#include "integrate.h"
#include "langhooks.h"
#include "target.h"
+#include "common/common-target.h"
#include "cfglayout.h"
#include "gimple.h"
#include "tree-pass.h"
#include "df.h"
#include "timevar.h"
#include "vecprim.h"
+#include "params.h"
+#include "bb-reorder.h"
/* So we can assign to cfun in this file. */
#undef cfun
can always export `prologue_epilogue_contains'. */
static void record_insns (rtx, rtx, htab_t *) ATTRIBUTE_UNUSED;
static bool contains (const_rtx, htab_t);
-#ifdef HAVE_return
-static void emit_return_into_block (basic_block);
-#endif
static void prepare_function_start (void);
static void do_clobber_return_reg (rtx, void *);
static void do_use_return_reg (rtx, void *);
prologue_insn_hash = NULL;
epilogue_insn_hash = NULL;
- if (crtl->emit.regno_pointer_align)
- free (crtl->emit.regno_pointer_align);
+ free (crtl->emit.regno_pointer_align);
memset (crtl, 0, sizeof (struct rtl_data));
f->eh = NULL;
frame_pointer_rtx. */
virtuals_instantiated = 1;
- /* See allocate_dynamic_stack_space for the rationale. */
-#ifdef SETJMP_VIA_SAVE_AREA
- if (flag_stack_usage && cfun->calls_setjmp)
- {
- int align = PREFERRED_STACK_BOUNDARY / BITS_PER_UNIT;
- dynamic_offset = (dynamic_offset + align - 1) / align * align;
- current_function_dynamic_stack_size
- += current_function_dynamic_alloc_count * dynamic_offset;
- }
-#endif
-
return 0;
}
0, /* properties_provided */
0, /* properties_destroyed */
0, /* todo_flags_start */
- TODO_dump_func /* todo_flags_finish */
+ 0 /* todo_flags_finish */
}
};
}
}
- return targetm.calls.pass_by_reference (ca, mode, type, named_arg);
+ return targetm.calls.pass_by_reference (pack_cumulative_args (ca), mode,
+ type, named_arg);
}
/* Return true if TYPE, which is passed by reference, should be callee
{
if (type && TREE_ADDRESSABLE (type))
return false;
- return targetm.calls.callee_copies (ca, mode, type, named_arg);
+ return targetm.calls.callee_copies (pack_cumulative_args (ca), mode, type,
+ named_arg);
}
/* Structures to communicate between the subroutines of assign_parms.
struct assign_parm_data_all
{
- CUMULATIVE_ARGS args_so_far;
+ /* When INIT_CUMULATIVE_ARGS gets revamped, allocating CUMULATIVE_ARGS
+ should become a job of the target or otherwise encapsulated. */
+ CUMULATIVE_ARGS args_so_far_v;
+ cumulative_args_t args_so_far;
struct args_size stack_args_size;
tree function_result_decl;
tree orig_fnargs;
fntype = TREE_TYPE (current_function_decl);
#ifdef INIT_CUMULATIVE_INCOMING_ARGS
- INIT_CUMULATIVE_INCOMING_ARGS (all->args_so_far, fntype, NULL_RTX);
+ INIT_CUMULATIVE_INCOMING_ARGS (all->args_so_far_v, fntype, NULL_RTX);
#else
- INIT_CUMULATIVE_ARGS (all->args_so_far, fntype, NULL_RTX,
+ INIT_CUMULATIVE_ARGS (all->args_so_far_v, fntype, NULL_RTX,
current_function_decl, -1);
#endif
+ all->args_so_far = pack_cumulative_args (&all->args_so_far_v);
#ifdef REG_PARM_STACK_SPACE
all->reg_parm_stack_space = REG_PARM_STACK_SPACE (current_function_decl);
data->named_arg = 1; /* No variadic parms. */
else if (DECL_CHAIN (parm))
data->named_arg = 1; /* Not the last non-variadic parm. */
- else if (targetm.calls.strict_argument_naming (&all->args_so_far))
+ else if (targetm.calls.strict_argument_naming (all->args_so_far))
data->named_arg = 1; /* Only variadic ones are unnamed. */
else
data->named_arg = 0; /* Treat as variadic. */
passed_type = TREE_TYPE (first_field (passed_type));
/* See if this arg was passed by invisible reference. */
- if (pass_by_reference (&all->args_so_far, passed_mode,
+ if (pass_by_reference (&all->args_so_far_v, passed_mode,
passed_type, data->named_arg))
{
passed_type = nominal_type = build_pointer_type (passed_type);
{
int varargs_pretend_bytes = 0;
- targetm.calls.setup_incoming_varargs (&all->args_so_far,
+ targetm.calls.setup_incoming_varargs (all->args_so_far,
data->promoted_mode,
data->passed_type,
&varargs_pretend_bytes, no_rtl);
return;
}
- entry_parm = targetm.calls.function_incoming_arg (&all->args_so_far,
+ entry_parm = targetm.calls.function_incoming_arg (all->args_so_far,
data->promoted_mode,
data->passed_type,
data->named_arg);
#endif
if (!in_regs && !data->named_arg)
{
- if (targetm.calls.pretend_outgoing_varargs_named (&all->args_so_far))
+ if (targetm.calls.pretend_outgoing_varargs_named (all->args_so_far))
{
rtx tem;
- tem = targetm.calls.function_incoming_arg (&all->args_so_far,
+ tem = targetm.calls.function_incoming_arg (all->args_so_far,
data->promoted_mode,
data->passed_type, true);
in_regs = tem != NULL;
{
int partial;
- partial = targetm.calls.arg_partial_bytes (&all->args_so_far,
+ partial = targetm.calls.arg_partial_bytes (all->args_so_far,
data->promoted_mode,
data->passed_type,
data->named_arg);
if (data->promoted_mode != BLKmode
&& data->promoted_mode != DECL_MODE (parm))
{
- set_mem_size (stack_parm,
- GEN_INT (GET_MODE_SIZE (data->promoted_mode)));
- if (MEM_EXPR (stack_parm) && MEM_OFFSET (stack_parm))
+ set_mem_size (stack_parm, GET_MODE_SIZE (data->promoted_mode));
+ if (MEM_EXPR (stack_parm) && MEM_OFFSET_KNOWN_P (stack_parm))
{
int offset = subreg_lowpart_offset (DECL_MODE (parm),
data->promoted_mode);
if (offset)
- set_mem_offset (stack_parm,
- plus_constant (MEM_OFFSET (stack_parm),
- -offset));
+ set_mem_offset (stack_parm, MEM_OFFSET (stack_parm) - offset);
}
}
}
int by = (UNITS_PER_WORD - size) * BITS_PER_UNIT;
rtx reg = gen_rtx_REG (word_mode, REGNO (entry_parm));
- x = expand_shift (LSHIFT_EXPR, word_mode, reg,
- build_int_cst (NULL_TREE, by),
- NULL_RTX, 1);
+ x = expand_shift (LSHIFT_EXPR, word_mode, reg, by, NULL_RTX, 1);
tem = change_address (mem, word_mode, 0);
emit_move_insn (tem, x);
}
SET_DECL_RTL (parm, stack_parm);
}
-/* A subroutine of assign_parm_setup_reg, called through note_stores.
