This file is consumed by genmatch which produces gimple-match.c
and generic-match.c from it.
- Copyright (C) 2014-2018 Free Software Foundation, Inc.
+ Copyright (C) 2014-2019 Free Software Foundation, Inc.
Contributed by Richard Biener <rguenther@suse.de>
and Prathamesh Kulkarni <bilbotheelffriend@gmail.com>
integer_each_onep integer_truep integer_nonzerop
real_zerop real_onep real_minus_onep
zerop
+ initializer_each_zero_or_onep
CONSTANT_CLASS_P
tree_expr_nonnegative_p
tree_expr_nonzero_p
integer_valued_real_p
integer_pow2p
- HONOR_NANS)
+ uniform_integer_cst_p
+ HONOR_NANS
+ uniform_vector_p)
/* Operator lists. */
(define_operator_list tcc_comparison
DEFINE_INT_AND_FLOAT_ROUND_FN (CEIL)
DEFINE_INT_AND_FLOAT_ROUND_FN (ROUND)
DEFINE_INT_AND_FLOAT_ROUND_FN (RINT)
-
+
+/* Binary operations and their associated IFN_COND_* function. */
+(define_operator_list UNCOND_BINARY
+ plus minus
+ mult trunc_div trunc_mod rdiv
+ min max
+ bit_and bit_ior bit_xor
+ lshift rshift)
+(define_operator_list COND_BINARY
+ IFN_COND_ADD IFN_COND_SUB
+ IFN_COND_MUL IFN_COND_DIV IFN_COND_MOD IFN_COND_RDIV
+ IFN_COND_MIN IFN_COND_MAX
+ IFN_COND_AND IFN_COND_IOR IFN_COND_XOR
+ IFN_COND_SHL IFN_COND_SHR)
+
+/* Same for ternary operations. */
+(define_operator_list UNCOND_TERNARY
+ IFN_FMA IFN_FMS IFN_FNMA IFN_FNMS)
+(define_operator_list COND_TERNARY
+ IFN_COND_FMA IFN_COND_FMS IFN_COND_FNMA IFN_COND_FNMS)
+
/* As opposed to convert?, this still creates a single pattern, so
it is not a suitable replacement for convert? in all cases. */
(match (nop_convert @0)
&& tree_nop_conversion_p (TREE_TYPE (type), TREE_TYPE (TREE_TYPE (@0))))))
/* This one has to be last, or it shadows the others. */
(match (nop_convert @0)
- @0)
+ @0)
+
+/* Transform likes of (char) ABS_EXPR <(int) x> into (char) ABSU_EXPR <x>
+ ABSU_EXPR returns unsigned absolute value of the operand and the operand
+ of the ABSU_EXPR will have the corresponding signed type. */
+(simplify (abs (convert @0))
+ (if (ANY_INTEGRAL_TYPE_P (TREE_TYPE (@0))
+ && !TYPE_UNSIGNED (TREE_TYPE (@0))
+ && element_precision (type) > element_precision (TREE_TYPE (@0)))
+ (with { tree utype = unsigned_type_for (TREE_TYPE (@0)); }
+ (convert (absu:utype @0)))))
+
/* Simplifications of operations with one constant operand and
simplifications to constants or single values. */
(if (fold_real_zero_addition_p (type, @1, 1))
(non_lvalue @0)))
+/* Even if the fold_real_zero_addition_p can't simplify X + 0.0
+ into X, we can optimize (X + 0.0) + 0.0 or (X + 0.0) - 0.0
+ or (X - 0.0) + 0.0 into X + 0.0 and (X - 0.0) - 0.0 into X - 0.0
+ if not -frounding-math. For sNaNs the first operation would raise
+ exceptions but turn the result into qNan, so the second operation
+ would not raise it. */
+(for inner_op (plus minus)
+ (for outer_op (plus minus)
+ (simplify
+ (outer_op (inner_op@3 @0 REAL_CST@1) REAL_CST@2)
+ (if (real_zerop (@1)
+ && real_zerop (@2)
+ && !HONOR_SIGN_DEPENDENT_ROUNDING (type))
+ (with { bool inner_plus = ((inner_op == PLUS_EXPR)
+ ^ REAL_VALUE_MINUS_ZERO (TREE_REAL_CST (@1)));
+ bool outer_plus
+ = ((outer_op == PLUS_EXPR)
+ ^ REAL_VALUE_MINUS_ZERO (TREE_REAL_CST (@2))); }
+ (if (outer_plus && !inner_plus)
+ (outer_op @0 @2)
+ @3))))))
+
/* Simplify x - x.
