/* Target-dependent code for GDB, the GNU debugger.
Copyright (C) 1986, 1987, 1989, 1991, 1992, 1993, 1994, 1995, 1996,
- 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004, 2005, 2006
+ 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007
Free Software Foundation, Inc.
This file is part of GDB.
return sp;
}
-/* PowerOpen always puts structures in memory. Vectors, which were
- added later, do get returned in a register though. */
-
-static int
-rs6000_use_struct_convention (int gcc_p, struct type *value_type)
-{
- if ((TYPE_LENGTH (value_type) == 16 || TYPE_LENGTH (value_type) == 8)
- && TYPE_VECTOR (value_type))
- return 0;
- return 1;
-}
-
-static void
-rs6000_extract_return_value (struct type *valtype, gdb_byte *regbuf,
- gdb_byte *valbuf)
+static enum return_value_convention
+rs6000_return_value (struct gdbarch *gdbarch, struct type *valtype,
+ struct regcache *regcache, gdb_byte *readbuf,
+ const gdb_byte *writebuf)
{
- int offset = 0;
struct gdbarch_tdep *tdep = gdbarch_tdep (current_gdbarch);
+ gdb_byte buf[8];
/* The calling convention this function implements assumes the
processor has floating-point registers. We shouldn't be using it
- on PPC variants that lack them. */
+ on PowerPC variants that lack them. */
gdb_assert (ppc_floating_point_unit_p (current_gdbarch));
- if (TYPE_CODE (valtype) == TYPE_CODE_FLT)
+ /* AltiVec extension: Functions that declare a vector data type as a
+ return value place that return value in VR2. */
+ if (TYPE_CODE (valtype) == TYPE_CODE_ARRAY && TYPE_VECTOR (valtype)
+ && TYPE_LENGTH (valtype) == 16)
{
+ if (readbuf)
+ regcache_cooked_read (regcache, tdep->ppc_vr0_regnum + 2, readbuf);
+ if (writebuf)
+ regcache_cooked_write (regcache, tdep->ppc_vr0_regnum + 2, writebuf);
- /* floats and doubles are returned in fpr1. fpr's have a size of 8 bytes.
- We need to truncate the return value into float size (4 byte) if
- necessary. */
-
- convert_typed_floating (®buf[DEPRECATED_REGISTER_BYTE
- (tdep->ppc_fp0_regnum + 1)],
- builtin_type_double,
- valbuf,
- valtype);
+ return RETURN_VALUE_REGISTER_CONVENTION;
}
- else if (TYPE_CODE (valtype) == TYPE_CODE_ARRAY
- && TYPE_LENGTH (valtype) == 16
- && TYPE_VECTOR (valtype))
+
+ /* If the called subprogram returns an aggregate, there exists an
+ implicit first argument, whose value is the address of a caller-
+ allocated buffer into which the callee is assumed to store its
+ return value. All explicit parameters are appropriately
+ relabeled. */
+ if (TYPE_CODE (valtype) == TYPE_CODE_STRUCT
+ || TYPE_CODE (valtype) == TYPE_CODE_UNION
+ || TYPE_CODE (valtype) == TYPE_CODE_ARRAY)
+ return RETURN_VALUE_STRUCT_CONVENTION;
+
+ /* Scalar floating-point values are returned in FPR1 for float or
+ double, and in FPR1:FPR2 for quadword precision. Fortran
+ complex*8 and complex*16 are returned in FPR1:FPR2, and
+ complex*32 is returned in FPR1:FPR4. */
+ if (TYPE_CODE (valtype) == TYPE_CODE_FLT
+ && (TYPE_LENGTH (valtype) == 4 || TYPE_LENGTH (valtype) == 8))
+ {
+ struct type *regtype = register_type (gdbarch, tdep->ppc_fp0_regnum);
+ gdb_byte regval[8];
+
