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 options (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 \htmlurl{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 where include, library and documention files are stored
230 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 \htmlurl{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}
333 Install and search paths
336 Linux (and other gcc-builds like Solaris, Cygwin, Mingw and OSX) by default
337 install in /usr/local.
338 You can override this when configuring with ---prefix-path.
339 Subdirs used will be bin, share/sdcc/include, share/sdcc/lib and share/sdcc/doc.
341 Windows MSVC and Borland builds will install in one single tree (e.g.
342 /sdcc) with subdirs bin, lib, include and doc.
346 The paths searched when running the compiler are as follows (the first catch
350 Binary files (preprocessor, assembler and linker):
352 - the path of argv[0] (if available)
355 \begin_inset Quotes sld
359 \begin_inset Quotes srd
365 \begin_inset Quotes sld
369 \begin_inset Quotes srd
382 \begin_inset Quotes sld
386 \begin_inset Quotes srd
392 \begin_inset Quotes sld
397 - /usr/local/share/sdcc/include (gcc builds)
399 - path(arv[0])/../include and then /sdcc/include (as a last resort for windoze
400 msvc and borland builds)
407 is auto-appended by the compiler, e.g.
408 small, large, z80, ds390 etc.):
413 \begin_inset Quotes sld
417 \begin_inset Quotes srd
427 \begin_inset Quotes sld
431 \begin_inset Quotes srd
440 - /usr/local/share/sdcc/lib/
446 - path(argv[0])/../lib/
454 (as a last resort for windoze msvc and borland builds)
457 Documentation (although never really searched for, you have to do that yourself
461 \begin_inset Quotes sld
465 \begin_inset Quotes srd
470 - /usr/local/share/sdcc/doc (gcc builds)
472 - /sdcc/doc (windoze msvc and borland builds)
475 So, for windoze it is highly recommended to set the environment variable
476 SDCCHOME to prevent needless usage of -I and -L.
479 Linux and other gcc-based systems (cygwin, mingw, osx)
484 Download the source package
486 either from the SDCC CVS repository or from the
487 \begin_inset LatexCommand \url[nightly snapshots]{http://sdcc.sourceforge.net/snap.php}
493 , it will be named something like sdcc
502 Bring up a command line terminal, such as xterm.
507 Unpack the file using a command like:
510 "tar -xzf sdcc.src.tgz
515 , this will create a sub-directory called sdcc with all of the sources.
518 Change directory into the main SDCC directory, for example type:
535 This configures the package for compilation on your system.
551 All of the source packages will compile, this can take a while.
567 This copies the binary executables, the include files, the libraries and
568 the documentation to the install directories.
572 \layout Subsubsection
574 Windows Install Using a Binary Package
577 Download the binary package and unpack it using your favorite unpacking
578 tool (gunzip, WinZip, etc).
579 This should unpack to a group of sub-directories.
580 An example directory structure after unpacking the mingw package is: c:
586 bin for the executables, c:
606 lib for the include and libraries.
609 Adjust your environment variable PATH to include the location of the bin
610 directory or start sdcc using the full path.
611 \layout Subsubsection
613 Windows Install Using Cygwin and Mingw
616 Follow the instruction in
618 Linux and other gcc-based systems
621 \layout Subsubsection
623 Windows Install Using Microsoft Visual C++ 6.0/NET
628 Download the source package
630 either from the SDCC CVS repository or from the
631 \begin_inset LatexCommand \url[nightly snapshots]{http://sdcc.sourceforge.net/snap.php}
637 , it will be named something like sdcc
644 SDCC is distributed with all the projects, workspaces, and files you need
645 to build it using Visual C++ 6.0/NET.
646 The workspace name is 'sdcc.dsw'.
647 Please note that as it is now, all the executables are created in a folder
651 Once built you need to copy the executables from sdcc
655 bin before runnng SDCC.
660 In order to build SDCC with Visual C++ 6.0/NET you need win32 executables
661 of bison.exe, flex.exe, and gawk.exe.
662 One good place to get them is
663 \begin_inset LatexCommand \url[here]{http://unxutils.sourceforge.net}
671 Download the file UnxUtils.zip.
672 Now you have to install the utilities and setup Visual C++ so it can locate
673 the required programs.
674 Here there are two alternatives (choose one!):
681 a) Extract UnxUtils.zip to your C:
683 hard disk PRESERVING the original paths, otherwise bison won't work.
684 (If you are using WinZip make certain that 'Use folder names' is selected)
688 b) In the Visual C++ IDE click Tools, Options, select the Directory tab,
689 in 'Show directories for:' select 'Executable files', and in the directories
690 window add a new path: 'C:
700 (As a side effect, you get a bunch of Unix utilities that could be useful,
701 such as diff and patch.)
708 This one avoids extracting a bunch of files you may not use, but requires
713 a) Create a directory were to put the tools needed, or use a directory already
721 b) Extract 'bison.exe', 'bison.hairy', 'bison.simple', 'flex.exe', and gawk.exe
722 to such directory WITHOUT preserving the original paths.
723 (If you are using WinZip make certain that 'Use folder names' is not selected)
727 c) Rename bison.exe to '_bison.exe'.
731 d) Create a batch file 'bison.bat' in 'C:
735 ' and add these lines:
755 _bison %1 %2 %3 %4 %5 %6 %7 %8 %9
759 Steps 'c' and 'd' are needed because bison requires by default that the
760 files 'bison.simple' and 'bison.hairy' reside in some weird Unix directory,
761 '/usr/local/share/' I think.
762 So it is necessary to tell bison where those files are located if they
763 are not in such directory.
764 That is the function of the environment variables BISON_SIMPLE and BISON_HAIRY.
768 e) In the Visual C++ IDE click Tools, Options, select the Directory tab,
769 in 'Show directories for:' select 'Executable files', and in the directories
770 window add a new path: 'c:
773 Note that you can use any other path instead of 'c:
775 util', even the path where the Visual C++ tools are, probably: 'C:
779 Microsoft Visual Studio
784 So you don't have to execute step 'e' :)
788 Open 'sdcc.dsw' in Visual Studio, click 'build all', when it finishes copy
789 the executables from sdcc
793 bin, and you can compile using sdcc.
794 \layout Subsubsection
796 Windows Install Using Borland ......
804 Testing out the SDCC Compiler
807 The first thing you should do after installing your SDCC compiler is to
815 at the prompt, and the program should run and tell you the version.
816 If it doesn't run, or gives a message about not finding sdcc program, then
817 you need to check over your installation.
818 Make sure that the sdcc bin directory is in your executable search path
819 defined by the PATH environment setting (see the Trouble-shooting section
821 Make sure that the sdcc program is in the bin folder, if not perhaps something
822 did not install correctly.
828 SDCC binaries are commonly installed in a directory arrangement like this:
836 <lyxtabular version="3" rows="3" columns="2">
838 <column alignment="left" valignment="top" leftline="true" width="0(null)">
839 <column alignment="left" valignment="top" leftline="true" rightline="true" width="0(null)">
840 <row topline="true" bottomline="true">
841 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
851 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
858 Holds executables(sdcc, s51, aslink,
867 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
874 usr/local/share/sdcc/lib
877 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
890 <row topline="true" bottomline="true">
891 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
898 usr/local/share/sdcc/include
901 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
908 Holds common C header files
922 Make sure the compiler works on a very simple example.
923 Type in the following test.c program using your favorite editor:
954 Compile this using the following command:
963 If all goes well, the compiler will generate a test.asm and test.rel file.
964 Congratulations, you've just compiled your first program with SDCC.
965 We used the -c option to tell SDCC not to link the generated code, just
966 to keep things simple for this step.
974 The next step is to try it with the linker.
984 If all goes well the compiler will link with the libraries and produce
985 a test.ihx output file.
990 (no test.ihx, and the linker generates warnings), then the problem is most
991 likely that sdcc cannot find the
995 usr/local/share/sdcc/lib directory
999 (see the Install trouble-shooting section for suggestions).
1007 The final test is to ensure sdcc can use the
1011 header files and libraries.
1012 Edit test.c and change it to the following:
1032 strcpy(str1, "testing");
1041 Compile this by typing
1048 This should generate a test.ihx output file, and it should give no warnings
1049 such as not finding the string.h file.
1050 If it cannot find the string.h file, then the problem is that sdcc cannot
1051 find the /usr/local/share/sdcc/include directory
1055 (see the Install trouble-shooting section for suggestions).
1058 Install Trouble-shooting
1059 \layout Subsubsection
1061 SDCC does not build correctly.
1064 A thing to try is starting from scratch by unpacking the .tgz source package
1065 again in an empty directory.
1072 ./configure 2>&1 | tee configure.log
1084 make 2>&1 | tee make.log
1090 If anything goes wrong, you can review the log files to locate the problem.
1091 Or a relevant part of this can be attached to an email that could be helpful
1092 when requesting help from the mailing list.
1093 \layout Subsubsection
1096 \begin_inset Quotes sld
1100 \begin_inset Quotes srd
1107 \begin_inset Quotes sld
1111 \begin_inset Quotes srd
1114 command is a script that analyzes your system and performs some configuration
1115 to ensure the source package compiles on your system.
1116 It will take a few minutes to run, and will compile a few tests to determine
1117 what compiler features are installed.
1118 \layout Subsubsection
1121 \begin_inset Quotes sld
1125 \begin_inset Quotes srd
1131 This runs the GNU make tool, which automatically compiles all the source
1132 packages into the final installed binary executables.
1133 \layout Subsubsection
1136 \begin_inset Quotes sld
1140 \begin_inset Quotes erd
1146 This will install the compiler, other executables and libraries in to the
1147 appropriate system directories.
1148 The default is to copy the executables to /usr/local/bin and the libraries
1149 and header files to /usr/local/share/sdcc/lib and /usr/local/share/sdcc/include.
1150 On most systems you will need super-user privilages to do this.
1153 Advanced Install Options
1157 \begin_inset Quotes eld
1161 \begin_inset Quotes erd
1164 command has several options.
1165 The most commonly used option is ---prefix=<directory name>, where <directory
1166 name> is the final location for the sdcc executables and libraries, (default
1167 location is /usr/local).
1168 The installation process will create the following directory structure
1169 under the <directory name> specified (if they do not already exist).
1174 bin/ - binary exectables (add to PATH environment variable)
1178 bin/share/sdcc/include/ - include header files
1182 bin/share/sdcc/lib/small/ - Object & library files for small model library
1184 bin/share/sdcc/lib/large/ - Object & library files for large model library
1186 bin/share/sdcc/lib/ds390/ - Object & library files for DS80C390 library
1188 bin/share/sdcc/lib/z80/ - Object & library files for Z80 library
1196 \begin_inset Quotes sld
1199 ./configure ---prefix=/usr/local
1200 \begin_inset Quotes erd
1206 will configure the compiler to be installed in directory /usr/local.
1212 SDCC is not just a compiler, but a collection of tools by various developers.
1213 These include linkers, assemblers, simulators and other components.
1214 Here is a summary of some of the components.
1215 Note that the included simulator and assembler have separate documentation
1216 which you can find in the source package in their respective directories.
1217 As SDCC grows to include support for other processors, other packages from
1218 various developers are included and may have their own sets of documentation.
1222 You might want to look at the files which are installed in <installdir>.
1223 At the time of this writing, we find the following programs:
1227 In <installdir>/bin:
1230 sdcc - The compiler.
1233 sdcpp - The C preprocessor.
1236 asx8051 - The assembler for 8051 type processors.
1243 as-gbz80 - The Z80 and GameBoy Z80 assemblers.
1246 aslink -The linker for 8051 type processors.
1253 link-gbz80 - The Z80 and GameBoy Z80 linkers.
1256 s51 - The ucSim 8051 simulator.
1259 sdcdb - The source debugger.
1262 packihx - A tool to pack (compress) Intel hex files.
1265 In <installdir>/share/sdcc/include
1271 In <installdir>/share/sdcc/lib
1274 the sources of the runtime library and the subdirs small large and ds390
1275 with the precompiled relocatables.
1278 In <installdir>/share/sdcc/doc
1284 As development for other processors proceeds, this list will expand to include
1285 executables to support processors like AVR, PIC, etc.
