936 *compiler, preprocessor, assembler, and linker
1942 is auto-appended by the compiler, e.g.
1943 small, large, z80, ds390 etc
1946 The install paths can still be changed during `make install` with e.g.:
1949 make install prefix=$(HOME)/local/sdcc
1952 Of course this doesn't change the search paths compiled into the binaries.
1956 \begin_inset LatexCommand \label{sub:Search-Paths}
1961 \begin_inset LatexCommand \index{Search path}
1968 Some search paths or parts of them are determined by configure variables
1973 , see section above).
1974 Further search paths are determined by environment variables during runtime.
1977 The paths searched when running the compiler are as follows (the first catch
1983 Binary files (preprocessor, assembler and linker)
1989 \begin_inset Tabular
1990 <lyxtabular version="3" rows="4" columns="3">
1992 <column alignment="block" valignment="top" leftline="true" width="0in">
1993 <column alignment="block" valignment="top" leftline="true" width="0in">
1994 <column alignment="block" valignment="top" leftline="true" rightline="true" width="0in">
1995 <row topline="true" bottomline="true">
1996 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
2004 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
2012 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
2021 <row topline="true">
2022 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
2032 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
2040 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
2051 <row topline="true">
2052 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
2057 Path of argv[0] (if available)
2060 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
2068 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
2077 <row topline="true" bottomline="true">
2078 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
2086 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
2094 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
2119 \begin_inset Tabular
2120 <lyxtabular version="3" rows="6" columns="3">
2122 <column alignment="block" valignment="top" leftline="true" width="1.5in">
2123 <column alignment="block" valignment="top" leftline="true" width="1.5in">
2124 <column alignment="block" valignment="top" leftline="true" rightline="true" width="0in">
2125 <row topline="true" bottomline="true">
2126 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
2134 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
2142 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
2151 <row topline="true">
2152 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
2170 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
2188 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
2207 <row topline="true">
2208 <cell alignment="left" valignment="top" topline="true" leftline="true" usebox="none">
2216 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
2224 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
2233 <row topline="true">
2234 <cell alignment="left" valignment="top" topline="true" leftline="true" usebox="none">
2248 <cell alignment="left" valignment="top" topline="true" leftline="true" usebox="none">
2260 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
2271 <row topline="true">
2272 <cell alignment="left" valignment="top" topline="true" leftline="true" usebox="none">
2290 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
2340 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
2353 <row topline="true" bottomline="true">
2354 <cell alignment="left" valignment="top" topline="true" leftline="true" usebox="none">
2370 <cell alignment="left" valignment="top" topline="true" leftline="true" usebox="none">
2375 /usr/local/share/sdcc/
2380 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
2408 -nostdinc disables the last two search paths.
2418 With the exception of
2419 \begin_inset Quotes sld
2433 \begin_inset Quotes srd
2440 is auto-appended by the compiler (e.g.
2441 small, large, z80, ds390 etc.).
2448 \begin_inset Tabular
2449 <lyxtabular version="3" rows="6" columns="3">
2451 <column alignment="block" valignment="top" leftline="true" width="1.7in">
2452 <column alignment="block" valignment="top" leftline="true" width="1.2in">
2453 <column alignment="block" valignment="top" leftline="true" rightline="true" width="1.2in">
2454 <row topline="true" bottomline="true">
2455 <cell alignment="left" valignment="top" topline="true" leftline="true" usebox="none">
2463 <cell alignment="left" valignment="top" topline="true" leftline="true" usebox="none">
2471 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
2480 <row topline="true">
2481 <cell alignment="left" valignment="top" topline="true" leftline="true" usebox="none">
2499 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
2517 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
2536 <row topline="true">
2537 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
2549 <cell alignment="left" valignment="top" topline="true" leftline="true" usebox="none">
2561 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
2576 <row topline="true">
2577 <cell alignment="left" valignment="top" topline="true" leftline="true" usebox="none">
2588 $LIB_DIR_SUFFIX/<model>
2591 <cell alignment="left" valignment="top" topline="true" leftline="true" usebox="none">
2605 <cell alignment="left" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
2622 <row topline="true">
2623 <cell alignment="left" valignment="top" topline="true" leftline="true" usebox="none">
2638 $LIB_DIR_SUFFIX/<model>
2641 <cell alignment="left" valignment="top" topline="true" leftline="true" usebox="none">
2694 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
2750 <row topline="true" bottomline="true">
2751 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
2760 $LIB_DIR_SUFFIX/<model>
2763 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
2768 /usr/local/share/sdcc/
2775 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
2793 Don't delete any of the stray spaces in the table above without checking
2794 the HTML output (last line)!
2810 -nostdlib disables the last two search paths.
2814 \begin_inset LatexCommand \index{Building SDCC}
2821 Building SDCC on Linux
2822 \begin_inset LatexCommand \label{sub:Building-SDCC-on-Linux}
2831 Download the source package
2833 either from the SDCC CVS repository or from the nightly snapshots
2835 , it will be named something like sdcc
2846 \begin_inset LatexCommand \url{http://sdcc.sourceforge.net/snap.php}
2855 Bring up a command line terminal, such as xterm.
2860 Unpack the file using a command like:
2863 "tar -xvzf sdcc.src.tar.gz
2868 , this will create a sub-directory called sdcc with all of the sources.
2871 Change directory into the main SDCC directory, for example type:
2888 This configures the package for compilation on your system.
2904 All of the source packages will compile, this can take a while.
2920 This copies the binary executables, the include files, the libraries and
2921 the documentation to the install directories.
2922 Proceed with section
2923 \begin_inset LatexCommand \ref{sec:Testing-the-SDCC}
2930 Building SDCC on OSX 2.x
2933 Follow the instruction for Linux.
2937 On OSX 2.x it was reported, that the default gcc (version 3.1 20020420 (prerelease
2938 )) fails to compile SDCC.
2939 Fortunately there's also gcc 2.9.x installed, which works fine.
2940 This compiler can be selected by running 'configure' with:
2943 ./configure CC=gcc2 CXX=g++2
2946 Cross compiling SDCC on Linux for Windows
2949 With the Mingw32 gcc cross compiler it's easy to compile SDCC for Win32.
2950 See section 'Configure Options'.
2953 Building SDCC on Windows
2956 With the exception of Cygwin the SDCC binaries uCsim and sdcdb can't be
2958 They use Unix-sockets, which are not available on Win32.
2961 Building SDCC using Cygwin and Mingw32
2964 For building and installing a Cygwin executable follow the instructions
2970 \begin_inset Quotes sld
2974 \begin_inset Quotes srd
2977 Win32-binary can be built, which will not need the Cygwin-DLL.
2978 For the necessary 'configure' options see section 'configure options' or
2979 the script 'sdcc/support/scripts/sdcc_cygwin_mingw32'.
2983 In order to install Cygwin on Windows download setup.exe from
2984 \begin_inset LatexCommand \url[www.cygwin.com]{http://www.cygwin.com/}
2990 \begin_inset Quotes sld
2993 default text file type
2994 \begin_inset Quotes srd
2998 \begin_inset Quotes sld
3002 \begin_inset Quotes srd
3005 and download/install at least the following packages.
3006 Some packages are selected by default, others will be automatically selected
3007 because of dependencies with the manually selected packages.
3008 Never deselect these packages!
3017 gcc ; version 3.x is fine, no need to use the old 2.9x
3020 binutils ; selected with gcc
3026 rxvt ; a nice console, which makes life much easier under windoze (see below)
3029 man ; not really needed for building SDCC, but you'll miss it sooner or
3033 less ; not really needed for building SDCC, but you'll miss it sooner or
3037 cvs ; only if you use CVS access
3040 If you want to develop something you'll need:
3043 python ; for the regression tests
3046 gdb ; the gnu debugger, together with the nice GUI
3047 \begin_inset Quotes sld
3051 \begin_inset Quotes srd
3057 openssh ; to access the CF or commit changes
3060 autoconf and autoconf-devel ; if you want to fight with 'configure', don't
3061 use autoconf-stable!
3064 rxvt is a nice console with history.
3065 Replace in your cygwin.bat the line
3084 rxvt -sl 1000 -fn "Lucida Console-12" -sr -cr red
3087 -bg black -fg white -geometry 100x65 -e bash -
3100 Text selected with the mouse is automatically copied to the clipboard, pasting
3101 works with shift-insert.
3105 The other good tip is to make sure you have no //c/-style paths anywhere,
3106 use /cygdrive/c/ instead.
3107 Using // invokes a network lookup which is very slow.
3109 \begin_inset Quotes sld
3113 \begin_inset Quotes srd
3116 is too long, you can change it with e.g.
3122 SDCC sources use the unix line ending LF.
3123 Life is much easier, if you store the source tree on a drive which is mounted
3125 And use an editor which can handle LF-only line endings.
3126 Make sure not to commit files with windows line endings.
3127 The tabulator spacing
3128 \begin_inset LatexCommand \index{tabulator spacing (8)}
3132 used in the project is 8.
3135 Building SDCC Using Microsoft Visual C++ 6.0/NET (MSVC)
3140 Download the source package
3142 either from the SDCC CVS repository or from the
3143 \begin_inset LatexCommand \url[nightly snapshots]{http://sdcc.sourceforge.net/snap.php}
3149 , it will be named something like sdcc
3156 SDCC is distributed with all the projects, workspaces, and files you need
3157 to build it using Visual C++ 6.0/NET (except for sdcdb.exe which currently
3158 doesn't build under MSVC).
3159 The workspace name is 'sdcc.dsw'.
3160 Please note that as it is now, all the executables are created in a folder
3164 Once built you need to copy the executables from sdcc
3168 bin before running SDCC.
3173 In order to build SDCC with MSVC you need win32 executables of bison.exe,
3174 flex.exe, and gawk.exe.
3175 One good place to get them is
3176 \begin_inset LatexCommand \url[here]{http://unxutils.sourceforge.net}
3184 Download the file UnxUtils
3185 \begin_inset LatexCommand \index{UnxUtils}
3190 Now you have to install the utilities and setup MSVC so it can locate the
3192 Here there are two alternatives (choose one!):
3199 a) Extract UnxUtils.zip to your C:
3201 hard disk PRESERVING the original paths, otherwise bison won't work.
3202 (If you are using WinZip make certain that 'Use folder names' is selected)
3206 b) In the Visual C++ IDE click Tools, Options, select the Directory tab,
3207 in 'Show directories for:' select 'Executable files', and in the directories
3208 window add a new path: 'C:
3218 (As a side effect, you get a bunch of Unix utilities that could be useful,
3219 such as diff and patch.)
3226 This one avoids extracting a bunch of files you may not use, but requires
3231 a) Create a directory were to put the tools needed, or use a directory already
3239 b) Extract 'bison.exe', 'bison.hairy', 'bison.simple', 'flex.exe', and gawk.exe
3240 to such directory WITHOUT preserving the original paths.
3241 (If you are using WinZip make certain that 'Use folder names' is not selected)
3245 c) Rename bison.exe to '_bison.exe'.
3249 d) Create a batch file 'bison.bat' in 'C:
3253 ' and add these lines:
3273 _bison %1 %2 %3 %4 %5 %6 %7 %8 %9
3277 Steps 'c' and 'd' are needed because bison requires by default that the
3278 files 'bison.simple' and 'bison.hairy' reside in some weird Unix directory,
3279 '/usr/local/share/' I think.
3280 So it is necessary to tell bison where those files are located if they
3281 are not in such directory.
3282 That is the function of the environment variables BISON_SIMPLE and BISON_HAIRY.
3286 e) In the Visual C++ IDE click Tools, Options, select the Directory tab,
3287 in 'Show directories for:' select 'Executable files', and in the directories
3288 window add a new path: 'c:
3291 Note that you can use any other path instead of 'c:
3293 util', even the path where the Visual C++ tools are, probably: 'C:
3297 Microsoft Visual Studio
3302 So you don't have to execute step 'e' :)
3306 Open 'sdcc.dsw' in Visual Studio, click 'build all', when it finishes copy
3307 the executables from sdcc
3311 bin, and you can compile using SDCC.
3314 Building SDCC Using Borland
3317 From the sdcc directory, run the command "make -f Makefile.bcc".
3318 This should regenerate all the .exe files in the bin directory except for
3319 sdcdb.exe (which currently doesn't build under Borland C++).
3322 If you modify any source files and need to rebuild, be aware that the dependenci
3323 es may not be correctly calculated.
3324 The safest option is to delete all .obj files and run the build again.
3325 From a Cygwin BASH prompt, this can easily be done with the command (be
3326 sure you are in the sdcc directory):
3336 ( -name '*.obj' -o -name '*.lib' -o -name '*.rul'
3338 ) -print -exec rm {}
3347 or on Windows NT/2000/XP from the command prompt with the command:
3354 del /s *.obj *.lib *.rul
3357 from the sdcc directory.
3360 Windows Install Using a Binary Package
3361 \begin_inset LatexCommand \label{sub:Windows-Install}
3368 Download the binary package from
3369 \begin_inset LatexCommand \url{http://sdcc.sourceforge.net/snap.php}
3373 and unpack it using your favorite unpacking tool (gunzip, WinZip, etc).
3374 This should unpack to a group of sub-directories.
3375 An example directory structure after unpacking the mingw32 package is:
3380 bin for the executables, c:
3388 lib for the include and libraries.
3391 Adjust your environment variable PATH to include the location of the bin
3392 directory or start sdcc using the full path.
3395 Building the Documentation
3398 If the necessary tools (LyX, LaTeX, LaTeX2HTML) are installed it is as easy
3399 as changing into the doc directory and typing
3403 \begin_inset Quotes srd
3407 \begin_inset Quotes srd
3414 You're invited to make changes and additions to this manual (sdcc/doc/sdccman.ly
3417 \begin_inset LatexCommand \url{www.lyx.org}
3421 as editor this is straightforward.
3422 If you want to avoid installing the tools you will have some success with
3423 a bootable Knoppix CD
3424 \begin_inset LatexCommand \url{http://www.knoppix.net}
3429 Prebuilt documentation in html and pdf format is available from
3430 \begin_inset LatexCommand \url{http://sdcc.sourceforge.net/snap.php}
3437 Reading the Documentation
3440 Currently reading the document in pdf format is recommended, as for unknown
3441 reason the hyperlinks are working there whereas in the html version they
3448 If you should know why please drop us a note
3452 You'll find the pdf version at
3453 \begin_inset LatexCommand \url{http://sdcc.sourceforge.net/doc/sdccman.pdf}
3459 This documentation is in some aspects different from a commercial documentation:
3463 It tries to document SDCC for several processor architectures in one document
3464 (commercially these probably would be separate documents/products).
3466 \begin_inset LatexCommand \index{Status of documentation}
3470 currently matches SDCC for mcs51 and DS390 best and does give too few informati
3475 There are many references pointing away from this documentation.
3476 Don't let this distract you.
3478 was a reference like
3479 \begin_inset LatexCommand \url{www.opencores.org}
3483 together with a statement
3484 \begin_inset Quotes sld
3487 some processors which are targetted by SDCC can be implemented in a
3504 \begin_inset LatexCommand \index{fpga (field programmable gate array)}
3509 \begin_inset Quotes srd
3512 we expect you to have a quick look there and come back.
3513 If you read this you are on the right track.
3516 Some sections attribute more space to problems, restrictions and warnings
3517 than to the solution.
3520 The installation section and the section about the debugger is intimidating.
3523 There are still lots of typos and there are more different writing styles
3527 Testing the SDCC Compiler
3528 \begin_inset LatexCommand \label{sec:Testing-the-SDCC}
3535 The first thing you should do after installing your SDCC compiler is to
3551 \begin_inset LatexCommand \index{version}
3558 at the prompt, and the program should run and tell you the version.
3559 If it doesn't run, or gives a message about not finding sdcc program, then
3560 you need to check over your installation.
3561 Make sure that the sdcc bin directory is in your executable search path
3562 defined by the PATH environment setting (
3567 \begin_inset LatexCommand \ref{sub:Install-Trouble-shooting}
3574 Install trouble-shooting for suggestions
3577 Make sure that the sdcc program is in the bin folder, if not perhaps something
3578 did not install correctly.
3586 is commonly installed as described in section
3587 \begin_inset Quotes sld
3590 Install and search paths
3591 \begin_inset Quotes srd
3600 Make sure the compiler works on a very simple example.
3601 Type in the following test.c program using your favorite
3627 Compile this using the following command:
3636 If all goes well, the compiler will generate a test.asm and test.rel file.
3637 Congratulations, you've just compiled your first program with SDCC.
3638 We used the -c option to tell SDCC not to link the generated code, just
3639 to keep things simple for this step.
3647 The next step is to try it with the linker.
3657 If all goes well the compiler will link with the libraries and produce
3658 a test.ihx output file.
3663 (no test.ihx, and the linker generates warnings), then the problem is most
3672 usr/local/share/sdcc/lib directory
3679 \begin_inset LatexCommand \ref{sub:Install-Trouble-shooting}
3686 Install trouble-shooting for suggestions).
3694 The final test is to ensure
3702 header files and libraries.
3703 Edit test.c and change it to the following:
3720 strcpy(str1, "testing");
3727 Compile this by typing
3734 This should generate a test.ihx output file, and it should give no warnings
3735 such as not finding the string.h file.
3736 If it cannot find the string.h file, then the problem is that
3740 cannot find the /usr/local/share/sdcc/include directory
3747 \begin_inset LatexCommand \ref{sub:Install-Trouble-shooting}
3754 Install trouble-shooting section for suggestions).
3772 \begin_inset LatexCommand \index{-\/-print-search-dirs}
3776 to find exactly where SDCC is looking for the include and lib files.
3779 Install Trouble-shooting
3780 \begin_inset LatexCommand \label{sub:Install-Trouble-shooting}
3785 \begin_inset LatexCommand \index{Install trouble-shooting}
3792 If SDCC does not build correctly
3795 A thing to try is starting from scratch by unpacking the .tgz source package
3796 again in an empty directory.
3804 ./configure 2>&1 | tee configure.log
3818 make 2>&1 | tee make.log
3825 If anything goes wrong, you can review the log files to locate the problem.
3826 Or a relevant part of this can be attached to an email that could be helpful
3827 when requesting help from the mailing list.
3831 \begin_inset Quotes sld
3835 \begin_inset Quotes srd
3842 \begin_inset Quotes sld
3846 \begin_inset Quotes srd
3849 command is a script that analyzes your system and performs some configuration
3850 to ensure the source package compiles on your system.
3851 It will take a few minutes to run, and will compile a few tests to determine
3852 what compiler features are installed.
3856 \begin_inset Quotes sld
3860 \begin_inset Quotes srd
3866 This runs the GNU make tool, which automatically compiles all the source
3867 packages into the final installed binary executables.
3871 \begin_inset Quotes sld
3875 \begin_inset Quotes erd
3881 This will install the compiler, other executables libraries and include
3882 files into the appropriate directories.
3884 \begin_inset LatexCommand \ref{sub:Install-paths}
3890 \begin_inset LatexCommand \ref{sub:Search-Paths}
3895 about install and search paths.
3897 On most systems you will need super-user privileges to do this.
3903 SDCC is not just a compiler, but a collection of tools by various developers.
3904 These include linkers, assemblers, simulators and other components.
3905 Here is a summary of some of the components.
3906 Note that the included simulator and assembler have separate documentation
3907 which you can find in the source package in their respective directories.
3908 As SDCC grows to include support for other processors, other packages from
3909 various developers are included and may have their own sets of documentation.
3913 You might want to look at the files which are installed in <installdir>.
3914 At the time of this writing, we find the following programs for gcc-builds:
3918 In <installdir>/bin:
3921 sdcc - The compiler.
3924 sdcpp - The C preprocessor.
3927 asx8051 - The assembler for 8051 type processors.
3934 as-gbz80 - The Z80 and GameBoy Z80 assemblers.
3937 aslink -The linker for 8051 type processors.
3944 link-gbz80 - The Z80 and GameBoy Z80 linkers.
3947 s51 - The ucSim 8051 simulator.
3950 sdcdb - The source debugger.
3953 packihx - A tool to pack (compress) Intel hex files.
3956 In <installdir>/share/sdcc/include
3962 In <installdir>/share/sdcc/lib
3965 the subdirs src and small, large, z80, gbz80 and ds390 with the precompiled
3969 In <installdir>/share/sdcc/doc
3975 As development for other processors proceeds, this list will expand to include
3976 executables to support processors like AVR, PIC, etc.
3982 This is the actual compiler, it in turn uses the c-preprocessor and invokes
3983 the assembler and linkage editor.
3986 sdcpp - The C-Preprocessor
3990 \begin_inset LatexCommand \index{sdcpp (preprocessor)}
3994 is a modified version of the GNU preprocessor.
3995 The C preprocessor is used to pull in #include sources, process #ifdef
3996 statements, #defines and so on.
4007 - The Assemblers and Linkage Editors
4010 This is retargettable assembler & linkage editor, it was developed by Alan
4012 John Hartman created the version for 8051, and I (Sandeep) have made some
4013 enhancements and bug fixes for it to work properly with SDCC.
4020 \begin_inset LatexCommand \index{s51}
4024 is a freeware, opensource simulator developed by Daniel Drotos (
4025 \begin_inset LatexCommand \url{mailto:drdani@mazsola.iit.uni-miskolc.hu}
4030 The simulator is built as part of the build process.
4031 For more information visit Daniel's web site at:
4032 \begin_inset LatexCommand \url{http://mazsola.iit.uni-miskolc.hu/~drdani/embedded/s51}
4037 It currently supports the core mcs51, the Dallas DS80C390 and the Phillips
4041 sdcdb - Source Level Debugger
4045 \begin_inset LatexCommand \index{sdcdb (debugger)}
4049 is the companion source level debugger.
4050 More about sdcdb in section
4051 \begin_inset LatexCommand \ref{cha:Debugging-with-SDCDB}
4056 The current version of the debugger uses Daniel's Simulator S51
4057 \begin_inset LatexCommand \index{s51}
4061 , but can be easily changed to use other simulators.
4071 Single Source File Projects
4074 For single source file 8051 projects the process is very simple.
4075 Compile your programs with the following command
4078 "sdcc sourcefile.c".
4082 This will compile, assemble and link your source file.
4083 Output files are as follows:
4087 \begin_inset LatexCommand \index{<file>.asm}
4092 \begin_inset LatexCommand \index{Assembler source}
4096 file created by the compiler
4100 \begin_inset LatexCommand \index{<file>.lst}
4105 \begin_inset LatexCommand \index{Assembler listing}
4109 file created by the Assembler
4113 \begin_inset LatexCommand \index{<file>.rst}
4118 \begin_inset LatexCommand \index{Assembler listing}
4122 file updated with linkedit information, created by linkage editor
4126 \begin_inset LatexCommand \index{<file>.sym}
4131 \begin_inset LatexCommand \index{Symbol listing}
4135 for the sourcefile, created by the assembler
4139 \begin_inset LatexCommand \index{<file>.rel}
4144 \begin_inset LatexCommand \index{<file>.o}
4149 \begin_inset LatexCommand \index{Object file}
4153 created by the assembler, input to Linkage editor
4157 \begin_inset LatexCommand \index{<file>.map}
4162 \begin_inset LatexCommand \index{Memory map}
4166 for the load module, created by the Linker
4170 \begin_inset LatexCommand \index{<file>.mem}
4174 - A file with a summary of the memory usage
4178 \begin_inset LatexCommand \index{<file>.ihx}
4182 - The load module in Intel hex format
4183 \begin_inset LatexCommand \index{Intel hex format}
4187 (you can select the Motorola S19 format
4188 \begin_inset LatexCommand \index{Motorola S19 format}
4203 \begin_inset LatexCommand \index{-\/-out-fmt-s19}
4208 If you need another format you might want to use
4215 \begin_inset LatexCommand \index{objdump (tool)}
4226 \begin_inset LatexCommand \index{srecord (tool)}
4231 Both formats are documented in the documentation of srecord
4232 \begin_inset LatexCommand \index{srecord (tool)}
4240 \begin_inset LatexCommand \index{<file>.adb}
4244 - An intermediate file containing debug information needed to create the
4256 \begin_inset LatexCommand \index{-\/-debug}
4264 \begin_inset LatexCommand \index{<file>.cdb}
4268 - An optional file (with -
4278 -debug) containing debug information.
