895 lines
28 KiB
C
895 lines
28 KiB
C
/* Perform doloop optimizations
|
|
Copyright (C) 1999, 2000, 2001, 2002 Free Software Foundation, Inc.
|
|
Contributed by Michael P. Hayes (m.hayes@elec.canterbury.ac.nz)
|
|
|
|
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 2, 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 COPYING. If not, write to the Free
|
|
Software Foundation, 59 Temple Place - Suite 330, Boston, MA
|
|
02111-1307, USA. */
|
|
|
|
#include "config.h"
|
|
#include "system.h"
|
|
#include "rtl.h"
|
|
#include "flags.h"
|
|
#include "expr.h"
|
|
#include "loop.h"
|
|
#include "hard-reg-set.h"
|
|
#include "basic-block.h"
|
|
#include "toplev.h"
|
|
#include "tm_p.h"
|
|
|
|
|
|
/* This module is used to modify loops with a determinable number of
|
|
iterations to use special low-overhead looping instructions.
|
|
|
|
It first validates whether the loop is well behaved and has a
|
|
determinable number of iterations (either at compile or run-time).
|
|
It then modifies the loop to use a low-overhead looping pattern as
|
|
follows:
|
|
|
|
1. A pseudo register is allocated as the loop iteration counter.
|
|
|
|
2. The number of loop iterations is calculated and is stored
|
|
in the loop counter.
|
|
|
|
3. At the end of the loop, the jump insn is replaced by the
|
|
doloop_end pattern. The compare must remain because it might be
|
|
used elsewhere. If the loop-variable or condition register are
|
|
used elsewhere, they will be eliminated by flow.
|
|
|
|
4. An optional doloop_begin pattern is inserted at the top of the
|
|
loop.
|
|
*/
|
|
|
|
|
|
#ifdef HAVE_doloop_end
|
|
|
|
static rtx doloop_condition_get
|
|
PARAMS ((rtx));
|
|
static unsigned HOST_WIDE_INT doloop_iterations_max
|
|
PARAMS ((const struct loop_info *, enum machine_mode, int));
|
|
static int doloop_valid_p
|
|
PARAMS ((const struct loop *, rtx));
|
|
static int doloop_modify
|
|
PARAMS ((const struct loop *, rtx, rtx, rtx, rtx, rtx));
|
|
static int doloop_modify_runtime
|
|
PARAMS ((const struct loop *, rtx, rtx, rtx, enum machine_mode, rtx));
|
|
|
|
|
|
/* Return the loop termination condition for PATTERN or zero
|
|
if it is not a decrement and branch jump insn. */
|
|
static rtx
|
|
doloop_condition_get (pattern)
|
|
rtx pattern;
|
|
{
|
|
rtx cmp;
|
|
rtx inc;
|
|
rtx reg;
|
|
rtx condition;
|
|
|
|
/* The canonical doloop pattern we expect is:
|
|
|
|
(parallel [(set (pc) (if_then_else (condition)
|
|
(label_ref (label))
|
|
(pc)))
|
|
(set (reg) (plus (reg) (const_int -1)))
|
|
(additional clobbers and uses)])
|
|
|
|
Some machines (IA-64) make the decrement conditional on
|
|
the condition as well, so we don't bother verifying the
|
|
actual decrement. In summary, the branch must be the
|
|
first entry of the parallel (also required by jump.c),
|
|
and the second entry of the parallel must be a set of
|
|
the loop counter register. */
|
|
|
|
if (GET_CODE (pattern) != PARALLEL)
|
|
return 0;
|
|
|
|
cmp = XVECEXP (pattern, 0, 0);
|
|
inc = XVECEXP (pattern, 0, 1);
|
|
|
|
/* Check for (set (reg) (something)). */
|
|
if (GET_CODE (inc) != SET || ! REG_P (SET_DEST (inc)))
|
|
return 0;
|
|
|
|
/* Extract loop counter register. */
|
|
reg = SET_DEST (inc);
|
|
|
|
/* Check for (set (pc) (if_then_else (condition)
|
|
(label_ref (label))
|
|
(pc))). */
|
|
if (GET_CODE (cmp) != SET
|
|
|| SET_DEST (cmp) != pc_rtx
|
|
|| GET_CODE (SET_SRC (cmp)) != IF_THEN_ELSE
|
|
|| GET_CODE (XEXP (SET_SRC (cmp), 1)) != LABEL_REF
|
|
|| XEXP (SET_SRC (cmp), 2) != pc_rtx)
|
|
return 0;
|
|
|
|
/* Extract loop termination condition. */
|
|
condition = XEXP (SET_SRC (cmp), 0);
|
|
|
|
if ((GET_CODE (condition) != GE && GET_CODE (condition) != NE)
|
|
|| GET_CODE (XEXP (condition, 1)) != CONST_INT)
|
|
return 0;
|
|
|
|
if (XEXP (condition, 0) == reg)
|
|
return condition;
|
|
|
|
if (GET_CODE (XEXP (condition, 0)) == PLUS
|
|
&& XEXP (XEXP (condition, 0), 0) == reg)
|
|
return condition;
|
|
|
|
/* ??? If a machine uses a funny comparison, we could return a
|
|
canonicalised form here. */
|
|
|
|
return 0;
|
|
}
|
|
|
|
|
|
/* Return an estimate of the maximum number of loop iterations for the
|
|
loop specified by LOOP or zero if the loop is not normal.
