Age | Commit message (Collapse) | Author | Files | Lines |
|
i386os9k.mt tm-i386os9k.h
* i386os9k.mt: newly add i386/os9000 make file.
* tm-i386os9k.h: newly add i386/os9000 target dependent .h file.
|
|
ChangeLog config.sub
* config.sub: Add os9k checking.
|
|
(STEP_SKIPS_DELAY): Added.
* infrun.c (proceed) [STEP_SKIPS_DELAY]: Check for a breakpoint in
the delay slot.
|
|
ChangeLog
|
|
ChangeLog targets.c Makefile.in cache.c
Added Files:
i386os9k.c
* i386os9k.c : new file to handle os9k format bfd.
* Makefile.in : Handle new file i386os9k.c
* bfd-in2.h : Add bfd_target_os9k_flavour.
* targets.c : Add bfd_target_os9k_flavour and i386os9k_vec.
* cache.c : Initialize cache_sentinel to 0.
|
|
an error, regardless of force_ellipsis. In the non-EIO case,
just print the error message rather than calling error(). Don't
access *(bufptr-1) if bufptr points to the start of the buffer.
When looking for '\0', don't increment bufptr and addr if bufptr
started out already at limit. If an error happens on fetching the
first character, don't print the string.
|
|
|
|
(line_comment_chars): Add ! to list.
(md_apply_fix): Do process 32- and 64-bit relocations.
|
|
ic960coff.
* config/ic960coff.mt: New file.
* config/obj-coffbfd.h [TC_I960]: Include coff/i960.h.
(TARGET_FORMAT) [TC_I960]: Use coff-Intel-little.
* config/te-ic960.h (CROSS_COMPILE): Don't undef this. We'll always build
little-endian object files.
* config/tc-i960.c (md_reloc_size): Don't define at all if BFD or
BFD_ASSEMBLER is defined.
(mem_fmt): Since COFF doesn't handle callx relocations yet, treat them like
normal 32-bit relocations.
(md_apply_fix): For callx relocations, store zero.
(tc_bout_fix_to_chars): Store symbol idx for all callx relocations, regardless
of link-relax setting.
(tc_coff_fix2rtype, tc_coff_sizemachdep): New functions.
(i960_handle_align) [! OBJ_BOUT]: If link-relax option is selected, print an
error message and clear it.
* config/tc-i960.h (BFD_ARCH, COFF_FLAGS, COFF_MAGIC, TC_COUNT_RELOC,
TC_COFF_FIX2RTYPE, TC_COFF_SIZEMACHDEP, tc_fix_adjustable): New macros.
(tc_coff_fix2rtype, tc_coff_sizemachdep): Declare.
|
|
(obj_coff_init_stab_section): Cast alloca result.
|
|
(objdump_start): Deleted unused proc.
(objdump_start_common): Merged into objdump_start_no_subdir.
* gas/alpha/fp.exp: Use objdump instead of objdump_start_no_subdir, since the
former actually waits for objdump to finish. Specify .rdata section only.
Make comment indicate Alpha architecture rather than SPARC.
* gas/alpha/fp.d: Omit .reginfo patterns. Just use "." to match
against ASCII code 0x2a ("*", special in regexp).
* gas/sun4/addend.exp: Use objdump instead of objdump_start_no_subdir.
|
|
* mpw-xconfig.in (emulname): Define.
|
|
|
|
* mpw-build.in: Remove, subsumed in mpw-make.in.
* mpw-make.in (bindir): Fix pathname.
(install): Move here from mpw-build.in.
|
|
* ldlang.c (cat): Define using ANSI style if ALMOST_STDC defined.
|
|
obsolete.
* i387-tdep.c, i386-tdep.c i386v-nat.c, i386aix-nat.c,
i386m3-nat.c, config/m68k/tm-m68k.h, i960-tdep.c
config/i960/tm-i960.h, remote-nindy.c, config/m88k/tm-m88k.h,
m88k-tdep.c: Use floatformat.h instead of ieee-float.h.
* sparc-tdep.c: Remove now-obsolete ieee-float.h stuff
* findvar.c: Update comment regarding ieee-float.h.
|
|
|
|
i386m3-nat.c, config/m68k/tm-m68k.h, i960-tdep.c
config/i960/tm-i960.h, remote-nindy.c, config/m88k/tm-m88k.h,
m88k-tdep.c: Use floatformat.h instead of ieee-float.h.
* sparc-tdep.c: Remove now-obsolete ieee-float.h stuff
* findvar.c: Update comment regarding ieee-float.h.
|
|
obsolete.
* i387-tdep.c, i386-tdep.c i386v-nat.c, i386aix-nat.c,
i386m3-nat.c, config/m68k/tm-m68k.h, i960-tdep.c
config/i960/tm-i960.h, remote-nindy.c, config/m88k/tm-m88k.h,
m88k-tdep.c: Use floatformat.h instead of ieee-float.h.