4279 The format is documented in cdbfileformat.pdf.
4284 \begin_inset LatexCommand \index{<file> (no extension)}
4288 An optional AOMF or AOMF51
4289 \begin_inset LatexCommand \index{AOMF, AOMF51}
4293 file containing debug information (generated with option -
4320 ormat is commonly used by third party tools (debuggers
4321 \begin_inset LatexCommand \index{Debugger}
4325 , simulators, emulators)
4329 \begin_inset LatexCommand \index{<file>.dump*}
4333 - Dump file to debug the compiler it self (generated with option -
4343 -dumpall) (see section
4344 \begin_inset LatexCommand \ref{sub:Intermediate-Dump-Options}
4350 \begin_inset LatexCommand \ref{sub:The-anatomy-of}
4356 \begin_inset Quotes sld
4359 Anatomy of the compiler
4360 \begin_inset Quotes srd
4366 Projects with Multiple Source Files
4369 SDCC can compile only ONE file at a time.
4370 Let us for example assume that you have a project containing the following
4375 foo1.c (contains some functions)
4377 foo2.c (contains some more functions)
4379 foomain.c (contains more functions and the function main)
4387 The first two files will need to be compiled separately with the commands:
4419 Then compile the source file containing the
4424 \begin_inset LatexCommand \index{Linker}
4428 the files together with the following command:
4436 foomain.c\SpecialChar ~
4437 foo1.rel\SpecialChar ~
4442 \begin_inset LatexCommand \index{<file>.rel}
4454 can be separately compiled as well:
4465 sdcc foomain.rel foo1.rel foo2.rel
4472 The file containing the
4487 file specified in the command line, since the linkage editor processes
4488 file in the order they are presented to it.
4489 The linker is invoked from SDCC using a script file with extension .lnk
4490 \begin_inset LatexCommand \index{<file>.lnk}
4495 You can view this file to troubleshoot linking problems such as those arising
4496 from missing libraries.
4499 Projects with Additional Libraries
4500 \begin_inset LatexCommand \index{Libraries}
4507 Some reusable routines may be compiled into a library, see the documentation
4508 for the assembler and linkage editor (which are in <installdir>/share/sdcc/doc)
4512 \begin_inset LatexCommand \index{<file>.lib}
4519 Libraries created in this manner can be included in the command line.
4520 Make sure you include the -L <library-path> option to tell the linker where
4521 to look for these files if they are not in the current directory.
4522 Here is an example, assuming you have the source file
4534 (if that is not the same as your current project):
4541 sdcc foomain.c foolib.lib -L mylib
4552 must be an absolute path name.
4556 The most efficient way to use libraries is to keep separate modules in separate
4558 The lib file now should name all the modules.rel
4559 \begin_inset LatexCommand \index{<file>.rel}
4564 For an example see the standard library file
4568 in the directory <installdir>/share/lib/small.
4571 Using sdcclib to Create and Manage Libraries
4572 \begin_inset LatexCommand \index{sdcclib}
4579 Alternatively, instead of having a .rel file for each entry on the library
4580 file as described in the preceding section, sdcclib can be used to embed
4581 all the modules belonging to such library in the library file itself.
4582 This results in a larger library file, but it greatly reduces the number
4583 of disk files accessed by the linker.
4584 Additionally, the packed library file contains an index of all include
4585 modules and symbols that significantly speeds up the linking process.
4586 To display a list of options supported by sdcclib type:
4595 \begin_inset LatexCommand \index{sdcclib}
4606 To create a new library file, start by compiling all the required modules.
4639 This will create files _divsint.rel, _divuint.rel, _modsint.rel, _moduint.rel,
4641 The next step is to add the .rel files to the library file:
4648 sdcclib libint.lib _divsint.rel
4649 \begin_inset LatexCommand \index{sdcclib}
4658 sdcclib libint.lib _divuint.rel
4663 sdcclib libint.lib _modsint.rel
4668 sdcclib libint.lib _moduint.rel
4673 sdcclib libint.lib _mulint.rel
4680 If the file already exists in the library, it will be replaced.
4681 To see what modules and symbols are included in the library, options -s
4682 and -m are available.
4689 sdcclib -s libint.lib
4690 \begin_inset LatexCommand \index{sdcclib}
4794 If the source files are compiled using --debug, the corresponding debug
4795 information file .adb will be include in the library file as well.
4796 The library files created with sdcclib are plain text files, so they can
4797 be viewed with a text editor.
4798 It is not recomended to modify a library file created with sdcclib using
4799 a text editor, as there are file indexes numbers located accross the file
4800 used by the linker to quickly locate the required module to link.
4801 Once a .rel file (as well as a .adb file) is added to a library using sdcclib,
4802 it can be safely deleted, since all the information required for linking
4803 is embedded in the library file itself.
4804 Library files created using sdcclib are used as described in the preceding
4808 Command Line Options
4809 \begin_inset LatexCommand \index{Command Line Options}
4816 Processor Selection Options
4817 \begin_inset LatexCommand \index{Options processor selection}
4822 \begin_inset LatexCommand \index{Processor selection options}
4828 \labelwidthstring 00.00.0000
4833 \begin_inset LatexCommand \index{-mmcs51}
4839 Generate code for the Intel MCS51
4840 \begin_inset LatexCommand \index{MCS51}
4844 family of processors.
4845 This is the default processor target.
4847 \labelwidthstring 00.00.0000
4852 \begin_inset LatexCommand \index{-mds390}
4858 Generate code for the Dallas DS80C390
4859 \begin_inset LatexCommand \index{DS80C390}
4865 \labelwidthstring 00.00.0000
4870 \begin_inset LatexCommand \index{-mds400}
4876 Generate code for the Dallas DS80C400
4877 \begin_inset LatexCommand \index{DS80C400}
4883 \labelwidthstring 00.00.0000
4888 \begin_inset LatexCommand \index{-mhc08}
4894 Generate code for the Motorola HC08
4895 \begin_inset LatexCommand \index{HC08}
4899 family of processors (added Oct 2003).
4901 \labelwidthstring 00.00.0000
4906 \begin_inset LatexCommand \index{-mz80}
4912 Generate code for the Zilog Z80
4913 \begin_inset LatexCommand \index{Z80}
4917 family of processors.
4919 \labelwidthstring 00.00.0000
4924 \begin_inset LatexCommand \index{-mgbz80}
4930 Generate code for the GameBoy Z80
4931 \begin_inset LatexCommand \index{gbz80 (GameBoy Z80)}
4935 processor (Not actively maintained).
4937 \labelwidthstring 00.00.0000
4942 \begin_inset LatexCommand \index{-mavr}
4948 Generate code for the Atmel AVR
4949 \begin_inset LatexCommand \index{AVR}
4953 processor (In development, not complete).
4954 AVR users should probably have a look at avr-gcc
4955 \begin_inset LatexCommand \url{ http://savannah.nongnu.org/download/avr-libc/snapshots/}
4962 I think it is fair to direct users there for now.
4963 Open source is also about avoiding unnecessary work .
4964 But I didn't find the 'official' link.
4966 \labelwidthstring 00.00.0000
4971 \begin_inset LatexCommand \index{-mpic14}
4977 Generate code for the Microchip PIC 14
4978 \begin_inset LatexCommand \index{PIC14}
4982 -bit processors (p16f84 and variants.
4983 In development, not complete).
4986 p16f627 p16f628 p16f84 p16f873 p16f877?
4988 \labelwidthstring 00.00.0000
4993 \begin_inset LatexCommand \index{-mpic16}
4999 Generate code for the Microchip PIC 16
5000 \begin_inset LatexCommand \index{PIC16}
5004 -bit processors (p18f452 and variants.
5005 In development, not complete).
5007 \labelwidthstring 00.00.0000
5013 Generate code for the Toshiba TLCS-900H
5014 \begin_inset LatexCommand \index{TLCS-900H}
5018 processor (Not maintained, not complete).
5020 \labelwidthstring 00.00.0000
5025 \begin_inset LatexCommand \index{-mxa51}
5031 Generate code for the Phillips XA51
5032 \begin_inset LatexCommand \index{XA51}
5036 processor (Not maintained, not complete).
5039 Preprocessor Options
5040 \begin_inset LatexCommand \index{Options preprocessor}
5045 \begin_inset LatexCommand \index{Preprocessor options}
5050 \begin_inset LatexCommand \index{sdcpp (preprocessor)}
5056 \labelwidthstring 00.00.0000
5061 \begin_inset LatexCommand \index{-I<path>}
5067 The additional location where the pre processor will look for <..h> or
5068 \begin_inset Quotes eld
5072 \begin_inset Quotes erd
5077 \labelwidthstring 00.00.0000
5082 \begin_inset LatexCommand \index{-D<macro[=value]>}
5088 Command line definition of macros.
5089 Passed to the preprocessor.
5091 \labelwidthstring 00.00.0000
5096 \begin_inset LatexCommand \index{-M}
5102 Tell the preprocessor to output a rule suitable for make describing the
5103 dependencies of each object file.
5104 For each source file, the preprocessor outputs one make-rule whose target
5105 is the object file name for that source file and whose dependencies are
5106 all the files `#include'd in it.
5107 This rule may be a single line or may be continued with `
5109 '-newline if it is long.
5110 The list of rules is printed on standard output instead of the preprocessed
5113 \begin_inset LatexCommand \index{-E}
5119 \labelwidthstring 00.00.0000
5124 \begin_inset LatexCommand \index{-C}
5130 Tell the preprocessor not to discard comments.
5131 Used with the `-E' option.
5133 \labelwidthstring 00.00.0000
5138 \begin_inset LatexCommand \index{-MM}
5149 Like `-M' but the output mentions only the user header files included with
5151 \begin_inset Quotes eld
5155 System header files included with `#include <file>' are omitted.
5157 \labelwidthstring 00.00.0000
5162 \begin_inset LatexCommand \index{-Aquestion(answer)}
5168 Assert the answer answer for question, in case it is tested with a preprocessor
5169 conditional such as `#if #question(answer)'.
5170 `-A-' disables the standard assertions that normally describe the target
5173 \labelwidthstring 00.00.0000
5178 \begin_inset LatexCommand \index{-Umacro}
5184 Undefine macro macro.
5185 `-U' options are evaluated after all `-D' options, but before any `-include'
5186 and `-imacros' options.
5188 \labelwidthstring 00.00.0000
5193 \begin_inset LatexCommand \index{-dM}
5199 Tell the preprocessor to output only a list of the macro definitions that
5200 are in effect at the end of preprocessing.
5201 Used with the `-E' option.
5203 \labelwidthstring 00.00.0000
5208 \begin_inset LatexCommand \index{-dD}
5214 Tell the preprocessor to pass all macro definitions into the output, in
5215 their proper sequence in the rest of the output.
5217 \labelwidthstring 00.00.0000
5222 \begin_inset LatexCommand \index{-dN}
5233 Like `-dD' except that the macro arguments and contents are omitted.
5234 Only `#define name' is included in the output.
5236 \labelwidthstring 00.00.0000
5241 preprocessorOption[,preprocessorOption]
5244 \begin_inset LatexCommand \index{-Wp preprocessorOption[,preprocessorOption]}
5249 Pass the preprocessorOption to the preprocessor
5254 \begin_inset LatexCommand \index{sdcpp (preprocessor)}
5259 SDCC uses an adapted version of the preprocessor cpp of the GNU Compiler
5260 Collection (gcc), if you need more dedicated options please refer to the
5262 \begin_inset LatexCommand \htmlurl{http://www.gnu.org/software/gcc/onlinedocs/}
5270 \begin_inset LatexCommand \index{Options linker}
5275 \begin_inset LatexCommand \index{Linker options}
5281 \labelwidthstring 00.00.0000
5301 \begin_inset LatexCommand \index{-\/-lib-path <path>}
5306 \begin_inset LatexCommand \index{-L -\/-lib-path}
5313 <absolute path to additional libraries> This option is passed to the linkage
5314 editor's additional libraries
5315 \begin_inset LatexCommand \index{Libraries}
5320 The path name must be absolute.
5321 Additional library files may be specified in the command line.
5322 See section Compiling programs for more details.
5324 \labelwidthstring 00.00.0000
5341 \begin_inset LatexCommand \index{-\/-xram-loc <Value>}
5346 <Value> The start location of the external ram
5347 \begin_inset LatexCommand \index{xdata (mcs51, ds390 storage class)}
5351 , default value is 0.
5352 The value entered can be in Hexadecimal or Decimal format, e.g.: -
5362 -xram-loc 0x8000 or -
5374 \labelwidthstring 00.00.0000
5391 \begin_inset LatexCommand \index{-\/-code-loc <Value>}
5396 <Value> The start location of the code
5397 \begin_inset LatexCommand \index{code}
5401 segment, default value 0.
5402 Note when this option is used the interrupt vector table is also relocated
5403 to the given address.
5404 The value entered can be in Hexadecimal or Decimal format, e.g.: -
5414 -code-loc 0x8000 or -
5426 \labelwidthstring 00.00.0000
5443 \begin_inset LatexCommand \index{-\/-stack-loc <Value>}
5448 <Value> By default the stack
5449 \begin_inset LatexCommand \index{stack}
5453 is placed after the data segment.
5454 Using this option the stack can be placed anywhere in the internal memory
5456 The value entered can be in Hexadecimal or Decimal format, e.g.
5467 -stack-loc 0x20 or -
5478 Since the sp register is incremented before a push or call, the initial
5479 sp will be set to one byte prior the provided value.
5480 The provided value should not overlap any other memory areas such as used
5481 register banks or the data segment and with enough space for the current
5484 \labelwidthstring 00.00.0000
5501 \begin_inset LatexCommand \index{-\/-data-loc <Value>}
5506 <Value> The start location of the internal ram data
5507 \begin_inset LatexCommand \index{data (mcs51, ds390 storage class)}
5512 The value entered can be in Hexadecimal or Decimal format, eg.
5534 (By default, the start location of the internal ram data segment is set
5535 as low as possible in memory, taking into account the used register banks
5536 and the bit segment at address 0x20.
5537 For example if register banks 0 and 1 are used without bit variables, the
5538 data segment will be set, if -
5548 -data-loc is not used, to location 0x10.)
5550 \labelwidthstring 00.00.0000
5567 \begin_inset LatexCommand \index{-\/-idata-loc <Value>}
5572 <Value> The start location of the indirectly addressable internal ram
5573 \begin_inset LatexCommand \index{idata (mcs51, ds390 storage class)}
5577 of the 8051, default value is 0x80.
5578 The value entered can be in Hexadecimal or Decimal format, eg.
5589 -idata-loc 0x88 or -
5601 \labelwidthstring 00.00.0000
5618 <Value> The start location of the bit
5619 \begin_inset LatexCommand \index{bit}
5623 addressable internal ram of the 8051.
5629 Instead an option can be passed directly to the linker: -Wl\SpecialChar ~
5632 \labelwidthstring 00.00.0000
5647 \begin_inset LatexCommand \index{-\/-out-fmt-ihx}
5656 The linker output (final object code) is in Intel Hex format.
5657 \begin_inset LatexCommand \index{Intel hex format}
5661 This is the default option.
5662 The format itself is documented in the documentation of srecord
5663 \begin_inset LatexCommand \index{srecord (tool)}
5669 \labelwidthstring 00.00.0000
5684 \begin_inset LatexCommand \index{-\/-out-fmt-s19}
5693 The linker output (final object code) is in Motorola S19 format
5694 \begin_inset LatexCommand \index{Motorola S19 format}
5699 The format itself is documented in the documentation of srecord.
5701 \labelwidthstring 00.00.0000
5706 linkOption[,linkOption]
5709 \begin_inset LatexCommand \index{-Wl linkOption[,linkOption]}
5714 Pass the linkOption to the linker.
5715 See file sdcc/as/doc/asxhtm.html for more on linker options.
5719 \begin_inset LatexCommand \index{Options MCS51}
5724 \begin_inset LatexCommand \index{MCS51 options}
5730 \labelwidthstring 00.00.0000
5745 \begin_inset LatexCommand \index{-\/-model-small}
5756 Generate code for Small Model programs, see section Memory Models for more
5758 This is the default model.
5760 \labelwidthstring 00.00.0000
5775 \begin_inset LatexCommand \index{-\/-model-large}
5781 Generate code for Large model programs, see section Memory Models for more
5783 If this option is used all source files in the project have to be compiled
5786 \labelwidthstring 00.00.0000
5801 \begin_inset LatexCommand \index{-\/-xstack}
5807 Uses a pseudo stack in the first 256 bytes in the external ram for allocating
5808 variables and passing parameters.
5810 \begin_inset LatexCommand \ref{sub:External-Stack}
5815 External Stack for more details.
5817 \labelwidthstring 00.00.0000
5835 \begin_inset LatexCommand \index{-\/-iram-size <Value>}
5839 Causes the linker to check if the internal ram usage is within limits of
5842 \labelwidthstring 00.00.0000
5860 \begin_inset LatexCommand \index{-\/-xram-size <Value>}
5864 Causes the linker to check if the external ram usage is within limits of
5867 \labelwidthstring 00.00.0000
5885 \begin_inset LatexCommand \index{-\/-code-size <Value>}
5889 Causes the linker to check if the code memory usage is within limits of
5893 DS390 / DS400 Options
5894 \begin_inset LatexCommand \index{Options DS390}
5899 \begin_inset LatexCommand \index{DS390 options}
5905 \labelwidthstring 00.00.0000
5922 \begin_inset LatexCommand \index{-\/-model-flat24}
5932 Generate 24-bit flat mode code.
5933 This is the one and only that the ds390 code generator supports right now
5934 and is default when using
5939 See section Memory Models for more details.
5941 \labelwidthstring 00.00.0000
5956 \begin_inset LatexCommand \index{-\/-protect-sp-update}
5962 disable interrupts during ESP:SP updates.
5964 \labelwidthstring 00.00.0000
5981 \begin_inset LatexCommand \index{-\/-stack-10bit}
5985 Generate code for the 10 bit stack mode of the Dallas DS80C390 part.
5986 This is the one and only that the ds390 code generator supports right now
5987 and is default when using
5992 In this mode, the stack is located in the lower 1K of the internal RAM,
5993 which is mapped to 0x400000.
5994 Note that the support is incomplete, since it still uses a single byte
5995 as the stack pointer.
5996 This means that only the lower 256 bytes of the potential 1K stack space
5997 will actually be used.
5998 However, this does allow you to reclaim the precious 256 bytes of low RAM
5999 for use for the DATA and IDATA segments.
6000 The compiler will not generate any code to put the processor into 10 bit
6002 It is important to ensure that the processor is in this mode before calling
6003 any re-entrant functions compiled with this option.
6004 In principle, this should work with the
6017 \begin_inset LatexCommand \index{-\/-stack-auto}
6023 option, but that has not been tested.
6024 It is incompatible with the
6037 \begin_inset LatexCommand \index{-\/-xstack}
6044 It also only makes sense if the processor is in 24 bit contiguous addressing
6057 -model-flat24 option
6061 \labelwidthstring 00.00.0000
6076 \begin_inset LatexCommand \index{-\/-stack-probe}
6082 insert call to function __stack_probe at each function prologue.
6084 \labelwidthstring 00.00.0000
6099 \begin_inset LatexCommand \index{-\/-tini-libid}
6105 <nnnn> LibraryID used in -mTININative.
6108 \labelwidthstring 00.00.0000
6123 \begin_inset LatexCommand \index{-\/-use-accelerator}
6129 generate code for DS390 Arithmetic Accelerator.
6134 \begin_inset LatexCommand \index{Options PIC}
6139 \begin_inset LatexCommand \index{PIC options}
6145 \labelwidthstring 00.00.0000
6160 \begin_inset LatexCommand \index{-\/-gen-banksel}
6166 enable the generation of banksel assembler directives in the PIC16
6167 \begin_inset LatexCommand \index{PIC16}
6175 \begin_inset LatexCommand \index{Options Z80}
6180 \begin_inset LatexCommand \index{Z80 options}
6186 \labelwidthstring 00.00.0000
6203 \begin_inset LatexCommand \index{-\/-callee-saves-bc}
6213 Force a called function to always save BC.
6215 \labelwidthstring 00.00.0000
6232 \begin_inset LatexCommand \index{-\/-no-std-crt0}
6236 When linking, skip the standard crt0.o object file.
6237 You must provide your own crt0.o for your system when linking.
6241 Optimization Options
6242 \begin_inset LatexCommand \index{Options optimization}
6247 \begin_inset LatexCommand \index{Optimization options}
6253 \labelwidthstring 00.00.0000
6268 \begin_inset LatexCommand \index{-\/-nogcse}
6274 Will not do global subexpression elimination, this option may be used when
6275 the compiler creates undesirably large stack/data spaces to store compiler
6277 A warning message will be generated when this happens and the compiler
6278 will indicate the number of extra bytes it allocated.
6279 It is recommended that this option NOT be used, #pragma\SpecialChar ~
6281 \begin_inset LatexCommand \index{\#pragma nogcse}
6285 can be used to turn off global subexpression elimination
6286 \begin_inset LatexCommand \index{Subexpression elimination}
6290 for a given function only.
6292 \labelwidthstring 00.00.0000
6307 \begin_inset LatexCommand \index{-\/-noinvariant}
6313 Will not do loop invariant optimizations, this may be turned off for reasons
6314 explained for the previous option.
6315 For more details of loop optimizations performed see section Loop Invariants.
6316 It is recommended that this option NOT be used, #pragma\SpecialChar ~
6318 \begin_inset LatexCommand \index{\#pragma noinvariant}
6322 can be used to turn off invariant optimizations for a given function only.
6324 \labelwidthstring 00.00.0000
6339 \begin_inset LatexCommand \index{-\/-noinduction}
6345 Will not do loop induction optimizations, see section strength reduction
6347 It is recommended that this option is NOT used, #pragma\SpecialChar ~
6349 \begin_inset LatexCommand \index{\#pragma noinduction}
6353 can be used to turn off induction optimizations for a given function only.
6355 \labelwidthstring 00.00.0000
6370 \begin_inset LatexCommand \index{-\/-nojtbound}
6381 Will not generate boundary condition check when switch statements
6382 \begin_inset LatexCommand \index{switch statement}
6386 are implemented using jump-tables.
6388 \begin_inset LatexCommand \ref{sub:'switch'-Statements}
6393 Switch Statements for more details.
6394 It is recommended that this option is NOT used, #pragma\SpecialChar ~
6396 \begin_inset LatexCommand \index{\#pragma nojtbound}
6400 can be used to turn off boundary checking for jump tables for a given function
6403 \labelwidthstring 00.00.0000
6418 \begin_inset LatexCommand \index{-\/-noloopreverse}
6427 Will not do loop reversal
6428 \begin_inset LatexCommand \index{Loop reversing}
6434 \labelwidthstring 00.00.0000
6451 \begin_inset LatexCommand \index{-\/-nolabelopt }
6455 Will not optimize labels (makes the dumpfiles more readable).
6457 \labelwidthstring 00.00.0000
6472 \begin_inset LatexCommand \index{-\/-no-xinit-opt}
6478 Will not memcpy initialized data from code space into xdata space.
6479 This saves a few bytes in code space if you don't have initialized data.
6481 \labelwidthstring 00.00.0000
6496 \begin_inset LatexCommand \index{-\/-nooverlay}
6502 The compiler will not overlay parameters and local variables of any function,
6503 see section Parameters and local variables for more details.
6505 \labelwidthstring 00.00.0000
6520 \begin_inset LatexCommand \index{-\/-no-peep}
6526 Disable peep-hole optimization.
6528 \labelwidthstring 00.00.0000
6545 \begin_inset LatexCommand \index{-\/-peep-file}
6550 <filename> This option can be used to use additional rules to be used by
6551 the peep hole optimizer.
6553 \begin_inset LatexCommand \ref{sub:Peephole-Optimizer}
6558 Peep Hole optimizations for details on how to write these rules.
6560 \labelwidthstring 00.00.0000
6575 \begin_inset LatexCommand \index{-\/-peep-asm}
6581 Pass the inline assembler code through the peep hole optimizer.
6582 This can cause unexpected changes to inline assembler code, please go through
6583 the peephole optimizer
6584 \begin_inset LatexCommand \index{Peephole optimizer}
6588 rules defined in the source file tree '<target>/peeph.def' before using
6593 \begin_inset LatexCommand \index{Options other}
6599 \labelwidthstring 00.00.0000
6615 \begin_inset LatexCommand \index{-\/-compile-only}
6620 \begin_inset LatexCommand \index{-c -\/-compile-only}
6626 will compile and assemble the source, but will not call the linkage editor.