|
|
MODE is the mode of the iteration count and NONNEG is non-zero if
|
|
the iteration count has been proved to be non-negative. */
|
|
static unsigned HOST_WIDE_INT
|
|
doloop_iterations_max (loop_info, mode, nonneg)
|
|
const struct loop_info *loop_info;
|
|
enum machine_mode mode;
|
|
int nonneg;
|
|
{
|
|
unsigned HOST_WIDE_INT n_iterations_max;
|
|
enum rtx_code code;
|
|
rtx min_value;
|
|
rtx max_value;
|
|
HOST_WIDE_INT abs_inc;
|
|
int neg_inc;
|
|
|
|
neg_inc = 0;
|
|
abs_inc = INTVAL (loop_info->increment);
|
|
if (abs_inc < 0)
|
|
{
|
|
abs_inc = -abs_inc;
|
|
neg_inc = 1;
|
|
}
|
|
|
|
if (neg_inc)
|
|
{
|
|
code = swap_condition (loop_info->comparison_code);
|
|
min_value = loop_info->final_equiv_value;
|
|
max_value = loop_info->initial_equiv_value;
|
|
}
|
|
else
|
|
{
|
|
code = loop_info->comparison_code;
|
|
min_value = loop_info->initial_equiv_value;
|
|
max_value = loop_info->final_equiv_value;
|
|
}
|
|
|
|
/* Since the loop has a VTOP, we know that the initial test will be
|
|
true and thus the value of max_value should be greater than the
|
|
value of min_value. Thus the difference should always be positive
|
|
and the code must be LT, LE, LTU, LEU, or NE. Otherwise the loop is
|
|
not normal, e.g., `for (i = 0; i < 10; i--)'. */
|
|
switch (code)
|
|
{
|
|
case LTU:
|
|
case LEU:
|
|
{
|
|
unsigned HOST_WIDE_INT umax;
|
|
unsigned HOST_WIDE_INT umin;
|
|
|
|
if (GET_CODE (min_value) == CONST_INT)
|
|
umin = INTVAL (min_value);
|
|
else
|
|
umin = 0;
|
|
|
|
if (GET_CODE (max_value) == CONST_INT)
|
|
umax = INTVAL (max_value);
|
|
else
|
|
umax = ((unsigned) 2 << (GET_MODE_BITSIZE (mode) - 1)) - 1;
|
|
|
|
n_iterations_max = umax - umin;
|
|
break;
|
|
}
|
|
|
|
case LT:
|
|
case LE:
|
|
{
|
|
HOST_WIDE_INT smax;
|
|
HOST_WIDE_INT smin;
|
|
|
|
if (GET_CODE (min_value) == CONST_INT)
|
|
smin = INTVAL (min_value);
|
|
else
|
|
smin = -((unsigned) 1 << (GET_MODE_BITSIZE (mode) - 1));
|
|
|
|
if (GET_CODE (max_value) == CONST_INT)
|
|
smax = INTVAL (max_value);
|
|
else
|
|
smax = ((unsigned) 1 << (GET_MODE_BITSIZE (mode) - 1)) - 1;
|
|
|
|
n_iterations_max = smax - smin;
|
|
break;
|
|
}
|
|
|
|
case NE:
|
|
if (GET_CODE (min_value) == CONST_INT
|
|
&& GET_CODE (max_value) == CONST_INT)
|
|
n_iterations_max = INTVAL (max_value) - INTVAL (min_value);
|
|
else
|
|
/* We need to conservatively assume that we might have the maximum
|
|
number of iterations without any additional knowledge. */
|
|
n_iterations_max = ((unsigned) 2 << (GET_MODE_BITSIZE (mode) - 1)) - 1;
|
|
break;
|
|
|
|
default:
|
|
return 0;
|
|
}
|
|
|
|
n_iterations_max /= abs_inc;
|
|
|
|
/* If we know that the iteration count is non-negative then adjust
|
|
n_iterations_max if it is so large that it appears negative. */
|
|
if (nonneg
|
|
&& n_iterations_max > ((unsigned) 1 << (GET_MODE_BITSIZE (mode) - 1)))
|
|
n_iterations_max = ((unsigned) 1 << (GET_MODE_BITSIZE (mode) - 1)) - 1;
|
|
|
|
return n_iterations_max;
|
|
}
|
|
|
|
|
|
/* Return non-zero if the loop specified by LOOP is suitable for
|
|
the use of special low-overhead looping instructions. */
|
|
static int
|
|
doloop_valid_p (loop, jump_insn)
|
|
const struct loop *loop;
|
|
rtx jump_insn;
|
|
{
|
|
const struct loop_info *loop_info = LOOP_INFO (loop);
|
|
|
|
