1 #LyX 1.2 created this file. For more info see http://www.lyx.org/
15 \use_numerical_citations 0
16 \paperorientation portrait
19 \paragraph_separation indent
21 \quotes_language swedish
29 Please note: double dashed longoptions (e.g.
30 --version) need three dashes in this document to be visable in html and
34 SDCC Compiler User Guide
38 \begin_inset LatexCommand \tableofcontents{}
55 is a Freeware, retargettable, optimizing ANSI-C compiler by
59 designed for 8 bit Microprocessors.
60 The current version targets Intel MCS51 based Microprocessors(8051,8052,
61 etc), Zilog Z80 based MCUs, and the Dallas DS80C390 variant.
62 It can be retargetted for other microprocessors, support for PIC, AVR and
63 186 is under development.
64 The entire source code for the compiler is distributed under GPL.
65 SDCC uses ASXXXX & ASLINK, a Freeware, retargettable assembler & linker.
66 SDCC has extensive language extensions suitable for utilizing various microcont
67 rollers and underlying hardware effectively.
72 In addition to the MCU specific optimizations SDCC also does a host of standard
76 global sub expression elimination,
79 loop optimizations (loop invariant, strength reduction of induction variables
83 constant folding & propagation,
99 For the back-end SDCC uses a global register allocation scheme which should
100 be well suited for other 8 bit MCUs.
105 The peep hole optimizer uses a rule based substitution mechanism which is
111 Supported data-types are:
114 char (8 bits, 1 byte),
117 short and int (16 bits, 2 bytes),
120 long (32 bit, 4 bytes)
127 The compiler also allows
129 inline assembler code
131 to be embedded anywhere in a function.
132 In addition, routines developed in assembly can also be called.
136 SDCC also provides an option (--cyclomatic) to report the relative complexity
138 These functions can then be further optimized, or hand coded in assembly
144 SDCC also comes with a companion source level debugger SDCDB, the debugger
145 currently uses ucSim a freeware simulator for 8051 and other micro-controllers.
150 The latest version can be downloaded from
151 \begin_inset LatexCommand \url{http://sdcc.sourceforge.net/}
163 All packages used in this compiler system are
171 ; source code for all the sub-packages (pre-processor, assemblers, linkers
172 etc) is distributed with the package.
173 This documentation is maintained using a freeware word processor (LyX).
175 This program is free software; you can redistribute it and/or modify it
176 under the terms of the GNU General Public License as published by the Free
177 Software Foundation; either version 2, or (at your option) any later version.
178 This program is distributed in the hope that it will be useful, but WITHOUT
179 ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS
180 FOR A PARTICULAR PURPOSE.
181 See the GNU General Public License for more details.
182 You should have received a copy of the GNU General Public License along
183 with this program; if not, write to the Free Software Foundation, 59 Temple
184 Place - Suite 330, Boston, MA 02111-1307, USA.
185 In other words, you are welcome to use, share and improve this program.
186 You are forbidden to forbid anyone else to use, share and improve what
188 Help stamp out software-hoarding!
191 Typographic conventions
194 Throughout this manual, we will use the following convention.
195 Commands you have to type in are printed in
203 Code samples are printed in
208 Interesting items and new terms are printed in
213 Compatibility with previous versions
216 This version has numerous bug fixes compared with the previous version.
217 But we also introduced some incompatibilities with older versions.
218 Not just for the fun of it, but to make the compiler more stable, efficient
225 short is now equivalent to int (16 bits), it used to be equivalent to char
226 (8 bits) which is not ANSI compliant
229 the default directory for gcc-builds where include, library and documention
230 files are stored is now in /usr/local/share
233 char type parameters to vararg functions are casted to int unless explicitly
250 will push a as an int and as a char resp.
253 option ---regextend has been removed
256 option ---noregparms has been removed
259 option ---stack-after-data has been removed
264 <pending: more incompatibilities?>
270 What do you need before you start installation of SDCC? A computer, and
272 The preferred method of installation is to compile SDCC from source using
274 For Windows some pre-compiled binary distributions are available for your
276 You should have some experience with command line tools and compiler use.
282 The SDCC home page at
283 \begin_inset LatexCommand \url{http://sdcc.sourceforge.net/}
287 is a great place to find distribution sets.
288 You can also find links to the user mailing lists that offer help or discuss
289 SDCC with other SDCC users.
290 Web links to other SDCC related sites can also be found here.
291 This document can be found in the DOC directory of the source package as
293 Some of the other tools (simulator and assembler) included with SDCC contain
294 their own documentation and can be found in the source distribution.
295 If you want the latest unreleased software, the complete source package
296 is available directly by anonymous CVS on cvs.sdcc.sourceforge.net.
299 Wishes for the future
302 There are (and always will be) some things that could be done.
303 Here are some I can think of:
310 char KernelFunction3(char p) at 0x340;
316 If you can think of some more, please send them to the list.
322 <pending: And then of course a proper index-table
323 \begin_inset LatexCommand \index{index}
336 The install paths, search paths and other options are defined when running
338 The defaults can be overriden by:
340 \labelwidthstring 00.00.0000
342 ---prefix see tabel below
344 \labelwidthstring 00.00.0000
346 ---exec_prefix see tabel below
348 \labelwidthstring 00.00.0000
350 ---bindir see tabel below
352 \labelwidthstring 00.00.0000
354 ---datadir see tabel below
356 \labelwidthstring 00.00.0000
358 docdir environment variable, see tabel below
360 \labelwidthstring 00.00.0000
362 include_dir_suffix environment variable, see tabel below
364 \labelwidthstring 00.00.0000
366 lib_dir_suffix environment variable, see tabel below
368 \labelwidthstring 00.00.0000
370 sdccconf_h_dir_separator environment variable, either / or
375 This character will only be used in sdccconf.h; don't forget it's a C-header,
376 therefore a double-backslash is needed there.
378 \labelwidthstring 00.00.0000
380 ---disable-mcs51-port Excludes the Intel mcs51 port
382 \labelwidthstring 00.00.0000
384 ---disable-gbz80-port Excludes the Gameboy gbz80 port
386 \labelwidthstring 00.00.0000
388 ---disable-z80-port Excludes the z80 port
390 \labelwidthstring 00.00.0000
392 ---disable-avr-port Excludes the AVR port
394 \labelwidthstring 00.00.0000
396 ---disable-ds390-port Excludes the DS390 port
398 \labelwidthstring 00.00.0000
400 ---disable-pic-port Excludes the PIC port
402 \labelwidthstring 00.00.0000
404 ---disable-xa51-port Excludes the XA51 port
406 \labelwidthstring 00.00.0000
408 ---disable-ucsim Disables configuring and building of ucsim
410 \labelwidthstring 00.00.0000
412 ---disable-device-lib-build Disables automatically building device libraries
414 \labelwidthstring 00.00.0000
416 ---disable-packihx Disables building packihx
418 \labelwidthstring 00.00.0000
420 ---enable-libgc Use the Bohem memory allocator.
421 Lower runtime footprint.
424 Furthermore the environment variables CC, CFLAGS, ...
425 the tools and their arguments can be influenced.
426 Please see `configure ---help` and the man/info pages of `configure` for
431 The names of the standard libraries STD_LIB, STD_INT_LIB, STD_LONG_LIB,
432 STD_FP_LIB, STD_DS390_LIB, STD_XA51_LIB and the environment variables SDCC_DIR_
433 NAME, SDCC_INCLUDE_NAME, SDCC_LIB_NAME are defined by `configure` too.
434 At the moment it's not possible to change the default settings (it was
435 simply never required.
439 These configure options are compiled into the binaries, and can only be
440 changed by rerunning 'configure' and recompiling SDCC.
441 The configure options are written in
445 to distinguish them from run time environment variables (see section search
451 \begin_inset Quotes sld
455 \begin_inset Quotes srd
458 are used by the SDCC team to build the official Win32 binaries.
459 The SDCC team uses Mingw32 to build the official Windows binaries, because
466 a gcc compiler and last but not least
469 the binaries can be built by cross compiling on Sourceforge's compile farm.
472 See the examples, how to pass the Win32 settings to 'configure'.
473 The other Win32 builds using Borland, VC or whatever don't use 'configure',
474 but a header file sdcc_vc_in.h is the same as sdccconf.h built by 'configure'
485 <lyxtabular version="3" rows="8" columns="3">
487 <column alignment="left" valignment="top" leftline="true" width="0in">
488 <column alignment="left" valignment="top" leftline="true" width="0in">
489 <column alignment="left" valignment="top" leftline="true" rightline="true" width="0in">
490 <row topline="true" bottomline="true">
491 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
499 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
507 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
517 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
527 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
535 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
547 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
557 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
567 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
579 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
589 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
601 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
617 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
627 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
639 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
651 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
661 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
673 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
689 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
699 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
707 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
716 <row topline="true" bottomline="true">
717 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
727 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
735 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
754 'configure' also computes relative paths.
755 This is needed for full relocatability of a binary package and to complete
756 search paths (see section search paths below):
762 <lyxtabular version="3" rows="4" columns="3">
764 <column alignment="left" valignment="top" leftline="true" width="0in">
765 <column alignment="left" valignment="top" leftline="true" width="0in">
766 <column alignment="left" valignment="top" leftline="true" rightline="true" width="0in">
767 <row topline="true" bottomline="true">
768 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
776 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
784 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
793 <row topline="true" bottomline="true">
794 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
804 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
812 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
823 <row bottomline="true">
824 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
834 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
842 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
851 <row bottomline="true">
852 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
862 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
870 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
892 ./configure ---prefix=
893 \begin_inset Quotes srd
897 \begin_inset Quotes srd
901 \begin_inset Quotes srd
905 \begin_inset Quotes srd
911 ./configure ---disable-avr-port ---disable-xa51-port
914 To crosscompile on linux for Mingw32 (see also 'sdcc/support/scripts/sdcc_mingw3
924 \begin_inset Quotes srd
928 \begin_inset Quotes srd
932 \begin_inset Quotes srd
936 \begin_inset Quotes srd
945 \begin_inset Quotes srd
948 i586-mingw32msvc-ranlib
949 \begin_inset Quotes srd
958 \begin_inset Quotes srd
961 i586-mingw32msvc-strip
962 \begin_inset Quotes srd
971 \begin_inset Quotes srd
975 \begin_inset Quotes srd
984 \begin_inset Quotes srd
988 \begin_inset Quotes srd
997 \begin_inset Quotes srd
1001 \begin_inset Quotes srd
1010 \begin_inset Quotes srd
1014 \begin_inset Quotes srd
1023 \begin_inset Quotes srd
1027 \begin_inset Quotes srd
1035 sdccconf_h_dir_separator=
1036 \begin_inset Quotes srd
1048 \begin_inset Quotes srd
1056 ---disable-device-lib-build
1066 ---host=i586-mingw32msvc ---build=unknown-unknown-linux-gnu
1070 \begin_inset Quotes sld
1074 \begin_inset Quotes srd
1077 compile on Cygwin for Mingw32(see also sdcc/support/scripts/sdcc_cygwin_mingw32)
1087 \begin_inset Quotes srd
1091 \begin_inset Quotes srd
1100 \begin_inset Quotes srd
1104 \begin_inset Quotes srd
1113 \begin_inset Quotes srd
1117 \begin_inset Quotes srd
1126 \begin_inset Quotes srd
1130 \begin_inset Quotes srd
1139 \begin_inset Quotes srd
1143 \begin_inset Quotes srd
1152 \begin_inset Quotes srd
1156 \begin_inset Quotes srd
1165 \begin_inset Quotes srd
1169 \begin_inset Quotes srd
1177 sdccconf_h_dir_separator=
1178 \begin_inset Quotes srd
1190 \begin_inset Quotes srd
1201 'configure' is quite slow on Cygwin (at least on windows before Win2000/XP).
1202 The option '--C' turns on caching, which gives a little bit extra speed.
1203 However if options are changed, it can be necessary to delete the config.cache
1211 Binary files (preprocessor, assembler and linker)
1215 \begin_inset Tabular
1216 <lyxtabular version="3" rows="2" columns="3">
1218 <column alignment="left" valignment="top" leftline="true" width="0in">
1219 <column alignment="left" valignment="top" leftline="true" width="0in">
1220 <column alignment="left" valignment="top" leftline="true" rightline="true" width="0in">
1221 <row topline="true" bottomline="true">
1222 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1230 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1238 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
1247 <row topline="true" bottomline="true">
1248 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1258 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1266 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
1292 \begin_inset Tabular
1293 <lyxtabular version="3" rows="2" columns="3">
1295 <column alignment="block" valignment="top" leftline="true" width="1.6in">
1296 <column alignment="left" valignment="top" leftline="true" width="0in">
1297 <column alignment="center" valignment="top" leftline="true" rightline="true" width="0in">
1298 <row topline="true" bottomline="true">
1299 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1307 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1315 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
1324 <row topline="true" bottomline="true">
1325 <cell alignment="left" valignment="top" topline="true" leftline="true" usebox="none">
1337 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1342 /usr/local/share/sdcc/include
1345 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
1371 is auto-appended by the compiler, e.g.
1372 small, large, z80, ds390 etc.)
1376 \begin_inset Tabular
1377 <lyxtabular version="3" rows="2" columns="3">
1379 <column alignment="left" valignment="top" leftline="true" width="0in">
1380 <column alignment="left" valignment="top" leftline="true" width="0in">
1381 <column alignment="left" valignment="top" leftline="true" rightline="true" width="0in">
1382 <row topline="true" bottomline="true">
1383 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1391 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1399 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
1408 <row topline="true" bottomline="true">
1409 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1416 $DATADIR/$LIB_DIR_SUFFIX
1419 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1424 /usr/local/share/sdcc/lib
1427 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
1453 \begin_inset Tabular
1454 <lyxtabular version="3" rows="2" columns="3">
1456 <column alignment="left" valignment="top" leftline="true" width="0in">
1457 <column alignment="left" valignment="top" leftline="true" width="0in">
1458 <column alignment="left" valignment="top" leftline="true" rightline="true" width="0in">
1459 <row topline="true" bottomline="true">
1460 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1468 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1476 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
1485 <row topline="true" bottomline="true">
1486 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1496 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1501 /usr/local/share/sdcc/doc
1504 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
1525 The install paths can still be changed during `make install` with e.g.:
1528 make install prefix=$(HOME)/local/sdcc
1531 Of course this doesn't change the search paths compiled into the binaries.
1537 Some search paths or parts of them are determined by configure variables
1542 , see section above).