6628 \labelwidthstring 00.00.0000
6647 \begin_inset LatexCommand \index{-\/-c1mode}
6653 reads the preprocessed source from standard input and compiles it.
6654 The file name for the assembler output must be specified using the -o option.
6656 \labelwidthstring 00.00.0000
6661 \begin_inset LatexCommand \index{-E}
6667 Run only the C preprocessor.
6668 Preprocess all the C source files specified and output the results to standard
6671 \labelwidthstring 00.00.0000
6677 \begin_inset LatexCommand \index{-o <path/file>}
6683 The output path resp.
6684 file where everything will be placed.
6685 If the parameter is a path, it must have a trailing slash (or backslash
6686 for the Windows binaries) to be recognized as a path.
6689 \labelwidthstring 00.00.0000
6704 \begin_inset LatexCommand \index{-\/-stack-auto}
6715 All functions in the source file will be compiled as
6720 \begin_inset LatexCommand \index{reentrant}
6725 the parameters and local variables will be allocated on the stack
6726 \begin_inset LatexCommand \index{stack}
6731 see section Parameters and Local Variables for more details.
6732 If this option is used all source files in the project should be compiled
6736 \labelwidthstring 00.00.0000
6751 \begin_inset LatexCommand \index{-\/-callee-saves}
6755 function1[,function2][,function3]....
6758 The compiler by default uses a caller saves convention for register saving
6759 across function calls, however this can cause unnecessary register pushing
6760 & popping when calling small functions from larger functions.
6761 This option can be used to switch the register saving convention for the
6762 function names specified.
6763 The compiler will not save registers when calling these functions, no extra
6764 code will be generated at the entry & exit (function prologue
6767 \begin_inset LatexCommand \index{function prologue}
6776 \begin_inset LatexCommand \index{function epilogue}
6782 ) for these functions to save & restore the registers used by these functions,
6783 this can SUBSTANTIALLY reduce code & improve run time performance of the
6785 In the future the compiler (with inter procedural analysis) will be able
6786 to determine the appropriate scheme to use for each function call.
6787 DO NOT use this option for built-in functions such as _mulint..., if this
6788 option is used for a library function the appropriate library function
6789 needs to be recompiled with the same option.
6790 If the project consists of multiple source files then all the source file
6791 should be compiled with the same -
6801 -callee-saves option string.
6802 Also see #pragma\SpecialChar ~
6804 \begin_inset LatexCommand \index{\#pragma callee\_saves}
6810 \labelwidthstring 00.00.0000
6825 \begin_inset LatexCommand \index{-\/-debug}
6834 When this option is used the compiler will generate debug information.
6835 The debug information collected in a file with .cdb extension can be used
6837 For more information see documentation for SDCDB.
6838 Another file with no extension contains debug information in AOMF or AOMF51
6839 \begin_inset LatexCommand \index{AOMF, AOMF51}
6843 format which is commonly used by third party tools.
6845 \labelwidthstring 00.00.0000
6850 \begin_inset LatexCommand \index{-S}
6861 Stop after the stage of compilation proper; do not assemble.
6862 The output is an assembler code file for the input file specified.
6864 \labelwidthstring 00.00.0000
6879 \begin_inset LatexCommand \index{-\/-int-long-reent}
6885 Integer (16 bit) and long (32 bit) libraries have been compiled as reentrant.
6886 Note by default these libraries are compiled as non-reentrant.
6887 See section Installation for more details.
6889 \labelwidthstring 00.00.0000
6904 \begin_inset LatexCommand \index{-\/-cyclomatic}
6913 This option will cause the compiler to generate an information message for
6914 each function in the source file.
6915 The message contains some
6919 information about the function.
6920 The number of edges and nodes the compiler detected in the control flow
6921 graph of the function, and most importantly the
6923 cyclomatic complexity
6924 \begin_inset LatexCommand \index{Cyclomatic complexity}
6930 see section on Cyclomatic Complexity for more details.
6932 \labelwidthstring 00.00.0000
6947 \begin_inset LatexCommand \index{-\/-float-reent}
6953 Floating point library is compiled as reentrant
6954 \begin_inset LatexCommand \index{reentrant}
6959 See section Installation for more details.
6961 \labelwidthstring 00.00.0000
6976 \begin_inset LatexCommand \index{-\/-main-return}
6982 This option can be used when the code generated is called by a monitor
6984 The compiler will generate a 'ret' upon return from the 'main'
6985 \begin_inset LatexCommand \index{main return}
6990 The default setting is to lock up i.e.
6997 \labelwidthstring 00.00.0000
7012 \begin_inset LatexCommand \index{-\/-nostdincl}
7018 This will prevent the compiler from passing on the default include path
7019 to the preprocessor.
7021 \labelwidthstring 00.00.0000
7036 \begin_inset LatexCommand \index{-\/-nostdlib}
7042 This will prevent the compiler from passing on the default library
7043 \begin_inset LatexCommand \index{Libraries}
7049 \labelwidthstring 00.00.0000
7064 \begin_inset LatexCommand \index{-\/-verbose}
7070 Shows the various actions the compiler is performing.
7072 \labelwidthstring 00.00.0000
7077 \begin_inset LatexCommand \index{-V}
7083 Shows the actual commands the compiler is executing.
7085 \labelwidthstring 00.00.0000
7100 \begin_inset LatexCommand \index{-\/-no-c-code-in-asm}
7106 Hides your ugly and inefficient c-code from the asm file, so you can always
7107 blame the compiler :).
7109 \labelwidthstring 00.00.0000
7124 \begin_inset LatexCommand \index{-\/-i-code-in-asm}
7130 Include i-codes in the asm file.
7131 Sounds like noise but is most helpful for debugging the compiler itself.
7133 \labelwidthstring 00.00.0000
7148 \begin_inset LatexCommand \index{-\/-less-pedantic}
7154 Disable some of the more pedantic warnings
7155 \begin_inset LatexCommand \index{Warnings}
7159 (jwk burps: please be more specific here, please!).
7160 If you want rather more than less warnings you should consider using a
7161 separate tool dedicated to syntax checking like splint
7162 \begin_inset LatexCommand \url{www.splint.org}
7168 \labelwidthstring 00.00.0000
7183 \begin_inset LatexCommand \index{-\/-print-search-dirs}
7189 Display the directories in the compiler's search path
7191 \labelwidthstring 00.00.0000
7206 \begin_inset LatexCommand \index{-\/-vc}
7212 Display errors and warnings using MSVC style, so you can use SDCC with
7215 \labelwidthstring 00.00.0000
7230 \begin_inset LatexCommand \index{-\/-use-stdout}
7236 Send errors and warnings to stdout instead of stderr.
7238 \labelwidthstring 00.00.0000
7243 asmOption[,asmOption]
7246 \begin_inset LatexCommand \index{-Wa asmOption[,asmOption]}
7251 Pass the asmOption to the assembler
7252 \begin_inset LatexCommand \index{Options assembler}
7257 \begin_inset LatexCommand \index{Assembler options}
7262 See file sdcc/as/doc/asxhtm.html for assembler options.
7265 Intermediate Dump Options
7266 \begin_inset LatexCommand \label{sub:Intermediate-Dump-Options}
7271 \begin_inset LatexCommand \index{Options intermediate dump}
7276 \begin_inset LatexCommand \index{Intermediate dump options}
7283 The following options are provided for the purpose of retargetting and debugging
7285 These provided a means to dump the intermediate code (iCode
7286 \begin_inset LatexCommand \index{iCode}
7290 ) generated by the compiler in human readable form at various stages of
7291 the compilation process.
7292 More on iCodes see chapter
7293 \begin_inset LatexCommand \ref{sub:The-anatomy-of}
7298 \begin_inset Quotes srd
7301 The anatomy of the compiler
7302 \begin_inset Quotes srd
7307 \labelwidthstring 00.00.0000
7322 \begin_inset LatexCommand \index{-\/-dumpraw}
7328 This option will cause the compiler to dump the intermediate code into
7331 <source filename>.dumpraw
7333 just after the intermediate code has been generated for a function, i.e.
7334 before any optimizations are done.
7336 \begin_inset LatexCommand \index{Basic blocks}
7340 at this stage ordered in the depth first number, so they may not be in
7341 sequence of execution.
7343 \labelwidthstring 00.00.0000
7358 \begin_inset LatexCommand \index{-\/-dumpgcse}
7364 Will create a dump of iCode's, after global subexpression elimination
7365 \begin_inset LatexCommand \index{Global subexpression elimination}
7371 <source filename>.dumpgcse.
7373 \labelwidthstring 00.00.0000
7388 \begin_inset LatexCommand \index{-\/-dumpdeadcode}
7394 Will create a dump of iCode's, after deadcode elimination
7395 \begin_inset LatexCommand \index{Dead-code elimination}
7401 <source filename>.dumpdeadcode.
7403 \labelwidthstring 00.00.0000
7418 \begin_inset LatexCommand \index{-\/-dumploop}
7427 Will create a dump of iCode's, after loop optimizations
7428 \begin_inset LatexCommand \index{Loop optimization}
7434 <source filename>.dumploop.
7436 \labelwidthstring 00.00.0000
7451 \begin_inset LatexCommand \index{-\/-dumprange}
7460 Will create a dump of iCode's, after live range analysis
7461 \begin_inset LatexCommand \index{Live range analysis}
7467 <source filename>.dumprange.
7469 \labelwidthstring 00.00.0000
7484 \begin_inset LatexCommand \index{-\/-dumlrange}
7490 Will dump the life ranges
7491 \begin_inset LatexCommand \index{Live range analysis}
7497 \labelwidthstring 00.00.0000
7512 \begin_inset LatexCommand \index{-\/-dumpregassign}
7521 Will create a dump of iCode's, after register assignment
7522 \begin_inset LatexCommand \index{Register assignment}
7528 <source filename>.dumprassgn.
7530 \labelwidthstring 00.00.0000
7545 \begin_inset LatexCommand \index{-\/-dumplrange}
7551 Will create a dump of the live ranges of iTemp's
7553 \labelwidthstring 00.00.0000
7568 \begin_inset LatexCommand \index{-\/-dumpall}
7579 Will cause all the above mentioned dumps to be created.
7582 Redirecting output on Windows Shells
7585 By default SDCC writes it's error messages to
7586 \begin_inset Quotes sld
7590 \begin_inset Quotes srd
7594 To force all messages to
7595 \begin_inset Quotes sld
7599 \begin_inset Quotes srd
7623 \begin_inset LatexCommand \index{-\/-use-stdout}
7628 Additionally, if you happen to have visual studio installed in your windows
7629 machine, you can use it to compile your sources using a custom build and
7645 \begin_inset LatexCommand \index{-\/-vc}
7650 Something like this should work:
7694 -model-large -c $(InputPath)
7697 Environment variables
7698 \begin_inset LatexCommand \index{Environment variables}
7705 SDCC recognizes the following environment variables:
7707 \labelwidthstring 00.00.0000
7712 \begin_inset LatexCommand \index{SDCC\_LEAVE\_SIGNALS}
7718 SDCC installs a signal handler
7719 \begin_inset LatexCommand \index{signal handler}
7723 to be able to delete temporary files after an user break (^C) or an exception.
7724 If this environment variable is set, SDCC won't install the signal handler
7725 in order to be able to debug SDCC.
7727 \labelwidthstring 00.00.0000
7734 \begin_inset LatexCommand \index{TMP, TEMP, TMPDIR}
7740 Path, where temporary files will be created.
7741 The order of the variables is the search order.
7742 In a standard *nix environment these variables are not set, and there's
7743 no need to set them.
7744 On Windows it's recommended to set one of them.
7746 \labelwidthstring 00.00.0000
7751 \begin_inset LatexCommand \index{SDCC\_HOME}
7758 \begin_inset LatexCommand \ref{sub:Install-paths}
7764 \begin_inset Quotes sld
7768 \begin_inset Quotes srd
7773 \labelwidthstring 00.00.0000
7778 \begin_inset LatexCommand \index{SDCC\_INCLUDE}
7785 \begin_inset LatexCommand \ref{sub:Search-Paths}
7791 \begin_inset Quotes sld
7795 \begin_inset Quotes srd
7800 \labelwidthstring 00.00.0000
7805 \begin_inset LatexCommand \index{SDCC\_LIB}
7812 \begin_inset LatexCommand \ref{sub:Search-Paths}
7818 \begin_inset Quotes sld
7822 \begin_inset Quotes srd
7828 There are some more environment variables recognized by SDCC, but these
7829 are solely used for debugging purposes.
7830 They can change or disappear very quickly, and will never be documented.
7833 Storage Class Language Extensions
7836 MCS51/DS390 Storage Class
7837 \begin_inset LatexCommand \index{Storage class}
7844 In addition to the ANSI storage classes SDCC allows the following MCS51
7845 specific storage classes:
7846 \layout Subsubsection
7849 \begin_inset LatexCommand \index{data (mcs51, ds390 storage class)}
7854 \begin_inset LatexCommand \index{near (storage class)}
7865 storage class for the Small Memory model (
7873 can be used synonymously).
7874 Variables declared with this storage class will be allocated in the directly
7875 addressable portion of the internal RAM of a 8051, e.g.:
7880 data unsigned char test_data;
7883 Writing 0x01 to this variable generates the assembly code:
7888 75*00 01\SpecialChar ~
7894 \layout Subsubsection
7897 \begin_inset LatexCommand \index{xdata (mcs51, ds390 storage class)}
7902 \begin_inset LatexCommand \index{far (storage class)}
7909 Variables declared with this storage class will be placed in the external
7915 storage class for the Large Memory model, e.g.:
7920 xdata unsigned char test_xdata;
7923 Writing 0x01 to this variable generates the assembly code:
7928 90s00r00\SpecialChar ~
7957 \layout Subsubsection
7960 \begin_inset LatexCommand \index{idata (mcs51, ds390 storage class)}
7967 Variables declared with this storage class will be allocated into the indirectly
7968 addressable portion of the internal ram of a 8051, e.g.:
7973 idata unsigned char test_idata;
7976 Writing 0x01 to this variable generates the assembly code:
8005 Please note, the first 128 byte of idata physically access the same RAM
8007 The original 8051 had 128 byte idata memory, nowadays most devices have
8008 256 byte idata memory.
8010 \begin_inset LatexCommand \index{stack}
8014 is located in idata memory.
8015 \layout Subsubsection
8018 \begin_inset LatexCommand \index{pdata (mcs51, ds390 storage class)}
8025 Paged xdata access is currently not as straightforward as using the other
8026 addressing modes of a 8051.
8027 The following example writes 0x01 to the address pointed to.
8028 Please note, pdata access physically accesses xdata memory.
8029 The high byte of the address is determined by port P2
8030 \begin_inset LatexCommand \index{P2 (mcs51 sfr)}
8034 (or in case of some 8051 variants by a separate Special Function Register,
8036 \begin_inset LatexCommand \ref{sub:MCS51-variants}
8045 pdata unsigned char *test_pdata_ptr;
8057 test_pdata_ptr = (pdata *)0xfe;
8063 *test_pdata_ptr = 1;
8068 Generates the assembly code:
8073 75*01 FE\SpecialChar ~
8077 _test_pdata_ptr,#0xFE
8109 Be extremely carefull if you use pdata together with the -
8120 \begin_inset LatexCommand \index{-\/-xstack}
8125 \layout Subsubsection
8128 \begin_inset LatexCommand \index{code}
8135 'Variables' declared with this storage class will be placed in the code
8141 code unsigned char test_code;
8144 Read access to this variable generates the assembly code:
8149 90s00r6F\SpecialChar ~
8152 mov dptr,#_test_code
8181 indexed arrays of characters in code memory can be accessed efficiently:
8186 code char test_array[] = {'c','h','e','a','p'};
8189 Read access to this array using an 8-bit unsigned index generates the assembly
8206 90s00r41\SpecialChar ~
8209 mov dptr,#_test_array
8224 \layout Subsubsection
8227 \begin_inset LatexCommand \index{bit}
8234 This is a data-type and a storage class specifier.
8235 When a variable is declared as a bit, it is allocated into the bit addressable
8236 memory of 8051, e.g.:
8244 Writing 1 to this variable generates the assembly code:
8260 The bit addressable memory consists of 128 bits which are located from 0x20
8261 to 0x2f in data memory.
8265 Apart from this 8051 specific storage class most architectures support ANSI-C
8267 \begin_inset LatexCommand \index{bitfields}
8277 Not really meant as examples, but nevertheless showing what bitfields are
8278 about: device/include/mc68hc908qy.h and support/regression/tests/bitfields.c
8282 \layout Subsubsection
8285 \begin_inset LatexCommand \index{sfr}
8290 \begin_inset LatexCommand \index{sbit}
8297 Like the bit keyword,
8301 signifies both a data-type and storage class, they are used to describe
8322 variables of a 8051, eg:
8328 \begin_inset LatexCommand \index{at}
8332 0x80 P0;\SpecialChar ~
8333 /* special function register P0 at location 0x80 */
8335 sbit at 0xd7 CY; /* CY (Carry Flag
8336 \begin_inset LatexCommand \index{Flags}
8341 \begin_inset LatexCommand \index{Carry flag}
8348 Special function registers which are located on an address dividable by
8349 8 are bit-addressable, an
8353 addresses a specific bit within these sfr.
8354 \layout Subsubsection
8357 \begin_inset LatexCommand \index{Pointer}
8361 to MCS51/DS390 specific memory spaces
8364 SDCC allows (via language extensions) pointers to explicitly point to any
8365 of the memory spaces
8366 \begin_inset LatexCommand \index{Memory model}
8371 In addition to the explicit pointers, the compiler uses (by default) generic
8372 pointers which can be used to point to any of the memory spaces.
8376 Pointer declaration examples:
8381 /* pointer physically in internal ram pointing to object in external ram
8384 xdata unsigned char * data p;
8388 /* pointer physically in external ram pointing to object in internal ram
8391 data unsigned char * xdata p;
8395 /* pointer physically in code rom pointing to data in xdata space */
8397 xdata unsigned char * code p;
8401 /* pointer physically in code space pointing to data in code space */
8403 code unsigned char * code p;
8407 /* the following is a generic pointer physically located in xdata space
8413 Well you get the idea.
8418 All unqualified pointers are treated as 3-byte (4-byte for the ds390)
8431 The highest order byte of the
8435 pointers contains the data space information.
8436 Assembler support routines are called whenever data is stored or retrieved
8442 These are useful for developing reusable library
8443 \begin_inset LatexCommand \index{Libraries}
8448 Explicitly specifying the pointer type will generate the most efficient
8450 \layout Subsubsection
8452 Notes on MCS51 memory
8453 \begin_inset LatexCommand \index{MCS51 memory}
8460 The 8051 family of microcontrollers have a minimum of 128 bytes of internal
8461 RAM memory which is structured as follows:
8465 - Bytes 00-1F - 32 bytes to hold up to 4 banks of the registers R0 to R7,
8468 - Bytes 20-2F - 16 bytes to hold 128 bit
8469 \begin_inset LatexCommand \index{bit}
8475 - Bytes 30-7F - 80 bytes for general purpose use.
8480 Additionally some members of the MCS51 family may have up to 128 bytes of
8481 additional, indirectly addressable, internal RAM memory (
8486 \begin_inset LatexCommand \index{idata (mcs51, ds390 storage class)}
8491 Furthermore, some chips may have some built in external memory (
8496 \begin_inset LatexCommand \index{xdata (mcs51, ds390 storage class)}
8500 ) which should not be confused with the internal, directly addressable RAM
8506 \begin_inset LatexCommand \index{data (mcs51, ds390 storage class)}
8511 Sometimes this built in
8515 memory has to be activated before using it (you can probably find this
8516 information on the datasheet of the microcontroller your are using, see
8518 \begin_inset LatexCommand \ref{sub:Startup-Code}
8526 Normally SDCC will only use the first bank
8527 \begin_inset LatexCommand \index{register bank (mcs51, ds390)}
8531 of registers (register bank 0), but it is possible to specify that other
8532 banks of registers should be used in interrupt
8533 \begin_inset LatexCommand \index{interrupt}
8538 By default, the compiler will place the stack after the last byte of allocated
8539 memory for variables.
8540 For example, if the first 2 banks of registers are used, and only four
8545 variables, it will position the base of the internal stack at address 20
8547 This implies that as the stack
8548 \begin_inset LatexCommand \index{stack}
8552 grows, it will use up the remaining register banks, and the 16 bytes used
8553 by the 128 bit variables, and 80 bytes for general purpose use.
8554 If any bit variables are used, the data variables will be placed after
8555 the byte holding the last bit variable.
8556 For example, if register banks 0 and 1 are used, and there are 9 bit variables
8561 variables will be placed starting at address 0x22.
8573 \begin_inset LatexCommand \index{-\/-data-loc<Value>}
8577 to specify the start address of the
8591 -iram-size to specify the size of the total internal RAM (
8603 By default the 8051 linker will place the stack after the last byte of data
8616 \begin_inset LatexCommand \index{-\/-stack-loc<Value>}
8620 allows you to specify the start of the stack, i.e.
8621 you could start it after any data in the general purpose area.
8622 If your microcontroller has additional indirectly addressable internal
8627 ) you can place the stack on it.
8628 You may also need to use -
8639 \begin_inset LatexCommand \index{-\/-xdata-loc<Value>}
8643 to set the start address of the external RAM (
8658 \begin_inset LatexCommand \index{-\/-data-loc}
8662 to specify its size.
8663 Same goes for the code memory, using -
8674 \begin_inset LatexCommand \index{-\/-data-loc}
8689 \begin_inset LatexCommand \index{-\/-data-loc}
8694 If in doubt, don't specify any options and see if the resulting memory
8695 layout is appropriate, then you can adjust it.
8698 The linker generates two files with memory allocation information.
8699 The first, with extension .map
8700 \begin_inset LatexCommand \index{<file>.map}
8704 shows all the variables and segments.
8705 The second with extension .mem
8706 \begin_inset LatexCommand \index{<file>.mem}
8710 shows the final memory layout.
8711 The linker will complain either if memory segments overlap, there is not
8712 enough memory, or there is not enough space for stack.
8713 If you get any linking warnings and/or errors related to stack or segments
8714 allocation, take a look at either the .map or .mem files to find out what
8716 The .mem file may even suggest a solution to the problem.
8719 Z80/Z180 Storage Class
8720 \begin_inset LatexCommand \index{Storage class}
8725 \layout Subsubsection
8728 \begin_inset LatexCommand \index{sfr}
8732 (in/out to 8-bit addresses)
8736 \begin_inset LatexCommand \index{Z80}
8740 family has separate address spaces for memory and
8750 \begin_inset LatexCommand \index{I/O memory (Z80, Z180)}
8754 is accessed with special instructions, e.g.:
8759 sfr at 0x78 IoPort;\SpecialChar ~
8761 /* define a var in I/O space at 78h called IoPort */
8765 Writing 0x01 to this variable generates the assembly code:
8785 \layout Subsubsection
8788 \begin_inset LatexCommand \index{sfr}
8792 (in/out to 16-bit addresses)
8799 is used to support 16 bit addresses in I/O memory e.g.:
8805 \begin_inset LatexCommand \index{at}
8812 Writing 0x01 to this variable generates the assembly code:
8817 01 23 01\SpecialChar ~
8837 \layout Subsubsection
8840 \begin_inset LatexCommand \index{sfr}
8844 (in0/out0 to 8 bit addresses on Z180
8845 \begin_inset LatexCommand \index{Z180}
8850 \begin_inset LatexCommand \index{HD64180}
8857 The compiler option -
8867 -portmode=180 (80) and a compiler #pragma\SpecialChar ~
8869 \begin_inset LatexCommand \index{\#pragma portmode}
8873 =z180 (z80) is used to turn on (off) the Z180/HD64180 port addressing instructio
8883 If you include the file z180.h this will be set automatically.
8887 \begin_inset LatexCommand \index{Absolute addressing}
8894 Data items can be assigned an absolute address with the
8897 \begin_inset LatexCommand \index{at}
8903 keyword, in addition to a storage class, e.g.:
8909 \begin_inset LatexCommand \index{xdata (mcs51, ds390 storage class)}
8914 \begin_inset LatexCommand \index{at}
8918 0x7ffe unsigned int chksum;
8921 In the above example the variable chksum will located at 0x7ffe and 0x7fff
8922 of the external ram.
