f40751dd34
* fwprop.c (should_replace_address): Add speed attribute. (PR_OPTIMIZE_FOR_SPEED): New flag. (propagate_rtx_1): Use it. (propagate_rtx): Set it. (try_fwprop_subst): Update call of rtx_costs. (forward_propagate_and_simplify): LIkewise. * hooks.c (hook_int_rtx_bool_0): New (hook_bool_rtx_int_int_intp_false): Replace by ... (hook_bool_rtx_int_int_intp_bool_false): .. thisone. * hooks.h (hook_int_rtx_bool_0): New (hook_bool_rtx_int_int_intp_false): Replace by ... (hook_bool_rtx_int_int_intp_bool_false): .. thisone. * optabs.c (avoid_expensive_constant): UPdate call of rtx_cost. (prepare_cmp_insn): UPdate call of rtx_cost. * postreload.c (reload_cse_simplify_set): Update call of rtx_cost. (reload_cse_simplify_operands): Update call of rtx_cost. (reload_cse_move2add): call of rtx_cost. * target.h (struct gcc_target): Update rtx_costs and address_costs. * rtlanal.c (rtx_cost): Add speed argument. (address_cost): Add speed argument (default_address_cost): Likewise. (insn_rtx_cost): Likewise. * cfgloopanal.c (seq_cost): Add speed argument. (target_reg_cost, target_spill_cost): Turn to array. (init_set_costs): Update for speed. (estimate_reg_pressure_cost): Add speed argument. * auto-inc-dec.c (attempt_change): Update call of rtx_cost. * dojump.c (prefer_and_bit_test): UPdate call of rtx_cost. * tree-ssa-loop-ivopts.c (struct ivopts_data): New field speed. (seq_cost): Add speed argument. (computation_cost): Add speed arugment. (add_cost, multiply_by_const, get_address_cost): add speed argument. (force_expr_to_var_cost): Update for profile info. (force_var_cost): Likewise. (split_address_cost): Likewise. (ptr_difference_cost): Likewise. (difference_cost): Likewise. (get_computation_cost_at): Likewise. (determine_iv_cost): Likewise. (ivopts_global_cost_for_size): Likewise. (rewrite_use_address): Likewise. (tree_ssa_iv_optimize_loop): Initialize speed field. * cse.c (optimize_this_for_speed_p): New static var. (notreg_cost): Update call of rtx_cost. (cse_extended_basic_block): set optimize_this_for_speed_p. * ifcvt.c (cheap_bb_rtx_cost_p): Update call of rtx_cost. (noce_try_cmove_arith): Likewise. (noce_try_sign_mask): LIkewise. * expr.c (compress_float_constant): Update rtx_cost calls. * tree-ssa-address.c (most_expensive_mult_to_index): Add speed argument. (addr_to_parts): Likewise. (create_mem_ref): Likewise. * dse.c (find_shift_sequence): Add speed argument. (replace_read): Update call. * calls.c (precompute_register_parameters): Update call of rtx_cost. * expmed.c (sdiv_pow2_cheap, smod_pow2_cheap, zero_cost, add_cost, * neg_cost, shift_cost, shiftadd_cost, shiftsub_cost, mul_cost, sdiv_cost, udiv_cost ,mul_widen_cost, mul_highpart_cost): Increase dimension. (init_expmed): Initialize for both size and speed. (expand_shift): Use profile. (synth_mult): Use profile. (choose_mult_variant): Use profile. (expand_mult): Use profile. (expand_mult_highpart_optab): Use profile. (expand_mult_highpart): Use profile. (expand_smod_pow2): Use profile. (expand_divmod): Use profile. * simplify-rtx.c (simplify_binary_operation_1): Update call of rtx_cost. * loop-invariant.c (create_new_invariant): Use profile. (gain_for_invariant): Add speed parameter. (best_gain_for_invariant): Likewise. (find_invariants_to_move): Likewise. (move_single_loop_invariants): Set it. * target-def.h (TARGET_RTX_COSTS): Use hook. * rtl.h (rtx_cost, address_cost, insn_rtx_cost): Update prototpe. (optimize_insn_for_size_p, optimize_insn_for_speed_p): Declare. * output.h (default_address_cost): Update prototype. * combine.c (optimize_this_for_speed_p): New static var. (combine_validate_cost): Update call of rtx_cost. (combine_instructions): Set optimize_this_for_speed_p. (expand_compound_operation): Update call of rtx_cost. (make_extraction):Update call of rtx_cost. (force_to_mode):Update call of rtx_cost. (distribute_and_simplify_rtx):Update call of rtx_cost. * cfgloop.h (target_reg_cost, target_spill_cost): Turn to array. (estimate_reg_pressure_cost): Update prototype. * tree-flow.h (multiply_by_cost, create_mem_ref): Update prototype. * basic-block.h (optimize_insn_for_size_p, optimize_insn_for_speed_p): Remove. * config/alpha/alpha.c (alpha_rtx_costs): Update. (alpha_rtx_costs): Update. * config/frv/frv.c (frv_rtx_costs): Update. * config/s390/s390.c (s390_rtx_costs): Update. * config/m32c/m32c.c (m32c_memory_move_cost): Update. (m32c_rtx_costs): Update. * config/spu/spu.c (TARGET_ADDRESS_COST): Upate. (spu_rtx_costs): Update. * config/sparc/sparc.c (sparc_rtx_costs): Update. * config/m32r/m32r.c (m32r_rtx_costs): Update. * config/i386/i386.c (:ix86_address_cost): Update. (ix86_rtx_costs): Update. * config/sh/sh.c (sh_rtx_costs, sh_address_cost): Update. * config/pdp11/pdp11.c (pdp11_rtx_costs): Update. * config/avr/avr.c (avr_rtx_costs, avr_address_cost): Update. * config/crx/crx.c (crx_address_cost): Update. * config/xtensa/xtensa.c (xtensa_rtx_costs): Update. * config/stormy16/stormy16.c (xstormy16_address_cost, xstormy16_rtx_costs): Update. * config/m68hc11/m68hc11.c (m68hc11_address_cost, m68hc11_rtx_costs): Update. * config/cris/cris.c (cris_rtx_costs, cris_address_cost): Update. * config/iq2000/iq2000.c (iq2000_rtx_costs, iq2000_address_cost): Update. * config/mn10300/mn10300.c (mn10300_address_cost, mn10300_rtx_costs): Update * config/ia64/ia64.c (ia64_rtx_costs): Update. * config/m68k/m68k.c (m68k_rtx_costs): Update. * config/rs6000/rs6000.c (rs6000_rtx_costs): Update. * config/arc/arc.c (arc_rtx_costs, arc_address_cost): Update. * config/mcore/mcore.c (TARGET_ADDRESS_COST): Update. (mcore_rtx_costs): update. * config/score/score3.c (score3_rtx_costs): Update. * config/score/score7.c (score7_rtx_costs): Update. * config/score/score3.h (score3_rtx_costs):Update. * config/score/score7.h (score7_rtx_costs): Update. * config/score/score.c (score_rtx_costs): Update. * config/arm/arm.c (arm_address_cost): Update. (arm_rtx_costs_1): Update. (arm_rtx_costs_1): Update. (arm_size_rtx_costs): Update. (arm_size_rtx_costs): Update. (arm_size_rtx_costs): Update. (arm_xscale_rtx_costs): Update. (arm_thumb_address_cost): Update. * config/pa/pa.c (hppa_address_cost): Update. * config/mips/mips.c (mips_rtx_costs): Update. * config/vax/vax.c (vax_address_cost): Update. * config/h8300/h8300.c (h8300_shift_costs): Update. (h8300_rtx_costs): Update. * config/v850/v850.c (TARGET_ADDRESS_COST): Update. (v850_rtx_costs): Update. * config/mmix/mmix.c (mmix_rtx_costs, mmix_rtx_costs): Update. * config/bfin/bfin.c (bfin_address_cost): Update. (bfin_rtx_costs): Update. * stmt.c (lshift_cheap_p): Update. From-SVN: r139821
816 lines
20 KiB
C
816 lines
20 KiB
C
/* Memory address lowering and addressing mode selection.
|
|
Copyright (C) 2004, 2006, 2007 Free Software Foundation, Inc.
|
|
|
|
This file is part of GCC.
|
|
|
|
GCC is free software; you can redistribute it and/or modify it
|
|
under the terms of the GNU General Public License as published by the
|
|
Free Software Foundation; either version 3, or (at your option) any
|
|
later version.
|
|
|
|
GCC is distributed in the hope that it will be useful, but WITHOUT
|
|
ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
|
|
FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
|
|
for more details.
|
|
|
|
You should have received a copy of the GNU General Public License
|
|
along with GCC; see the file COPYING3. If not see
|
|
<http://www.gnu.org/licenses/>. */
|
|
|
|
/* Utility functions for manipulation with TARGET_MEM_REFs -- tree expressions
|
|
that directly map to addressing modes of the target. */
|
|
|
|
#include "config.h"
|
|
#include "system.h"
|
|
#include "coretypes.h"
|
|
#include "tm.h"
|
|
#include "tree.h"
|
|
#include "rtl.h"
|
|
#include "tm_p.h"
|
|
#include "hard-reg-set.h"
|
|
#include "basic-block.h"
|
|
#include "output.h"
|
|
#include "diagnostic.h"
|
|
#include "tree-flow.h"
|
|
#include "tree-dump.h"
|
|
#include "tree-pass.h"
|
|
#include "timevar.h"
|
|
#include "flags.h"
|
|
#include "tree-inline.h"
|
|
#include "insn-config.h"
|
|
#include "recog.h"
|
|
#include "expr.h"
|
|
#include "ggc.h"
|
|
#include "tree-affine.h"
|
|
|
|
/* TODO -- handling of symbols (according to Richard Hendersons
|
|
comments, http://gcc.gnu.org/ml/gcc-patches/2005-04/msg00949.html):
|
|
|
|
There are at least 5 different kinds of symbols that we can run up against:
|
|
|
|
(1) binds_local_p, small data area.