1286 \layout Subsubsection
1291 This is the actual compiler, it in turn uses the c-preprocessor and invokes
1292 the assembler and linkage editor.
1293 \layout Subsubsection
1295 sdcpp - The C-Preprocessor
1298 The preprocessor is a modified version of the GNU preprocessor.
1299 The C preprocessor is used to pull in #include sources, process #ifdef
1300 statements, #defines and so on.
1301 \layout Subsubsection
1303 asx8051, as-z80, as-gbz80, aslink, link-z80, link-gbz80 - The Assemblers
1307 This is retargettable assembler & linkage editor, it was developed by Alan
1309 John Hartman created the version for 8051, and I (Sandeep) have made some
1310 enhancements and bug fixes for it to work properly with the SDCC.
1311 \layout Subsubsection
1316 S51 is a freeware, opensource simulator developed by Daniel Drotos (
1317 \begin_inset LatexCommand \url{mailto:drdani@mazsola.iit.uni-miskolc.hu}
1322 The simulator is built as part of the build process.
1323 For more information visit Daniel's website at:
1324 \begin_inset LatexCommand \url{http://mazsola.iit.uni-miskolc.hu/~drdani/embedded/s51}
1329 It currently support the core mcs51, the Dallas DS80C390 and the Philips
1331 \layout Subsubsection
1333 sdcdb - Source Level Debugger
1339 <todo: is this thing alive?>
1346 Sdcdb is the companion source level debugger.
1347 The current version of the debugger uses Daniel's Simulator S51, but can
1348 be easily changed to use other simulators.
1355 \layout Subsubsection
1357 Single Source File Projects
1360 For single source file 8051 projects the process is very simple.
1361 Compile your programs with the following command
1364 "sdcc sourcefile.c".
1368 This will compile, assemble and link your source file.
1369 Output files are as follows
1373 sourcefile.asm - Assembler source file created by the compiler
1375 sourcefile.lst - Assembler listing file created by the Assembler
1377 sourcefile.rst - Assembler listing file updated with linkedit information,
1378 created by linkage editor
1380 sourcefile.sym - symbol listing for the sourcefile, created by the assembler
1382 sourcefile.rel - Object file created by the assembler, input to Linkage editor
1384 sourcefile.map - The memory map for the load module, created by the Linker
1386 sourcefile.ihx - The load module in Intel hex format (you can select the
1387 Motorola S19 format with ---out-fmt-s19)
1389 sourcefile.cdb - An optional file (with ---debug) containing debug information
1392 \layout Subsubsection
1394 Projects with Multiple Source Files
1397 SDCC can compile only ONE file at a time.
1398 Let us for example assume that you have a project containing the following
1403 foo1.c (contains some functions)
1405 foo2.c (contains some more functions)
1407 foomain.c (contains more functions and the function main)
1415 The first two files will need to be compiled separately with the commands:
1447 Then compile the source file containing the
1451 function and link the files together with the following command:
1459 foomain.c\SpecialChar ~
1460 foo1.rel\SpecialChar ~
1472 can be separately compiled as well:
1483 sdcc foomain.rel foo1.rel foo2.rel
1490 The file containing the
1505 file specified in the command line, since the linkage editor processes
1506 file in the order they are presented to it.
1507 \layout Subsubsection
1509 Projects with Additional Libraries
1512 Some reusable routines may be compiled into a library, see the documentation
1513 for the assembler and linkage editor (which are in <installdir>/share/sdcc/doc)
1519 Libraries created in this manner can be included in the command line.
1520 Make sure you include the -L <library-path> option to tell the linker where
1521 to look for these files if they are not in the current directory.
1522 Here is an example, assuming you have the source file
1534 (if that is not the same as your current project):
1541 sdcc foomain.c foolib.lib -L mylib
1552 must be an absolute path name.
1556 The most efficient way to use libraries is to keep seperate modules in seperate
1558 The lib file now should name all the modules.rel files.
1559 For an example see the standard library file
1563 in the directory <installdir>/share/lib/small.
1566 Command Line Options
1567 \layout Subsubsection
1569 Processor Selection Options
1571 \labelwidthstring 00.00.0000
1577 Generate code for the MCS51 (8051) family of processors.
1578 This is the default processor target.
1580 \labelwidthstring 00.00.0000
1586 Generate code for the DS80C390 processor.
1588 \labelwidthstring 00.00.0000
1594 Generate code for the Z80 family of processors.
1596 \labelwidthstring 00.00.0000
1602 Generate code for the GameBoy Z80 processor.
1604 \labelwidthstring 00.00.0000
1610 Generate code for the Atmel AVR processor (In development, not complete).
1612 \labelwidthstring 00.00.0000
1618 Generate code for the PIC 14-bit processors (In development, not complete).
1620 \labelwidthstring 00.00.0000
1626 Generate code for the Toshiba TLCS-900H processor (In development, not
1629 \labelwidthstring 00.00.0000
1635 Generate code for the Philips XA51 processor (In development, not complete).
1636 \layout Subsubsection
1638 Preprocessor Options
1640 \labelwidthstring 00.00.0000
1646 The additional location where the pre processor will look for <..h> or
1647 \begin_inset Quotes eld
1651 \begin_inset Quotes erd
1656 \labelwidthstring 00.00.0000
1662 Command line definition of macros.
1663 Passed to the pre processor.
1665 \labelwidthstring 00.00.0000
1671 Tell the preprocessor to output a rule suitable for make describing the
1672 dependencies of each object file.
1673 For each source file, the preprocessor outputs one make-rule whose target
1674 is the object file name for that source file and whose dependencies are
1675 all the files `#include'd in it.
1676 This rule may be a single line or may be continued with `
1678 '-newline if it is long.
1679 The list of rules is printed on standard output instead of the preprocessed
1683 \labelwidthstring 00.00.0000
1689 Tell the preprocessor not to discard comments.
1690 Used with the `-E' option.
1692 \labelwidthstring 00.00.0000
1703 Like `-M' but the output mentions only the user header files included with
1705 \begin_inset Quotes eld
1709 System header files included with `#include <file>' are omitted.
1711 \labelwidthstring 00.00.0000
1717 Assert the answer answer for question, in case it is tested with a preprocessor
1718 conditional such as `#if #question(answer)'.
1719 `-A-' disables the standard assertions that normally describe the target
1722 \labelwidthstring 00.00.0000
1728 (answer) Assert the answer answer for question, in case it is tested with
1729 a preprocessor conditional such as `#if #question(answer)'.
1730 `-A-' disables the standard assertions that normally describe the target
1733 \labelwidthstring 00.00.0000
1739 Undefine macro macro.
1740 `-U' options are evaluated after all `-D' options, but before any `-include'
1741 and `-imacros' options.
1743 \labelwidthstring 00.00.0000
1749 Tell the preprocessor to output only a list of the macro definitions that
1750 are in effect at the end of preprocessing.
1751 Used with the `-E' option.
1753 \labelwidthstring 00.00.0000
1759 Tell the preprocessor to pass all macro definitions into the output, in
1760 their proper sequence in the rest of the output.
1762 \labelwidthstring 00.00.0000
1773 Like `-dD' except that the macro arguments and contents are omitted.
1774 Only `#define name' is included in the output.
1775 \layout Subsubsection
1779 \labelwidthstring 00.00.0000
1786 the output path resp.
1787 file where everything will be placed
1789 \labelwidthstring 00.00.0000
1799 <absolute path to additional libraries> This option is passed to the linkage
1800 editor's additional libraries search path.
1801 The path name must be absolute.
1802 Additional library files may be specified in the command line.
1803 See section Compiling programs for more details.
1805 \labelwidthstring 00.00.0000
1811 <Value> The start location of the external ram, default value is 0.
1812 The value entered can be in Hexadecimal or Decimal format, e.g.: ---xram-loc
1813 0x8000 or ---xram-loc 32768.
1815 \labelwidthstring 00.00.0000
1821 <Value> The start location of the code segment, default value 0.
1822 Note when this option is used the interrupt vector table is also relocated
1823 to the given address.
1824 The value entered can be in Hexadecimal or Decimal format, e.g.: ---code-loc
1825 0x8000 or ---code-loc 32768.
1827 \labelwidthstring 00.00.0000
1833 <Value> By default the stack is placed after the data segment.
1834 Using this option the stack can be placed anywhere in the internal memory
1836 The value entered can be in Hexadecimal or Decimal format, e.g.
1837 ---stack-loc 0x20 or ---stack-loc 32.
1838 Since the sp register is incremented before a push or call, the initial
1839 sp will be set to one byte prior the provided value.
1840 The provided value should not overlap any other memory areas such as used
1841 register banks or the data segment and with enough space for the current
1844 \labelwidthstring 00.00.0000
1850 <Value> The start location of the internal ram data segment.
1851 The value entered can be in Hexadecimal or Decimal format, eg.
1852 ---data-loc 0x20 or ---data-loc 32.
1853 (By default, the start location of the internal ram data segment is set
1854 as low as possible in memory, taking into account the used register banks
1855 and the bit segment at address 0x20.
1856 For example if register banks 0 and 1 are used without bit variables, the
1857 data segment will be set, if ---data-loc is not used, to location 0x10.)
1859 \labelwidthstring 00.00.0000
1865 <Value> The start location of the indirectly addressable internal ram, default
1867 The value entered can be in Hexadecimal or Decimal format, eg.
1868 ---idata-loc 0x88 or ---idata-loc 136.
1870 \labelwidthstring 00.00.0000
1879 The linker output (final object code) is in Intel Hex format.
1880 (This is the default option).
1882 \labelwidthstring 00.00.0000
1891 The linker output (final object code) is in Motorola S19 format.
1892 \layout Subsubsection
1896 \labelwidthstring 00.00.0000
1902 Generate code for Large model programs see section Memory Models for more
1904 If this option is used all source files in the project should be compiled
1906 In addition the standard library routines are compiled with small model,
1907 they will need to be recompiled.
1909 \labelwidthstring 00.00.0000
1920 Generate code for Small Model programs see section Memory Models for more
1922 This is the default model.
1923 \layout Subsubsection
1927 \labelwidthstring 00.00.0000
1938 Generate 24-bit flat mode code.
1939 This is the one and only that the ds390 code generator supports right now
1940 and is default when using
1945 See section Memory Models for more details.
1947 \labelwidthstring 00.00.0000
1953 Generate code for the 10 bit stack mode of the Dallas DS80C390 part.
1954 This is the one and only that the ds390 code generator supports right now
1955 and is default when using
1960 In this mode, the stack is located in the lower 1K of the internal RAM,
1961 which is mapped to 0x400000.
1962 Note that the support is incomplete, since it still uses a single byte
1963 as the stack pointer.
1964 This means that only the lower 256 bytes of the potential 1K stack space
1965 will actually be used.
1966 However, this does allow you to reclaim the precious 256 bytes of low RAM
1967 for use for the DATA and IDATA segments.
1968 The compiler will not generate any code to put the processor into 10 bit
1970 It is important to ensure that the processor is in this mode before calling
1971 any re-entrant functions compiled with this option.
1972 In principle, this should work with the
1976 option, but that has not been tested.
1977 It is incompatible with the
1982 It also only makes sense if the processor is in 24 bit contiguous addressing
1985 ---model-flat24 option
1988 \layout Subsubsection
1990 Optimization Options
1992 \labelwidthstring 00.00.0000
1998 Will not do global subexpression elimination, this option may be used when
1999 the compiler creates undesirably large stack/data spaces to store compiler
2001 A warning message will be generated when this happens and the compiler
2002 will indicate the number of extra bytes it allocated.
2003 It recommended that this option NOT be used, #pragma\SpecialChar ~
2005 to turn off global subexpression elimination for a given function only.
2007 \labelwidthstring 00.00.0000
2013 Will not do loop invariant optimizations, this may be turned off for reasons
2014 explained for the previous option.
2015 For more details of loop optimizations performed see section Loop Invariants.It
2016 recommended that this option NOT be used, #pragma\SpecialChar ~
2017 NOINVARIANT can be used
2018 to turn off invariant optimizations for a given function only.
2020 \labelwidthstring 00.00.0000
2026 Will not do loop induction optimizations, see section strength reduction
2027 for more details.It is recommended that this option is NOT used, #pragma\SpecialChar ~
2029 ION can be used to turn off induction optimizations for a given function
2032 \labelwidthstring 00.00.0000
2043 Will not generate boundary condition check when switch statements are implement
2044 ed using jump-tables.
