* i386-tdep.c (i386_register_virtual_type): Return
builtin_type_vec128i for SSE registers.
* gdbtypes.h (builtin_type_vec128i): Declare.
* gdbtypes.c (build_builtin_type_vec128i): New function.
(builtin_type_v2_double, builtin_type_v4_int64): New types.
(builtin_type_vec128i): New type for SSE2 128-bit registers.
(build_gdbtypes): Initialize new builtin vector types.
(_initialize_gdbtypes): Register new vector types with gdbarch.
(struct gdbarch_tdep): Replace os_ident member with abi.
(i386_gdbarch_register_os_abi): New prototype.
* i386-tdep.c (i386_abi_names): New array.
(process_note_abi_tag_sections): Removed.
(process_note_sections): New function.
(i386_elf_abi_from_note, i386_elf_abi): New functions.
(struct i386_abi_handler): New struct.
(i386_abi_handler_list): New variable.
(i386_gdbarch_register_os_abi): New function.
(i386_gdbarch_init): Adapt for the changes given above.
* i386-tdep.c: Include "elf-bfd.h".
(process_note_abi_tag_sections): New function.
(i386_gdbarch_init): Add code to recognize various OS/ABI
combinations.
FCTRL_REGNUM, FPC_REGNUM, FSTAT_REGNUM, FTAG_REGNUM, FCS_REGNUM,
FCOFF_REGNUM, FDS_REGNUM, FDOFF_REGNUM, FOP_REGNUM,
LAST_FPU_CTRL_REGNUM, XMM0_REGNUM, XMM7_REGNUM, MXCSR_REGNUM,
IS_FP_REGNUM, IS_SSE_REGNUM): Removed.
(FP0_REGNUM): Define conditionally depending on HAVE_I387_REGS.
(SIZEOF_FPU_CTRL_REGS): Hardcode value.
* i386-tdep.h (struct gdbarch_tdep): Change such that it contains
a single member `num_xmm_regs'.
(FPC_REGNUM): New macro.
(FIRST_FPU_REGNUM, LAST_FPU_REGNUM, FISRT_XMM_REGNUM,
LAST_XMM_REGNUM, MXCSR_REGNUM, FIRST_FPU_CTRL_REGNUM,
LAST_FPU_CTRL_REGNUM): Removed.
(FCTRL_REGNUM, FSTAT_REGNUM, FTAG_REGNUM, FOP_REGNUM, XMM0_REGNUM,
MXCSR_REGNUM): Define unconditionally. Change macros to match the
comment describing the register layout.
(FISEG_REGNUM, FIOFF_REGNUM, FOSEG_REGNUM, FOOFF_REGNUM): New macros.
(FP_REGNUM_P, FPC_REGNUM_P, SSE_REGNUM_P): New macros.
(IS_FP_REGNUM, IS_FPU_CTRL_REGNUM, IS_SSE_REGNUM): Make obsolete,
unconditionally define in terms of FP_REGNUM_P, FPC_REGNUM_P and
SSE_REGNUM_P).
(FCS_REGNUM, FCOFF_REGNUM, FDS_REGNUM, FDOFF_REGNUM): Make
obsolete, unconditionally define in terms of FISEG_REGNUM,
FIOFF_REGNUM, FOSEG_REGNUM, FOOFF_REGNUM.
* i386-tdep.c (i386_gdbarch_init): Initialize `num_xmm_regs'
member of `struct gdbarch_tdep'.
* x86-64-tdep.c (i386_gdbarch_init): Change initialization of
`struct gdbarch_tdep'.
* i387-nat.c (FCS_REGNUM, FCOFF_REGNUM, FDS_REGNUM, FDOFF_REGNUM):
Replace with FISEG_REGNUM, FIOFF_REGNUM, FOSEG_REGNUM and
FOOFF_REGNUM. Use FPC_REGNUM instead of FIRST_FPU_CTRL_REGNUM.
Use XMM0_REGNUM instead of LAST_FPU_CTRL_REGNUM.
i386_register_byte. Made static.
(i386_register_size): Renamed from i386_register_raw_size. Made
static.
(i386_register_virtual_size): Removed.
(i386_register_byte, i386_register_raw_size,
i386_register_virtual_size): New functions.
(_initialize_i386_tdep): Initialize i386_register_offset instead
of i386_register_byte. Remove code to initialize
i386_register_virtual_size.
* config/i386/tm-i386.h (REGISTER_BYTE): Redefine to use
i386_register_byte function.
(REGISTER_RAW_SIZE): Redefine to use i386_register_raw_size
function.
(REGISTER_VIRTUAL_SIZE): Redefine to use
i386_register_virtual_size function.
(i386_register_byte, i386_register_raw_size,
i386_register_virtual_size): New functions.
(REGISTER_NAME): New define.
(i386_register_name): New prototype.
* i386-tdep.c (i386_register_names): New variable.
(i386_register_name): New function.
