summaryrefslogtreecommitdiff
path: root/gnu
AgeCommit message (Collapse)Author
2010-05-24Link against libc_p and libpthread_p when -p or -pg is specified, like we didMark Kettenis
for gcc3. ok espie@
2010-05-23GCC PR C++/33094.Marc Espie
C++ semantics fix. Member constants can be file-local, if they're not external. Fixes build of py-tagpy, which barfs in Boost otherwise. (original fix, different text from gcc 4.2.4) okay deraadt@, kettenis@
2010-05-20We have a copy of git_version.h in src so that we can build libperlTodd C. Miller
with out own rules but apparently this causes problems after git_version.h is generated in the obj dir. This will only ever be a stub as we build from a release, not a git checkout. Add a pre-generated lib/Config_git.pl and prevent git_version.h and Config_git.pl from being removed by the clean target. Appears to fix parallel make issues. OK oga@
2010-05-20turn off strict overflow for all optimizations on OpenBSD, also disable wrapv.Dale Rahn
ok kettenis@ robert@, tested by marco@, kevlo@ myself
2010-05-17always provide "includes" targetChristian Weisgerber
2010-05-16the cp-demangler is part of libsupc++ as wellMarc Espie
2010-05-16Don't include -ftree-vrp in -O2 and higher as it has known bugs in gcc 4.2.Jonathan Gray
Fixes jdk builds among other things. ok robert@ espie@
2010-05-16install unwind.h, like for gcc3.Marc Espie
problem spotted by matthieu/naddy okay kettenis@, matthieu@
2010-05-15We really want position independant objects in libsupc++.a.Matthieu Herrb
ok espie@
2010-05-15Add libsupc++-v3Robert Nagy
2010-05-15Add Makefile for libsupc++-v3 that pulls common stuff fromRobert Nagy
libstdc++-v3/Makefile.common. ok espie@
2010-05-15Move libsupc++ sources and other common stuff to Makefile.commonRobert Nagy
to be used by gnu/lib/libsupc++-v3 and gnu/lib/libstdc++-v3 ok espie@
2010-05-14Fix ICE when checking a zero-size array for __bounded__ (TYPE_MAX_VALUENicholas Marriott
is now NULL for zero-length arrays so check for that). ok avsm
2010-05-12wint_t is int on OpenBSD; ok kettenis@Christian Weisgerber
2010-05-12port -ftrampolines/-Wtrampolines support from gcc3.Marc Espie
comments and okay kettenis@
2010-05-12options.[ch] will auto-define those.Marc Espie
okay kettenis@
2010-05-11default to march=i486 for gcc4 as we no longer run on 386 class machinesJonathan Gray
and this lets gcc emit code with xadd/cmpxchg instructions. ok kettenis@
2010-05-10Use OPENBSD_OS_CPP_BUILTINS(), and add a few missing defines to make sureMark Kettenis
the types used by GCC match those in machine/_types.h.
2010-05-10Switch to using COMPILER_VERSION instead of USE_GCC3, allows for more flexibleDale Rahn
complier picking, eg supporting gcc2, gcc3, and gcc4. based on diff by Marco, with fixes from espie@. ok espie@ general mumbling of approval of others.
2010-05-10fix a few strcpy issues.Marc Espie
okay millert@, nicm@ ... (thx others for pointing out stupidity in intermediate patches...)
2010-05-10Fix tags targetMarco Peereboom
ok oga kettenis committed in the skys over the continental us :-)
2010-05-10reflect reality, those types must match libc.Marc Espie
and then, we can have a default facet when locales are not defined. (this lets std::istringstream work again)
2010-05-09Set TARGET_CPU_DEFAULT to MASK_PA_11 on hppa to match what we had for gcc3.Mark Kettenis
2010-05-09-Wstack-larger-than-N for hppa/hppa64.Mark Kettenis
2010-05-09Makefile for libobjc for gcc4Robert Nagy
2010-05-09-Wstack-larger-than-N for sparc/sparc64.Mark Kettenis
2010-05-09Accept and ignore -Wbounded, fully implementing this is in the worksJonathan Gray
but this makes things less painful for people for the time being. ok kettenis@ robert@
2010-05-09This file won't be needed, espie pointed out how stupid I was.Robert Nagy
2010-05-09Modift -fstack-protector support such that it works with the supportMark Kettenis
functions in our libc and continues to provide the function name to __stack_smash_handler(). ok miod@
2010-05-09fix obvious typo, unbreak build on i386Marc Espie
2010-05-09Only #define _GLIBCXX_ATOMIC_BUILTINS on architectures where we use GCC'sMark Kettenis
builtin atomic operations. ok robert@
2010-05-08expand lvalue macro and remove an unnecessary and illegal cast inChristian Weisgerber
the process; ok kettenis@ miod@
2010-05-08Fixes to build with gcc4, mostly function prototypes and missingNicholas Marriott
headers. ok robert kettenis jsg
2010-05-08Disable -Wpointer-sign in -WallRobert Nagy
2010-05-08add i386 to the atomicity via builtins list as wellJonathan Gray
2010-05-08Use "builtins" for atomicity on sparc64 as well.Mark Kettenis
2010-05-08Some fixes from upstream to make this compile with gcc4.Mark Kettenis
ok jsg@, miod@
2010-05-08Install libsupc++ header files as well.Mark Kettenis
ok robert@
2010-05-07Add the bits of code I wrote for the ports gcc that are necessary to handleMark Kettenis
StackGhost on sparc64.
2010-05-07Add some missing HAVE_AS_SPARC defines necessary to make exception handlingMark Kettenis
work on sparc64.
2010-05-07Bring over changes from gcc3 (and gcc4 in ports) to make C++ exceptionMark Kettenis
handling work with shared libraries.
2010-05-07Add infrastructure to build libstdc++-v3 for gcc4Robert Nagy
ok kettenis@, marco@
2010-05-07Add openbsd config headers for libstdc++-v3Robert Nagy
2010-05-06Don't try to create /usr/bin; mtree will have taken care of that.Christian Weisgerber
ok kettenis@
2010-05-06Add missing cleandir targetRobert Nagy
2010-05-06Our libc provides ssp, so use that.Robert Nagy
2010-05-05Compile the additional sources to handle exceptionsRobert Nagy
2010-05-04We don't include gettext, dgettext or dcgettext as gcc's builtins, so don'tFederico G. Schwindt
check the arguments. Fixes autocrap tests in c++, which otherwise fail with "error: nonnull argument with out-of-range operand number". miod@ ok
2010-05-04fix .asm handling on powerpc and unify with alpha; ok drahn@Christian Weisgerber
2010-05-04add infrastructure to build GCC 4.2.1 for OpenBSD/alpha; ok kettenis@Christian Weisgerber