2045 See section Switch Statements for more details.
2046 It is recommended that this option is NOT used, #pragma\SpecialChar ~
2048 used to turn off boundary checking for jump tables for a given function
2051 \labelwidthstring 00.00.0000
2060 Will not do loop reversal optimization.
2062 \labelwidthstring 00.00.0000
2068 This will disable the memcpy of initialized data in far space from code
2070 \layout Subsubsection
2074 \labelwidthstring 00.00.0000
2081 will compile and assemble the source, but will not call the linkage editor.
2083 \labelwidthstring 00.00.0000
2089 Run only the C preprocessor.
2090 Preprocess all the C source files specified and output the results to standard
2093 \labelwidthstring 00.00.0000
2104 All functions in the source file will be compiled as
2109 the parameters and local variables will be allocated on the stack.
2110 see section Parameters and Local Variables for more details.
2111 If this option is used all source files in the project should be compiled
2115 \labelwidthstring 00.00.0000
2121 Uses a pseudo stack in the first 256 bytes in the external ram for allocating
2122 variables and passing parameters.
2123 See section on external stack for more details.
2125 \labelwidthstring 00.00.0000
2129 ---callee-saves function1[,function2][,function3]....
2132 The compiler by default uses a caller saves convention for register saving
2133 across function calls, however this can cause unneccessary register pushing
2134 & popping when calling small functions from larger functions.
2135 This option can be used to switch the register saving convention for the
2136 function names specified.
2137 The compiler will not save registers when calling these functions, no extra
2138 code will be generated at the entry & exit for these functions to save
2139 & restore the registers used by these functions, this can SUBSTANTIALLY
2140 reduce code & improve run time performance of the generated code.
2141 In the future the compiler (with interprocedural analysis) will be able
2142 to determine the appropriate scheme to use for each function call.
2143 DO NOT use this option for built-in functions such as _muluint..., if this
2144 option is used for a library function the appropriate library function
2145 needs to be recompiled with the same option.
2146 If the project consists of multiple source files then all the source file
2147 should be compiled with the same ---callee-saves option string.
2148 Also see #pragma\SpecialChar ~
2151 \labelwidthstring 00.00.0000
2160 When this option is used the compiler will generate debug information, that
2161 can be used with the SDCDB.
2162 The debug information is collected in a file with .cdb extension.
2163 For more information see documentation for SDCDB.
2165 \labelwidthstring 00.00.0000
2171 <filename> This option can be used to use additional rules to be used by
2172 the peep hole optimizer.
2173 See section Peep Hole optimizations for details on how to write these rules.
2175 \labelwidthstring 00.00.0000
2186 Stop after the stage of compilation proper; do not assemble.
2187 The output is an assembler code file for the input file specified.
2189 \labelwidthstring 00.00.0000
2193 -Wa_asmOption[,asmOption]
2196 Pass the asmOption to the assembler.
2198 \labelwidthstring 00.00.0000
2202 -Wl_linkOption[,linkOption]
2205 Pass the linkOption to the linker.
2207 \labelwidthstring 00.00.0000
2216 Integer (16 bit) and long (32 bit) libraries have been compiled as reentrant.
2217 Note by default these libraries are compiled as non-reentrant.
2218 See section Installation for more details.
2220 \labelwidthstring 00.00.0000
2229 This option will cause the compiler to generate an information message for
2230 each function in the source file.
2231 The message contains some
2235 information about the function.
2236 The number of edges and nodes the compiler detected in the control flow
2237 graph of the function, and most importantly the
2239 cyclomatic complexity
2241 see section on Cyclomatic Complexity for more details.
2243 \labelwidthstring 00.00.0000
2252 Floating point library is compiled as reentrant.See section Installation
2255 \labelwidthstring 00.00.0000
2261 The compiler will not overlay parameters and local variables of any function,
2262 see section Parameters and local variables for more details.
2264 \labelwidthstring 00.00.0000
2270 This option can be used when the code generated is called by a monitor
2272 The compiler will generate a 'ret' upon return from the 'main' function.
2273 The default option is to lock up i.e.
2276 \labelwidthstring 00.00.0000
2282 Disable peep-hole optimization.
2284 \labelwidthstring 00.00.0000
2290 Pass the inline assembler code through the peep hole optimizer.
2291 This can cause unexpected changes to inline assembler code, please go through
2292 the peephole optimizer rules defined in the source file tree '<target>/peeph.def
2293 ' before using this option.
2295 \labelwidthstring 00.00.0000
2301 <Value> Causes the linker to check if the internal ram usage is within limits
2304 \labelwidthstring 00.00.0000
2310 <Value> Causes the linker to check if the external ram usage is within limits
2313 \labelwidthstring 00.00.0000
2319 <Value> Causes the linker to check if the code usage is within limits of
2322 \labelwidthstring 00.00.0000
2328 This will prevent the compiler from passing on the default include path
2329 to the preprocessor.
2331 \labelwidthstring 00.00.0000
2337 This will prevent the compiler from passing on the default library path
2340 \labelwidthstring 00.00.0000
2346 Shows the various actions the compiler is performing.
2348 \labelwidthstring 00.00.0000
2354 Shows the actual commands the compiler is executing.
2355 \layout Subsubsection
2357 Intermediate Dump Options
2360 The following options are provided for the purpose of retargetting and debugging
2362 These provided a means to dump the intermediate code (iCode) generated
2363 by the compiler in human readable form at various stages of the compilation
2367 \labelwidthstring 00.00.0000
2373 This option will cause the compiler to dump the intermediate code into
2376 <source filename>.dumpraw
2378 just after the intermediate code has been generated for a function, i.e.
2379 before any optimizations are done.
2380 The basic blocks at this stage ordered in the depth first number, so they
2381 may not be in sequence of execution.
2383 \labelwidthstring 00.00.0000
2389 Will create a dump of iCode's, after global subexpression elimination,
2392 <source filename>.dumpgcse.
2394 \labelwidthstring 00.00.0000
2400 Will create a dump of iCode's, after deadcode elimination, into a file
2403 <source filename>.dumpdeadcode.
2405 \labelwidthstring 00.00.0000
2414 Will create a dump of iCode's, after loop optimizations, into a file named
2417 <source filename>.dumploop.
2419 \labelwidthstring 00.00.0000
2428 Will create a dump of iCode's, after live range analysis, into a file named
2431 <source filename>.dumprange.
2433 \labelwidthstring 00.00.0000
2439 Will dump the life ranges for all symbols.
2441 \labelwidthstring 00.00.0000
2450 Will create a dump of iCode's, after register assignment, into a file named
2453 <source filename>.dumprassgn.
2455 \labelwidthstring 00.00.0000
2461 Will create a dump of the live ranges of iTemp's
2463 \labelwidthstring 00.00.0000
2474 Will cause all the above mentioned dumps to be created.
2477 MCS51/DS390 Storage Class Language Extensions
2480 In addition to the ANSI storage classes SDCC allows the following MCS51
2481 specific storage classes.
2482 \layout Subsubsection
2487 Variables declared with this storage class will be placed in the extern
2493 storage class for Large Memory model, e.g.:
2499 xdata unsigned char xduc;
2500 \layout Subsubsection
2509 storage class for Small Memory model.
2510 Variables declared with this storage class will be allocated in the internal
2518 \layout Subsubsection
2523 Variables declared with this storage class will be allocated into the indirectly
2524 addressable portion of the internal ram of a 8051, e.g.:
2531 \layout Subsubsection
2536 This is a data-type and a storage class specifier.
2537 When a variable is declared as a bit, it is allocated into the bit addressable
2538 memory of 8051, e.g.:
2545 \layout Subsubsection
2550 Like the bit keyword,
2554 signifies both a data-type and storage class, they are used to describe
2555 the special function registers and special bit variables of a 8051, eg:
2561 sfr at 0x80 P0; /* special function register P0 at location 0x80 */
2563 sbit at 0xd7 CY; /* CY (Carry Flag) */
2569 SDCC allows (via language extensions) pointers to explicitly point to any
2570 of the memory spaces of the 8051.
2571 In addition to the explicit pointers, the compiler uses (by default) generic
2572 pointers which can be used to point to any of the memory spaces.
2576 Pointer declaration examples:
2585 /* pointer physically in xternal ram pointing to object in internal ram
2588 data unsigned char * xdata p;
2592 /* pointer physically in code rom pointing to data in xdata space */
2594 xdata unsigned char * code p;
2598 /* pointer physically in code space pointing to data in code space */
2600 code unsigned char * code p;
2604 /* the folowing is a generic pointer physically located in xdata space */
2615 Well you get the idea.
2620 All unqualified pointers are treated as 3-byte (4-byte for the ds390)
2633 The highest order byte of the
2637 pointers contains the data space information.
2638 Assembler support routines are called whenever data is stored or retrieved
2644 These are useful for developing reusable library routines.
2645 Explicitly specifying the pointer type will generate the most efficient
2649 Parameters & Local Variables
2652 Automatic (local) variables and parameters to functions can either be placed
2653 on the stack or in data-space.
2654 The default action of the compiler is to place these variables in the internal
2655 RAM (for small model) or external RAM (for large model).
2656 This in fact makes them
2660 so by default functions are non-reentrant.
2664 They can be placed on the stack either by using the
2668 option or by using the
2672 keyword in the function declaration, e.g.:
2681 unsigned char foo(char i) reentrant
2694 Since stack space on 8051 is limited, the
2702 option should be used sparingly.
2703 Note that the reentrant keyword just means that the parameters & local
2704 variables will be allocated to the stack, it
2708 mean that the function is register bank independent.
2712 Local variables can be assigned storage classes and absolute addresses,
2719 unsigned char foo() {
2725 xdata unsigned char i;
2737 data at 0x31 unsiged char j;
2752 In the above example the variable
2756 will be allocated in the external ram,
2760 in bit addressable space and
2769 or when a function is declared as
2773 this should only be done for static variables.
2776 Parameters however are not allowed any storage class, (storage classes for
2777 parameters will be ignored), their allocation is governed by the memory
2778 model in use, and the reentrancy options.
2784 For non-reentrant functions SDCC will try to reduce internal ram space usage
2785 by overlaying parameters and local variables of a function (if possible).
2786 Parameters and local variables of a function will be allocated to an overlayabl
2787 e segment if the function has
2789 no other function calls and the function is non-reentrant and the memory
2793 If an explicit storage class is specified for a local variable, it will
2797 Note that the compiler (not the linkage editor) makes the decision for overlayin
2799 Functions that are called from an interrupt service routine should be preceded
2800 by a #pragma\SpecialChar ~
2801 NOOVERLAY if they are not reentrant.
2804 Also note that the compiler does not do any processing of inline assembler
2805 code, so the compiler might incorrectly assign local variables and parameters
2806 of a function into the overlay segment if the inline assembler code calls
2807 other c-functions that might use the overlay.
2808 In that case the #pragma\SpecialChar ~
2809 NOOVERLAY should be used.
2812 Parameters and Local variables of functions that contain 16 or 32 bit multiplica
2813 tion or division will NOT be overlayed since these are implemented using
2814 external functions, e.g.:
2824 void set_error(unsigned char errcd)
2840 void some_isr () interrupt 2 using 1
2869 In the above example the parameter
2877 would be assigned to the overlayable segment if the #pragma\SpecialChar ~
2879 not present, this could cause unpredictable runtime behavior when called
2881 The #pragma\SpecialChar ~
2882 NOOVERLAY ensures that the parameters and local variables for
2883 the function are NOT overlayed.
2886 Interrupt Service Routines
2889 SDCC allows interrupt service routines to be coded in C, with some extended
2896 void timer_isr (void) interrupt 2 using 1
2909 The number following the
2913 keyword is the interrupt number this routine will service.
2914 The compiler will insert a call to this routine in the interrupt vector
2915 table for the interrupt number specified.
2920 keyword is used to tell the compiler to use the specified register bank
2921 (8051 specific) when generating code for this function.