|
|
(2) binds_local_p, eg local statics
|
|
(3) !binds_local_p, eg global variables
|
|
(4) thread local, local_exec
|
|
(5) thread local, !local_exec
|
|
|
|
Now, (1) won't appear often in an array context, but it certainly can.
|
|
All you have to do is set -GN high enough, or explicitly mark any
|
|
random object __attribute__((section (".sdata"))).
|
|
|
|
All of these affect whether or not a symbol is in fact a valid address.
|
|
The only one tested here is (3). And that result may very well
|
|
be incorrect for (4) or (5).
|
|
|
|
An incorrect result here does not cause incorrect results out the
|
|
back end, because the expander in expr.c validizes the address. However
|
|
it would be nice to improve the handling here in order to produce more
|
|
precise results. */
|
|
|
|
/* A "template" for memory address, used to determine whether the address is
|
|
valid for mode. */
|
|
|
|
struct mem_addr_template GTY (())
|
|
{
|
|
rtx ref; /* The template. */
|
|
rtx * GTY ((skip)) step_p; /* The point in template where the step should be
|
|
filled in. */
|
|
rtx * GTY ((skip)) off_p; /* The point in template where the offset should
|
|
be filled in. */
|
|
};
|
|
|
|
/* The templates. Each of the five bits of the index corresponds to one
|
|
component of TARGET_MEM_REF being present, see TEMPL_IDX. */
|
|
|
|
static GTY (()) struct mem_addr_template templates[32];
|
|
|
|
#define TEMPL_IDX(SYMBOL, BASE, INDEX, STEP, OFFSET) \
|
|
(((SYMBOL != 0) << 4) \
|
|
| ((BASE != 0) << 3) \
|
|
| ((INDEX != 0) << 2) \
|
|
| ((STEP != 0) << 1) \
|
|
| (OFFSET != 0))
|
|
|
|
/* Stores address for memory reference with parameters SYMBOL, BASE, INDEX,
|
|
STEP and OFFSET to *ADDR. Stores pointers to where step is placed to
|
|
*STEP_P and offset to *OFFSET_P. */
|
|
|
|
static void
|
|
gen_addr_rtx (rtx symbol, rtx base, rtx index, rtx step, rtx offset,
|
|
rtx *addr, rtx **step_p, rtx **offset_p)
|
|
{
|
|
rtx act_elem;
|
|
|
|
*addr = NULL_RTX;
|
|
if (step_p)
|
|
*step_p = NULL;
|
|
if (offset_p)
|
|
*offset_p = NULL;
|
|
|
|
if (index)
|
|
{
|
|
act_elem = index;
|
|
if (step)
|
|
{
|
|
act_elem = gen_rtx_MULT (Pmode, act_elem, step);
|
|
|
|
if (step_p)
|
|
*step_p = &XEXP (act_elem, 1);
|
|
}
|
|
|
|
*addr = act_elem;
|
|
}
|
|
|
|
if (base)
|
|
{
|
|
if (*addr)
|
|
*addr = simplify_gen_binary (PLUS, Pmode, base, *addr);
|
|
else
|
|
*addr = base;
|
|
}
|
|
|
|
if (symbol)
|
|
{
|
|
act_elem = symbol;
|
|
if (offset)
|
|
{
|
|
act_elem = gen_rtx_PLUS (Pmode, act_elem, offset);
|
|
|
|
if (offset_p)
|
|
*offset_p = &XEXP (act_elem, 1);
|
|
|
|
if (GET_CODE (symbol) == SYMBOL_REF
|
|
|| GET_CODE (symbol) == LABEL_REF
|
|
|| GET_CODE (symbol) == CONST)
|
|
act_elem = gen_rtx_CONST (Pmode, act_elem);
|
|
}
|
|
|
|
if (*addr)
|
|
*addr = gen_rtx_PLUS (Pmode, *addr, act_elem);
|
|
else
|
|
*addr = act_elem;
|
|
}
|
|
else if (offset)
|
|
{
|
|
if (*addr)
|
|
{
|
|
*addr = gen_rtx_PLUS (Pmode, *addr, offset);
|
|
if (offset_p)
|
|
*offset_p = &XEXP (*addr, 1);
|
|
}
|
|
else
|
|
{
|
|
*addr = offset;
|
|
if (offset_p)
|
|
*offset_p = addr;
|
|
}
|
|
}
|
|
|
|
if (!*addr)
|
|
*addr = const0_rtx;
|
|
}
|
|
|
|
/* Returns address for TARGET_MEM_REF with parameters given by ADDR.
