summaryrefslogtreecommitdiff
path: root/mk/wrapper
AgeCommit message (Collapse)AuthorFilesLines
2008-06-17Add a bunch of supported options.sketch1-1/+8
2008-04-21Revert rev 1.3 as it breaks e.g. -Wl,--whole-archive. See PR 38476 forjoerg1-2/+2
details.
2008-04-20Keep libraries specified as .a files in their original position amongkim1-2/+2
other libraries (specified with -l). Fixes a linking problem in net-snmp, where the Perl DynaLoader.a library becomes out-of-order when all -l arguments have been moved to the end of the command line.
2008-02-19Add IRIX command sinks, needed to get rpath fixup when using GCC.tnn3-1/+57
2008-02-19Pass gcc's -mabi=* option (used on mips) without warning.tnn1-1/+2
2007-11-28Do not load the $cache file for every argument. Instead, load it oncerillig2-5/+8
and then call it as a shell function. Note that the shell function is not called "cache", since some shells have problems when a function has the same name as a variable. This speeds up the wrapper by as little as 75 percent for the final link command of editors/abiword on SunOS-5.10-sparc (before: 20 seconds, after: 5 seconds).
2007-11-28Oops, don't overflow the terminal.rillig1-2/+2
2007-11-28sunpro doesn't need -lstdc++.rillig1-1/+4
2007-11-28Using a single awk process instead of a pipe of four grep processesrillig1-4/+6
should be both faster and easier to read.
2007-11-19Sunpro doesn't need -rdynamic, since the linker is sensible by default.rillig1-1/+2
This change also works around shells/scsh's broken check for -rdynamic.
2007-11-08Preserve the -Wl,-Bstatic / -Wl,-Bdynamic arguments, but alsohe1-1/+11
duplicate them into the list of libraries. Someone may want to mix static and dynamic linking. Fixes PR#37228, approved by jlam. jlam says that a better long-term fix would be to ensure that all the -L specifications come before the -l specifications, instead of moving all the -l specifications to the end of the command line.
2007-11-01sunpro doesn't know -ggdb.rillig1-1/+4
2007-10-17Handle more variants of the position independent code flag from Solaris.is1-1/+3
This is needed for Xaw3d. Fix suggested by joerg@.
2007-10-09Remove trailing spaces.martti1-2/+2
2007-10-05When linking with XLC on AIX, force -brtl to get linkage behaviour wejoerg1-1/+8
expect from other platforms.
2007-09-20When CPP is defined as "cc -E", the wrapper framework created therillig1-1/+2
wrapper for gcc and cc as type "CPP", although these aren't C preprocessors. By preventing the code from overwriting files, this works now.
2007-09-19Fixed incorrect quoting of arguments. This behavior had been unnoticedrillig2-9/+4
for several years now, since the directories in the -I... and -L... options usually don't need quoting. But when the -D... option was processed using the same code, the bug had become visible.
2007-09-19Before and after the transformations, the commands are written to therillig1-3/+4
log file in a format that allows copy-and-paste into a shell. Before, it wasn't possible to see from the log file whether a space was a delimiter or part of a shell word.
2007-09-13The word "skipped" is ambiguous. Use "untransformed" instead.rillig1-2/+2
2007-09-11Merge -D and the following word, so -D UNIX -D WHATEVER doesn't getrillig1-2/+2
transformed to -D UNIX WHATEVER (because the wrapper eliminates duplicate arguments). Now it becomes -DUNIX -DWHATEVER. This fixes a build problem reported in http://mail-index.netbsd.org/pkgsrc-users/2007/09/10/0004.html.
2007-09-08gcc knows -fpic.rillig1-1/+2
2007-08-02Add core of the infrastructure support for cross-compilation.joerg4-2/+177
- USE_CROSS_COMPILATION activates it, CROSS_DESTDIR specifies root of the target filesystem - derive _CROSS_DESTDIR from CROSS_DESTDIR or MAKEOBJDIR - buildlink3.mk prefixes the files to symlink with _CROSS_DESTDIR - compiler/gcc.mk knows about the target prefix (e.g. i386--netbsdelf) - PKG_DBDIR is prefixed with _CROSS_DESTDIR - package-install and bin-install are not called with su - install and strip are redirected to the tool version - links for the target specific ar, as, ld, nm, objdump, ranlib and strip are added - compiler wrapper detect if linking is requested or not - special command sinks for CPP and CC/CXX add the cross-compile magic: - modify include dirs to get the target /usr/include - modify linker dirs and runpath to use target /usr/lib at link time, but keep correct rpath entries Supported-by: Google SoC 2007 Basic tests by he@ on Sparc. Review from jlam@.