This is unsafe for certain floats even in non-IEEE formats.
In IEEE, it is unsafe because it does wrong for NaNs.
|| !COMPLEX_FLOAT_TYPE_P (type)))
(negate @0)))
+/* Transform { 0 or 1 } * { 0 or 1 } into { 0 or 1 } & { 0 or 1 } */
+(simplify
+ (mult SSA_NAME@1 SSA_NAME@2)
+ (if (INTEGRAL_TYPE_P (type)
+ && get_nonzero_bits (@1) == 1
+ && get_nonzero_bits (@2) == 1)
+ (bit_and @1 @2)))
+
+/* Transform x * { 0 or 1, 0 or 1, ... } into x & { 0 or -1, 0 or -1, ...},
+ unless the target has native support for the former but not the latter. */
+(simplify
+ (mult @0 VECTOR_CST@1)
+ (if (initializer_each_zero_or_onep (@1)
+ && !HONOR_SNANS (type)
+ && !HONOR_SIGNED_ZEROS (type))
+ (with { tree itype = FLOAT_TYPE_P (type) ? unsigned_type_for (type) : type; }
+ (if (itype
+ && (!VECTOR_MODE_P (TYPE_MODE (type))
+ || (VECTOR_MODE_P (TYPE_MODE (itype))
+ && optab_handler (and_optab,
+ TYPE_MODE (itype)) != CODE_FOR_nothing)))
+ (view_convert (bit_and:itype (view_convert @0)
+ (ne @1 { build_zero_cst (type); })))))))
+
(for cmp (gt ge lt le)
outp (convert convert negate negate)
outn (negate negate convert convert)
And not for _Fract types where we can't build 1. */
(if (!integer_zerop (@0) && !ALL_FRACT_MODE_P (TYPE_MODE (type)))
{ build_one_cst (type); }))
- /* X / abs (X) is X < 0 ? -1 : 1. */
+ /* X / abs (X) is X < 0 ? -1 : 1. */
(simplify
(div:C @0 (abs @0))
(if (INTEGRAL_TYPE_P (type)
and floor_div is trickier and combining round_div even more so. */
(for div (trunc_div exact_div)
(simplify
- (div (div @0 INTEGER_CST@1) INTEGER_CST@2)
+ (div (div@3 @0 INTEGER_CST@1) INTEGER_CST@2)
(with {
- bool overflow_p;
+ wi::overflow_type overflow;
wide_int mul = wi::mul (wi::to_wide (@1), wi::to_wide (@2),
- TYPE_SIGN (type), &overflow_p);
+ TYPE_SIGN (type), &overflow);
}
- (if (!overflow_p)
- (div @0 { wide_int_to_tree (type, mul); })
- (if (TYPE_UNSIGNED (type)
- || mul != wi::min_value (TYPE_PRECISION (type), SIGNED))
- { build_zero_cst (type); })))))
+ (if (div == EXACT_DIV_EXPR
+ || optimize_successive_divisions_p (@2, @3))
+ (if (!overflow)
+ (div @0 { wide_int_to_tree (type, mul); })
+ (if (TYPE_UNSIGNED (type)
+ || mul != wi::min_value (TYPE_PRECISION (type), SIGNED))
+ { build_zero_cst (type); }))))))
/* Combine successive multiplications. Similar to above, but handling
overflow is different. */
(simplify
(mult (mult @0 INTEGER_CST@1) INTEGER_CST@2)
(with {
- bool overflow_p;
+ wi::overflow_type overflow;
wide_int mul = wi::mul (wi::to_wide (@1), wi::to_wide (@2),
- TYPE_SIGN (type), &overflow_p);
+ TYPE_SIGN (type), &overflow);
}
/* Skip folding on overflow: the only special case is @1 * @2 == -INT_MIN,
otherwise undefined overflow implies that @0 must be zero. */
- (if (!overflow_p || TYPE_OVERFLOW_WRAPS (type))
+ (if (!overflow || TYPE_OVERFLOW_WRAPS (type))
(mult @0 { wide_int_to_tree (type, mul); }))))
/* Optimize A / A to 1.0 if we don't care about
(rdiv @0 (negate @1))
(rdiv (negate @0) @1))
+(if (flag_unsafe_math_optimizations)
+ /* Simplify (C / x op 0.0) to x op 0.0 for C != 0, C != Inf/Nan.