1543 Further search paths are determined by environment variables during runtime.
1546 The paths searched when running the compiler are as follows (the first catch
1552 Binary files (preprocessor, assembler and linker)
1555 \begin_inset Tabular
1556 <lyxtabular version="3" rows="4" columns="3">
1558 <column alignment="left" valignment="top" leftline="true" width="0in">
1559 <column alignment="left" valignment="top" leftline="true" width="0in">
1560 <column alignment="left" valignment="top" leftline="true" rightline="true" width="0in">
1561 <row topline="true" bottomline="true">
1562 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1570 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1578 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
1587 <row topline="true">
1588 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1598 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1606 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
1617 <row topline="true">
1618 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1623 Path of argv[0] (if available)
1626 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1634 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
1643 <row topline="true" bottomline="true">
1644 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1652 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1660 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
1681 \begin_inset Tabular
1682 <lyxtabular version="3" rows="6" columns="3">
1684 <column alignment="block" valignment="top" leftline="true" width="1.5in">
1685 <column alignment="block" valignment="top" leftline="true" width="1.5in">
1686 <column alignment="left" valignment="top" leftline="true" rightline="true" width="0in">
1687 <row topline="true" bottomline="true">
1688 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1696 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1704 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
1713 <row topline="true">
1714 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1722 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1730 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
1739 <row topline="true">
1740 <cell alignment="left" valignment="top" topline="true" leftline="true" usebox="none">
1748 <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">
1765 <row topline="true">
1766 <cell alignment="left" valignment="top" topline="true" leftline="true" usebox="none">
1780 <cell alignment="left" valignment="top" topline="true" leftline="true" usebox="none">
1792 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
1803 <row topline="true">
1804 <cell alignment="left" valignment="top" topline="true" leftline="true" usebox="none">
1822 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
1872 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
1885 <row topline="true" bottomline="true">
1886 <cell alignment="left" valignment="top" topline="true" leftline="true" usebox="none">
1902 <cell alignment="left" valignment="top" topline="true" leftline="true" usebox="none">
1907 /usr/local/share/sdcc/
1912 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
1929 The option ---nostdinc disables the last two search paths.
1936 With the exception of
1937 \begin_inset Quotes sld
1941 \begin_inset Quotes srd
1948 is auto-appended by the compiler (e.g.
1949 small, large, z80, ds390 etc.).
1953 \begin_inset Tabular
1954 <lyxtabular version="3" rows="6" columns="3">
1956 <column alignment="block" valignment="top" leftline="true" width="1.7in">
1957 <column alignment="left" valignment="top" leftline="true" width="1.2in">
1958 <column alignment="block" valignment="top" leftline="true" rightline="true" width="1.2in">
1959 <row topline="true" bottomline="true">
1960 <cell alignment="left" valignment="top" topline="true" leftline="true" usebox="none">
1968 <cell alignment="left" valignment="top" topline="true" leftline="true" usebox="none">
1976 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
1985 <row topline="true">
1986 <cell alignment="left" valignment="top" topline="true" leftline="true" usebox="none">
1994 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
2002 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
2011 <row topline="true">
2012 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
2024 <cell alignment="left" valignment="top" topline="true" leftline="true" usebox="none">
2036 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
2051 <row topline="true">
2052 <cell alignment="left" valignment="top" topline="true" leftline="true" usebox="none">
2063 $LIB_DIR_SUFFIX/<model>
2066 <cell alignment="left" valignment="top" topline="true" leftline="true" usebox="none">
2080 <cell alignment="left" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
2097 <row topline="true">
2098 <cell alignment="left" valignment="top" topline="true" leftline="true" usebox="none">
2113 $LIB_DIR_SUFFIX/<model>
2116 <cell alignment="left" valignment="top" topline="true" leftline="true" usebox="none">
2169 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
2225 <row topline="true" bottomline="true">
2226 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
2235 $LIB_DIR_SUFFIX/<model>
2238 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
2243 /usr/local/share/sdcc/
2250 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
2266 Don't delete any of the stray spaces in the table above without checking
2267 the HTML output (last line)!
2273 The option ---nostdlib disables the last two search paths.
2277 \layout Subsubsection
2279 Building SDCC on Linux
2284 Download the source package
2286 either from the SDCC CVS repository or from the
2287 \begin_inset LatexCommand \url[nightly snapshots]{http://sdcc.sourceforge.net/snap.php}
2293 , it will be named something like sdcc
2306 Bring up a command line terminal, such as xterm.
2311 Unpack the file using a command like:
2314 "tar -xzf sdcc.src.tar.gz
2319 , this will create a sub-directory called sdcc with all of the sources.
2322 Change directory into the main SDCC directory, for example type:
2339 This configures the package for compilation on your system.
2355 All of the source packages will compile, this can take a while.
2371 This copies the binary executables, the include files, the libraries and
2372 the documentation to the install directories.
2373 \layout Subsubsection
2375 Building SDCC on OSX 2.x
2378 Follow the instruction for Linux.
2382 On OSX 2.x it was reported, that the default gcc (version 3.1 20020420 (prerelease
2383 )) fails to compile SDCC.
2384 Fortunately there's also gcc 2.9.x installed, which works fine.
2385 This compiler can be selected by running 'configure' with:
2388 ./configure CC=gcc2 CXX=g++2
2389 \layout Subsubsection
2391 Crosscompiling SDCC on Linux for Windows
2394 With the Mingw32 gcc crosscompiler it's easy to compile SDCC for Win32.
2395 See section 'Configure Options'.
2396 \layout Subsubsection
2398 Building SDCC on Windows
2401 With the exception of Cygwin the SDCC binaries uCsim and sdcdb can't be
2403 They use Unix-sockets, which are not available on Win32.
2404 \layout Subsubsection
2406 Windows Install Using a Binary Package
2409 Download the binary package and unpack it using your favorite unpacking
2410 tool (gunzip, WinZip, etc).
2411 This should unpack to a group of sub-directories.
2412 An example directory structure after unpacking the mingw32 package is:
2417 bin for the executables, c:
2425 lib for the include and libraries.
2428 Adjust your environment variable PATH to include the location of the bin
2429 directory or start sdcc using the full path.
2430 \layout Subsubsection
2432 Building SDCC using Cygwin and Mingw32
2435 For building and installing a Cygwin executable follow the instructions
2441 \begin_inset Quotes sld
2445 \begin_inset Quotes srd
2448 Win32-binary can be built, which will not need the Cygwin-DLL.
2449 For the necessary 'configure' options see section 'configure options' or
2450 the script 'sdcc/support/scripts/sdcc_cygwinmingw32'.
2454 In order to install Cygwin on Windows download setup.exe from
2455 \begin_inset LatexCommand \url[www.cygwin.com]{http://www.cygwin.com/}
2461 \begin_inset Quotes sld
2464 default text file type
2465 \begin_inset Quotes srd
2469 \begin_inset Quotes sld
2473 \begin_inset Quotes srd
2476 and download/install at least the following packages.
2477 Some packages are selected by default, others will be automatically selected
2478 because of dependencies with the manually selected packages.
2479 Never deselect these packages!
2488 gcc ; version 3.x is fine, no need to use the old 2.9x
2491 binutils ; selected with gcc
2497 rxvt ; a nice console, which makes life much easier under windoze (see below)
2500 man ; not really needed for building SDCC, but you'll miss it sooner or
2504 less ; not really needed for building SDCC, but you'll miss it sooner or
2508 cvs ; only if you use CVS access
2511 If you want to develop something you'll need:
2514 python ; for the regression tests
2517 gdb ; the gnu debugger, together with the nice GUI
2518 \begin_inset Quotes sld
2522 \begin_inset Quotes srd
2528 openssh ; to access the CF or commit changes
2531 autoconf and autoconf-devel ; if you want to fight with 'configure', don't
2532 use autoconf-stable!
2535 rxvt is a nice console with history.
2536 Replace in your cygwin.bat the line
2545 rxvt -sl 1000 -fn "Lucida Console-12" -sr -cr red
2548 -bg black -fg white -geometry 100x65 -e bash --login
2551 Text selected with the mouse is automatically copied to the clipboard, pasting
2552 works with shift-insert.
2556 The other good tip is to make sure you have no //c/-style paths anywhere,
2557 use /cygdrive/c/ instead.
2558 Using // invokes a network lookup which is very slow.
2560 \begin_inset Quotes sld
2564 \begin_inset Quotes srd
2567 is too long, you can change it with e.g.
2573 SDCC sources use the unix line ending LF.
2574 Life is much easier, if you store the source tree on a drive, which is
2575 mount in binary mode.
2576 And use an editor which can handle LF-only line endings.
2577 Make sure not to commit files with windows line endings.
2578 \layout Subsubsection
2580 Windows Install Using Microsoft Visual C++ 6.0/NET
2585 Download the source package
2587 either from the SDCC CVS repository or from the
2588 \begin_inset LatexCommand \url[nightly snapshots]{http://sdcc.sourceforge.net/snap.php}
2594 , it will be named something like sdcc
2601 SDCC is distributed with all the projects, workspaces, and files you need
2602 to build it using Visual C++ 6.0/NET.
2603 The workspace name is 'sdcc.dsw'.
2604 Please note that as it is now, all the executables are created in a folder
2608 Once built you need to copy the executables from sdcc
2612 bin before runnng SDCC.
2617 In order to build SDCC with Visual C++ 6.0/NET you need win32 executables
2618 of bison.exe, flex.exe, and gawk.exe.
2619 One good place to get them is
2620 \begin_inset LatexCommand \url[here]{http://unxutils.sourceforge.net}
2628 Download the file UnxUtils.zip.
2629 Now you have to install the utilities and setup Visual C++ so it can locate
2630 the required programs.
2631 Here there are two alternatives (choose one!):
2638 a) Extract UnxUtils.zip to your C:
2640 hard disk PRESERVING the original paths, otherwise bison won't work.
2641 (If you are using WinZip make certain that 'Use folder names' is selected)
2645 b) In the Visual C++ IDE click Tools, Options, select the Directory tab,
2646 in 'Show directories for:' select 'Executable files', and in the directories
2647 window add a new path: 'C:
2657 (As a side effect, you get a bunch of Unix utilities that could be useful,
2658 such as diff and patch.)
2665 This one avoids extracting a bunch of files you may not use, but requires
2670 a) Create a directory were to put the tools needed, or use a directory already
2678 b) Extract 'bison.exe', 'bison.hairy', 'bison.simple', 'flex.exe', and gawk.exe
2679 to such directory WITHOUT preserving the original paths.
2680 (If you are using WinZip make certain that 'Use folder names' is not selected)
2684 c) Rename bison.exe to '_bison.exe'.
2688 d) Create a batch file 'bison.bat' in 'C:
2692 ' and add these lines:
2712 _bison %1 %2 %3 %4 %5 %6 %7 %8 %9
2716 Steps 'c' and 'd' are needed because bison requires by default that the
2717 files 'bison.simple' and 'bison.hairy' reside in some weird Unix directory,
2718 '/usr/local/share/' I think.
2719 So it is necessary to tell bison where those files are located if they
2720 are not in such directory.
2721 That is the function of the environment variables BISON_SIMPLE and BISON_HAIRY.
2725 e) In the Visual C++ IDE click Tools, Options, select the Directory tab,
2726 in 'Show directories for:' select 'Executable files', and in the directories
2727 window add a new path: 'c:
2730 Note that you can use any other path instead of 'c:
2732 util', even the path where the Visual C++ tools are, probably: 'C:
2736 Microsoft Visual Studio
2741 So you don't have to execute step 'e' :)
2745 Open 'sdcc.dsw' in Visual Studio, click 'build all', when it finishes copy
2746 the executables from sdcc
2750 bin, and you can compile using sdcc.
2751 \layout Subsubsection
2753 Windows Install Using Borland
2756 From the sdcc directory, run the command "make -f Makefile.bcc".
2757 This should regenerate all the .exe files in the bin directory except for
2758 sdcdb.exe (which currently doesn't build under Borland C++).
2761 If you modify any source files and need to rebuild, be aware that the dependanci
2762 es may not be correctly calculated.
2763 The safest option is to delete all .obj files and run the build again.
2764 From a Cygwin BASH prompt, this can easily be done with the commmand:
2774 ( -name '*.obj' -o -name '*.lib' -o -name '*.rul'
2776 ) -print -exec rm {}
2785 or on Windows NT/2000/XP from the command prompt with the commmand:
2792 del /s *.obj *.lib *.rul
2795 from the sdcc directory.
2798 Building the Documentation
2805 Testing out the SDCC Compiler
2808 The first thing you should do after installing your SDCC compiler is to
2816 at the prompt, and the program should run and tell you the version.
2817 If it doesn't run, or gives a message about not finding sdcc program, then
2818 you need to check over your installation.
2819 Make sure that the sdcc bin directory is in your executable search path
2820 defined by the PATH environment setting (see the Trouble-shooting section
2822 Make sure that the sdcc program is in the bin folder, if not perhaps something
2823 did not install correctly.
2831 is commonly installed as described in section
2832 \begin_inset Quotes sld
2835 Install and search paths
2836 \begin_inset Quotes srd
2845 Make sure the compiler works on a very simple example.
2846 Type in the following test.c program using your favorite
2881 Compile this using the following command:
2890 If all goes well, the compiler will generate a test.asm and test.rel file.
2891 Congratulations, you've just compiled your first program with SDCC.
2892 We used the -c option to tell SDCC not to link the generated code, just
2893 to keep things simple for this step.
2901 The next step is to try it with the linker.
2911 If all goes well the compiler will link with the libraries and produce
2912 a test.ihx output file.
2917 (no test.ihx, and the linker generates warnings), then the problem is most
2918 likely that sdcc cannot find the
2922 usr/local/share/sdcc/lib directory
2926 (see the Install trouble-shooting section for suggestions).
2934 The final test is to ensure sdcc can use the
2938 header files and libraries.
2939 Edit test.c and change it to the following:
2959 strcpy(str1, "testing");
2968 Compile this by typing
2975 This should generate a test.ihx output file, and it should give no warnings
2976 such as not finding the string.h file.
2977 If it cannot find the string.h file, then the problem is that sdcc cannot
2978 find the /usr/local/share/sdcc/include directory
2982 (see the Install trouble-shooting section for suggestions).
2985 Install Trouble-shooting
2986 \layout Subsubsection
2988 SDCC does not build correctly.
2991 A thing to try is starting from scratch by unpacking the .tgz source package
2992 again in an empty directory.
3000 ./configure 2>&1 | tee configure.log
3014 make 2>&1 | tee make.log
3021 If anything goes wrong, you can review the log files to locate the problem.
3022 Or a relevant part of this can be attached to an email that could be helpful
3023 when requesting help from the mailing list.
3024 \layout Subsubsection
3027 \begin_inset Quotes sld
3031 \begin_inset Quotes srd
3038 \begin_inset Quotes sld
3042 \begin_inset Quotes srd
3045 command is a script that analyzes your system and performs some configuration
3046 to ensure the source package compiles on your system.
3047 It will take a few minutes to run, and will compile a few tests to determine
3048 what compiler features are installed.
3049 \layout Subsubsection
3052 \begin_inset Quotes sld
3056 \begin_inset Quotes srd
3062 This runs the GNU make tool, which automatically compiles all the source
3063 packages into the final installed binary executables.
3064 \layout Subsubsection
3067 \begin_inset Quotes sld
3071 \begin_inset Quotes erd
3077 This will install the compiler, other executables libraries and include
3078 files in to the appropriate directories.