8927 reserve any space for variables declared in this way (they are implemented
8928 with an equate in the assembler).
8929 Thus it is left to the programmer to make sure there are no overlaps with
8930 other variables that are declared without the absolute address.
8931 The assembler listing file (.lst
8932 \begin_inset LatexCommand \index{<file>.lst}
8936 ) and the linker output files (.rst
8937 \begin_inset LatexCommand \index{<file>.rst}
8942 \begin_inset LatexCommand \index{<file>.map}
8946 ) are good places to look for such overlaps.
8947 Variables with an absolute address are
8952 \begin_inset LatexCommand \index{Variable initialization}
8959 In case of memory mapped I/O devices the keyword
8963 should be used to tell the compiler that accesses might not be optimized
8970 \begin_inset LatexCommand \index{volatile}
8975 \begin_inset LatexCommand \index{xdata (mcs51, ds390 storage class)}
8980 \begin_inset LatexCommand \index{at}
8984 0x8000 unsigned char PORTA_8255;
8987 For some architectures (mcs51) array accesses are more efficient if an (xdata/fa
8992 \begin_inset LatexCommand \index{Aligned array}
8999 starts at a block (256 byte) boundary
9000 \begin_inset LatexCommand \index{block boundary}
9005 \begin_inset LatexCommand \ref{sub:A-Step-by Assembler Introduction}
9011 Absolute addresses can be specified for variables in all storage classes,
9018 \begin_inset LatexCommand \index{bit}
9023 \begin_inset LatexCommand \index{at}
9030 The above example will allocate the variable at offset 0x02 in the bit-addressab
9032 There is no real advantage to assigning absolute addresses to variables
9033 in this manner, unless you want strict control over all the variables allocated.
9034 One possible use would be to write hardware portable code.
9035 For example, if you have a routine that uses one or more of the microcontroller
9036 I/O pins, and such pins are different for two different hardwares, you
9037 can declare the I/O pins in your routine using:
9042 extern volatile bit SDI;
9044 extern volatile bit SCLK;
9046 extern volatile bit CPOL;
9050 void DS1306_put(unsigned char value)
9058 unsigned char mask=0x80;
9082 SDI=(value & mask)?1:0;
9123 Then, someplace in the code for the first hardware you would use
9128 bit at 0x80 SDI;\SpecialChar ~
9132 /* I/O port 0, bit 0 */
9134 bit at 0x81 SCLK;\SpecialChar ~
9137 /* I/O port 0, bit 1 */
9139 bit CPOL;\SpecialChar ~
9150 /* This is a variable, let the linker allocate this one */
9153 Similarly, for the second hardware you would use
9158 bit at 0x83 SDI;\SpecialChar ~
9162 /* I/O port 0, bit 3 */
9164 bit at 0x91 SCLK;\SpecialChar ~
9167 /* I/O port 1, bit 1 */
9170 \begin_inset LatexCommand \index{bit}
9185 /* This is a variable, let the linker allocate this one */
9188 and you can use the same hardware dependent routine without changes, as
9189 for example in a library.
9190 This is somehow similar to sbit, but only one absolute address has to be
9191 specified in the whole project.
9195 \begin_inset LatexCommand \index{Parameters}
9200 \begin_inset LatexCommand \index{function parameter}
9205 \begin_inset LatexCommand \index{local variables}
9212 Automatic (local) variables and parameters to functions can either be placed
9213 on the stack or in data-space.
9214 The default action of the compiler is to place these variables in the internal
9215 RAM (for small model) or external RAM (for large model).
9216 This in fact makes them similar to
9219 \begin_inset LatexCommand \index{static}
9225 so by default functions are non-reentrant
9226 \begin_inset LatexCommand \index{reentrant}
9235 They can be placed on the stack
9236 \begin_inset LatexCommand \index{stack}
9253 \begin_inset LatexCommand \index{-\/-stack-auto}
9259 option or by using the
9262 \begin_inset LatexCommand \index{reentrant}
9268 keyword in the function declaration, e.g.:
9273 unsigned char foo(char i) reentrant
9287 Since stack space on 8051 is limited, the
9305 option should be used sparingly.
9306 Note that the reentrant keyword just means that the parameters & local
9307 variables will be allocated to the stack, it
9311 mean that the function is register bank
9312 \begin_inset LatexCommand \index{register bank (mcs51, ds390)}
9321 \begin_inset LatexCommand \index{local variables}
9325 can be assigned storage classes and absolute
9326 \begin_inset LatexCommand \index{Absolute addressing}
9343 xdata unsigned char i;
9356 \begin_inset LatexCommand \index{at}
9360 0x31 unsigned char j;
9372 In the above example the variable
9376 will be allocated in the external ram,
9380 in bit addressable space and
9399 or when a function is declared as
9403 this should only be done for static variables.
9407 \begin_inset LatexCommand \index{function parameter}
9411 however are not allowed any storage class
9412 \begin_inset LatexCommand \index{Storage class}
9416 , (storage classes for parameters will be ignored), their allocation is
9417 governed by the memory model in use, and the reentrancy options.
9421 \begin_inset LatexCommand \label{sub:Overlaying}
9426 \begin_inset LatexCommand \index{Overlaying}
9434 \begin_inset LatexCommand \index{reentrant}
9438 functions SDCC will try to reduce internal ram space usage by overlaying
9439 parameters and local variables of a function (if possible).
9440 Parameters and local variables
9441 \begin_inset LatexCommand \index{local variables}
9445 of a function will be allocated to an overlayable segment if the function
9448 no other function calls and the function is non-reentrant and the memory
9450 \begin_inset LatexCommand \index{Memory model}
9457 If an explicit storage class
9458 \begin_inset LatexCommand \index{Storage class}
9462 is specified for a local variable, it will NOT be overlayed.
9465 Note that the compiler (not the linkage editor) makes the decision for overlayin
9467 Functions that are called from an interrupt service routine should be preceded
9468 by a #pragma\SpecialChar ~
9470 \begin_inset LatexCommand \index{\#pragma nooverlay}
9474 if they are not reentrant.
9477 Also note that the compiler does not do any processing of inline assembler
9478 code, so the compiler might incorrectly assign local variables and parameters
9479 of a function into the overlay segment if the inline assembler code calls
9480 other c-functions that might use the overlay.
9481 In that case the #pragma\SpecialChar ~
9482 nooverlay should be used.
9485 Parameters and local variables of functions that contain 16 or 32 bit multiplica
9487 \begin_inset LatexCommand \index{Multiplication}
9492 \begin_inset LatexCommand \index{Division}
9496 will NOT be overlayed since these are implemented using external functions,
9505 \begin_inset LatexCommand \index{\#pragma nooverlay}
9511 void set_error(unsigned char errcd)
9527 void some_isr () interrupt
9528 \begin_inset LatexCommand \index{interrupt}
9558 In the above example the parameter
9566 would be assigned to the overlayable segment if the #pragma\SpecialChar ~
9568 not present, this could cause unpredictable runtime behavior when called
9569 from an interrupt service routine.
9570 The #pragma\SpecialChar ~
9571 nooverlay ensures that the parameters and local variables for
9572 the function are NOT overlayed.
9575 Interrupt Service Routines
9576 \begin_inset LatexCommand \label{sub:Interrupt-Service-Routines}
9595 outines to be coded in C, with some extended keywords.
9600 void timer_isr (void) interrupt 1 using 1
9614 The optional number following the
9617 \begin_inset LatexCommand \index{interrupt}
9623 keyword is the interrupt number this routine will service.
9624 When present, the compiler will insert a call to this routine in the interrupt
9625 vector table for the interrupt number specified.
9626 If you have multiple source files in your project, interrupt service routines
9627 can be present in any of them, but a prototype of the isr MUST be present
9628 or included in the file that contains the function
9637 keyword can be used to tell the compiler to use the specified register
9638 bank (8051 specific) when generating code for this function.
9644 Interrupt numbers and the corresponding address & descriptions for the Standard
9645 8051/8052 are listed below.
9646 SDCC will automatically adjust the interrupt vector table to the maximum
9647 interrupt number specified.
9653 \begin_inset Tabular
9654 <lyxtabular version="3" rows="7" columns="3">
9656 <column alignment="center" valignment="top" leftline="true" width="0in">
9657 <column alignment="center" valignment="top" leftline="true" width="0in">
9658 <column alignment="center" valignment="top" leftline="true" rightline="true" width="0in">
9659 <row topline="true" bottomline="true">
9660 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
9668 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
9676 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
9685 <row topline="true">
9686 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
9694 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
9702 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
9711 <row topline="true">
9712 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
9720 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
9728 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
9737 <row topline="true">
9738 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
9746 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
9754 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
9763 <row topline="true">
9764 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
9772 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
9780 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
9789 <row topline="true">
9790 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
9798 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
9806 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
9815 <row topline="true" bottomline="true">
9816 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
9824 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
9832 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
9850 If the interrupt service routine is defined without
9853 \begin_inset LatexCommand \index{using (mcs51, ds390 register bank)}
9859 a register bank or with register bank 0 (
9863 0), the compiler will save the registers used by itself on the stack upon
9864 entry and restore them at exit, however if such an interrupt service routine
9865 calls another function then the entire register bank will be saved on the
9867 This scheme may be advantageous for small interrupt service routines which
9868 have low register usage.
9871 If the interrupt service routine is defined to be using a specific register
9876 & psw are saved and restored, if such an interrupt service routine calls
9877 another function (using another register bank) then the entire register
9878 bank of the called function will be saved on the stack.
9879 This scheme is recommended for larger interrupt service routines.
9882 Interrupt service routines open the door for some very interesting bugs:
9886 If the interrupt service routines changes variables which are accessed by
9887 other functions these variables should be declared
9892 \begin_inset LatexCommand \index{volatile}
9900 If the access to these variables is not
9903 \begin_inset LatexCommand \index{atomic access}
9910 the processor needs more than one instruction for the access and could
9911 be interrupted while accessing the variable) the interrupt must disabled
9912 during the access to avoid inconsistent data.
9913 Access to 16 or 32 bit variables is obviously not atomic on 8 bit CPUs
9914 and should be protected by disabling interrupts.
9915 You're not automatically on the safe side if you use 8 bit variables though.
9916 We need an example here: f.e.
9917 on the 8051 the harmless looking
9918 \begin_inset Quotes srd
9928 \begin_inset Quotes sld
9937 \begin_inset Quotes srd
9947 \begin_inset Quotes sld
9950 from within an interrupt routine might get lost if the interrupt occurs
9953 \begin_inset Quotes sld
9958 counter\SpecialChar ~
9963 \begin_inset Quotes srd
9966 is not atomic on the 8051 even if
9970 is located in data memory.
9971 Bugs like these are hard to reproduce and can cause a lot of trouble.
9975 A special note here, int (16 bit) and long (32 bit) integer division
9976 \begin_inset LatexCommand \index{Division}
9981 \begin_inset LatexCommand \index{Multiplication}
9986 \begin_inset LatexCommand \index{Modulus}
9991 \begin_inset LatexCommand \index{Floating point support}
9995 operations are implemented using external support routines developed in
9997 If an interrupt service routine needs to do any of these operations then
9998 the support routines (as mentioned in a following section) will have to
9999 be recompiled using the
10012 \begin_inset LatexCommand \index{-\/-stack-auto}
10018 option and the source file will need to be compiled using the
10033 \begin_inset LatexCommand \index{-\/-int-long-reent}
10040 Calling other functions from an interrupt service routine is not recommended,
10041 avoid it if possible.
10042 Note that when some function is called from an interrupt service routine
10043 it should be preceded by a #pragma\SpecialChar ~
10045 \begin_inset LatexCommand \index{\#pragma nooverlay}
10049 if it is not reentrant.
10050 Furthermore nonreentrant functions should not be called from the main program
10051 while the interrupt service routine might be active.
10057 \begin_inset LatexCommand \ref{sub:Overlaying}
10062 about Overlaying and section
10063 \begin_inset LatexCommand \ref{sub:Functions-using-private-banks}
10068 about Functions using private register banks.
10071 Enabling and Disabling Interrupts
10074 Critical Functions and Critical Statements
10077 A special keyword may be associated with a block or a function declaring
10083 SDCC will generate code to disable all interrupts
10084 \begin_inset LatexCommand \index{interrupt}
10088 upon entry to a critical function and restore the interrupt enable to the
10089 previous state before returning.
10090 Nesting critical functions will need one additional byte on the stack
10091 \begin_inset LatexCommand \index{stack}
10100 int foo () critical
10101 \begin_inset LatexCommand \index{critical}
10126 The critical attribute maybe used with other attributes like
10136 may also be used to disable interrupts more locally:
10144 More than one statement could have been included in the block.
10147 Enabling and Disabling Interrupts directly
10151 \begin_inset LatexCommand \index{interrupt}
10155 can also be disabled and enabled directly (8051):
10160 EA = 0;\SpecialChar ~
10223 EA = 1;\SpecialChar ~
10290 On other architectures which have seperate opcodes for enabling and disabling
10291 interrupts you might want to make use of defines with inline assembly
10292 \begin_inset LatexCommand \index{Assembler routines}
10302 \begin_inset LatexCommand \index{\_asm}
10311 \begin_inset LatexCommand \index{\_endasm}
10320 #define SEI _asm\SpecialChar ~
10332 Note: it is sometimes sufficient to disable only a specific interrupt source
10334 a timer or serial interrupt by manipulating an
10337 \begin_inset LatexCommand \index{interrupt mask}
10347 Usually the time during which interrupts are disabled should be kept as
10349 This minimizes both
10354 \begin_inset LatexCommand \index{interrupt latency}
10358 (the time between the occurrence of the interrupt and the execution of
10359 the first code in the interrupt routine) and
10364 \begin_inset LatexCommand \index{interrupt jitter}
10368 (the difference between the shortest and the longest interrupt latency).
10369 These really are something different, f.e.
10370 a serial interrupt has to be served before its buffer overruns so it cares
10371 for the maximum interrupt latency, whereas it does not care about jitter.
10372 On a loudspeaker driven via a digital to analog converter which is fed
10373 by an interrupt a latency of a few milliseconds might be tolerable, whereas
10374 a much smaller jitter will be very audible.
10377 You can reenable interrupts within an interrupt routine and on some architecture
10378 s you can make use of two (or more) levels of
10380 interrupt priorities
10383 \begin_inset LatexCommand \index{interrupt priority}
10388 On some architectures which don't support interrupt priorities these can
10389 be implemented by manipulating the interrupt mask and reenabling interrupts
10390 within the interrupt routine.
10391 Don't add complexity unless you have to.
10395 Functions using private register banks
10396 \begin_inset LatexCommand \label{sub:Functions-using-private-banks}
10403 Some architectures have support for quickly changing register sets.
10404 SDCC supports this feature with the
10407 \begin_inset LatexCommand \index{using (mcs51, ds390 register bank)}
10413 attribute (which tells the compiler to use a register bank
10414 \begin_inset LatexCommand \index{register bank (mcs51, ds390)}
10418 other than the default bank zero).
10419 It should only be applied to
10422 \begin_inset LatexCommand \index{interrupt}
10428 functions (see footnote below).
10429 This will in most circumstances make the generated ISR code more efficient
10430 since it will not have to save registers on the stack.
10437 attribute will have no effect on the generated code for a
10441 function (but may occasionally be useful anyway
10447 possible exception: if a function is called ONLY from 'interrupt' functions
10448 using a particular bank, it can be declared with the same 'using' attribute
10449 as the calling 'interrupt' functions.
10450 For instance, if you have several ISRs using bank one, and all of them
10451 call memcpy(), it might make sense to create a specialized version of memcpy()
10452 'using 1', since this would prevent the ISR from having to save bank zero
10453 to the stack on entry and switch to bank zero before calling the function
10460 (pending: I don't think this has been done yet)
10467 function using a non-zero bank will assume that it can trash that register
10468 bank, and will not save it.
10469 Since high-priority interrupts
10470 \begin_inset LatexCommand \index{interrupt priority}
10474 can interrupt low-priority ones on the 8051 and friends, this means that
10475 if a high-priority ISR
10479 a particular bank occurs while processing a low-priority ISR
10483 the same bank, terrible and bad things can happen.
10484 To prevent this, no single register bank should be
10488 by both a high priority and a low priority ISR.
10489 This is probably most easily done by having all high priority ISRs use
10490 one bank and all low priority ISRs use another.
10491 If you have an ISR which can change priority at runtime, you're on your
10492 own: I suggest using the default bank zero and taking the small performance
10496 It is most efficient if your ISR calls no other functions.
10497 If your ISR must call other functions, it is most efficient if those functions
10498 use the same bank as the ISR (see note 1 below); the next best is if the
10499 called functions use bank zero.
10500 It is very inefficient to call a function using a different, non-zero bank
10506 \begin_inset LatexCommand \label{sub:Startup-Code}
10511 \begin_inset LatexCommand \index{Startup code}
10518 MCS51/DS390 Startup Code
10521 The compiler inserts a call to the C routine
10523 _sdcc_external_startup()
10524 \begin_inset LatexCommand \index{\_sdcc\_external\_startup()}
10533 at the start of the CODE area.
10534 This routine is in the runtime library
10535 \begin_inset LatexCommand \index{Runtime library}
10540 By default this routine returns 0, if this routine returns a non-zero value,
10541 the static & global variable initialization will be skipped and the function
10542 main will be invoked.
10543 Otherwise static & global variables will be initialized before the function
10547 _sdcc_external_startup()
10549 routine to your program to override the default if you need to setup hardware
10550 or perform some other critical operation prior to static & global variable
10552 On some mcs51 variants xdata has to be explicitly enabled before it can
10553 be accessed, this is the place to do it.
10554 See also the compiler option
10573 \begin_inset LatexCommand \index{-\/-no-xinit-opt}
10578 \begin_inset LatexCommand \ref{sub:MCS51-variants}
10583 about MCS51-variants.
10589 The HC08 startup code follows the same scheme as the MCS51 startup code.
10595 On the Z80 the startup code is inserted by linking with crt0.o which is generated
10596 from sdcc/device/lib/z80/crt0.s.
10597 If you need a different startup code you can use the compiler option
10618 \begin_inset LatexCommand \index{-\/-no-std-crt0}
10622 and provide your own crt0.o.
10626 Inline Assembler Code
10627 \begin_inset LatexCommand \index{Assembler routines}
10634 A Step by Step Introduction
10635 \begin_inset LatexCommand \label{sub:A-Step-by Assembler Introduction}
10642 Starting from a small snippet of c-code this example shows for the MCS51
10643 how to use inline assembly, access variables, a function parameter and
10644 an array in xdata memory.
10645 The example uses an MCS51 here but is easily adapted for other architectures.
10646 This is a buffer routine which should be optimized:
10653 \begin_inset LatexCommand \index{far (storage class)}
10658 \begin_inset LatexCommand \index{at}
10663 \begin_inset LatexCommand \index{Aligned array}
10669 unsigned char head,tail;
10673 void to_buffer( unsigned char c )
10681 if( head != tail-1 )
10696 If the code snippet (assume it is saved in buffer.c) is compiled with SDCC
10697 then a corresponding buffer.asm file is generated.
10698 We define a new function
10702 in file buffer.c in which we cut and paste the generated code, removing
10703 unwanted comments and some ':'.
10705 \begin_inset Quotes sld
10709 \begin_inset Quotes srd
10713 \begin_inset Quotes sld
10717 \begin_inset Quotes srd
10720 to the beginning and the end of the function body:
10726 /* With a cut and paste from the .asm file, we have something to start with.
10731 The function is not yet OK! (registers aren't saved) */
10733 void to_buffer_asm( unsigned char c )
10742 \begin_inset LatexCommand \index{\_asm}
10756 ;buffer.c if( head != tail-1 )
10804 ;buffer.c buf[ head++ ] = c; /* access to a 256 byte aligned array */
10805 \begin_inset LatexCommand \index{Aligned array}
10874 The new file buffer.c should compile with only one warning about the unreferenced
10875 function argument 'c'.
10876 Now we hand-optimize the assembly code and insert an #define USE_ASSEMBLY
10877 (1) and finally have:
10883 unsigned char far at 0x7f00 buf[0x100];
10885 unsigned char head,tail;
10887 #define USE_ASSEMBLY (1)
10895 void to_buffer( unsigned char c )
10903 if( head != tail-1 )
10923 void to_buffer( unsigned char c )
10931 c; // to avoid warning: unreferenced function argument
10938 \begin_inset LatexCommand \index{\_asm}
10952 ; save used registers here.
10963 ; If we were still using r2,r3 we would have to push them here.
10966 ; if( head != tail-1 )
11009 ; we could do an ANL a,#0x0f here to use a smaller buffer (see below)
11033 ; buf[ head++ ] = c;
11044 a,dpl \SpecialChar ~
11051 ; dpl holds lower byte of function argument
11062 dpl,_head \SpecialChar ~
11065 ; buf is 0x100 byte aligned so head can be used directly
11107 ; we could do an ANL _head,#0x0f here to use a smaller buffer (see above)
11119 ; restore used registers here
11132 The inline assembler code can contain any valid code understood by the assembler
11133 , this includes any assembler directives and comment lines
11139 The assembler does not like some characters like ':' or ''' in comments.
11140 You'll find an 100+ pages assembler manual in sdcc/as/doc/asxhtm.html
11144 The compiler does not do any validation of the code within the
11147 \begin_inset LatexCommand \index{\_asm}
11155 Specifically it will not know which registers are used and thus register
11157 \begin_inset LatexCommand \index{push/pop}
11161 has to be done manually.
11165 It is recommended that each assembly instruction (including labels) be placed
11166 in a separate line (as the example shows).
11180 \begin_inset LatexCommand \index{-\/-peep-asm}
11186 command line option is used, the inline assembler code will be passed through
11187 the peephole optimizer
11188 \begin_inset LatexCommand \index{Peephole optimizer}
11193 There are only a few (if any) cases where this option makes sense, it might
11194 cause some unexpected changes in the inline assembler code.
11195 Please go through the peephole optimizer rules defined in file
11199 before using this option.
11203 \begin_inset LatexCommand \label{sub:Naked-Functions}
11208 \begin_inset LatexCommand \index{Naked functions}
11215 A special keyword may be associated with a function declaring it as
11218 \begin_inset LatexCommand \index{\_naked}
11229 function modifier attribute prevents the compiler from generating prologue
11230 \begin_inset LatexCommand \index{function prologue}
11235 \begin_inset LatexCommand \index{function epilogue}
11239 code for that function.
11240 This means that the user is entirely responsible for such things as saving
11241 any registers that may need to be preserved, selecting the proper register
11242 bank, generating the
11246 instruction at the end, etc.
11247 Practically, this means that the contents of the function must be written
11248 in inline assembler.
11249 This is particularly useful for interrupt functions, which can have a large
11250 (and often unnecessary) prologue/epilogue.
11251 For example, compare the code generated by these two functions:
11257 \begin_inset LatexCommand \index{volatile}
11261 data unsigned char counter;
11265 void simpleInterrupt(void) interrupt
11266 \begin_inset LatexCommand \index{interrupt}
11284 void nakedInterrupt(void) interrupt 2 _naked
11293 \begin_inset LatexCommand \index{\_asm}
11310 _counter ; does not change flags, no need to save psw
11322 ; MUST explicitly include ret or reti in _naked function.
11329 \begin_inset LatexCommand \index{\_endasm}
11338 For an 8051 target, the generated simpleInterrupt looks like:
11479 whereas nakedInterrupt looks like:
11494 _counter ; does not change flags, no need to save psw
11512 ; MUST explicitly include ret or reti in _naked function
11515 The related directive #pragma exclude
11516 \begin_inset LatexCommand \index{\#pragma exclude}
11520 allows a more fine grained control over pushing & popping
11521 \begin_inset LatexCommand \index{push/pop}
11528 While there is nothing preventing you from writing C code inside a
11532 function, there are many ways to shoot yourself in the foot doing this,
11533 and it is recommended that you stick to inline assembler.
11536 Use of Labels within Inline Assembler
11539 SDCC allows the use of in-line assembler with a few restrictions regarding
11541 In older versions of the compiler all labels defined within inline assembler
11550 where nnnn is a number less than 100 (which implies a limit of utmost 100
11551 inline assembler labels
11565 \begin_inset LatexCommand \index{\_asm}
11595 \begin_inset LatexCommand \index{\_endasm}
11602 Inline assembler code cannot reference any C-Labels, however it can reference
11604 \begin_inset LatexCommand \index{Labels}
11608 defined by the inline assembler, e.g.:
11633 ; some assembler code
11653 /* some more c code */
11655 clabel:\SpecialChar ~
11657 /* inline assembler cannot reference this label */
11669 $0003: ;label (can be referenced by inline assembler only)
11681 /* some more c code */
11686 In other words inline assembly code can access labels defined in inline
11687 assembly within the scope of the function.