+ Since C / x may underflow to zero, do this only for unsafe math. */
+ (for op (lt le gt ge)
+ neg_op (gt ge lt le)
+ (simplify
+ (op (rdiv REAL_CST@0 @1) real_zerop@2)
+ (if (!HONOR_SIGNED_ZEROS (@1) && !HONOR_INFINITIES (@1))
+ (switch
+ (if (real_less (&dconst0, TREE_REAL_CST_PTR (@0)))
+ (op @1 @2))
+ /* For C < 0, use the inverted operator. */
+ (if (real_less (TREE_REAL_CST_PTR (@0), &dconst0))
+ (neg_op @1 @2)))))))
+
/* Optimize (X & (-A)) / A where A is a power of 2, to X >> log2(A) */
(for div (trunc_div ceil_div floor_div round_div exact_div)
(simplify
&& TYPE_OVERFLOW_UNDEFINED (type)
&& wi::multiple_of_p (wi::to_wide (@1), wi::to_wide (@2),
TYPE_SIGN (type)))
- { build_zero_cst (type); })))
+ { build_zero_cst (type); }))
+ /* For (X % C) == 0, if X is signed and C is power of 2, use unsigned
+ modulo and comparison, since it is simpler and equivalent. */
+ (for cmp (eq ne)
+ (simplify
+ (cmp (mod @0 integer_pow2p@2) integer_zerop@1)
+ (if (!TYPE_UNSIGNED (TREE_TYPE (@0)))
+ (with { tree utype = unsigned_type_for (TREE_TYPE (@0)); }
+ (cmp (mod (convert:utype @0) (convert:utype @2)) (convert:utype @1)))))))
/* X % -C is the same as X % C. */
(simplify
/* Simplify (t * 2) / 2) -> t. */
(for div (trunc_div ceil_div floor_div round_div exact_div)
(simplify
- (div (mult @0 @1) @1)
+ (div (mult:c @0 @1) @1)
(if (ANY_INTEGRAL_TYPE_P (type)
&& TYPE_OVERFLOW_UNDEFINED (type))
@0)))
(mult (abs@1 @0) @1)
(mult @0 @0))
+/* Convert absu(x)*absu(x) -> x*x. */
+(simplify
+ (mult (absu@1 @0) @1)
+ (mult (convert@2 @0) @2))
+
/* cos(copysign(x, y)) -> cos(x). Similarly for cosh. */
(for coss (COS COSH)
copysigns (COPYSIGN)
(bit_not (bit_and:cs (bit_not @0) @1))
(bit_ior @0 (bit_not @1)))
+/* ~(~a | b) --> a & ~b */
+(simplify
+ (bit_not (bit_ior:cs (bit_not @0) @1))
+ (bit_and @0 (bit_not @1)))
+
/* Simplify (~X & Y) to X ^ Y if we know that (X & ~Y) is 0. */
#if GIMPLE
(simplify
(bit_xor @0 @1)))
#endif
+/* For constants M and N, if M == (1LL << cst) - 1 && (N & M) == M,
+ ((A & N) + B) & M -> (A + B) & M
+ Similarly if (N & M) == 0,
+ ((A | N) + B) & M -> (A + B) & M
+ and for - instead of + (or unary - instead of +)
+ and/or ^ instead of |.