+ /* FIXME: kettenis/2007-01-01: Add support for quadword
+ precision and complex. */
+
+ if (readbuf)
+ {
+ regcache_cooked_read (regcache, tdep->ppc_fp0_regnum + 1, regval);
+ convert_typed_floating (regval, regtype, readbuf, valtype);
+ }
+ if (writebuf)
+ {
+ convert_typed_floating (writebuf, valtype, regval, regtype);
+ regcache_cooked_write (regcache, tdep->ppc_fp0_regnum + 1, regval);
+ }
+
+ return RETURN_VALUE_REGISTER_CONVENTION;
+ }
+
+ /* Values of the types int, long, short, pointer, and char (length
+ is less than or equal to four bytes), as well as bit values of
+ lengths less than or equal to 32 bits, must be returned right
+ justified in GPR3 with signed values sign extended and unsigned
+ values zero extended, as necessary. */
+ if (TYPE_LENGTH (valtype) <= tdep->wordsize)
{
- memcpy (valbuf, regbuf + DEPRECATED_REGISTER_BYTE (tdep->ppc_vr0_regnum + 2),
- TYPE_LENGTH (valtype));
+ if (readbuf)
+ {
+ ULONGEST regval;
+
+ /* For reading we don't have to worry about sign extension. */
+ regcache_cooked_read_unsigned (regcache, tdep->ppc_gp0_regnum + 3,
+ ®val);
+ store_unsigned_integer (readbuf, TYPE_LENGTH (valtype), regval);
+ }
+ if (writebuf)
+ {
+ /* For writing, use unpack_long since that should handle any
+ required sign extension. */
+ regcache_cooked_write_unsigned (regcache, tdep->ppc_gp0_regnum + 3,
+ unpack_long (valtype, writebuf));
+ }
+
+ return RETURN_VALUE_REGISTER_CONVENTION;
}
- else
+
+ /* Eight-byte non-floating-point scalar values must be returned in
+ GPR3:GPR4. */
+
+ if (TYPE_LENGTH (valtype) == 8)
{
- /* return value is copied starting from r3. */
- if (TARGET_BYTE_ORDER == BFD_ENDIAN_BIG
- && TYPE_LENGTH (valtype) < register_size (current_gdbarch, 3))
- offset = register_size (current_gdbarch, 3) - TYPE_LENGTH (valtype);
-
- memcpy (valbuf,
- regbuf + DEPRECATED_REGISTER_BYTE (3) + offset,
- TYPE_LENGTH (valtype));
+ gdb_assert (TYPE_CODE (valtype) != TYPE_CODE_FLT);
+ gdb_assert (tdep->wordsize == 4);
+
+ if (readbuf)
+ {
+ gdb_byte regval[8];
+
+ regcache_cooked_read (regcache, tdep->ppc_gp0_regnum + 3, regval);
+ regcache_cooked_read (regcache, tdep->ppc_gp0_regnum + 4,
+ regval + 4);
+ memcpy (readbuf, regval, 8);
+ }
+ if (writebuf)
+ {
+ regcache_cooked_write (regcache, tdep->ppc_gp0_regnum + 3, writebuf);
+ regcache_cooked_write (regcache, tdep->ppc_gp0_regnum + 4,
+ writebuf + 4);
+ }
+
+ return RETURN_VALUE_REGISTER_CONVENTION;
}
+
+ return RETURN_VALUE_STRUCT_CONVENTION;
}
/* Return whether handle_inferior_event() should proceed through code
}
}
-
-static void
-rs6000_store_return_value (struct type *type,
- struct regcache *regcache,
- const gdb_byte *valbuf)
-{
- struct gdbarch *gdbarch = get_regcache_arch (regcache);
- struct gdbarch_tdep *tdep = gdbarch_tdep (gdbarch);
- int regnum = -1;
-
- /* The calling convention this function implements assumes the
- processor has floating-point registers. We shouldn't be using it
- on PPC variants that lack them. */
- gdb_assert (ppc_floating_point_unit_p (gdbarch));
-
- if (TYPE_CODE (type) == TYPE_CODE_FLT)
- /* Floating point values are returned starting from FPR1 and up.