2922 Note that when some function is called from an interrupt service routine
2923 it should be preceded by a #pragma\SpecialChar ~
2924 NOOVERLAY if it is not reentrant.
2925 A special note here, int (16 bit) and long (32 bit) integer division, multiplic
2926 ation & modulus operations are implemented using external support routines
2927 developed in ANSI-C, if an interrupt service routine needs to do any of
2928 these operations then the support routines (as mentioned in a following
2929 section) will have to be recompiled using the
2933 option and the source file will need to be compiled using the
2940 If you have multiple source files in your project, interrupt service routines
2941 can be present in any of them, but a prototype of the isr MUST be present
2942 or included in the file that contains the function
2949 Interrupt Numbers and the corresponding address & descriptions for the Standard
2950 8051 are listed below.
2951 SDCC will automatically adjust the interrupt vector table to the maximum
2952 interrupt number specified.
2958 \begin_inset Tabular
2959 <lyxtabular version="3" rows="6" columns="3">
2961 <column alignment="center" valignment="top" leftline="true" width="0(null)">
2962 <column alignment="center" valignment="top" leftline="true" width="0(null)">
2963 <column alignment="center" valignment="top" leftline="true" rightline="true" width="0(null)">
2964 <row topline="true" bottomline="true">
2965 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
2973 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
2981 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
2990 <row topline="true">
2991 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
2999 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
3007 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
3016 <row topline="true">
3017 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
3025 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
3033 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
3042 <row topline="true">
3043 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
3051 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
3059 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
3068 <row topline="true">
3069 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
3077 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
3085 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
3094 <row topline="true" bottomline="true">
3095 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
3103 <cell alignment="center" valignment="top" topline="true" leftline="true" usebox="none">
3111 <cell alignment="center" valignment="top" topline="true" leftline="true" rightline="true" usebox="none">
3128 If the interrupt service routine is defined without
3132 a register bank or with register bank 0 (using 0), the compiler will save
3133 the registers used by itself on the stack upon entry and restore them at
3134 exit, however if such an interrupt service routine calls another function
3135 then the entire register bank will be saved on the stack.
3136 This scheme may be advantageous for small interrupt service routines which
3137 have low register usage.
3140 If the interrupt service routine is defined to be using a specific register
3145 are save and restored, if such an interrupt service routine calls another
3146 function (using another register bank) then the entire register bank of
3147 the called function will be saved on the stack.
3148 This scheme is recommended for larger interrupt service routines.
3151 Calling other functions from an interrupt service routine is not recommended,
3152 avoid it if possible.
3156 Also see the _naked modifier.
3164 <TODO: this isn't implemented at all!>
3170 A special keyword may be associated with a function declaring it as
3175 SDCC will generate code to disable all interrupts upon entry to a critical
3176 function and enable them back before returning.
3177 Note that nesting critical functions may cause unpredictable results.
3202 The critical attribute maybe used with other attributes like
3210 A special keyword may be associated with a function declaring it as
3219 function modifier attribute prevents the compiler from generating prologue
3220 and epilogue code for that function.
3221 This means that the user is entirely responsible for such things as saving
3222 any registers that may need to be preserved, selecting the proper register
3223 bank, generating the
3227 instruction at the end, etc.
3228 Practically, this means that the contents of the function must be written
3229 in inline assembler.
3230 This is particularly useful for interrupt functions, which can have a large
3231 (and often unnecessary) prologue/epilogue.
3232 For example, compare the code generated by these two functions:
3238 data unsigned char counter;
3240 void simpleInterrupt(void) interrupt 1
3254 void nakedInterrupt(void) interrupt 2 _naked
3287 ; MUST explicitly include ret in _naked function.
3301 For an 8051 target, the generated simpleInterrupt looks like:
3446 whereas nakedInterrupt looks like:
3471 ; MUST explicitly include ret(i) in _naked function.
3477 While there is nothing preventing you from writing C code inside a _naked
3478 function, there are many ways to shoot yourself in the foot doing this,
3479 and it is recommended that you stick to inline assembler.
3482 Functions using private banks
3489 attribute (which tells the compiler to use a register bank other than the
3490 default bank zero) should only be applied to
3494 functions (see note 1 below).
3495 This will in most circumstances make the generated ISR code more efficient
3496 since it will not have to save registers on the stack.
3503 attribute will have no effect on the generated code for a
3507 function (but may occasionally be useful anyway
3513 possible exception: if a function is called ONLY from 'interrupt' functions
3514 using a particular bank, it can be declared with the same 'using' attribute
3515 as the calling 'interrupt' functions.
3516 For instance, if you have several ISRs using bank one, and all of them
3517 call memcpy(), it might make sense to create a specialized version of memcpy()
3518 'using 1', since this would prevent the ISR from having to save bank zero
3519 to the stack on entry and switch to bank zero before calling the function
3526 (pending: I don't think this has been done yet)
3533 function using a non-zero bank will assume that it can trash that register
3534 bank, and will not save it.
3535 Since high-priority interrupts can interrupt low-priority ones on the 8051
3536 and friends, this means that if a high-priority ISR
3540 a particular bank occurs while processing a low-priority ISR
3544 the same bank, terrible and bad things can happen.
3545 To prevent this, no single register bank should be
3549 by both a high priority and a low priority ISR.
3550 This is probably most easily done by having all high priority ISRs use
3551 one bank and all low priority ISRs use another.
3552 If you have an ISR which can change priority at runtime, you're on your
3553 own: I suggest using the default bank zero and taking the small performance
3557 It is most efficient if your ISR calls no other functions.
3558 If your ISR must call other functions, it is most efficient if those functions
3559 use the same bank as the ISR (see note 1 below); the next best is if the
3560 called functions use bank zero.
3561 It is very inefficient to call a function using a different, non-zero bank
3569 Data items can be assigned an absolute address with the
3573 keyword, in addition to a storage class, e.g.:
3579 xdata at 0x8000 unsigned char PORTA_8255 ;
3585 In the above example the PORTA_8255 will be allocated to the location 0x8000
3586 of the external ram.
3587 Note that this feature is provided to give the programmer access to
3591 devices attached to the controller.
3592 The compiler does not actually reserve any space for variables declared
3593 in this way (they are implemented with an equate in the assembler).
3594 Thus it is left to the programmer to make sure there are no overlaps with
3595 other variables that are declared without the absolute address.
3596 The assembler listing file (.lst) and the linker output files (.rst) and
3597 (.map) are a good places to look for such overlaps.
3601 Absolute address can be specified for variables in all storage classes,
3614 The above example will allocate the variable at offset 0x02 in the bit-addressab
3616 There is no real advantage to assigning absolute addresses to variables
3617 in this manner, unless you want strict control over all the variables allocated.
3623 The compiler inserts a call to the C routine
3625 _sdcc__external__startup()
3630 at the start of the CODE area.
3631 This routine is in the runtime library.
3632 By default this routine returns 0, if this routine returns a non-zero value,
3633 the static & global variable initialization will be skipped and the function
3634 main will be invoked Other wise static & global variables will be initialized
3635 before the function main is invoked.
3638 _sdcc__external__startup()
3640 routine to your program to override the default if you need to setup hardware
3641 or perform some other critical operation prior to static & global variable
3645 Inline Assembler Code
3648 SDCC allows the use of in-line assembler with a few restriction as regards
3650 All labels defined within inline assembler code
3658 where nnnn is a number less than 100 (which implies a limit of utmost 100
3659 inline assembler labels
3667 It is strongly recommended that each assembly instruction (including labels)
3668 be placed in a separate line (as the example shows).
3673 command line option is used, the inline assembler code will be passed through
3674 the peephole optimizer.
3675 This might cause some unexpected changes in the inline assembler code.
3676 Please go throught the peephole optimizer rules defined in file
3680 carefully before using this option.
3720 The inline assembler code can contain any valid code understood by the assembler
3721 , this includes any assembler directives and comment lines.
3722 The compiler does not do any validation of the code within the
3732 Inline assembler code cannot reference any C-Labels, however it can reference
3733 labels defined by the inline assembler, e.g.:
3759 ; some assembler code
3779 /* some more c code */
3781 clabel:\SpecialChar ~
3783 /* inline assembler cannot reference this label */
3795 $0003: ;label (can be reference by inline assembler only)
3807 /* some more c code */
3815 In other words inline assembly code can access labels defined in inline
3816 assembly within the scope of the funtion.
3820 The same goes the other way, ie.
3821 labels defines in inline assembly CANNOT be accessed by C statements.
3824 int(16 bit) and long (32 bit) Support
3827 For signed & unsigned int (16 bit) and long (32 bit) variables, division,
3828 multiplication and modulus operations are implemented by support routines.
3829 These support routines are all developed in ANSI-C to facilitate porting
3830 to other MCUs, although some model specific assembler optimations are used.
3831 The following files contain the described routine, all of them can be found
3832 in <installdir>/share/sdcc/lib.
3838 <pending: tabularise this>
3844 _mulsint.c - signed 16 bit multiplication (calls _muluint)
3846 _muluint.c - unsigned 16 bit multiplication
3848 _divsint.c - signed 16 bit division (calls _divuint)
3850 _divuint.c - unsigned 16 bit division
3852 _modsint.c - signed 16 bit modulus (call _moduint)
3854 _moduint.c - unsigned 16 bit modulus
3856 _mulslong.c - signed 32 bit multiplication (calls _mululong)
3858 _mululong.c - unsigned32 bit multiplication
3860 _divslong.c - signed 32 division (calls _divulong)
3862 _divulong.c - unsigned 32 division
3864 _modslong.c - signed 32 bit modulus (calls _modulong)
3866 _modulong.c - unsigned 32 bit modulus
3874 Since they are compiled as
3878 , interrupt service routines should not do any of the above operations.
3879 If this is unavoidable then the above routines will need to be compiled
3884 option, after which the source program will have to be compiled with
3891 Floating Point Support
3894 SDCC supports IEEE (single precision 4bytes) floating point numbers.The floating
3895 point support routines are derived from gcc's floatlib.c and consists of
3896 the following routines:
3902 <pending: tabularise this>
3908 _fsadd.c - add floating point numbers
3910 _fssub.c - subtract floating point numbers
3912 _fsdiv.c - divide floating point numbers
3914 _fsmul.c - multiply floating point numbers
3916 _fs2uchar.c - convert floating point to unsigned char
3918 _fs2char.c - convert floating point to signed char
3920 _fs2uint.c - convert floating point to unsigned int
3922 _fs2int.c - convert floating point to signed int
3924 _fs2ulong.c - convert floating point to unsigned long
3926 _fs2long.c - convert floating point to signed long
3928 _uchar2fs.c - convert unsigned char to floating point
3930 _char2fs.c - convert char to floating point number
3932 _uint2fs.c - convert unsigned int to floating point
3934 _int2fs.c - convert int to floating point numbers
3936 _ulong2fs.c - convert unsigned long to floating point number
3938 _long2fs.c - convert long to floating point number
3946 Note if all these routines are used simultaneously the data space might
3948 For serious floating point usage it is strongly recommended that the large
3955 SDCC allows two memory models for MCS51 code, small and large.
3956 Modules compiled with different memory models should
3960 be combined together or the results would be unpredictable.
3961 The library routines supplied with the compiler are compiled as both small
3963 The compiled library modules are contained in seperate directories as small
3964 and large so that you can link to either set.
3968 When the large model is used all variables declared without a storage class
3969 will be allocated into the external ram, this includes all parameters and
3970 local variables (for non-reentrant functions).
3971 When the small model is used variables without storage class are allocated
3972 in the internal ram.
3975 Judicious usage of the processor specific storage classes and the 'reentrant'
3976 function type will yield much more efficient code, than using the large
3978 Several optimizations are disabled when the program is compiled using the
3979 large model, it is therefore strongly recommdended that the small model
3980 be used unless absolutely required.
3986 The only model supported is Flat 24.
3987 This generates code for the 24 bit contiguous addressing mode of the Dallas
3989 In this mode, up to four meg of external RAM or code space can be directly
3991 See the data sheets at www.dalsemi.com for further information on this part.
3995 In older versions of the compiler, this option was used with the MCS51 code
4001 Now, however, the '390 has it's own code generator, selected by the
4010 Note that the compiler does not generate any code to place the processor
4011 into 24 bitmode (although
4015 in the ds390 libraries will do that for you).