- This collects sets and clobbers of hard registers in a HARD_REG_SET,
- which is pointed to by DATA. */
-static void
-record_hard_reg_sets (rtx x, const_rtx pat ATTRIBUTE_UNUSED, void *data)
-{
- HARD_REG_SET *pset = (HARD_REG_SET *)data;
- if (REG_P (x) && HARD_REGISTER_P (x))
- add_to_hard_reg_set (pset, GET_MODE (x), REGNO (x));
-}
-
/* A subroutine of assign_parms. Allocate a pseudo to hold the current
parameter. Get it there. Perform all ABI specified conversions. */
/* ??? This may need a big-endian conversion on sparc64. */
data->stack_parm
= adjust_address (data->stack_parm, data->nominal_mode, 0);
- if (offset && MEM_OFFSET (data->stack_parm))
+ if (offset && MEM_OFFSET_KNOWN_P (data->stack_parm))
set_mem_offset (data->stack_parm,
- plus_constant (MEM_OFFSET (data->stack_parm),
- offset));
+ MEM_OFFSET (data->stack_parm) + offset);
}
}
set_decl_incoming_rtl (parm, data.entry_parm, false);
/* Update info on where next arg arrives in registers. */
- targetm.calls.function_arg_advance (&all.args_so_far, data.promoted_mode,
+ targetm.calls.function_arg_advance (all.args_so_far, data.promoted_mode,
data.passed_type, data.named_arg);
assign_parm_adjust_stack_rtl (&data);
/* For stdarg.h function, save info about
regs and stack space used by the named args. */
- crtl->args.info = all.args_so_far;
+ crtl->args.info = all.args_so_far_v;
/* Set the rtx used for the function return value. Put this in its
own variable so any optimizers that need this information don't have
continue;
/* Update info on where next arg arrives in registers. */
- targetm.calls.function_arg_advance (&all.args_so_far, data.promoted_mode,
+ targetm.calls.function_arg_advance (all.args_so_far, data.promoted_mode,
data.passed_type, data.named_arg);
/* ??? Once upon a time variable_size stuffed parameter list
if (data.passed_pointer)
{
tree type = TREE_TYPE (data.passed_type);
- if (reference_callee_copied (&all.args_so_far, TYPE_MODE (type),
+ if (reference_callee_copied (&all.args_so_far_v, TYPE_MODE (type),
type, data.named_arg))
{
tree local, t;
&& compare_tree_int (DECL_SIZE_UNIT (parm),
STACK_CHECK_MAX_VAR_SIZE) > 0))
{
- local = create_tmp_reg (type, get_name (parm));
+ local = create_tmp_var (type, get_name (parm));
DECL_IGNORED_P (local) = 0;
/* If PARM was addressable, move that flag over
to the local copy, as its address will be taken,
as we'll query that flag during gimplification. */
if (TREE_ADDRESSABLE (parm))
TREE_ADDRESSABLE (local) = 1;
+ else if (TREE_CODE (type) == COMPLEX_TYPE
+ || TREE_CODE (type) == VECTOR_TYPE)
+ DECL_GIMPLE_REG_P (local) = 1;
}
else
{
DECL_IGNORED_P (addr) = 0;
local = build_fold_indirect_ref (addr);
- t = built_in_decls[BUILT_IN_ALLOCA];
- t = build_call_expr (t, 1, DECL_SIZE_UNIT (parm));
+ t = builtin_decl_explicit (BUILT_IN_ALLOCA_WITH_ALIGN);
+ t = build_call_expr (t, 2, DECL_SIZE_UNIT (parm),
+ size_int (DECL_ALIGN (parm)));
+
/* The call has been built for a variable-sized object. */
- ALLOCA_FOR_VAR_P (t) = 1;
+ CALL_ALLOCA_FOR_VAR_P (t) = 1;
t = fold_convert (ptr_type, t);
t = build2 (MODIFY_EXPR, TREE_TYPE (addr), addr, t);
gimplify_and_add (t, &stmts);
{
tree sizetree;
enum direction where_pad;
- unsigned int boundary;
+ unsigned int boundary, round_boundary;
int reg_parm_stack_space = 0;
int part_size_in_regs;
= type ? size_in_bytes (type) : size_int (GET_MODE_SIZE (passed_mode));
where_pad = FUNCTION_ARG_PADDING (passed_mode, type);
boundary = targetm.calls.function_arg_boundary (passed_mode, type);
+ round_boundary = targetm.calls.function_arg_round_boundary (passed_mode,
+ type);
locate->where_pad = where_pad;
/* Alignment can't exceed MAX_SUPPORTED_STACK_ALIGNMENT. */
tree s2 = sizetree;
if (where_pad != none
&& (!host_integerp (sizetree, 1)
- || (tree_low_cst (sizetree, 1) * BITS_PER_UNIT) % PARM_BOUNDARY))
- s2 = round_up (s2, PARM_BOUNDARY / BITS_PER_UNIT);
+ || (tree_low_cst (sizetree, 1) * BITS_PER_UNIT) % round_boundary))
+ s2 = round_up (s2, round_boundary / BITS_PER_UNIT);
SUB_PARM_SIZE (locate->slot_offset, s2);
}
if (where_pad != none
&& (!host_integerp (sizetree, 1)
- || (tree_low_cst (sizetree, 1) * BITS_PER_UNIT) % PARM_BOUNDARY))
- sizetree = round_up (sizetree, PARM_BOUNDARY / BITS_PER_UNIT);
+ || (tree_low_cst (sizetree, 1) * BITS_PER_UNIT) % round_boundary))
+ sizetree = round_up (sizetree, round_boundary / BITS_PER_UNIT);
ADD_PARM_SIZE (locate->size, sizetree);
return funcdef_no++;
}
+/* Return value of funcdef. */
+int
+get_last_funcdef_no (void)
+{
+ return funcdef_no;
+}
+
/* Allocate a function structure for FNDECL and set its contents
to the defaults. Set cfun to the newly-allocated object.
Some of the helper functions invoked during initialization assume
init_expr ();
default_rtl_profile ();
- if (flag_stack_usage)
+ if (flag_stack_usage_info)
{
cfun->su = ggc_alloc_cleared_stack_usage ();
cfun->su->static_stack_size = -1;
else
allocate_struct_function (subr, false);
prepare_function_start ();
+ decide_function_section (subr);
/* Warn if this value is an aggregate type,
regardless of which calling convention we are using for it. */
if (!DECL_RTL_SET_P (var))
expand_decl (var);
- t_save = build4 (ARRAY_REF, ptr_type_node,
+ t_save = build4 (ARRAY_REF,
+ TREE_TYPE (TREE_TYPE (cfun->nonlocal_goto_save_area)),
cfun->nonlocal_goto_save_area,
integer_zero_node, NULL_TREE, NULL_TREE);
r_save = expand_expr (t_save, NULL_RTX, VOIDmode, EXPAND_WRITE);
- r_save = convert_memory_address (Pmode, r_save);
+ gcc_assert (GET_MODE (r_save) == Pmode);
emit_move_insn (r_save, targetm.builtin_setjmp_frame_value ());
update_nonlocal_goto_save_area ();
#endif
}
- /* After the display initializations is where the stack checking
- probe should go. */
- if(flag_stack_check)
+ /* If we are doing generic stack checking, the probe should go here. */
+ if (flag_stack_check == GENERIC_STACK_CHECK)
stack_check_probe_note = emit_note (NOTE_INSN_DELETED);
/* Make sure there is a line number after the function entry setup code. */
/* Output the label for the actual return from the function. */
emit_label (return_label);
- if (targetm.except_unwind_info (&global_options) == UI_SJLJ)
+ if (targetm_common.except_unwind_info (&global_options) == UI_SJLJ)
{
/* Let except.c know where it should emit the call to unregister
the function context for sjlj exceptions. */
may trap are not moved into the epilogue by scheduling, because
we don't always emit unwind information for the epilogue. */
if (cfun->can_throw_non_call_exceptions
- && targetm.except_unwind_info (&global_options) != UI_SJLJ)
+ && targetm_common.except_unwind_info (&global_options) != UI_SJLJ)
emit_insn (gen_blockage ());
/* If stack protection is enabled for this function, check the guard. */
return 0;
}
+#ifdef HAVE_simple_return
+
+/* Return true if INSN requires the stack frame to be set up.