/* The loop must have a conditional jump at the end. */
|
|
if (! any_condjump_p (jump_insn)
|
|
|| ! onlyjump_p (jump_insn))
|
|
{
|
|
if (loop_dump_stream)
|
|
fprintf (loop_dump_stream,
|
|
"Doloop: Invalid jump at loop end.\n");
|
|
return 0;
|
|
}
|
|
|
|
/* Give up if a loop has been completely unrolled. */
|
|
if (loop_info->n_iterations == loop_info->unroll_number)
|
|
{
|
|
if (loop_dump_stream)
|
|
fprintf (loop_dump_stream,
|
|
"Doloop: Loop completely unrolled.\n");
|
|
return 0;
|
|
}
|
|
|
|
/* The loop must have a single exit target. A break or return
|
|
statement within a loop will generate multiple loop exits.
|
|
Another example of a loop that currently generates multiple exit
|
|
targets is for (i = 0; i < (foo ? 8 : 4); i++) { }. */
|
|
if (loop_info->has_multiple_exit_targets || loop->exit_count)
|
|
{
|
|
if (loop_dump_stream)
|
|
fprintf (loop_dump_stream,
|
|
"Doloop: Loop has multiple exit targets.\n");
|
|
return 0;
|
|
}
|
|
|
|
/* An indirect jump may jump out of the loop. */
|
|
if (loop_info->has_indirect_jump)
|
|
{
|
|
if (loop_dump_stream)
|
|
fprintf (loop_dump_stream,
|
|
"Doloop: Indirect jump in function.\n");
|
|
return 0;
|
|
}
|
|
|
|
/* A called function may clobber any special registers required for
|
|
low-overhead looping. */
|
|
if (loop_info->has_call)
|
|
{
|
|
if (loop_dump_stream)
|
|
fprintf (loop_dump_stream,
|
|
"Doloop: Function call in loop.\n");
|
|
return 0;
|
|
}
|
|
|
|
/* Some targets (eg, PPC) use the count register for branch on table
|
|
instructions. ??? This should be a target specific check. */
|
|
if (loop_info->has_tablejump)
|
|
{
|
|
if (loop_dump_stream)
|
|
fprintf (loop_dump_stream,
|
|
"Doloop: Computed branch in the loop.\n");
|
|
return 0;
|
|
}
|
|
|
|
if (! loop_info->increment)
|
|
{
|
|
if (loop_dump_stream)
|
|
fprintf (loop_dump_stream,
|
|
"Doloop: Could not determine iteration info.\n");
|
|
return 0;
|
|
}
|
|
|
|
if (GET_CODE (loop_info->increment) != CONST_INT)
|
|
{
|
|
if (loop_dump_stream)
|
|
fprintf (loop_dump_stream,
|
|
"Doloop: Increment not an integer constant.\n");
|
|
return 0;
|
|
}
|
|
|
|
/* There is no guarantee that a NE loop will terminate if the
|
|
absolute increment is not unity. ??? We could compute this
|
|
condition at run-time and have an additional jump around the loop
|
|
to ensure an infinite loop. */
|
|
if (loop_info->comparison_code == NE
|
|
&& INTVAL (loop_info->increment) != -1
|
|
&& INTVAL (loop_info->increment) != 1)
|
|
{
|
|
if (loop_dump_stream)
|
|
fprintf (loop_dump_stream,
|
|
"Doloop: NE loop with non-unity increment.\n");
|
|
return 0;
|
|
}
|
|
|
|
/* Check for loops that may not terminate under special conditions. */
|
|
if (! loop_info->n_iterations
|
|
&& ((loop_info->comparison_code == LEU
|
|
&& INTVAL (loop_info->increment) > 0)
|
|
|| (loop_info->comparison_code == GEU
|
|
&& INTVAL (loop_info->increment) < 0)
|
|
|| (loop_info->comparison_code == LTU
|
|
&& INTVAL (loop_info->increment) > 1)
|
|
|| (loop_info->comparison_code == GTU
|
|
&& INTVAL (loop_info->increment) < -1)))
|
|
{
|
|
/* If the comparison is LEU and the comparison value is UINT_MAX
|
|
then the loop will not terminate. Similarly, if the
|
|
comparison code is GEU and the initial value is 0, the loop
|
|
will not terminate.