|
|
If REALLY_EXPAND is false, just make fake registers instead
|
|
of really expanding the operands, and perform the expansion in-place
|
|
by using one of the "templates". */
|
|
|
|
rtx
|
|
addr_for_mem_ref (struct mem_address *addr, bool really_expand)
|
|
{
|
|
rtx address, sym, bse, idx, st, off;
|
|
static bool templates_initialized = false;
|
|
struct mem_addr_template *templ;
|
|
|
|
if (addr->step && !integer_onep (addr->step))
|
|
st = immed_double_const (TREE_INT_CST_LOW (addr->step),
|
|
TREE_INT_CST_HIGH (addr->step), Pmode);
|
|
else
|
|
st = NULL_RTX;
|
|
|
|
if (addr->offset && !integer_zerop (addr->offset))
|
|
off = immed_double_const (TREE_INT_CST_LOW (addr->offset),
|
|
TREE_INT_CST_HIGH (addr->offset), Pmode);
|
|
else
|
|
off = NULL_RTX;
|
|
|
|
if (!really_expand)
|
|
{
|
|
/* Reuse the templates for addresses, so that we do not waste memory. */
|
|
if (!templates_initialized)
|
|
{
|
|
unsigned i;
|
|
|
|
templates_initialized = true;
|
|
sym = gen_rtx_SYMBOL_REF (Pmode, ggc_strdup ("test_symbol"));
|
|
bse = gen_raw_REG (Pmode, LAST_VIRTUAL_REGISTER + 1);
|
|
idx = gen_raw_REG (Pmode, LAST_VIRTUAL_REGISTER + 2);
|
|
|
|
for (i = 0; i < 32; i++)
|
|
gen_addr_rtx ((i & 16 ? sym : NULL_RTX),
|
|
(i & 8 ? bse : NULL_RTX),
|
|
(i & 4 ? idx : NULL_RTX),
|
|
(i & 2 ? const0_rtx : NULL_RTX),
|
|
(i & 1 ? const0_rtx : NULL_RTX),
|
|
&templates[i].ref,
|
|
&templates[i].step_p,
|
|
&templates[i].off_p);
|
|
}
|
|
|
|
templ = templates + TEMPL_IDX (addr->symbol, addr->base, addr->index,
|
|
st, off);
|
|
if (st)
|
|
*templ->step_p = st;
|
|
if (off)
|
|
*templ->off_p = off;
|
|
|
|
return templ->ref;
|
|
}
|
|
|
|
/* Otherwise really expand the expressions. */
|
|
sym = (addr->symbol
|
|
? expand_expr (build_addr (addr->symbol, current_function_decl),
|
|
NULL_RTX, Pmode, EXPAND_NORMAL)
|
|
: NULL_RTX);
|
|
bse = (addr->base
|
|
? expand_expr (addr->base, NULL_RTX, Pmode, EXPAND_NORMAL)
|
|
: NULL_RTX);
|
|
idx = (addr->index
|
|
? expand_expr (addr->index, NULL_RTX, Pmode, EXPAND_NORMAL)
|
|
: NULL_RTX);
|
|
|
|
gen_addr_rtx (sym, bse, idx, st, off, &address, NULL, NULL);
|
|
return address;
|
|
}
|
|
|
|
/* Returns address of MEM_REF in TYPE. */
|
|
|
|
tree
|
|
tree_mem_ref_addr (tree type, tree mem_ref)
|
|
{
|
|
tree addr;
|
|
tree act_elem;
|
|
tree step = TMR_STEP (mem_ref), offset = TMR_OFFSET (mem_ref);
|
|
tree sym = TMR_SYMBOL (mem_ref), base = TMR_BASE (mem_ref);
|
|
tree addr_base = NULL_TREE, addr_off = NULL_TREE;
|
|
|
|
if (sym)
|
|
addr_base = fold_convert (type, build_addr (sym, current_function_decl));
|
|
else if (base && POINTER_TYPE_P (TREE_TYPE (base)))
|
|
{
|
|
addr_base = fold_convert (type, base);
|
|
base = NULL_TREE;
|
|
}
|
|
|
|
act_elem = TMR_INDEX (mem_ref);
|
|
if (act_elem)
|
|
{
|
|
if (step)
|
|
act_elem = fold_build2 (MULT_EXPR, sizetype, act_elem, step);
|
|
addr_off = act_elem;
|
|
}
|
|
|
|
act_elem = base;
|
|
if (act_elem)
|
|
{
|
|
if (addr_off)
|
|
addr_off = fold_build2 (PLUS_EXPR, sizetype, addr_off, act_elem);
|
|
else
|
|
addr_off = act_elem;
|
|
}
|
|
|
|
if (offset && !integer_zerop (offset))
|
|
{
|
|
if (addr_off)
|
|
addr_off = fold_build2 (PLUS_EXPR, sizetype, addr_off, offset);
|
|
else
|
|
addr_off = offset;
|
|
}
|
|
|
|
if (addr_off)
|
|
{
|
|
if (addr_base)
|
|
addr = fold_build2 (POINTER_PLUS_EXPR, type, addr_base, addr_off);