+ PROLOGUE_USED contains the hard registers used in the function
+ prologue. SET_UP_BY_PROLOGUE is the set of registers we expect the
+ prologue to set up for the function. */
+bool
+requires_stack_frame_p (rtx insn, HARD_REG_SET prologue_used,
+ HARD_REG_SET set_up_by_prologue)
+{
+ df_ref *df_rec;
+ HARD_REG_SET hardregs;
+ unsigned regno;
+
+ if (CALL_P (insn))
+ return !SIBLING_CALL_P (insn);
+
+ CLEAR_HARD_REG_SET (hardregs);
+ for (df_rec = DF_INSN_DEFS (insn); *df_rec; df_rec++)
+ {
+ rtx dreg = DF_REF_REG (*df_rec);
+
+ if (!REG_P (dreg))
+ continue;
+
+ add_to_hard_reg_set (&hardregs, GET_MODE (dreg),
+ REGNO (dreg));
+ }
+ if (hard_reg_set_intersect_p (hardregs, prologue_used))
+ return true;
+ AND_COMPL_HARD_REG_SET (hardregs, call_used_reg_set);
+ for (regno = 0; regno < FIRST_PSEUDO_REGISTER; regno++)
+ if (TEST_HARD_REG_BIT (hardregs, regno)
+ && df_regs_ever_live_p (regno))
+ return true;
+
+ for (df_rec = DF_INSN_USES (insn); *df_rec; df_rec++)
+ {
+ rtx reg = DF_REF_REG (*df_rec);
+
+ if (!REG_P (reg))
+ continue;
+
+ add_to_hard_reg_set (&hardregs, GET_MODE (reg),
+ REGNO (reg));
+ }
+ if (hard_reg_set_intersect_p (hardregs, set_up_by_prologue))
+ return true;
+
+ return false;
+}
+
+/* Look for sets of call-saved registers in the first block of the
+ function, and move them down into successor blocks if the register
+ is used only on one path. This exposes more opportunities for
+ shrink-wrapping.
+ These kinds of sets often occur when incoming argument registers are
+ moved to call-saved registers because their values are live across
+ one or more calls during the function. */
+
+static void
+prepare_shrink_wrap (basic_block entry_block)
+{
+ rtx insn, curr;
+ FOR_BB_INSNS_SAFE (entry_block, insn, curr)
+ {
+ basic_block next_bb;
+ edge e, live_edge;
+ edge_iterator ei;
+ rtx set, scan;
+ unsigned destreg, srcreg;
+
+ if (!NONDEBUG_INSN_P (insn))
+ continue;
+ set = single_set (insn);
+ if (!set)
+ continue;
+
+ if (!REG_P (SET_SRC (set)) || !REG_P (SET_DEST (set)))
+ continue;
+ srcreg = REGNO (SET_SRC (set));
+ destreg = REGNO (SET_DEST (set));
+ if (hard_regno_nregs[srcreg][GET_MODE (SET_SRC (set))] > 1
+ || hard_regno_nregs[destreg][GET_MODE (SET_DEST (set))] > 1)
+ continue;
+
+ next_bb = entry_block;
+ scan = insn;
+
+ for (;;)
+ {
+ live_edge = NULL;
+ /* Try to find a single edge across which the register is live.
+ If we find one, we'll try to move the set across this edge. */
+ FOR_EACH_EDGE (e, ei, next_bb->succs)
+ {
+ if (REGNO_REG_SET_P (df_get_live_in (e->dest), destreg))
+ {
+ if (live_edge)
+ {
+ live_edge = NULL;
+ break;
+ }
+ live_edge = e;
+ }
+ }
+ if (!live_edge)
+ break;
+ /* We can sometimes encounter dead code. Don't try to move it
+ into the exit block. */
+ if (live_edge->dest == EXIT_BLOCK_PTR)
+ break;
+ if (EDGE_COUNT (live_edge->dest->preds) > 1)
+ break;
+ while (scan != BB_END (next_bb))
+ {
+ scan = NEXT_INSN (scan);
+ if (NONDEBUG_INSN_P (scan))
+ {
+ rtx link;
+ HARD_REG_SET set_regs;
+
+ CLEAR_HARD_REG_SET (set_regs);
+ note_stores (PATTERN (scan), record_hard_reg_sets,
+ &set_regs);
+ if (CALL_P (scan))
+ IOR_HARD_REG_SET (set_regs, call_used_reg_set);
+ for (link = REG_NOTES (scan); link; link = XEXP (link, 1))
+ if (REG_NOTE_KIND (link) == REG_INC)
+ record_hard_reg_sets (XEXP (link, 0), NULL, &set_regs);
+
+ if (TEST_HARD_REG_BIT (set_regs, srcreg)
+ || reg_referenced_p (SET_DEST (set),
+ PATTERN (scan)))
+ {
+ scan = NULL_RTX;
+ break;
+ }
+ if (CALL_P (scan))
+ {
+ rtx link = CALL_INSN_FUNCTION_USAGE (scan);
+ while (link)
+ {
+ rtx tmp = XEXP (link, 0);
+ if (GET_CODE (tmp) == USE
+ && reg_referenced_p (SET_DEST (set), tmp))
+ break;
+ link = XEXP (link, 1);
+ }
+ if (link)
+ {
+ scan = NULL_RTX;
+ break;
+ }
+ }
+ }
+ }
+ if (!scan)
+ break;
+ next_bb = live_edge->dest;
+ }
+
+ if (next_bb != entry_block)
+ {
+ rtx after = BB_HEAD (next_bb);
+ while (!NOTE_P (after)
+ || NOTE_KIND (after) != NOTE_INSN_BASIC_BLOCK)
+ after = NEXT_INSN (after);
+ emit_insn_after (PATTERN (insn), after);
+ delete_insn (insn);
+ }
+ }
+}
+
+#endif
+
#ifdef HAVE_return
/* Insert use of return register before the end of BB. */
emit_insn_before (seq, BB_END (bb));
}
-/* Insert gen_return at the end of block BB. This also means updating
- block_for_insn appropriately. */
+
+/* Create a return pattern, either simple_return or return, depending on
+ simple_p. */
+
+static rtx
+gen_return_pattern (bool simple_p)
+{
+#ifdef HAVE_simple_return
+ return simple_p ? gen_simple_return () : gen_return ();
+#else
+ gcc_assert (!simple_p);
+ return gen_return ();
+#endif
+}
+
+/* Insert an appropriate return pattern at the end of block BB. This
+ also means updating block_for_insn appropriately. SIMPLE_P is
+ the same as in gen_return_pattern and passed to it. */
static void
-emit_return_into_block (basic_block bb)
+emit_return_into_block (bool simple_p, basic_block bb)
{
- emit_jump_insn_after (gen_return (), BB_END (bb));
+ rtx jump, pat;
+ jump = emit_jump_insn_after (gen_return_pattern (simple_p), BB_END (bb));
+ pat = PATTERN (jump);
+ if (GET_CODE (pat) == PARALLEL)
+ pat = XVECEXP (pat, 0, 0);
+ gcc_assert (ANY_RETURN_P (pat));
+ JUMP_LABEL (jump) = pat;
}
-#endif /* HAVE_return */
+#endif
+
+/* Set JUMP_LABEL for a return insn. */
+
+void
+set_return_jump_label (rtx returnjump)
+{
+ rtx pat = PATTERN (returnjump);
+ if (GET_CODE (pat) == PARALLEL)
+ pat = XVECEXP (pat, 0, 0);
+ if (ANY_RETURN_P (pat))
+ JUMP_LABEL (returnjump) = pat;
+ else
+ JUMP_LABEL (returnjump) = ret_rtx;
+}
+
+#ifdef HAVE_simple_return
+/* Create a copy of BB instructions and insert at BEFORE. Redirect
+ preds of BB to COPY_BB if they don't appear in NEED_PROLOGUE. */
+static void
+dup_block_and_redirect (basic_block bb, basic_block copy_bb, rtx before,
+ bitmap_head *need_prologue)
+{
+ edge_iterator ei;
+ edge e;
+ rtx insn = BB_END (bb);
+
+ /* We know BB has a single successor, so there is no need to copy a
+ simple jump at the end of BB. */
+ if (simplejump_p (insn))
+ insn = PREV_INSN (insn);
+
+ start_sequence ();
+ duplicate_insn_chain (BB_HEAD (bb), insn);
+ if (dump_file)
+ {
+ unsigned count = 0;
+ for (insn = get_insns (); insn; insn = NEXT_INSN (insn))
+ if (active_insn_p (insn))
+ ++count;
+ fprintf (dump_file, "Duplicating bb %d to bb %d, %u active insns.\n",
+ bb->index, copy_bb->index, count);
+ }
+ insn = get_insns ();
+ end_sequence ();
+ emit_insn_before (insn, before);
+
+ /* Redirect all the paths that need no prologue into copy_bb. */
+ for (ei = ei_start (bb->preds); (e = ei_safe_edge (ei)); )
+ if (!bitmap_bit_p (need_prologue, e->src->index))
+ {
+ redirect_edge_and_branch_force (e, copy_bb);
+ continue;
+ }
+ else
+ ei_next (&ei);
+}
+#endif
+
+#if defined (HAVE_return) || defined (HAVE_simple_return)
+/* Return true if there are any active insns between HEAD and TAIL. */
+static bool
+active_insn_between (rtx head, rtx tail)
+{
+ while (tail)
+ {
+ if (active_insn_p (tail))
+ return true;
+ if (tail == head)
+ return false;
+ tail = PREV_INSN (tail);
+ }
+ return false;
+}
+
+/* LAST_BB is a block that exits, and empty of active instructions.
