Warmcat homepage andy@warmcat.com
libwebsockets
{"schema":"libjg2-1", "vpath":"/git/", "avatar":"/git/avatar/", "alang":"", "gen_ut":1753165668, "reponame":"openssl", "desc":"OpenSSL", "owner": { "name": "Andy Green", "email": "andy@warmcat.com", "md5": "c50933ca2aa61e0fe2c43d46bb6b59cb" },"url":"https://warmcat.com/repo/openssl", "f":3, "items": [ { "schema":"libjg2-1", "oid":{ "oid": "ec9cbe21519f85e65e2f128d3f04bfd904642079", "alias": []},"tree": [ { "name": "00-base-templates.conf","mode": "33188", "size":11578}, { "name": "10-main.conf","mode": "33188", "size":79939}, { "name": "15-android.conf","mode": "33188", "size":10842}, { "name": "15-ios.conf","mode": "33188", "size":2411}, { "name": "50-djgpp.conf","mode": "33188", "size":624}, { "name": "50-haiku.conf","mode": "33188", "size":1218}, { "name": "50-masm.conf","mode": "33188", "size":1076}, { "name": "50-win-onecore.conf","mode": "33188", "size":3030}, { "name": "90-team.norelease.conf","mode": "33188", "size":5509}, { "name": "INTERNALS.Configure","mode": "33188", "size":7883}, { "name": "README","mode": "33188", "size":35494}, { "name": "README.design","mode": "33188", "size":24884}, { "name": "common.tmpl","mode": "33188", "size":8450}, { "name": "common0.tmpl","mode": "33188", "size":1324}, { "name": "descrip.mms.tmpl","mode": "33188", "size":50831}, { "name": "shared-info.pl","mode": "33188", "size":2791}, { "name": "unix-Makefile.tmpl","mode": "33188", "size":54514}, { "name": "unix-checker.pm","mode": "33188", "size":617}, { "name": "windows-checker.pm","mode": "33188", "size":649}, { "name": "windows-makefile.tmpl","mode": "33188", "size":30459}],"s":{"c":1753165668,"u": 5520}} ,{"schema":"libjg2-1", "cid":"80d2486d5f3fa6495cf3afbadfe58a68", "oid":{ "oid": "ec9cbe21519f85e65e2f128d3f04bfd904642079", "alias": []},"blobname": "Configurations/README", "blob": "Intro\n\u003d\u003d\u003d\u003d\u003d\n\nThis directory contains a few sets of files that are used for\nconfiguration in diverse ways:\n\n *.conf Target platform configurations, please read\n 'Configurations of OpenSSL target platforms' for more\n information.\n *.tmpl Build file templates, please read 'Build-file\n programming with the \u0022unified\u0022 build system' as well\n as 'Build info files' for more information.\n *.pm Helper scripts / modules for the main `Configure`\n script. See 'Configure helper scripts for more\n information.\n\n\nConfigurations of OpenSSL target platforms\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nConfiguration targets are a collection of facts that we know about\ndifferent platforms and their capabilities. We organise them in a\nhash table, where each entry represent a specific target.\n\nNote that configuration target names must be unique across all config\nfiles. The Configure script does check that a config file doesn't\nhave config targets that shadow config targets from other files.\n\nIn each table entry, the following keys are significant:\n\n inherit_from \u003d\u003e Other targets to inherit values from.\n Explained further below. [1]\n template \u003d\u003e Set to 1 if this isn't really a platform\n target. Instead, this target is a template\n upon which other targets can be built.\n Explained further below. [1]\n\n sys_id \u003d\u003e System identity for systems where that\n is difficult to determine automatically.\n\n enable \u003d\u003e Enable specific configuration features.\n This MUST be an array of words.\n disable \u003d\u003e Disable specific configuration features.\n This MUST be an array of words.\n Note: if the same feature is both enabled\n and disabled, disable wins.\n\n as \u003d\u003e The assembler command. This is not always\n used (for example on Unix, where the C\n compiler is used instead).\n asflags \u003d\u003e Default assembler command flags [4].\n cpp \u003d\u003e The C preprocessor command, normally not\n given, as the build file defaults are\n usually good enough.\n cppflags \u003d\u003e Default C preprocessor flags [4].