|
|
else
|
|
addr = fold_convert (type, addr_off);
|
|
}
|
|
else if (addr_base)
|
|
addr = addr_base;
|
|
else
|
|
addr = build_int_cst (type, 0);
|
|
|
|
return addr;
|
|
}
|
|
|
|
/* Returns true if a memory reference in MODE and with parameters given by
|
|
ADDR is valid on the current target. */
|
|
|
|
static bool
|
|
valid_mem_ref_p (enum machine_mode mode, struct mem_address *addr)
|
|
{
|
|
rtx address;
|
|
|
|
address = addr_for_mem_ref (addr, false);
|
|
if (!address)
|
|
return false;
|
|
|
|
return memory_address_p (mode, address);
|
|
}
|
|
|
|
/* Checks whether a TARGET_MEM_REF with type TYPE and parameters given by ADDR
|
|
is valid on the current target and if so, creates and returns the
|
|
TARGET_MEM_REF. */
|
|
|
|
static tree
|
|
create_mem_ref_raw (tree type, struct mem_address *addr)
|
|
{
|
|
if (!valid_mem_ref_p (TYPE_MODE (type), addr))
|
|
return NULL_TREE;
|
|
|
|
if (addr->step && integer_onep (addr->step))
|
|
addr->step = NULL_TREE;
|
|
|
|
if (addr->offset && integer_zerop (addr->offset))
|
|
addr->offset = NULL_TREE;
|
|
|
|
return build7 (TARGET_MEM_REF, type,
|
|
addr->symbol, addr->base, addr->index,
|
|
addr->step, addr->offset, NULL, NULL);
|
|
}
|
|
|
|
/* Returns true if OBJ is an object whose address is a link time constant. */
|
|
|
|
static bool
|
|
fixed_address_object_p (tree obj)
|
|
{
|
|
return (TREE_CODE (obj) == VAR_DECL
|
|
&& (TREE_STATIC (obj)
|
|
|| DECL_EXTERNAL (obj))
|
|
&& ! DECL_DLLIMPORT_P (obj));
|
|
}
|
|
|
|
/* If ADDR contains an address of object that is a link time constant,
|
|
move it to PARTS->symbol. */
|
|
|
|
static void
|
|
move_fixed_address_to_symbol (struct mem_address *parts, aff_tree *addr)
|
|
{
|
|
unsigned i;
|
|
tree val = NULL_TREE;
|
|
|
|
for (i = 0; i < addr->n; i++)
|
|
{
|
|
if (!double_int_one_p (addr->elts[i].coef))
|
|
continue;
|
|
|
|
val = addr->elts[i].val;
|
|
if (TREE_CODE (val) == ADDR_EXPR
|
|
&& fixed_address_object_p (TREE_OPERAND (val, 0)))
|
|
break;
|
|
}
|
|
|
|
if (i == addr->n)
|
|
return;
|
|
|
|
parts->symbol = TREE_OPERAND (val, 0);
|
|
aff_combination_remove_elt (addr, i);
|
|
}
|
|
|
|
/* If ADDR contains an address of a dereferenced pointer, move it to
|
|
PARTS->base. */
|
|
|
|
static void
|
|
move_pointer_to_base (struct mem_address *parts, aff_tree *addr)
|
|
{
|
|
unsigned i;
|
|
tree val = NULL_TREE;
|
|
|
|
for (i = 0; i < addr->n; i++)
|
|
{
|
|
if (!double_int_one_p (addr->elts[i].coef))
|
|
continue;
|
|
|
|
val = addr->elts[i].val;
|
|
if (POINTER_TYPE_P (TREE_TYPE (val)))
|
|
break;
|
|
}
|
|
|
|
if (i == addr->n)
|
|
return;
|
|
|
|
parts->base = val;
|
|
aff_combination_remove_elt (addr, i);
|
|
}
|
|
|
|
/* Adds ELT to PARTS. */
|
|
|
|
static void
|
|
add_to_parts (struct mem_address *parts, tree elt)
|
|
{
|
|
tree type;
|
|
|
|
if (!parts->index)
|
|
{
|
|
parts->index = fold_convert (sizetype, elt);
|
|
return;
|
|
}
|
|
|
|
if (!parts->base)
|
|
{
|
|
parts->base = elt;
|
|
return;
|
|
}
|
|
|
|
/* Add ELT to base. */
|
|
type = TREE_TYPE (parts->base);
|
|
if (POINTER_TYPE_P (type))
|
|
parts->base = fold_build2 (POINTER_PLUS_EXPR, type,
|
|
parts->base,
|
|
fold_convert (sizetype, elt));
|
|
else
|
|
parts->base = fold_build2 (PLUS_EXPR, type,
|
|
parts->base, elt);
|
|
}
|
|
|
|
/* Finds the most expensive multiplication in ADDR that can be
|
|