+ Examine its predecessors for jumps that can be converted to
+ (conditional) returns. */
+static VEC (edge, heap) *
+convert_jumps_to_returns (basic_block last_bb, bool simple_p,
+ VEC (edge, heap) *unconverted ATTRIBUTE_UNUSED)
+{
+ int i;
+ basic_block bb;
+ rtx label;
+ edge_iterator ei;
+ edge e;
+ VEC(basic_block,heap) *src_bbs;
+
+ src_bbs = VEC_alloc (basic_block, heap, EDGE_COUNT (last_bb->preds));
+ FOR_EACH_EDGE (e, ei, last_bb->preds)
+ if (e->src != ENTRY_BLOCK_PTR)
+ VEC_quick_push (basic_block, src_bbs, e->src);
+
+ label = BB_HEAD (last_bb);
+
+ FOR_EACH_VEC_ELT (basic_block, src_bbs, i, bb)
+ {
+ rtx jump = BB_END (bb);
+
+ if (!JUMP_P (jump) || JUMP_LABEL (jump) != label)
+ continue;
+
+ e = find_edge (bb, last_bb);
+
+ /* If we have an unconditional jump, we can replace that
+ with a simple return instruction. */
+ if (simplejump_p (jump))
+ {
+ /* The use of the return register might be present in the exit
+ fallthru block. Either:
+ - removing the use is safe, and we should remove the use in
+ the exit fallthru block, or
+ - removing the use is not safe, and we should add it here.
+ For now, we conservatively choose the latter. Either of the
+ 2 helps in crossjumping. */
+ emit_use_return_register_into_block (bb);
+
+ emit_return_into_block (simple_p, bb);
+ delete_insn (jump);
+ }
+
+ /* If we have a conditional jump branching to the last
+ block, we can try to replace that with a conditional
+ return instruction. */
+ else if (condjump_p (jump))
+ {
+ rtx dest;
+
+ if (simple_p)
+ dest = simple_return_rtx;
+ else
+ dest = ret_rtx;
+ if (!redirect_jump (jump, dest, 0))
+ {
+#ifdef HAVE_simple_return
+ if (simple_p)
+ {
+ if (dump_file)
+ fprintf (dump_file,
+ "Failed to redirect bb %d branch.\n", bb->index);
+ VEC_safe_push (edge, heap, unconverted, e);
+ }
+#endif
+ continue;
+ }
+
+ /* See comment in simplejump_p case above. */
+ emit_use_return_register_into_block (bb);
+
+ /* If this block has only one successor, it both jumps
+ and falls through to the fallthru block, so we can't
+ delete the edge. */
+ if (single_succ_p (bb))
+ continue;
+ }
+ else
+ {
+#ifdef HAVE_simple_return
+ if (simple_p)
+ {
+ if (dump_file)
+ fprintf (dump_file,
+ "Failed to redirect bb %d branch.\n", bb->index);
+ VEC_safe_push (edge, heap, unconverted, e);
+ }
+#endif
+ continue;
+ }
+
+ /* Fix up the CFG for the successful change we just made. */
+ redirect_edge_succ (e, EXIT_BLOCK_PTR);
+ }
+ VEC_free (basic_block, heap, src_bbs);
+ return unconverted;
+}
+
+/* Emit a return insn for the exit fallthru block. */
+static basic_block
+emit_return_for_exit (edge exit_fallthru_edge, bool simple_p)
+{
+ basic_block last_bb = exit_fallthru_edge->src;
+
+ if (JUMP_P (BB_END (last_bb)))
+ {
+ last_bb = split_edge (exit_fallthru_edge);
+ exit_fallthru_edge = single_succ_edge (last_bb);
+ }
+ emit_barrier_after (BB_END (last_bb));
+ emit_return_into_block (simple_p, last_bb);
+ exit_fallthru_edge->flags &= ~EDGE_FALLTHRU;
+ return last_bb;
+}
+#endif
+
/* Generate the prologue and epilogue RTL if the machine supports it. Thread
this into place with notes indicating where the prologue ends and where
- the epilogue begins. Update the basic block information when possible. */
+ the epilogue begins. Update the basic block information when possible.
+
+ Notes on epilogue placement:
+ There are several kinds of edges to the exit block:
+ * a single fallthru edge from LAST_BB
+ * possibly, edges from blocks containing sibcalls
+ * possibly, fake edges from infinite loops
+
+ The epilogue is always emitted on the fallthru edge from the last basic
+ block in the function, LAST_BB, into the exit block.
+
+ If LAST_BB is empty except for a label, it is the target of every
+ other basic block in the function that ends in a return. If a
+ target has a return or simple_return pattern (possibly with
+ conditional variants), these basic blocks can be changed so that a
+ return insn is emitted into them, and their target is adjusted to
+ the real exit block.
+
+ Notes on shrink wrapping: We implement a fairly conservative
+ version of shrink-wrapping rather than the textbook one. We only
+ generate a single prologue and a single epilogue. This is
+ sufficient to catch a number of interesting cases involving early
+ exits.
+
+ First, we identify the blocks that require the prologue to occur before
+ them. These are the ones that modify a call-saved register, or reference
+ any of the stack or frame pointer registers. To simplify things, we then
+ mark everything reachable from these blocks as also requiring a prologue.
+ This takes care of loops automatically, and avoids the need to examine
+ whether MEMs reference the frame, since it is sufficient to check for
+ occurrences of the stack or frame pointer.
+
+ We then compute the set of blocks for which the need for a prologue
+ is anticipatable (borrowing terminology from the shrink-wrapping
+ description in Muchnick's book). These are the blocks which either
+ require a prologue themselves, or those that have only successors
+ where the prologue is anticipatable. The prologue needs to be
+ inserted on all edges from BB1->BB2 where BB2 is in ANTIC and BB1
+ is not. For the moment, we ensure that only one such edge exists.
