| Age | Commit message (Collapse) | Author | |
|---|---|---|---|
| 2006-11-22 | Improved cross build support for gcc on OpenBSD. allows the cross built | Dale Rahn | |
| compiler to have the same default options as native. The changes to cross build a native compiler is not quite complete. "Get this in" miod@ | |||
| 2006-11-16 | hack at build process some more, so that files are not rebuilt under src, | Marc Espie | |
| especially since we have the tools. problem reported by Ryan McBride, okay deraadt@, mcbride@ | |||
| 2006-11-14 | Do not align jump targets to cache line boundaries at -O2 on OpenBSD. | Miod Vallat | |
| 2006-11-11 | Provide size and type of symbols so that it doesn't confuse ld when building | Dale Rahn | |
| shared binaries/libraries. | |||
| 2006-11-11 | OpenBSD defaults to dynamic linker of /usr/libexec/ld.so | Dale Rahn | |
| 2006-11-10 | Yes, openbsd does not support shared libraries on sh yet, but why make it hard? | Dale Rahn | |
| 2006-11-09 | Dont build all of the extra libgcc.a files, they are not needed for | Dale Rahn | |
| native OpenBSD. | |||
| 2006-11-06 | Move back to softfloat for now. | Dale Rahn | |
| 2006-11-03 | Use __start (the symbol we provide) not start. ok miod@ | Dale Rahn | |
| 2006-10-31 | Fix infinite loop when passing object with 3+ base classes by value: | Peter Stromberg | |
| http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18300 Needed by openoffice and boost on amd64 ok espie@ miod@ | |||
| 2006-10-28 | Make stack executable when doing trampolines on macppc | Pedro Martelletto | |
| Okay espie@, kettenis@ | |||
| 2006-10-18 | Under OpenBSD, do not force -msmall-data if -fpic and -mlarge-data if -fPIC. | Miod Vallat | |
| Instead, have both -fpic and -fPIC only select pic code generation, and use whatever memory mode is specified (with -mlarge-data being the implicit default). This means that -fpic and -fPIC will be equivalent by default under OpenBSD again, as was the case with gcc 2.x; this will help porter's lives. No consensus but a majority of developers in favour of this change (including martin@, pvalchev@, deraadt@ and I), so in it goes. | |||
| 2006-10-11 | Default to -m4. | Miod Vallat | |
| 2006-10-08 | Define superH obsd-specific emulations for ld, and use it by default. | Miod Vallat | |
| 2006-10-06 | Disable stack protector in OVERRIDE_OPTIONS regardless of the commandline | Miod Vallat | |
| settings for now. | |||
| 2006-10-06 | Early bits for a 32-bit SuperH toolchain, currently for little-endian | Miod Vallat | |
| systems only; more work is necessary in ld land. | |||
| 2006-09-29 | garbage collect obsolete or unused macros - TARGET_ENDIAN_DEFAULT, | Brad Smith | |
| PREFERRED_DEBUGGING_TYPE and ASM_OUTPUT_DEFINE_LABEL_DIFFERENCE_SYMBOL. ok kettenis@ martin@ | |||
| 2006-07-20 | Introduce a new compiler warning, -Wstack-larger-than-N, to report | Miod Vallat | |
| functions which are too greedy in stack variables. This is intended to be used for kernel compiles, where this warning will be enabled for a reasonable size (after a few weeks grace period so that people can upgrade their compiler). Please note that this warning relies upon md code, and as such is only available on platforms OpenBSD runs on; also, the stack size being warned on is only the local variables size, regardless of the ABI stack usage requirements and the callee-saved registers; which means a function may be warning-clean yet need more stack space than meets the eye; the actual size being checked on may change to include these extras in the future. | |||
| 2006-07-12 | makeinfo --no-split, simplifies sets immensely. | Marc Espie | |
| Approved deraadt@, kettenis@ | |||
| 2006-07-11 | Fix false-positive -Wbounded warnings for sscanf(3) format-string checking. | Anil Madhavapeddy | |
| They only happened on non-char arrays, so the new code restricts the checking to these valid types. ok miod@ | |||
| 2006-05-13 | Declare a symbol of __mips64__for 64-bit MIPS targets. | Brad Smith | |
| ok deraadt@ drahn@ | |||
| 2006-04-18 | Backport fix for gcc bug #10692. Fixes code generation for htons() | Otto Moerbeek | |
| on i386 with -march=i686 -O2. Problem report and nice test program by chefren at pi dot net. Note that the fix is in non-target specific code, even while the gcc people classified this as a m68k bug. ok deraadt@ | |||