expressed in an addressing mode and move the corresponding
|
|
element(s) to PARTS. */
|
|
|
|
static void
|
|
most_expensive_mult_to_index (struct mem_address *parts, aff_tree *addr,
|
|
bool speed)
|
|
{
|
|
HOST_WIDE_INT coef;
|
|
double_int best_mult, amult, amult_neg;
|
|
unsigned best_mult_cost = 0, acost;
|
|
tree mult_elt = NULL_TREE, elt;
|
|
unsigned i, j;
|
|
enum tree_code op_code;
|
|
|
|
best_mult = double_int_zero;
|
|
for (i = 0; i < addr->n; i++)
|
|
{
|
|
if (!double_int_fits_in_shwi_p (addr->elts[i].coef))
|
|
continue;
|
|
|
|
/* FIXME: Should use the correct memory mode rather than Pmode. */
|
|
|
|
coef = double_int_to_shwi (addr->elts[i].coef);
|
|
if (coef == 1
|
|
|| !multiplier_allowed_in_address_p (coef, Pmode))
|
|
continue;
|
|
|
|
acost = multiply_by_cost (coef, Pmode, speed);
|
|
|
|
if (acost > best_mult_cost)
|
|
{
|
|
best_mult_cost = acost;
|
|
best_mult = addr->elts[i].coef;
|
|
}
|
|
}
|
|
|
|
if (!best_mult_cost)
|
|
return;
|
|
|
|
/* Collect elements multiplied by best_mult. */
|
|
for (i = j = 0; i < addr->n; i++)
|
|
{
|
|
amult = addr->elts[i].coef;
|
|
amult_neg = double_int_ext_for_comb (double_int_neg (amult), addr);
|
|
|
|
if (double_int_equal_p (amult, best_mult))
|
|
op_code = PLUS_EXPR;
|
|
else if (double_int_equal_p (amult_neg, best_mult))
|
|
op_code = MINUS_EXPR;
|
|
else
|
|
{
|
|
addr->elts[j] = addr->elts[i];
|
|
j++;
|
|
continue;
|
|
}
|
|
|
|
elt = fold_convert (sizetype, addr->elts[i].val);
|
|
if (mult_elt)
|
|
mult_elt = fold_build2 (op_code, sizetype, mult_elt, elt);
|
|
else if (op_code == PLUS_EXPR)
|
|
mult_elt = elt;
|
|
else
|
|
mult_elt = fold_build1 (NEGATE_EXPR, sizetype, elt);
|
|
}
|
|
addr->n = j;
|
|
|
|
parts->index = mult_elt;
|
|
parts->step = double_int_to_tree (sizetype, best_mult);
|
|
}
|
|
|
|
/* Splits address ADDR into PARTS.
|
|
|
|
TODO -- be more clever about the distribution of the elements of ADDR
|
|
to PARTS. Some architectures do not support anything but single
|
|
register in address, possibly with a small integer offset; while
|
|
create_mem_ref will simplify the address to an acceptable shape
|
|
later, it would be more efficient to know that asking for complicated
|
|
addressing modes is useless. */
|
|
|
|
static void
|
|
addr_to_parts (aff_tree *addr, struct mem_address *parts, bool speed)
|
|
{
|
|
tree part;
|
|
unsigned i;
|
|
|
|
parts->symbol = NULL_TREE;
|
|
parts->base = NULL_TREE;
|
|
parts->index = NULL_TREE;
|
|
parts->step = NULL_TREE;
|
|
|
|
if (!double_int_zero_p (addr->offset))
|
|
parts->offset = double_int_to_tree (sizetype, addr->offset);
|
|
else
|
|
parts->offset = NULL_TREE;
|
|
|
|
/* Try to find a symbol. */
|
|
move_fixed_address_to_symbol (parts, addr);
|
|
|
|
/* First move the most expensive feasible multiplication
|
|
to index. */
|
|
most_expensive_mult_to_index (parts, addr, speed);
|
|
|
|
/* Try to find a base of the reference. Since at the moment
|
|
there is no reliable way how to distinguish between pointer and its
|
|
offset, this is just a guess. */
|
|
if (!parts->symbol)
|
|
move_pointer_to_base (parts, addr);
|
|
|
|
/* Then try to process the remaining elements. */
|
|
for (i = 0; i < addr->n; i++)
|
|
{
|
|
part = fold_convert (sizetype, addr->elts[i].val);
|
|
if (!double_int_one_p (addr->elts[i].coef))
|
|
part = fold_build2 (MULT_EXPR, sizetype, part,
|
|