\n defines \u003d\u003e As an alternative, macro definitions may be\n given here instead of in `cppflags' [4].\n If given here, they MUST be as an array of\n the string such as \u0022MACRO\u003dvalue\u0022, or just\n \u0022MACRO\u0022 for definitions without value.\n includes \u003d\u003e As an alternative, inclusion directories\n may be given here instead of in `cppflags'\n [4]. If given here, the MUST be an array\n of strings, one directory specification\n each.\n cc \u003d\u003e The C compiler command, usually one of \u0022cc\u0022,\n \u0022gcc\u0022 or \u0022clang\u0022. This command is normally\n also used to link object files and\n libraries into the final program.\n cxx \u003d\u003e The C++ compiler command, usually one of\n \u0022c++\u0022, \u0022g++\u0022 or \u0022clang++\u0022. This command is\n also used when linking a program where at\n least one of the object file is made from\n C++ source.\n cflags \u003d\u003e Defaults C compiler flags [4].\n cxxflags \u003d\u003e Default C++ compiler flags [4]. If unset,\n it gets the same value as cflags.\n\n (linking is a complex thing, see [3] below)\n ld \u003d\u003e Linker command, usually not defined\n (meaning the compiler command is used\n instead).\n (NOTE: this is here for future use, it's\n not implemented yet)\n lflags \u003d\u003e Default flags used when linking apps,\n shared libraries or DSOs [4].\n ex_libs \u003d\u003e Extra libraries that are needed when\n linking shared libraries, DSOs or programs.\n The value is also assigned to Libs.private\n in $(libdir)/pkgconfig/libcrypto.pc.\n\n shared_cppflags \u003d\u003e Extra C preprocessor flags used when\n processing C files for shared libraries.\n shared_cflag \u003d\u003e Extra C compiler flags used when compiling\n for shared libraries, typically something\n like \u0022-fPIC\u0022.\n shared_ldflag \u003d\u003e Extra linking flags used when linking\n shared libraries.\n module_cppflags\n module_cflags\n module_ldflags \u003d\u003e Has the same function as the corresponding\n `shared_' attributes, but for building DSOs.\n When unset, they get the same values as the\n corresponding `shared_' attributes.\n\n ar \u003d\u003e The library archive command, the default is\n \u0022ar\u0022.\n (NOTE: this is here for future use, it's\n not implemented yet)\n arflags \u003d\u003e Flags to be used with the library archive\n command. On Unix, this includes the\n command letter, 'r' by default.\n\n ranlib \u003d\u003e The library archive indexing command, the\n default is 'ranlib' it it exists.\n\n unistd \u003d\u003e An alternative header to the typical\n '\u003cunistd.h\u003e'. This is very rarely needed.\n\n shared_extension \u003d\u003e File name extension used for shared\n libraries.\n obj_extension \u003d\u003e File name extension used for object files.\n On unix, this defaults to \u0022.o\u0022 (NOTE: this\n is here for future use, it's not\n implemented yet)\n exe_extension \u003d\u003e File name extension used for executable\n files. On unix, this defaults to \u0022\u0022 (NOTE:\n this is here for future use, it's not\n implemented yet)\n shlib_variant \u003d\u003e A \u0022variant\u0022 identifier inserted between the base\n shared library name and the extension. On \u0022unixy\u0022\n platforms (BSD, Linux, Solaris, MacOS/X, ...) this\n supports installation of custom OpenSSL libraries\n that don't conflict with other builds of OpenSSL\n installed on the system. The variant identifier\n becomes part of the SONAME of the library and also\n any symbol versions (symbol versions are not used or\n needed with MacOS/X). For example, on a system\n where a default build would normally create the SSL\n shared library as 'libssl.so -\u003e libssl.so.1.1' with\n the value of the symlink as the SONAME, a target\n definition that sets 'shlib_variant \u003d\u003e \u0022-abc\u0022' will\n create 'libssl.so -\u003e libssl-abc.so.1.1', again with\n an SONAME equal to the value of the symlink. The\n symbol versions associated with the variant library\n would then be 'OPENSSL_ABC_\u003cversion\u003e' rather than\n the default 'OPENSSL_\u003cversion\u003e'. The string inserted\n into symbol versions is obtained by mapping all\n letters in the \u0022variant\u0022 identifier to upper case\n and all non-alphanumeric characters to '_'.