3080 \begin_inset Quotes sld
3083 Install and Search PATHS
3084 \begin_inset Quotes srd
3089 On most systems you will need super-user privilages to do this.
3095 SDCC is not just a compiler, but a collection of tools by various developers.
3096 These include linkers, assemblers, simulators and other components.
3097 Here is a summary of some of the components.
3098 Note that the included simulator and assembler have separate documentation
3099 which you can find in the source package in their respective directories.
3100 As SDCC grows to include support for other processors, other packages from
3101 various developers are included and may have their own sets of documentation.
3105 You might want to look at the files which are installed in <installdir>.
3106 At the time of this writing, we find the following programs for gcc-builds:
3110 In <installdir>/bin:
3113 sdcc - The compiler.
3116 sdcpp - The C preprocessor.
3119 asx8051 - The assembler for 8051 type processors.
3126 as-gbz80 - The Z80 and GameBoy Z80 assemblers.
3129 aslink -The linker for 8051 type processors.
3136 link-gbz80 - The Z80 and GameBoy Z80 linkers.
3139 s51 - The ucSim 8051 simulator.
3142 sdcdb - The source debugger.
3145 packihx - A tool to pack (compress) Intel hex files.
3148 In <installdir>/share/sdcc/include
3154 In <installdir>/share/sdcc/lib
3157 the subdirs src and small, large, z80, gbz80 and ds390 with the precompiled
3161 In <installdir>/share/sdcc/doc
3167 As development for other processors proceeds, this list will expand to include
3168 executables to support processors like AVR, PIC, etc.
3169 \layout Subsubsection
3174 This is the actual compiler, it in turn uses the c-preprocessor and invokes
3175 the assembler and linkage editor.
3176 \layout Subsubsection
3178 sdcpp - The C-Preprocessor
3181 The preprocessor is a modified version of the GNU preprocessor.
3182 The C preprocessor is used to pull in #include sources, process #ifdef
3183 statements, #defines and so on.
3184 \layout Subsubsection
3186 asx8051, as-z80, as-gbz80, aslink, link-z80, link-gbz80 - The Assemblers
3190 This is retargettable assembler & linkage editor, it was developed by Alan
3192 John Hartman created the version for 8051, and I (Sandeep) have made some
3193 enhancements and bug fixes for it to work properly with the SDCC.
3194 \layout Subsubsection
3199 S51 is a freeware, opensource simulator developed by Daniel Drotos (
3200 \begin_inset LatexCommand \url{mailto:drdani@mazsola.iit.uni-miskolc.hu}
3205 The simulator is built as part of the build process.
3206 For more information visit Daniel's website at:
3207 \begin_inset LatexCommand \url{http://mazsola.iit.uni-miskolc.hu/~drdani/embedded/s51}
3212 It currently support the core mcs51, the Dallas DS80C390 and the Philips
3214 \layout Subsubsection
3216 sdcdb - Source Level Debugger
3222 <todo: is this thing still alive?>
3229 Sdcdb is the companion source level debugger.
3230 The current version of the debugger uses Daniel's Simulator S51, but can
3231 be easily changed to use other simulators.
3238 \layout Subsubsection
3240 Single Source File Projects
3243 For single source file 8051 projects the process is very simple.
3244 Compile your programs with the following command
3247 "sdcc sourcefile.c".
3251 This will compile, assemble and link your source file.
3252 Output files are as follows
3256 sourcefile.asm - Assembler source file created by the compiler
3258 sourcefile.lst - Assembler listing file created by the Assembler
3260 sourcefile.rst - Assembler listing file updated with linkedit information,
3261 created by linkage editor
3263 sourcefile.sym - symbol listing for the sourcefile, created by the assembler
3265 sourcefile.rel - Object file created by the assembler, input to Linkage editor
3267 sourcefile.map - The memory map for the load module, created by the Linker
3269 sourcefile.ihx - The load module in Intel hex format (you can select the
3270 Motorola S19 format with ---out-fmt-s19)
3272 sourcefile.cdb - An optional file (with ---debug) containing debug information
3274 sourcefile.dump* - Dump file to debug the compiler it self (with ---dumpall)
3276 \begin_inset Quotes sld
3279 Anatomy of the compiler
3280 \begin_inset Quotes srd
3284 \layout Subsubsection
3286 Projects with Multiple Source Files
3289 SDCC can compile only ONE file at a time.
3290 Let us for example assume that you have a project containing the following
3295 foo1.c (contains some functions)
3297 foo2.c (contains some more functions)
3299 foomain.c (contains more functions and the function main)
3307 The first two files will need to be compiled separately with the commands:
3339 Then compile the source file containing the
3343 function and link the files together with the following command:
3351 foomain.c\SpecialChar ~
3352 foo1.rel\SpecialChar ~
3364 can be separately compiled as well:
3375 sdcc foomain.rel foo1.rel foo2.rel
3382 The file containing the
3397 file specified in the command line, since the linkage editor processes
3398 file in the order they are presented to it.
3399 \layout Subsubsection
3401 Projects with Additional Libraries
3404 Some reusable routines may be compiled into a library, see the documentation
3405 for the assembler and linkage editor (which are in <installdir>/share/sdcc/doc)
3411 Libraries created in this manner can be included in the command line.
3412 Make sure you include the -L <library-path> option to tell the linker where
3413 to look for these files if they are not in the current directory.
3414 Here is an example, assuming you have the source file
3426 (if that is not the same as your current project):
3433 sdcc foomain.c foolib.lib -L mylib
3444 must be an absolute path name.
3448 The most efficient way to use libraries is to keep seperate modules in seperate
3450 The lib file now should name all the modules.rel files.
3451 For an example see the standard library file
3455 in the directory <installdir>/share/lib/small.
3458 Command Line Options
3459 \layout Subsubsection
3461 Processor Selection Options
3463 \labelwidthstring 00.00.0000
3469 Generate code for the MCS51 (8051) family of processors.
3470 This is the default processor target.
3472 \labelwidthstring 00.00.0000
3478 Generate code for the DS80C390 processor.
3480 \labelwidthstring 00.00.0000
3486 Generate code for the Z80 family of processors.
3488 \labelwidthstring 00.00.0000
3494 Generate code for the GameBoy Z80 processor.
3496 \labelwidthstring 00.00.0000
3502 Generate code for the Atmel AVR processor (In development, not complete).
3504 \labelwidthstring 00.00.0000
3510 Generate code for the PIC 14-bit processors (In development, not complete).
3512 \labelwidthstring 00.00.0000
3518 Generate code for the Toshiba TLCS-900H processor (In development, not
3521 \labelwidthstring 00.00.0000
3527 Generate code for the Philips XA51 processor (In development, not complete).
3528 \layout Subsubsection
3530 Preprocessor Options
3532 \labelwidthstring 00.00.0000
3538 The additional location where the pre processor will look for <..h> or
3539 \begin_inset Quotes eld
3543 \begin_inset Quotes erd
3548 \labelwidthstring 00.00.0000
3554 Command line definition of macros.
3555 Passed to the pre processor.
3557 \labelwidthstring 00.00.0000
3563 Tell the preprocessor to output a rule suitable for make describing the
3564 dependencies of each object file.
3565 For each source file, the preprocessor outputs one make-rule whose target
3566 is the object file name for that source file and whose dependencies are
3567 all the files `#include'd in it.
3568 This rule may be a single line or may be continued with `
3570 '-newline if it is long.
3571 The list of rules is printed on standard output instead of the preprocessed
3575 \labelwidthstring 00.00.0000
3581 Tell the preprocessor not to discard comments.
3582 Used with the `-E' option.
3584 \labelwidthstring 00.00.0000
3595 Like `-M' but the output mentions only the user header files included with
3597 \begin_inset Quotes eld
3601 System header files included with `#include <file>' are omitted.
3603 \labelwidthstring 00.00.0000
3609 Assert the answer answer for question, in case it is tested with a preprocessor
3610 conditional such as `#if #question(answer)'.
3611 `-A-' disables the standard assertions that normally describe the target
3614 \labelwidthstring 00.00.0000
3620 (answer) Assert the answer answer for question, in case it is tested with
3621 a preprocessor conditional such as `#if #question(answer)'.
3622 `-A-' disables the standard assertions that normally describe the target
3625 \labelwidthstring 00.00.0000
3631 Undefine macro macro.
3632 `-U' options are evaluated after all `-D' options, but before any `-include'
3633 and `-imacros' options.
3635 \labelwidthstring 00.00.0000
3641 Tell the preprocessor to output only a list of the macro definitions that
3642 are in effect at the end of preprocessing.
3643 Used with the `-E' option.
3645 \labelwidthstring 00.00.0000
3651 Tell the preprocessor to pass all macro definitions into the output, in
3652 their proper sequence in the rest of the output.
3654 \labelwidthstring 00.00.0000
3665 Like `-dD' except that the macro arguments and contents are omitted.
3666 Only `#define name' is included in the output.
3667 \layout Subsubsection
3671 \labelwidthstring 00.00.0000
3681 <absolute path to additional libraries> This option is passed to the linkage
3682 editor's additional libraries search path.
3683 The path name must be absolute.
3684 Additional library files may be specified in the command line.
3685 See section Compiling programs for more details.
3687 \labelwidthstring 00.00.0000
3693 <Value> The start location of the external ram, default value is 0.
3694 The value entered can be in Hexadecimal or Decimal format, e.g.: ---xram-loc
3695 0x8000 or ---xram-loc 32768.
3697 \labelwidthstring 00.00.0000
3703 <Value> The start location of the code segment, default value 0.
3704 Note when this option is used the interrupt vector table is also relocated
3705 to the given address.
3706 The value entered can be in Hexadecimal or Decimal format, e.g.: ---code-loc
3707 0x8000 or ---code-loc 32768.
3709 \labelwidthstring 00.00.0000
3715 <Value> By default the stack is placed after the data segment.
3716 Using this option the stack can be placed anywhere in the internal memory
3718 The value entered can be in Hexadecimal or Decimal format, e.g.
3719 ---stack-loc 0x20 or ---stack-loc 32.
3720 Since the sp register is incremented before a push or call, the initial
3721 sp will be set to one byte prior the provided value.
3722 The provided value should not overlap any other memory areas such as used
3723 register banks or the data segment and with enough space for the current
3726 \labelwidthstring 00.00.0000
3732 <Value> The start location of the internal ram data segment.
3733 The value entered can be in Hexadecimal or Decimal format, eg.
3734 ---data-loc 0x20 or ---data-loc 32.
3735 (By default, the start location of the internal ram data segment is set
3736 as low as possible in memory, taking into account the used register banks
3737 and the bit segment at address 0x20.
3738 For example if register banks 0 and 1 are used without bit variables, the
3739 data segment will be set, if ---data-loc is not used, to location 0x10.)
3741 \labelwidthstring 00.00.0000
3747 <Value> The start location of the indirectly addressable internal ram, default
3749 The value entered can be in Hexadecimal or Decimal format, eg.
3750 ---idata-loc 0x88 or ---idata-loc 136.
3752 \labelwidthstring 00.00.0000
3761 The linker output (final object code) is in Intel Hex format.
3762 (This is the default option).
3764 \labelwidthstring 00.00.0000
3773 The linker output (final object code) is in Motorola S19 format.
3774 \layout Subsubsection
3778 \labelwidthstring 00.00.0000
3784 Generate code for Large model programs see section Memory Models for more
3786 If this option is used all source files in the project should be compiled
3788 In addition the standard library routines are compiled with small model,
3789 they will need to be recompiled.
3791 \labelwidthstring 00.00.0000
3802 Generate code for Small Model programs see section Memory Models for more
3804 This is the default model.
3805 \layout Subsubsection
3809 \labelwidthstring 00.00.0000
3820 Generate 24-bit flat mode code.
3821 This is the one and only that the ds390 code generator supports right now
3822 and is default when using
3827 See section Memory Models for more details.
3829 \labelwidthstring 00.00.0000
3835 Generate code for the 10 bit stack mode of the Dallas DS80C390 part.
3836 This is the one and only that the ds390 code generator supports right now
3837 and is default when using
3842 In this mode, the stack is located in the lower 1K of the internal RAM,
3843 which is mapped to 0x400000.
3844 Note that the support is incomplete, since it still uses a single byte
3845 as the stack pointer.
3846 This means that only the lower 256 bytes of the potential 1K stack space
3847 will actually be used.
3848 However, this does allow you to reclaim the precious 256 bytes of low RAM
3849 for use for the DATA and IDATA segments.
3850 The compiler will not generate any code to put the processor into 10 bit
3852 It is important to ensure that the processor is in this mode before calling
3853 any re-entrant functions compiled with this option.
3854 In principle, this should work with the
3858 option, but that has not been tested.
3859 It is incompatible with the
3864 It also only makes sense if the processor is in 24 bit contiguous addressing
3867 ---model-flat24 option
3870 \layout Subsubsection
3872 Optimization Options
3874 \labelwidthstring 00.00.0000
3880 Will not do global subexpression elimination, this option may be used when
3881 the compiler creates undesirably large stack/data spaces to store compiler
3883 A warning message will be generated when this happens and the compiler
3884 will indicate the number of extra bytes it allocated.
3885 It recommended that this option NOT be used, #pragma\SpecialChar ~
3887 to turn off global subexpression elimination for a given function only.
3889 \labelwidthstring 00.00.0000
3895 Will not do loop invariant optimizations, this may be turned off for reasons
3896 explained for the previous option.
3897 For more details of loop optimizations performed see section Loop Invariants.It
3898 recommended that this option NOT be used, #pragma\SpecialChar ~
3899 NOINVARIANT can be used
3900 to turn off invariant optimizations for a given function only.
3902 \labelwidthstring 00.00.0000
3908 Will not do loop induction optimizations, see section strength reduction
3909 for more details.It is recommended that this option is NOT used, #pragma\SpecialChar ~
3911 ION can be used to turn off induction optimizations for a given function
3914 \labelwidthstring 00.00.0000
3925 Will not generate boundary condition check when switch statements are implement
3926 ed using jump-tables.
3927 See section Switch Statements for more details.
3928 It is recommended that this option is NOT used, #pragma\SpecialChar ~
3930 used to turn off boundary checking for jump tables for a given function
3933 \labelwidthstring 00.00.0000
3942 Will not do loop reversal optimization.
3944 \labelwidthstring 00.00.0000
3950 Will not optimize labels (makes the dumpfiles more readable).
3952 \labelwidthstring 00.00.0000
3958 Will not memcpy initialized data in far space from code space.
3959 This saves a few bytes in code space if you don't have initialized data.
3960 \layout Subsubsection
3964 \labelwidthstring 00.00.0000
3971 will compile and assemble the source, but will not call the linkage editor.
3973 \labelwidthstring 00.00.0000
3979 reads the preprocessed source from standard input and compiles it.
3980 The file name for the assembler output must be specified using the -o option.
3982 \labelwidthstring 00.00.0000
3988 Run only the C preprocessor.
3989 Preprocess all the C source files specified and output the results to standard
3992 \labelwidthstring 00.00.0000
3999 The output path resp.
4000 file where everything will be placed.
