summaryrefslogtreecommitdiff
path: root/gnu/lib
AgeCommit message (Collapse)Author
2006-08-28regenerate. (largely from dsl@, thanks.) libstdc++ should supportmrg
wide characters properly again now, for everyone.
2006-08-24regenerate with many fixes in libstdc++-v3 configury.mrg
2006-08-23regenerate with recent wchar/iconv fixes.mrg
2006-08-23regenerate with partly-fixed gcc4/libstdc++-v3 configury.mrg
2006-07-23MKDEBUG -> MKDEBUGLIBtsutsui
2006-07-16Remove C++ sources which are linked into the build area in "cleandir"tron
target. This fixes problems when trying to build another architecture in the same source tree without using object dirs.
2006-07-07regenerate with fixed gcc configury and libcmrg
2006-07-06for PowerPC64, use -mlong-double-128 when building libgcc.mrg
2006-07-06regenerate with fixed armeb configurymrg
2006-07-04Remove "cs-tconfig.h" in "cleandir" target.tron
2006-07-03regeneratemrg
2006-07-02regeneratemrg
2006-07-02these aren't used and probably won't be so clean them up.mrg
2006-07-02regeneratemrg
2006-07-02regeneratemrg
2006-07-02regeneratemrg
2006-07-02regeneratemrg
2006-07-02regenerate.mrg
2006-07-02regeneratemrg
2006-07-01regeneratemrg
2006-07-01regenerate.mrg
2006-07-01regenerate these.mrg
2006-06-30mknative binutils for powerpc64mrg
2006-06-30don't build _mulsc3.o, _muldc3.o, _mulxc3.o, _multc3.o, _divsc3.o,mrg
_divdc3.o, _divxc3.o or _divtc3.o for vax, since they depend on __builtin_inf which does not exist there.
2006-06-30mknative files for vax / gcc4.mrg
2006-06-30regenerate mknative for armeb.mrg
2006-06-30run mknative for powerpc64.mrg
2006-06-26regenerate for armmrg
2006-06-26regenerate arm, sh3el (now properly working) and sh3eb (not yetmrg
properly working.)
2006-06-24regenerate hppa mknative files. this mostly works but i have tomrg
compile significant portions of the tree at -O1 instead of -O2, and sometimes -Os/-O0 is required, and others none works. i'd guess that some instructions in the .md file are generating more assembler than they claim to be since all these errors are due too large offsets. but fixing that is work for someone else, or at least another day :-)
2006-06-23factor out the unwind.h building code. use it for libgcc itself (neededmrg
when there is no libgcc_s.) add dependancies for the generated arch/defs.mk to optionlist.
2006-06-23add GCC4 mknative files for sun2. seems to build ok..mrg
2006-06-12More files changes in mknative re-run. I'm still cheating withskrll
src/tools/gcc/Makefile
2006-06-12Re-run mknative for the gcc 3.3.6 import. There's a bit of cheating goingskrll
on here - mainly the use of an old src/tools/gcc/Makefile.
2006-06-04remove a comment that isn't really valid.mrg
2006-06-03actually use the local generated options.h rather than the empty versionmrg
from the libgcc build dir. this should fix powerpc build issues.
2006-06-02remove support for building (with) GCC 2.95. also:mrg
- always install <stdbool.h> - don't generate a fake one for vax / gettext.
2006-06-02make GCC3 work for vax - without shared library support.mrg
2006-05-27sync RCS IDs in comments, no functional changedrochner
2006-05-27reran native-gcc, fixing the COMDAT problemdrochner
2006-05-26regenerate. fixes mipsel HAVE_COMDAT_GROUPmrg
2006-05-24generate new libobjc4 files and regenerate many many others.mrg
XXX: re-running "make native-gcc" some times causes the XXX: gcc/configure test for COMDAT groups to fail. a fresh XXX: configure (ie, with no objdir/.native tree yet) seems to XXX: fix this. right now mipsel and alpha have the wrong XXX: definition in the checked in auto-host.h.
2006-05-23generate options.h like we do for gcc4/backend. so far only seems tomrg
be strictly necessary for powerpc so allow the old MI defs.mk to be used if a MD one does not exist. when all platforms are regenerated this support can be removed along with libobjc4/defs.mk.
2006-05-21mknative generated files for hppa and GCC4. seems to kinda of build, not ↵mrg
run-tested.
2006-05-20generated sh5 mknative GCC4 files. seems to build ok, but as-yet untested.mrg
2006-05-20add mknative files for sh3* and GCC4. these build the basics but bombmrg
out in libc due to the current implementation of va_arg() using lvalue casts and GCC4 gets very unhappy with lvalue casts.
2006-05-20also -I.../gcc/config. this fixes the sh3 builds.mrg
2006-05-20generated sparc and m68k GCC4 mknative files. these are only buildmrg
tested and mvme68k was the m68k platform has been built.
2006-05-19generated mknative files for amd64 and GCC4. not tested but a releasemrg
build completes.
2006-05-19add alpha native supportdrochner
(a kernel compiled with gcc4 runs for me, userland untested)