\n\n thread_scheme \u003d\u003e The type of threads is used on the\n configured platform. Currently known\n values are \u0022(unknown)\u0022, \u0022pthreads\u0022,\n \u0022uithreads\u0022 (a.k.a solaris threads) and\n \u0022winthreads\u0022. Except for \u0022(unknown)\u0022, the\n actual value is currently ignored but may\n be used in the future. See further notes\n below [2].\n dso_scheme \u003d\u003e The type of dynamic shared objects to build\n for. This mostly comes into play with\n engines, but can be used for other purposes\n as well. Valid values are \u0022DLFCN\u0022\n (dlopen() et al), \u0022DLFCN_NO_H\u0022 (for systems\n that use dlopen() et al but do not have\n fcntl.h), \u0022DL\u0022 (shl_load() et al), \u0022WIN32\u0022\n and \u0022VMS\u0022.\n perlasm_scheme \u003d\u003e The perlasm method used to create the\n assembler files used when compiling with\n assembler implementations.\n shared_target \u003d\u003e The shared library building method used.\n This is a target found in Makefile.shared.\n build_scheme \u003d\u003e The scheme used to build up a Makefile.\n In its simplest form, the value is a string\n with the name of the build scheme.\n The value may also take the form of a list\n of strings, if the build_scheme is to have\n some options. In this case, the first\n string in the list is the name of the build\n scheme.\n Currently recognised build scheme is \u0022unified\u0022.\n For the \u0022unified\u0022 build scheme, this item\n *must* be an array with the first being the\n word \u0022unified\u0022 and the second being a word\n to identify the platform family.\n\n multilib \u003d\u003e On systems that support having multiple\n implementations of a library (typically a\n 32-bit and a 64-bit variant), this is used\n to have the different variants in different\n directories.\n\n bn_ops \u003d\u003e Building options (was just bignum options in\n the earlier history of this option, hence the\n name). This is a string of words that describe\n algorithms' implementation parameters that\n are optimal for the designated target platform,\n such as the type of integers used to build up\n the bignum, different ways to implement certain\n ciphers and so on. To fully comprehend the\n meaning, the best is to read the affected\n source.\n The valid words are:\n\n THIRTY_TWO_BIT bignum limbs are 32 bits,\n this is default if no\n option is specified, it\n works on any supported\n system [unless \u0022wider\u0022\n limb size is implied in\n assembly code];\n BN_LLONG bignum limbs are 32 bits,\n but 64-bit 'unsigned long\n long' is used internally\n in calculations;\n SIXTY_FOUR_BIT_LONG bignum limbs are 64 bits\n and sizeof(long) is 8;\n SIXTY_FOUR_BIT bignums limbs are 64 bits,\n but execution environment\n is ILP32;\n RC4_CHAR RC4 key schedule is made\n up of 'unsigned char's;\n RC4_INT RC4 key schedule is made\n up of 'unsigned int's;\n EXPORT_VAR_AS_FN for shared libraries,\n export vars as\n accessor functions.\n\n apps_aux_src \u003d\u003e Extra source to build apps/openssl and other\n apps, as needed by the target and that can be\n collected in a library.\n apps_init_src \u003d\u003e Init source to build apps/openssl and other\n apps, as needed by the target. This code\n cannot be placed in a library, as the rest\n of the code isn't expected to link to it\n explicitly.\n cpuid_asm_src \u003d\u003e assembler implementation of cpuid code as\n well as OPENSSL_cleanse().\n Default to mem_clr.c\n bn_asm_src \u003d\u003e Assembler implementation of core bignum\n functions.\n Defaults to bn_asm.c\n ec_asm_src \u003d\u003e Assembler implementation of core EC\n functions.\n des_asm_src \u003d\u003e Assembler implementation of core DES\n encryption functions.