4001 If the parameter is a path, it must have a trailing slash (or backslash
4002 for the Windows binaries) to be recognized as a path.
4005 \labelwidthstring 00.00.0000
4016 All functions in the source file will be compiled as
4021 the parameters and local variables will be allocated on the stack.
4022 see section Parameters and Local Variables for more details.
4023 If this option is used all source files in the project should be compiled
4027 \labelwidthstring 00.00.0000
4033 Uses a pseudo stack in the first 256 bytes in the external ram for allocating
4034 variables and passing parameters.
4035 See section on external stack for more details.
4037 \labelwidthstring 00.00.0000
4041 ---callee-saves function1[,function2][,function3]....
4044 The compiler by default uses a caller saves convention for register saving
4045 across function calls, however this can cause unneccessary register pushing
4046 & popping when calling small functions from larger functions.
4047 This option can be used to switch the register saving convention for the
4048 function names specified.
4049 The compiler will not save registers when calling these functions, no extra
4050 code will be generated at the entry & exit for these functions to save
4051 & restore the registers used by these functions, this can SUBSTANTIALLY
4052 reduce code & improve run time performance of the generated code.
4053 In the future the compiler (with interprocedural analysis) will be able
4054 to determine the appropriate scheme to use for each function call.
4055 DO NOT use this option for built-in functions such as _muluint..., if this
4056 option is used for a library function the appropriate library function
4057 needs to be recompiled with the same option.
4058 If the project consists of multiple source files then all the source file
4059 should be compiled with the same ---callee-saves option string.
4060 Also see #pragma\SpecialChar ~
4063 \labelwidthstring 00.00.0000
4072 When this option is used the compiler will generate debug information, that
4073 can be used with the SDCDB.
4074 The debug information is collected in a file with .cdb extension.
4075 For more information see documentation for SDCDB.
4077 \labelwidthstring 00.00.0000
4083 <filename> This option can be used to use additional rules to be used by
4084 the peep hole optimizer.
4085 See section Peep Hole optimizations for details on how to write these rules.
4087 \labelwidthstring 00.00.0000
4098 Stop after the stage of compilation proper; do not assemble.
4099 The output is an assembler code file for the input file specified.
4101 \labelwidthstring 00.00.0000
4105 -Wa_asmOption[,asmOption]
4108 Pass the asmOption to the assembler.
4110 \labelwidthstring 00.00.0000
4114 -Wl_linkOption[,linkOption]
4117 Pass the linkOption to the linker.
4119 \labelwidthstring 00.00.0000
4128 Integer (16 bit) and long (32 bit) libraries have been compiled as reentrant.
4129 Note by default these libraries are compiled as non-reentrant.
4130 See section Installation for more details.
4132 \labelwidthstring 00.00.0000
4141 This option will cause the compiler to generate an information message for
4142 each function in the source file.
4143 The message contains some
4147 information about the function.
4148 The number of edges and nodes the compiler detected in the control flow
4149 graph of the function, and most importantly the
4151 cyclomatic complexity
4153 see section on Cyclomatic Complexity for more details.
4155 \labelwidthstring 00.00.0000
4164 Floating point library is compiled as reentrant.See section Installation
4167 \labelwidthstring 00.00.0000
4173 The compiler will not overlay parameters and local variables of any function,
4174 see section Parameters and local variables for more details.
4176 \labelwidthstring 00.00.0000
4182 This option can be used when the code generated is called by a monitor
4184 The compiler will generate a 'ret' upon return from the 'main' function.
4185 The default option is to lock up i.e.
4188 \labelwidthstring 00.00.0000
4194 Disable peep-hole optimization.
4196 \labelwidthstring 00.00.0000
4202 Pass the inline assembler code through the peep hole optimizer.
4203 This can cause unexpected changes to inline assembler code, please go through
4204 the peephole optimizer rules defined in the source file tree '<target>/peeph.def
4205 ' before using this option.
4207 \labelwidthstring 00.00.0000
4213 <Value> Causes the linker to check if the internal ram usage is within limits
4216 \labelwidthstring 00.00.0000
4222 <Value> Causes the linker to check if the external ram usage is within limits
4225 \labelwidthstring 00.00.0000
4231 <Value> Causes the linker to check if the code usage is within limits of
4234 \labelwidthstring 00.00.0000
4240 This will prevent the compiler from passing on the default include path
4241 to the preprocessor.
4243 \labelwidthstring 00.00.0000
4249 This will prevent the compiler from passing on the default library path
4252 \labelwidthstring 00.00.0000
4258 Shows the various actions the compiler is performing.
4260 \labelwidthstring 00.00.0000
4266 Shows the actual commands the compiler is executing.
4268 \labelwidthstring 00.00.0000
4274 Hides your ugly and inefficient c-code from the asm file, so you can always
4275 blame the compiler :).
4277 \labelwidthstring 00.00.0000
4283 Include i-codes in the asm file.
4284 Sounds like noise but is most helpfull for debugging the compiler itself.
4286 \labelwidthstring 00.00.0000
4292 Disable some of the more pedantic warnings (jwk burps: please be more specific
4295 \labelwidthstring 00.00.0000
4299 ---print-search-dirs
4301 Display the directories in the compiler's search path
4302 \layout Subsubsection
4304 Intermediate Dump Options
4307 The following options are provided for the purpose of retargetting and debugging
4309 These provided a means to dump the intermediate code (iCode) generated
4310 by the compiler in human readable form at various stages of the compilation
4314 \labelwidthstring 00.00.0000
4320 This option will cause the compiler to dump the intermediate code into
4323 <source filename>.dumpraw
4325 just after the intermediate code has been generated for a function, i.e.
4326 before any optimizations are done.
4327 The basic blocks at this stage ordered in the depth first number, so they
4328 may not be in sequence of execution.
4330 \labelwidthstring 00.00.0000
4336 Will create a dump of iCode's, after global subexpression elimination,
4339 <source filename>.dumpgcse.
4341 \labelwidthstring 00.00.0000
4347 Will create a dump of iCode's, after deadcode elimination, into a file
4350 <source filename>.dumpdeadcode.
4352 \labelwidthstring 00.00.0000
4361 Will create a dump of iCode's, after loop optimizations, into a file named
4364 <source filename>.dumploop.
4366 \labelwidthstring 00.00.0000
4375 Will create a dump of iCode's, after live range analysis, into a file named
4378 <source filename>.dumprange.
4380 \labelwidthstring 00.00.0000
4386 Will dump the life ranges for all symbols.
4388 \labelwidthstring 00.00.0000
4397 Will create a dump of iCode's, after register assignment, into a file named
4400 <source filename>.dumprassgn.
4402 \labelwidthstring 00.00.0000
4408 Will create a dump of the live ranges of iTemp's
4410 \labelwidthstring 00.00.0000
4421 Will cause all the above mentioned dumps to be created.
4424 Environment variables
4427 SDCC recognizes the following environment variables:
4429 \labelwidthstring 00.00.0000
4435 SDCC installs a signal handler to be able to delete temporary files after
4436 an user break (^C) or an exception.
4437 If this environment variable is set, SDCC won't install the signal handler
4438 in order to be able to debug SDCC.
4440 \labelwidthstring 00.00.0000
4448 Path, where temporary files will be created.
4449 The order of the variables is the search order.
4450 In a standard *nix environment these variables are not set, and there's
4451 no need to set them.
4452 On Windows it's recommended to set one of them.
4454 \labelwidthstring 00.00.0000
4461 \begin_inset Quotes sld
4464 2.3 Install and search paths
4465 \begin_inset Quotes srd
4470 \labelwidthstring 00.00.0000
4477 \begin_inset Quotes sld
4480 2.3 Install and search paths
4481 \begin_inset Quotes srd
4486 \labelwidthstring 00.00.0000
4493 \begin_inset Quotes sld
4496 2.3 Install and search paths
4497 \begin_inset Quotes srd
4503 There are some more environment variables recognized by SDCC, but these
4504 are solely used for debugging purposes.
4505 They can change or disappear very quickly, and will never be documentated.
4508 MCS51/DS390 Storage Class Language Extensions
4511 In addition to the ANSI storage classes SDCC allows the following MCS51
4512 specific storage classes.
4513 \layout Subsubsection
4518 Variables declared with this storage class will be placed in the extern
4524 storage class for Large Memory model, e.g.:
4530 xdata unsigned char xduc;
4531 \layout Subsubsection
4540 storage class for Small Memory model.
4541 Variables declared with this storage class will be allocated in the internal
4549 \layout Subsubsection
4554 Variables declared with this storage class will be allocated into the indirectly
4555 addressable portion of the internal ram of a 8051, e.g.:
4562 \layout Subsubsection
4567 This is a data-type and a storage class specifier.
4568 When a variable is declared as a bit, it is allocated into the bit addressable
4569 memory of 8051, e.g.:
4576 \layout Subsubsection
4581 Like the bit keyword,
4585 signifies both a data-type and storage class, they are used to describe
4586 the special function registers and special bit variables of a 8051, eg:
4592 sfr at 0x80 P0; /* special function register P0 at location 0x80 */
4594 sbit at 0xd7 CY; /* CY (Carry Flag) */
4600 SDCC allows (via language extensions) pointers to explicitly point to any
4601 of the memory spaces of the 8051.
4602 In addition to the explicit pointers, the compiler uses (by default) generic
4603 pointers which can be used to point to any of the memory spaces.
4607 Pointer declaration examples:
4616 /* pointer physically in xternal ram pointing to object in internal ram
4619 data unsigned char * xdata p;
4623 /* pointer physically in code rom pointing to data in xdata space */
4625 xdata unsigned char * code p;
4629 /* pointer physically in code space pointing to data in code space */
4631 code unsigned char * code p;
4635 /* the folowing is a generic pointer physically located in xdata space */
4646 Well you get the idea.
4651 All unqualified pointers are treated as 3-byte (4-byte for the ds390)
4664 The highest order byte of the
4668 pointers contains the data space information.
4669 Assembler support routines are called whenever data is stored or retrieved
4675 These are useful for developing reusable library routines.
4676 Explicitly specifying the pointer type will generate the most efficient
4680 Parameters & Local Variables
4683 Automatic (local) variables and parameters to functions can either be placed
4684 on the stack or in data-space.
4685 The default action of the compiler is to place these variables in the internal
4686 RAM (for small model) or external RAM (for large model).
4687 This in fact makes them
4691 so by default functions are non-reentrant.
4695 They can be placed on the stack either by using the
4699 option or by using the
4703 keyword in the function declaration, e.g.:
4712 unsigned char foo(char i) reentrant
4725 Since stack space on 8051 is limited, the
4733 option should be used sparingly.
4734 Note that the reentrant keyword just means that the parameters & local
4735 variables will be allocated to the stack, it
4739 mean that the function is register bank independent.
4743 Local variables can be assigned storage classes and absolute addresses,
4750 unsigned char foo() {
4756 xdata unsigned char i;
4768 data at 0x31 unsiged char j;
4783 In the above example the variable
4787 will be allocated in the external ram,
4791 in bit addressable space and
4800 or when a function is declared as
4804 this should only be done for static variables.
4807 Parameters however are not allowed any storage class, (storage classes for
4808 parameters will be ignored), their allocation is governed by the memory
4809 model in use, and the reentrancy options.
4815 For non-reentrant functions SDCC will try to reduce internal ram space usage
4816 by overlaying parameters and local variables of a function (if possible).
4817 Parameters and local variables of a function will be allocated to an overlayabl
4818 e segment if the function has
4820 no other function calls and the function is non-reentrant and the memory
4824 If an explicit storage class is specified for a local variable, it will
4828 Note that the compiler (not the linkage editor) makes the decision for overlayin
4830 Functions that are called from an interrupt service routine should be preceded
4831 by a #pragma\SpecialChar ~
4832 NOOVERLAY if they are not reentrant.
4835 Also note that the compiler does not do any processing of inline assembler
4836 code, so the compiler might incorrectly assign local variables and parameters
4837 of a function into the overlay segment if the inline assembler code calls
4838 other c-functions that might use the overlay.
4839 In that case the #pragma\SpecialChar ~
4840 NOOVERLAY should be used.
4843 Parameters and Local variables of functions that contain 16 or 32 bit multiplica
4844 tion or division will NOT be overlayed since these are implemented using
4845 external functions, e.g.:
4855 void set_error(unsigned char errcd)
4871 void some_isr () interrupt 2 using 1
4900 In the above example the parameter
4908 would be assigned to the overlayable segment if the #pragma\SpecialChar ~
4910 not present, this could cause unpredictable runtime behavior when called
4912 The #pragma\SpecialChar ~
4913 NOOVERLAY ensures that the parameters and local variables for
4914 the function are NOT overlayed.
4917 Interrupt Service Routines
4920 SDCC allows interrupt service routines to be coded in C, with some extended
4927 void timer_isr (void) interrupt 2 using 1
4940 The number following the
4944 keyword is the interrupt number this routine will service.
4945 The compiler will insert a call to this routine in the interrupt vector
4946 table for the interrupt number specified.
4951 keyword is used to tell the compiler to use the specified register bank
4952 (8051 specific) when generating code for this function.
4953 Note that when some function is called from an interrupt service routine
4954 it should be preceded by a #pragma\SpecialChar ~
4955 NOOVERLAY if it is not reentrant.
4956 A special note here, int (16 bit) and long (32 bit) integer division, multiplic
4957 ation & modulus operations are implemented using external support routines
4958 developed in ANSI-C, if an interrupt service routine needs to do any of
4959 these operations then the support routines (as mentioned in a following
4960 section) will have to be recompiled using the
4964 option and the source file will need to be compiled using the
4971 If you have multiple source files in your project, interrupt service routines
4972 can be present in any of them, but a prototype of the isr MUST be present
4973 or included in the file that contains the function
4980 Interrupt Numbers and the corresponding address & descriptions for the Standard
4981 8051 are listed below.
4982 SDCC will automatically adjust the interrupt vector table to the maximum
4983 interrupt number specified.
4989 \begin_inset Tabular
4990 <lyxtabular version="3" rows="6" columns="3">
4992 <column alignment="center" valignment="top" leftline="true" width="0in">
4993 <column alignment="center" valignment="top" leftline="true" width="0in">
4994 <column alignment="center" valignment="top" leftline="true" rightline="true" width="0in">
4995 <row topline="true" bottomline="true">
4996 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
5004 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
5012 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
5021 <row topline="true">
5022 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
5030 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
5038 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
5047 <row topline="true">
5048 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
5056 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
5064 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
5073 <row topline="true">
5074 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
5082 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
5090 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
5099 <row topline="true">
5100 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
5108 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
5116 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
5125 <row topline="true" bottomline="true">
5126 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
5134 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
5142 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
5159 If the interrupt service routine is defined without
5163 a register bank or with register bank 0 (using 0), the compiler will save
5164 the registers used by itself on the stack upon entry and restore them at
5165 exit, however if such an interrupt service routine calls another function
5166 then the entire register bank will be saved on the stack.