4020 , the boot loader or similar code must ensure that the processor is in 24
4021 bit contiguous addressing mode before calling the SDCC startup code.
4029 option, variables will by default be placed into the XDATA segment.
4034 Segments may be placed anywhere in the 4 meg address space using the usual
4036 Note that if any segments are located above 64K, the -r flag must be passed
4037 to the linker to generate the proper segment relocations, and the Intel
4038 HEX output format must be used.
4039 The -r flag can be passed to the linker by using the option
4043 on the sdcc command line.
4044 However, currently the linker can not handle code segments > 64k.
4047 Defines Created by the Compiler
4050 The compiler creates the following #defines.
4053 SDCC - this Symbol is always defined.
4056 SDCC_mcs51 or SDCC_ds390 or SDCC_z80, etc - depending on the model used
4060 __mcs51 or __ds390 or __z80, etc - depending on the model used (e.g.
4064 SDCC_STACK_AUTO - this symbol is defined when
4071 SDCC_MODEL_SMALL - when
4078 SDCC_MODEL_LARGE - when
4085 SDCC_USE_XSTACK - when
4092 SDCC_STACK_TENBIT - when
4099 SDCC_MODEL_FLAT24 - when
4112 SDCC performs a host of standard optimizations in addition to some MCU specific
4115 \layout Subsubsection
4117 Sub-expression Elimination
4120 The compiler does local and global common subexpression elimination, e.g.:
4135 will be translated to
4151 Some subexpressions are not as obvious as the above example, e.g.:
4165 In this case the address arithmetic a->b[i] will be computed only once;
4166 the equivalent code in C would be.
4182 The compiler will try to keep these temporary variables in registers.
4183 \layout Subsubsection
4185 Dead-Code Elimination
4200 i = 1; \SpecialChar ~
4205 global = 1;\SpecialChar ~
4218 global = 3;\SpecialChar ~
4233 int global; void f ()
4246 \layout Subsubsection
4307 Note: the dead stores created by this copy propagation will be eliminated
4308 by dead-code elimination.
4309 \layout Subsubsection
4314 Two types of loop optimizations are done by SDCC loop invariant lifting
4315 and strength reduction of loop induction variables.
4316 In addition to the strength reduction the optimizer marks the induction
4317 variables and the register allocator tries to keep the induction variables
4318 in registers for the duration of the loop.
4319 Because of this preference of the register allocator, loop induction optimizati
4320 on causes an increase in register pressure, which may cause unwanted spilling
4321 of other temporary variables into the stack / data space.
4322 The compiler will generate a warning message when it is forced to allocate
4323 extra space either on the stack or data space.
4324 If this extra space allocation is undesirable then induction optimization
4325 can be eliminated either for the entire source file (with ---noinduction
4326 option) or for a given function only using #pragma\SpecialChar ~
4337 for (i = 0 ; i < 100 ; i ++)
4355 for (i = 0; i < 100; i++)
4365 As mentioned previously some loop invariants are not as apparent, all static
4366 address computations are also moved out of the loop.
4370 Strength Reduction, this optimization substitutes an expression by a cheaper
4377 for (i=0;i < 100; i++)
4397 for (i=0;i< 100;i++) {
4401 ar[itemp1] = itemp2;
4417 The more expensive multiplication is changed to a less expensive addition.
4418 \layout Subsubsection
4423 This optimization is done to reduce the overhead of checking loop boundaries
4424 for every iteration.
4425 Some simple loops can be reversed and implemented using a
4426 \begin_inset Quotes eld
4429 decrement and jump if not zero
4430 \begin_inset Quotes erd
4434 SDCC checks for the following criterion to determine if a loop is reversible
4435 (note: more sophisticated compilers use data-dependency analysis to make
4436 this determination, SDCC uses a more simple minded analysis).
4439 The 'for' loop is of the form
4445 for (<symbol> = <expression> ; <sym> [< | <=] <expression> ; [<sym>++ |
4455 The <for body> does not contain
4456 \begin_inset Quotes eld
4460 \begin_inset Quotes erd
4464 \begin_inset Quotes erd
4470 All goto's are contained within the loop.
4473 No function calls within the loop.
4476 The loop control variable <sym> is not assigned any value within the loop
4479 The loop control variable does NOT participate in any arithmetic operation
4483 There are NO switch statements in the loop.
4484 \layout Subsubsection
4486 Algebraic Simplifications
4489 SDCC does numerous algebraic simplifications, the following is a small sub-set
4490 of these optimizations.
4496 i = j + 0 ; /* changed to */ i = j;
4498 i /= 2; /* changed to */ i >>= 1;
4500 i = j - j ; /* changed to */ i = 0;
4502 i = j / 1 ; /* changed to */ i = j;
4508 Note the subexpressions given above are generally introduced by macro expansions
4509 or as a result of copy/constant propagation.
4510 \layout Subsubsection
4515 SDCC changes switch statements to jump tables when the following conditions
4520 The case labels are in numerical sequence, the labels need not be in order,
4521 and the starting number need not be one or zero.
4527 switch(i) {\SpecialChar ~
4634 Both the above switch statements will be implemented using a jump-table.
4637 The number of case labels is at least three, since it takes two conditional
4638 statements to handle the boundary conditions.
4641 The number of case labels is less than 84, since each label takes 3 bytes
4642 and a jump-table can be utmost 256 bytes long.
4646 Switch statements which have gaps in the numeric sequence or those that
4647 have more that 84 case labels can be split into more than one switch statement
4648 for efficient code generation, e.g.:
4686 If the above switch statement is broken down into two switch statements
4720 case 9: \SpecialChar ~
4730 case 12:\SpecialChar ~
4740 then both the switch statements will be implemented using jump-tables whereas
4741 the unmodified switch statement will not be.
4742 \layout Subsubsection
4744 Bit-shifting Operations.
4747 Bit shifting is one of the most frequently used operation in embedded programmin
4749 SDCC tries to implement bit-shift operations in the most efficient way
4769 generates the following code:
4787 In general SDCC will never setup a loop if the shift count is known.
4827 Note that SDCC stores numbers in little-endian format (i.e.
4828 lowest order first).
4829 \layout Subsubsection
4834 A special case of the bit-shift operation is bit rotation, SDCC recognizes
4835 the following expression to be a left bit-rotation:
4846 i = ((i << 1) | (i >> 7));
4854 will generate the following code:
4870 SDCC uses pattern matching on the parse tree to determine this operation.Variatio
4871 ns of this case will also be recognized as bit-rotation, i.e.:
4877 i = ((i >> 7) | (i << 1)); /* left-bit rotation */
4878 \layout Subsubsection
4883 It is frequently required to obtain the highest order bit of an integral
4884 type (long, int, short or char types).
4885 SDCC recognizes the following expression to yield the highest order bit
4886 and generates optimized code for it, e.g.:
4907 hob = (gint >> 15) & 1;
4920 will generate the following code:
4959 000A E5*01\SpecialChar ~
4987 000C 33\SpecialChar ~
5018 000D E4\SpecialChar ~
5049 000E 13\SpecialChar ~
5080 000F F5*02\SpecialChar ~
5110 Variations of this case however will
5115 It is a standard C expression, so I heartily recommend this be the only
5116 way to get the highest order bit, (it is portable).
5117 Of course it will be recognized even if it is embedded in other expressions,
5124 xyz = gint + ((gint >> 15) & 1);
5130 will still be recognized.
5131 \layout Subsubsection
5136 The compiler uses a rule based, pattern matching and re-writing mechanism
5137 for peep-hole optimization.
5142 a peep-hole optimizer by Christopher W.
5143 Fraser (cwfraser@microsoft.com).
5144 A default set of rules are compiled into the compiler, additional rules
5145 may be added with the
5147 ---peep-file <filename>
5150 The rule language is best illustrated with examples.
5178 The above rule will change the following assembly sequence:
5208 Note: All occurrences of a
5212 (pattern variable) must denote the same string.
5213 With the above rule, the assembly sequence:
5231 will remain unmodified.
5235 Other special case optimizations may be added by the user (via
5241 some variants of the 8051 MCU allow only
5250 The following two rules will change all
5272 replace { lcall %1 } by { acall %1 }
5274 replace { ljmp %1 } by { ajmp %1 }
5282 inline-assembler code
5284 is also passed through the peep hole optimizer, thus the peephole optimizer
5285 can also be used as an assembly level macro expander.
5286 The rules themselves are MCU dependent whereas the rule language infra-structur
5287 e is MCU independent.
5288 Peephole optimization rules for other MCU can be easily programmed using
5293 The syntax for a rule is as follows:
5299 rule := replace [ restart ] '{' <assembly sequence> '
5337 <assembly sequence> '
5355 '}' [if <functionName> ] '
5363 <assembly sequence> := assembly instruction (each instruction including
5364 labels must be on a separate line).
5368 The optimizer will apply to the rules one by one from the top in the sequence
5369 of their appearance, it will terminate when all rules are exhausted.
5370 If the 'restart' option is specified, then the optimizer will start matching
5371 the rules again from the top, this option for a rule is expensive (performance)
5372 , it is intended to be used in situations where a transformation will trigger
5373 the same rule again.
5374 An example of this (not a good one, it has side effects) is the following
5401 Note that the replace pattern cannot be a blank, but can be a comment line.
5402 Without the 'restart' option only the inner most 'pop' 'push' pair would
5403 be eliminated, i.e.:
5455 the restart option the rule will be applied again to the resulting code
5456 and then all the pop-push pairs will be eliminated to yield:
5474 A conditional function can be attached to a rule.
5475 Attaching rules are somewhat more involved, let me illustrate this with
5506 The optimizer does a look-up of a function name table defined in function
5511 in the source file SDCCpeeph.c, with the name
5516 If it finds a corresponding entry the function is called.
5517 Note there can be no parameters specified for these functions, in this
5522 is crucial, since the function
5526 expects to find the label in that particular variable (the hash table containin
5527 g the variable bindings is passed as a parameter).
5528 If you want to code more such functions, take a close look at the function
5529 labelInRange and the calling mechanism in source file SDCCpeeph.c.
5530 I know this whole thing is a little kludgey, but maybe some day we will
5531 have some better means.
5532 If you are looking at this file, you will also see the default rules that
5533 are compiled into the compiler, you can add your own rules in the default
5534 set there if you get tired of specifying the ---peep-file option.
5540 SDCC supports the following #pragma directives.
5541 This directives are applicable only at a function level.
5544 SAVE - this will save all the current options.
5547 RESTORE - will restore the saved options from the last save.
5548 Note that SAVES & RESTOREs cannot be nested.
5549 SDCC uses the same buffer to save the options each time a SAVE is called.
5552 NOGCSE - will stop global subexpression elimination.
5555 NOINDUCTION - will stop loop induction optimizations.
5558 NOJTBOUND - will not generate code for boundary value checking, when switch
5559 statements are turned into jump-tables.
5562 NOOVERLAY - the compiler will not overlay the parameters and local variables
5566 NOLOOPREVERSE - Will not do loop reversal optimization
5569 EXCLUDE NONE | {acc[,b[,dpl[,dph]]] - The exclude pragma disables generation
5570 of pair of push/pop instruction in ISR function (using interrupt keyword).
5571 The directive should be placed immediately before the ISR function definition
5572 and it affects ALL ISR functions following it.
5573 To enable the normal register saving for ISR functions use #pragma\SpecialChar ~
5574 EXCLUDE\SpecialChar ~
5578 NOIV - Do not generate interrupt vector table entries for all ISR functions
5579 defined after the pragma.
5580 This is useful in cases where the interrupt vector table must be defined
5581 manually, or when there is a secondary, manually defined interrupt vector
5583 for the autovector feature of the Cypress EZ-USB FX2).
5586 CALLEE-SAVES function1[,function2[,function3...]] - The compiler by default
5587 uses a caller saves convention for register saving across function calls,
5588 however this can cause unneccessary register pushing & popping when calling
5589 small functions from larger functions.
5590 This option can be used to switch the register saving convention for the
5591 function names specified.
5592 The compiler will not save registers when calling these functions, extra
5593 code will be generated at the entry & exit for these functions to save
5594 & restore the registers used by these functions, this can SUBSTANTIALLY
5595 reduce code & improve run time performance of the generated code.