\n Defaults to 'des_enc.c fcrypt_b.c'\n aes_asm_src \u003d\u003e Assembler implementation of core AES\n functions.\n Defaults to 'aes_core.c aes_cbc.c'\n bf_asm_src \u003d\u003e Assembler implementation of core BlowFish\n functions.\n Defaults to 'bf_enc.c'\n md5_asm_src \u003d\u003e Assembler implementation of core MD5\n functions.\n sha1_asm_src \u003d\u003e Assembler implementation of core SHA1,\n functions, and also possibly SHA256 and\n SHA512 ones.\n cast_asm_src \u003d\u003e Assembler implementation of core CAST\n functions.\n Defaults to 'c_enc.c'\n rc4_asm_src \u003d\u003e Assembler implementation of core RC4\n functions.\n Defaults to 'rc4_enc.c rc4_skey.c'\n rmd160_asm_src \u003d\u003e Assembler implementation of core RMD160\n functions.\n rc5_asm_src \u003d\u003e Assembler implementation of core RC5\n functions.\n Defaults to 'rc5_enc.c'\n wp_asm_src \u003d\u003e Assembler implementation of core WHIRLPOOL\n functions.\n cmll_asm_src \u003d\u003e Assembler implementation of core CAMELLIA\n functions.\n Defaults to 'camellia.c cmll_misc.c cmll_cbc.c'\n modes_asm_src \u003d\u003e Assembler implementation of cipher modes,\n currently the functions gcm_gmult_4bit and\n gcm_ghash_4bit.\n padlock_asm_src \u003d\u003e Assembler implementation of core parts of\n the padlock engine. This is mandatory on\n any platform where the padlock engine might\n actually be built.\n\n\n[1] as part of the target configuration, one can have a key called\n 'inherit_from' that indicate what other configurations to inherit\n data from. These are resolved recursively.\n\n Inheritance works as a set of default values that can be overridden\n by corresponding key values in the inheriting configuration.\n\n Note 1: any configuration table can be used as a template.\n Note 2: pure templates have the attribute 'template \u003d\u003e 1' and\n cannot be used as build targets.\n\n If several configurations are given in the 'inherit_from' array,\n the values of same attribute are concatenated with space\n separation. With this, it's possible to have several smaller\n templates for different configuration aspects that can be combined\n into a complete configuration.\n\n instead of a scalar value or an array, a value can be a code block\n of the form 'sub { /* your code here */ }'. This code block will\n be called with the list of inherited values for that key as\n arguments. In fact, the concatenation of strings is really done\n by using 'sub { join(\u0022 \u0022,@_) }' on the list of inherited values.\n\n An example:\n\n \u0022foo\u0022 \u003d\u003e {\n template \u003d\u003e 1,\n haha \u003d\u003e \u0022ha ha\u0022,\n hoho \u003d\u003e \u0022ho\u0022,\n ignored \u003d\u003e \u0022This should not appear in the end result\u0022,\n },\n \u0022bar\u0022 \u003d\u003e {\n template \u003d\u003e 1,\n haha \u003d\u003e \u0022ah\u0022,\n hoho \u003d\u003e \u0022haho\u0022,\n hehe \u003d\u003e \u0022hehe\u0022\n },\n \u0022laughter\u0022 \u003d\u003e {\n inherit_from \u003d\u003e [ \u0022foo\u0022, \u0022bar\u0022 ],\n hehe \u003d\u003e sub { join(\u0022 \u0022,(@_,\u0022!!!\u0022)) },\n ignored \u003d\u003e \u0022\u0022,\n }\n\n The entry for \u0022laughter\u0022 will become as follows after processing:\n\n \u0022laughter\u0022 \u003d\u003e {\n haha \u003d\u003e \u0022ha ha ah\u0022,\n hoho \u003d\u003e \u0022ho haho\u0022,\n hehe \u003d\u003e \u0022hehe !!!\u0022,\n ignored \u003d\u003e \u0022\u0022\n }\n\n[2] OpenSSL is built with threading capabilities unless the user\n specifies 'no-threads'. The value of the key 'thread_scheme' may\n be \u0022(unknown)\u0022, in which case the user MUST give some compilation\n flags to Configure.\n\n[3] OpenSSL has three types of things to link from object files or\n static libraries:\n\n - shared libraries; that would be libcrypto and libssl.\n - shared objects (sometimes called dynamic libraries); that would\n be the engines.\n - applications; those are apps/openssl and all the test apps.