5167 This scheme may be advantageous for small interrupt service routines which
5168 have low register usage.
5171 If the interrupt service routine is defined to be using a specific register
5176 are save and restored, if such an interrupt service routine calls another
5177 function (using another register bank) then the entire register bank of
5178 the called function will be saved on the stack.
5179 This scheme is recommended for larger interrupt service routines.
5182 Calling other functions from an interrupt service routine is not recommended,
5183 avoid it if possible.
5187 Also see the _naked modifier.
5195 <TODO: this isn't implemented at all!>
5201 A special keyword may be associated with a function declaring it as
5206 SDCC will generate code to disable all interrupts upon entry to a critical
5207 function and enable them back before returning.
5208 Note that nesting critical functions may cause unpredictable results.
5233 The critical attribute maybe used with other attributes like
5241 A special keyword may be associated with a function declaring it as
5250 function modifier attribute prevents the compiler from generating prologue
5251 and epilogue code for that function.
5252 This means that the user is entirely responsible for such things as saving
5253 any registers that may need to be preserved, selecting the proper register
5254 bank, generating the
5258 instruction at the end, etc.
5259 Practically, this means that the contents of the function must be written
5260 in inline assembler.
5261 This is particularly useful for interrupt functions, which can have a large
5262 (and often unnecessary) prologue/epilogue.
5263 For example, compare the code generated by these two functions:
5269 data unsigned char counter;
5271 void simpleInterrupt(void) interrupt 1
5285 void nakedInterrupt(void) interrupt 2 _naked
5318 ; MUST explicitly include ret in _naked function.
5332 For an 8051 target, the generated simpleInterrupt looks like:
5477 whereas nakedInterrupt looks like:
5502 ; MUST explicitly include ret(i) in _naked function.
5508 While there is nothing preventing you from writing C code inside a _naked
5509 function, there are many ways to shoot yourself in the foot doing this,
5510 and it is recommended that you stick to inline assembler.
5513 Functions using private banks
5520 attribute (which tells the compiler to use a register bank other than the
5521 default bank zero) should only be applied to
5525 functions (see note 1 below).
5526 This will in most circumstances make the generated ISR code more efficient
5527 since it will not have to save registers on the stack.
5534 attribute will have no effect on the generated code for a
5538 function (but may occasionally be useful anyway
5544 possible exception: if a function is called ONLY from 'interrupt' functions
5545 using a particular bank, it can be declared with the same 'using' attribute
5546 as the calling 'interrupt' functions.
5547 For instance, if you have several ISRs using bank one, and all of them
5548 call memcpy(), it might make sense to create a specialized version of memcpy()
5549 'using 1', since this would prevent the ISR from having to save bank zero
5550 to the stack on entry and switch to bank zero before calling the function
5557 (pending: I don't think this has been done yet)
5564 function using a non-zero bank will assume that it can trash that register
5565 bank, and will not save it.
5566 Since high-priority interrupts can interrupt low-priority ones on the 8051
5567 and friends, this means that if a high-priority ISR
5571 a particular bank occurs while processing a low-priority ISR
5575 the same bank, terrible and bad things can happen.
5576 To prevent this, no single register bank should be
5580 by both a high priority and a low priority ISR.
5581 This is probably most easily done by having all high priority ISRs use
5582 one bank and all low priority ISRs use another.
5583 If you have an ISR which can change priority at runtime, you're on your
5584 own: I suggest using the default bank zero and taking the small performance
5588 It is most efficient if your ISR calls no other functions.
5589 If your ISR must call other functions, it is most efficient if those functions
5590 use the same bank as the ISR (see note 1 below); the next best is if the
5591 called functions use bank zero.
5592 It is very inefficient to call a function using a different, non-zero bank
5600 Data items can be assigned an absolute address with the
5604 keyword, in addition to a storage class, e.g.:
5610 xdata at 0x8000 unsigned char PORTA_8255 ;
5616 In the above example the PORTA_8255 will be allocated to the location 0x8000
5617 of the external ram.
5618 Note that this feature is provided to give the programmer access to
5622 devices attached to the controller.
5623 The compiler does not actually reserve any space for variables declared
5624 in this way (they are implemented with an equate in the assembler).
5625 Thus it is left to the programmer to make sure there are no overlaps with
5626 other variables that are declared without the absolute address.
5627 The assembler listing file (.lst) and the linker output files (.rst) and
5628 (.map) are a good places to look for such overlaps.
5632 Absolute address can be specified for variables in all storage classes,
5645 The above example will allocate the variable at offset 0x02 in the bit-addressab
5647 There is no real advantage to assigning absolute addresses to variables
5648 in this manner, unless you want strict control over all the variables allocated.
5654 The compiler inserts a call to the C routine
5656 _sdcc_external_startup()
5661 at the start of the CODE area.
5662 This routine is in the runtime library.
5663 By default this routine returns 0, if this routine returns a non-zero value,
5664 the static & global variable initialization will be skipped and the function
5665 main will be invoked Other wise static & global variables will be initialized
5666 before the function main is invoked.
5669 _sdcc_external_startup()
5671 routine to your program to override the default if you need to setup hardware
5672 or perform some other critical operation prior to static & global variable
5676 Inline Assembler Code
5679 SDCC allows the use of in-line assembler with a few restriction as regards
5681 All labels defined within inline assembler code
5689 where nnnn is a number less than 100 (which implies a limit of utmost 100
5690 inline assembler labels
5698 It is strongly recommended that each assembly instruction (including labels)
5699 be placed in a separate line (as the example shows).
5704 command line option is used, the inline assembler code will be passed through
5705 the peephole optimizer.
5706 This might cause some unexpected changes in the inline assembler code.
5707 Please go throught the peephole optimizer rules defined in file
5711 carefully before using this option.
5751 The inline assembler code can contain any valid code understood by the assembler
5752 , this includes any assembler directives and comment lines.
5753 The compiler does not do any validation of the code within the
5763 Inline assembler code cannot reference any C-Labels, however it can reference
5764 labels defined by the inline assembler, e.g.:
5790 ; some assembler code
5810 /* some more c code */
5812 clabel:\SpecialChar ~
5814 /* inline assembler cannot reference this label */
5826 $0003: ;label (can be reference by inline assembler only)
5838 /* some more c code */
5846 In other words inline assembly code can access labels defined in inline
5847 assembly within the scope of the funtion.
5851 The same goes the other way, ie.
5852 labels defines in inline assembly CANNOT be accessed by C statements.
5855 int (16 bit) and long (32 bit) Support
5858 For signed & unsigned int (16 bit) and long (32 bit) variables, division,
5859 multiplication and modulus operations are implemented by support routines.
5860 These support routines are all developed in ANSI-C to facilitate porting
5861 to other MCUs, although some model specific assembler optimations are used.
5862 The following files contain the described routine, all of them can be found
5863 in <installdir>/share/sdcc/lib.
5869 <pending: tabularise this>
5875 _mulsint.c - signed 16 bit multiplication (calls _muluint)
5877 _muluint.c - unsigned 16 bit multiplication
5879 _divsint.c - signed 16 bit division (calls _divuint)
5881 _divuint.c - unsigned 16 bit division
5883 _modsint.c - signed 16 bit modulus (call _moduint)
5885 _moduint.c - unsigned 16 bit modulus
5887 _mulslong.c - signed 32 bit multiplication (calls _mululong)
5889 _mululong.c - unsigned32 bit multiplication
5891 _divslong.c - signed 32 division (calls _divulong)
5893 _divulong.c - unsigned 32 division
5895 _modslong.c - signed 32 bit modulus (calls _modulong)
5897 _modulong.c - unsigned 32 bit modulus
5905 Since they are compiled as
5909 , interrupt service routines should not do any of the above operations.
5910 If this is unavoidable then the above routines will need to be compiled
5915 option, after which the source program will have to be compiled with
5922 Floating Point Support
5925 SDCC supports IEEE (single precision 4bytes) floating point numbers.The floating
5926 point support routines are derived from gcc's floatlib.c and consists of
5927 the following routines:
5933 <pending: tabularise this>
5939 _fsadd.c - add floating point numbers
5941 _fssub.c - subtract floating point numbers
5943 _fsdiv.c - divide floating point numbers
5945 _fsmul.c - multiply floating point numbers
5947 _fs2uchar.c - convert floating point to unsigned char
5949 _fs2char.c - convert floating point to signed char
5951 _fs2uint.c - convert floating point to unsigned int
5953 _fs2int.c - convert floating point to signed int
5955 _fs2ulong.c - convert floating point to unsigned long
5957 _fs2long.c - convert floating point to signed long
5959 _uchar2fs.c - convert unsigned char to floating point
5961 _char2fs.c - convert char to floating point number
5963 _uint2fs.c - convert unsigned int to floating point
5965 _int2fs.c - convert int to floating point numbers
5967 _ulong2fs.c - convert unsigned long to floating point number
5969 _long2fs.c - convert long to floating point number
5977 Note if all these routines are used simultaneously the data space might
5979 For serious floating point usage it is strongly recommended that the large
5986 SDCC allows two memory models for MCS51 code, small and large.
5987 Modules compiled with different memory models should
5991 be combined together or the results would be unpredictable.
5992 The library routines supplied with the compiler are compiled as both small
5994 The compiled library modules are contained in seperate directories as small
5995 and large so that you can link to either set.
5999 When the large model is used all variables declared without a storage class
6000 will be allocated into the external ram, this includes all parameters and
6001 local variables (for non-reentrant functions).
6002 When the small model is used variables without storage class are allocated
6003 in the internal ram.
6006 Judicious usage of the processor specific storage classes and the 'reentrant'
6007 function type will yield much more efficient code, than using the large
6009 Several optimizations are disabled when the program is compiled using the
6010 large model, it is therefore strongly recommdended that the small model
6011 be used unless absolutely required.
6017 The only model supported is Flat 24.
6018 This generates code for the 24 bit contiguous addressing mode of the Dallas
6020 In this mode, up to four meg of external RAM or code space can be directly
6022 See the data sheets at www.dalsemi.com for further information on this part.
6026 In older versions of the compiler, this option was used with the MCS51 code
6032 Now, however, the '390 has it's own code generator, selected by the
6041 Note that the compiler does not generate any code to place the processor
6042 into 24 bitmode (although
6046 in the ds390 libraries will do that for you).
6051 , the boot loader or similar code must ensure that the processor is in 24
6052 bit contiguous addressing mode before calling the SDCC startup code.
6060 option, variables will by default be placed into the XDATA segment.
6065 Segments may be placed anywhere in the 4 meg address space using the usual
6067 Note that if any segments are located above 64K, the -r flag must be passed
6068 to the linker to generate the proper segment relocations, and the Intel
6069 HEX output format must be used.
6070 The -r flag can be passed to the linker by using the option
6074 on the sdcc command line.
6075 However, currently the linker can not handle code segments > 64k.
6078 Defines Created by the Compiler
6081 The compiler creates the following #defines.
6084 SDCC - this Symbol is always defined.
6087 SDCC_mcs51 or SDCC_ds390 or SDCC_z80, etc - depending on the model used
6091 __mcs51 or __ds390 or __z80, etc - depending on the model used (e.g.
6095 SDCC_STACK_AUTO - this symbol is defined when
6102 SDCC_MODEL_SMALL - when
6109 SDCC_MODEL_LARGE - when
6116 SDCC_USE_XSTACK - when
6123 SDCC_STACK_TENBIT - when
6130 SDCC_MODEL_FLAT24 - when
6143 SDCC performs a host of standard optimizations in addition to some MCU specific
6146 \layout Subsubsection
6148 Sub-expression Elimination
6151 The compiler does local and global common subexpression elimination, e.g.:
6166 will be translated to
6182 Some subexpressions are not as obvious as the above example, e.g.:
6196 In this case the address arithmetic a->b[i] will be computed only once;
6197 the equivalent code in C would be.
6213 The compiler will try to keep these temporary variables in registers.
6214 \layout Subsubsection
6216 Dead-Code Elimination
6231 i = 1; \SpecialChar ~
6236 global = 1;\SpecialChar ~
6249 global = 3;\SpecialChar ~
6264 int global; void f ()
6277 \layout Subsubsection
6338 Note: the dead stores created by this copy propagation will be eliminated
6339 by dead-code elimination.
6340 \layout Subsubsection
6345 Two types of loop optimizations are done by SDCC loop invariant lifting
6346 and strength reduction of loop induction variables.
6347 In addition to the strength reduction the optimizer marks the induction
6348 variables and the register allocator tries to keep the induction variables
6349 in registers for the duration of the loop.
6350 Because of this preference of the register allocator, loop induction optimizati
6351 on causes an increase in register pressure, which may cause unwanted spilling
6352 of other temporary variables into the stack / data space.
6353 The compiler will generate a warning message when it is forced to allocate
6354 extra space either on the stack or data space.
6355 If this extra space allocation is undesirable then induction optimization
6356 can be eliminated either for the entire source file (with ---noinduction
6357 option) or for a given function only using #pragma\SpecialChar ~
6368 for (i = 0 ; i < 100 ; i ++)
6386 for (i = 0; i < 100; i++)
6396 As mentioned previously some loop invariants are not as apparent, all static
6397 address computations are also moved out of the loop.
6401 Strength Reduction, this optimization substitutes an expression by a cheaper
6408 for (i=0;i < 100; i++)
6428 for (i=0;i< 100;i++) {
6432 ar[itemp1] = itemp2;
6448 The more expensive multiplication is changed to a less expensive addition.
6449 \layout Subsubsection
6454 This optimization is done to reduce the overhead of checking loop boundaries
6455 for every iteration.
6456 Some simple loops can be reversed and implemented using a
6457 \begin_inset Quotes eld
6460 decrement and jump if not zero
6461 \begin_inset Quotes erd
6465 SDCC checks for the following criterion to determine if a loop is reversible
6466 (note: more sophisticated compilers use data-dependency analysis to make
6467 this determination, SDCC uses a more simple minded analysis).
6470 The 'for' loop is of the form
6476 for (<symbol> = <expression> ; <sym> [< | <=] <expression> ; [<sym>++ |
6486 The <for body> does not contain
6487 \begin_inset Quotes eld
6491 \begin_inset Quotes erd
6495 \begin_inset Quotes erd
6501 All goto's are contained within the loop.
6504 No function calls within the loop.
6507 The loop control variable <sym> is not assigned any value within the loop
6510 The loop control variable does NOT participate in any arithmetic operation
6514 There are NO switch statements in the loop.
6515 \layout Subsubsection
6517 Algebraic Simplifications
6520 SDCC does numerous algebraic simplifications, the following is a small sub-set
6521 of these optimizations.
6527 i = j + 0 ; /* changed to */ i = j;
6529 i /= 2; /* changed to */ i >>= 1;
6531 i = j - j ; /* changed to */ i = 0;
6533 i = j / 1 ; /* changed to */ i = j;
6539 Note the subexpressions given above are generally introduced by macro expansions
6540 or as a result of copy/constant propagation.