- Say a double_double_double type could be returned in
- FPR1/FPR2/FPR3 triple. */
- regnum = tdep->ppc_fp0_regnum + 1;
- else if (TYPE_CODE (type) == TYPE_CODE_ARRAY)
- {
- if (TYPE_LENGTH (type) == 16
- && TYPE_VECTOR (type))
- regnum = tdep->ppc_vr0_regnum + 2;
- else
- internal_error (__FILE__, __LINE__,
- _("rs6000_store_return_value: "
- "unexpected array return type"));
- }
- else
- /* Everything else is returned in GPR3 and up. */
- regnum = tdep->ppc_gp0_regnum + 3;
-
- {
- size_t bytes_written = 0;
-
- while (bytes_written < TYPE_LENGTH (type))
- {
- /* How much of this value can we write to this register? */
- size_t bytes_to_write = min (TYPE_LENGTH (type) - bytes_written,
- register_size (gdbarch, regnum));
- regcache_cooked_write_part (regcache, regnum,
- 0, bytes_to_write,
- valbuf + bytes_written);
- regnum++;
- bytes_written += bytes_to_write;
- }
- }
-}
-
-
-/* Extract from an array REGBUF containing the (raw) register state
- the address in which a function should return its structure value,
- as a CORE_ADDR (or an expression that can be used as one). */
-
-static CORE_ADDR
-rs6000_extract_struct_value_address (struct regcache *regcache)
-{
- /* FIXME: cagney/2002-09-26: PR gdb/724: When making an inferior
- function call GDB knows the address of the struct return value
- and hence, should not need to call this function. Unfortunately,
- the current call_function_by_hand() code only saves the most
- recent struct address leading to occasional calls. The code
- should instead maintain a stack of such addresses (in the dummy
- frame object). */
- /* NOTE: cagney/2002-09-26: Return 0 which indicates that we've
- really got no idea where the return value is being stored. While
- r3, on function entry, contained the address it will have since
- been reused (scratch) and hence wouldn't be valid */
- return 0;
-}
-
/* Hook called when a new child process is started. */
void
else if (sysv_abi && wordsize == 4)
set_gdbarch_return_value (gdbarch, ppc_sysv_abi_return_value);
else
- {
- set_gdbarch_deprecated_extract_return_value (gdbarch, rs6000_extract_return_value);
- set_gdbarch_store_return_value (gdbarch, rs6000_store_return_value);
- }
+ set_gdbarch_return_value (gdbarch, rs6000_return_value);
/* Set lr_frame_offset. */
if (wordsize == 8)
set_gdbarch_stab_reg_to_regnum (gdbarch, rs6000_stab_reg_to_regnum);
set_gdbarch_dwarf2_reg_to_regnum (gdbarch, rs6000_dwarf2_reg_to_regnum);
- /* Note: kevinb/2002-04-12: I'm not convinced that rs6000_push_arguments()
- is correct for the SysV ABI when the wordsize is 8, but I'm also
- fairly certain that ppc_sysv_abi_push_arguments() will give even
- worse results since it only works for 32-bit code. So, for the moment,
- we're better off calling rs6000_push_arguments() since it works for
- 64-bit code. At some point in the future, this matter needs to be
- revisited. */
+
if (sysv_abi && wordsize == 4)
set_gdbarch_push_dummy_call (gdbarch, ppc_sysv_abi_push_dummy_call);
else if (sysv_abi && wordsize == 8)
else
set_gdbarch_push_dummy_call (gdbarch, rs6000_push_dummy_call);
- set_gdbarch_deprecated_extract_struct_value_address (gdbarch, rs6000_extract_struct_value_address);
-
set_gdbarch_skip_prologue (gdbarch, rs6000_skip_prologue);
set_gdbarch_in_function_epilogue_p (gdbarch, rs6000_in_function_epilogue_p);
/* Not sure on this. FIXMEmgo */
set_gdbarch_frame_args_skip (gdbarch, 8);
- if (!sysv_abi)
- set_gdbarch_deprecated_use_struct_convention (gdbarch, rs6000_use_struct_convention);
-
if (!sysv_abi)
{
/* Handle RS/6000 function pointers (which are really function
projects / binutils-gdb.git / commitdiff