+ If B is constant and (B & M) == 0, fold into A & M. */
+(for op (plus minus)
+ (for bitop (bit_and bit_ior bit_xor)
+ (simplify
+ (bit_and (op:s (bitop:s@0 @3 INTEGER_CST@4) @1) INTEGER_CST@2)
+ (with
+ { tree pmop[2];
+ tree utype = fold_bit_and_mask (TREE_TYPE (@0), @2, op, @0, bitop,
+ @3, @4, @1, ERROR_MARK, NULL_TREE,
+ NULL_TREE, pmop); }
+ (if (utype)
+ (convert (bit_and (op (convert:utype { pmop[0]; })
+ (convert:utype { pmop[1]; }))
+ (convert:utype @2))))))
+ (simplify
+ (bit_and (op:s @0 (bitop:s@1 @3 INTEGER_CST@4)) INTEGER_CST@2)
+ (with
+ { tree pmop[2];
+ tree utype = fold_bit_and_mask (TREE_TYPE (@0), @2, op, @0, ERROR_MARK,
+ NULL_TREE, NULL_TREE, @1, bitop, @3,
+ @4, pmop); }
+ (if (utype)
+ (convert (bit_and (op (convert:utype { pmop[0]; })
+ (convert:utype { pmop[1]; }))
+ (convert:utype @2)))))))
+ (simplify
+ (bit_and (op:s @0 @1) INTEGER_CST@2)
+ (with
+ { tree pmop[2];
+ tree utype = fold_bit_and_mask (TREE_TYPE (@0), @2, op, @0, ERROR_MARK,
+ NULL_TREE, NULL_TREE, @1, ERROR_MARK,
+ NULL_TREE, NULL_TREE, pmop); }
+ (if (utype)
+ (convert (bit_and (op (convert:utype { pmop[0]; })
+ (convert:utype { pmop[1]; }))
+ (convert:utype @2)))))))
+(for bitop (bit_and bit_ior bit_xor)
+ (simplify
+ (bit_and (negate:s (bitop:s@0 @2 INTEGER_CST@3)) INTEGER_CST@1)
+ (with
+ { tree pmop[2];
+ tree utype = fold_bit_and_mask (TREE_TYPE (@0), @1, NEGATE_EXPR, @0,
+ bitop, @2, @3, NULL_TREE, ERROR_MARK,
+ NULL_TREE, NULL_TREE, pmop); }
+ (if (utype)
+ (convert (bit_and (negate (convert:utype { pmop[0]; }))
+ (convert:utype @1)))))))
+
/* X % Y is smaller than Y. */
(for cmp (lt ge)
(simplify
(bitop:c @0 (bit_not (bitop:cs @0 @1)))
(bitop @0 (bit_not @1))))
+/* (~x & y) | ~(x | y) -> ~x */
+(simplify
+ (bit_ior:c (bit_and:c (bit_not@2 @0) @1) (bit_not (bit_ior:c @0 @1)))
+ @2)
+
+/* (x | y) ^ (x | ~y) -> ~x */
+(simplify
+ (bit_xor:c (bit_ior:c @0 @1) (bit_ior:c @0 (bit_not @1)))
+ (bit_not @0))
+
+/* (x & y) | ~(x | y) -> ~(x ^ y) */
+(simplify
+ (bit_ior:c (bit_and:s @0 @1) (bit_not:s (bit_ior:s @0 @1)))
+ (bit_not (bit_xor @0 @1)))
+
+/* (~x | y) ^ (x ^ y) -> x | ~y */
+(simplify
+ (bit_xor:c (bit_ior:cs (bit_not @0) @1) (bit_xor:s @0 @1))
+ (bit_ior @0 (bit_not @1)))
+
+/* (x ^ y) | ~(x | y) -> ~(x & y) */
+(simplify
+ (bit_ior:c (bit_xor:s @0 @1) (bit_not:s (bit_ior:s @0 @1)))
+ (bit_not (bit_and @0 @1)))
+
/* (x | y) & ~x -> y & ~x */
/* (x & y) | ~x -> y | ~x */
(for bitop (bit_and bit_ior)
(bit_and:c (bit_ior:c @0 @1) (bit_xor:c @1 (bit_not @0)))
(bit_and @0 @1))
+/* (~x | y) & (x | ~y) -> ~(x ^ y) */
+(simplify
+ (bit_and (bit_ior:cs (bit_not @0) @1) (bit_ior:cs @0 (bit_not @1)))
+ (bit_not (bit_xor @0 @1)))
+
+/* (~x | y) ^ (x | ~y) -> x ^ y */
+(simplify
+ (bit_xor (bit_ior:c (bit_not @0) @1) (bit_ior:c @0 (bit_not @1)))
+ (bit_xor @0 @1))
+
/* ~x & ~y -> ~(x | y)
~x | ~y -> ~(x & y) */
(for op (bit_and bit_ior)
(for opo (bit_and bit_xor)
opi (bit_xor bit_and)
(simplify
- (opo:c (opi:c @0 @1) @1)
+ (opo:c (opi:cs @0 @1) @1)
(bit_and (bit_not @0) @1)))
/* Given a bit-wise operation CODE applied to ARG0 and ARG1, see if both
&& tree_nop_conversion_p (type, TREE_TYPE (@2)))
(bit_xor (convert @1) (convert @2))))
+/* Convert abs (abs (X)) into abs (X).