6541 \layout Subsubsection
6546 SDCC changes switch statements to jump tables when the following conditions
6551 The case labels are in numerical sequence, the labels need not be in order,
6552 and the starting number need not be one or zero.
6558 switch(i) {\SpecialChar ~
6665 Both the above switch statements will be implemented using a jump-table.
6668 The number of case labels is at least three, since it takes two conditional
6669 statements to handle the boundary conditions.
6672 The number of case labels is less than 84, since each label takes 3 bytes
6673 and a jump-table can be utmost 256 bytes long.
6677 Switch statements which have gaps in the numeric sequence or those that
6678 have more that 84 case labels can be split into more than one switch statement
6679 for efficient code generation, e.g.:
6717 If the above switch statement is broken down into two switch statements
6751 case 9: \SpecialChar ~
6761 case 12:\SpecialChar ~
6771 then both the switch statements will be implemented using jump-tables whereas
6772 the unmodified switch statement will not be.
6773 \layout Subsubsection
6775 Bit-shifting Operations.
6778 Bit shifting is one of the most frequently used operation in embedded programmin
6780 SDCC tries to implement bit-shift operations in the most efficient way
6800 generates the following code:
6818 In general SDCC will never setup a loop if the shift count is known.
6858 Note that SDCC stores numbers in little-endian format (i.e.
6859 lowest order first).
6860 \layout Subsubsection
6865 A special case of the bit-shift operation is bit rotation, SDCC recognizes
6866 the following expression to be a left bit-rotation:
6877 i = ((i << 1) | (i >> 7));
6885 will generate the following code:
6901 SDCC uses pattern matching on the parse tree to determine this operation.Variatio
6902 ns of this case will also be recognized as bit-rotation, i.e.:
6908 i = ((i >> 7) | (i << 1)); /* left-bit rotation */
6909 \layout Subsubsection
6914 It is frequently required to obtain the highest order bit of an integral
6915 type (long, int, short or char types).
6916 SDCC recognizes the following expression to yield the highest order bit
6917 and generates optimized code for it, e.g.:
6938 hob = (gint >> 15) & 1;
6951 will generate the following code:
6990 000A E5*01\SpecialChar ~
7018 000C 33\SpecialChar ~
7049 000D E4\SpecialChar ~
7080 000E 13\SpecialChar ~
7111 000F F5*02\SpecialChar ~
7141 Variations of this case however will
7146 It is a standard C expression, so I heartily recommend this be the only
7147 way to get the highest order bit, (it is portable).
7148 Of course it will be recognized even if it is embedded in other expressions,
7155 xyz = gint + ((gint >> 15) & 1);
7161 will still be recognized.
7162 \layout Subsubsection
7167 The compiler uses a rule based, pattern matching and re-writing mechanism
7168 for peep-hole optimization.
7173 a peep-hole optimizer by Christopher W.
7174 Fraser (cwfraser@microsoft.com).
7175 A default set of rules are compiled into the compiler, additional rules
7176 may be added with the
7178 ---peep-file <filename>
7181 The rule language is best illustrated with examples.
7209 The above rule will change the following assembly sequence:
7239 Note: All occurrences of a
7243 (pattern variable) must denote the same string.
7244 With the above rule, the assembly sequence:
7262 will remain unmodified.
7266 Other special case optimizations may be added by the user (via
7272 some variants of the 8051 MCU allow only
7281 The following two rules will change all
7303 replace { lcall %1 } by { acall %1 }
7305 replace { ljmp %1 } by { ajmp %1 }
7313 inline-assembler code
7315 is also passed through the peep hole optimizer, thus the peephole optimizer
7316 can also be used as an assembly level macro expander.
7317 The rules themselves are MCU dependent whereas the rule language infra-structur
7318 e is MCU independent.
7319 Peephole optimization rules for other MCU can be easily programmed using
7324 The syntax for a rule is as follows:
7330 rule := replace [ restart ] '{' <assembly sequence> '
7368 <assembly sequence> '
7386 '}' [if <functionName> ] '
7394 <assembly sequence> := assembly instruction (each instruction including
7395 labels must be on a separate line).
7399 The optimizer will apply to the rules one by one from the top in the sequence
7400 of their appearance, it will terminate when all rules are exhausted.
7401 If the 'restart' option is specified, then the optimizer will start matching
7402 the rules again from the top, this option for a rule is expensive (performance)
7403 , it is intended to be used in situations where a transformation will trigger
7404 the same rule again.
7405 An example of this (not a good one, it has side effects) is the following
7432 Note that the replace pattern cannot be a blank, but can be a comment line.
7433 Without the 'restart' option only the inner most 'pop' 'push' pair would
7434 be eliminated, i.e.:
7486 the restart option the rule will be applied again to the resulting code
7487 and then all the pop-push pairs will be eliminated to yield:
7505 A conditional function can be attached to a rule.
7506 Attaching rules are somewhat more involved, let me illustrate this with
7537 The optimizer does a look-up of a function name table defined in function
7542 in the source file SDCCpeeph.c, with the name
7547 If it finds a corresponding entry the function is called.
7548 Note there can be no parameters specified for these functions, in this
7553 is crucial, since the function
7557 expects to find the label in that particular variable (the hash table containin
7558 g the variable bindings is passed as a parameter).
7559 If you want to code more such functions, take a close look at the function
7560 labelInRange and the calling mechanism in source file SDCCpeeph.c.
7561 I know this whole thing is a little kludgey, but maybe some day we will
7562 have some better means.
7563 If you are looking at this file, you will also see the default rules that
7564 are compiled into the compiler, you can add your own rules in the default
7565 set there if you get tired of specifying the ---peep-file option.
7571 SDCC supports the following #pragma directives.
7574 SAVE - this will save all current options to the SAVE/RESTORE stack.
7578 RESTORE - will restore saved options from the last save.
7579 SAVEs & RESTOREs can be nested.
7580 SDCC uses a SAVE/RESTORE stack: SAVE pushes current options to the stack,
7581 RESTORE pulls current options from the stack.
7585 NOGCSE - will stop global subexpression elimination.
7588 NOINDUCTION - will stop loop induction optimizations.
7591 NOJTBOUND - will not generate code for boundary value checking, when switch
7592 statements are turned into jump-tables.
7595 NOOVERLAY - the compiler will not overlay the parameters and local variables
7599 LESS_PEDANTIC - the compiler will not warn you anymore for obvious mistakes,
7600 you'r on your own now ;-(
7603 NOLOOPREVERSE - Will not do loop reversal optimization
7606 EXCLUDE NONE | {acc[,b[,dpl[,dph]]] - The exclude pragma disables generation
7607 of pair of push/pop instruction in ISR function (using interrupt keyword).
7608 The directive should be placed immediately before the ISR function definition
7609 and it affects ALL ISR functions following it.
7610 To enable the normal register saving for ISR functions use #pragma\SpecialChar ~
7611 EXCLUDE\SpecialChar ~
7615 NOIV - Do not generate interrupt vector table entries for all ISR functions
7616 defined after the pragma.
7617 This is useful in cases where the interrupt vector table must be defined
7618 manually, or when there is a secondary, manually defined interrupt vector
7620 for the autovector feature of the Cypress EZ-USB FX2).
7623 CALLEE-SAVES function1[,function2[,function3...]] - The compiler by default
7624 uses a caller saves convention for register saving across function calls,
7625 however this can cause unneccessary register pushing & popping when calling
7626 small functions from larger functions.
7627 This option can be used to switch off the register saving convention for
7628 the function names specified.
7629 The compiler will not save registers when calling these functions, extra
7630 code need to be manually inserted at the entry & exit for these functions
7631 to save & restore the registers used by these functions, this can SUBSTANTIALLY
7632 reduce code & improve run time performance of the generated code.
7633 In the future the compiler (with interprocedural analysis) may be able
7634 to determine the appropriate scheme to use for each function call.
7635 If ---callee-saves command line option is used, the function names specified
7636 in #pragma\SpecialChar ~
7637 CALLEE-SAVES is appended to the list of functions specified in
7641 The pragma's are intended to be used to turn-off certain optimizations which
7642 might cause the compiler to generate extra stack / data space to store
7643 compiler generated temporary variables.
7644 This usually happens in large functions.
7645 Pragma directives should be used as shown in the following example, they
7646 are used to control options & optimizations for a given function; pragmas
7647 should be placed before and/or after a function, placing pragma's inside
7648 a function body could have unpredictable results.
7654 #pragma SAVE /* save the current settings */
7656 #pragma NOGCSE /* turnoff global subexpression elimination */
7658 #pragma NOINDUCTION /* turn off induction optimizations */
7680 #pragma RESTORE /* turn the optimizations back on */
7686 The compiler will generate a warning message when extra space is allocated.
7687 It is strongly recommended that the SAVE and RESTORE pragma's be used when
7688 changing options for a function.
7693 <pending: this is messy and incomplete>
7698 Compiler support routines (_gptrget, _mulint etc)
7701 Stdclib functions (puts, printf, strcat etc)
7704 Math functions (sin, pow, sqrt etc)
7707 Interfacing with Assembly Routines
7708 \layout Subsubsection
7710 Global Registers used for Parameter Passing
7713 The compiler always uses the global registers
7721 to pass the first parameter to a routine.
7722 The second parameter onwards is either allocated on the stack (for reentrant
7723 routines or if ---stack-auto is used) or in the internal / external ram
7724 (depending on the memory model).
7726 \layout Subsubsection
7728 Assembler Routine(non-reentrant)
7731 In the following example the function cfunc calls an assembler routine asm_func,
7732 which takes two parameters.
7738 extern int asm_func(unsigned char, unsigned char);
7742 int c_func (unsigned char i, unsigned char j)
7750 return asm_func(i,j);
7764 return c_func(10,9);
7772 The corresponding assembler function is:
7778 .globl _asm_func_PARM_2
7842 add a,_asm_func_PARM_2
7878 Note here that the return values are placed in 'dpl' - One byte return value,
7879 'dpl' LSB & 'dph' MSB for two byte values.
7880 'dpl', 'dph' and 'b' for three byte values (generic pointers) and 'dpl','dph','
7881 b' & 'acc' for four byte values.
7884 The parameter naming convention is _<function_name>_PARM_<n>, where n is
7885 the parameter number starting from 1, and counting from the left.
7886 The first parameter is passed in
7887 \begin_inset Quotes eld
7891 \begin_inset Quotes erd
7894 for One bye parameter,
7895 \begin_inset Quotes eld
7899 \begin_inset Quotes erd
7903 \begin_inset Quotes eld
7907 \begin_inset Quotes erd
7911 \begin_inset Quotes eld
7915 \begin_inset Quotes erd
7918 for four bytes, the varible name for the second parameter will be _<function_na
7923 Assemble the assembler routine with the following command:
7930 asx8051 -losg asmfunc.asm
7937 Then compile and link the assembler routine to the C source file with the
7945 sdcc cfunc.c asmfunc.rel
7946 \layout Subsubsection
7948 Assembler Routine(reentrant)
7951 In this case the second parameter onwards will be passed on the stack, the
7952 parameters are pushed from right to left i.e.
7953 after the call the left most parameter will be on the top of the stack.
7960 extern int asm_func(unsigned char, unsigned char);
7964 int c_func (unsigned char i, unsigned char j) reentrant
7972 return asm_func(i,j);
7986 return c_func(10,9);
7994 The corresponding assembler routine is:
8104 The compiling and linking procedure remains the same, however note the extra
8105 entry & exit linkage required for the assembler code, _bp is the stack
8106 frame pointer and is used to compute the offset into the stack for parameters
8107 and local variables.
8113 The external stack is located at the start of the external ram segment,
8114 and is 256 bytes in size.
8115 When ---xstack option is used to compile the program, the parameters and
8116 local variables of all reentrant functions are allocated in this area.
8117 This option is provided for programs with large stack space requirements.
8118 When used with the ---stack-auto option, all parameters and local variables
8119 are allocated on the external stack (note support libraries will need to
8120 be recompiled with the same options).
8123 The compiler outputs the higher order address byte of the external ram segment
8124 into PORT P2, therefore when using the External Stack option, this port
8125 MAY NOT be used by the application program.
8131 Deviations from the compliancy.
8134 functions are not always reentrant.
8137 structures cannot be assigned values directly, cannot be passed as function
8138 parameters or assigned to each other and cannot be a return value from
8165 s1 = s2 ; /* is invalid in SDCC although allowed in ANSI */
8176 struct s foo1 (struct s parms) /* is invalid in SDCC although allowed in
8198 return rets;/* is invalid in SDCC although allowed in ANSI */
8203 'long long' (64 bit integers) not supported.
8206 'double' precision floating point not supported.
8209 No support for setjmp and longjmp (for now).
8212 Old K&R style function declarations are NOT allowed.
8218 foo(i,j) /* this old style of function declarations */
8220 int i,j; /* are valid in ANSI but not valid in SDCC */
8234 functions declared as pointers must be dereferenced during the call.
8245 /* has to be called like this */
8247 (*foo)(); /* ansi standard allows calls to be made like 'foo()' */
8250 Cyclomatic Complexity
8253 Cyclomatic complexity of a function is defined as the number of independent
8254 paths the program can take during execution of the function.
8255 This is an important number since it defines the number test cases you
8256 have to generate to validate the function.
8257 The accepted industry standard for complexity number is 10, if the cyclomatic
8258 complexity reported by SDCC exceeds 10 you should think about simplification
8259 of the function logic.
8260 Note that the complexity level is not related to the number of lines of
8262 Large functions can have low complexity, and small functions can have large
8268 SDCC uses the following formula to compute the complexity:
8273 complexity = (number of edges in control flow graph) - (number of nodes
8274 in control flow graph) + 2;
8278 Having said that the industry standard is 10, you should be aware that in
8279 some cases it be may unavoidable to have a complexity level of less than
8281 For example if you have switch statement with more than 10 case labels,
8282 each case label adds one to the complexity level.
8283 The complexity level is by no means an absolute measure of the algorithmic
8284 complexity of the function, it does however provide a good starting point
8285 for which functions you might look at for further optimization.
8291 Here are a few guidelines that will help the compiler generate more efficient
8292 code, some of the tips are specific to this compiler others are generally
8293 good programming practice.
8296 Use the smallest data type to represent your data-value.
8297 If it is known in advance that the value is going to be less than 256 then
8298 use an 'unsigned char' instead of a 'short' or 'int'.
8301 Use unsigned when it is known in advance that the value is not going to
8303 This helps especially if you are doing division or multiplication.
8306 NEVER jump into a LOOP.
8309 Declare the variables to be local whenever possible, especially loop control
8310 variables (induction).
8313 Since the compiler does not always do implicit integral promotion, the programme
8314 r should do an explicit cast when integral promotion is required.
8317 Reducing the size of division, multiplication & modulus operations can reduce
8318 code size substantially.