* sparc-tdep.c: Remove now-obsolete ieee-float.h stuff
* findvar.c: Update comment regarding ieee-float.h.
|
|
i386m3-nat.c, config/m68k/tm-m68k.h, i960-tdep.c
config/i960/tm-i960.h, remote-nindy.c, config/m88k/tm-m88k.h,
m88k-tdep.c: Use floatformat.h instead of ieee-float.h.
* sparc-tdep.c: Remove now-obsolete ieee-float.h stuff
* findvar.c: Update comment regarding ieee-float.h.
|
|
action in case where caller specified no argument relocation.
(hppa_elf_build_linker_stub): Try again to get the sym_ptr_ptr
right in the original relocation and the stub's relocation.
|
|
works again. Attach some output symbols to the stub file bfd.
|
|
unused symbols argument.
* elf32-hppa.c (hppa_elf_stub_reloc): Accept asymbol ** rather
than asymbol * for original target symbol. All callers changed.
Set reloc->sym_ptr_ptr appropriately.
(hppa_elf_build_linker_stub): Set reloc->sym_ptr_ptr correctly.
(hppa_elf_look_for_stubs_in_section): No longer need symbols
argument. Use the output symbols when canonicalizing the relocs,
creating them if necessary.
* linker.c (_bfd_generic_link_output_symbols): Do not rebuild/clobber
the output symbols if they already exist.
|
|
(file_chain): Add decl.
(hppa_look_for_stubs_in_section): Delete decl.
(hppaelf_finish): Reenable code. Do not pass symbols
down to hppa_look_for_stubs_in_section.
|
|
|
|
|
|
* mpw-configure (prefix, mpw_prefix): Handle it.
(ARCHDEFS): Add to makefile editing.
(TDEPFILES, TM_FILE, XDEPFILES, XM_FILE, NAT_FILE): Add to
makefile if defined.
(mpw-edit-prefix): Use during makefile editing.
* mpw-config.in (mmalloc, readline): Don't configure.
* mpw-build.in (thisscript): Rename to ThisScript.
Use mpw-build instead of BuildProgram everywhere.
(mmalloc, readline): Don't build.
* mpw-README: New file, basic documentation about the MPW port.
|
|
* mpw-make.in (DEBUG_BFD_SEND): Define in default compilation
rule.
|
|
* targets.c (BFD_SEND, BFD_SEND_FMT): Add debugging versions that
check all the pointer dereferences. Enabled via DEBUG_BFD_SEND.
* bfd-in2.h: Rebuilt.
|
|
* srec.c (hex_value): Always set to a size of 256 bytes.
(srec_init): Cosmetic changes.
|
|
FRAME_CHAIN, FRAMELESS_FUNCTION_INVOCATION, FRAME_SAVED_PC):
Define to make backtracing through the various sigtramp handlers
work.
|
|
FRAME_CHAIN, FRAMELESS_FUNCTION_INVOCATION, FRAME_SAVED_PC):
Define to make backtracing through the various sigtramp handlers
work.
* i386-tdep.c (i386v4_sigtramp_saved_pc): New routine to fetch
the saved pc from ucontext on the stack for SVR4 signal handling.
|
|
PARAMize and staticize rest of functions. Delete unused
functions. Delete unused/unnecessary arguments to some functions.
Group static vars together. Abort for bad errors until we have
error code propogation working. Work on spacing and indention.
Add FIXMEs for unresolved problems. Use enums rather than
#defines for lots of things. Merge two functions which build
linker stubs into a single function (so they can easily share a
ton of common code).
|
|
|
|
|
|
|
|
* mpw-make (install): Replace with MPW-specific actions.
* mpw-build.in: Remove, functionality subsumed in mpw-make.in.
|
|
* ar.c (do_quick_append): Pad with a genuine character 10,
rather than whatever '\n' might happen to be.
|
|
* mpw-make.in (install): Moved here from mpw-build.in.
* mpw-build.in: Removed, functionality in mpw-make.in
|
|
Just use map_file or map_filename instead.
(add_archive_element): Use minfo to write map information, not
info_msg.
(constructor_callback): Use fprintf to write map information, not
info_msg.
* ldmain.h (write_map): Don't declare.
* ldgram.y (mri_script_command): Removed reference to write_map.
* ldlang.c (lang_one_common): Likewise.
* lexsup.c (parse_args): Likewise.
|
|
|
|
|
|
* ecoff.c, ecofflink.c, ecoffswap.h, srec.c: Likewise.
* elf32-i386.c: Likewise.
* elfcode.h (bfd_section_from_shdr): Make i unsigned; remove old
#if 0 code.
(elf_write_phdrs): Make i unsigned.
(map_program_segments): Make i and n_left unsigned.
(assign_file_positions_except_relocs): Make i unsigned.
(write_shdrs_and_ehdr): Make count unsigned.
(assign_file_positions_for_relocs): Make i unsigned.
(NAME(bfd,elf_write_object_contents)): Make count unsigned.
(section_from_elf_index): Make index argument unsigned.
|
|
avoid depending upon right shifts of signed numbers, and to
correct handling of src_mask with lower bits zero.
|
|
|
|
|
|
break in switch.
|
|
(swap_out_syms): Use BFD_ASSERT rather than assert.
|
|
declarations.
|
|
|