11688 The same goes the other way, i.e.
11689 labels defines in inline assembly can not be accessed by C statements.
11692 Interfacing with Assembler Code
11693 \begin_inset LatexCommand \index{Assembler routines}
11700 Global Registers used for Parameter Passing
11701 \begin_inset LatexCommand \index{Parameter passing}
11708 The compiler always uses the global registers
11711 \begin_inset LatexCommand \index{DPTR, DPH, DPL}
11716 \begin_inset LatexCommand \index{DPTR}
11721 \begin_inset LatexCommand \index{B (mcs51, ds390 register)}
11730 \begin_inset LatexCommand \index{ACC (mcs51, ds390 register)}
11736 to pass the first parameter to a routine.
11737 The second parameter onwards is either allocated on the stack (for reentrant
11748 -stack-auto is used) or in data / xdata memory (depending on the memory
11753 Assembler Routine (non-reentrant)
11756 In the following example
11757 \begin_inset LatexCommand \index{reentrant}
11762 \begin_inset LatexCommand \index{Assembler routines (non-reentrant)}
11766 the function c_func calls an assembler routine asm_func, which takes two
11768 \begin_inset LatexCommand \index{function parameter}
11777 extern int asm_func(unsigned char, unsigned char);
11781 int c_func (unsigned char i, unsigned char j)
11789 return asm_func(i,j);
11803 return c_func(10,9);
11808 The corresponding assembler function is:
11813 .globl _asm_func_PARM_2
11914 \begin_inset LatexCommand \index{DPTR, DPH, DPL}
11931 Note here that the return values
11932 \begin_inset LatexCommand \index{return value}
11936 are placed in 'dpl' - One byte return value, 'dpl' LSB & 'dph' MSB for
11938 'dpl', 'dph' and 'b' for three byte values (generic pointers) and 'dpl','dph','
11939 b' & 'acc' for four byte values.
11942 The parameter naming convention is _<function_name>_PARM_<n>, where n is
11943 the parameter number starting from 1, and counting from the left.
11944 The first parameter is passed in
11945 \begin_inset Quotes eld
11949 \begin_inset Quotes erd
11952 for a one byte parameter,
11953 \begin_inset Quotes eld
11957 \begin_inset Quotes erd
11961 \begin_inset Quotes eld
11965 \begin_inset Quotes erd
11968 for three bytes and
11969 \begin_inset Quotes eld
11973 \begin_inset Quotes erd
11976 for a four bytes parameter.
11977 The variable name for the second parameter will be _<function_name>_PARM_2.
11981 Assemble the assembler routine with the following command:
11988 asx8051 -losg asmfunc.asm
11995 Then compile and link the assembler routine to the C source file with the
12003 sdcc cfunc.c asmfunc.rel
12006 Assembler Routine (reentrant)
12010 \begin_inset LatexCommand \index{reentrant}
12015 \begin_inset LatexCommand \index{Assembler routines (reentrant)}
12019 the second parameter
12020 \begin_inset LatexCommand \index{function parameter}
12024 onwards will be passed on the stack, the parameters are pushed from right
12026 after the call the leftmost parameter will be on the top of the stack.
12027 Here is an example:
12032 extern int asm_func(unsigned char, unsigned char);
12036 int c_func (unsigned char i, unsigned char j) reentrant
12044 return asm_func(i,j);
12058 return c_func(10,9);
12063 The corresponding assembler routine is:
12163 The compiling and linking procedure remains the same, however note the extra
12164 entry & exit linkage required for the assembler code, _bp is the stack
12165 frame pointer and is used to compute the offset into the stack for parameters
12166 and local variables.
12170 \begin_inset LatexCommand \index{int (16 bit)}
12175 \begin_inset LatexCommand \index{long (32 bit)}
12182 For signed & unsigned int (16 bit) and long (32 bit) variables, division,
12183 multiplication and modulus operations are implemented by support routines.
12184 These support routines are all developed in ANSI-C to facilitate porting
12185 to other MCUs, although some model specific assembler optimizations are
12187 The following files contain the described routines, all of them can be
12188 found in <installdir>/share/sdcc/lib.
12194 \begin_inset Tabular
12195 <lyxtabular version="3" rows="11" columns="2">
12197 <column alignment="center" valignment="top" leftline="true" width="0">
12198 <column alignment="center" valignment="top" leftline="true" rightline="true" width="0">
12199 <row topline="true" bottomline="true">
12200 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
12210 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
12221 <row topline="true">
12222 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
12230 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
12235 16 bit multiplication
12239 <row topline="true">
12240 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
12248 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
12253 signed 16 bit division (calls _divuint)
12257 <row topline="true">
12258 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
12266 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
12271 unsigned 16 bit division
12275 <row topline="true">
12276 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
12284 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
12289 signed 16 bit modulus (calls _moduint)
12293 <row topline="true">
12294 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
12302 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
12307 unsigned 16 bit modulus
12311 <row topline="true">
12312 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
12320 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
12325 32 bit multiplication
12329 <row topline="true">
12330 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
12338 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
12343 signed 32 division (calls _divulong)
12347 <row topline="true">
12348 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
12356 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
12361 unsigned 32 division
12365 <row topline="true">
12366 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
12374 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
12379 signed 32 bit modulus (calls _modulong)
12383 <row topline="true" bottomline="true">
12384 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
12392 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
12397 unsigned 32 bit modulus
12410 Since they are compiled as
12415 \begin_inset LatexCommand \index{reentrant}
12420 \begin_inset LatexCommand \index{interrupt}
12424 service routines should not do any of the above operations.
12425 If this is unavoidable then the above routines will need to be compiled
12439 \begin_inset LatexCommand \index{-\/-stack-auto}
12445 option, after which the source program will have to be compiled with
12458 \begin_inset LatexCommand \index{-\/-int-long-reent}
12465 Notice that you don't have to call these routines directly.
12466 The compiler will use them automatically every time an integer operation
12470 Floating Point Support
12471 \begin_inset LatexCommand \index{Floating point support}
12478 SDCC supports IEEE (single precision 4 bytes) floating point numbers.The
12479 floating point support routines are derived from gcc's floatlib.c and consist
12480 of the following routines:
12488 \begin_inset Tabular
12489 <lyxtabular version="3" rows="17" columns="2">
12491 <column alignment="center" valignment="top" leftline="true" width="0">
12492 <column alignment="center" valignment="top" leftline="true" rightline="true" width="0">
12493 <row topline="true" bottomline="true">
12494 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
12511 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
12520 <row topline="true">
12521 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
12538 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
12552 add floating point numbers
12556 <row topline="true">
12557 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
12574 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
12588 subtract floating point numbers
12592 <row topline="true">
12593 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
12610 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
12624 divide floating point numbers
12628 <row topline="true">
12629 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
12646 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
12660 multiply floating point numbers
12664 <row topline="true">
12665 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
12682 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
12696 convert floating point to unsigned char
12700 <row topline="true">
12701 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
12718 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
12732 convert floating point to signed char
12736 <row topline="true">
12737 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
12754 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
12768 convert floating point to unsigned int
12772 <row topline="true">
12773 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
12790 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
12804 convert floating point to signed int
12808 <row topline="true">
12809 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
12835 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
12849 convert floating point to unsigned long
12853 <row topline="true">
12854 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
12871 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
12885 convert floating point to signed long
12889 <row topline="true">
12890 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
12907 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
12921 convert unsigned char to floating point
12925 <row topline="true">
12926 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
12943 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
12957 convert char to floating point number
12961 <row topline="true">
12962 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
12979 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
12993 convert unsigned int to floating point
12997 <row topline="true">
12998 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
13015 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
13029 convert int to floating point numbers
13033 <row topline="true">
13034 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
13051 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
13065 convert unsigned long to floating point number
13069 <row topline="true" bottomline="true">
13070 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
13087 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
13101 convert long to floating point number
13114 These support routines are developed in ANSI-C so there is room for space
13115 and speed improvement.
13116 Note if all these routines are used simultaneously the data space might
13118 For serious floating point usage it is recommended that the large model
13120 Also notice that you don't have to call this routines directly.
13121 The compiler will use them automatically every time a floating point operation
13126 \begin_inset LatexCommand \index{Libraries}
13135 <pending: this is messy and incomplete - a little more information is in
13136 sdcc/doc/libdoc.txt
13141 Compiler support routines (_gptrget, _mulint etc.)
13144 Stdclib functions (puts, printf, strcat etc.)
13145 \layout Subsubsection
13151 \begin_inset LatexCommand \index{<stdio.h>}
13155 As usual on embedded systems you have to provide your own
13158 \begin_inset LatexCommand \index{getchar()}
13167 \begin_inset LatexCommand \index{putchar()}
13174 SDCC does not know whether the system connects to a serial line with or
13175 without handshake, LCD, keyboard or other device.
13176 You'll find examples for serial routines f.e.
13177 in sdcc/device/lib.
13180 If you're short on memory you might want to use
13191 \begin_inset LatexCommand \index{printf()}
13198 For the mcs51 there is an assembly version
13202 which should fit the requirements of many embedded systems (by unsetting
13203 #defines it can be customized to
13207 support long variables and field widths).
13210 Math functions (sin, pow, sqrt etc.)
13217 \begin_inset LatexCommand \index{Libraries}
13221 included in SDCC should have a license at least as liberal as the GNU Lesser
13222 General Public License
13223 \begin_inset LatexCommand \index{GNU Lesser General Public License, LGPL}
13234 license statements for the libraries are missing.
13235 sdcc/device/lib/ser_ir.c
13239 come with a GPL (as opposed to LGPL) License - this will not be liberal
13240 enough for many embedded programmers.
13243 If you have ported some library or want to share experience about some code
13245 falls into any of these categories Busses (I
13246 \begin_inset Formula $^{\textrm{2}}$
13249 C, CAN, Ethernet, Profibus, Modbus, USB, SPI, JTAG ...), Media (IDE, Memory
13250 cards, eeprom, flash...), En-/Decryption, Remote debugging, Realtime kernel,
13251 Keyboard, LCD, RTC, FPGA, PID then the sdcc-user mailing list
13252 \begin_inset LatexCommand \url{http://sourceforge.net/mail/?group_id=599}
13257 would certainly like to hear about it.
13258 Programmers coding for embedded systems are not especially famous for being
13259 enthusiastic, so don't expect a big hurray but as the mailing list is searchabl
13260 e these references are very valuable.
13261 Let's help to create a climate where information is shared.
13267 MCS51 Memory Models
13268 \begin_inset LatexCommand \index{Memory model}
13273 \begin_inset LatexCommand \index{MCS51 memory model}
13278 \layout Subsubsection
13283 SDCC allows two memory models for MCS51 code,
13292 Modules compiled with different memory models should
13296 be combined together or the results would be unpredictable.
13297 The library routines supplied with the compiler are compiled as both small
13299 The compiled library modules are contained in separate directories as small
13300 and large so that you can link to either set.
13304 When the large model is used all variables declared without a storage class
13305 will be allocated into the external ram, this includes all parameters and
13306 local variables (for non-reentrant
13307 \begin_inset LatexCommand \index{reentrant}
13312 When the small model is used variables without storage class are allocated
13313 in the internal ram.
13316 Judicious usage of the processor specific storage classes
13317 \begin_inset LatexCommand \index{Storage class}
13321 and the 'reentrant' function type will yield much more efficient code,
13322 than using the large model.
13323 Several optimizations are disabled when the program is compiled using the
13324 large model, it is therefore recommended that the small model be used unless
13325 absolutely required.
13326 \layout Subsubsection
13329 \begin_inset LatexCommand \label{sub:External-Stack}
13334 \begin_inset LatexCommand \index{stack}
13339 \begin_inset LatexCommand \index{External stack (mcs51)}
13346 The external stack (-
13357 \begin_inset LatexCommand \index{-\/-xstack}
13361 ) is located in pdata
13362 \begin_inset LatexCommand \index{pdata (mcs51, ds390 storage class)}
13366 memory (usually at the start of the external ram segment) and is 256 bytes
13378 -xstack option is used to compile the program, the parameters and local
13380 \begin_inset LatexCommand \index{local variables}
13384 of all reentrant functions are allocated in this area.
13385 This option is provided for programs with large stack space requirements.
13386 When used with the -
13397 \begin_inset LatexCommand \index{-\/-stack-auto}
13401 option, all parameters and local variables are allocated on the external
13402 stack (note: support libraries will need to be recompiled with the same
13406 The compiler outputs the higher order address byte of the external ram segment
13408 \begin_inset LatexCommand \index{P2 (mcs51 sfr)}
13413 \begin_inset LatexCommand \ref{sub:MCS51-variants}
13417 ), therefore when using the External Stack option, this port
13421 be used by the application program.
13425 \begin_inset LatexCommand \index{Memory model}
13430 \begin_inset LatexCommand \index{DS390 memory model}
13437 The only model supported is Flat 24
13438 \begin_inset LatexCommand \index{Flat 24 (DS390 memory model)}
13443 This generates code for the 24 bit contiguous addressing mode of the Dallas
13445 In this mode, up to four meg of external RAM or code space can be directly
13447 See the data sheets at www.dalsemi.com for further information on this part.
13451 Note that the compiler does not generate any code to place the processor
13452 into 24 bitmode (although
13456 in the ds390 libraries will do that for you).
13462 \begin_inset LatexCommand \index{Tinibios (DS390)}
13466 , the boot loader or similar code must ensure that the processor is in 24
13467 bit contiguous addressing mode before calling the SDCC startup code.
13485 option, variables will by default be placed into the XDATA segment.
13490 Segments may be placed anywhere in the 4 meg address space using the usual
13502 Note that if any segments are located above 64K, the -r flag must be passed
13503 to the linker to generate the proper segment relocations, and the Intel
13504 HEX output format must be used.
13505 The -r flag can be passed to the linker by using the option
13509 on the SDCC command line.
13510 However, currently the linker can not handle code segments > 64k.
13514 \begin_inset LatexCommand \index{Pragmas}
13521 SDCC supports the following #pragma directives:
13525 \begin_inset LatexCommand \index{\#pragma save}
13529 - this will save all current options to the save/restore stack.
13534 \begin_inset LatexCommand \index{\#pragma restore}
13538 - will restore saved options from the last save.
13539 saves & restores can be nested.
13540 SDCC uses a save/restore stack: save pushes current options to the stack,
13541 restore pulls current options from the stack.
13546 \begin_inset LatexCommand \index{\#pragma nogcse}
13550 - will stop global common subexpression elimination.
13554 \begin_inset LatexCommand \index{\#pragma noinduction}
13558 - will stop loop induction optimizations.
13562 \begin_inset LatexCommand \index{\#pragma nojtbound}
13566 - will not generate code for boundary value checking, when switch statements
13567 are turned into jump-tables (dangerous).
13568 For more details see section
13569 \begin_inset LatexCommand \ref{sub:'switch'-Statements}
13577 \begin_inset LatexCommand \index{\#pragma nooverlay}
13581 - the compiler will not overlay the parameters and local variables of a
13586 \begin_inset LatexCommand \index{\#pragma less\_pedantic}
13590 - the compiler will not warn you anymore for obvious mistakes, you'r on
13595 \begin_inset LatexCommand \index{\#pragma noloopreverse}
13599 - Will not do loop reversal optimization
13603 \begin_inset LatexCommand \index{\#pragma exclude}
13607 none | {acc[,b[,dpl[,dph]]] - The exclude pragma disables generation of
13609 \begin_inset LatexCommand \index{push/pop}
13613 instruction in ISR function (using interrupt
13614 \begin_inset LatexCommand \index{interrupt}
13619 The directive should be placed immediately before the ISR function definition
13620 and it affects ALL ISR functions following it.
13621 To enable the normal register saving for ISR functions use #pragma\SpecialChar ~
13622 exclude\SpecialChar ~
13624 \begin_inset LatexCommand \index{\#pragma exclude}
13632 \begin_inset LatexCommand \index{\#pragma noiv}
13636 - Do not generate interrupt
13637 \begin_inset LatexCommand \index{interrupt}
13641 vector table entries for all ISR functions defined after the pragma.
13642 This is useful in cases where the interrupt vector table must be defined
13643 manually, or when there is a secondary, manually defined interrupt vector
13645 for the autovector feature of the Cypress EZ-USB FX2).
13646 More elegantly this can be achieved by obmitting the optional interrupt
13647 number after the interrupt keyword, see section
13648 \begin_inset LatexCommand \ref{sub:Interrupt-Service-Routines}
13657 \begin_inset LatexCommand \index{\#pragma callee\_saves}
13662 \begin_inset LatexCommand \index{function prologue}
13666 function1[,function2[,function3...]] - The compiler by default uses a caller
13667 saves convention for register saving across function calls, however this
13668 can cause unnecessary register pushing & popping
13669 \begin_inset LatexCommand \index{push/pop}
13673 when calling small functions from larger functions.
13674 This option can be used to switch off the register saving convention for
13675 the function names specified.
13676 The compiler will not save registers when calling these functions, extra
13677 code need to be manually inserted at the entry & exit for these functions
13678 to save & restore the registers used by these functions, this can SUBSTANTIALLY
13679 reduce code & improve run time performance of the generated code.
13680 In the future the compiler (with inter procedural analysis) may be able
13681 to determine the appropriate scheme to use for each function call.
13692 -callee-saves command line option is used, the function names specified
13693 in #pragma\SpecialChar ~
13695 \begin_inset LatexCommand \index{\#pragma callee\_saves}
13699 is appended to the list of functions specified in the command line.
13702 SDCPP supports the following #pragma directives:
13706 \begin_inset LatexCommand \index{\#pragma preproc\_asm}
13710 (+ | -) - switch _asm _endasm block preprocessing on / off.
13714 The pragma's are intended to be used to turn-on or off certain optimizations
13715 which might cause the compiler to generate extra stack / data space to
13716 store compiler generated temporary variables.
13717 This usually happens in large functions.
13718 Pragma directives should be used as shown in the following example, they
13719 are used to control options & optimizations for a given function; pragmas
13720 should be placed before and/or after a function, placing pragma's inside
13721 a function body could have unpredictable results.
13727 \begin_inset LatexCommand \index{\#pragma save}
13738 /* save the current settings */
13741 \begin_inset LatexCommand \index{\#pragma nogcse}
13750 /* turnoff global subexpression elimination */
13752 #pragma noinduction
13753 \begin_inset LatexCommand \index{\#pragma noinduction}
13757 /* turn off induction optimizations */
13780 \begin_inset LatexCommand \index{\#pragma restore}
13784 /* turn the optimizations back on */
13787 The compiler will generate a warning message when extra space is allocated.
13788 It is strongly recommended that the save and restore pragma's be used when
13789 changing options for a function.
13792 Defines Created by the Compiler
13795 The compiler creates the following #defines
13796 \begin_inset LatexCommand \index{\#defines}
13801 \begin_inset LatexCommand \index{Defines created by the compiler}
13811 \begin_inset Tabular
13812 <lyxtabular version="3" rows="10" columns="2">
13814 <column alignment="center" valignment="top" leftline="true" width="0">
13815 <column alignment="center" valignment="top" leftline="true" rightline="true" width="0">
13816 <row topline="true" bottomline="true">
13817 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
13827 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
13838 <row topline="true">
13839 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
13845 \begin_inset LatexCommand \index{SDCC}
13852 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
13857 this Symbol is always defined
13861 <row topline="true">
13862 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
13868 \begin_inset LatexCommand \index{SDCC\_mcs51}
13873 \begin_inset LatexCommand \index{SDCC\_ds390}
13878 \begin_inset LatexCommand \index{SDCC\_z80}
13885 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
13890 depending on the model used (e.g.: -mds390
13894 <row topline="true">
13895 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
13901 \begin_inset LatexCommand \index{\_\_mcs51}
13906 \begin_inset LatexCommand \index{\_\_ds390}
13911 \begin_inset LatexCommand \index{\_\_hc08}
13916 \begin_inset LatexCommand \index{\_\_z80}
13923 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
13928 depending on the model used (e.g.
13933 <row topline="true">
13934 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
13940 \begin_inset LatexCommand \index{SDCC\_STACK\_AUTO}
13947 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
13970 <row topline="true">
13971 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
13977 \begin_inset LatexCommand \index{SDCC\_MODEL\_SMALL}
13984 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
14007 <row topline="true">
14008 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
14014 \begin_inset LatexCommand \index{SDCC\_MODEL\_LARGE}
14021 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
14044 <row topline="true">
14045 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
14051 \begin_inset LatexCommand \index{SDCC\_USE\_XSTACK}
14058 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
14081 <row topline="true">
14082 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
14088 \begin_inset LatexCommand \index{SDCC\_STACK\_TENBIT}
14095 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
14108 <row topline="true" bottomline="true">
14109 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
14115 \begin_inset LatexCommand \index{SDCC\_MODEL\_FLAT24}
14122 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
14142 Notes on supported Processors
14146 \begin_inset LatexCommand \label{sub:MCS51-variants}
14151 \begin_inset LatexCommand \index{MCS51 variants}
14158 MCS51 processors are available from many vendors and come in many different
14160 While they might differ considerably in respect to Special Function Registers
14161 the core MCS51 is usually not modified or is kept compatible.
14165 pdata access by SFR
14168 With the upcome of devices with internal xdata and flash memory devices
14170 \begin_inset LatexCommand \index{P2 (mcs51 sfr)}
14174 as dedicated I/O port is becoming more popular.
14175 Switching the high byte for pdata
14176 \begin_inset LatexCommand \index{pdata (mcs51, ds390 storage class)}
14180 access which was formerly done by port P2 is then achieved by a Special
14182 \begin_inset LatexCommand \index{sfr}
14187 In well-established MCS51 tradition the address of this
14191 is where the chip designers decided to put it.
14192 Needless to say that they didn't agree on a common name either.
14193 As pdata addressing is used in the startup code for the initialization
14194 of xdata variables a separate startup code should be used as described
14196 \begin_inset LatexCommand \ref{sub:Startup-Code}
14203 Other Features available by SFR
14206 Some MCS51 variants offer features like Double DPTR
14207 \begin_inset LatexCommand \index{DPTR}
14211 , multiple DPTR, decrementing DPTR, 16x16 Multiply.
14212 These are currently not used for the MCS51 port.
14213 If you absolutely need them you can fall back to inline assembly or submit
14217 The Z80 and gbz80 port
14220 SDCC can target both the Zilog
14221 \begin_inset LatexCommand \index{Z80}
14225 and the Nintendo Gameboy's Z80-like gbz80
14226 \begin_inset LatexCommand \index{gbz80 (GameBoy Z80)}
14231 The Z80 port is passed through the same
14234 \begin_inset LatexCommand \index{Regression test}
14240 as the MCS51 and DS390 ports, so floating point support, support for long
14241 variables and bitfield support is fine.
14242 See mailing lists and forums about interrupt routines.
14245 As always, the code is the authoritative reference - see z80/ralloc.c and
14248 \begin_inset LatexCommand \index{stack}
14252 frame is similar to that generated by the IAR Z80 compiler.
14253 IX is used as the base pointer, HL is used as a temporary register, and
14254 BC and DE are available for holding variables.
14255 IY is currently unused.
14257 \begin_inset LatexCommand \index{return value}
14262 One bad side effect of using IX as the base pointer is that a functions
14263 stack frame is limited to 127 bytes - this will be fixed in a later version.
14269 The port to the Motorola HC08
14270 \begin_inset LatexCommand \index{HC08}
14274 family has been added in October 2003, thank you Erik!
14281 \begin_inset LatexCommand \index{PIC14}
14285 port still requires a major effort from the development community.
14286 However it can work for very simple code.
14289 C code and 14bit PIC code page
14290 \begin_inset LatexCommand \index{code page (pic14)}
14295 \begin_inset LatexCommand \index{RAM bank (pic14)}
14302 The linker organizes allocation for the code page and RAM banks.
14303 It does not have intimate knowledge of the code flow.
14304 It will put all the code section of a single asm file into a single code
14306 In order to make use of multiple code pages, separate asm files must be
14308 The compiler treats all functions of a single C file as being in the same
14309 code page unless it is non static.