+ also absu (absu (X)) into absu (X). */
(simplify
(abs (abs@1 @0))
@1)
+
+(simplify
+ (absu (convert@2 (absu@1 @0)))
+ (if (tree_nop_conversion_p (TREE_TYPE (@2), TREE_TYPE (@1)))
+ @1))
+
+/* Convert abs[u] (-X) -> abs[u] (X). */
(simplify
(abs (negate @0))
(abs @0))
+
+(simplify
+ (absu (negate @0))
+ (absu @0))
+
+/* Convert abs[u] (X) where X is nonnegative -> (X). */
(simplify
(abs tree_expr_nonnegative_p@0)
@0)
+(simplify
+ (absu tree_expr_nonnegative_p@0)
+ (convert @0))
+
/* A few cases of fold-const.c negate_expr_p predicate. */
(match negate_expr_p
INTEGER_CST
(if (tree_nop_conversion_p (type, TREE_TYPE (@0))
&& tree_nop_conversion_p (type, TREE_TYPE (@1)))
(mult (convert @0) (convert (negate @1)))))
-
+
/* -(A + B) -> (-B) - A. */
(simplify
(negate (plus:c @0 negate_expr_p@1))
/* X / 4 < Y / 4 iff X < Y when the division is known to be exact. */
(for cmp (simple_comparison)
(simplify
- (cmp (exact_div @0 INTEGER_CST@2) (exact_div @1 @2))
- (if (wi::gt_p (wi::to_wide (@2), 0, TYPE_SIGN (TREE_TYPE (@2))))
- (cmp @0 @1))))
+ (cmp (convert?@3 (exact_div @0 INTEGER_CST@2)) (convert? (exact_div @1 @2)))
+ (if (element_precision (@3) >= element_precision (@0)
+ && types_match (@0, @1))
+ (if (wi::lt_p (wi::to_wide (@2), 0, TYPE_SIGN (TREE_TYPE (@2))))
+ (if (!TYPE_UNSIGNED (TREE_TYPE (@3)))
+ (cmp @1 @0)
+ (if (tree_expr_nonzero_p (@0) && tree_expr_nonzero_p (@1))
+ (with
+ {
+ tree utype = unsigned_type_for (TREE_TYPE (@0));
+ }
+ (cmp (convert:utype @1) (convert:utype @0)))))
+ (if (wi::gt_p (wi::to_wide (@2), 1, TYPE_SIGN (TREE_TYPE (@2))))
+ (if (TYPE_UNSIGNED (TREE_TYPE (@0)) || !TYPE_UNSIGNED (TREE_TYPE (@3)))
+ (cmp @0 @1)
+ (with
+ {
+ tree utype = unsigned_type_for (TREE_TYPE (@0));
+ }
+ (cmp (convert:utype @0) (convert:utype @1)))))))))
/* X / C1 op C2 into a simple range test. */
(for cmp (simple_comparison)
(op:c (plus:c@2 @0 @1) @1)
(if (ANY_INTEGRAL_TYPE_P (TREE_TYPE (@0))
&& TYPE_OVERFLOW_UNDEFINED (TREE_TYPE (@0))
+ && !TYPE_OVERFLOW_SANITIZED (TREE_TYPE (@0))
&& (CONSTANT_CLASS_P (@0) || single_use (@2)))
(op @0 { build_zero_cst (TREE_TYPE (@0)); }))))
/* For equality, this is also true with wrapping overflow. */
(op @1 { build_zero_cst (TREE_TYPE (@1)); }))))
/* Transform:
- * (X / Y) == 0 -> X < Y if X, Y are unsigned.