8319 Take the following code for example.
8325 foobar(unsigned int p1, unsigned char ch)
8329 unsigned char ch1 = p1 % ch ;
8340 For the modulus operation the variable ch will be promoted to unsigned int
8341 first then the modulus operation will be performed (this will lead to a
8342 call to support routine _moduint()), and the result will be casted to a
8344 If the code is changed to
8350 foobar(unsigned int p1, unsigned char ch)
8354 unsigned char ch1 = (unsigned char)p1 % ch ;
8365 It would substantially reduce the code generated (future versions of the
8366 compiler will be smart enough to detect such optimization oppurtunities).
8369 Notes on MCS51 memory layout
8372 The 8051 family of micro controller have a minimum of 128 bytes of internal
8373 memory which is structured as follows
8377 - Bytes 00-1F - 32 bytes to hold up to 4 banks of the registers R7 to R7
8380 - Bytes 20-2F - 16 bytes to hold 128 bit variables and
8382 - Bytes 30-7F - 60 bytes for general purpose use.
8386 Normally the SDCC compiler will only utilise the first bank of registers,
8387 but it is possible to specify that other banks of registers should be used
8388 in interrupt routines.
8389 By default, the compiler will place the stack after the last bank of used
8391 if the first 2 banks of registers are used, it will position the base of
8392 the internal stack at address 16 (0X10).
8393 This implies that as the stack grows, it will use up the remaining register
8394 banks, and the 16 bytes used by the 128 bit variables, and 60 bytes for
8395 general purpose use.
8398 By default, the compiler uses the 60 general purpose bytes to hold "near
8400 The compiler/optimiser may also declare some Local Variables in this area
8405 If any of the 128 bit variables are used, or near data is being used then
8406 care needs to be taken to ensure that the stack does not grow so much that
8407 it starts to over write either your bit variables or "near data".
8408 There is no runtime checking to prevent this from happening.
8411 The amount of stack being used is affected by the use of the "internal stack"
8412 to save registers before a subroutine call is made (---stack-auto will
8413 declare parameters and local variables on the stack) and the number of
8417 If you detect that the stack is over writing you data, then the following
8419 ---xstack will cause an external stack to be used for saving registers
8420 and (if ---stack-auto is being used) storing parameters and local variables.
8421 However this will produce more code which will be slower to execute.
8425 ---stack-loc will allow you specify the start of the stack, i.e.
8426 you could start it after any data in the general purpose area.
8427 However this may waste the memory not used by the register banks and if
8428 the size of the "near data" increases, it may creep into the bottom of
8432 ---stack-after-data, similar to the ---stack-loc, but it automatically places
8433 the stack after the end of the "near data".
8434 Again this could waste any spare register space.
8437 ---data-loc allows you to specify the start address of the near data.
8438 This could be used to move the "near data" further away from the stack
8439 giving it more room to grow.
8440 This will only work if no bit variables are being used and the stack can
8441 grow to use the bit variable space.
8449 If you find that the stack is over writing your bit variables or "near data"
8450 then the approach which best utilised the internal memory is to position
8451 the "near data" after the last bank of used registers or, if you use bit
8452 variables, after the last bit variable by using the ---data-loc, e.g.
8453 if two register banks are being used and no bit variables, ---data-loc
8454 16, and use the ---stack-after-data option.
8457 If bit variables are being used, another method would be to try and squeeze
8458 the data area in the unused register banks if it will fit, and start the
8459 stack after the last bit variable.
8462 Retargetting for other MCUs.
8465 The issues for retargetting the compiler are far too numerous to be covered
8467 What follows is a brief description of each of the seven phases of the
8468 compiler and its MCU dependency.
8471 Parsing the source and building the annotated parse tree.
8472 This phase is largely MCU independent (except for the language extensions).
8473 Syntax & semantic checks are also done in this phase, along with some initial
8474 optimizations like back patching labels and the pattern matching optimizations
8475 like bit-rotation etc.
8478 The second phase involves generating an intermediate code which can be easy
8479 manipulated during the later phases.
8480 This phase is entirely MCU independent.
8481 The intermediate code generation assumes the target machine has unlimited
8482 number of registers, and designates them with the name iTemp.
8483 The compiler can be made to dump a human readable form of the code generated
8484 by using the ---dumpraw option.
8487 This phase does the bulk of the standard optimizations and is also MCU independe
8489 This phase can be broken down into several sub-phases:
8493 Break down intermediate code (iCode) into basic blocks.
8495 Do control flow & data flow analysis on the basic blocks.
8497 Do local common subexpression elimination, then global subexpression elimination
8499 Dead code elimination
8503 If loop optimizations caused any changes then do 'global subexpression eliminati
8504 on' and 'dead code elimination' again.
8507 This phase determines the live-ranges; by live range I mean those iTemp
8508 variables defined by the compiler that still survive after all the optimization
8510 Live range analysis is essential for register allocation, since these computati
8511 on determines which of these iTemps will be assigned to registers, and for
8515 Phase five is register allocation.
8516 There are two parts to this process.
8520 The first part I call 'register packing' (for lack of a better term).
8521 In this case several MCU specific expression folding is done to reduce
8526 The second part is more MCU independent and deals with allocating registers
8527 to the remaining live ranges.
8528 A lot of MCU specific code does creep into this phase because of the limited
8529 number of index registers available in the 8051.
8532 The Code generation phase is (unhappily), entirely MCU dependent and very
8533 little (if any at all) of this code can be reused for other MCU.
8534 However the scheme for allocating a homogenized assembler operand for each
8535 iCode operand may be reused.
8538 As mentioned in the optimization section the peep-hole optimizer is rule
8539 based system, which can reprogrammed for other MCUs.
8542 SDCDB - Source Level Debugger
8545 SDCC is distributed with a source level debugger.
8546 The debugger uses a command line interface, the command repertoire of the
8547 debugger has been kept as close to gdb (the GNU debugger) as possible.
8548 The configuration and build process is part of the standard compiler installati
8549 on, which also builds and installs the debugger in the target directory
8550 specified during configuration.
8551 The debugger allows you debug BOTH at the C source and at the ASM source
8555 Compiling for Debugging
8560 debug option must be specified for all files for which debug information
8562 The complier generates a .cdb file for each of these files.
8563 The linker updates the .cdb file with the address information.
8564 This .cdb is used by the debugger.
8567 How the Debugger Works
8570 When the ---debug option is specified the compiler generates extra symbol
8571 information some of which are put into the the assembler source and some
8572 are put into the .cdb file, the linker updates the .cdb file with the address
8573 information for the symbols.
8574 The debugger reads the symbolic information generated by the compiler &
8575 the address information generated by the linker.
8576 It uses the SIMULATOR (Daniel's S51) to execute the program, the program
8577 execution is controlled by the debugger.
8578 When a command is issued for the debugger, it translates it into appropriate
8579 commands for the simulator.
8582 Starting the Debugger
8585 The debugger can be started using the following command line.
8586 (Assume the file you are debugging has the file name foo).
8600 The debugger will look for the following files.
8603 foo.c - the source file.
8606 foo.cdb - the debugger symbol information file.
8609 foo.ihx - the intel hex format object file.
8612 Command Line Options.
8615 ---directory=<source file directory> this option can used to specify the
8616 directory search list.
8617 The debugger will look into the directory list specified for source, cdb
8619 The items in the directory list must be separated by ':', e.g.
8620 if the source files can be in the directories /home/src1 and /home/src2,
8621 the ---directory option should be ---directory=/home/src1:/home/src2.
8622 Note there can be no spaces in the option.
8626 -cd <directory> - change to the <directory>.
8629 -fullname - used by GUI front ends.
8632 -cpu <cpu-type> - this argument is passed to the simulator please see the
8633 simulator docs for details.
8636 -X <Clock frequency > this options is passed to the simulator please see
8637 the simulator docs for details.
8640 -s <serial port file> passed to simulator see the simulator docs for details.
8643 -S <serial in,out> passed to simulator see the simulator docs for details.
8649 As mention earlier the command interface for the debugger has been deliberately
8650 kept as close the GNU debugger gdb, as possible.
8651 This will help the integration with existing graphical user interfaces
8652 (like ddd, xxgdb or xemacs) existing for the GNU debugger.
8653 \layout Subsubsection
8655 break [line | file:line | function | file:function]
8658 Set breakpoint at specified line or function:
8667 sdcdb>break foo.c:100
8671 sdcdb>break foo.c:funcfoo
8672 \layout Subsubsection
8674 clear [line | file:line | function | file:function ]
8677 Clear breakpoint at specified line or function:
8686 sdcdb>clear foo.c:100
8690 sdcdb>clear foo.c:funcfoo
8691 \layout Subsubsection
8696 Continue program being debugged, after breakpoint.
8697 \layout Subsubsection
8702 Execute till the end of the current function.
8703 \layout Subsubsection
8708 Delete breakpoint number 'n'.
8709 If used without any option clear ALL user defined break points.
8710 \layout Subsubsection
8712 info [break | stack | frame | registers ]
8715 info break - list all breakpoints
8718 info stack - show the function call stack.
8721 info frame - show information about the current execution frame.
8724 info registers - show content of all registers.
8725 \layout Subsubsection
8730 Step program until it reaches a different source line.
8731 \layout Subsubsection
8736 Step program, proceeding through subroutine calls.
8737 \layout Subsubsection
8742 Start debugged program.
8743 \layout Subsubsection
8748 Print type information of the variable.
8749 \layout Subsubsection
8754 print value of variable.
8755 \layout Subsubsection
8760 load the given file name.
8761 Note this is an alternate method of loading file for debugging.
8762 \layout Subsubsection
8767 print information about current frame.
8768 \layout Subsubsection
8773 Toggle between C source & assembly source.
8774 \layout Subsubsection
8779 Send the string following '!' to the simulator, the simulator response is
8781 Note the debugger does not interpret the command being sent to the simulator,
8782 so if a command like 'go' is sent the debugger can loose its execution
8783 context and may display incorrect values.
8784 \layout Subsubsection
8791 My name is Bobby Brown"
8794 Interfacing with XEmacs.
8797 Two files (in emacs lisp) are provided for the interfacing with XEmacs,
8798 sdcdb.el and sdcdbsrc.el.
8799 These two files can be found in the $(prefix)/bin directory after the installat
8801 These files need to be loaded into XEmacs for the interface to work.
8802 This can be done at XEmacs startup time by inserting the following into
8803 your '.xemacs' file (which can be found in your HOME directory):
8809 (load-file sdcdbsrc.el)
8815 .xemacs is a lisp file so the () around the command is REQUIRED.
8816 The files can also be loaded dynamically while XEmacs is running, set the
8817 environment variable 'EMACSLOADPATH' to the installation bin directory
8818 (<installdir>/bin), then enter the following command ESC-x load-file sdcdbsrc.
8819 To start the interface enter the following command:
8833 You will prompted to enter the file name to be debugged.
8838 The command line options that are passed to the simulator directly are bound
8839 to default values in the file sdcdbsrc.el.
8840 The variables are listed below, these values maybe changed as required.
8843 sdcdbsrc-cpu-type '51
8846 sdcdbsrc-frequency '11059200
8852 The following is a list of key mapping for the debugger interface.
8860 ;; Current Listing ::
8877 binding\SpecialChar ~
8916 ------\SpecialChar ~
8956 sdcdb-next-from-src\SpecialChar ~
8982 sdcdb-back-from-src\SpecialChar ~
9008 sdcdb-cont-from-src\SpecialChar ~
9018 SDCDB continue command
9034 sdcdb-step-from-src\SpecialChar ~
9060 sdcdb-whatis-c-sexp\SpecialChar ~
9070 SDCDB ptypecommand for data at
9134 sdcdbsrc-delete\SpecialChar ~
9148 SDCDB Delete all breakpoints if no arg
9196 given or delete arg (C-u arg x)
9212 sdcdbsrc-frame\SpecialChar ~
9227 SDCDB Display current frame if no arg,
9276 given or display frame arg
9341 sdcdbsrc-goto-sdcdb\SpecialChar ~
9351 Goto the SDCDB output buffer
9367 sdcdb-print-c-sexp\SpecialChar ~
9378 SDCDB print command for data at
9442 sdcdbsrc-goto-sdcdb\SpecialChar ~
9452 Goto the SDCDB output buffer
9468 sdcdbsrc-mode\SpecialChar ~
9484 Toggles Sdcdbsrc mode (turns it off)
9488 ;; C-c C-f\SpecialChar ~
9496 sdcdb-finish-from-src\SpecialChar ~
9504 SDCDB finish command
9508 ;; C-x SPC\SpecialChar ~
9516 sdcdb-break\SpecialChar ~
9534 Set break for line with point
9536 ;; ESC t\SpecialChar ~
9546 sdcdbsrc-mode\SpecialChar ~
9562 Toggle Sdcdbsrc mode
9564 ;; ESC m\SpecialChar ~
9574 sdcdbsrc-srcmode\SpecialChar ~
9598 The Z80 and gbz80 port
9601 SDCC can target both the Zilog Z80 and the Nintendo Gameboy's Z80-like gbz80.
9602 The port is incomplete - long support is incomplete (mul, div and mod are
9603 unimplimented), and both float and bitfield support is missing.
9604 Apart from that the code generated is correct.
9607 As always, the code is the authoritave reference - see z80/ralloc.c and z80/gen.c.
9608 The stack frame is similar to that generated by the IAR Z80 compiler.
9609 IX is used as the base pointer, HL is used as a temporary register, and
9610 BC and DE are available for holding varibles.
9611 IY is currently unusued.
9612 Return values are stored in HL.
9613 One bad side effect of using IX as the base pointer is that a functions
9614 stack frame is limited to 127 bytes - this will be fixed in a later version.
9620 SDCC has grown to be a large project.
9621 The compiler alone (without the preprocessor, assembler and linker) is
9622 about 40,000 lines of code (blank stripped).
9623 The open source nature of this project is a key to its continued growth
9625 You gain the benefit and support of many active software developers and
9627 Is SDCC perfect? No, that's why we need your help.
9628 The developers take pride in fixing reported bugs.
9629 You can help by reporting the bugs and helping other SDCC users.
9630 There are lots of ways to contribute, and we encourage you to take part
9631 in making SDCC a great software package.
9637 Send an email to the mailing list at 'user-sdcc@sdcc.sourceforge.net' or 'devel-sd
9638 cc@sdcc.sourceforge.net'.
9639 Bugs will be fixed ASAP.
9640 When reporting a bug, it is very useful to include a small test program
9641 which reproduces the problem.
9642 If you can isolate the problem by looking at the generated assembly code,
9643 this can be very helpful.
9644 Compiling your program with the ---dumpall option can sometimes be useful
9645 in locating optimization problems.
9651 The anatomy of the compiler
9656 This is an excerpt from an atricle published in Circuit Cellar MagaZine
9658 It's a little outdated (the compiler is much more efficient now and user/devell
9659 oper friendly), but pretty well exposes the guts of it all.
9665 The current version of SDCC can generate code for Intel 8051 and Z80 MCU.
9666 It is fairly easy to retarget for other 8-bit MCU.
9667 Here we take a look at some of the internals of the compiler.
9674 Parsing the input source file and creating an AST (Annotated Syntax Tree).
9675 This phase also involves propagating types (annotating each node of the
9676 parse tree with type information) and semantic analysis.
