summaryrefslogtreecommitdiff
path: root/gnu
AgeCommit message (Collapse)Author
2013-01-31Enable shared libraries on m88k platforms.Miod Vallat
And there was much rejoicing.
2013-01-30Revert m88k.md 1.7 and m88k.c 1.27, 1.26 and 1.24 (partially) to restore theMiod Vallat
existing behaviour of wrapping big PIC addresses in `unspec's to prevent the cse pass from modifying them. However, instead of building the unspec ourselves in the C code, rely upon explicitely named define_insn() constructs in the .md description. Then, fix the ambiguity this introduces between the locate1 and movsi_high_pic insns by defining symbolic constants as the unspec selector, and using them, instead of 0.
2013-01-30Revert link-time resolution of R_88K_GOT_16[HL] modified in 1.5, but keep theMiod Vallat
error test. The logic was correct, I was tricked by bogus compiler generated assembly.
2013-01-28Correctly compute the relocation value of GOT_16[HL] relocations when they canMiod Vallat
be resolved at link time, such as when linking a binary from objects which have been compiled PIC (one such case being perl's miniperl). While there, borrow the check and error message for GOT_16[HL] with nonzero addend (which can only happen with badly crafted .s files and do not behave like one may naively think they do) from elf32-ppc.
2013-01-28Use symbolic_address_p() instead of simply checking whether GET_CODE() isMiod Vallat
SYMBOL_REF or LABEL_REF, when deciding to put #got_rel relocations. This allows (symbol + constant) to be correctly relocated when building PIC.
2013-01-27The removal of the wrapping of PIC symbol references in an UNSPEC vectorMiod Vallat
caused the code responsible for adding the proper #got_rel relocation to the second part of the PIC symbol address load. Instead of using an UNSPEC for that purpose (which violates assumptions made by the MI code anyway), check for the the second operand of a LO_SUM being either a SYMBOL_REF or a LABEL_REF, to output the #got_rel relocation, and drop the remaining bits of UNSPEC handling.
2013-01-24Fix unused local and missing initializer warnings.Miod Vallat
2013-01-24Get rid of a few warnings by pouring prototypes and punctuation.Miod Vallat
2013-01-23m88k_expand_prologue(): add a REG_MAYBE_DEAD note to the insns loading theMiod Vallat
PIC base register, it case it turns out to not be necessary. This should not be needed because this sequence is only generated if current_function_uses_pic_offset_table is set, but it can't hurt to expect the worse (from rs6000). legitimize_address(): do not coerce the address using (unspec:SI gen_rtvec() ), but rather use it unmodified, in the -fPIC case. This lets libstdc++.so build (and probably anything else being compiled with -fPIC rather than -fpic).
2013-01-20Enable multilib to build -fpic and -fPIC versions of libgcc.a, in addition toMiod Vallat
the fully static one. Update the compiler specs to allow for dynamic linking of libraries and binaries.
2013-01-20Generate linker scripts for shared libraries. Force padding sections aroundMiod Vallat
plt, because it may need to be written to during initial relocation.
2013-01-20Rewrite. This is heavily based upon elf32-m68k.c, with a few things borrowedMiod Vallat
from elf32-ppc.c. The plt entry setup matches what can be found in Motorola's SVR4/m88k binaries.
2013-01-20Implement tc_fix_adjustable() on ELF, and use it to prevent #got_relMiod Vallat
relocations to local symbols to be rewritten as relocations to the beginning of the section they are in, plus an addend, as addend can't work for these relocations, the way the got is set up. Inspired by tc-ppc.
2013-01-20Backport two fixes for amd64 from binutils 2.16.Brad Smith
- For DefaultSize instructions, don't guess a 'q' suffix if the instruction doesn't support it. - Corrections for x86_64 assembly. ok miod@
2013-01-15Simplify OBSD_LIB_SPEC a bit for gcc3/4.Brad Smith
ok miod@
2013-01-15Bring back symbol visibility support. This had been done, then backed out,Miod Vallat
a long time ago, because I had not realized symbol visibility required a lot of shared library major bumps to avoid mixed binaries. However, at the moment, none of the platforms using gcc 3 supports shared libraries (this is being worked on), and the remaining gcc 2 platforms won't support gcc 3 until switching from a.out to ELF, which makes the risk of mixing visibility-unaware binaries with visibility-aware binaries moot. Reminded by brad@.