5596 In future the compiler (with interprocedural analysis) will be able to
5597 determine the appropriate scheme to use for each function call.
5598 If ---callee-saves command line option is used, the function names specified
5599 in #pragma\SpecialChar ~
5600 CALLEE-SAVES is appended to the list of functions specified inthe
5604 The pragma's are intended to be used to turn-off certain optimizations which
5605 might cause the compiler to generate extra stack / data space to store
5606 compiler generated temporary variables.
5607 This usually happens in large functions.
5608 Pragma directives should be used as shown in the following example, they
5609 are used to control options & optimizations for a given function; pragmas
5610 should be placed before and/or after a function, placing pragma's inside
5611 a function body could have unpredictable results.
5617 #pragma SAVE /* save the current settings */
5619 #pragma NOGCSE /* turnoff global subexpression elimination */
5621 #pragma NOINDUCTION /* turn off induction optimizations */
5643 #pragma RESTORE /* turn the optimizations back on */
5649 The compiler will generate a warning message when extra space is allocated.
5650 It is strongly recommended that the SAVE and RESTORE pragma's be used when
5651 changing options for a function.
5656 <pending: this is messy and incomplete>
5661 Compiler support routines (_gptrget, _mulint etc)
5664 Stdclib functions (puts, printf, strcat etc)
5667 Math functions (sin, pow, sqrt etc)
5670 Interfacing with Assembly Routines
5671 \layout Subsubsection
5673 Global Registers used for Parameter Passing
5676 The compiler always uses the global registers
5684 to pass the first parameter to a routine.
5685 The second parameter onwards is either allocated on the stack (for reentrant
5686 routines or if ---stack-auto is used) or in the internal / external ram
5687 (depending on the memory model).
5689 \layout Subsubsection
5691 Assembler Routine(non-reentrant)
5694 In the following example the function cfunc calls an assembler routine asm_func,
5695 which takes two parameters.
5701 extern int asm_func(unsigned char, unsigned char);
5705 int c_func (unsigned char i, unsigned char j)
5713 return asm_func(i,j);
5727 return c_func(10,9);
5735 The corresponding assembler function is:
5741 .globl _asm_func_PARM_2
5805 add a,_asm_func_PARM_2
5841 Note here that the return values are placed in 'dpl' - One byte return value,
5842 'dpl' LSB & 'dph' MSB for two byte values.
5843 'dpl', 'dph' and 'b' for three byte values (generic pointers) and 'dpl','dph','
5844 b' & 'acc' for four byte values.
5847 The parameter naming convention is _<function_name>_PARM_<n>, where n is
5848 the parameter number starting from 1, and counting from the left.
5849 The first parameter is passed in
5850 \begin_inset Quotes eld
5854 \begin_inset Quotes erd
5857 for One bye parameter,
5858 \begin_inset Quotes eld
5862 \begin_inset Quotes erd
5866 \begin_inset Quotes eld
5870 \begin_inset Quotes erd
5874 \begin_inset Quotes eld
5878 \begin_inset Quotes erd
5881 for four bytes, the varible name for the second parameter will be _<function_na
5886 Assemble the assembler routine with the following command:
5893 asx8051 -losg asmfunc.asm
5900 Then compile and link the assembler routine to the C source file with the
5908 sdcc cfunc.c asmfunc.rel
5909 \layout Subsubsection
5911 Assembler Routine(reentrant)
5914 In this case the second parameter onwards will be passed on the stack, the
5915 parameters are pushed from right to left i.e.
5916 after the call the left most parameter will be on the top of the stack.
5923 extern int asm_func(unsigned char, unsigned char);
5927 int c_func (unsigned char i, unsigned char j) reentrant
5935 return asm_func(i,j);
5949 return c_func(10,9);
5957 The corresponding assembler routine is:
6067 The compiling and linking procedure remains the same, however note the extra
6068 entry & exit linkage required for the assembler code, _bp is the stack
6069 frame pointer and is used to compute the offset into the stack for parameters
6070 and local variables.
6076 The external stack is located at the start of the external ram segment,
6077 and is 256 bytes in size.
6078 When ---xstack option is used to compile the program, the parameters and
6079 local variables of all reentrant functions are allocated in this area.
6080 This option is provided for programs with large stack space requirements.
6081 When used with the ---stack-auto option, all parameters and local variables
6082 are allocated on the external stack (note support libraries will need to
6083 be recompiled with the same options).
6086 The compiler outputs the higher order address byte of the external ram segment
6087 into PORT P2, therefore when using the External Stack option, this port
6088 MAY NOT be used by the application program.
6094 Deviations from the compliancy.
6097 functions are not always reentrant.
6100 structures cannot be assigned values directly, cannot be passed as function
6101 parameters or assigned to each other and cannot be a return value from
6128 s1 = s2 ; /* is invalid in SDCC although allowed in ANSI */
6139 struct s foo1 (struct s parms) /* is invalid in SDCC although allowed in
6161 return rets;/* is invalid in SDCC although allowed in ANSI */
6166 'long long' (64 bit integers) not supported.
6169 'double' precision floating point not supported.
6172 No support for setjmp and longjmp (for now).
6175 Old K&R style function declarations are NOT allowed.
6181 foo(i,j) /* this old style of function declarations */
6183 int i,j; /* are valid in ANSI but not valid in SDCC */
6197 functions declared as pointers must be dereferenced during the call.
6208 /* has to be called like this */
6210 (*foo)(); /* ansi standard allows calls to be made like 'foo()' */
6213 Cyclomatic Complexity
6216 Cyclomatic complexity of a function is defined as the number of independent
6217 paths the program can take during execution of the function.
6218 This is an important number since it defines the number test cases you
6219 have to generate to validate the function.
6220 The accepted industry standard for complexity number is 10, if the cyclomatic
6221 complexity reported by SDCC exceeds 10 you should think about simplification
6222 of the function logic.
6223 Note that the complexity level is not related to the number of lines of
6225 Large functions can have low complexity, and small functions can have large
6231 SDCC uses the following formula to compute the complexity:
6236 complexity = (number of edges in control flow graph) - (number of nodes
6237 in control flow graph) + 2;
6241 Having said that the industry standard is 10, you should be aware that in
6242 some cases it be may unavoidable to have a complexity level of less than
6244 For example if you have switch statement with more than 10 case labels,
6245 each case label adds one to the complexity level.
6246 The complexity level is by no means an absolute measure of the algorithmic
6247 complexity of the function, it does however provide a good starting point
6248 for which functions you might look at for further optimization.
6254 Here are a few guidelines that will help the compiler generate more efficient
6255 code, some of the tips are specific to this compiler others are generally
6256 good programming practice.
6259 Use the smallest data type to represent your data-value.
6260 If it is known in advance that the value is going to be less than 256 then
6261 use a 'char' instead of a 'short' or 'int'.
6264 Use unsigned when it is known in advance that the value is not going to
6266 This helps especially if you are doing division or multiplication.
6269 NEVER jump into a LOOP.
6272 Declare the variables to be local whenever possible, especially loop control
6273 variables (induction).
6276 Since the compiler does not do implicit integral promotion, the programmer
6277 should do an explicit cast when integral promotion is required.
6280 Reducing the size of division, multiplication & modulus operations can reduce
6281 code size substantially.
6282 Take the following code for example.
6288 foobar(unsigned int p1, unsigned char ch)
6292 unsigned char ch1 = p1 % ch ;
6303 For the modulus operation the variable ch will be promoted to unsigned int
6304 first then the modulus operation will be performed (this will lead to a
6305 call to support routine _moduint()), and the result will be casted to a
6307 If the code is changed to
6313 foobar(unsigned int p1, unsigned char ch)
6317 unsigned char ch1 = (unsigned char)p1 % ch ;
6328 It would substantially reduce the code generated (future versions of the
6329 compiler will be smart enough to detect such optimization oppurtunities).
6332 Notes on MCS51 memory layout
6335 The 8051 family of micro controller have a minimum of 128 bytes of internal
6336 memory which is structured as follows
6340 - Bytes 00-1F - 32 bytes to hold up to 4 banks of the registers R7 to R7
6343 - Bytes 20-2F - 16 bytes to hold 128 bit variables and
6345 - Bytes 30-7F - 60 bytes for general purpose use.
6349 Normally the SDCC compiler will only utilise the first bank of registers,
6350 but it is possible to specify that other banks of registers should be used
6351 in interrupt routines.
6352 By default, the compiler will place the stack after the last bank of used
6354 if the first 2 banks of registers are used, it will position the base of
6355 the internal stack at address 16 (0X10).
6356 This implies that as the stack grows, it will use up the remaining register
6357 banks, and the 16 bytes used by the 128 bit variables, and 60 bytes for
6358 general purpose use.
6361 By default, the compiler uses the 60 general purpose bytes to hold "near
6363 The compiler/optimiser may also declare some Local Variables in this area
6368 If any of the 128 bit variables are used, or near data is being used then
6369 care needs to be taken to ensure that the stack does not grow so much that
6370 it starts to over write either your bit variables or "near data".
6371 There is no runtime checking to prevent this from happening.
6374 The amount of stack being used is affected by the use of the "internal stack"
6375 to save registers before a subroutine call is made (---stack-auto will
6376 declare parameters and local variables on the stack) and the number of
6380 If you detect that the stack is over writing you data, then the following
6382 ---xstack will cause an external stack to be used for saving registers
6383 and (if ---stack-auto is being used) storing parameters and local variables.
6384 However this will produce more code which will be slower to execute.
6388 ---stack-loc will allow you specify the start of the stack, i.e.
6389 you could start it after any data in the general purpose area.
6390 However this may waste the memory not used by the register banks and if
6391 the size of the "near data" increases, it may creep into the bottom of
6395 ---stack-after-data, similar to the ---stack-loc, but it automatically places
6396 the stack after the end of the "near data".
6397 Again this could waste any spare register space.
6400 ---data-loc allows you to specify the start address of the near data.
6401 This could be used to move the "near data" further away from the stack
6402 giving it more room to grow.
6403 This will only work if no bit variables are being used and the stack can
6404 grow to use the bit variable space.
6412 If you find that the stack is over writing your bit variables or "near data"
6413 then the approach which best utilised the internal memory is to position
6414 the "near data" after the last bank of used registers or, if you use bit
6415 variables, after the last bit variable by using the ---data-loc, e.g.
6416 if two register banks are being used and no bit variables, ---data-loc
6417 16, and use the ---stack-after-data option.
6420 If bit variables are being used, another method would be to try and squeeze
6421 the data area in the unused register banks if it will fit, and start the
6422 stack after the last bit variable.
6425 Retargetting for other MCUs.
6428 The issues for retargetting the compiler are far too numerous to be covered
6430 What follows is a brief description of each of the seven phases of the
6431 compiler and its MCU dependency.
6434 Parsing the source and building the annotated parse tree.
6435 This phase is largely MCU independent (except for the language extensions).
6436 Syntax & semantic checks are also done in this phase, along with some initial
6437 optimizations like back patching labels and the pattern matching optimizations
6438 like bit-rotation etc.
6441 The second phase involves generating an intermediate code which can be easy
6442 manipulated during the later phases.
6443 This phase is entirely MCU independent.
6444 The intermediate code generation assumes the target machine has unlimited
6445 number of registers, and designates them with the name iTemp.
6446 The compiler can be made to dump a human readable form of the code generated
6447 by using the ---dumpraw option.
6450 This phase does the bulk of the standard optimizations and is also MCU independe
6452 This phase can be broken down into several sub-phases:
6456 Break down intermediate code (iCode) into basic blocks.
6458 Do control flow & data flow analysis on the basic blocks.
6460 Do local common subexpression elimination, then global subexpression elimination
6462 Dead code elimination
6466 If loop optimizations caused any changes then do 'global subexpression eliminati
6467 on' and 'dead code elimination' again.
6470 This phase determines the live-ranges; by live range I mean those iTemp
6471 variables defined by the compiler that still survive after all the optimization
6473 Live range analysis is essential for register allocation, since these computati
6474 on determines which of these iTemps will be assigned to registers, and for
6478 Phase five is register allocation.
6479 There are two parts to this process.
6483 The first part I call 'register packing' (for lack of a better term).
6484 In this case several MCU specific expression folding is done to reduce
6489 The second part is more MCU independent and deals with allocating registers
6490 to the remaining live ranges.