\n\n Very roughly speaking, linking is done like this (words in braces\n represent the configuration settings documented at the beginning\n of this file):\n\n shared libraries:\n {ld} $(CFLAGS) {lflags} {shared_ldflag} -o libfoo.so \u005c\n foo/something.o foo/somethingelse.o {ex_libs}\n\n shared objects:\n {ld} $(CFLAGS) {lflags} {module_ldflags} -o libeng.so \u005c\n blah1.o blah2.o -lcrypto {ex_libs}\n\n applications:\n {ld} $(CFLAGS) {lflags} -o app \u005c\n app1.o utils.o -lssl -lcrypto {ex_libs}\n\n[4] There are variants of these attribute, prefixed with `lib_',\n `dso_' or `bin_'. Those variants replace the unprefixed attribute\n when building library, DSO or program modules specifically.\n\nHistorically, the target configurations came in form of a string with\nvalues separated by colons. This use is deprecated. The string form\nlooked like this:\n\n \u0022target\u0022 \u003d\u003e \u0022{cc}:{cflags}:{unistd}:{thread_cflag}:{sys_id}:{lflags}:{bn_ops}:{cpuid_obj}:{bn_obj}:{ec_obj}:{des_obj}:{aes_obj}:{bf_obj}:{md5_obj}:{sha1_obj}:{cast_obj}:{rc4_obj}:{rmd160_obj}:{rc5_obj}:{wp_obj}:{cmll_obj}:{modes_obj}:{padlock_obj}:{perlasm_scheme}:{dso_scheme}:{shared_target}:{shared_cflag}:{shared_ldflag}:{shared_extension}:{ranlib}:{arflags}:{multilib}\u0022\n\n\nBuild info files\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nThe build.info files that are spread over the source tree contain the\nminimum information needed to build and distribute OpenSSL. It uses a\nsimple and yet fairly powerful language to determine what needs to be\nbuilt, from what sources, and other relationships between files.\n\nFor every build.info file, all file references are relative to the\ndirectory of the build.info file for source files, and the\ncorresponding build directory for built files if the build tree\ndiffers from the source tree.\n\nWhen processed, every line is processed with the perl module\nText::Template, using the delimiters \u0022{-\u0022 and \u0022-}\u0022. The hashes\n%config and %target are passed to the perl fragments, along with\n$sourcedir and $builddir, which are the locations of the source\ndirectory for the current build.info file and the corresponding build\ndirectory, all relative to the top of the build tree.\n\nTo begin with, things to be built are declared by setting specific\nvariables:\n\n PROGRAMS\u003dfoo bar\n LIBS\u003dlibsomething\n ENGINES\u003dlibeng\n SCRIPTS\u003dmyhack\n EXTRA\u003dfile1 file2\n\nNote that the files mentioned for PROGRAMS, LIBS and ENGINES *must* be\nwithout extensions. The build file templates will figure them out.\n\nFor each thing to be built, it is then possible to say what sources\nthey are built from:\n\n PROGRAMS\u003dfoo bar\n SOURCE[foo]\u003dfoo.c common.c\n SOURCE[bar]\u003dbar.c extra.c common.c\n\nIt's also possible to tell some other dependencies:\n\n DEPEND[foo]\u003dlibsomething\n DEPEND[libbar]\u003dlibsomethingelse\n\n(it could be argued that 'libsomething' and 'libsomethingelse' are\nsource as well. However, the files given through SOURCE are expected\nto be located in the source tree while files given through DEPEND are\nexpected to be located in the build tree)\n\nIt's also possible to depend on static libraries explicitly:\n\n DEPEND[foo]\u003dlibsomething.a\n DEPEND[libbar]\u003dlibsomethingelse.a\n\nThis should be rarely used, and care should be taken to make sure it's\nonly used when supported. For example, native Windows build doesn't\nsupport building static libraries and DLLs at the same time, so using\nstatic libraries on Windows can only be done when configured\n'no-shared'.\n\nOne some platforms, shared libraries come with a name that's different\nfrom their static counterpart. That's declared as follows:\n\n SHARED_NAME[libfoo]\u003dcygfoo-{- $config{shlibver} -}\n\nThe example is from Cygwin, which has a required naming convention.