2013-01-15When configured for ELF object format output, recognize #abdiff, #got_rel andMiod Vallat
#plt symbol modifiers, and produce the appropriate relocation types. This allows as(1) to correctly assemble PIC code.
2013-01-13Change an instance of submit.cf that should now be openbsd-submit.cf.Todd C. Miller
From grunk@
2013-01-11Automagically define __PIC__ and __pic__ to 1 or 2, when building withMiod Vallat
-fpic or -fPIC. Same for __PIE__ and __pie__; matches the behaviour of the in-tree gcc 4.
2013-01-09In the WANT_LDAP=yes case, always link libsasl2Matthieu Herrb
(libdap depends on it). ok millert@ ajacoutot@
2013-01-07Teach as(1) how to handle mips symbolic register names.bcallah
Backported from a commit made after binutils 2.17 but before the GPLv3 switch. ok miod@
2013-01-05Replace the list of platforms enabled (or not) to build the whole binutilsMiod Vallat
with a test of ELF_TOOLCHAIN provided by bsd.own.mk. No functional change.
2013-01-05Switch m88k ports to ELF.Miod Vallat
2013-01-05Remove the -mno-underscores, -msvr3 and -msvr4 options. They suppose theMiod Vallat
compiler can be built on one particular m88k system and output code for another with the proper options, but it only does half the job since there is too much knowledge of the native toolchain embedded (at compile-time) in the compiler. Building code for a different operating system and toolchain is what cross-compiler are for. Remove support for SDB debug information (only used by SVR3) and tdesc debug information (only used by SVR4). The latter is questionable, but this helps making the code less entangled. Tell m88k.h to not rewrite almost all ASM_WHATEVER macros gratuitously on the false pretense that `all m88k assemblers use the same syntax for pseudo-ops'. This was already not true with SVR4, and is getting worse with our ELF toolchain.
2013-01-05Rename m88k_pound_sign to m88k_register_prefix (even though the prefix isMiod Vallat
used for more than registers). Convert all output of registers to asm_fprintf with %R. Stop altering reg_names[] at runtime depending on the selected register prefix.
2013-01-04Remove extern declarations redundant with "flags.h"Miod Vallat
2013-01-03The removal of sdata support also removed the only use of SYMBOL_REF_FLAGMiod Vallat
in the m88k backend; as a result, SHORT_ADDRESS_P() will always return 0, therefore remove all code depending on it being nonzero.
2013-01-02More dead features recycling: -mwarn-passed-struct doesn't do anything, andMiod Vallat
-m{no-,}optimize-arg-area doesn't do anything either since the arg area has been removed in gcc 2.95.
2013-01-02Help the DWARF code to handle our prologues correctly:Miod Vallat
- add the frame pointer to EPILOGUE_USES when profiling - when the initial stack adjustement is larger than 64KB and requires a temporary register to be used, add a REG_FRAME_RELATED_EXPR of the original computation (not constrained by the adjustement size) to the REG_NOTES of the stack pointer adjustement sequence. Borrowed from alpha which has a similar issue.
2013-01-02Remove last boolean argumont of output_function_profiler(), it is always invokedMiod Vallat
with the same value.
2013-01-02Start removing obsolete features from the m88k backend (it's not like we'llMiod Vallat
need to sync with upstream anymore) to make things a bit simpler and get rid of unintended bitrotting. This time: - remove the small-data feature. It was only supported for COFF binaries anyway. This allows us to stop providing TARGET_ASM_SELECT_SECTION and use the default flavour, from which this backend had derived a bit in addition to sdata support, so this is a good thing... - remove the -midentify-version option - remove the -mbig-pic option, it has been superseded by -fPIC since years and was supposed to be removed `in a future version' since a long time as well. - remove the old MONITOR_GCC debug hooks.