+
+ The epilogue is placed as described above, but we make a
+ distinction between inserting return and simple_return patterns
+ when modifying other blocks that end in a return. Blocks that end
+ in a sibcall omit the sibcall_epilogue if the block is not in
+ ANTIC. */
static void
thread_prologue_and_epilogue_insns (void)
{
bool inserted;
+#ifdef HAVE_simple_return
+ VEC (edge, heap) *unconverted_simple_returns = NULL;
+ bool nonempty_prologue;
+ bitmap_head bb_flags;
+ unsigned max_grow_size;
+#endif
+ rtx returnjump;
rtx seq ATTRIBUTE_UNUSED, epilogue_end ATTRIBUTE_UNUSED;
- edge entry_edge, e;
+ rtx prologue_seq ATTRIBUTE_UNUSED, split_prologue_seq ATTRIBUTE_UNUSED;
+ edge e, entry_edge, orig_entry_edge, exit_fallthru_edge;
edge_iterator ei;
+ df_analyze ();
+
rtl_profile_for_bb (ENTRY_BLOCK_PTR);
inserted = false;
seq = NULL_RTX;
epilogue_end = NULL_RTX;
+ returnjump = NULL_RTX;
/* Can't deal with multiple successors of the entry block at the
moment. Function should always have at least one entry
point. */
gcc_assert (single_succ_p (ENTRY_BLOCK_PTR));
entry_edge = single_succ_edge (ENTRY_BLOCK_PTR);
+ orig_entry_edge = entry_edge;
+ split_prologue_seq = NULL_RTX;
if (flag_split_stack
&& (lookup_attribute ("no_split_stack", DECL_ATTRIBUTES (cfun->decl))
== NULL))
start_sequence ();
emit_insn (gen_split_stack_prologue ());
- seq = get_insns ();
+ split_prologue_seq = get_insns ();
end_sequence ();
- record_insns (seq, NULL, &prologue_insn_hash);
- set_insn_locators (seq, prologue_locator);
-
- insert_insn_on_edge (seq, entry_edge);
- inserted = true;
+ record_insns (split_prologue_seq, NULL, &prologue_insn_hash);
+ set_insn_locators (split_prologue_seq, prologue_locator);
#endif
}
+ prologue_seq = NULL_RTX;
#ifdef HAVE_prologue
if (HAVE_prologue)
{
if (!targetm.profile_before_prologue () && crtl->profile)
emit_insn (gen_blockage ());
- seq = get_insns ();
+ prologue_seq = get_insns ();
end_sequence ();
- set_insn_locators (seq, prologue_locator);
-
- insert_insn_on_edge (seq, entry_edge);
- inserted = true;
+ set_insn_locators (prologue_seq, prologue_locator);
}
#endif
- /* If the exit block has no non-fake predecessors, we don't need
- an epilogue. */
- FOR_EACH_EDGE (e, ei, EXIT_BLOCK_PTR->preds)
- if ((e->flags & EDGE_FAKE) == 0)
- break;
- if (e == NULL)
- goto epilogue_done;
+#ifdef HAVE_simple_return
+ bitmap_initialize (&bb_flags, &bitmap_default_obstack);
- rtl_profile_for_bb (EXIT_BLOCK_PTR);
-#ifdef HAVE_return
- if (optimize && HAVE_return)
- {
- /* If we're allowed to generate a simple return instruction,
- then by definition we don't need a full epilogue. Examine
- the block that falls through to EXIT. If it does not
- contain any code, examine its predecessors and try to
- emit (conditional) return instructions. */
+ /* Try to perform a kind of shrink-wrapping, making sure the
+ prologue/epilogue is emitted only around those parts of the
+ function that require it. */
- basic_block last;
- rtx label;
+ nonempty_prologue = false;
+ for (seq = prologue_seq; seq; seq = NEXT_INSN (seq))
+ if (!NOTE_P (seq) || NOTE_KIND (seq) != NOTE_INSN_PROLOGUE_END)
+ {
+ nonempty_prologue = true;
+ break;
+ }
+
+ if (flag_shrink_wrap && HAVE_simple_return
+ && (targetm.profile_before_prologue () || !crtl->profile)
+ && nonempty_prologue && !crtl->calls_eh_return)
+ {
+ HARD_REG_SET prologue_clobbered, prologue_used, live_on_edge;
+ HARD_REG_SET set_up_by_prologue;
+ rtx p_insn;
+ VEC(basic_block, heap) *vec;
+ basic_block bb;
+ bitmap_head bb_antic_flags;
+ bitmap_head bb_on_list;
+ bitmap_head bb_tail;
+
+ if (dump_file)
+ fprintf (dump_file, "Attempting shrink-wrapping optimization.\n");
+
+ /* Compute the registers set and used in the prologue. */
+ CLEAR_HARD_REG_SET (prologue_clobbered);
+ CLEAR_HARD_REG_SET (prologue_used);
+ for (p_insn = prologue_seq; p_insn; p_insn = NEXT_INSN (p_insn))
+ {
+ HARD_REG_SET this_used;
+ if (!NONDEBUG_INSN_P (p_insn))
+ continue;
- e = find_fallthru_edge (EXIT_BLOCK_PTR->preds);
- if (e == NULL)
- goto epilogue_done;
- last = e->src;
+ CLEAR_HARD_REG_SET (this_used);
+ note_uses (&PATTERN (p_insn), record_hard_reg_uses,
+ &this_used);
+ AND_COMPL_HARD_REG_SET (this_used, prologue_clobbered);
+ IOR_HARD_REG_SET (prologue_used, this_used);
+ note_stores (PATTERN (p_insn), record_hard_reg_sets,
+ &prologue_clobbered);
+ }
+
+ prepare_shrink_wrap (entry_edge->dest);
+
+ bitmap_initialize (&bb_antic_flags, &bitmap_default_obstack);
+ bitmap_initialize (&bb_on_list, &bitmap_default_obstack);
+ bitmap_initialize (&bb_tail, &bitmap_default_obstack);
+
+ /* Find the set of basic blocks that require a stack frame,
+ and blocks that are too big to be duplicated. */
- /* Verify that there are no active instructions in the last block. */
- label = BB_END (last);
- while (label && !LABEL_P (label))
+ vec = VEC_alloc (basic_block, heap, n_basic_blocks);
+
+ CLEAR_HARD_REG_SET (set_up_by_prologue);
+ add_to_hard_reg_set (&set_up_by_prologue, Pmode, STACK_POINTER_REGNUM);
+ add_to_hard_reg_set (&set_up_by_prologue, Pmode, ARG_POINTER_REGNUM);
+ if (frame_pointer_needed)
+ add_to_hard_reg_set (&set_up_by_prologue, Pmode,
+ HARD_FRAME_POINTER_REGNUM);
+ if (pic_offset_table_rtx)
+ add_to_hard_reg_set (&set_up_by_prologue, Pmode,
+ PIC_OFFSET_TABLE_REGNUM);
+
+ /* We don't use a different max size depending on
+ optimize_bb_for_speed_p because increasing shrink-wrapping
+ opportunities by duplicating tail blocks can actually result
+ in an overall decrease in code size. */
+ max_grow_size = get_uncond_jump_length ();
+ max_grow_size *= PARAM_VALUE (PARAM_MAX_GROW_COPY_BB_INSNS);
+
+ FOR_EACH_BB (bb)
{
- if (active_insn_p (label))
- break;
- label = PREV_INSN (label);
+ rtx insn;
+ unsigned size = 0;
+
+ FOR_BB_INSNS (bb, insn)
+ if (NONDEBUG_INSN_P (insn))
+ {
+ if (requires_stack_frame_p (insn, prologue_used,
+ set_up_by_prologue))
+ {
+ if (bb == entry_edge->dest)
+ goto fail_shrinkwrap;
+ bitmap_set_bit (&bb_flags, bb->index);
+ VEC_quick_push (basic_block, vec, bb);
+ break;
+ }
+ else if (size <= max_grow_size)
+ {
+ size += get_attr_min_length (insn);
+ if (size > max_grow_size)
+ bitmap_set_bit (&bb_on_list, bb->index);
+ }
+ }
}
- if (BB_HEAD (last) == label && LABEL_P (label))
