summaryrefslogtreecommitdiffstats
path: root/toolchain
Commit message (Collapse)AuthorAgeFilesLines
* Add proper target support for newfangled sh targets.Paul Mundt2005-05-022-0/+125
| | | | | | | | | | | | | GCC folk in their infinite wisdom have added half-assed target support for newer sh cores. The support is there, but of course we have no way to actually enable the thing, so we do that with this patch. Likewise, binutils has no idea about any of these new targets either (not even a half-assed attempt, sadly), so we fix that up too.. now we're able to actually build a real toolchain for sh2a_nofpu- and other more ineptly named toolchains (and yes, there are more inept targets than that one, really. Go look, I promise).
* No need to patch this when there's a configuration parameter.Manuel Novoa III2005-05-021-15/+0
|
* sh64 no longer implies sh as far as uClibc is concerned..Paul Mundt2005-05-021-1/+0
|
* Obsolete.Manuel Novoa III2005-05-021-41/+0
|
* This is the 100 patch for 2.16.90.0.2. Note that 2.16.90.0.2 seems to be ↵Manuel Novoa III2005-05-021-30/+107
| | | | broken for mips.
* Ugly hack to force the configuration for the cross libstdc++ to fail for ↵Manuel Novoa III2005-05-011-0/+12
| | | | complex funcs we don't yet support.
* No longer needed and actually breaks the gcc java build.Manuel Novoa III2005-05-011-41/+0
|
* Prune out a bunch of binutils versions that are no longerEric Andersen2005-05-0153-10583/+1
| | | | the latest and greatest in their respective binutils series
* Weed out older, no longer maintained versions of gccEric Andersen2005-05-0128-14772/+0
|
* It is now official, the cause of the mysterious gcc 3.3.5 exceptionEric Andersen2005-05-011-1/+2
| | | | | | handling failure is definately sjlj. Only enable it for the apparently broken gcc 3.3.3 toolchain, which _does_ need it for exception handling to work properly.
* Fix soft float options for gcc 4.Manuel Novoa III2005-04-301-0/+7
|
* Someone forgot mipsel for soft float...Manuel Novoa III2005-04-301-1/+1
|
* Add 4.0.0 support. Some arches may still need some tweaking and/or ↵Manuel Novoa III2005-04-306-0/+3817
| | | | additional patches. But I successfully built cross and target toolchains for x86 and mipsel supporting c/c++/objective-c/java. Of course I haven't tested them yet...
* Support gcc 4.[01]. Only tried for 4.0.0 at this point.Manuel Novoa III2005-04-301-0/+7
|
* Add gcc 4.0.0 support. Note... I haven't needed the sjlj exception support ↵Manuel Novoa III2005-04-301-0/+4
| | | | in some time, but haven't touched the default in this commit.
* Fill in a couple of omissions.Manuel Novoa III2005-04-301-30/+107
|
* Needed --disable-werror to get 2.16.mumble build from dying on missing ↵Manuel Novoa III2005-04-301-0/+5
| | | | prototypes in lex-generated files.
* grab the -linux-gnu* -> -linux-* config changes from mainline and backport ↵Mike Frysinger2005-04-303-156/+727
| | | | them to the 2.16 branch
* 2.16.90.0.2 releasedMike Frysinger2005-04-308-0/+447
|
* Add support for soon to be released binutils-2.16 as well as"Steven J. Hill"2005-04-2815-0/+4170
| | | | | | GCC-4.1.0. These tool versions are needed so that NPTL support can be added into uClibc. Backporting NPTL from GCC-4.1.0 into GCC-4.0 may be possible, but not recommended.
* Don't forget about libgcc_s_nof.so when processingJoakim Tjernlund2005-04-251-4/+4
| | | | libgcc_s.so
* Fix broken patch. Was this even tested?"Steven J. Hill"2005-04-231-1/+1
|
* Change all instances of 'ln -sf' to 'ln -snf'Eric Andersen2005-04-152-6/+6
|
* Add a bunch of patches snarfed from the debian gdb packageEric Andersen2005-04-1421-0/+3566
|
* delete older 2.15.94.0.2 and add new 2.16.90.0.1Mike Frysinger2005-04-1110-801/+155
|
* update gdb a bit. remove support for ancient versions.Eric Andersen2005-03-267-618/+68
|
* Overwrite any existing link should there be oneEric Andersen2005-03-261-1/+1
|
* Add missing definition for 'struct ip_nat_multi_range_compat' soEric Andersen2005-03-221-0/+18
| | | | that iptables can compile
* Bug #110, acmay writes: The gdb version depends on BR2_HOST_GDB, but it ↵Mike Frysinger2005-03-211-2/+2
| | | | should be BR2_PACKAGE_GDB_CLIENT.
* Bump available sanitized kernel headersEric Andersen2005-03-182-16/+20
|
* Be extra careful about where the uClibc config file comes fromEric Andersen2005-03-091-0/+2
|
* Fixup gcc 3.3.3 and 3.3.4 builds with patches backportedEric Andersen2005-03-0810-0/+800
| | | | from stuff I added to gcc-3.3.5
* gcc 3.4.0 also needs to not emit embedded pic relocs for mipsEric Andersen2005-03-021-0/+83
|
* Fix things the other way round. Instead of restoring the mipsEric Andersen2005-03-023-119/+59
| | | | | embedded pic support into binutils, instead fixup gcc to not emit any embedded pic relocations.
* bother. Fix the filenameEric Andersen2005-03-021-0/+0
|
* 117_mips_symbolic_link is from debianEric Andersen2005-03-021-0/+44
|
* 450-mips-restore-embedded-pic is I believe the minimum necessary to get gccEric Andersen2005-03-021-0/+75
| | | | | | | 3.3.x to compile again on mips. As was, compiling gcc 3.3.x for mips would cause gcc to blow up whining about combining symbols from different segments, since stock binutils no longer supports the PC relative expressions emitted by gcc 3.3.x for mips when generating embedded PIC code.
* Always use CFLAGS_FOR_BUILD for using CC_FOR_BUILDEric Andersen2005-03-0211-0/+352
|
* kill off support for gdb 5.3Eric Andersen2005-03-024-803/+1
|
* simplify target language selection to remove nasty nested ifeqEric Andersen2005-03-021-6/+10
|
* teach BR2_INSTALL_LIBSTDCPP to actually install libstdc++Eric Andersen2005-03-021-5/+12
|
* On mips, default to 4k pages to make binaries smallerEric Andersen2005-03-0114-0/+364
|
* use tested patch from upstreamMike Frysinger2005-03-011-1/+1
|
* 3.3.5 needs arm big endian lovin tooMike Frysinger2005-03-012-0/+93
|
* Added support for gdb 6.3. Rediffed 100-uclibc-conf.patch to fixJoakim Tjernlund2005-02-274-0/+331
| | | | obvious rejects.
* pdinoto writes:Eric Andersen2005-02-252-0/+11
| | | | | | This patch adds an option in GCC config to compile Objective-C language support in the toolchain. It is disabled by default, so it is not intrusive.
* Trivial updatesEric Andersen2005-02-251-2/+2
|
* This doesn't need ncurses headers installed in the targetEric Andersen2005-02-241-1/+1
|
* 2.15.94.0.2.2 has been releasedMike Frysinger2005-02-197-3/+1035
|
* oldMike Frysinger2005-02-196-1016/+0
|