double_int_to_tree (sizetype, addr->elts[i].coef));
|
|
add_to_parts (parts, part);
|
|
}
|
|
if (addr->rest)
|
|
add_to_parts (parts, fold_convert (sizetype, addr->rest));
|
|
}
|
|
|
|
/* Force the PARTS to register. */
|
|
|
|
static void
|
|
gimplify_mem_ref_parts (gimple_stmt_iterator *gsi, struct mem_address *parts)
|
|
{
|
|
if (parts->base)
|
|
parts->base = force_gimple_operand_gsi (gsi, parts->base,
|
|
true, NULL_TREE,
|
|
true, GSI_SAME_STMT);
|
|
if (parts->index)
|
|
parts->index = force_gimple_operand_gsi (gsi, parts->index,
|
|
true, NULL_TREE,
|
|
true, GSI_SAME_STMT);
|
|
}
|
|
|
|
/* Creates and returns a TARGET_MEM_REF for address ADDR. If necessary
|
|
computations are emitted in front of GSI. TYPE is the mode
|
|
of created memory reference. */
|
|
|
|
tree
|
|
create_mem_ref (gimple_stmt_iterator *gsi, tree type, aff_tree *addr,
|
|
bool speed)
|
|
{
|
|
tree mem_ref, tmp;
|
|
tree atype;
|
|
struct mem_address parts;
|
|
|
|
addr_to_parts (addr, &parts, speed);
|
|
gimplify_mem_ref_parts (gsi, &parts);
|
|
mem_ref = create_mem_ref_raw (type, &parts);
|
|
if (mem_ref)
|
|
return mem_ref;
|
|
|
|
/* The expression is too complicated. Try making it simpler. */
|
|
|
|
if (parts.step && !integer_onep (parts.step))
|
|
{
|
|
/* Move the multiplication to index. */
|
|
gcc_assert (parts.index);
|
|
parts.index = force_gimple_operand_gsi (gsi,
|
|
fold_build2 (MULT_EXPR, sizetype,
|
|
parts.index, parts.step),
|
|
true, NULL_TREE, true, GSI_SAME_STMT);
|
|
parts.step = NULL_TREE;
|
|
|
|
mem_ref = create_mem_ref_raw (type, &parts);
|
|
if (mem_ref)
|
|
return mem_ref;
|
|
}
|
|
|
|
if (parts.symbol)
|
|
{
|
|
tmp = build_addr (parts.symbol, current_function_decl);
|
|
gcc_assert (is_gimple_val (tmp));
|
|
|
|
/* Add the symbol to base, eventually forcing it to register. */
|
|
if (parts.base)
|
|
{
|
|
gcc_assert (useless_type_conversion_p
|
|
(sizetype, TREE_TYPE (parts.base)));
|
|
|
|
if (parts.index)
|
|
{
|
|
atype = TREE_TYPE (tmp);
|
|
parts.base = force_gimple_operand_gsi (gsi,
|
|
fold_build2 (PLUS_EXPR, atype,
|
|
fold_convert (atype, parts.base),
|
|
tmp),
|
|
true, NULL_TREE, true, GSI_SAME_STMT);
|
|
}
|
|
else
|
|
{
|
|
parts.index = parts.base;
|
|
parts.base = tmp;
|
|
}
|
|
}
|
|
else
|
|
parts.base = tmp;
|
|
parts.symbol = NULL_TREE;
|
|
|
|
mem_ref = create_mem_ref_raw (type, &parts);
|
|
if (mem_ref)
|
|
return mem_ref;
|
|
}
|
|
|
|
if (parts.index)
|
|
{
|
|
/* Add index to base. */
|
|
if (parts.base)
|
|
{
|
|
atype = TREE_TYPE (parts.base);
|
|
parts.base = force_gimple_operand_gsi (gsi,
|
|
fold_build2 (POINTER_PLUS_EXPR, atype,
|
|
parts.base,
|
|
parts.index),
|
|
true, NULL_TREE, true, GSI_SAME_STMT);
|
|
}
|
|
else
|
|
parts.base = parts.index;
|
|
parts.index = NULL_TREE;
|
|
|
|
mem_ref = create_mem_ref_raw (type, &parts);
|
|
if (mem_ref)
|
|
return mem_ref;
|
|
}
|
|
|
|
if (parts.offset && !integer_zerop (parts.offset))
|
|
{
|
|
/* Try adding offset to base. */
|
|
if (parts.base)
|
|
{
|
|
atype = TREE_TYPE (parts.base);
|
|
parts.base = force_gimple_operand_gsi (gsi,
|
|
fold_build2 (POINTER_PLUS_EXPR, atype,
|
|
parts.base,
|
|
fold_convert (sizetype, parts.offset)),
|
|
true, NULL_TREE, true, GSI_SAME_STMT);
|
|
}
|
|
else
|
|
parts.base = parts.offset;
|
|
|
|
parts.offset = NULL_TREE;
|
|
|
|
mem_ref = create_mem_ref_raw (type, &parts);