+ /* Blocks that really need a prologue, or are too big for tails. */
+ bitmap_ior_into (&bb_on_list, &bb_flags);
+
+ /* For every basic block that needs a prologue, mark all blocks
+ reachable from it, so as to ensure they are also seen as
+ requiring a prologue. */
+ while (!VEC_empty (basic_block, vec))
{
- edge_iterator ei2;
+ basic_block tmp_bb = VEC_pop (basic_block, vec);
- for (ei2 = ei_start (last->preds); (e = ei_safe_edge (ei2)); )
- {
- basic_block bb = e->src;
- rtx jump;
+ FOR_EACH_EDGE (e, ei, tmp_bb->succs)
+ if (e->dest != EXIT_BLOCK_PTR
+ && bitmap_set_bit (&bb_flags, e->dest->index))
+ VEC_quick_push (basic_block, vec, e->dest);
+ }
- if (bb == ENTRY_BLOCK_PTR)
- {
- ei_next (&ei2);
- continue;
- }
+ /* Find the set of basic blocks that need no prologue, have a
+ single successor, can be duplicated, meet a max size
+ requirement, and go to the exit via like blocks. */
+ VEC_quick_push (basic_block, vec, EXIT_BLOCK_PTR);
+ while (!VEC_empty (basic_block, vec))
+ {
+ basic_block tmp_bb = VEC_pop (basic_block, vec);
+
+ FOR_EACH_EDGE (e, ei, tmp_bb->preds)
+ if (single_succ_p (e->src)
+ && !bitmap_bit_p (&bb_on_list, e->src->index)
+ && can_duplicate_block_p (e->src)
+ && bitmap_set_bit (&bb_tail, e->src->index))
+ VEC_quick_push (basic_block, vec, e->src);
+ }
- jump = BB_END (bb);
- if (!JUMP_P (jump) || JUMP_LABEL (jump) != label)
- {
- ei_next (&ei2);
- continue;
- }
+ /* Now walk backwards from every block that is marked as needing
+ a prologue to compute the bb_antic_flags bitmap. Exclude
+ tail blocks; They can be duplicated to be used on paths not
+ needing a prologue. */
+ bitmap_clear (&bb_on_list);
+ bitmap_and_compl (&bb_antic_flags, &bb_flags, &bb_tail);
+ FOR_EACH_BB (bb)
+ {
+ if (!bitmap_bit_p (&bb_antic_flags, bb->index))
+ continue;
+ FOR_EACH_EDGE (e, ei, bb->preds)
+ if (!bitmap_bit_p (&bb_antic_flags, e->src->index)
+ && bitmap_set_bit (&bb_on_list, e->src->index))
+ VEC_quick_push (basic_block, vec, e->src);
+ }
+ while (!VEC_empty (basic_block, vec))
+ {
+ basic_block tmp_bb = VEC_pop (basic_block, vec);
+ bool all_set = true;
- /* If we have an unconditional jump, we can replace that
- with a simple return instruction. */
- if (simplejump_p (jump))
- {
- /* The use of the return register might be present in the exit
- fallthru block. Either:
- - removing the use is safe, and we should remove the use in
- the exit fallthru block, or
- - removing the use is not safe, and we should add it here.
- For now, we conservatively choose the latter. Either of the
- 2 helps in crossjumping. */
- emit_use_return_register_into_block (bb);
-
- emit_return_into_block (bb);
- delete_insn (jump);
- }
+ bitmap_clear_bit (&bb_on_list, tmp_bb->index);
+ FOR_EACH_EDGE (e, ei, tmp_bb->succs)
+ if (!bitmap_bit_p (&bb_antic_flags, e->dest->index))
+ {
+ all_set = false;
+ break;
+ }
- /* If we have a conditional jump, we can try to replace
- that with a conditional return instruction. */
- else if (condjump_p (jump))
+ if (all_set)
+ {
+ bitmap_set_bit (&bb_antic_flags, tmp_bb->index);
+ FOR_EACH_EDGE (e, ei, tmp_bb->preds)
+ if (!bitmap_bit_p (&bb_antic_flags, e->src->index)
+ && bitmap_set_bit (&bb_on_list, e->src->index))
+ VEC_quick_push (basic_block, vec, e->src);
+ }
+ }
+ /* Find exactly one edge that leads to a block in ANTIC from
+ a block that isn't. */
+ if (!bitmap_bit_p (&bb_antic_flags, entry_edge->dest->index))
+ FOR_EACH_BB (bb)
+ {
+ if (!bitmap_bit_p (&bb_antic_flags, bb->index))
+ continue;
+ FOR_EACH_EDGE (e, ei, bb->preds)
+ if (!bitmap_bit_p (&bb_antic_flags, e->src->index))
{
- if (! redirect_jump (jump, 0, 0))
+ if (entry_edge != orig_entry_edge)
{
- ei_next (&ei2);
- continue;
+ entry_edge = orig_entry_edge;
+ if (dump_file)
+ fprintf (dump_file, "More than one candidate edge.\n");
+ goto fail_shrinkwrap;
}
+ if (dump_file)
+ fprintf (dump_file, "Found candidate edge for "
+ "shrink-wrapping, %d->%d.\n", e->src->index,
+ e->dest->index);
+ entry_edge = e;
+ }
+ }
+
+ if (entry_edge != orig_entry_edge)
+ {
+ /* Test whether the prologue is known to clobber any register
+ (other than FP or SP) which are live on the edge. */
+ CLEAR_HARD_REG_BIT (prologue_clobbered, STACK_POINTER_REGNUM);
+ if (frame_pointer_needed)
+ CLEAR_HARD_REG_BIT (prologue_clobbered, HARD_FRAME_POINTER_REGNUM);
+ CLEAR_HARD_REG_SET (live_on_edge);
+ reg_set_to_hard_reg_set (&live_on_edge,
+ df_get_live_in (entry_edge->dest));
+ if (hard_reg_set_intersect_p (live_on_edge, prologue_clobbered))
+ {
+ entry_edge = orig_entry_edge;
+ if (dump_file)
+ fprintf (dump_file,
+ "Shrink-wrapping aborted due to clobber.\n");
+ }
+ }
+ if (entry_edge != orig_entry_edge)
+ {
+ crtl->shrink_wrapped = true;
+ if (dump_file)
+ fprintf (dump_file, "Performing shrink-wrapping.\n");
+
+ /* Find tail blocks reachable from both blocks needing a
+ prologue and blocks not needing a prologue. */
+ if (!bitmap_empty_p (&bb_tail))
+ FOR_EACH_BB (bb)
+ {
+ bool some_pro, some_no_pro;
+ if (!bitmap_bit_p (&bb_tail, bb->index))
+ continue;
+ some_pro = some_no_pro = false;
+ FOR_EACH_EDGE (e, ei, bb->preds)
+ {
+ if (bitmap_bit_p (&bb_flags, e->src->index))
+ some_pro = true;
+ else
+ some_no_pro = true;
+ }
+ if (some_pro && some_no_pro)
+ VEC_quick_push (basic_block, vec, bb);
+ else
+ bitmap_clear_bit (&bb_tail, bb->index);
+ }
+ /* Find the head of each tail. */
+ while (!VEC_empty (basic_block, vec))
+ {
+ basic_block tbb = VEC_pop (basic_block, vec);
- /* See comment in simple_jump_p case above. */
- emit_use_return_register_into_block (bb);
+ if (!bitmap_bit_p (&bb_tail, tbb->index))
+ continue;
- /* If this block has only one successor, it both jumps
- and falls through to the fallthru block, so we can't
- delete the edge. */
- if (single_succ_p (bb))
- {
- ei_next (&ei2);
- continue;
- }
- }
- else
+ while (single_succ_p (tbb))
{
- ei_next (&ei2);
- continue;
+ tbb = single_succ (tbb);
+ bitmap_clear_bit (&bb_tail, tbb->index);
}
+ }
+ /* Now duplicate the tails. */
+ if (!bitmap_empty_p (&bb_tail))
+ FOR_EACH_BB_REVERSE (bb)
+ {
+ basic_block copy_bb, tbb;
+ rtx insert_point;
+ int eflags;
+
+ if (!bitmap_clear_bit (&bb_tail, bb->index))
+ continue;
+
+ /* Create a copy of BB, instructions and all, for
+ use on paths that don't need a prologue.