|
|
|
|
If the absolute increment is not 1, the loop can be infinite
|
|
even with LTU/GTU, e.g. for (i = 3; i > 0; i -= 2)
|
|
|
|
Note that with LE and GE, the loop behaviour is undefined
|
|
(C++ standard section 5 clause 5) if an overflow occurs, say
|
|
between INT_MAX and INT_MAX + 1. We thus don't have to worry
|
|
about these two cases.
|
|
|
|
??? We could compute these conditions at run-time and have a
|
|
additional jump around the loop to ensure an infinite loop.
|
|
However, it is very unlikely that this is the intended
|
|
behaviour of the loop and checking for these rare boundary
|
|
conditions would pessimize all other code.
|
|
|
|
If the loop is executed only a few times an extra check to
|
|
restart the loop could use up most of the benefits of using a
|
|
count register loop. Note however, that normally, this
|
|
restart branch would never execute, so it could be predicted
|
|
well by the CPU. We should generate the pessimistic code by
|
|
default, and have an option, e.g. -funsafe-loops that would
|
|
enable count-register loops in this case. */
|
|
if (loop_dump_stream)
|
|
fprintf (loop_dump_stream,
|
|
"Doloop: Possible infinite iteration case ignored.\n");
|
|
}
|
|
|
|
return 1;
|
|
}
|
|
|
|
|
|
/* Modify the loop to use the low-overhead looping insn where LOOP
|
|
describes the loop, ITERATIONS is an RTX containing the desired
|
|
number of loop iterations, ITERATIONS_MAX is a CONST_INT specifying
|
|
the maximum number of loop iterations, and DOLOOP_INSN is the
|
|
low-overhead looping insn to emit at the end of the loop. This
|
|
returns non-zero if it was successful. */
|
|
static int
|
|
doloop_modify (loop, iterations, iterations_max,
|
|
doloop_seq, start_label, condition)
|
|
const struct loop *loop;
|
|
rtx iterations;
|
|
rtx iterations_max;
|
|
rtx doloop_seq;
|
|
rtx start_label;
|
|
rtx condition;
|
|
{
|
|
rtx counter_reg;
|
|
rtx count;
|
|
rtx sequence;
|
|
rtx jump_insn;
|
|
int nonneg = 0;
|
|
int decrement_count;
|
|
|
|
jump_insn = prev_nonnote_insn (loop->end);
|
|
|
|
if (loop_dump_stream)
|
|
{
|
|
fprintf (loop_dump_stream, "Doloop: Inserting doloop pattern (");
|
|
if (GET_CODE (iterations) == CONST_INT)
|
|
fprintf (loop_dump_stream, HOST_WIDE_INT_PRINT_DEC,
|
|
INTVAL (iterations));
|
|
else
|
|
fputs ("runtime", loop_dump_stream);
|
|
fputs (" iterations).", loop_dump_stream);
|
|
}
|
|
|
|
/* Emit the label that will delimit the top of the loop.
|
|
This has to be done before the delete_insn call below, to prevent
|
|
delete_insn from deleting too much. */
|
|
emit_label_after (start_label, loop->top ? loop->top : loop->start);
|
|
LABEL_NUSES (start_label)++;
|
|
|
|
/* Discard original jump to continue loop. The original compare
|
|
result may still be live, so it cannot be discarded explicitly. */
|
|
delete_related_insns (jump_insn);
|
|
|
|
counter_reg = XEXP (condition, 0);
|
|
if (GET_CODE (counter_reg) == PLUS)
|
|
counter_reg = XEXP (counter_reg, 0);
|
|
|
|
start_sequence ();
|
|
|
|
count = iterations;
|
|
decrement_count = 0;
|
|
switch (GET_CODE (condition))
|
|
{
|
|
case NE:
|
|
/* Currently only NE tests against zero and one are supported. */
|
|
if (XEXP (condition, 1) == const0_rtx)
|
|
decrement_count = 1;
|
|
else if (XEXP (condition, 1) != const1_rtx)
|
|
abort ();
|
|
break;
|
|
|
|
case GE:
|
|
/* Currently only GE tests against zero are supported. */
|
|
if (XEXP (condition, 1) != const0_rtx)
|
|
abort ();
|
|
|
|
/* The iteration count needs decrementing for a GE test. */
|
|
decrement_count = 1;
|
|
|
|
/* Determine if the iteration counter will be non-negative.