6491 A lot of MCU specific code does creep into this phase because of the limited
6492 number of index registers available in the 8051.
6495 The Code generation phase is (unhappily), entirely MCU dependent and very
6496 little (if any at all) of this code can be reused for other MCU.
6497 However the scheme for allocating a homogenized assembler operand for each
6498 iCode operand may be reused.
6501 As mentioned in the optimization section the peep-hole optimizer is rule
6502 based system, which can reprogrammed for other MCUs.
6505 SDCDB - Source Level Debugger
6508 SDCC is distributed with a source level debugger.
6509 The debugger uses a command line interface, the command repertoire of the
6510 debugger has been kept as close to gdb (the GNU debugger) as possible.
6511 The configuration and build process is part of the standard compiler installati
6512 on, which also builds and installs the debugger in the target directory
6513 specified during configuration.
6514 The debugger allows you debug BOTH at the C source and at the ASM source
6518 Compiling for Debugging
6523 debug option must be specified for all files for which debug information
6525 The complier generates a .cdb file for each of these files.
6526 The linker updates the .cdb file with the address information.
6527 This .cdb is used by the debugger.
6530 How the Debugger Works
6533 When the ---debug option is specified the compiler generates extra symbol
6534 information some of which are put into the the assembler source and some
6535 are put into the .cdb file, the linker updates the .cdb file with the address
6536 information for the symbols.
6537 The debugger reads the symbolic information generated by the compiler &
6538 the address information generated by the linker.
6539 It uses the SIMULATOR (Daniel's S51) to execute the program, the program
6540 execution is controlled by the debugger.
6541 When a command is issued for the debugger, it translates it into appropriate
6542 commands for the simulator.
6545 Starting the Debugger
6548 The debugger can be started using the following command line.
6549 (Assume the file you are debugging has the file name foo).
6563 The debugger will look for the following files.
6566 foo.c - the source file.
6569 foo.cdb - the debugger symbol information file.
6572 foo.ihx - the intel hex format object file.
6575 Command Line Options.
6578 ---directory=<source file directory> this option can used to specify the
6579 directory search list.
6580 The debugger will look into the directory list specified for source, cdb
6582 The items in the directory list must be separated by ':', e.g.
6583 if the source files can be in the directories /home/src1 and /home/src2,
6584 the ---directory option should be ---directory=/home/src1:/home/src2.
6585 Note there can be no spaces in the option.
6589 -cd <directory> - change to the <directory>.
6592 -fullname - used by GUI front ends.
6595 -cpu <cpu-type> - this argument is passed to the simulator please see the
6596 simulator docs for details.
6599 -X <Clock frequency > this options is passed to the simulator please see
6600 the simulator docs for details.
6603 -s <serial port file> passed to simulator see the simulator docs for details.
6606 -S <serial in,out> passed to simulator see the simulator docs for details.
6612 As mention earlier the command interface for the debugger has been deliberately
6613 kept as close the GNU debugger gdb, as possible.
6614 This will help the integration with existing graphical user interfaces
6615 (like ddd, xxgdb or xemacs) existing for the GNU debugger.
6616 \layout Subsubsection
6618 break [line | file:line | function | file:function]
6621 Set breakpoint at specified line or function:
6630 sdcdb>break foo.c:100
6634 sdcdb>break foo.c:funcfoo
6635 \layout Subsubsection
6637 clear [line | file:line | function | file:function ]
6640 Clear breakpoint at specified line or function:
6649 sdcdb>clear foo.c:100
6653 sdcdb>clear foo.c:funcfoo
6654 \layout Subsubsection
6659 Continue program being debugged, after breakpoint.
6660 \layout Subsubsection
6665 Execute till the end of the current function.
6666 \layout Subsubsection
6671 Delete breakpoint number 'n'.
6672 If used without any option clear ALL user defined break points.
6673 \layout Subsubsection
6675 info [break | stack | frame | registers ]
6678 info break - list all breakpoints
6681 info stack - show the function call stack.
6684 info frame - show information about the current execution frame.
6687 info registers - show content of all registers.
6688 \layout Subsubsection
6693 Step program until it reaches a different source line.
6694 \layout Subsubsection
6699 Step program, proceeding through subroutine calls.
6700 \layout Subsubsection
6705 Start debugged program.
6706 \layout Subsubsection
6711 Print type information of the variable.
6712 \layout Subsubsection
6717 print value of variable.
6718 \layout Subsubsection
6723 load the given file name.
6724 Note this is an alternate method of loading file for debugging.
6725 \layout Subsubsection
6730 print information about current frame.
6731 \layout Subsubsection
6736 Toggle between C source & assembly source.
6737 \layout Subsubsection
6742 Send the string following '!' to the simulator, the simulator response is
6744 Note the debugger does not interpret the command being sent to the simulator,
6745 so if a command like 'go' is sent the debugger can loose its execution
6746 context and may display incorrect values.
6747 \layout Subsubsection
6754 My name is Bobby Brown"
6757 Interfacing with XEmacs.
6760 Two files (in emacs lisp) are provided for the interfacing with XEmacs,
6761 sdcdb.el and sdcdbsrc.el.
6762 These two files can be found in the $(prefix)/bin directory after the installat
6764 These files need to be loaded into XEmacs for the interface to work.
6765 This can be done at XEmacs startup time by inserting the following into
6766 your '.xemacs' file (which can be found in your HOME directory):
6772 (load-file sdcdbsrc.el)
6778 .xemacs is a lisp file so the () around the command is REQUIRED.
6779 The files can also be loaded dynamically while XEmacs is running, set the
6780 environment variable 'EMACSLOADPATH' to the installation bin directory
6781 (<installdir>/bin), then enter the following command ESC-x load-file sdcdbsrc.
6782 To start the interface enter the following command:
6796 You will prompted to enter the file name to be debugged.
6801 The command line options that are passed to the simulator directly are bound
6802 to default values in the file sdcdbsrc.el.
6803 The variables are listed below, these values maybe changed as required.
6806 sdcdbsrc-cpu-type '51
6809 sdcdbsrc-frequency '11059200
6815 The following is a list of key mapping for the debugger interface.
6823 ;; Current Listing ::
6840 binding\SpecialChar ~
6879 ------\SpecialChar ~
6919 sdcdb-next-from-src\SpecialChar ~
6945 sdcdb-back-from-src\SpecialChar ~
6971 sdcdb-cont-from-src\SpecialChar ~
6981 SDCDB continue command
6997 sdcdb-step-from-src\SpecialChar ~
7023 sdcdb-whatis-c-sexp\SpecialChar ~
7033 SDCDB ptypecommand for data at
7097 sdcdbsrc-delete\SpecialChar ~
7111 SDCDB Delete all breakpoints if no arg
7159 given or delete arg (C-u arg x)
7175 sdcdbsrc-frame\SpecialChar ~
7190 SDCDB Display current frame if no arg,
7239 given or display frame arg
7304 sdcdbsrc-goto-sdcdb\SpecialChar ~
7314 Goto the SDCDB output buffer
7330 sdcdb-print-c-sexp\SpecialChar ~
7341 SDCDB print command for data at
7405 sdcdbsrc-goto-sdcdb\SpecialChar ~
7415 Goto the SDCDB output buffer
7431 sdcdbsrc-mode\SpecialChar ~
7447 Toggles Sdcdbsrc mode (turns it off)
7451 ;; C-c C-f\SpecialChar ~
7459 sdcdb-finish-from-src\SpecialChar ~
7467 SDCDB finish command
7471 ;; C-x SPC\SpecialChar ~
7479 sdcdb-break\SpecialChar ~
7497 Set break for line with point
7499 ;; ESC t\SpecialChar ~
7509 sdcdbsrc-mode\SpecialChar ~
7525 Toggle Sdcdbsrc mode
7527 ;; ESC m\SpecialChar ~
7537 sdcdbsrc-srcmode\SpecialChar ~
7561 The Z80 and gbz80 port
7564 SDCC can target both the Zilog Z80 and the Nintendo Gameboy's Z80-like gbz80.
7565 The port is incomplete - long support is incomplete (mul, div and mod are
7566 unimplimented), and both float and bitfield support is missing.
7567 Apart from that the code generated is correct.
7570 As always, the code is the authoritave reference - see z80/ralloc.c and z80/gen.c.
7571 The stack frame is similar to that generated by the IAR Z80 compiler.
7572 IX is used as the base pointer, HL is used as a temporary register, and
7573 BC and DE are available for holding varibles.
7574 IY is currently unusued.
7575 Return values are stored in HL.
7576 One bad side effect of using IX as the base pointer is that a functions
7577 stack frame is limited to 127 bytes - this will be fixed in a later version.
7583 SDCC has grown to be a large project.
7584 The compiler alone (without the preprocessor, assembler and linker) is
7585 about 40,000 lines of code (blank stripped).
7586 The open source nature of this project is a key to its continued growth
7588 You gain the benefit and support of many active software developers and
7590 Is SDCC perfect? No, that's why we need your help.
7591 The developers take pride in fixing reported bugs.
7592 You can help by reporting the bugs and helping other SDCC users.
7593 There are lots of ways to contribute, and we encourage you to take part
7594 in making SDCC a great software package.
7600 Send an email to the mailing list at 'user-sdcc@sdcc.sourceforge.net' or 'devel-sd
7601 cc@sdcc.sourceforge.net'.
7602 Bugs will be fixed ASAP.
7603 When reporting a bug, it is very useful to include a small test program
7604 which reproduces the problem.
7605 If you can isolate the problem by looking at the generated assembly code,
7606 this can be very helpful.
7607 Compiling your program with the ---dumpall option can sometimes be useful
7608 in locating optimization problems.
7611 The anatomy of the compiler
7616 This is an excerpt from an atricle published in Circuit Cellar MagaZine
7618 It's a little outdated (the compiler is much more efficient now and user/devell
7619 oper friendly), but pretty well exposes the guts of it all.
7625 The current version of SDCC can generate code for Intel 8051 and Z80 MCU.
7626 It is fairly easy to retarget for other 8-bit MCU.
7627 Here we take a look at some of the internals of the compiler.
7634 Parsing the input source file and creating an AST (Annotated Syntax Tree).
7635 This phase also involves propagating types (annotating each node of the
7636 parse tree with type information) and semantic analysis.
7637 There are some MCU specific parsing rules.
7638 For example the storage classes, the extended storage classes are MCU specific
7639 while there may be a xdata storage class for 8051 there is no such storage
7640 class for z80 or Atmel AVR.
7641 SDCC allows MCU specific storage class extensions, i.e.
7642 xdata will be treated as a storage class specifier when parsing 8051 C
7643 code but will be treated as a C identifier when parsing z80 or ATMEL AVR
7650 Intermediate code generation.
7651 In this phase the AST is broken down into three-operand form (iCode).
7652 These three operand forms are represented as doubly linked lists.
7653 ICode is the term given to the intermediate form generated by the compiler.
7654 ICode example section shows some examples of iCode generated for some simple
7661 Bulk of the target independent optimizations is performed in this phase.
7662 The optimizations include constant propagation, common sub-expression eliminati
7663 on, loop invariant code movement, strength reduction of loop induction variables
7664 and dead-code elimination.
7670 During intermediate code generation phase, the compiler assumes the target
7671 machine has infinite number of registers and generates a lot of temporary
7673 The live range computation determines the lifetime of each of these compiler-ge
7674 nerated temporaries.
7675 A picture speaks a thousand words.
7676 ICode example sections show the live range annotations for each of the
7678 It is important to note here, each iCode is assigned a number in the order
7679 of its execution in the function.
7680 The live ranges are computed in terms of these numbers.
7681 The from number is the number of the iCode which first defines the operand
7682 and the to number signifies the iCode which uses this operand last.
7688 The register allocation determines the type and number of registers needed
7690 In most MCUs only a few registers can be used for indirect addressing.
7691 In case of 8051 for example the registers R0 & R1 can be used to indirectly
7692 address the internal ram and DPTR to indirectly address the external ram.
7693 The compiler will try to allocate the appropriate register to pointer variables
7695 ICode example section shows the operands annotated with the registers assigned
7697 The compiler will try to keep operands in registers as much as possible;
7698 there are several schemes the compiler uses to do achieve this.