\n\nSometimes, it makes sense to rename an output file, for example a\nlibrary:\n\n RENAME[libfoo]\u003dlibbar\n\nThat line has \u0022libfoo\u0022 renamed to \u0022libbar\u0022. While it makes no\nsense at all to just have a rename like that (why not just use\n\u0022libbar\u0022 everywhere?), it does make sense when it can be used\nconditionally. See a little further below for an example.\n\nIn some cases, it's desirable to include some source files in the\nshared form of a library only:\n\n SHARED_SOURCE[libfoo]\u003ddllmain.c\n\nFor any file to be built, it's also possible to tell what extra\ninclude paths the build of their source files should use:\n\n INCLUDE[foo]\u003dinclude\n\nIn some cases, one might want to generate some source files from\nothers, that's done as follows:\n\n GENERATE[foo.s]\u003dasm/something.pl $(CFLAGS)\n GENERATE[bar.s]\u003dasm/bar.S\n\nThe value of each GENERATE line is a command line or part of it.\nConfigure places no rules on the command line, except that the first\nitem must be the generator file. It is, however, entirely up to the\nbuild file template to define exactly how those command lines should\nbe handled, how the output is captured and so on.\n\nSometimes, the generator file itself depends on other files, for\nexample if it is a perl script that depends on other perl modules.\nThis can be expressed using DEPEND like this:\n\n DEPEND[asm/something.pl]\u003d../perlasm/Foo.pm\n\nThere may also be cases where the exact file isn't easily specified,\nbut an inclusion directory still needs to be specified. INCLUDE can\nbe used in that case:\n\n INCLUDE[asm/something.pl]\u003d../perlasm\n\nNOTE: GENERATE lines are limited to one command only per GENERATE.\n\nAs a last resort, it's possible to have raw build file lines, between\nBEGINRAW and ENDRAW lines as follows:\n\n BEGINRAW[Makefile(unix)]\n haha.h: {- $builddir -}/Makefile\n echo \u0022/* haha */\u0022 \u003e haha.h\n ENDRAW[Makefile(unix)]\n\nThe word within square brackets is the build_file configuration item\nor the build_file configuration item followed by the second word in the\nbuild_scheme configuration item for the configured target within\nparenthesis as shown above. For example, with the following relevant\nconfiguration items:\n\n build_file \u003d\u003e \u0022build.ninja\u0022\n build_scheme \u003d\u003e [ \u0022unified\u0022, \u0022unix\u0022 ]\n\n... these lines will be considered:\n\n BEGINRAW[build.ninja]\n build haha.h: echo \u0022/* haha */\u0022 \u003e haha.h\n ENDRAW[build.ninja]\n\n BEGINRAW[build.ninja(unix)]\n build hoho.h: echo \u0022/* hoho */\u0022 \u003e hoho.h\n ENDRAW[build.ninja(unix)]\n\nShould it be needed because the recipes within a RAW section might\nclash with those generated by Configure, it's possible to tell it\nnot to generate them with the use of OVERRIDES, for example:\n\n SOURCE[libfoo]\u003dfoo.c bar.c\n\n OVERRIDES\u003dbar.o\n BEGINRAW[Makefile(unix)]\n bar.o: bar.c\n \t$(CC) $(CFLAGS) -DSPECIAL -c -o $@ $\u003c\n ENDRAW[Makefile(unix)]\n\nSee the documentation further up for more information on configuration\nitems.\n\nFinally, you can have some simple conditional use of the build.info\ninformation, looking like this:\n\n IF[1]\n something\n ELSIF[2]\n something other\n ELSE\n something else\n ENDIF\n\nThe expression in square brackets is interpreted as a string in perl,\nand will be seen as true if perl thinks it is, otherwise false. For\nexample, the above would have \u0022something\u0022 used, since 1 is true.\n\nTogether with the use of Text::Template, this can be used as\nconditions based on something in the passed variables, for example:\n\n IF[{- $disabled{shared} -}]\n LIBS\u003dlibcrypto\n SOURCE[libcrypto]\u003d...\n ELSE\n LIBS\u003dlibfoo\n SOURCE[libfoo]\u003d...\n ENDIF\n\nor:\n\n # VMS has a cultural standard where all libraries are prefixed.\n # For OpenSSL, the choice is 'ossl_'\n IF[{- $config{target} \u003d~ /^vms/ -}]\n RENAME[libcrypto]\u003dossl_libcrypto\n RENAME[libssl]\u003dossl_libssl\n ENDIF\n\n\nBuild-file programming with the \u0022unified\u0022 build system\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\n\u0022Build files\u0022 are called \u0022Makefile\u0022 on Unix-like operating systems,\n\u0022descrip.mms\u0022 for MMS on VMS, \u0022makefile\u0022 for nmake on Windows, etc.