|
|
Note that the maximum value loaded is iterations_max - 1. */
|
|
if ((unsigned HOST_WIDE_INT) INTVAL (iterations_max)
|
|
<= ((unsigned) 1 << (GET_MODE_BITSIZE (GET_MODE (counter_reg)) - 1)))
|
|
nonneg = 1;
|
|
break;
|
|
|
|
/* Abort if an invalid doloop pattern has been generated. */
|
|
default:
|
|
abort ();
|
|
}
|
|
|
|
if (decrement_count)
|
|
{
|
|
if (GET_CODE (count) == CONST_INT)
|
|
count = GEN_INT (INTVAL (count) - 1);
|
|
else
|
|
count = expand_simple_binop (GET_MODE (counter_reg), MINUS,
|
|
count, GEN_INT (1),
|
|
0, 0, OPTAB_LIB_WIDEN);
|
|
}
|
|
|
|
/* Insert initialization of the count register into the loop header. */
|
|
convert_move (counter_reg, count, 1);
|
|
sequence = gen_sequence ();
|
|
end_sequence ();
|
|
emit_insn_before (sequence, loop->start);
|
|
|
|
/* Some targets (eg, C4x) need to initialize special looping
|
|
registers. */
|
|
#ifdef HAVE_doloop_begin
|
|
{
|
|
rtx init;
|
|
|
|
init = gen_doloop_begin (counter_reg,
|
|
GET_CODE (iterations) == CONST_INT
|
|
? iterations : const0_rtx, iterations_max,
|
|
GEN_INT (loop->level));
|
|
if (init)
|
|
{
|
|
start_sequence ();
|
|
emit_insn (init);
|
|
sequence = gen_sequence ();
|
|
end_sequence ();
|
|
emit_insn_after (sequence, loop->start);
|
|
}
|
|
}
|
|
#endif
|
|
|
|
/* Insert the new low-overhead looping insn. */
|
|
emit_jump_insn_before (doloop_seq, loop->end);
|
|
jump_insn = prev_nonnote_insn (loop->end);
|
|
JUMP_LABEL (jump_insn) = start_label;
|
|
|
|
/* Add a REG_NONNEG note if the actual or estimated maximum number
|
|
of iterations is non-negative. */
|
|
if (nonneg)
|
|
{
|
|
REG_NOTES (jump_insn)
|
|
= gen_rtx_EXPR_LIST (REG_NONNEG, NULL_RTX, REG_NOTES (jump_insn));
|
|
}
|
|
return 1;
|
|
}
|
|
|
|
|
|
/* Handle the more complex case, where the bounds are not known at
|
|
compile time. In this case we generate a run_time calculation of
|
|
the number of iterations. We rely on the existence of a run-time
|
|
guard to ensure that the loop executes at least once, i.e.,
|
|
initial_value obeys the loop comparison condition. If a guard is
|
|
not present, we emit one. The loop to modify is described by LOOP.
|
|
ITERATIONS_MAX is a CONST_INT specifying the estimated maximum
|
|
number of loop iterations. DOLOOP_INSN is the low-overhead looping
|
|
insn to insert. Returns non-zero if loop successfully modified. */
|
|
static int
|
|
doloop_modify_runtime (loop, iterations_max,
|
|
doloop_seq, start_label, mode, condition)
|
|
const struct loop *loop;
|
|
rtx iterations_max;
|
|
rtx doloop_seq;
|
|
rtx start_label;
|
|
enum machine_mode mode;
|
|
rtx condition;
|
|
{
|
|
const struct loop_info *loop_info = LOOP_INFO (loop);
|
|
HOST_WIDE_INT abs_inc;
|
|
HOST_WIDE_INT abs_loop_inc;
|
|
int neg_inc;
|
|
rtx diff;
|
|
rtx sequence;
|
|
rtx iterations;
|
|
rtx initial_value;
|
|
rtx final_value;
|
|
rtx increment;
|
|
int unsigned_p;
|
|
enum rtx_code comparison_code;
|
|
|
|
increment = loop_info->increment;
|
|
initial_value = loop_info->initial_value;
|
|
final_value = loop_info->final_value;
|
|
|
|
neg_inc = 0;
|
|
abs_inc = INTVAL (increment);
|
|
if (abs_inc < 0)
|
|
{
|
|
abs_inc = -abs_inc;
|
|
neg_inc = 1;
|
|
}
|
|
|
|
comparison_code = loop_info->comparison_code;
|
|
unsigned_p = (comparison_code == LTU
|
|
|| comparison_code == LEU
|
|
|| comparison_code == GTU
|
|
|| comparison_code == GEU
|
|
|| comparison_code == NE);
|
|
|
|
/* The number of iterations (prior to any loop unrolling) is given by:
|
|
|
|
n = (abs (final - initial) + abs_inc - 1) / abs_inc.