14310 The compiler treats all local variables of a single C file as being in
14311 the same RAM bank unless it is an extern.
14315 To get the best follow these guide lines:
14318 make local functions static, as non static functions require code page selection
14322 Make local variables static as extern variables require RAM bank selection
14326 For devices that have multiple code pages it is more efficient to use the
14327 same number of files as pages, i.e.
14328 for the 16F877 use 4 separate files and i.e.
14329 for the 16F874 use 2 separate files.
14330 This way the linker can put the code for each file into different code
14331 pages and the compiler can allocate reusable variables more efficiently
14332 and there's less page selection overhead.
14333 And as for any 8 bit micro (especially for PIC 14 as they have a very simple
14334 instruction set) use 'unsigned char' whereever possible instead of 'int'.
14337 Creating a device include file
14340 For generating a device include file use the support perl script inc2h.pl
14341 kept in directory support/script.
14347 For the interrupt function, use the keyword 'interrupt'
14348 \begin_inset LatexCommand \index{interrupt}
14352 with level number of 0 (PIC14 only has 1 interrupt so this number is only
14353 there to avoid a syntax error - it ought to be fixed).
14359 void Intr(void) interrupt 0
14365 T0IF = 0; /* Clear timer interrupt */
14370 Linking and assembling
14372 For assembling you can use either GPUTILS'
14373 \begin_inset LatexCommand \index{gputils (pic tools)}
14377 gpasm.exe or MPLAB's mpasmwin.exe.
14378 For linking you can use either GPUTIL's gplink or MPLAB's mplink.exe.
14379 If you use MPLAB and an interrupt function then the linker script file
14380 vectors section will need to be enlarged to link with mplink.
14403 sdcc -S -V -mpic14 -p16F877 $<
14417 $(PRJ).hex: $(OBJS)
14427 gplink -m -s $(PRJ).lkr -o $(PRJ).hex $(OBJS)
14449 sdcc -S -V -mpic14 -p16F877 $<
14459 mpasmwin /q /o $*.asm
14463 $(PRJ).hex: $(OBJS)
14473 mplink /v $(PRJ).lkr /m $(PRJ).map /o $(PRJ).hex $(OBJS)
14477 \begin_inset LatexCommand \index{PIC16}
14485 \begin_inset LatexCommand \index{PIC16}
14489 port is the portion of SDCC that is responsible to produce code for the
14491 \begin_inset LatexCommand \index{Microchip}
14495 (TM) microcontrollers with 16 bit core.
14496 Currently this family of microcontrollers contains the PIC18Fxxx and PIC18Fxxxx.
14502 PIC16 port supports the standard command line arguments as supposed, with
14503 the exception of certain cases that will be mentioned in the following
14506 \labelwidthstring 00.00.0000
14518 -stack-auto Auto variables that are function parameters, will be saved on
14522 There is no need to specify this in the command line.
14524 \labelwidthstring 00.00.0000
14536 -float-reent All floating point functions are reentrant by default.
14539 There is no need to specifiy this in the command line.
14541 \labelwidthstring 00.00.0000
14553 -callee-saves See -
14565 \labelwidthstring 00.00.0000
14577 -all-callee-saves All function arguments are passed on stack by default.
14580 There is no need to specify this in the command line.
14582 \labelwidthstring 00.00.0000
14594 -fommit-frame-pointer Frame pointer will be omitted when the function uses
14595 no local variables.
14598 PIC16 Port Specific Options
14601 The port specific options appear after the global options in the sdcc --help
14603 \layout Subsubsection
14608 General options enable certain port features and optimizations.
14610 \labelwidthstring 00.00.0000
14622 -pgen-bank Instructs the port to insert BANKSEL directives before instructions
14623 that use the Bank Select Register (BSR).
14625 \labelwidthstring 00.00.0000
14637 -pomit-config-words Instructs the port to omit the generation of the configurati
14640 \labelwidthstring 00.00.0000
14652 -pomit-ivt Instructs the port to omit the generation of the interrupt vectors
14654 \labelwidthstring 00.00.0000
14666 -pleave-reset-vector Used in conjuction with the previous command, instructs
14667 the port NOT to omit the reset vector.
14669 \labelwidthstring 00.00.0000
14681 -penable-stack Enables stack usage.
14682 All new development is done with stack enabled.
14683 This command line soon will be deprecated and stack will be enabled by
14685 For the time being it must be entered if one wants to have stack.
14689 This option is deprecated.
14690 Stack is enabled by default in the port and there is no way to disable
14692 It is left here only for reference.
14694 \labelwidthstring 00.00.0000
14706 -stack-model=[model] Used in conjuction with the command above.
14707 Defines the stack model to be used, valid stack models are :
14710 \labelwidthstring 00.00.0000
14716 Selects small stack model.
14717 8 bit stack and frame pointers.
14718 Supports 256 bytes stack size.
14720 \labelwidthstring 00.00.0000
14726 Selects large stack model.
14727 16 bit stack and frame pointers.
14728 Supports 65536 bytes stack size.
14731 \labelwidthstring 00.00.0000
14743 -preplace-udata-with=[kword] Replaces the default udata keyword for allocating
14744 unitialized data variables with [kword].
14745 Valid keywords are: "udata_acs", "udata_shr", "udata_ovr".
14746 \layout Subsubsection
14751 Debugging options enable extra debugging information in the output files.
14753 \labelwidthstring 00.00.0000
14765 -debug-xtra Similar to --debug, but dumps more information.
14767 \labelwidthstring 00.00.0000
14779 -debug-ralloc Force register allocator to dump <source>.d file with debugging
14781 <source> is the name of the file compiled.
14783 \labelwidthstring 00.00.0000
14795 -pcode-verbose Enable pcode debugging information in translation.
14798 Preprocessor Macros
14801 PIC16 port defines the following preprocessor macros while translating a
14806 \begin_inset Tabular
14807 <lyxtabular version="3" rows="2" columns="2">
14809 <column alignment="center" valignment="top" leftline="true" width="0">
14810 <column alignment="center" valignment="top" leftline="true" rightline="true" width="0">
14811 <row topline="true" bottomline="true">
14812 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
14820 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
14829 <row topline="true" bottomline="true">
14830 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
14838 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
14843 MCU Identification.
14848 is the microcontrol identification number, i.e.
14863 PIC16 port uses the following directories for searching header files and
14868 \begin_inset Tabular
14869 <lyxtabular version="3" rows="3" columns="4">
14871 <column alignment="center" valignment="top" leftline="true" width="0">
14872 <column alignment="center" valignment="top" leftline="true" rightline="true" width="0">
14873 <column alignment="center" valignment="top" leftline="true" rightline="true" width="0">
14874 <column alignment="center" valignment="top" leftline="true" rightline="true" width="0">
14875 <row topline="true" bottomline="true">
14876 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
14884 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
14892 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
14900 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
14909 <row topline="true" bottomline="true">
14910 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
14915 PREFIX/sdcc/include/pic16
14918 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
14923 PIC16 specific headers
14926 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
14934 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
14943 <row topline="true" bottomline="true">
14944 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
14949 PREFIX/sdcc/lib/pic16
14952 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
14957 PIC16 specific libraries
14960 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
14968 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
14987 PIC16 port currently supports the following pragmas:
14989 \labelwidthstring 00.00.0000
14991 stack pragma stack forces the code generator to initialize the stack & frame
14992 pointers at a specific address.
14993 This is an adhoc solution since gplink does not support yet stack.
14994 When the gplink issue is resolved the pragma will be deprecated
15002 It is important to initialize the stack, otherwise strange things can happen.
15003 Stack is not initialized by default because there are some sources that
15005 (like library sources)
15010 If you omit setting the pragma the port emits a warning message before linking.
15011 If not initializing the stack is desired ignore the message.
15019 /* initializes stack at RAM address 0x5ff*/
15022 #pragma stack 0x5ff
15030 There is one main header file that can be included to the source files using
15037 This header file contains the definitions for the processor special registers,
15038 so it is necessery if the soruce accesses them.
15039 It can be included by adding the following line in the beginning of the
15043 #include <pic18fregs.h>
15046 The specific microcontroller is selected within the pic18fregs.h automatically,
15047 so the same source can be used with a variety of devices.
15053 The libraries that pic16 port depends on are the microcontroller device
15054 libraries which contain the symbol definitions for the microcontroller
15055 special function registers.
15056 These libraries have the format pic18fxxxx.lib, where
15060 is the microcontroller identification number.
15061 The specific library is selected automatically by the compiler at link
15062 stage according to the selected device.
15065 Libraries are created with gplib which is part of the gputils package
15066 \begin_inset LatexCommand \url{http://gputils.sourceforge.net}
15076 The following memory models are supported by PIC16 port:
15085 Memory model affects the default size of pointers within the source.
15086 The sizes are shown in the next table:
15090 \begin_inset Tabular
15091 <lyxtabular version="3" rows="3" columns="3">
15093 <column alignment="center" valignment="top" leftline="true" rightline="true" width="0">
15094 <column alignment="center" valignment="top" leftline="true" width="0">
15095 <column alignment="center" valignment="top" leftline="true" rightline="true" width="0">
15096 <row topline="true" bottomline="true">
15097 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
15102 Pointer sizes according to memory model
15105 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
15113 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
15122 <row topline="true" bottomline="true">
15123 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
15131 <cell multicolumn="1" alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
15139 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
15148 <row topline="true" bottomline="true">
15149 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
15157 <cell multicolumn="1" alignment="center" valignment="top" topline="true" bottomline="true" leftline="true" usebox="none">
15165 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
15181 It is advisable that all sources within a project are compiled with the
15183 If one wants to override the default memory model, this can be done by
15184 declaring a pointer as
15193 Far selects large memory model's pointers, while near selects small memory
15197 The standard device libraries (see 4.5.6) contain no reference to pointers,
15198 so they can be used with both memory models.
15204 The stack implementation for the PIC16 port uses two indirect registers,
15207 \labelwidthstring 00.00.0000
15209 FSR1 is assigned as stack pointer
15211 \labelwidthstring 00.00.0000
15213 FSR2 is assigned as frame pointer
15216 The following stack models are supported by PIC16 port
15229 model means that only the FSRxL byte is used to access stack and frame,
15236 uses both FSRxL and FSRxH registers.
15237 The following table shows the stack/frame pointers sizes according to stack
15238 model and the maximum space they can address:
15242 \begin_inset Tabular
15243 <lyxtabular version="3" rows="3" columns="3">
15245 <column alignment="center" valignment="top" leftline="true" rightline="true" width="0">
15246 <column alignment="center" valignment="top" leftline="true" width="0">
15247 <column alignment="center" valignment="top" leftline="true" rightline="true" width="0">
15248 <row topline="true" bottomline="true">
15249 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
15254 Stack & Frame pointer sizes according to stack model
15257 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
15265 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
15274 <row topline="true">
15275 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
15283 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
15291 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
15300 <row topline="true" bottomline="true">
15301 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
15309 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
15317 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
15333 Function return values
15336 Return values from functions are placed to the appropriate registers following
15337 a modified Microchip policy optimized for SDCC.
15338 The following table shows these registers:
15342 \begin_inset Tabular
15343 <lyxtabular version="3" rows="6" columns="2">
15345 <column alignment="center" valignment="top" leftline="true" width="0">
15346 <column alignment="center" valignment="top" leftline="true" rightline="true" width="0">
15347 <row topline="true" bottomline="true">
15348 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
15356 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
15361 destination register
15365 <row topline="true">
15366 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
15374 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
15383 <row topline="true">
15384 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
15392 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
15401 <row topline="true">
15402 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
15410 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
15419 <row topline="true">
15420 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
15428 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
15433 FSR0L:PRODH:PRODL:WREG
15437 <row topline="true" bottomline="true">
15438 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
15446 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
15451 on stack, FSR0 points to the beginning
15465 When entering an interrupt, currently the PIC16 port automatically saves
15466 the following registers:
15478 PROD (PRODL and PRODH)
15481 FSR0 (FSR0L and FSR0H)
15484 These registers are restored upon return from the interrupt routine
15490 NOTE that when the _naked attribute is specified for an interrupt routine,
15491 then NO registers are stored or restored.
15497 Currently interrupt enable flags are left unaffected when entering an interrupt
15499 This may change in the future.
15502 Debugging with SDCDB
15503 \begin_inset LatexCommand \label{cha:Debugging-with-SDCDB}
15508 \begin_inset LatexCommand \index{sdcdb (debugger)}
15515 SDCC is distributed with a source level debugger
15516 \begin_inset LatexCommand \index{Debugger}
15521 The debugger uses a command line interface, the command repertoire of the
15522 debugger has been kept as close to gdb
15523 \begin_inset LatexCommand \index{gdb}
15527 (the GNU debugger) as possible.
15528 The configuration and build process is part of the standard compiler installati
15529 on, which also builds and installs the debugger in the target directory
15530 specified during configuration.
15531 The debugger allows you debug BOTH at the C source and at the ASM source
15533 Sdcdb is available on Unix platforms only.
15536 Compiling for Debugging
15539 The \SpecialChar \-
15541 debug option must be specified for all files for which debug information
15542 is to be generated.
15543 The complier generates a .adb file for each of these files.
15544 The linker creates the .cdb
15545 \begin_inset LatexCommand \index{<file>.cdb}
15550 \begin_inset LatexCommand \index{<file>.adb}
15554 files and the address information.
15555 This .cdb is used by the debugger.
15558 How the Debugger Works
15571 -debug option is specified the compiler generates extra symbol information
15572 some of which are put into the assembler source and some are put into the
15574 Then the linker creates the .cdb file from the individual .adb files with
15575 the address information for the symbols.
15576 The debugger reads the symbolic information generated by the compiler &
15577 the address information generated by the linker.
15578 It uses the SIMULATOR (Daniel's S51) to execute the program, the program
15579 execution is controlled by the debugger.
15580 When a command is issued for the debugger, it translates it into appropriate
15581 commands for the simulator.
15584 Starting the Debugger
15587 The debugger can be started using the following command line.
15588 (Assume the file you are debugging has the file name foo).
15602 The debugger will look for the following files.
15605 foo.c - the source file.
15608 foo.cdb - the debugger symbol information file.
15611 foo.ihx - the Intel hex format
15612 \begin_inset LatexCommand \index{Intel hex format}
15619 Command Line Options.
15632 -directory=<source file directory> this option can used to specify the directory
15634 The debugger will look into the directory list specified for source, cdb
15636 The items in the directory list must be separated by ':', e.g.
15637 if the source files can be in the directories /home/src1 and /home/src2,
15648 -directory option should be -
15658 -directory=/home/src1:/home/src2.
15659 Note there can be no spaces in the option.
15663 -cd <directory> - change to the <directory>.
15666 -fullname - used by GUI front ends.
15669 -cpu <cpu-type> - this argument is passed to the simulator please see the
15670 simulator docs for details.
15673 -X <Clock frequency > this options is passed to the simulator please see
15674 the simulator docs for details.
15677 -s <serial port file> passed to simulator see the simulator docs for details.
15680 -S <serial in,out> passed to simulator see the simulator docs for details.
15683 -k <port number> passed to simulator see the simulator docs for details.
15689 As mentioned earlier the command interface for the debugger has been deliberatel
15690 y kept as close the GNU debugger gdb, as possible.
15691 This will help the integration with existing graphical user interfaces
15692 (like ddd, xxgdb or xemacs) existing for the GNU debugger.
15693 If you use a graphical user interface for the debugger you can skip the
15695 \layout Subsubsection*
15697 break [line | file:line | function | file:function]
15700 Set breakpoint at specified line or function:
15709 sdcdb>break foo.c:100
15711 sdcdb>break funcfoo
15713 sdcdb>break foo.c:funcfoo
15714 \layout Subsubsection*
15716 clear [line | file:line | function | file:function ]
15719 Clear breakpoint at specified line or function:
15728 sdcdb>clear foo.c:100
15730 sdcdb>clear funcfoo
15732 sdcdb>clear foo.c:funcfoo
15733 \layout Subsubsection*
15738 Continue program being debugged, after breakpoint.
15739 \layout Subsubsection*
15744 Execute till the end of the current function.
15745 \layout Subsubsection*
15750 Delete breakpoint number 'n'.
15751 If used without any option clear ALL user defined break points.
15752 \layout Subsubsection*
15754 info [break | stack | frame | registers ]
15757 info break - list all breakpoints
15760 info stack - show the function call stack.
15763 info frame - show information about the current execution frame.
15766 info registers - show content of all registers.
15767 \layout Subsubsection*
15772 Step program until it reaches a different source line.
15773 Note: pressing <return> repeats the last command.
15774 \layout Subsubsection*
15779 Step program, proceeding through subroutine calls.
15780 \layout Subsubsection*
15785 Start debugged program.
15786 \layout Subsubsection*
15791 Print type information of the variable.
15792 \layout Subsubsection*
15797 print value of variable.
15798 \layout Subsubsection*
15803 load the given file name.
15804 Note this is an alternate method of loading file for debugging.
15805 \layout Subsubsection*
15810 print information about current frame.
15811 \layout Subsubsection*
15816 Toggle between C source & assembly source.
15817 \layout Subsubsection*
15819 ! simulator command
15822 Send the string following '!' to the simulator, the simulator response is
15824 Note the debugger does not interpret the command being sent to the simulator,
15825 so if a command like 'go' is sent the debugger can loose its execution
15826 context and may display incorrect values.
15827 \layout Subsubsection*
15834 My name is Bobby Brown"
15837 Interfacing with XEmacs
15838 \begin_inset LatexCommand \index{XEmacs}
15843 \begin_inset LatexCommand \index{Emacs}
15850 Two files (in emacs lisp) are provided for the interfacing with XEmacs,
15851 sdcdb.el and sdcdbsrc.el.
15852 These two files can be found in the $(prefix)/bin directory after the installat
15854 These files need to be loaded into XEmacs for the interface to work.
15855 This can be done at XEmacs startup time by inserting the following into
15856 your '.xemacs' file (which can be found in your HOME directory):
15862 (load-file sdcdbsrc.el)
15868 .xemacs is a lisp file so the () around the command is REQUIRED.
15869 The files can also be loaded dynamically while XEmacs is running, set the
15870 environment variable 'EMACSLOADPATH' to the installation bin directory
15871 (<installdir>/bin), then enter the following command ESC-x load-file sdcdbsrc.
15872 To start the interface enter the following command:
15886 You will prompted to enter the file name to be debugged.
15891 The command line options that are passed to the simulator directly are bound
15892 to default values in the file sdcdbsrc.el.
15893 The variables are listed below, these values maybe changed as required.
15896 sdcdbsrc-cpu-type '51
15899 sdcdbsrc-frequency '11059200
15902 sdcdbsrc-serial nil
15905 The following is a list of key mapping for the debugger interface.
15913 ;; Current Listing ::
15915 ;;key\SpecialChar ~
15930 binding\SpecialChar ~
15954 ;;---\SpecialChar ~
15969 ------\SpecialChar ~
16009 sdcdb-next-from-src\SpecialChar ~
16035 sdcdb-back-from-src\SpecialChar ~
16061 sdcdb-cont-from-src\SpecialChar ~
16071 SDCDB continue command
16087 sdcdb-step-from-src\SpecialChar ~
16113 sdcdb-whatis-c-sexp\SpecialChar ~
16123 SDCDB ptypecommand for data at
16187 sdcdbsrc-delete\SpecialChar ~
16201 SDCDB Delete all breakpoints if no arg
16249 given or delete arg (C-u arg x)
16265 sdcdbsrc-frame\SpecialChar ~
16280 SDCDB Display current frame if no arg,
16329 given or display frame arg
16394 sdcdbsrc-goto-sdcdb\SpecialChar ~
16404 Goto the SDCDB output buffer
16420 sdcdb-print-c-sexp\SpecialChar ~
16431 SDCDB print command for data at
16495 sdcdbsrc-goto-sdcdb\SpecialChar ~
16505 Goto the SDCDB output buffer
16521 sdcdbsrc-mode\SpecialChar ~
16537 Toggles Sdcdbsrc mode (turns it off)
16541 ;; C-c C-f\SpecialChar ~
16549 sdcdb-finish-from-src\SpecialChar ~
16557 SDCDB finish command
16561 ;; C-x SPC\SpecialChar ~
16569 sdcdb-break\SpecialChar ~
16587 Set break for line with point
16589 ;; ESC t\SpecialChar ~
16599 sdcdbsrc-mode\SpecialChar ~
16615 Toggle Sdcdbsrc mode
16617 ;; ESC m\SpecialChar ~
16627 sdcdbsrc-srcmode\SpecialChar ~
16649 Here are a few guidelines that will help the compiler generate more efficient
16650 code, some of the tips are specific to this compiler others are generally
16651 good programming practice.
16654 Use the smallest data type to represent your data-value.
16655 If it is known in advance that the value is going to be less than 256 then
16656 use an 'unsigned char' instead of a 'short' or 'int'.
16657 Please note, that ANSI C requires both signed and unsigned chars to be
16658 promoted to 'signed int' before doing any operation.
16659 This promotion can be omitted, if the result is the same.
16660 The effect of the promotion rules together with the sign-extension is often
16667 unsigned char uc = 0xfe;
16669 if (uc * uc < 0) /* this is true! */
16688 (int) uc * (int) uc = (int) 0xfe * (int) 0xfe = (int) 0xfc04 = -1024
16698 (unsigned char) -12 / (signed char) -3 = ...
16701 No, the result is not 4:
16706 (int) (unsigned char) -12 / (int) (signed char) -3 =
16708 (int) (unsigned char) 0xf4 / (int) (signed char) 0xfd =
16710 (int) 0x00f4 / (int) 0xfffd =
16712 (int) 0x00f4 / (int) 0xfffd =
16714 (int) 244 / (int) -3 =
16716 (int) -81 = (int) 0xffaf;
16719 Don't complain, that gcc gives you a different result.
16720 gcc uses 32 bit ints, while SDCC uses 16 bit ints.
16721 Therefore the results are different.
16724 \begin_inset Quotes sld
16728 \begin_inset Quotes srd
16734 If well-defined overflow characteristics are important and negative values
16735 are not, or if you want to steer clear of sign-extension problems when
16736 manipulating bits or bytes, use one of the corresponding unsigned types.
16737 (Beware when mixing signed and unsigned values in expressions, though.)
16739 Although character types (especially unsigned char) can be used as "tiny"
16740 integers, doing so is sometimes more trouble than it's worth, due to unpredicta
16741 ble sign extension and increased code size.
16745 Use unsigned when it is known in advance that the value is not going to
16747 This helps especially if you are doing division or multiplication, bit-shifting
16748 or are using an array index.
16751 NEVER jump into a LOOP.
16754 Declare the variables to be local
16755 \begin_inset LatexCommand \index{local variables}
16759 whenever possible, especially loop control variables (induction).
16762 Since the compiler does not always do implicit integral promotion, the programme
16763 r should do an explicit cast when integral promotion is required.
16766 Reducing the size of division, multiplication & modulus operations can reduce
16767 code size substantially.
16768 Take the following code for example.
16774 foobar(unsigned int p1, unsigned char ch)
16782 unsigned char ch1 = p1 % ch ;
16793 For the modulus operation the variable ch will be promoted to unsigned int
16794 first then the modulus operation will be performed (this will lead to a
16795 call to support routine _moduint()), and the result will be casted to a
16797 If the code is changed to
16802 foobar(unsigned int p1, unsigned char ch)
16810 unsigned char ch1 = (unsigned char)p1 % ch ;
16821 It would substantially reduce the code generated (future versions of the
16822 compiler will be smart enough to detect such optimization opportunities).
16826 Have a look at the assembly listing to get a
16827 \begin_inset Quotes sld
16831 \begin_inset Quotes srd
16834 for the code generation.