+ Ideal placement of the copy is on a fall-thru edge
+ or after a block that would jump to the copy. */
+ FOR_EACH_EDGE (e, ei, bb->preds)
+ if (!bitmap_bit_p (&bb_flags, e->src->index)
+ && single_succ_p (e->src))
+ break;
+ if (e)
+ {
+ copy_bb = create_basic_block (NEXT_INSN (BB_END (e->src)),
+ NULL_RTX, e->src);
+ BB_COPY_PARTITION (copy_bb, e->src);
+ }
+ else
+ {
+ /* Otherwise put the copy at the end of the function. */
+ copy_bb = create_basic_block (NULL_RTX, NULL_RTX,
+ EXIT_BLOCK_PTR->prev_bb);
+ BB_COPY_PARTITION (copy_bb, bb);
+ }
+
+ insert_point = emit_note_after (NOTE_INSN_DELETED,
+ BB_END (copy_bb));
+ emit_barrier_after (BB_END (copy_bb));
+
+ tbb = bb;
+ while (1)
+ {
+ dup_block_and_redirect (tbb, copy_bb, insert_point,
+ &bb_flags);
+ tbb = single_succ (tbb);
+ if (tbb == EXIT_BLOCK_PTR)
+ break;
+ e = split_block (copy_bb, PREV_INSN (insert_point));
+ copy_bb = e->dest;
+ }
+
+ /* Quiet verify_flow_info by (ab)using EDGE_FAKE.
+ We have yet to add a simple_return to the tails,
+ as we'd like to first convert_jumps_to_returns in
+ case the block is no longer used after that. */
+ eflags = EDGE_FAKE;
+ if (CALL_P (PREV_INSN (insert_point))
+ && SIBLING_CALL_P (PREV_INSN (insert_point)))
+ eflags = EDGE_SIBCALL | EDGE_ABNORMAL;
+ make_single_succ_edge (copy_bb, EXIT_BLOCK_PTR, eflags);
+
+ /* verify_flow_info doesn't like a note after a
+ sibling call. */
+ delete_insn (insert_point);
+ if (bitmap_empty_p (&bb_tail))
+ break;
+ }
+ }
- /* Fix up the CFG for the successful change we just made. */
- redirect_edge_succ (e, EXIT_BLOCK_PTR);
+ fail_shrinkwrap:
+ bitmap_clear (&bb_tail);
+ bitmap_clear (&bb_antic_flags);
+ bitmap_clear (&bb_on_list);
+ VEC_free (basic_block, heap, vec);
+ }
+#endif
+
+ if (split_prologue_seq != NULL_RTX)
+ {
+ insert_insn_on_edge (split_prologue_seq, orig_entry_edge);
+ inserted = true;
+ }
+ if (prologue_seq != NULL_RTX)
+ {
+ insert_insn_on_edge (prologue_seq, entry_edge);
+ inserted = true;
+ }
+
+ /* If the exit block has no non-fake predecessors, we don't need
+ an epilogue. */
+ FOR_EACH_EDGE (e, ei, EXIT_BLOCK_PTR->preds)
+ if ((e->flags & EDGE_FAKE) == 0)
+ break;
+ if (e == NULL)
+ goto epilogue_done;
+
+ rtl_profile_for_bb (EXIT_BLOCK_PTR);
+
+ exit_fallthru_edge = find_fallthru_edge (EXIT_BLOCK_PTR->preds);
+
+ /* If we're allowed to generate a simple return instruction, then by
+ definition we don't need a full epilogue. If the last basic
+ block before the exit block does not contain active instructions,
+ examine its predecessors and try to emit (conditional) return
+ instructions. */
+#ifdef HAVE_simple_return
+ if (entry_edge != orig_entry_edge)
+ {
+ if (optimize)
+ {
+ unsigned i, last;
+
+ /* convert_jumps_to_returns may add to EXIT_BLOCK_PTR->preds
+ (but won't remove). Stop at end of current preds. */
+ last = EDGE_COUNT (EXIT_BLOCK_PTR->preds);
+ for (i = 0; i < last; i++)
+ {
+ e = EDGE_I (EXIT_BLOCK_PTR->preds, i);
+ if (LABEL_P (BB_HEAD (e->src))
+ && !bitmap_bit_p (&bb_flags, e->src->index)
+ && !active_insn_between (BB_HEAD (e->src), BB_END (e->src)))
+ unconverted_simple_returns
+ = convert_jumps_to_returns (e->src, true,
+ unconverted_simple_returns);
}
+ }
+
+ if (exit_fallthru_edge != NULL
+ && EDGE_COUNT (exit_fallthru_edge->src->preds) != 0
+ && !bitmap_bit_p (&bb_flags, exit_fallthru_edge->src->index))
+ {
+ basic_block last_bb;
- /* Emit a return insn for the exit fallthru block. Whether
- this is still reachable will be determined later. */
+ last_bb = emit_return_for_exit (exit_fallthru_edge, true);
+ returnjump = BB_END (last_bb);
+ exit_fallthru_edge = NULL;
+ }
+ }
+#endif
+#ifdef HAVE_return
+ if (HAVE_return)
+ {
+ if (exit_fallthru_edge == NULL)
+ goto epilogue_done;
+
+ if (optimize)
+ {
+ basic_block last_bb = exit_fallthru_edge->src;
- emit_barrier_after (BB_END (last));
- emit_return_into_block (last);
- epilogue_end = BB_END (last);
- single_succ_edge (last)->flags &= ~EDGE_FALLTHRU;
- goto epilogue_done;
+ if (LABEL_P (BB_HEAD (last_bb))
+ && !active_insn_between (BB_HEAD (last_bb), BB_END (last_bb)))
+ convert_jumps_to_returns (last_bb, false, NULL);
+
+ if (EDGE_COUNT (last_bb->preds) != 0
+ && single_succ_p (last_bb))
+ {
+ last_bb = emit_return_for_exit (exit_fallthru_edge, false);
+ epilogue_end = returnjump = BB_END (last_bb);
+#ifdef HAVE_simple_return
+ /* Emitting the return may add a basic block.
+ Fix bb_flags for the added block. */
+ if (last_bb != exit_fallthru_edge->src)
+ bitmap_set_bit (&bb_flags, last_bb->index);
+#endif
+ goto epilogue_done;
+ }
}
}
#endif
}
#endif
- /* Find the edge that falls through to EXIT. Other edges may exist
- due to RETURN instructions, but those don't need epilogues.
- There really shouldn't be a mixture -- either all should have
- been converted or none, however... */
+ /* If nothing falls through into the exit block, we don't need an
+ epilogue. */
- e = find_fallthru_edge (EXIT_BLOCK_PTR->preds);
- if (e == NULL)
+ if (exit_fallthru_edge == NULL)
goto epilogue_done;
#ifdef HAVE_epilogue
set_insn_locators (seq, epilogue_locator);
seq = get_insns ();
+ returnjump = get_last_insn ();
end_sequence ();
- insert_insn_on_edge (seq, e);
+ insert_insn_on_edge (seq, exit_fallthru_edge);
inserted = true;
+
+ if (JUMP_P (returnjump))
+ set_return_jump_label (returnjump);
}
else
#endif
{
basic_block cur_bb;
- if (! next_active_insn (BB_END (e->src)))
+ if (! next_active_insn (BB_END (exit_fallthru_edge->src)))
goto epilogue_done;
/* We have a fall-through edge to the exit block, the source is not
at the end of the function, and there will be an assembler epilogue
at the end of the function.
