Get Gentoo!
gentoo.org sites
gentoo.org Wiki Bugs Forums Packages
Planet Archives Sources
Infra Status
summaryrefslogtreecommitdiff
path: root/eclass/toolchain.eclass
Commit message (Collapse)AuthorAgeFilesLines
* enable thumb mode by default for arm targets that are using the ↵Mike Frysinger2013-05-241-1/+4
| | | | microcontroller profile #418209 by Jiri Pittner & Boris Vingradov
* adjust case indentation so it is consistent in the whole fileMike Frysinger2013-05-211-79/+79
|
* convert confgcc to an array to avoid overloading $@ and to clean things up a ↵Mike Frysinger2013-05-211-100/+116
| | | | bit -- hopefully no real functional changes here
* Explicitly disable lto in 4.5 to stop configure from helpfully re-enabling ↵Ryan Hill2013-05-151-2/+6
| | | | it when libelf is present.
* Rename test USE flag to regression-test.Ryan Hill2013-05-151-6/+13
|
* Add lto USE flag for all versions. Drop LTO support for 4.5.Ryan Hill2013-05-151-11/+3
|
* disable libatomic when building a stage1 gcc compilerMike Frysinger2013-05-141-3/+5
|
* Drop graphite support for 4.4/4.5.Ryan Hill2013-05-091-23/+18
|
* Snapshots/vcs don't contain .info files.Ryan Hill2013-05-031-2/+6
|
* Extend workaround for bug #411333 to Via C3-2 processors (bug #466454 by ↵Ryan Hill2013-04-201-2/+3
| | | | Andrew Church).
* Since this is conditional code 4.6.4 will have to be handled here after all.Ryan Hill2013-04-171-3/+3
|
* Multiarch got backported to 4.6/4.7. 4.6.4 will be handled by patch since ↵Ryan Hill2013-04-171-2/+2
| | | | things are even more broken there.
* handle new gcc-4.8 MULTILIB_OSDIRNAMES logic that includes multiarch junkMike Frysinger2013-04-101-5/+9
|
* fix S clobbering that is not needed anymore to unbreak info page installs ↵Mike Frysinger2013-04-081-8/+4
| | | | #464008
* Fix for gcc info page installation #464008Patrick Lauer2013-04-081-2/+2
|
* make sure gcc/doc/ in the output exists, and copy all info files #464008Mike Frysinger2013-04-071-2/+3
|
* use simpler find statementsMike Frysinger2013-04-071-5/+5
|
* make sure we install the info pages after we disabled generation #464008Mike Frysinger2013-04-071-1/+4
|
* disable info page generation since gcc includes generated pages already ↵Mike Frysinger2013-04-061-1/+5
| | | | #464008 by Max Steel
* Work around bootstrap comparison failures on ancient systems (bug #411333).Ryan Hill2013-04-051-4/+9
|
* Revert previous due to breakage.Ryan Hill2013-04-021-5/+1
|
* avoid regenerating info pages due to changed version/bugurl fields #464008Mike Frysinger2013-04-011-2/+5
|
* drop --enable-target-optspace for uClibc targets ... there is no real need ↵Mike Frysinger2013-04-011-2/+2
| | | | for this, and if people want to build with -Os, they can. this also avoids build failures on ppc (PR43810).
* Add cloog/isl support for gcc 4.8 (bug # 434816, patch by Kacper Kowalik).Ryan Hill2013-03-161-8/+21
|
* update snapshot location to use gcc.gnu.org #455152Mike Frysinger2013-02-091-7/+6
|
* drop USE="bootstrap build" as it does very little and the existing logic ↵Mike Frysinger2013-01-241-12/+5
| | | | should handle things correctly #440224
* Drop go support for 4.6 - broken by newer glibc versions and upstream ↵Ryan Hill2013-01-151-3/+3
| | | | recommends it not be used.
* enable libffi with gcc-3.0+ (i did not have a local copy at the time to ↵Mike Frysinger2012-12-291-2/+2
| | | | check against)
* change `use doc` to `use_if_iuse doc` since not all gcc versions have USE=docMike Frysinger2012-12-291-2/+2
|
* only mess with libffi stuff with gcc-3.1+Mike Frysinger2012-12-291-3/+5
|
* restrict multilib logic in a few more places to gcc-3.2+, and libstdc++-v3 ↵Mike Frysinger2012-12-291-18/+25
| | | | to gcc-3+
* do not mention active gcc-config if we are removing a version that is not ↵Mike Frysinger2012-12-241-10/+13
| | | | selected #446830
* pass --enable-libstdcxx-time so that it tries to locate support in the C lib ↵Mike Frysinger2012-12-241-1/+8
| | | | -- if it does not find it, it will skip support like normal and not fail or fall back to -lrt #411681
* disable X configure flags setting up bad -I/-L flags that will mess with ↵Mike Frysinger2012-11-291-1/+7
| | | | multilib builds that need 3rd party libs (such as libjava and gtk/etc...)
* if uClibc is built w/out shared lib support (like nommu), then pass in ↵Mike Frysinger2012-11-241-2/+10
| | | | --disable-shared #291870 by Petric Frank
* depend on gtk/x emul packages when USE="amd64 multilib gcj gtk" #444350 by ↵Mike Frysinger2012-11-241-1/+5
| | | | David Kredba
* fix typo in env.d generation after recent rewrite so we store all ldpaths ↵Mike Frysinger2012-11-221-2/+2
| | | | and not just the last one #444193 by Robert Cernansky
* drop USE=nocxx support since everyone should be migrated nowMike Frysinger2012-11-221-11/+2
|
* stop disabling multilib libjava #425154 #435728Mike Frysinger2012-11-191-48/+40
|
* set FCFLAGS/FFLAGS when cross-compilingMike Frysinger2012-11-191-1/+3
|
* do not sed libstdc++ python dir with gcc-4.7+ since the issue has been fixed ↵Mike Frysinger2012-11-021-2/+2
| | | | in those versions
* depend on virtual/pkgconfig for gcj/gtk deps #436724 by Diego Elio PettenòMike Frysinger2012-10-011-2/+4
|
* use --enable-poison-system-directories with cross-compilers when the patch ↵Mike Frysinger2012-08-191-1/+5
| | | | is available
* update path to fix_libtool_files.sh in toolchain_pkg_postrm #420731 by Piotr ↵Mike Frysinger2012-06-111-3/+3
| | | | Karbowski
* add *-softfp-* support to the vendor field for people who want this for ↵Mike Frysinger2012-06-091-4/+12
| | | | binary compat with closed sourced drivers #416727 by Raúl Porcel; merge the tc-is-hardfloat helper into the one place where it is called in the whole tree to avoid further confusion between when the return values of tc-is-softfloat and tc-is-hardfloat
* restore pkg_postrm export -- commit 1.542 was meant to delete pkg_prerm, not ↵Mike Frysinger2012-06-051-2/+2
| | | | pkg_postrm as pointed out by Arfrever
* always build up the multilib list for gcc even when --disable-multilib as it ↵Mike Frysinger2012-06-031-3/+5
| | | | needs that info when picking the default target (to make native x32 work again)
* teach gcc how to handle CTARGET abis independent of CHOST abisMike Frysinger2012-06-031-9/+13
|
* punt pkg_preinst as it is merely a stub and no one calls it anymoreMike Frysinger2012-06-021-6/+2
|
* delete pkg_prerm from export_functions since its no longer definedMike Frysinger2012-06-021-2/+2
|