* config/i386/tm-i386os9k.h, config/i386/tm-ptx.h,
config/i386/tm-symmetry.h: Undefine REGISTER_NAME instead of
REGISTER_NAMES.
convert_typed_floating to extract floating-point value from
register buffer.
(i386_store_return_value): Use convert_typed_floating to store
floating-point return balue in register buffer
(i386_register_virtual_type): Change type of floating-point
registers to builtin_type_i387_ext.
(i386_register_convert_to_virtual): Use convert_typed_floating to
do the conversion.
(i386_register_convert_to_raw): Likewise.
assertion with a warning if we're asked to convert towards a
non-floating-point type. Zero out the the buffer where the data
is supposed to be stored in that case.
(i386_register_convert_to_virtual): Fix such that it can handle
conversion to any floating-point type. Assert that we are dealing
with a floating-point first.
(i386_register_convert_to_raw): Assert that TYPE is a
floating-point type with length 12.
(i386_register_convertible): New function.
* config/i386/tm-i386.h (REGISTER_VIRTUAL_TYPE): Redefine in terms
of i386_register_virtual_type.
(REGISTER_CONVERTIBLE): Redefine in terms of
i386_register_convertible.
(i386_register_virtual_type, i386_register_convertible): New
prototypes.
* config/i386/tm-i386.h (FRAMELESS_FUNCTION_INVOCATION): Redefine
in terms of i386_frameless_function_invocation. Adjust comment.
(i386_frameless_function_invocation): New prototype.
* config/i386/tm-i386.h (EXTRACT_STRUCT_VALUE_ADDRESS): Redefine
in terms of i386_extract_struct_value_address.
(i386_extract_struct_value_address): New prototype.
return value is TYPE_STRUCT and the number of fields is one, call
ourselves with TYPE set tp the type of the first field.
(i386_store_return_value): Likewise.
This fixes a problem with returning structs consisting of a single
`float' or `double' on *BSD.
functions.
* config/i386/tm-i386.h (PUSH_ARGUMENTS): New macro.
(STORE_STRUCT_RETURN): Redefine in terms of
i386_store_struct_return.
(i386_push_arguments, i386_store_struct_return): New prototypes.
* config/i386/tm-i386v.h (STORE_STRUCT_RETURN): Remove. It's
definition was identical to the definition in "i386/tm-i386.h" so
the new definition should suffice too.
(i386_extract_return_value): Return directly after issuing the
warning and filling *VALBUF with zeroes if we cannot get at the
floating-point registers.
(i386_store_return_value): New function.
* config/i386/tm-i386.h (STORE_RETURN_VALUE): Simply call
i386_store_return_value.
Add prototype for i386_store_return_value.
(FIX_CALL_DUMMY): Redefined to call i386_fix_call_dummy.
(i386_fix_call_dummy): Add prototype.
* i386-tdep.c (i386_fix_call_dummy): New function based on the
code from the old FIX_CALL_DUMMY macro.
arch-utils.{h,c}. The ``set architecutre'' and ``set endian''
commands (part of the move) were implemented to use add_set_enum_cmd()
so that ``set architecture <tab>'' works.
* config/i386/tm-i386aix.h (I386_AIX_TARGET): Remove.
* config/i386/tm-linux.h (LOW_RETURN_REGNUM, HIGH_RETURN_REGNUM):
Remove
* i386-tdep.c (LOW_RETURN_REGNUM, HIGH_RETURN_REGNUM): New defines.
(i386_extract_return_value): Rewritten. Correctly support all
floating-point types and large integer types on targets that use
the standard i386 GDB register layout and return floating-point
values in the FPU.
LINUX_SIGTRAMP_INSN1, LINUX_SIGTRAMP_OFFSET1,
LINUX_SIGTRAMP_INSN2, LINUX_SIGTRAMP_OFFSET2, linux_sigtramp_code,
LINUX_SIGTRAMP_LEN, i386_linux_sigtramp_start,
LINUX_RT_SIGTRAMP_INSN0, LINUX_RT_SIGTRAMP_OFFSET0,
LINUX_RT_SIGTRAMP_INSN1, LINUX_RT_SIGTRAMP_OFFSET1,
linux_rt_sigtramp_code, LINUX_RT_SIGTRAMP_LEN,
i386_linux_rt_sigtramp_start, i386_linux_in_sigtramp,
i386_linux_sigcontext_addr, LINUX_SIGCONTEXT_PC_OFFSET,
i386_linux_sigtramp_saved_pc, LINUX_SIGCONTEXT_SP_OFFSET,
i386_linux_sigtramp_saved_sp): Deleted. These all implement
Linux-specific signal trampoline detection, and should be moved to...
* i386-linux-nat.c: ... here.
* config/i386/tm-linux.h (I386_LINUX_SIGTRAMP): No need to define
this any more, since we're not enabling OS-specific code in a
OS-independent file.
Martin Hunt
Mark Kettenis
Andrew Cagney
Peter Schauer
Pierre Muller
Kevin Buettner
Jim Blandy