9677 There are some MCU specific parsing rules.
9678 For example the storage classes, the extended storage classes are MCU specific
9679 while there may be a xdata storage class for 8051 there is no such storage
9680 class for z80 or Atmel AVR.
9681 SDCC allows MCU specific storage class extensions, i.e.
9682 xdata will be treated as a storage class specifier when parsing 8051 C
9683 code but will be treated as a C identifier when parsing z80 or ATMEL AVR
9690 Intermediate code generation.
9691 In this phase the AST is broken down into three-operand form (iCode).
9692 These three operand forms are represented as doubly linked lists.
9693 ICode is the term given to the intermediate form generated by the compiler.
9694 ICode example section shows some examples of iCode generated for some simple
9701 Bulk of the target independent optimizations is performed in this phase.
9702 The optimizations include constant propagation, common sub-expression eliminati
9703 on, loop invariant code movement, strength reduction of loop induction variables
9704 and dead-code elimination.
9710 During intermediate code generation phase, the compiler assumes the target
9711 machine has infinite number of registers and generates a lot of temporary
9713 The live range computation determines the lifetime of each of these compiler-ge
9714 nerated temporaries.
9715 A picture speaks a thousand words.
9716 ICode example sections show the live range annotations for each of the
9718 It is important to note here, each iCode is assigned a number in the order
9719 of its execution in the function.
9720 The live ranges are computed in terms of these numbers.
9721 The from number is the number of the iCode which first defines the operand
9722 and the to number signifies the iCode which uses this operand last.
9728 The register allocation determines the type and number of registers needed
9730 In most MCUs only a few registers can be used for indirect addressing.
9731 In case of 8051 for example the registers R0 & R1 can be used to indirectly
9732 address the internal ram and DPTR to indirectly address the external ram.
9733 The compiler will try to allocate the appropriate register to pointer variables
9735 ICode example section shows the operands annotated with the registers assigned
9737 The compiler will try to keep operands in registers as much as possible;
9738 there are several schemes the compiler uses to do achieve this.
9739 When the compiler runs out of registers the compiler will check to see
9740 if there are any live operands which is not used or defined in the current
9741 basic block being processed, if there are any found then it will push that
9742 operand and use the registers in this block, the operand will then be popped
9743 at the end of the basic block.
9747 There are other MCU specific considerations in this phase.
9748 Some MCUs have an accumulator; very short-lived operands could be assigned
9749 to the accumulator instead of general-purpose register.
9755 Figure II gives a table of iCode operations supported by the compiler.
9756 The code generation involves translating these operations into corresponding
9757 assembly code for the processor.
9758 This sounds overly simple but that is the essence of code generation.
9759 Some of the iCode operations are generated on a MCU specific manner for
9760 example, the z80 port does not use registers to pass parameters so the
9761 SEND and RECV iCode operations will not be generated, and it also does
9762 not support JUMPTABLES.
9769 <Where is Figure II ?>
9775 This section shows some details of iCode.
9776 The example C code does not do anything useful; it is used as an example
9777 to illustrate the intermediate code generated by the compiler.
9790 /* This function does nothing useful.
9797 for the purpose of explaining iCode */
9800 short function (data int *x)
9808 short i=10; /* dead initialization eliminated */
9813 short sum=10; /* dead initialization eliminated */
9826 while (*x) *x++ = *p++;
9840 /* compiler detects i,j to be induction variables */
9844 for (i = 0, j = 10 ; i < 10 ; i++, j---) {
9856 mul += i * 3; /* this multiplication remains */
9862 gint += j * 3;/* this multiplication changed to addition */
9879 In addition to the operands each iCode contains information about the filename
9880 and line it corresponds to in the source file.
9881 The first field in the listing should be interpreted as follows:
9886 Filename(linenumber: iCode Execution sequence number : ICode hash table
9887 key : loop depth of the iCode).
9892 Then follows the human readable form of the ICode operation.
9893 Each operand of this triplet form can be of three basic types a) compiler
9894 generated temporary b) user defined variable c) a constant value.
9895 Note that local variables and parameters are replaced by compiler generated
9897 Live ranges are computed only for temporaries (i.e.
9898 live ranges are not computed for global variables).
9899 Registers are allocated for temporaries only.
9900 Operands are formatted in the following manner:
9905 Operand Name [lr live-from : live-to ] { type information } [ registers
9911 As mentioned earlier the live ranges are computed in terms of the execution
9912 sequence number of the iCodes, for example
9914 the iTemp0 is live from (i.e.
9915 first defined in iCode with execution sequence number 3, and is last used
9916 in the iCode with sequence number 5).
9917 For induction variables such as iTemp21 the live range computation extends
9918 the lifetime from the start to the end of the loop.
9920 The register allocator used the live range information to allocate registers,
9921 the same registers may be used for different temporaries if their live
9922 ranges do not overlap, for example r0 is allocated to both iTemp6 and to
9923 iTemp17 since their live ranges do not overlap.
9924 In addition the allocator also takes into consideration the type and usage
9925 of a temporary, for example itemp6 is a pointer to near space and is used
9926 as to fetch data from (i.e.
9927 used in GET_VALUE_AT_ADDRESS) so it is allocated a pointer registers (r0).
9928 Some short lived temporaries are allocated to special registers which have
9929 meaning to the code generator e.g.
9930 iTemp13 is allocated to a pseudo register CC which tells the back end that
9931 the temporary is used only for a conditional jump the code generation makes
9932 use of this information to optimize a compare and jump ICode.
9934 There are several loop optimizations performed by the compiler.
9935 It can detect induction variables iTemp21(i) and iTemp23(j).
9936 Also note the compiler does selective strength reduction, i.e.
9937 the multiplication of an induction variable in line 18 (gint = j * 3) is
9938 changed to addition, a new temporary iTemp17 is allocated and assigned
9939 a initial value, a constant 3 is then added for each iteration of the loop.
9940 The compiler does not change the multiplication in line 17 however since
9941 the processor does support an 8 * 8 bit multiplication.
9943 Note the dead code elimination optimization eliminated the dead assignments
9944 in line 7 & 8 to I and sum respectively.
9951 Sample.c (5:1:0:0) _entry($9) :
9956 Sample.c(5:2:1:0) proc _function [lr0:0]{function short}
9961 Sample.c(11:3:2:0) iTemp0 [lr3:5]{_near * int}[r2] = recv
9966 Sample.c(11:4:53:0) preHeaderLbl0($11) :
9971 Sample.c(11:5:55:0) iTemp6 [lr5:16]{_near * int}[r0] := iTemp0 [lr3:5]{_near
9977 Sample.c(11:6:5:1) _whilecontinue_0($1) :
9982 Sample.c(11:7:7:1) iTemp4 [lr7:8]{int}[r2 r3] = @[iTemp6 [lr5:16]{_near *
9988 Sample.c(11:8:8:1) if iTemp4 [lr7:8]{int}[r2 r3] == 0 goto _whilebreak_0($3)
9993 Sample.c(11:9:14:1) iTemp7 [lr9:13]{_far * int}[DPTR] := _p [lr0:0]{_far
9999 Sample.c(11:10:15:1) _p [lr0:0]{_far * int} = _p [lr0:0]{_far * int} + 0x2
10005 Sample.c(11:13:18:1) iTemp10 [lr13:14]{int}[r2 r3] = @[iTemp7 [lr9:13]{_far
10011 Sample.c(11:14:19:1) *(iTemp6 [lr5:16]{_near * int}[r0]) := iTemp10 [lr13:14]{int
10017 Sample.c(11:15:12:1) iTemp6 [lr5:16]{_near * int}[r0] = iTemp6 [lr5:16]{_near
10018 * int}[r0] + 0x2 {short}
10023 Sample.c(11:16:20:1) goto _whilecontinue_0($1)
10028 Sample.c(11:17:21:0)_whilebreak_0($3) :
10033 Sample.c(12:18:22:0) iTemp2 [lr18:40]{short}[r2] := 0x0 {short}
10038 Sample.c(13:19:23:0) iTemp11 [lr19:40]{short}[r3] := 0x0 {short}
10043 Sample.c(15:20:54:0)preHeaderLbl1($13) :
10048 Sample.c(15:21:56:0) iTemp21 [lr21:38]{short}[r4] := 0x0 {short}
10053 Sample.c(15:22:57:0) iTemp23 [lr22:38]{int}[r5 r6] := 0xa {int}
10058 Sample.c(15:23:58:0) iTemp17 [lr23:38]{int}[r7 r0] := 0x1e {int}
10063 Sample.c(15:24:26:1)_forcond_0($4) :
10068 Sample.c(15:25:27:1) iTemp13 [lr25:26]{char}[CC] = iTemp21 [lr21:38]{short}[r4]
10074 Sample.c(15:26:28:1) if iTemp13 [lr25:26]{char}[CC] == 0 goto _forbreak_0($7)
10079 Sample.c(16:27:31:1) iTemp2 [lr18:40]{short}[r2] = iTemp2 [lr18:40]{short}[r2]
10080 + ITemp21 [lr21:38]{short}[r4]
10085 Sample.c(17:29:33:1) iTemp15 [lr29:30]{short}[r1] = iTemp21 [lr21:38]{short}[r4]
10091 Sample.c(17:30:34:1) iTemp11 [lr19:40]{short}[r3] = iTemp11 [lr19:40]{short}[r3]
10092 + iTemp15 [lr29:30]{short}[r1]
10097 Sample.c(18:32:36:1:1) iTemp17 [lr23:38]{int}[r7 r0]= iTemp17 [lr23:38]{int}[r7
10103 Sample.c(18:33:37:1) _gint [lr0:0]{int} = _gint [lr0:0]{int} + iTemp17 [lr23:38]{
10109 Sample.c(15:36:42:1) iTemp21 [lr21:38]{short}[r4] = iTemp21 [lr21:38]{short}[r4]
10115 Sample.c(15:37:45:1) iTemp23 [lr22:38]{int}[r5 r6]= iTemp23 [lr22:38]{int}[r5
10121 Sample.c(19:38:47:1) goto _forcond_0($4)
10126 Sample.c(19:39:48:0)_forbreak_0($7) :
10131 Sample.c(20:40:49:0) iTemp24 [lr40:41]{short}[DPTR] = iTemp2 [lr18:40]{short}[r2]
10132 + ITemp11 [lr19:40]{short}[r3]
10137 Sample.c(20:41:50:0) ret iTemp24 [lr40:41]{short}
10142 Sample.c(20:42:51:0)_return($8) :
10147 Sample.c(20:43:52:0) eproc _function [lr0:0]{ ia0 re0 rm0}{function short}
10153 Finally the code generated for this function:
10194 ; ----------------------------------------------
10199 ; function function
10204 ; ----------------------------------------------
10214 ; iTemp0 [lr3:5]{_near * int}[r2] = recv
10226 ; iTemp6 [lr5:16]{_near * int}[r0] := iTemp0 [lr3:5]{_near * int}[r2]
10238 ;_whilecontinue_0($1) :
10248 ; iTemp4 [lr7:8]{int}[r2 r3] = @[iTemp6 [lr5:16]{_near * int}[r0]]
10253 ; if iTemp4 [lr7:8]{int}[r2 r3] == 0 goto _whilebreak_0($3)
10312 ; iTemp7 [lr9:13]{_far * int}[DPTR] := _p [lr0:0]{_far * int}
10331 ; _p [lr0:0]{_far * int} = _p [lr0:0]{_far * int} + 0x2 {short}
10378 ; iTemp10 [lr13:14]{int}[r2 r3] = @[iTemp7 [lr9:13]{_far * int}[DPTR]]
10418 ; *(iTemp6 [lr5:16]{_near * int}[r0]) := iTemp10 [lr13:14]{int}[r2 r3]
10444 ; iTemp6 [lr5:16]{_near * int}[r0] =
10449 ; iTemp6 [lr5:16]{_near * int}[r0] +
10466 ; goto _whilecontinue_0($1)
10478 ; _whilebreak_0($3) :
10488 ; iTemp2 [lr18:40]{short}[r2] := 0x0 {short}
10500 ; iTemp11 [lr19:40]{short}[r3] := 0x0 {short}
10512 ; iTemp21 [lr21:38]{short}[r4] := 0x0 {short}
10524 ; iTemp23 [lr22:38]{int}[r5 r6] := 0xa {int}
10543 ; iTemp17 [lr23:38]{int}[r7 r0] := 0x1e {int}
10572 ; iTemp13 [lr25:26]{char}[CC] = iTemp21 [lr21:38]{short}[r4] < 0xa {short}
10577 ; if iTemp13 [lr25:26]{char}[CC] == 0 goto _forbreak_0($7)
10622 ; iTemp2 [lr18:40]{short}[r2] = iTemp2 [lr18:40]{short}[r2] +
10627 ; iTemp21 [lr21:38]{short}[r4]
10653 ; iTemp15 [lr29:30]{short}[r1] = iTemp21 [lr21:38]{short}[r4] * 0x3 {short}
10686 ; iTemp11 [lr19:40]{short}[r3] = iTemp11 [lr19:40]{short}[r3] +
10691 ; iTemp15 [lr29:30]{short}[r1]
10710 ; iTemp17 [lr23:38]{int}[r7 r0]= iTemp17 [lr23:38]{int}[r7 r0]- 0x3 {short}
10757 ; _gint [lr0:0]{int} = _gint [lr0:0]{int} + iTemp17 [lr23:38]{int}[r7 r0]
10804 ; iTemp21 [lr21:38]{short}[r4] = iTemp21 [lr21:38]{short}[r4] + 0x1 {short}
10816 ; iTemp23 [lr22:38]{int}[r5 r6]= iTemp23 [lr22:38]{int}[r5 r6]- 0x1 {short}
10830 cjne r5,#0xff,00104$
10842 ; goto _forcond_0($4)
10854 ; _forbreak_0($7) :
10864 ; ret iTemp24 [lr40:41]{short}
10907 A few words about basic block successors, predecessors and dominators
10910 Successors are basic blocks that might execute after this basic block.
10912 Predecessors are basic blocks that might execute before reaching this basic
10915 Dominators are basic blocks that WILL execute before reaching this basic
10941 a) succList of [BB2] = [BB4], of [BB3] = [BB4], of [BB1] = [BB2,BB3]
10944 b) predList of [BB2] = [BB1], of [BB3] = [BB1], of [BB4] = [BB2,BB3]
10947 c) domVect of [BB4] = BB1 ...
10948 here we are not sure if BB2 or BB3 was executed but we are SURE that BB1
10956 \begin_inset LatexCommand \url{http://sdcc.sourceforge.net#Who}
10966 Thanks to all the other volunteer developers who have helped with coding,
10967 testing, web-page creation, distribution sets, etc.
10968 You know who you are :-)
10975 This document was initially written by Sandeep Dutta
10978 All product names mentioned herein may be trademarks of their respective
10984 \begin_inset LatexCommand \printindex{}