Liren Large Software Subsidy 7

home *** CD-ROM | disk | FTP | other *** search

/ Liren Large Software Subsidy 7 / 07.iso / c / c065 / 1.ddi / TCRTL.DOC < prev next >

Wrap

Text File | 1990-06-07 | 20.7 KB | 511 lines

TURBO C RUNTIME LIBRARY SOURCE CODE 3.0 DOCUMENTATION ───────────────────────────────────────────────────── Please read this document carefully. It contains information on how to install the Turbo C++ Runtime Library source files, an overview of what the Runtime Library contains, instructions for modifying Runtime Library files, and Borland's licensing agreement. TABLE OF CONTENTS ───────────────── 1. Installing Library Sources 2. Runtime Library Overview 3. Modifying Library Files 4. License Agreement 5. More information on the batch files 1. INSTALLING LIBRARY SOURCES ───────────────────────────── The following diagram shows the recommended directory structure for the libraries on a hard disk. This is also the directory structure created by the hard disk installation program, INSTALL. Most of the batch files that build the RTL files (except CLIPREPL.BAT and MATHREPL.BAT) assume this directory structure. In the diagram, and in the text that follows, we assume that you have installed the RTL directory structure in a subdirectory named \TCRTL. If you have put the RTL directory structure in a subdirectory other than \TCRTL, simply substitute the path to that directory wherever \TCRTL appears. RECOMMENDED LIBRARY DIRECTORY STRUCTURE ─────────────────────────────────────── \TCRTL library root directory \TCRTL\INCLUDE library header files (INCLUDE.ZIP) \TCRTL\CLIB general library sources (CLIB1.ZIP,CLIB2.ZIP) \TCRTL\CLIB\SMALL general object files: small model \TCRTL\CLIB\COMPACT general object files: compact model \TCRTL\CLIB\MEDIUM general object files: medium model \TCRTL\CLIB\LARGE general object files: large model \TCRTL\CLIB\HUGE general object files: huge model \TCRTL\MATH math library sources (MATH.ZIP) \TCRTL\MATH\OBJ math object files: model-independent \TCRTL\MATH\SMALL math object files: small model \TCRTL\MATH\COMPACT math object files: compact model \TCRTL\MATH\MEDIUM math object files: medium model \TCRTL\MATH\LARGE math object files: large model \TCRTL\MATH\HUGE math object files: huge model \TCRTL\EMU floating point initialization. (FPINIT.ASM) To install the library source files on your hard disk, insert the library source distribution disk 1 in your floppy drive, and enter the following command at the DOS prompt: A:INSTALL The INSTALL program will default to the library directories as shown above and extract the source and header/include files, placing them in the appropriate directories. It will also extract several batch files that may be used to recompile and/or rebuild the C and MATH libraries. It is a good idea to have a look at these batch files, to see how the C and MATH files are compiled and how the libraries are built. The following batch files will be placed in \TCRTL\CLIB and \TCRTL\MATH (replace the * with either CLIB or MATH, depending on the library): *.BAT - Compiles and assembles ALL modules in the library for the given memory model. *RCMP.BAT - Compiles (or assembles) a given module for the given memory model. *RLIB.BAT - Rebuilds the libraries for the given model from object files built by *.BAT or *RCMP.BAT. *REPL.BAT - Compiles/assembles a given module for a given memory model and replaces it in the C*.LIB or MATH*.LIB libraries. For usage information and an example, see section 5 or enter the name of the batch file without any parameters. These batch files assume the recommended directory structure. If you aren't using the recommended directory structure, the *REPL.BAT batch files can still be used. They need to be able to find your library file in which to replace the module. If you choose to compile or assemble a module without using the batch files, be aware that certain files need special switches. Make sure to check both the configuration and batch files to find the necessary switches. 2. RUNTIME LIBRARY OVERVIEW ─────────────────────────── Some of the library routines require .H files from Turbo C++, and you'll also need TLIB, the object file librarian. You will need to modify the TURBOC.CFG file in both the CLIB and MATH subdirectories to specify where your Turbo C++ header (.H) files are located. The runtime libraries are blocks of object code grouped together in several library files. These library files contain the general library routines and the standard math routines. FPINIT.ASM is provided to allow you to modify the floating-point initialization and termination process. If you used the installation batch file, this file will be found in a subdirectory called EMU. The CLIB* and MATH* batch