2013-01-02Oops, commited this by accident when adding the m88k ELF toolchain bits.Miod Vallat
2013-01-01m88k ELF toolchain, supporting static binaries only so far.Miod Vallat
Assembler supports a register prefix of `#' when configured for SVR4, and `%' when configured for OpenBSD. Pseudo-ops do not need a leading dot, except when configured for OpenBSD, allowing the `set' instruction to be recognized correctly. Based upon various unfinished or unreliable works, including smurph@'s early work in 2003, my own attempts in 2004 and 2005, and Michael Kato's binutils 2.15 diff from 2005. Support for register prefixes, as well as fixing an evil off-by-one in resolving 16-bit pc-relative relocations in ld (which would only affect kernel .S files branching to global symbols), came as part of the 2012 work. Not enabled yet (needs Makefile.bsd-wrapper, bfd/config.bfd and gdb/configure.tgt updates to completely enable) so as not to disturb the existing a.out toolchain builds; will hopefully be enabled very soon.
2012-12-31Correctly disassemble {ld,st}.*.usr opcodes.Miod Vallat
2004-05-17import binutils 2.14 (excluding testsuites, .info files, and .po files)Dale Rahn
2012-12-31Remove GLIBC_DYNAMIC_LINKER which is only relevant for Linux.Brad Smith
ok miod@
2012-12-31Make use of OPENBSD_OS_CPP_BUILTINS_COMMON here as well.Brad Smith
ok miod@
2012-12-30Stop including other files in m88k/openbsd.h; instead specific the complete listMiod Vallat
of files to include, in proper order, as tm_file at configure time. This requires openbsd.h to be split, as also encountered on other platforms, and will make further work easier. No functional change introduced.
2012-12-26Update to sendmail 8.14.6Todd C. Miller
2012-12-25Fix longstanding bug in FRAME_POINTER_REQUIRED; unless -fomit-frame-pointerMiod Vallat
is used, a frame pointer is required for all non-leaf functions, and for leaf functions unless -momit-leaf-frame-pointer is specified. Also, a frame pointer is mandatory when profiling. This used to be mostly harmless with gcc 2.95, but is required with gcc 3 to get profiling to work.
2012-12-25Define INCOMING_RETURN_ADDR_RTX and DWARF_FRAME_RETURN_COLUMN to allow forMiod Vallat
DWARF debugging information to be produced, but define DWARF2_UNWIND_INFO to zero to prevent it from being output yet. This exposes dire weaknesses in the way the dwarf unwinder recognizes stack frame setup instructions, so borrow code from rs6000 to rewrite and annotate them as dwarf-edible instructions. No functional change intended - although this enables one more optimization code path in gcc.
2012-12-20Bring stack smashing protector for MIPS; enabled by default.Martynas Venckus
Reorganize soft frame pointer so that locals are below it and grow downwards. Tested by miod@, jasper@. OK miod@.
2012-12-20Bring stack smashing protector for Alpha; enabled by default.Martynas Venckus
Reorganize soft frame pointer so that locals are below it and grow downwards. Thanks Nick for the access. OK miod@.
2012-12-17Make the SH register spill failure for R0 workaround work againMartynas Venckus
after switching to __guard_local. OK matthew@, miod@.
2012-12-12Previous commit introduced a bug; only return early from m88k_layout_frame()Miod Vallat
if reload_completed; INITIAL_ELIMINATION_OFFSET() invokes it (which is very likely a source for more bugs, to be investigated)
2012-12-10Fix c-bounded when arrays are declared without the size expression.Martynas Venckus
The array_size was uninitialized and used to work by accident. Spotted with SSP on MIPS. OK miod@. Tested by jasper@.
2012-12-10Fix unaligned memory loads on Alpha. GCC used to generate them inMartynas Venckus
the following manner: ([reg:A & -8] << (64 - (((reg:FP+reg:B) & 0x7) << 3))) >> 56 This fails when we're doing loads with the offset from the frame pointer %8. Since it's aligned, optimizer makes it a zero. The correct expression is: ([reg:A & -8] << (56 - (((reg:FP+reg:B-1) & 0x7) << 3))) >> 56 This is actually a 13-year-old bug. Checked by Miod; a few files in the kernel were affected. Spotted with SSP for Alpha. OK miod@. Tested by naddy@.
2012-12-09Remove vax-specific configuration constructs.Miod Vallat
2012-12-08Work-in-progress gcc 3.3.6 configuration for OpenBSD/vax.Miod Vallat
2012-12-08Bring in a couple of bugfixes from old (Aug 2000), but more recent than thisMiod Vallat
one, gas, to correctly output pc-relative jumps.
generated by cgit v1.2.3 (git 2.25.1) at 2025-01-16 02:03:55 +0000