|
|
if (mem_ref)
|
|
return mem_ref;
|
|
}
|
|
|
|
/* Verify that the address is in the simplest possible shape
|
|
(only a register). If we cannot create such a memory reference,
|
|
something is really wrong. */
|
|
gcc_assert (parts.symbol == NULL_TREE);
|
|
gcc_assert (parts.index == NULL_TREE);
|
|
gcc_assert (!parts.step || integer_onep (parts.step));
|
|
gcc_assert (!parts.offset || integer_zerop (parts.offset));
|
|
gcc_unreachable ();
|
|
}
|
|
|
|
/* Copies components of the address from OP to ADDR. */
|
|
|
|
void
|
|
get_address_description (tree op, struct mem_address *addr)
|
|
{
|
|
addr->symbol = TMR_SYMBOL (op);
|
|
addr->base = TMR_BASE (op);
|
|
addr->index = TMR_INDEX (op);
|
|
addr->step = TMR_STEP (op);
|
|
addr->offset = TMR_OFFSET (op);
|
|
}
|
|
|
|
/* Copies the additional information attached to target_mem_ref FROM to TO. */
|
|
|
|
void
|
|
copy_mem_ref_info (tree to, tree from)
|
|
{
|
|
/* Copy the annotation, to preserve the aliasing information. */
|
|
TMR_TAG (to) = TMR_TAG (from);
|
|
|
|
/* And the info about the original reference. */
|
|
TMR_ORIGINAL (to) = TMR_ORIGINAL (from);
|
|
}
|
|
|
|
/* Move constants in target_mem_ref REF to offset. Returns the new target
|
|
mem ref if anything changes, NULL_TREE otherwise. */
|
|
|
|
tree
|
|
maybe_fold_tmr (tree ref)
|
|
{
|
|
struct mem_address addr;
|
|
bool changed = false;
|
|
tree ret, off;
|
|
|
|
get_address_description (ref, &addr);
|
|
|
|
if (addr.base && TREE_CODE (addr.base) == INTEGER_CST)
|
|
{
|
|
if (addr.offset)
|
|
addr.offset = fold_binary_to_constant (PLUS_EXPR, sizetype,
|
|
addr.offset,
|
|
fold_convert (sizetype, addr.base));
|
|
else
|
|
addr.offset = addr.base;
|
|
|
|
addr.base = NULL_TREE;
|
|
changed = true;
|
|
}
|
|
|
|
if (addr.index && TREE_CODE (addr.index) == INTEGER_CST)
|
|
{
|
|
off = addr.index;
|
|
if (addr.step)
|
|
{
|
|
off = fold_binary_to_constant (MULT_EXPR, sizetype,
|
|
off, addr.step);
|
|
addr.step = NULL_TREE;
|
|
}
|
|
|
|
if (addr.offset)
|
|
{
|
|
addr.offset = fold_binary_to_constant (PLUS_EXPR, sizetype,
|
|
addr.offset, off);
|
|
}
|
|
else
|
|
addr.offset = off;
|
|
|
|
addr.index = NULL_TREE;
|
|
changed = true;
|
|
}
|
|
|
|
if (!changed)
|
|
return NULL_TREE;
|
|
|
|
ret = create_mem_ref_raw (TREE_TYPE (ref), &addr);
|
|
if (!ret)
|
|
return NULL_TREE;
|
|
|
|
copy_mem_ref_info (ret, ref);
|
|
return ret;
|
|
}
|
|
|
|
/* Dump PARTS to FILE. */
|
|
|
|
extern void dump_mem_address (FILE *, struct mem_address *);
|
|
void
|
|
dump_mem_address (FILE *file, struct mem_address *parts)
|
|
{
|
|
if (parts->symbol)
|
|
{
|
|
fprintf (file, "symbol: ");
|
|
print_generic_expr (file, parts->symbol, TDF_SLIM);
|
|
fprintf (file, "\n");
|
|
}
|
|
if (parts->base)
|
|
{
|
|
fprintf (file, "base: ");
|
|
print_generic_expr (file, parts->base, TDF_SLIM);
|
|
fprintf (file, "\n");
|
|
}
|
|
if (parts->index)
|
|
{
|
|
fprintf (file, "index: ");
|
|
print_generic_expr (file, parts->index, TDF_SLIM);
|
|
fprintf (file, "\n");
|
|
}
|
|
if (parts->step)
|
|
{
|
|
fprintf (file, "step: ");
|
|
print_generic_expr (file, parts->step, TDF_SLIM);
|
|
fprintf (file, "\n");
|
|
}
|
|
if (parts->offset)
|
|
{
|
|
fprintf (file, "offset: ");
|
|
print_generic_expr (file, parts->offset, TDF_SLIM);
|
|
fprintf (file, "\n");
|
|
}
|
|
}
|
|
|
|
#include "gt-tree-ssa-address.h"
|