2007-06-18Changed the plain "mkdir" into ${MKDIR}, like in all the other cases.rillig1-2/+2
Apparently, the tools directory isn't in the PATH at the point where this code is executed. Fixes PR 35487.
2007-06-15Added transformations for -R, -fPIC, -kPIC, -KPIC and -g3.rillig1-1/+5
2007-06-09Added a slight variation of the patch provided in PR 36187 by Naotorillig1-4/+7
Morishima.
2007-06-06Replaced ${_PKG_SILENT}${_PKG_DEBUG} with a simple ${RUN} to save somerillig1-49/+39
space.
2007-05-28Remove trailing spaces.martti2-4/+4
2007-05-19When a wrapped compiler or linker command fails, the transformed commandrillig1-2/+2
line is no longer printed by default. To get that behavior back, you can set the variable WRAPPER_DEBUG to "yes". This avoids some noise during the build. The old behavior was mostly useful on Solaris and IRIX, where a -std=c99 or -c99 option had been added to the compiler's arguments by the wrapper. This caused diagnostics that were hard to understand, since the user did not specify these flags and there was no sign that they had been added.
2007-05-06Back out previous. The packages that incorrectly assume that ldtnn2-7/+1
--version-script works are few enough to warrant fixing each one of them instead. Silenty dropping the flag makes it impossible to detect whether or not --version-script is supported. Pointed out by joerg@
2007-04-29--version-script is GNU ld specific, remove from command line.tnn2-1/+7
2007-04-28Correct comment about -g3 flag.tnn1-2/+2
2007-04-22Handle linker optimization flag.tnn1-1/+6
2007-04-21fix typotnn1-2/+2
2007-04-21add workaround for ld(1) run path problems and document thosetnn2-1/+33
2007-04-14Teach pkgsrc about HP-UX.tnn3-1/+106
2007-04-12Add OSF1 wrapper quirks to always enable posix sockets and iso C++ iostreams.tnn1-1/+4
Fixes PR 31444. Fixes PR 31447. Both problems reported by Rumi Szabolcs.
2007-04-10+ -march=*rillig1-1/+2
2007-04-09Remove special treatment of "DynaLoader.a" completely. It's not necessarytron1-14/+1
with current libtool version and breaks the build under Darwin 8.x and NetBSD 4.0_BETA2.
2007-03-15wrapper_type can be IMAKE, too.rillig1-2/+2
2007-03-07Added the flags to control diagnostics.rillig1-1/+7
2007-03-07Fixed the documentation of the opt:*:* transformation.rillig1-2/+2
2007-02-23The compiler option -Wl,foo,bar,baz is split into -Wl,foo -Wl,barrillig1-1/+26
-Wl,baz before processing the resulting options again. Fixes PR 32203.
2007-02-22Changed the default warning message to include the word "unknown", sincerillig1-2/+2
otherwise it doesn't sound like a warning.
2007-02-22More documentation.rillig1-3/+3
2007-02-21gcc understands -Wno-unused-parameter.rillig1-1/+2
2007-02-21Changed the failure into a warning when options such as "-I -L" appearrillig1-2/+2
next to each other.
2007-02-21The options -I, -L and -R must not be followed directly by anotherrillig2-2/+12
option. Instead of just merging them, the wrapper fails. Let's see which packages get broken by that and then fix them. I've also heard rumours that a plain -L is used as an alias to -nostdlib in some compilers. This would have been handled wrong up to now, since it definitely wasn't intended to be merged with the following argument.
2007-02-20Mapping command line options to the ABI is not as simple as I thought:rillig1-8/+15
-32 does _not_ belong to ABI 32. Instead, the mapping is: ABI Options (undef) -o32 -32 32 -n32 64 -64 While here, allowing -OPT:* to pass through to the compiler, since it will understand it. This should fix PR 35729.
2007-02-19The ABI for which the code should be generated can be queried from therillig4-5/+28
wrapper_abi variable. This is used in the MIPSpro wrapper to reject any package that tries to compile a file for the wrong ABI. This is not just a warning because such a package most likely has more severe problems. The gcc -Wall option seems to be similar to the MIPSpro -fullwarn option, so map them.
2007-02-19The argument -64 means to use the 64-bit ABI for Sun C and MIPSpro.rillig1-1/+2