| 2006-01-13 | Adapt to the machine/_types.h changes; size_t and ptrdiff_t are now | Todd C. Miller | |
| defined in terms of long, not int on all architectures. | |||
| 2006-01-06 | Include <sys/types.h> and don't define away __size_t | Todd C. Miller | |
| Based on existing bits for FreeBSD 5 | |||
| 2005-12-22 | Install header files for MMX/SSE intrinsics on amd64 too. | Mark Kettenis | |
| ok pval@ | |||
| 2005-12-13 | Don't define _BIG_ENDIAN on macpcc, only __BIG_ENDIAN__. | Todd C. Miller | |
| Adapted from netbsd.h. OK drahn@ deraadt@ | |||
| 2005-10-19 | Teach unwinder about StackGhost. | Mark Kettenis | |
| ok deraadt@ | |||
| 2005-08-01 | Change the compilers default scheduling settings from PA-8000 (4 instructions | Miod Vallat | |
| in parallel) to PA-700 (no parallel instructions) on 32-bit hppa. This causes gcc to produce slightly faster code on most supported hppa, no change in code size. Tested by marcop@, mickey@, I and a few others; ok mickey@ | |||
| 2005-05-25 | spelling fixes | David Krause | |
| 2005-04-23 | fix WCHAR_TYPE* | Marc Espie | |
| okay drahn@, tested alek@ | |||
| 2005-04-16 | Tell gcc -Wformat that our kprintf implementation understands %q and %ll | Chad Loder | |
| for quads. OK deraadt@, some comments millert@ | |||
| 2005-04-15 | pedantic fix for wchar type inside gcc. | Marc Espie | |
| checked through a make release by drahn@ | |||
| 2005-04-08 | amd64 canonicalization. | Mark Kettenis | |
| ok deraadt@, pvalchev@ | |||
| 2005-03-29 | Switch i386 over to DWARF-based exception handling. | Mark Kettenis | |
| ok deraadt@ | |||
| 2005-03-29 | Plug a memory leak. From gcc 3.4 branch. | Miod Vallat | |
| 2005-02-10 | -ftrampolines, not -ftrampoline; | Jason McIntyre | |
| from moritz grimm; ok espie@ | |||
| 2005-02-07 | zap sp64elf.h, get the only config part really used (PREFERRED_DEBUGGING_TYPE) | Marc Espie | |
| into openbsd64.h. Keep the other stuff in #if 0 for now. okay miod@, brad@, pval@... | |||
| 2005-02-06 | disallow trampolines by default. okay miod@, jmc@ | Marc Espie | |
| 2005-02-05 | -dynamic-linker ld.so is NOT a normal object file. Fixes issues in trying | Marc Espie | |
| to build a gcc3 sparc. (reviewed and accepted upstream) | |||
| 2005-02-04 | unfuck amd64 | Peter Valchev | |
| 2005-01-25 | hppa64-unknown-openbsd build support; espie@ miod@ brad@ ok | Michael Shalayeff | |
| 2005-01-25 | simplify OpenBSD case by using arch(1) and sed to grab | Brad Smith | |
| the machine architecture. We now output amd64 instead of x86_64 as it should be. ok deraadt@ pvalchev@ | |||
| 2005-01-24 | get rid of i386elf.h which is an embedded platform include. | Marc Espie | |
| Get the only define we want, namely the debugger registers. with this, i386 emits dwarf2 by default. Considering how well stabs are tested in recent gcc (not), this is a good thing. Sleuthing work by kettenis@ and me. Tried out by lots of people, including me, naddy, pvalchev, kettenis, henning, hshoexer... | |||
| 2005-01-24 | add the availability check of reg_renumber array. | Hiroaki Etoh | |
| ok pvalchev@ | |||
| 2005-01-14 | handle sentinel for real. Code, taken from gcc 4.0, which was based on | Marc Espie | |
| my design. okay millert@, otto@. | |||
| 2005-01-02 | add zaurus, fix cats entry. ok jakob@ pvalchev@ deraadt@ | Dale Rahn | |
| 2005-01-02 | -fzero-initialized-in-bss is wrong, meaning code that | Ted Unangst | |
| depends on zero initialized variables showing up in data fails. turn it off by default. ok deraadt@ miod@ | |||
| 2004-12-30 | gcc 3.3.5 loop optimization will produce incorrect output on powerpc; a | Miod Vallat | |
| victim of this being awk(1), which in turn caused some manual pages to be generated without dates at the bottom. Since this is fixed by the (invasive) loop optimization rewrite done for 3.4, we can not expect a 3.3 fix anytime soon, and the fix is non-trivial, so disable the failing section of code on powerpc only. This has the net effect of magically turning -O1 and -O2 into -Os as far as -floop-optimize is concerned. Problem spotted by deraadt@, researched by otto@ and yours truly; this workaround agreed by many. | |||
| 2004-12-26 | Remove duplicate definition of LONG_DOUBLE_TYPE_SIZE, and shrink WINT_TYPE | Miod Vallat | |
| to 32 bits, as done on all other OpenBSD platforms. ok espie@ | |||
| 2004-12-25 | zap this too | Marc Espie | |
 |
index : src | |
| OpenBSD base system |
| summaryrefslogtreecommitdiff |