16838 \begin_inset LatexCommand \index{Tools}
16842 included in the distribution
16846 \begin_inset Tabular
16847 <lyxtabular version="3" rows="12" columns="3">
16849 <column alignment="center" valignment="top" leftline="true" width="0pt">
16850 <column alignment="center" valignment="top" leftline="true" width="0pt">
16851 <column alignment="center" valignment="top" leftline="true" rightline="true" width="0pt">
16852 <row topline="true" bottomline="true">
16853 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
16861 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
16869 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
16878 <row topline="true">
16879 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
16887 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
16892 Simulator for various architectures
16895 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
16904 <row topline="true">
16905 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
16913 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
16918 header file conversion
16921 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
16926 sdcc/support/scripts
16930 <row topline="true">
16931 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
16939 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
16944 header file conversion
16947 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
16952 sdcc/support/scripts
16956 <row topline="true">
16957 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
16965 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
16973 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
16991 <row topline="true">
16992 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17000 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17008 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
17026 <row topline="true">
17027 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17035 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17043 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
17061 <row topline="true">
17062 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17070 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17078 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
17096 <row topline="true">
17097 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17105 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17113 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
17131 <row topline="true">
17132 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17140 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17148 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
17166 <row topline="true">
17167 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17175 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17183 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
17201 <row topline="true" bottomline="true">
17202 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17210 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17218 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
17246 \begin_inset LatexCommand \index{Documentation}
17250 included in the distribution
17254 \begin_inset Tabular
17255 <lyxtabular version="3" rows="10" columns="2">
17257 <column alignment="left" valignment="top" leftline="true" width="0">
17258 <column alignment="left" valignment="top" leftline="true" rightline="true" width="0pt">
17259 <row topline="true" bottomline="true">
17260 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17268 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
17273 Where to get / filename
17277 <row topline="true">
17278 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17283 SDCC Compiler User Guide
17286 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
17291 You're reading it right now
17295 <row topline="true">
17296 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17304 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
17313 <row topline="true">
17314 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17319 ASXXXX Assemblers and ASLINK Relocating Linker
17322 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
17327 sdcc/as/doc/asxhtm.html
17331 <row topline="true">
17332 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17337 SDCC regression test
17338 \begin_inset LatexCommand \index{Regression test}
17345 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
17350 sdcc/doc/test_suite_spec.pdf
17354 <row topline="true">
17355 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17363 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
17372 <row topline="true">
17373 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17378 Notes on debugging with sdcdb
17379 \begin_inset LatexCommand \index{sdcdb (debugger)}
17386 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
17391 sdcc/debugger/README
17395 <row topline="true">
17396 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17401 Software simulator for microcontrollers
17404 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
17431 <row topline="true">
17432 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17437 Temporary notes on the pic16
17438 \begin_inset LatexCommand \index{PIC16}
17445 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
17450 sdcc/src/pic16/NOTES
17454 <row topline="true" bottomline="true">
17455 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17460 SDCC internal documentation (debugging file format)
17463 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
17499 Related open source tools
17500 \begin_inset LatexCommand \index{Related tools}
17508 \begin_inset Tabular
17509 <lyxtabular version="3" rows="11" columns="3">
17511 <column alignment="center" valignment="top" leftline="true" width="0pt">
17512 <column alignment="block" valignment="top" leftline="true" width="30line%">
17513 <column alignment="center" valignment="top" leftline="true" rightline="true" width="0pt">
17514 <row topline="true" bottomline="true">
17515 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17523 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17531 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
17540 <row topline="true">
17541 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17547 \begin_inset LatexCommand \index{gpsim (pic simulator)}
17554 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17562 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
17568 \begin_inset LatexCommand \url{http://www.dattalo.com/gnupic/gpsim.html}
17576 <row topline="true">
17577 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17583 \begin_inset LatexCommand \index{gputils (pic tools)}
17590 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17598 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
17604 \begin_inset LatexCommand \url{http://gputils.sourceforge.net/}
17612 <row topline="true">
17613 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17621 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17629 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
17635 \begin_inset LatexCommand \url{http://digilander.libero.it/fbradasc/FLP5.html}
17643 <row topline="true">
17644 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17650 \begin_inset LatexCommand \index{indent (source formatting tool)}
17657 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17662 Formats C source - Master of the white spaces
17665 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
17671 \begin_inset LatexCommand \url{http://home.hccnet.nl/d.ingamells/beautify.html}
17679 <row topline="true">
17680 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17686 \begin_inset LatexCommand \index{srecord (tool)}
17693 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17698 Object file conversion, checksumming, ...
17701 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
17707 \begin_inset LatexCommand \url{http://srecord.sourceforge.net/}
17715 <row topline="true">
17716 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17722 \begin_inset LatexCommand \index{objdump (tool)}
17729 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17734 Object file conversion, ...
17737 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
17742 Part of binutils (should be there anyway)
17746 <row topline="true">
17747 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17753 \begin_inset LatexCommand \index{doxygen (source documentation tool)}
17760 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17765 Source code documentation system
17768 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
17774 \begin_inset LatexCommand \url{http://www.doxygen.org}
17782 <row topline="true">
17783 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17791 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17796 IDE (has anyone tried integrating SDCC & sdcdb? Unix only)
17799 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
17805 \begin_inset LatexCommand \url{http://www.kdevelop.org}
17813 <row topline="true">
17814 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17820 \begin_inset LatexCommand \index{splint (syntax checking tool)}
17827 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17832 Statically checks c sources
17835 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
17841 \begin_inset LatexCommand \url{http://www.splint.org}
17849 <row topline="true" bottomline="true">
17850 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17856 \begin_inset LatexCommand \index{ddd (debugger)}
17863 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17868 Debugger, serves nicely as GUI to sdcdb
17869 \begin_inset LatexCommand \index{sdcdb (debugger)}
17876 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
17882 \begin_inset LatexCommand \url{http://www.gnu.org/software/ddd/}
17899 Related documentation / recommended reading
17903 \begin_inset Tabular
17904 <lyxtabular version="3" rows="6" columns="3">
17906 <column alignment="center" valignment="top" leftline="true" width="0pt">
17907 <column alignment="block" valignment="top" leftline="true" width="30line%">
17908 <column alignment="center" valignment="top" leftline="true" rightline="true" width="0pt">
17909 <row topline="true" bottomline="true">
17910 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17918 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17926 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
17935 <row topline="true">
17936 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17953 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17959 \begin_inset LatexCommand \index{C Reference card}
17966 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
17972 \begin_inset LatexCommand \url{http://www.refcards.com/about/c.html}
17980 <row topline="true">
17981 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17989 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
17997 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
18003 \begin_inset LatexCommand \url{http://www.eskimo.com/~scs/C-faq/top.html}
18011 <row topline="true">
18012 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
18019 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
18024 Latest datasheet of the target CPU
18027 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
18036 <row topline="true">
18037 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
18044 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
18049 Revision history of datasheet
18052 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
18061 <row topline="true" bottomline="true">
18062 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
18072 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
18077 Advanced Compiler Design and Implementation
18080 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
18085 bookstore (very dedicated, probably read other books first)
18101 Some questions answered, some pointers given - it might be time to in turn
18109 can you solve your project with the selected microcontroller? Would you
18110 find out early or rather late that your target is too small/slow/whatever?
18111 Can you switch to a slightly better device if it doesn't fit?
18114 should you solve the problem with an 8 bit CPU? Or would a 16/32 bit CPU
18115 and/or another programming language be more adequate? Would an operating
18116 system on the target device help?
18119 if you solved the problem, will the marketing department be happy?
18122 if the marketing department is happy, will customers be happy?
18125 if you're the project manager, marketing department and maybe even the customer
18126 in one person, have you tried to see the project from the outside?
18129 is the project done if you think it is done? Or is just that other interface/pro
18130 tocol/feature/configuration/option missing? How about website, manual(s),
18131 internationali(z|s)ation, packaging, labels, 2nd source for components,
18132 electromagnetic compatability/interference, documentation for production,
18133 production test software, update mechanism, patent issues?
18136 is your project adequately positioned in that magic triangle: fame, fortune,
18140 Maybe not all answers to these questions are known and some answers may
18145 , nevertheless knowing these questions may help you to avoid burnout
18151 burnout is bad for electronic devices, programmers and motorcycle tyres
18155 Chances are you didn't want to hear some of them...
18159 \begin_inset LatexCommand \index{Support}
18166 SDCC has grown to be a large project.
18167 The compiler alone (without the preprocessor, assembler and linker) is
18168 well over 100,000 lines of code (blank stripped).
18169 The open source nature of this project is a key to its continued growth
18171 You gain the benefit and support of many active software developers and
18173 Is SDCC perfect? No, that's why we need your help.
18174 The developers take pride in fixing reported bugs.
18175 You can help by reporting the bugs and helping other SDCC users.
18176 There are lots of ways to contribute, and we encourage you to take part
18177 in making SDCC a great software package.
18181 The SDCC project is hosted on the SDCC sourceforge site at
18182 \begin_inset LatexCommand \htmlurl{http://sourceforge.net/projects/sdcc}
18187 You'll find the complete set of mailing lists
18188 \begin_inset LatexCommand \index{Mailing list}
18192 , forums, bug reporting system, patch submission
18193 \begin_inset LatexCommand \index{Patch submission}
18198 \begin_inset LatexCommand \index{download}
18202 area and cvs code repository
18203 \begin_inset LatexCommand \index{cvs code repository}
18211 \begin_inset LatexCommand \index{Bug reporting}
18216 \begin_inset LatexCommand \index{Reporting bugs}
18223 The recommended way of reporting bugs is using the infrastructure of the
18225 You can follow the status of bug reports there and have an overview about
18229 Bug reports are automatically forwarded to the developer mailing list and
18230 will be fixed ASAP.
18231 When reporting a bug, it is very useful to include a small test program
18232 (the smaller the better) which reproduces the problem.
18233 If you can isolate the problem by looking at the generated assembly code,
18234 this can be very helpful.
18235 Compiling your program with the -
18246 \begin_inset LatexCommand \index{-\/-dumpall}
18250 option can sometimes be useful in locating optimization problems.
18251 When reporting a bug please maker sure you:
18254 Attach the code you are compiling with SDCC.
18258 Specify the exact command you use to run SDCC, or attach your Makefile.
18262 Specify the SDCC version (type "
18268 "), your platform, and operating system.
18272 Provide an exact copy of any error message or incorrect output.
18276 Put something meaningful in the subject of your message.
18279 Please attempt to include these 5 important parts, as applicable, in all
18280 requests for support or when reporting any problems or bugs with SDCC.
18281 Though this will make your message lengthy, it will greatly improve your
18282 chance that SDCC users and developers will be able to help you.
18283 Some SDCC developers are frustrated by bug reports without code provided
18284 that they can use to reproduce and ultimately fix the problem, so please
18285 be sure to provide sample code if you are reporting a bug!
18288 Please have a short check that you are using a recent version of SDCC and
18289 the bug is not yet known.
18290 This is the link for reporting bugs:
18291 \begin_inset LatexCommand \htmlurl{http://sourceforge.net/tracker/?group_id=599&atid=100599}
18298 Requesting Features
18299 \begin_inset LatexCommand \label{sub:Requesting-Features}
18304 \begin_inset LatexCommand \index{Feature request}
18309 \begin_inset LatexCommand \index{Requesting features}
18316 Like bug reports feature requests are forwarded to the developer mailing
18318 This is the link for requesting features:
18319 \begin_inset LatexCommand \htmlurl{http://sourceforge.net/tracker/?group_id=599&atid=350599}
18329 Like bug reports contributed patches are forwarded to the developer mailing
18331 This is the link for submitting patches
18332 \begin_inset LatexCommand \index{Patch submission}
18337 \begin_inset LatexCommand \url{http://sourceforge.net/tracker/?group_id=599&atid=300599}
18344 You need to specify some parameters to the
18348 command for the patches to be useful.
18349 If you modified more than one file a patch created f.e.
18354 \begin_inset Quotes sld
18357 diff -Naur unmodified_directory modified_directory >my_changes.patch
18358 \begin_inset Quotes srd
18364 will be fine, otherwise
18368 \begin_inset Quotes sld
18371 diff -u sourcefile.c.orig sourcefile.c >my_changes.patch
18372 \begin_inset Quotes srd
18385 These links should take you directly to the
18386 \begin_inset LatexCommand \url[Mailing lists]{http://sourceforge.net/mail/?group_id=599}
18396 Traffic on sdcc-devel and sdcc-user is about 100 mails/month each not counting
18397 automated messages (mid 2003)
18401 \begin_inset LatexCommand \url[Forums]{http://sourceforge.net/forum/?group_id=599}
18405 , lists and forums are archived and searchable so if you are lucky someone
18406 already had a similar problem.
18412 You can follow the status of the cvs version
18413 \begin_inset LatexCommand \index{version}
18417 of SDCC by watching the Changelog
18418 \begin_inset LatexCommand \index{Changelog}
18422 in the cvs-repository
18427 \begin_inset LatexCommand \htmlurl{http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/*checkout*/sdcc/sdcc/ChangeLog?rev=HEAD&content-type=text/plain}
18435 \begin_inset LatexCommand \index{Release policy}
18442 Historically there often were long delays between official releases and
18443 the sourceforge download area tends to get not updated at all.
18444 Excuses in the past might have referred to problems with live range analysis,
18445 but as this was fixed a while ago, the current problem is that another
18446 excuse has to be found.
18447 Kidding aside, we have to get better there! On the other hand there are
18448 daily snapshots available at
18449 \begin_inset LatexCommand \htmlurl[snap]{http://sdcc.sourceforge.net/snap.php}
18453 , and you can always build the very last version (hopefully with many bugs
18454 fixed, and features added) from the source code available at
18455 \begin_inset LatexCommand \htmlurl[Source]{http://sdcc.sourceforge.net/snap.php#Source}
18463 \begin_inset LatexCommand \index{Examples}
18470 You'll find some small examples in the directory
18472 sdcc/device/examples/.
18475 More examples and libraries are available at
18477 The SDCC Open Knowledge Resource
18478 \begin_inset LatexCommand \url{http://www.qsl.net/dl9sec/SDCC_OKR.html}
18485 \begin_inset LatexCommand \url{http://www.pjrc.com/tech/8051/}
18492 I did insert a reference to Paul's web site here although it seems rather
18493 dedicated to a specific 8032 board (I think it's okay because it f.e.
18494 shows LCD/Harddisc interface and has a free 8051 monitor.
18495 Independent 8032 board vendors face hard competition of heavily subsidized
18496 development boards anyway).
18499 Maybe we should include some links to real world applications.
18500 Preferably pointer to pointers (one for each architecture) so this stays
18505 \begin_inset LatexCommand \index{Quality control}
18512 The compiler is passed through nightly compile and build checks.
18518 \begin_inset LatexCommand \index{Regression test}
18522 check that SDCC itself compiles flawlessly on several platforms and checks
18523 the quality of the code generated by SDCC by running the code through simulator
18525 There is a separate document
18528 \begin_inset LatexCommand \index{Test suite}
18537 You'll find the test code in the directory
18539 sdcc/support/regression
18542 You can run these tests manually by running
18546 in this directory (or f.e.
18551 \begin_inset Quotes sld
18555 \begin_inset Quotes srd
18561 if you don't want to run the complete tests).
18562 The test code might also be interesting if you want to look for examples
18563 \begin_inset LatexCommand \index{Examples}
18567 checking corner cases of SDCC or if you plan to submit patches
18568 \begin_inset LatexCommand \index{Patch submission}
18575 The pic port uses a different set of regression tests, you'll find them
18578 sdcc/src/regression
18583 SDCC Technical Data
18587 \begin_inset LatexCommand \index{Optimizations}
18594 SDCC performs a host of standard optimizations in addition to some MCU specific
18599 Sub-expression Elimination
18600 \begin_inset LatexCommand \index{Subexpression elimination}
18607 The compiler does local and
18633 will be translated to
18645 Some subexpressions are not as obvious as the above example, e.g.:
18655 In this case the address arithmetic a->b[i] will be computed only once;
18656 the equivalent code in C would be.
18668 The compiler will try to keep these temporary variables in registers.
18671 Dead-Code Elimination
18672 \begin_inset LatexCommand \index{Dead-code elimination}
18693 i = 1; \SpecialChar ~
18702 global = 1;\SpecialChar ~
18715 global = 3;\SpecialChar ~
18744 \begin_inset LatexCommand \index{Copy propagation}
18800 Note: the dead stores created by this copy propagation will be eliminated
18801 by dead-code elimination.
18805 \begin_inset LatexCommand \index{Loop optimization}
18812 Two types of loop optimizations are done by SDCC loop invariant lifting
18813 and strength reduction of loop induction variables.
18814 In addition to the strength reduction the optimizer marks the induction
18815 variables and the register allocator tries to keep the induction variables
18816 in registers for the duration of the loop.
18817 Because of this preference of the register allocator
18818 \begin_inset LatexCommand \index{Register allocation}
18822 , loop induction optimization causes an increase in register pressure, which
18823 may cause unwanted spilling of other temporary variables into the stack
18824 \begin_inset LatexCommand \index{stack}
18829 The compiler will generate a warning message when it is forced to allocate
18830 extra space either on the stack or data space.
18831 If this extra space allocation is undesirable then induction optimization
18832 can be eliminated either for the entire source file (with -
18842 -noinduction option) or for a given function only using #pragma\SpecialChar ~
18844 \begin_inset LatexCommand \index{\#pragma noinduction}
18857 for (i = 0 ; i < 100 ; i ++)
18873 for (i = 0; i < 100; i++)
18882 As mentioned previously some loop invariants are not as apparent, all static
18883 address computations are also moved out of the loop.
18888 \begin_inset LatexCommand \index{Strength reduction}
18892 , this optimization substitutes an expression by a cheaper expression:
18897 for (i=0;i < 100; i++)
18915 for (i=0;i< 100;i++) {
18921 ar[itemp1] = itemp2;
18938 The more expensive multiplication
18939 \begin_inset LatexCommand \index{Multiplication}
18943 is changed to a less expensive addition.
18947 \begin_inset LatexCommand \index{Loop reversing}
18954 This optimization is done to reduce the overhead of checking loop boundaries
18955 for every iteration.
18956 Some simple loops can be reversed and implemented using a
18957 \begin_inset Quotes eld
18960 decrement and jump if not zero
18961 \begin_inset Quotes erd
18965 SDCC checks for the following criterion to determine if a loop is reversible
18966 (note: more sophisticated compilers use data-dependency analysis to make
18967 this determination, SDCC uses a more simple minded analysis).
18970 The 'for' loop is of the form
18976 for(<symbol> = <expression>; <sym> [< | <=] <expression>; [<sym>++ | <sym>
18986 The <for body> does not contain
18987 \begin_inset Quotes eld
18991 \begin_inset Quotes erd
18995 \begin_inset Quotes erd
19001 All goto's are contained within the loop.
19004 No function calls within the loop.
19007 The loop control variable <sym> is not assigned any value within the loop
19010 The loop control variable does NOT participate in any arithmetic operation
19014 There are NO switch statements in the loop.
19017 Algebraic Simplifications
19020 SDCC does numerous algebraic simplifications, the following is a small sub-set
19021 of these optimizations.
19026 i = j + 0;\SpecialChar ~
19030 /* changed to: */\SpecialChar ~
19036 i /= 2;\SpecialChar ~
19043 /* changed to: */\SpecialChar ~
19049 i = j - j;\SpecialChar ~
19053 /* changed to: */\SpecialChar ~
19059 i = j / 1;\SpecialChar ~
19063 /* changed to: */\SpecialChar ~
19070 Note the subexpressions
19071 \begin_inset LatexCommand \index{Subexpression}
19075 given above are generally introduced by macro expansions or as a result
19076 of copy/constant propagation.
19079 'switch' Statements
19080 \begin_inset LatexCommand \label{sub:'switch'-Statements}
19085 \begin_inset LatexCommand \index{switch statement}
19092 SDCC changes switch statements to jump tables
19093 \begin_inset LatexCommand \index{jump tables}
19097 when the following conditions are true.
19101 The case labels are in numerical sequence, the labels need not be in order,
19102 and the starting number need not be one or zero.
19108 switch(i) {\SpecialChar ~
19139 case 4: ...\SpecialChar ~
19171 case 5: ...\SpecialChar ~
19203 case 3: ...\SpecialChar ~
19235 case 6: ...\SpecialChar ~
19303 Both the above switch statements will be implemented using a jump-table.
19304 The example to the right side is slightly more efficient as the check for
19305 the lower boundary of the jump-table is not needed.
19309 The number of case labels is at least three, since it takes two conditional
19310 statements to handle the boundary conditions.
19313 The number of case labels is less than 84, since each label takes 3 bytes
19314 and a jump-table can be utmost 256 bytes long.
19317 Switch statements which have gaps in the numeric sequence or those that
19318 have more that 84 case labels can be split into more than one switch statement
19319 for efficient code generation, e.g.:
19369 If the above switch statement is broken down into two switch statements
19408 case 9:\SpecialChar ~
19415 case 10:\SpecialChar ~
19421 case 11:\SpecialChar ~
19427 case 12:\SpecialChar ~
19434 then both the switch statements will be implemented using jump-tables whereas
19435 the unmodified switch statement will not be.
19436 You might also consider inserting dummy cases 0 and 5 to 8 in this example.
19439 The pragma nojtbound
19440 \begin_inset LatexCommand \index{\#pragma nojtbound}
19444 can be used to turn off checking the
19457 It has no effect if a default label is supplied.
19458 Use of this pragma is dangerous: if the switch argument is not matched
19459 by a case statement the processor will happily jump into Nirvana.
19462 Bit-shifting Operations
19463 \begin_inset LatexCommand \index{Bit shifting}
19470 Bit shifting is one of the most frequently used operation in embedded programmin
19472 SDCC tries to implement bit-shift operations in the most efficient way
19488 generates the following code:
19505 In general SDCC will never setup a loop if the shift count is known.
19548 \begin_inset LatexCommand \index{Bit rotation}
19555 A special case of the bit-shift operation is bit rotation
19556 \begin_inset LatexCommand \index{rotating bits}
19560 , SDCC recognizes the following expression to be a left bit-rotation:
19570 char i;\SpecialChar ~
19581 /* unsigned is needed for rotation */
19586 i = ((i << 1) | (i >> 7));
19595 will generate the following code:
19614 SDCC uses pattern matching on the parse tree to determine this operation.Variatio
19615 ns of this case will also be recognized as bit-rotation, i.e.:
19620 i = ((i >> 7) | (i << 1)); /* left-bit rotation */
19623 Nibble and Byte Swapping
19626 Other special cases of the bit-shift operations are nibble or byte swapping
19627 \begin_inset LatexCommand \index{swapping nibbles/bytes}
19631 , SDCC recognizes the following expressions:
19654 i = ((i << 4) | (i >> 4));
19660 j = ((j << 8) | (j >> 8));
19663 and generates a swap instruction for the nibble swapping
19664 \begin_inset LatexCommand \index{Nibble swapping}
19668 or move instructions for the byte swapping
19669 \begin_inset LatexCommand \index{Byte swapping}
19675 \begin_inset Quotes sld
19679 \begin_inset Quotes srd
19682 example can be used to convert from little to big-endian or vice versa.
19683 If you want to change the endianness of a
19687 integer you have to cast to
19694 Note that SDCC stores numbers in little-endian
19700 Usually 8-bit processors don't care much about endianness.
19701 This is not the case for the standard 8051 which only has an instruction
19707 \begin_inset LatexCommand \index{DPTR}
19715 so little-endian is the more efficient byte order.
19719 \begin_inset LatexCommand \index{little-endian}
19724 \begin_inset LatexCommand \index{Endianness}
19729 lowest order first).
19733 \begin_inset LatexCommand \index{Highest Order Bit}
19740 It is frequently required to obtain the highest order bit of an integral
19741 type (long, int, short or char types).
19742 SDCC recognizes the following expression to yield the highest order bit
19743 and generates optimized code for it, e.g.:
19765 hob = (gint >> 15) & 1;
19775 will generate the following code:
19808 000A E5*01\SpecialChar ~
19835 000C 23\SpecialChar ~
19866 000D 54 01\SpecialChar ~
19893 000F F5*02\SpecialChar ~
19921 Variations of this case however will
19926 It is a standard C expression, so I heartily recommend this be the only
19927 way to get the highest order bit, (it is portable).
19928 Of course it will be recognized even if it is embedded in other expressions,
19934 xyz = gint + ((gint >> 15) & 1);
19937 will still be recognized.
19941 \begin_inset LatexCommand \label{sub:Peephole-Optimizer}
19946 \begin_inset LatexCommand \index{Peephole optimizer}
19953 The compiler uses a rule based, pattern matching and re-writing mechanism
19954 for peep-hole optimization.
19959 a peep-hole optimizer by Christopher W.
19960 Fraser (cwfraser@microsoft.com).
19961 A default set of rules are compiled into the compiler, additional rules
19962 may be added with the
19975 \begin_inset LatexCommand \index{-\/-peep-file}
19982 The rule language is best illustrated with examples.
20006 The above rule will change the following assembly
20007 \begin_inset LatexCommand \index{Assembler routines}
20029 Note: All occurrences of a
20033 (pattern variable) must denote the same string.