7699 When the compiler runs out of registers the compiler will check to see
7700 if there are any live operands which is not used or defined in the current
7701 basic block being processed, if there are any found then it will push that
7702 operand and use the registers in this block, the operand will then be popped
7703 at the end of the basic block.
7707 There are other MCU specific considerations in this phase.
7708 Some MCUs have an accumulator; very short-lived operands could be assigned
7709 to the accumulator instead of general-purpose register.
7715 Figure II gives a table of iCode operations supported by the compiler.
7716 The code generation involves translating these operations into corresponding
7717 assembly code for the processor.
7718 This sounds overly simple but that is the essence of code generation.
7719 Some of the iCode operations are generated on a MCU specific manner for
7720 example, the z80 port does not use registers to pass parameters so the
7721 SEND and RECV iCode operations will not be generated, and it also does
7722 not support JUMPTABLES.
7729 <Where is Figure II ?>
7735 This section shows some details of iCode.
7736 The example C code does not do anything useful; it is used as an example
7737 to illustrate the intermediate code generated by the compiler.
7750 /* This function does nothing useful.
7757 for the purpose of explaining iCode */
7760 short function (data int *x)
7768 short i=10; /* dead initialization eliminated */
7773 short sum=10; /* dead initialization eliminated */
7786 while (*x) *x++ = *p++;
7800 /* compiler detects i,j to be induction variables */
7804 for (i = 0, j = 10 ; i < 10 ; i++, j---) {
7816 mul += i * 3; /* this multiplication remains */
7822 gint += j * 3;/* this multiplication changed to addition */
7839 In addition to the operands each iCode contains information about the filename
7840 and line it corresponds to in the source file.
7841 The first field in the listing should be interpreted as follows:
7846 Filename(linenumber: iCode Execution sequence number : ICode hash table
7847 key : loop depth of the iCode).
7852 Then follows the human readable form of the ICode operation.
7853 Each operand of this triplet form can be of three basic types a) compiler
7854 generated temporary b) user defined variable c) a constant value.
7855 Note that local variables and parameters are replaced by compiler generated
7857 Live ranges are computed only for temporaries (i.e.
7858 live ranges are not computed for global variables).
7859 Registers are allocated for temporaries only.
7860 Operands are formatted in the following manner:
7865 Operand Name [lr live-from : live-to ] { type information } [ registers
7871 As mentioned earlier the live ranges are computed in terms of the execution
7872 sequence number of the iCodes, for example
7874 the iTemp0 is live from (i.e.
7875 first defined in iCode with execution sequence number 3, and is last used
7876 in the iCode with sequence number 5).
7877 For induction variables such as iTemp21 the live range computation extends
7878 the lifetime from the start to the end of the loop.
7880 The register allocator used the live range information to allocate registers,
7881 the same registers may be used for different temporaries if their live
7882 ranges do not overlap, for example r0 is allocated to both iTemp6 and to
7883 iTemp17 since their live ranges do not overlap.
7884 In addition the allocator also takes into consideration the type and usage
7885 of a temporary, for example itemp6 is a pointer to near space and is used
7886 as to fetch data from (i.e.
7887 used in GET_VALUE_AT_ADDRESS) so it is allocated a pointer registers (r0).
7888 Some short lived temporaries are allocated to special registers which have
7889 meaning to the code generator e.g.
7890 iTemp13 is allocated to a pseudo register CC which tells the back end that
7891 the temporary is used only for a conditional jump the code generation makes
7892 use of this information to optimize a compare and jump ICode.
7894 There are several loop optimizations performed by the compiler.
7895 It can detect induction variables iTemp21(i) and iTemp23(j).
7896 Also note the compiler does selective strength reduction, i.e.
7897 the multiplication of an induction variable in line 18 (gint = j * 3) is
7898 changed to addition, a new temporary iTemp17 is allocated and assigned
7899 a initial value, a constant 3 is then added for each iteration of the loop.
7900 The compiler does not change the multiplication in line 17 however since
7901 the processor does support an 8 * 8 bit multiplication.
7903 Note the dead code elimination optimization eliminated the dead assignments
7904 in line 7 & 8 to I and sum respectively.
7911 Sample.c (5:1:0:0) _entry($9) :
7916 Sample.c(5:2:1:0) proc _function [lr0:0]{function short}
7921 Sample.c(11:3:2:0) iTemp0 [lr3:5]{_near * int}[r2] = recv
7926 Sample.c(11:4:53:0) preHeaderLbl0($11) :
7931 Sample.c(11:5:55:0) iTemp6 [lr5:16]{_near * int}[r0] := iTemp0 [lr3:5]{_near
7937 Sample.c(11:6:5:1) _whilecontinue_0($1) :
7942 Sample.c(11:7:7:1) iTemp4 [lr7:8]{int}[r2 r3] = @[iTemp6 [lr5:16]{_near *
7948 Sample.c(11:8:8:1) if iTemp4 [lr7:8]{int}[r2 r3] == 0 goto _whilebreak_0($3)
7953 Sample.c(11:9:14:1) iTemp7 [lr9:13]{_far * int}[DPTR] := _p [lr0:0]{_far
7959 Sample.c(11:10:15:1) _p [lr0:0]{_far * int} = _p [lr0:0]{_far * int} + 0x2
7965 Sample.c(11:13:18:1) iTemp10 [lr13:14]{int}[r2 r3] = @[iTemp7 [lr9:13]{_far
7971 Sample.c(11:14:19:1) *(iTemp6 [lr5:16]{_near * int}[r0]) := iTemp10 [lr13:14]{int
7977 Sample.c(11:15:12:1) iTemp6 [lr5:16]{_near * int}[r0] = iTemp6 [lr5:16]{_near
7988 Sample.c(11:16:20:1) goto _whilecontinue_0($1)
7993 Sample.c(11:17:21:0)_whilebreak_0($3) :
7998 Sample.c(12:18:22:0) iTemp2 [lr18:40]{short}[r2] := 0x0 {short}
8003 Sample.c(13:19:23:0) iTemp11 [lr19:40]{short}[r3] := 0x0 {short}
8008 Sample.c(15:20:54:0)preHeaderLbl1($13) :
8013 Sample.c(15:21:56:0) iTemp21 [lr21:38]{short}[r4] := 0x0 {short}
8018 Sample.c(15:22:57:0) iTemp23 [lr22:38]{int}[r5 r6] := 0xa {int}
8023 Sample.c(15:23:58:0) iTemp17 [lr23:38]{int}[r7 r0] := 0x1e {int}
8028 Sample.c(15:24:26:1)_forcond_0($4) :
8033 Sample.c(15:25:27:1) iTemp13 [lr25:26]{char}[CC] = iTemp21 [lr21:38]{short}[r4]
8039 Sample.c(15:26:28:1) if iTemp13 [lr25:26]{char}[CC] == 0 goto _forbreak_0($7)
8044 Sample.c(16:27:31:1) iTemp2 [lr18:40]{short}[r2] = iTemp2 [lr18:40]{short}[r2]
8050 ITemp21 [lr21:38]{short}[r4]
8055 Sample.c(17:29:33:1) iTemp15 [lr29:30]{short}[r1] = iTemp21 [lr21:38]{short}[r4]
8061 Sample.c(17:30:34:1) iTemp11 [lr19:40]{short}[r3] = iTemp11 [lr19:40]{short}[r3]
8067 iTemp15 [lr29:30]{short}[r1]
8072 Sample.c(18:32:36:1:1) iTemp17 [lr23:38]{int}[r7 r0]= iTemp17 [lr23:38]{int}[r7
8078 Sample.c(18:33:37:1) _gint [lr0:0]{int} = _gint [lr0:0]{int} + iTemp17 [lr23:38]{
8084 Sample.c(15:36:42:1) iTemp21 [lr21:38]{short}[r4] = iTemp21 [lr21:38]{short}[r4]
8090 Sample.c(15:37:45:1) iTemp23 [lr22:38]{int}[r5 r6]= iTemp23 [lr22:38]{int}[r5
8096 Sample.c(19:38:47:1) goto _forcond_0($4)
8101 Sample.c(19:39:48:0)_forbreak_0($7) :
8106 Sample.c(20:40:49:0) iTemp24 [lr40:41]{short}[DPTR] = iTemp2 [lr18:40]{short}[r2]
8112 ITemp11 [lr19:40]{short}[r3]
8117 sample.c(20:41:50:0) ret iTemp24 [lr40:41]{short}
8122 sample.c(20:42:51:0)_return($8) :
8127 sample.c(20:43:52:0) eproc _function [lr0:0]{ ia0 re0 rm0}{function short}
8133 Finally the code generated for this function:
8174 ; ----------------------------------------------
8184 ; ----------------------------------------------
8194 ; iTemp0 [lr3:5]{_near * int}[r2] = recv
8206 ; iTemp6 [lr5:16]{_near * int}[r0] := iTemp0 [lr3:5]{_near * int}[r2]
8218 ;_whilecontinue_0($1) :
8228 ; iTemp4 [lr7:8]{int}[r2 r3] = @[iTemp6 [lr5:16]{_near * int}[r0]]
8233 ; if iTemp4 [lr7:8]{int}[r2 r3] == 0 goto _whilebreak_0($3)
8292 ; iTemp7 [lr9:13]{_far * int}[DPTR] := _p [lr0:0]{_far * int}
8311 ; _p [lr0:0]{_far * int} = _p [lr0:0]{_far * int} + 0x2 {short}
8358 ; iTemp10 [lr13:14]{int}[r2 r3] = @[iTemp7 [lr9:13]{_far * int}[DPTR]]
8398 ; *(iTemp6 [lr5:16]{_near * int}[r0]) := iTemp10 [lr13:14]{int}[r2 r3]
8424 ; iTemp6 [lr5:16]{_near * int}[r0] =
8429 ; iTemp6 [lr5:16]{_near * int}[r0] +
8446 ; goto _whilecontinue_0($1)
8458 ; _whilebreak_0($3) :
8468 ; iTemp2 [lr18:40]{short}[r2] := 0x0 {short}
8480 ; iTemp11 [lr19:40]{short}[r3] := 0x0 {short}
8492 ; iTemp21 [lr21:38]{short}[r4] := 0x0 {short}
8504 ; iTemp23 [lr22:38]{int}[r5 r6] := 0xa {int}
8523 ; iTemp17 [lr23:38]{int}[r7 r0] := 0x1e {int}
8552 ; iTemp13 [lr25:26]{char}[CC] = iTemp21 [lr21:38]{short}[r4] < 0xa {short}
8557 ; if iTemp13 [lr25:26]{char}[CC] == 0 goto _forbreak_0($7)
8602 ; iTemp2 [lr18:40]{short}[r2] = iTemp2 [lr18:40]{short}[r2] +
8607 ; iTemp21 [lr21:38]{short}[r4]
8633 ; iTemp15 [lr29:30]{short}[r1] = iTemp21 [lr21:38]{short}[r4] * 0x3 {short}
8666 ; iTemp11 [lr19:40]{short}[r3] = iTemp11 [lr19:40]{short}[r3] +
8671 ; iTemp15 [lr29:30]{short}[r1]
8690 ; iTemp17 [lr23:38]{int}[r7 r0]= iTemp17 [lr23:38]{int}[r7 r0]- 0x3 {short}
8737 ; _gint [lr0:0]{int} = _gint [lr0:0]{int} + iTemp17 [lr23:38]{int}[r7 r0]
8784 ; iTemp21 [lr21:38]{short}[r4] = iTemp21 [lr21:38]{short}[r4] + 0x1 {short}
8796 ; iTemp23 [lr22:38]{int}[r5 r6]= iTemp23 [lr22:38]{int}[r5 r6]- 0x1 {short}
8810 cjne r5,#0xff,00104$
8822 ; goto _forcond_0($4)
8844 ; ret iTemp24 [lr40:41]{short}
8893 \begin_inset LatexCommand \url{http://sdcc.sourceforge.net#Who}
8903 Thanks to all the other volunteer developers who have helped with coding,
8904 testing, web-page creation, distribution sets, etc.
8905 You know who you are :-)
8912 This document was initially written by Sandeep Dutta
8915 All product names mentioned herein may be trademarks of their respective
8921 \begin_inset LatexCommand \printindex{}