|
|
|
|
However, it is possible for the summation to overflow, and a
|
|
safer method is:
|
|
|
|
n = abs (final - initial) / abs_inc;
|
|
n += (abs (final - initial) % abs_inc) != 0;
|
|
|
|
But when abs_inc is a power of two, the summation won't overflow
|
|
except in cases where the loop never terminates. So we don't
|
|
need to use this more costly calculation.
|
|
|
|
If the loop has been unrolled, the full calculation is
|
|
|
|
t1 = abs_inc * unroll_number; increment per loop
|
|
n = abs (final - initial) / t1; full loops
|
|
n += (abs (final - initial) % t1) != 0; partial loop
|
|
|
|
However, in certain cases the unrolled loop will be preconditioned
|
|
by emitting copies of the loop body with conditional branches,
|
|
so that the unrolled loop is always a full loop and thus needs
|
|
no exit tests. In this case we don't want to add the partial
|
|
loop count. As above, when t1 is a power of two we don't need to
|
|
worry about overflow.
|
|
|
|
The division and modulo operations can be avoided by requiring
|
|
that the increment is a power of 2 (precondition_loop_p enforces
|
|
this requirement). Nevertheless, the RTX_COSTS should be checked
|
|
to see if a fast divmod is available. */
|
|
|
|
start_sequence ();
|
|
/* abs (final - initial) */
|
|
diff = expand_simple_binop (mode, MINUS,
|
|
copy_rtx (neg_inc ? initial_value : final_value),
|
|
copy_rtx (neg_inc ? final_value : initial_value),
|
|
NULL_RTX, unsigned_p, OPTAB_LIB_WIDEN);
|
|
|
|
/* Some code transformations can result in code akin to
|
|
|
|
tmp = i + 1;
|
|
...
|
|
goto scan_start;
|
|
top:
|
|
tmp = tmp + 1;
|
|
scan_start:
|
|
i = tmp;
|
|
if (i < n) goto top;
|
|
|
|
We'll have already detected this form of loop in scan_loop,
|
|
and set loop->top and loop->scan_start appropriately.
|
|
|
|
In this situation, we skip the increment the first time through
|
|
the loop, which results in an incorrect estimate of the number
|
|
of iterations. Adjust the difference to compensate. */
|
|
/* ??? Logically, it would seem this belongs in loop_iterations.
|
|
However, this causes regressions e.g. on x86 execute/20011008-3.c,
|
|
so I do not believe we've properly characterized the exact nature
|
|
of the problem. In the meantime, this fixes execute/20011126-2.c
|
|
on ia64 and some Ada front end miscompilation on ppc. */
|
|
|
|
if (loop->scan_start)
|
|
{
|
|
rtx iteration_var = loop_info->iteration_var;
|
|
struct loop_ivs *ivs = LOOP_IVS (loop);
|
|
struct iv_class *bl;
|
|
|
|
if (REG_IV_TYPE (ivs, REGNO (iteration_var)) == BASIC_INDUCT)
|
|
bl = REG_IV_CLASS (ivs, REGNO (iteration_var));
|
|
else if (REG_IV_TYPE (ivs, REGNO (iteration_var)) == GENERAL_INDUCT)
|
|
{
|
|
struct induction *v = REG_IV_INFO (ivs, REGNO (iteration_var));
|
|
bl = REG_IV_CLASS (ivs, REGNO (v->src_reg));
|
|
}
|
|
else
|
|
/* Iteration var must be an induction variable to get here. */
|
|
abort ();
|
|
|
|
if (INSN_UID (bl->biv->insn) < max_uid_for_loop
|
|
&& INSN_LUID (bl->biv->insn) < INSN_LUID (loop->scan_start))
|
|
{
|
|
if (loop_dump_stream)
|
|
fprintf (loop_dump_stream,
|
|
"Doloop: Basic induction var skips initial incr.\n");
|
|
|
|
diff = expand_simple_binop (mode, PLUS, diff, GEN_INT (abs_inc),
|
|
diff, unsigned_p, OPTAB_LIB_WIDEN);
|
|
}
|
|
}
|
|
|
|
abs_loop_inc = abs_inc * loop_info->unroll_number;
|
|
if (abs_loop_inc != 1)
|
|
{
|
|
int shift_count;
|
|
|
|
shift_count = exact_log2 (abs_loop_inc);
|
|
if (shift_count < 0)
|
|
abort ();
|
|
|
|
if (!loop_info->preconditioned)
|
|
diff = expand_simple_binop (GET_MODE (diff), PLUS,
|
|
diff, GEN_INT (abs_loop_inc - 1),
|
|
diff, 1, OPTAB_LIB_WIDEN);
|
|
|
|
/* (abs (final - initial) + abs_inc * unroll_number - 1)
|
|
/ (abs_inc * unroll_number) */
|
|
diff = expand_simple_binop (GET_MODE (diff), LSHIFTRT,
|
|
diff, GEN_INT (shift_count),
|
|
diff, 1, OPTAB_LIB_WIDEN);
|
|
}
|
|
iterations = diff;
|
|
|
|
/* If there is a NOTE_INSN_LOOP_VTOP, we have a `for' or `while'
|
|
style loop, with a loop exit test at the start. Thus, we can
|
|
assume that the loop condition was true when the loop was
|
|
entered.