files will not work with this file. FPINIT.ASM is used to create two OBJ files, one for the emulator (FPINIT.OBJ) and one for inline 8087 instructions (FP87INIT.OBJ). The following command lines will create these two files from the EMU subdirectory. TASM /MX /t /z fpinit TASM /MX /t /z /DFP87 fpinit, fp87init Most of the modules contained in the runtime libraries are object files produced by compiling the library source files with the Turbo C++ compiler. Some of the files contain inline assembly language code; therefore you will need a TASM 2.0 (or compatible) Macro Assembler to compile those source files. Finally, a few modules are written entirely in assembly language. Here is a key to some of the file extensions found in the RTL: .C - C source file .H - C header file .CAS - C source file with inline assembler .ASM - Assembler source file .ASI - Include file needed for .ASM files .CFG - Configuration files for Turbo C++ and Turbo Assembler. .RSP - TLIB response file. Contains a list of all files for a given library .ZIP - Archive file The library source files have been grouped together into several "archive" files (.ZIP extensions). An archive file contains a collection of files in a compressed form. You use the UNZIP utility to extract files from the archive. Here is the usage information for UNZIP: Usage: UNZIP [options] zipfile [destpath] [files...] Options: /o = overwrite existing files /v = verbose listing of archive(s) /c = extract file(s) to the console /p = extract file(s) to the printer /t = test archive integrity zipfile Archive file name, wildcards *,? ok. Default extension is .ZIP destpath Destination for extracted files. files Name(s) of files to extract. Wildcards *,? ok. Default is ALL files. Examples: UNZIP CLIB1 - Unpack all files in CLIB1 to current drive and directory. UNZIP MATH C:\TCRTL\MATH - Unpack all files in MATH.ZIP to C:\LIBRARY\MATH directory. UNZIP CLIB2 C: *.CPP *.CAS - Unpack all .CPP and .CAS files in CLIB2.ZIP to drive C. 3. MODIFYING LIBRARY FILES ────────────────────────── This section describes how to modify a source module, compile or assemble it, and replace it in a library. If you have installed the source using INSTALL, there are several batch files included to make your job easier; see section 5 for more information. You can modify a library file by first changing the corresponding source file, compiling (or assembling) the modified source file into an object file, and finally either replacing the changed object file in the appropriate library file(s) or rebuilding the library files from scratch. The latter method is somewhat slower and requires the presence of all the object modules that comprise a given library. This method may be preferable, however, when many modules have been changed or when you have lost track of which files have been modified. In order to compile most of the library source files, the standard header files and the library header files (contained in INCLUDE.ZIP) #included in most of the C source files must be present on your system. They must be made available to Turbo C++ through the -I option. The following steps are recommended for modifying a library module. You might want to modify all the modules you need, then rebuild the library (or libraries) from scratch. (Note: If you have installed all the source files on a hard disk using the installation procedure described in section 1, you may be able to skip some of the following steps.) A. Figure out which module you need to modify, and to which library file(s) it belongs. The library listing feature of TLIB will be helpful here, since it shows all the modules in a library file, together with all the functions that the modules define. Use "tlib lib-name,con" or "tlib lib-name,listfile-name." Note that the general routine libraries (CS.LIB, CC.LIB, CM.LIB, CL.LIB, and CH.LIB) all contain the same set of modules, compiled for different memory models; likewise for the math routine libraries (MATHS.LIB, MATHC.LIB, MATHM.LIB, MATHL.LIB, and MATHH.LIB). B. If you haven't done so already, extract the source file from the appropriate archive file using UNZIP, and place it in your working directory (for example, UNZIP A:CLIB2 CRTINIT.CAS). If you have not run INSTALL, unzip the header files from INCLUDE.ZIP and place them in either your working directory or the directory where you keep the standard C header files. Better yet install them in a separate subdirectory and add that subdirectory to the list of include subdirectories for TC or TCC. For example, to extract the header files and put them in \TCRTL\INCLUDE, do the following: 1. Create the directory, if it is not already present, like this: MKDIR \TCRTL\INCLUDE 2. Extract all files from