\n\nTo use the \u0022unified\u0022 build system, the target configuration needs to\nset the three items 'build_scheme', 'build_file' and 'build_command'.\nIn the rest of this section, we will assume that 'build_scheme' is set\nto \u0022unified\u0022 (see the configurations documentation above for the\ndetails).\n\nFor any name given by 'build_file', the \u0022unified\u0022 system expects a\ntemplate file in Configurations/ named like the build file, with\n\u0022.tmpl\u0022 appended, or in case of possible ambiguity, a combination of\nthe second 'build_scheme' list item and the 'build_file' name. For\nexample, if 'build_file' is set to \u0022Makefile\u0022, the template could be\nConfigurations/Makefile.tmpl or Configurations/unix-Makefile.tmpl.\nIn case both Configurations/unix-Makefile.tmpl and\nConfigurations/Makefile.tmpl are present, the former takes\nprecedence.\n\nThe build-file template is processed with the perl module\nText::Template, using \u0022{-\u0022 and \u0022-}\u0022 as delimiters that enclose the\nperl code fragments that generate configuration-dependent content.\nThose perl fragments have access to all the hash variables from\nconfigdata.pem.\n\nThe build-file template is expected to define at least the following\nperl functions in a perl code fragment enclosed with \u0022{-\u0022 and \u0022-}\u0022.\nThey are all expected to return a string with the lines they produce.\n\n generatesrc - function that produces build file lines to generate\n a source file from some input.\n\n It's called like this:\n\n generatesrc(src \u003d\u003e \u0022PATH/TO/tobegenerated\u0022,\n generator \u003d\u003e [ \u0022generatingfile\u0022, ... ]\n generator_incs \u003d\u003e [ \u0022INCL/PATH\u0022, ... ]\n generator_deps \u003d\u003e [ \u0022dep1\u0022, ... ]\n generator \u003d\u003e [ \u0022generatingfile\u0022, ... ]\n incs \u003d\u003e [ \u0022INCL/PATH\u0022, ... ],\n deps \u003d\u003e [ \u0022dep1\u0022, ... ],\n intent \u003d\u003e one of \u0022libs\u0022, \u0022dso\u0022, \u0022bin\u0022 );\n\n 'src' has the name of the file to be generated.\n 'generator' is the command or part of command to\n generate the file, of which the first item is\n expected to be the file to generate from.\n generatesrc() is expected to analyse and figure out\n exactly how to apply that file and how to capture\n the result. 'generator_incs' and 'generator_deps'\n are include directories and files that the generator\n file itself depends on. 'incs' and 'deps' are\n include directories and files that are used if $(CC)\n is used as an intermediary step when generating the\n end product (the file indicated by 'src'). 'intent'\n indicates what the generated file is going to be\n used for.\n\n src2obj - function that produces build file lines to build an\n object file from source files and associated data.\n\n It's called like this:\n\n src2obj(obj \u003d\u003e \u0022PATH/TO/objectfile\u0022,\n srcs \u003d\u003e [ \u0022PATH/TO/sourcefile\u0022, ... ],\n deps \u003d\u003e [ \u0022dep1\u0022, ... ],\n incs \u003d\u003e [ \u0022INCL/PATH\u0022, ... ]\n intent \u003d\u003e one of \u0022lib\u0022, \u0022dso\u0022, \u0022bin\u0022 );\n\n 'obj' has the intended object file *without*\n extension, src2obj() is expected to add that.\n 'srcs' has the list of source files to build the\n object file, with the first item being the source\n file that directly corresponds to the object file.\n 'deps' is a list of explicit dependencies. 'incs'\n is a list of include file directories. Finally,\n 'intent' indicates what this object file is going\n to be used for.\n\n obj2lib - function that produces build file lines to build a\n static library file (\u0022libfoo.a\u0022 in Unix terms) from\n object files.\n\n called like this:\n\n obj2lib(lib \u003d\u003e \u0022PATH/TO/libfile\u0022,\n objs \u003d\u003e [ \u0022PATH/TO/objectfile\u0022, ... ]);\n\n 'lib' has the intended library file name *without*\n extension, obj2lib is expected to add that. 