|
|
|
|
`do-while' loops require special treatment since the exit test is
|
|
not executed before the start of the loop. We need to determine
|
|
if the loop will terminate after the first pass and to limit the
|
|
iteration count to one if necessary. */
|
|
if (! loop->vtop)
|
|
{
|
|
if (loop_dump_stream)
|
|
fprintf (loop_dump_stream, "Doloop: Do-while loop.\n");
|
|
|
|
/* A `do-while' loop must iterate at least once. For code like
|
|
i = initial; do { ... } while (++i < final);
|
|
we will calculate a bogus iteration count if initial > final.
|
|
So detect this and set the iteration count to 1.
|
|
Note that if the loop has been unrolled, then the loop body
|
|
is guaranteed to execute at least once. Also, when the
|
|
comparison is NE, our calculated count will be OK. */
|
|
if (loop_info->unroll_number == 1 && comparison_code != NE)
|
|
{
|
|
rtx label;
|
|
|
|
/* Emit insns to test if the loop will immediately
|
|
terminate and to set the iteration count to 1 if true. */
|
|
label = gen_label_rtx();
|
|
emit_cmp_and_jump_insns (copy_rtx (initial_value),
|
|
copy_rtx (loop_info->comparison_value),
|
|
comparison_code, NULL_RTX, mode, 0,
|
|
label);
|
|
JUMP_LABEL (get_last_insn ()) = label;
|
|
LABEL_NUSES (label)++;
|
|
emit_move_insn (iterations, const1_rtx);
|
|
emit_label (label);
|
|
}
|
|
}
|
|
|
|
sequence = gen_sequence ();
|
|
end_sequence ();
|
|
emit_insn_before (sequence, loop->start);
|
|
|
|
return doloop_modify (loop, iterations, iterations_max, doloop_seq,
|
|
start_label, condition);
|
|
}
|
|
|
|
|
|
/* This is the main entry point. Process loop described by LOOP
|
|
validating that the loop is suitable for conversion to use a low
|
|
overhead looping instruction, replacing the jump insn where
|
|
suitable. We distinguish between loops with compile-time bounds
|
|
and those with run-time bounds. Information from LOOP is used to
|
|
compute the number of iterations and to determine whether the loop
|
|
is a candidate for this optimization. Returns non-zero if loop
|
|
successfully modified. */
|
|
int
|
|
doloop_optimize (loop)
|
|
const struct loop *loop;
|
|
{
|
|
struct loop_info *loop_info = LOOP_INFO (loop);
|
|
rtx initial_value;
|
|
rtx final_value;
|
|
rtx increment;
|
|
rtx jump_insn;
|
|
enum machine_mode mode;
|
|
unsigned HOST_WIDE_INT n_iterations;
|
|
unsigned HOST_WIDE_INT n_iterations_max;
|
|
rtx doloop_seq, doloop_pat, doloop_reg;
|
|
rtx iterations;
|
|
rtx iterations_max;
|
|
rtx start_label;
|
|
rtx condition;
|
|
|
|
if (loop_dump_stream)
|
|
fprintf (loop_dump_stream,
|
|
"Doloop: Processing loop %d, enclosed levels %d.\n",
|
|
loop->num, loop->level);
|
|
|
|
jump_insn = prev_nonnote_insn (loop->end);
|
|
|
|
/* Check that loop is a candidate for a low-overhead looping insn. */
|
|
if (! doloop_valid_p (loop, jump_insn))
|
|
return 0;
|
|
|
|
/* Determine if the loop can be safely, and profitably,
|
|
preconditioned. While we don't precondition the loop in a loop
|
|
unrolling sense, this test ensures that the loop is well behaved
|
|
and that the increment is a constant integer. */
|
|
if (! precondition_loop_p (loop, &initial_value, &final_value,
|
|
&increment, &mode))
|
|
{
|
|
if (loop_dump_stream)
|
|