INCLUDE.ZIP: UNZIP A:INCLUDE \TCRTL\INCLUDE If you did not run INSTALL, you also need to unzip the configuration files from CLIB2.ZIP and MATH.ZIP into their respective directories and modify them to include the directory for your Turbo C++ header files. For example: 1. Extract the configuration files: UNZIP A:CLIB2 \TCRTL\CLIB *.cfg UNZIP A:MATH \TCRTL\MATH *.cfg 2. Add the Turbo C++ include directory (assuming \TC\INCLUDE) to both TURBOC.CFG files: Edit TURBOC.CFG and add -I\TC\INCLUDE C. Apply and test all the changes you need to make. To test a C library module, write a simple testing program, and compile and link both modules at the same time; for example: TCC [options] tsrchp searchp.cas That way you make sure the tested module will not be linked in from the library. For an assembler file, assemble the library file, then link it with a test file using TCC; for example: TASM lrsh /MX; TCC [options] tstshf lrsh.obj D. Recompile the module and replace it in the library file(s). Two batch files are provided on the distribution disk to help with recompiling and replacing modules. These files will work even if you haven't installed the source with INSTALL. CLIBREPL.BAT - Compiles/assembles a given module for a given memory model and replaces it in the appropriate Cx.LIB library. MATHREPL.BAT - Compiles/assembles a given module for a given memory model and replaces it in the appropriate MATHx.LIB library. For usage information and an example, see section 5 or enter the name of the batch file without any parameters. If you are using a floppy-based system, don't use these batch files. At this point, it's easier just to update the library with the .OBJ file. Remember, TLIB creates a backup of the library file, so make sure you have plenty of space. 4. LICENSE AGREEMENT ──────────────────── Turbo C++ Runtime Library Souce Code Borland's No-Nonsense License Agreement Borland International, Inc. (Borland) is offering you a license to the source code to the Turbo C++ Runtime Library portion of Turbo C++ (the Source Programs), including updates that may later be supplied by Borland at additional cost, but not including the source code of the 8087 emulator or the graphics library. The Source Programs are protected by both United States copyright law and international treaty provisions. Therefore, you must treat the Source Programs just like a book, with the following single exception: Borland authorizes you to make archival copies of the Source Programs for the sole purpose of backing-up your programs and protecting your investments from loss. By "just like a book," Borland means, for example, that the Source Programs may be used by any number of people and may be freely moved from one computer workstation to another, so long as there is no possibility of them being used at one computer workstation at the same time they are being used at another. Just as a book can't be read by two different people in two different places at the same time, neither can the Source Programs be used by two different people in two different places at the same time. (Unless, of course, Borland's copyright has been violated.) You may modify the Source Programs, but the modified source code, regardless of the extent of the modifications, shall always remain Borland's source code. You may not remove or modify Borland's copyright and other proprietary copyright notices, and you may not transport any of the Source Programs to another computer operating system or environment. You are responsible for all claims, liability, and damages arising from your own modifications and the products which include them. You may use the Source Programs to support your licensed copies of Turbo C++. This means you may include all or some of the original or modified Source Programs routines in your own Turbo C++ based programs, but you may only distribute binary copies of such routines when linked into your executable programs. You may not distribute any source code for the original or modified Turbo C++ Runtime Library under any circumstances. You are not, of course, restricted from distributing your own source code. All rights not specifically granted in this license are reserved by Borland. 