'objs'\n has the list of object files (also *without*\n extension) to build this library.\n\n libobj2shlib - function that produces build file lines to build a\n shareable object library file (\u0022libfoo.so\u0022 in Unix\n terms) from the corresponding static library file\n or object files.\n\n called like this:\n\n libobj2shlib(shlib \u003d\u003e \u0022PATH/TO/shlibfile\u0022,\n lib \u003d\u003e \u0022PATH/TO/libfile\u0022,\n objs \u003d\u003e [ \u0022PATH/TO/objectfile\u0022, ... ],\n deps \u003d\u003e [ \u0022PATH/TO/otherlibfile\u0022, ... ]);\n\n 'lib' has the intended library file name *without*\n extension, libobj2shlib is expected to add that.\n 'shlib' has the corresponding shared library name\n *without* extension. 'deps' has the list of other\n libraries (also *without* extension) this library\n needs to be linked with. 'objs' has the list of\n object files (also *without* extension) to build\n this library.\n\n This function has a choice; it can use the\n corresponding static library as input to make the\n shared library, or the list of object files.\n\n obj2dso - function that produces build file lines to build a\n dynamic shared object file from object files.\n\n called like this:\n\n obj2dso(lib \u003d\u003e \u0022PATH/TO/libfile\u0022,\n objs \u003d\u003e [ \u0022PATH/TO/objectfile\u0022, ... ],\n deps \u003d\u003e [ \u0022PATH/TO/otherlibfile\u0022,\n ... ]);\n\n This is almost the same as libobj2shlib, but the\n intent is to build a shareable library that can be\n loaded in runtime (a \u0022plugin\u0022...). The differences\n are subtle, one of the most visible ones is that the\n resulting shareable library is produced from object\n files only.\n\n obj2bin - function that produces build file lines to build an\n executable file from object files.\n\n called like this:\n\n obj2bin(bin \u003d\u003e \u0022PATH/TO/binfile\u0022,\n objs \u003d\u003e [ \u0022PATH/TO/objectfile\u0022, ... ],\n deps \u003d\u003e [ \u0022PATH/TO/libfile\u0022, ... ]);\n\n 'bin' has the intended executable file name\n *without* extension, obj2bin is expected to add\n that. 'objs' has the list of object files (also\n *without* extension) to build this library. 'deps'\n has the list of library files (also *without*\n extension) that the programs needs to be linked\n with.\n\n in2script - function that produces build file lines to build a\n script file from some input.\n\n called like this:\n\n in2script(script \u003d\u003e \u0022PATH/TO/scriptfile\u0022,\n sources \u003d\u003e [ \u0022PATH/TO/infile\u0022, ... ]);\n\n 'script' has the intended script file name.\n 'sources' has the list of source files to build the\n resulting script from.\n\nIn all cases, file file paths are relative to the build tree top, and\nthe build file actions run with the build tree top as current working\ndirectory.\n\nMake sure to end the section with these functions with a string that\nyou thing is appropriate for the resulting build file. If nothing\nelse, end it like this:\n\n \u0022\u0022; # Make sure no lingering values end up in the Makefile\n -}\n\n\nConfigure helper scripts\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nConfigure uses helper scripts in this directory:\n\nChecker scripts\n---------------\n\nThese scripts are per platform family, to check the integrity of the\ntools used for configuration and building. The checker script used is\neither {build_platform}-{build_file}-checker.pm or\n{build_platform}-checker.pm, where {build_platform} is the second\n'build_scheme' list element from the configuration target data, and\n{build_file} is 'build_file' from the same target data.\n\nIf the check succeeds, the script is expected to end with a non-zero\nexpression. If the check fails, the script can end with a zero, or\nwith a `die`.\n","s":{"c":1753165668,"u": 2443}} ],"g": 19533,"chitpc": 0,"ehitpc": 0,"indexed":0 , "ab": 1, "si": 0, "db":0, "di":1, "sat":0, "lfc": "0000"}