fprintf (loop_dump_stream,
|
|
"Doloop: Cannot precondition loop.\n");
|
|
return 0;
|
|
}
|
|
|
|
/* Determine or estimate the maximum number of loop iterations. */
|
|
n_iterations = loop_info->n_iterations;
|
|
if (n_iterations)
|
|
{
|
|
/* This is the simple case where the initial and final loop
|
|
values are constants. */
|
|
n_iterations_max = n_iterations;
|
|
}
|
|
else
|
|
{
|
|
int nonneg = find_reg_note (jump_insn, REG_NONNEG, 0) != 0;
|
|
|
|
/* This is the harder case where the initial and final loop
|
|
values may not be constants. */
|
|
n_iterations_max = doloop_iterations_max (loop_info, mode, nonneg);
|
|
|
|
if (! n_iterations_max)
|
|
{
|
|
/* We have something like `for (i = 0; i < 10; i--)'. */
|
|
if (loop_dump_stream)
|
|
fprintf (loop_dump_stream,
|
|
"Doloop: Not normal loop.\n");
|
|
return 0;
|
|
}
|
|
}
|
|
|
|
/* Account for loop unrolling in the iteration count. This will
|
|
have no effect if loop_iterations could not determine the number
|
|
of iterations. */
|
|
n_iterations /= loop_info->unroll_number;
|
|
n_iterations_max /= loop_info->unroll_number;
|
|
|
|
if (n_iterations && n_iterations < 3)
|
|
{
|
|
if (loop_dump_stream)
|
|
fprintf (loop_dump_stream,
|
|
"Doloop: Too few iterations (%ld) to be profitable.\n",
|
|
(long int) n_iterations);
|
|
return 0;
|
|
}
|
|
|
|
iterations = GEN_INT (n_iterations);
|
|
iterations_max = GEN_INT (n_iterations_max);
|
|
|
|
/* Generate looping insn. If the pattern FAILs then give up trying
|
|
to modify the loop since there is some aspect the back-end does
|
|
not like. */
|
|
start_label = gen_label_rtx ();
|
|
doloop_reg = gen_reg_rtx (mode);
|
|
doloop_seq = gen_doloop_end (doloop_reg, iterations, iterations_max,
|
|
GEN_INT (loop->level), start_label);
|
|
if (! doloop_seq && mode != word_mode)
|
|
{
|
|
PUT_MODE (doloop_reg, word_mode);
|
|
doloop_seq = gen_doloop_end (doloop_reg, iterations, iterations_max,
|
|
GEN_INT (loop->level), start_label);
|
|
}
|
|
if (! doloop_seq)
|
|
{
|
|
if (loop_dump_stream)
|
|
fprintf (loop_dump_stream,
|
|
"Doloop: Target unwilling to use doloop pattern!\n");
|
|
return 0;
|
|
}
|
|
|
|
/* A raw define_insn may yield a plain pattern. If a sequence
|
|
was involved, the last must be the jump instruction. */
|
|
if (GET_CODE (doloop_seq) == SEQUENCE)
|
|
{
|
|
doloop_pat = XVECEXP (doloop_seq, 0, XVECLEN (doloop_seq, 0) - 1);
|
|
if (GET_CODE (doloop_pat) == JUMP_INSN)
|
|
doloop_pat = PATTERN (doloop_pat);
|
|
else
|
|
doloop_pat = NULL_RTX;
|
|
}
|
|
else
|
|
doloop_pat = doloop_seq;
|
|
|
|
if (! doloop_pat
|
|
|| ! (condition = doloop_condition_get (doloop_pat)))
|
|
{
|
|
if (loop_dump_stream)
|
|
fprintf (loop_dump_stream,
|
|
"Doloop: Unrecognizable doloop pattern!\n");
|
|
return 0;
|
|
}
|
|
|
|
if (n_iterations != 0)
|
|
/* Handle the simpler case, where we know the iteration count at
|
|
compile time. */
|
|
return doloop_modify (loop, iterations, iterations_max, doloop_seq,
|
|
start_label, condition);
|
|
else
|
|
/* Handle the harder case, where we must add additional runtime tests. */
|
|
return doloop_modify_runtime (loop, iterations_max, doloop_seq,
|
|
start_label, mode, condition);
|
|
}
|
|
|
|
#endif /* HAVE_doloop_end */
|