5. MORE INFORMATION ON THE BATCH FILES ────────────────────────────────────── The following descriptions are of various batch files you may find useful when modifying the library source files. All but the last two assume that you have installed the sources by using INSTALL. This information can also be found by entering the name of the batch file without any parameters. Note that CLIB.BAT, CLIBREPL.BAT, and CLIBRLIB.BAT invoke TLIB with the /0 option and that MATH.BAT, MATHREPL.BAT, and MATHRLIB.BAT invoke TLIB with the /E /0 options. If you want to build one or more library files to include debugging information (so that you can step through library calls with TC or TD), you have to remove both the /E and /0 options from the TLIB command line(s), as these options effectively remove all debugging information from library files. CLIB.BAT ──────── This batch file will recompile all the CLIB library files for the specifed memory model, placing the resultant object files in the appropriate directory (SMALL, COMPACT, MEDIUM, LARGE, or HUGE). The library file can then be built from these object files using the CLIBRLIB batch file. The "model" parameter specifies which memory model should be used. When ALL is specified, the library is recompiled for all memory models. Usage: clib model [tcc options] Examples: clib ALL -I\tc\include clib LARGE -I. -N -u- clib MEDIUM Note that recompiling all the source files using this batch file may take a considerable amount of time. CLIBRCMP.BAT ──────────── This batch file will recompile the specified source file for the selected memory model, and place the resultant object file in the SMALL, COMPACT, MEDIUM, LARGE, or HUGE subdirectory (for use with CLIBRLIB.BAT). The "model" parameter specifies which memory model should be used. When ALL is specified, the file is recompiled for all memory models. Usage: clibrcmp model filename extension [TCC or TASM switches] Examples: clibrcmp ALL screen cas clibrcmp HUGE fopen c -N clibrcmp LARGE setargv asm CLIBRLIB.BAT ──────────── This batch file rebuilds the specified CLIB library file from the object files contained in the SMALL, COMPACT, MEDIUM, LARGE, or HUGE subdirectories. The "model" parameter specifies which memory model library file should be rebuilt. When ALL is specified, all five library files will be rebuilt. Usage: clibrlib model Examples: clibrlib ALL clibrlib LARGE MATH.BAT ──────── This batch file will recompile all the MATH library files for the specifed memory model, placing the resultant object files in the appropriate directory (OBJ and SMALL, COMPACT, MEDIUM, LARGE, or HUGE). The library file can then be built from these object files using the MATHRLIB batch file. The "model" parameter specifies which memory model should be used. When ALL is specified, the library is recompiled for all memory models. Usage: math model [tcc options] Examples: math ALL -I\tc\include math LARGE -I. -p -N math MEDIUM Note that recompiling all the source files using this batch file may take a considerable amount of time. MATHRCMP.BAT ──────────── This batch file will recompile the specified source file for the selected memory model, and place the resultant object file in the SMALL, COMPACT, MEDIUM, LARGE, or HUGE subdirectory (for use with MATHRLIB.BAT). The "model" parameter specifies which memory model should be used. When ALL is specified, the file is recompiled for all memory models. Usage: mathrcmp model filename extension [TCC or TASM switches] Examples: mathrcmp ALL realcvt cas mathrcmp LARGE gcvt c -N MATHRLIB.BAT ──────────── This batch file rebuilds the specified MATH library file from the object files contained in the OBJ and SMALL, COMPACT, MEDIUM, LARGE, or HUGE subdirectories. The "model" parameter specifies which memory model library file should be rebuilt. When ALL is specified, all five library files will be rebuilt. Usage: mathrlib model Examples: mathrlib ALL mathrlib LARGE The last two batch files do not require any special directory setup. CLIBREPL.BAT ──────────── This batch file will recompile the specified module for the selected memory model, and replace the resultant object module in the SMALL, COMPACT, MEDIUM, LARGE, or HUGE library file in the specified library directory. The "model" parameter specifies which memory model should be used. When ALL is specified, the file is recompiled and replaced for all memory models. Usage: clibrepl model filename ext lib-dir [switches] Examples: clibrepl ALL screen cas \tc\lib -N clibrepl MEDIUM setargv asm \tc\lib Note that you should not include the trailing "\" in the destination library directory specification. MATHREPL.BAT ──────────── This batch file will recompile the specified module for the selected memory model, and replace the resultant object module in the SMALL, COMPACT, MEDIUM, LARGE, or HUGE library file in the specified library directory. The "model" parameter specifies which memory model should be used. When ALL is specified, the file is recompiled and replaced for all memory models. Usage: mathrepl model filename extension lib-dir [TCC switches] Examples: mathrepl ALL realcvt cas \tc\lib mathrepl LARGE ldexp cas \tc\lib -N Note that you should not include the trailing "\" in the destination library directory specification.