20034 With the above rule, the assembly sequence:
20044 will remain unmodified.
20048 Other special case optimizations may be added by the user (via
20064 some variants of the 8051 MCU allow only
20073 The following two rules will change all
20092 replace { lcall %1 } by { acall %1 }
20094 replace { ljmp %1 } by { ajmp %1 }
20099 inline-assembler code
20101 is also passed through the peep hole optimizer, thus the peephole optimizer
20102 can also be used as an assembly level macro expander.
20103 The rules themselves are MCU dependent whereas the rule language infra-structur
20104 e is MCU independent.
20105 Peephole optimization rules for other MCU can be easily programmed using
20110 The syntax for a rule is as follows:
20115 rule := replace [ restart ] '{' <assembly sequence> '
20153 <assembly sequence> '
20171 '}' [if <functionName> ] '
20176 <assembly sequence> := assembly instruction (each instruction including
20177 labels must be on a separate line).
20181 The optimizer will apply to the rules one by one from the top in the sequence
20182 of their appearance, it will terminate when all rules are exhausted.
20183 If the 'restart' option is specified, then the optimizer will start matching
20184 the rules again from the top, this option for a rule is expensive (performance)
20185 , it is intended to be used in situations where a transformation will trigger
20186 the same rule again.
20187 An example of this (not a good one, it has side effects) is the following
20210 Note that the replace pattern cannot be a blank, but can be a comment line.
20211 Without the 'restart' option only the innermost 'pop' 'push' pair would
20212 be eliminated, i.e.:
20242 the restart option the rule will be applied again to the resulting code
20243 and then all the pop-push pairs will be eliminated to yield:
20253 A conditional function can be attached to a rule.
20254 Attaching rules are somewhat more involved, let me illustrate this with
20281 The optimizer does a look-up of a function name table defined in function
20286 in the source file SDCCpeeph.c, with the name
20291 If it finds a corresponding entry the function is called.
20292 Note there can be no parameters specified for these functions, in this
20297 is crucial, since the function
20301 expects to find the label in that particular variable (the hash table containin
20302 g the variable bindings is passed as a parameter).
20303 If you want to code more such functions, take a close look at the function
20304 labelInRange and the calling mechanism in source file SDCCpeeph.c.
20305 Currently implemented are
20307 labelInRange, labelRefCount, labelIsReturnOnly, operandsNotSame, xramMovcOption,
20308 24bitMode, portIsDS390, 24bitModeAndPortDS390
20317 I know this whole thing is a little kludgey, but maybe some day we will
20318 have some better means.
20319 If you are looking at this file, you will see the default rules that are
20320 compiled into the compiler, you can add your own rules in the default set
20321 there if you get tired of specifying the -
20335 \begin_inset LatexCommand \index{ANSI-compliance}
20340 \begin_inset LatexCommand \label{sub:ANSI-Compliance}
20347 Deviations from the compliance:
20350 functions are not always reentrant
20351 \begin_inset LatexCommand \index{reentrant}
20358 structures cannot be assigned values directly, cannot be passed as function
20359 parameters or assigned to each other and cannot be a return value from
20386 s1 = s2 ; /* is invalid in SDCC although allowed in ANSI */
20397 struct s foo1 (struct s parms) /* invalid in SDCC although allowed in ANSI
20419 return rets;/* is invalid in SDCC although allowed in ANSI */
20426 \begin_inset LatexCommand \index{long long (not supported)}
20431 \begin_inset LatexCommand \index{int (64 bit) (not supported)}
20439 \begin_inset LatexCommand \index{double (not supported)}
20443 ' precision floating point
20444 \begin_inset LatexCommand \index{Floating point support}
20451 No support for setjmp
20452 \begin_inset LatexCommand \index{setjmp (not supported)}
20457 \begin_inset LatexCommand \index{longjmp (not supported)}
20465 \begin_inset LatexCommand \index{K\&R style}
20469 function declarations are NOT allowed.
20475 foo(i,j) /* this old style of function declarations */
20477 int i,j; /* are valid in ANSI but not valid in SDCC */
20492 functions declared as pointers
20493 \begin_inset LatexCommand \index{Pointer (to function)}
20498 \begin_inset LatexCommand \index{function pointers}
20502 must be dereferenced during the call.
20513 /* has to be called like this */
20515 (*foo)(); /* ANSI standard allows calls to be made like 'foo()' */
20519 Cyclomatic Complexity
20520 \begin_inset LatexCommand \index{Cyclomatic complexity}
20527 Cyclomatic complexity of a function is defined as the number of independent
20528 paths the program can take during execution of the function.
20529 This is an important number since it defines the number test cases you
20530 have to generate to validate the function.
20531 The accepted industry standard for complexity number is 10, if the cyclomatic
20532 complexity reported by SDCC exceeds 10 you should think about simplification
20533 of the function logic.
20534 Note that the complexity level is not related to the number of lines of
20535 code in a function.
20536 Large functions can have low complexity, and small functions can have large
20542 SDCC uses the following formula to compute the complexity:
20547 complexity = (number of edges in control flow graph) - (number of nodes
20548 in control flow graph) + 2;
20552 Having said that the industry standard is 10, you should be aware that in
20553 some cases it be may unavoidable to have a complexity level of less than
20555 For example if you have switch statement with more than 10 case labels,
20556 each case label adds one to the complexity level.
20557 The complexity level is by no means an absolute measure of the algorithmic
20558 complexity of the function, it does however provide a good starting point
20559 for which functions you might look at for further optimization.
20562 Retargetting for other Processors
20565 The issues for retargetting the compiler are far too numerous to be covered
20567 What follows is a brief description of each of the seven phases of the
20568 compiler and its MCU dependency.
20571 Parsing the source and building the annotated parse tree.
20572 This phase is largely MCU independent (except for the language extensions).
20573 Syntax & semantic checks are also done in this phase, along with some initial
20574 optimizations like back patching labels and the pattern matching optimizations
20575 like bit-rotation etc.
20578 The second phase involves generating an intermediate code which can be easy
20579 manipulated during the later phases.
20580 This phase is entirely MCU independent.
20581 The intermediate code generation assumes the target machine has unlimited
20582 number of registers, and designates them with the name iTemp.
20583 The compiler can be made to dump a human readable form of the code generated
20597 This phase does the bulk of the standard optimizations and is also MCU independe
20599 This phase can be broken down into several sub-phases:
20603 Break down intermediate code (iCode) into basic blocks.
20605 Do control flow & data flow analysis on the basic blocks.
20607 Do local common subexpression elimination, then global subexpression elimination
20609 Dead code elimination
20613 If loop optimizations caused any changes then do 'global subexpression eliminati
20614 on' and 'dead code elimination' again.
20617 This phase determines the live-ranges; by live range I mean those iTemp
20618 variables defined by the compiler that still survive after all the optimization
20620 Live range analysis
20621 \begin_inset LatexCommand \index{Live range analysis}
20625 is essential for register allocation, since these computation determines
20626 which of these iTemps will be assigned to registers, and for how long.
20629 Phase five is register allocation.
20630 There are two parts to this process.
20634 The first part I call 'register packing' (for lack of a better term).
20635 In this case several MCU specific expression folding is done to reduce
20640 The second part is more MCU independent and deals with allocating registers
20641 to the remaining live ranges.
20642 A lot of MCU specific code does creep into this phase because of the limited
20643 number of index registers available in the 8051.
20646 The Code generation phase is (unhappily), entirely MCU dependent and very
20647 little (if any at all) of this code can be reused for other MCU.
20648 However the scheme for allocating a homogenized assembler operand for each
20649 iCode operand may be reused.
20652 As mentioned in the optimization section the peep-hole optimizer is rule
20653 based system, which can reprogrammed for other MCUs.
20657 \begin_inset LatexCommand \index{Compiler internals}
20664 The anatomy of the compiler
20665 \begin_inset LatexCommand \label{sub:The-anatomy-of}
20674 This is an excerpt from an article published in Circuit Cellar Magazine
20676 It's a little outdated (the compiler is much more efficient now and user/develo
20677 per friendly), but pretty well exposes the guts of it all.
20683 The current version of SDCC can generate code for Intel 8051 and Z80 MCU.
20684 It is fairly easy to retarget for other 8-bit MCU.
20685 Here we take a look at some of the internals of the compiler.
20690 \begin_inset LatexCommand \index{Parsing}
20697 Parsing the input source file and creating an AST (Annotated Syntax Tree
20698 \begin_inset LatexCommand \index{Annotated syntax tree}
20703 This phase also involves propagating types (annotating each node of the
20704 parse tree with type information) and semantic analysis.
20705 There are some MCU specific parsing rules.
20706 For example the storage classes, the extended storage classes are MCU specific
20707 while there may be a xdata storage class for 8051 there is no such storage
20708 class for z80 or Atmel AVR.
20709 SDCC allows MCU specific storage class extensions, i.e.
20710 xdata will be treated as a storage class specifier when parsing 8051 C
20711 code but will be treated as a C identifier when parsing z80 or ATMEL AVR
20716 \begin_inset LatexCommand \index{iCode}
20723 Intermediate code generation.
20724 In this phase the AST is broken down into three-operand form (iCode).
20725 These three operand forms are represented as doubly linked lists.
20726 ICode is the term given to the intermediate form generated by the compiler.
20727 ICode example section shows some examples of iCode generated for some simple
20728 C source functions.
20732 \begin_inset LatexCommand \index{Optimizations}
20739 Bulk of the target independent optimizations is performed in this phase.
20740 The optimizations include constant propagation, common sub-expression eliminati
20741 on, loop invariant code movement, strength reduction of loop induction variables
20742 and dead-code elimination.
20745 Live range analysis
20746 \begin_inset LatexCommand \index{Live range analysis}
20753 During intermediate code generation phase, the compiler assumes the target
20754 machine has infinite number of registers and generates a lot of temporary
20756 The live range computation determines the lifetime of each of these compiler-ge
20757 nerated temporaries.
20758 A picture speaks a thousand words.
20759 ICode example sections show the live range annotations for each of the
20761 It is important to note here, each iCode is assigned a number in the order
20762 of its execution in the function.
20763 The live ranges are computed in terms of these numbers.
20764 The from number is the number of the iCode which first defines the operand
20765 and the to number signifies the iCode which uses this operand last.
20768 Register Allocation
20769 \begin_inset LatexCommand \index{Register allocation}
20776 The register allocation determines the type and number of registers needed
20778 In most MCUs only a few registers can be used for indirect addressing.
20779 In case of 8051 for example the registers R0 & R1 can be used to indirectly
20780 address the internal ram and DPTR to indirectly address the external ram.
20781 The compiler will try to allocate the appropriate register to pointer variables
20783 ICode example section shows the operands annotated with the registers assigned
20785 The compiler will try to keep operands in registers as much as possible;
20786 there are several schemes the compiler uses to do achieve this.
20787 When the compiler runs out of registers the compiler will check to see
20788 if there are any live operands which is not used or defined in the current
20789 basic block being processed, if there are any found then it will push that
20790 operand and use the registers in this block, the operand will then be popped
20791 at the end of the basic block.
20795 There are other MCU specific considerations in this phase.
20796 Some MCUs have an accumulator; very short-lived operands could be assigned
20797 to the accumulator instead of a general-purpose register.
20803 Figure II gives a table of iCode operations supported by the compiler.
20804 The code generation involves translating these operations into corresponding
20805 assembly code for the processor.
20806 This sounds overly simple but that is the essence of code generation.
20807 Some of the iCode operations are generated on a MCU specific manner for
20808 example, the z80 port does not use registers to pass parameters so the
20809 SEND and RECV iCode operations will not be generated, and it also does
20810 not support JUMPTABLES.
20817 <Where is Figure II ?>
20821 \begin_inset LatexCommand \index{iCode}
20828 This section shows some details of iCode.
20829 The example C code does not do anything useful; it is used as an example
20830 to illustrate the intermediate code generated by the compiler.
20842 /* This function does nothing useful.
20849 for the purpose of explaining iCode */
20852 short function (data int *x)
20860 short i=10; \SpecialChar ~
20862 /* dead initialization eliminated */
20867 short sum=10; /* dead initialization eliminated */
20880 while (*x) *x++ = *p++;
20894 /* compiler detects i,j to be induction variables */
20898 for (i = 0, j = 10 ; i < 10 ; i++, j
20924 mul += i * 3; \SpecialChar ~
20926 /* this multiplication remains */
20932 gint += j * 3;\SpecialChar ~
20934 /* this multiplication changed to addition */
20948 In addition to the operands each iCode contains information about the filename
20949 and line it corresponds to in the source file.
20950 The first field in the listing should be interpreted as follows:
20955 Filename(linenumber: iCode Execution sequence number : ICode hash table
20956 key : loop depth of the iCode).
20961 Then follows the human readable form of the ICode operation.
20962 Each operand of this triplet form can be of three basic types a) compiler
20963 generated temporary b) user defined variable c) a constant value.
20964 Note that local variables and parameters are replaced by compiler generated
20967 \begin_inset LatexCommand \index{Live range analysis}
20971 are computed only for temporaries (i.e.
20972 live ranges are not computed for global variables).
20974 \begin_inset LatexCommand \index{Register allocation}
20978 are allocated for temporaries only.
20979 Operands are formatted in the following manner:
20984 Operand Name [lr live-from : live-to ] { type information } [ registers
20990 As mentioned earlier the live ranges are computed in terms of the execution
20991 sequence number of the iCodes, for example
20993 the iTemp0 is live from (i.e.
20994 first defined in iCode with execution sequence number 3, and is last used
20995 in the iCode with sequence number 5).
20996 For induction variables such as iTemp21 the live range computation extends
20997 the lifetime from the start to the end of the loop.
20999 The register allocator used the live range information to allocate registers,
21000 the same registers may be used for different temporaries if their live
21001 ranges do not overlap, for example r0 is allocated to both iTemp6 and to
21002 iTemp17 since their live ranges do not overlap.
21003 In addition the allocator also takes into consideration the type and usage
21004 of a temporary, for example itemp6 is a pointer to near space and is used
21005 as to fetch data from (i.e.
21006 used in GET_VALUE_AT_ADDRESS) so it is allocated a pointer register (r0).
21007 Some short lived temporaries are allocated to special registers which have
21008 meaning to the code generator e.g.
21009 iTemp13 is allocated to a pseudo register CC which tells the back end that
21010 the temporary is used only for a conditional jump the code generation makes
21011 use of this information to optimize a compare and jump ICode.
21013 There are several loop optimizations
21014 \begin_inset LatexCommand \index{Loop optimization}
21018 performed by the compiler.
21019 It can detect induction variables iTemp21(i) and iTemp23(j).
21020 Also note the compiler does selective strength reduction
21021 \begin_inset LatexCommand \index{Strength reduction}
21026 the multiplication of an induction variable in line 18 (gint = j * 3) is
21027 changed to addition, a new temporary iTemp17 is allocated and assigned
21028 a initial value, a constant 3 is then added for each iteration of the loop.
21029 The compiler does not change the multiplication
21030 \begin_inset LatexCommand \index{Multiplication}
21034 in line 17 however since the processor does support an 8 * 8 bit multiplication.
21036 Note the dead code elimination
21037 \begin_inset LatexCommand \index{Dead-code elimination}
21041 optimization eliminated the dead assignments in line 7 & 8 to I and sum
21049 Sample.c (5:1:0:0) _entry($9) :
21054 Sample.c(5:2:1:0) proc _function [lr0:0]{function short}
21059 Sample.c(11:3:2:0) iTemp0 [lr3:5]{_near * int}[r2] = recv
21064 Sample.c(11:4:53:0) preHeaderLbl0($11) :
21069 Sample.c(11:5:55:0) iTemp6 [lr5:16]{_near * int}[r0] := iTemp0 [lr3:5]{_near
21075 Sample.c(11:6:5:1) _whilecontinue_0($1) :
21080 Sample.c(11:7:7:1) iTemp4 [lr7:8]{int}[r2 r3] = @[iTemp6 [lr5:16]{_near *
21086 Sample.c(11:8:8:1) if iTemp4 [lr7:8]{int}[r2 r3] == 0 goto _whilebreak_0($3)
21091 Sample.c(11:9:14:1) iTemp7 [lr9:13]{_far * int}[DPTR] := _p [lr0:0]{_far
21097 Sample.c(11:10:15:1) _p [lr0:0]{_far * int} = _p [lr0:0]{_far * int} + 0x2
21103 Sample.c(11:13:18:1) iTemp10 [lr13:14]{int}[r2 r3] = @[iTemp7 [lr9:13]{_far
21109 Sample.c(11:14:19:1) *(iTemp6 [lr5:16]{_near * int}[r0]) := iTemp10 [lr13:14]{int
21115 Sample.c(11:15:12:1) iTemp6 [lr5:16]{_near * int}[r0] = iTemp6 [lr5:16]{_near
21116 * int}[r0] + 0x2 {short}
21121 Sample.c(11:16:20:1) goto _whilecontinue_0($1)
21126 Sample.c(11:17:21:0)_whilebreak_0($3) :
21131 Sample.c(12:18:22:0) iTemp2 [lr18:40]{short}[r2] := 0x0 {short}
21136 Sample.c(13:19:23:0) iTemp11 [lr19:40]{short}[r3] := 0x0 {short}
21141 Sample.c(15:20:54:0)preHeaderLbl1($13) :
21146 Sample.c(15:21:56:0) iTemp21 [lr21:38]{short}[r4] := 0x0 {short}
21151 Sample.c(15:22:57:0) iTemp23 [lr22:38]{int}[r5 r6] := 0xa {int}
21156 Sample.c(15:23:58:0) iTemp17 [lr23:38]{int}[r7 r0] := 0x1e {int}
21161 Sample.c(15:24:26:1)_forcond_0($4) :
21166 Sample.c(15:25:27:1) iTemp13 [lr25:26]{char}[CC] = iTemp21 [lr21:38]{short}[r4]
21172 Sample.c(15:26:28:1) if iTemp13 [lr25:26]{char}[CC] == 0 goto _forbreak_0($7)
21177 Sample.c(16:27:31:1) iTemp2 [lr18:40]{short}[r2] = iTemp2 [lr18:40]{short}[r2]
21178 + ITemp21 [lr21:38]{short}[r4]
21183 Sample.c(17:29:33:1) iTemp15 [lr29:30]{short}[r1] = iTemp21 [lr21:38]{short}[r4]
21189 Sample.c(17:30:34:1) iTemp11 [lr19:40]{short}[r3] = iTemp11 [lr19:40]{short}[r3]
21190 + iTemp15 [lr29:30]{short}[r1]
21195 Sample.c(18:32:36:1:1) iTemp17 [lr23:38]{int}[r7 r0]= iTemp17 [lr23:38]{int}[r7
21201 Sample.c(18:33:37:1) _gint [lr0:0]{int} = _gint [lr0:0]{int} + iTemp17 [lr23:38]{
21207 Sample.c(15:36:42:1) iTemp21 [lr21:38]{short}[r4] = iTemp21 [lr21:38]{short}[r4]
21213 Sample.c(15:37:45:1) iTemp23 [lr22:38]{int}[r5 r6]= iTemp23 [lr22:38]{int}[r5
21219 Sample.c(19:38:47:1) goto _forcond_0($4)
21224 Sample.c(19:39:48:0)_forbreak_0($7) :
21229 Sample.c(20:40:49:0) iTemp24 [lr40:41]{short}[DPTR] = iTemp2 [lr18:40]{short}[r2]
21230 + ITemp11 [lr19:40]{short}[r3]
21235 Sample.c(20:41:50:0) ret iTemp24 [lr40:41]{short}
21240 Sample.c(20:42:51:0)_return($8) :
21245 Sample.c(20:43:52:0) eproc _function [lr0:0]{ ia0 re0 rm0}{function short}
21251 Finally the code generated for this function:
21292 ; ----------------------------------------------
21297 ; function function
21302 ; ----------------------------------------------
21312 ; iTemp0 [lr3:5]{_near * int}[r2] = recv
21324 ; iTemp6 [lr5:16]{_near * int}[r0] := iTemp0 [lr3:5]{_near * int}[r2]
21336 ;_whilecontinue_0($1) :
21346 ; iTemp4 [lr7:8]{int}[r2 r3] = @[iTemp6 [lr5:16]{_near * int}[r0]]
21351 ; if iTemp4 [lr7:8]{int}[r2 r3] == 0 goto _whilebreak_0($3)
21410 ; iTemp7 [lr9:13]{_far * int}[DPTR] := _p [lr0:0]{_far * int}
21429 ; _p [lr0:0]{_far * int} = _p [lr0:0]{_far * int} + 0x2 {short}
21476 ; iTemp10 [lr13:14]{int}[r2 r3] = @[iTemp7 [lr9:13]{_far * int}[DPTR]]
21516 ; *(iTemp6 [lr5:16]{_near * int}[r0]) := iTemp10 [lr13:14]{int}[r2 r3]
21542 ; iTemp6 [lr5:16]{_near * int}[r0] =
21547 ; iTemp6 [lr5:16]{_near * int}[r0] +
21564 ; goto _whilecontinue_0($1)
21576 ; _whilebreak_0($3) :
21586 ; iTemp2 [lr18:40]{short}[r2] := 0x0 {short}
21598 ; iTemp11 [lr19:40]{short}[r3] := 0x0 {short}
21610 ; iTemp21 [lr21:38]{short}[r4] := 0x0 {short}
21622 ; iTemp23 [lr22:38]{int}[r5 r6] := 0xa {int}
21641 ; iTemp17 [lr23:38]{int}[r7 r0] := 0x1e {int}
21670 ; iTemp13 [lr25:26]{char}[CC] = iTemp21 [lr21:38]{short}[r4] < 0xa {short}
21675 ; if iTemp13 [lr25:26]{char}[CC] == 0 goto _forbreak_0($7)
21720 ; iTemp2 [lr18:40]{short}[r2] = iTemp2 [lr18:40]{short}[r2] +
21725 ; iTemp21 [lr21:38]{short}[r4]
21751 ; iTemp15 [lr29:30]{short}[r1] = iTemp21 [lr21:38]{short}[r4] * 0x3 {short}
21784 ; iTemp11 [lr19:40]{short}[r3] = iTemp11 [lr19:40]{short}[r3] +
21789 ; iTemp15 [lr29:30]{short}[r1]
21808 ; iTemp17 [lr23:38]{int}[r7 r0]= iTemp17 [lr23:38]{int}[r7 r0]- 0x3 {short}
21855 ; _gint [lr0:0]{int} = _gint [lr0:0]{int} + iTemp17 [lr23:38]{int}[r7 r0]
21902 ; iTemp21 [lr21:38]{short}[r4] = iTemp21 [lr21:38]{short}[r4] + 0x1 {short}
21914 ; iTemp23 [lr22:38]{int}[r5 r6]= iTemp23 [lr22:38]{int}[r5 r6]- 0x1 {short}
21928 cjne r5,#0xff,00104$
21940 ; goto _forcond_0($4)
21952 ; _forbreak_0($7) :
21962 ; ret iTemp24 [lr40:41]{short}
22005 A few words about basic block successors, predecessors and dominators
22008 Successors are basic blocks
22009 \begin_inset LatexCommand \index{Basic blocks}
22013 that might execute after this basic block.
22015 Predecessors are basic blocks that might execute before reaching this basic
22018 Dominators are basic blocks that WILL execute before reaching this basic
22052 a) succList of [BB2] = [BB4], of [BB3] = [BB4], of [BB1] = [BB2,BB3]
22055 b) predList of [BB2] = [BB1], of [BB3] = [BB1], of [BB4] = [BB2,BB3]
22058 c) domVect of [BB4] = BB1 ...
22059 here we are not sure if BB2 or BB3 was executed but we are SURE that BB1
22067 \begin_inset LatexCommand \url{http://sdcc.sourceforge.net#Who}
22077 Thanks to all the other volunteer developers who have helped with coding,
22078 testing, web-page creation, distribution sets, etc.
22079 You know who you are :-)
22086 This document was initially written by Sandeep Dutta
22089 All product names mentioned herein may be trademarks
22090 \begin_inset LatexCommand \index{Trademarks}
22094 of their respective companies.
22101 To avoid confusion, the installation and building options for SDCC itself
22102 (chapter 2) are not part of the index.
22106 \begin_inset LatexCommand \printindex{}