We can't use force_nonfallthru here, because that would try to
- use return. Inserting a jump 'by hand' is extremely messy, so
+ use return. Inserting a jump 'by hand' is extremely messy, so
we take advantage of cfg_layout_finalize using
- fixup_fallthru_exit_predecessor. */
+ fixup_fallthru_exit_predecessor. */
cfg_layout_initialize (0);
FOR_EACH_BB (cur_bb)
if (cur_bb->index >= NUM_FIXED_BLOCKS
}
epilogue_done:
+
default_rtl_profile ();
if (inserted)
blocks = sbitmap_alloc (last_basic_block);
sbitmap_zero (blocks);
SET_BIT (blocks, entry_edge->dest->index);
+ SET_BIT (blocks, orig_entry_edge->dest->index);
find_many_sub_basic_blocks (blocks);
sbitmap_free (blocks);
}
}
+#ifdef HAVE_simple_return
+ /* If there were branches to an empty LAST_BB which we tried to
+ convert to conditional simple_returns, but couldn't for some
+ reason, create a block to hold a simple_return insn and redirect
+ those remaining edges. */
+ if (!VEC_empty (edge, unconverted_simple_returns))
+ {
+ basic_block simple_return_block_hot = NULL;
+ basic_block simple_return_block_cold = NULL;
+ edge pending_edge_hot = NULL;
+ edge pending_edge_cold = NULL;
+ basic_block exit_pred = EXIT_BLOCK_PTR->prev_bb;
+ int i;
+
+ gcc_assert (entry_edge != orig_entry_edge);
+
+ /* See if we can reuse the last insn that was emitted for the
+ epilogue. */
+ if (returnjump != NULL_RTX
+ && JUMP_LABEL (returnjump) == simple_return_rtx)
+ {
+ e = split_block (BLOCK_FOR_INSN (returnjump), PREV_INSN (returnjump));
+ if (BB_PARTITION (e->src) == BB_HOT_PARTITION)
+ simple_return_block_hot = e->dest;
+ else
+ simple_return_block_cold = e->dest;
+ }
+
+ /* Also check returns we might need to add to tail blocks. */
+ FOR_EACH_EDGE (e, ei, EXIT_BLOCK_PTR->preds)
+ if (EDGE_COUNT (e->src->preds) != 0
+ && (e->flags & EDGE_FAKE) != 0
+ && !bitmap_bit_p (&bb_flags, e->src->index))
+ {
+ if (BB_PARTITION (e->src) == BB_HOT_PARTITION)
+ pending_edge_hot = e;
+ else
+ pending_edge_cold = e;
+ }
+
+ FOR_EACH_VEC_ELT (edge, unconverted_simple_returns, i, e)
+ {
+ basic_block *pdest_bb;
+ edge pending;
+
+ if (BB_PARTITION (e->src) == BB_HOT_PARTITION)
+ {
+ pdest_bb = &simple_return_block_hot;
+ pending = pending_edge_hot;
+ }
+ else
+ {
+ pdest_bb = &simple_return_block_cold;
+ pending = pending_edge_cold;
+ }
+
+ if (*pdest_bb == NULL && pending != NULL)
+ {
+ emit_return_into_block (true, pending->src);
+ pending->flags &= ~(EDGE_FALLTHRU | EDGE_FAKE);
+ *pdest_bb = pending->src;
+ }
+ else if (*pdest_bb == NULL)
+ {
+ basic_block bb;
+ rtx start;
+
+ bb = create_basic_block (NULL, NULL, exit_pred);
+ BB_COPY_PARTITION (bb, e->src);
+ start = emit_jump_insn_after (gen_simple_return (),
+ BB_END (bb));
+ JUMP_LABEL (start) = simple_return_rtx;
+ emit_barrier_after (start);
+
+ *pdest_bb = bb;
+ make_edge (bb, EXIT_BLOCK_PTR, 0);
+ }
+ redirect_edge_and_branch_force (e, *pdest_bb);
+ }
+ VEC_free (edge, heap, unconverted_simple_returns);
+ }
+
+ if (entry_edge != orig_entry_edge)
+ {
+ FOR_EACH_EDGE (e, ei, EXIT_BLOCK_PTR->preds)
+ if (EDGE_COUNT (e->src->preds) != 0
+ && (e->flags & EDGE_FAKE) != 0
+ && !bitmap_bit_p (&bb_flags, e->src->index))
+ {
+ emit_return_into_block (true, e->src);
+ e->flags &= ~(EDGE_FALLTHRU | EDGE_FAKE);
+ }
+ }
+#endif
+
#ifdef HAVE_sibcall_epilogue
/* Emit sibling epilogues before any sibling call sites. */
for (ei = ei_start (EXIT_BLOCK_PTR->preds); (e = ei_safe_edge (ei)); )
{
basic_block bb = e->src;
rtx insn = BB_END (bb);
+ rtx ep_seq;
if (!CALL_P (insn)
- || ! SIBLING_CALL_P (insn))
+ || ! SIBLING_CALL_P (insn)
+#ifdef HAVE_simple_return
+ || (entry_edge != orig_entry_edge
+ && !bitmap_bit_p (&bb_flags, bb->index))
+#endif
+ )
{
ei_next (&ei);
continue;
}
- start_sequence ();
- emit_note (NOTE_INSN_EPILOGUE_BEG);
- emit_insn (gen_sibcall_epilogue ());
- seq = get_insns ();
- end_sequence ();
+ ep_seq = gen_sibcall_epilogue ();
+ if (ep_seq)
+ {
+ start_sequence ();
+ emit_note (NOTE_INSN_EPILOGUE_BEG);
+ emit_insn (ep_seq);
+ seq = get_insns ();
+ end_sequence ();
- /* Retain a map of the epilogue insns. Used in life analysis to
- avoid getting rid of sibcall epilogue insns. Do this before we
- actually emit the sequence. */
- record_insns (seq, NULL, &epilogue_insn_hash);
- set_insn_locators (seq, epilogue_locator);
+ /* Retain a map of the epilogue insns. Used in life analysis to
+ avoid getting rid of sibcall epilogue insns. Do this before we
+ actually emit the sequence. */
+ record_insns (seq, NULL, &epilogue_insn_hash);
+ set_insn_locators (seq, epilogue_locator);
- emit_insn_before (seq, insn);
+ emit_insn_before (seq, insn);
+ }
ei_next (&ei);
}
#endif
}
#endif
+#ifdef HAVE_simple_return
+ bitmap_clear (&bb_flags);
+#endif
+
/* Threading the prologue and epilogue changes the artificial refs
in the entry and exit blocks. */
epilogue_completed = 1;
thread_prologue_and_epilogue_insns ();
/* The stack usage info is finalized during prologue expansion. */
- if (flag_stack_usage)
+ if (flag_stack_usage_info)
output_stack_usage ();
return 0;
0, /* properties_provided */
0, /* properties_destroyed */
TODO_verify_flow, /* todo_flags_start */
- TODO_dump_func |
TODO_df_verify |
TODO_df_finish | TODO_verify_rtl_sharing |
TODO_ggc_collect /* todo_flags_finish */
0, /* properties_provided */
0, /* properties_destroyed */
0, /* todo_flags_start */
- TODO_dump_func /* todo_flags_finish */
+ 0 /* todo_flags_finish */
}
};
OSDN Git Service