- * (X / Y) != 0 -> X >= Y, if X, Y are unsigned.
- */
+ (X / Y) == 0 -> X < Y if X, Y are unsigned.
+ (X / Y) != 0 -> X >= Y, if X, Y are unsigned. */
(for cmp (eq ne)
ocmp (lt ge)
(simplify
(cmp (trunc_div @0 @1) integer_zerop)
(if (TYPE_UNSIGNED (TREE_TYPE (@0))
+ /* Complex ==/!= is allowed, but not </>=. */
+ && TREE_CODE (TREE_TYPE (@0)) != COMPLEX_TYPE
&& (VECTOR_TYPE_P (type) || !VECTOR_TYPE_P (TREE_TYPE (@0))))
(ocmp @0 @1))))
(if (ptr_difference_const (@0, @1, &diff))
{ build_int_cst_type (type, diff); }))))
(simplify
- (pointer_diff (convert?@2 ADDR_EXPR@0) (convert?@3 @1))
+ (pointer_diff (convert?@2 ADDR_EXPR@0) (convert1?@3 @1))
(if (tree_nop_conversion_p (TREE_TYPE(@2), TREE_TYPE (@0))
&& tree_nop_conversion_p (TREE_TYPE(@3), TREE_TYPE (@1)))
(with { poly_int64 diff; }
(if (ptr_difference_const (@0, @1, &diff))
{ build_int_cst_type (type, diff); }))))
(simplify
- (pointer_diff (convert?@2 @0) (convert?@3 ADDR_EXPR@1))
+ (pointer_diff (convert?@2 @0) (convert1?@3 ADDR_EXPR@1))
(if (tree_nop_conversion_p (TREE_TYPE(@2), TREE_TYPE (@0))
&& tree_nop_conversion_p (TREE_TYPE(@3), TREE_TYPE (@1)))
(with { poly_int64 diff; }
CONSTANT_CLASS_P@2)
/* If one of the types wraps, use that one. */
(if (!ANY_INTEGRAL_TYPE_P (type) || TYPE_OVERFLOW_WRAPS (type))
- (if (outer_op == PLUS_EXPR)
- (plus (view_convert @0) (inner_op @2 (view_convert @1)))
- (minus (view_convert @0) (neg_inner_op @2 (view_convert @1))))
+ /* If all 3 captures are CONSTANT_CLASS_P, punt, as we might recurse
+ forever if something doesn't simplify into a constant. */
+ (if (!CONSTANT_CLASS_P (@0))
+ (if (outer_op == PLUS_EXPR)
+ (plus (view_convert @0) (inner_op @2 (view_convert @1)))
+ (minus (view_convert @0) (neg_inner_op @2 (view_convert @1)))))
(if (!ANY_INTEGRAL_TYPE_P (TREE_TYPE (@0))
|| TYPE_OVERFLOW_WRAPS (TREE_TYPE (@0)))
(if (outer_op == PLUS_EXPR)
(neg_inner_op @0 { wide_int_to_tree (type, wi::to_wide (cst)); })
/* Last resort, use some unsigned type. */
(with { tree utype = unsigned_type_for (type); }
- (view_convert (inner_op
- (view_convert:utype @0)
- (view_convert:utype
- { drop_tree_overflow (cst); })))))))))))))
+ (if (utype)
+ (view_convert (inner_op
+ (view_convert:utype @0)
+ (view_convert:utype
+ { drop_tree_overflow (cst); }))))))))))))))
/* (CST1 - A) +- CST2 -> CST3 - A */
(for outer_op (plus minus)
/* The second argument of pointer_plus must be interpreted as signed, and
thus sign-extended if necessary. */
(with { tree stype = signed_type_for (TREE_TYPE (@1)); }
- (convert (convert:stype @1))))
+ /* Use view_convert instead of convert here, as POINTER_PLUS_EXPR
+ second arg is unsigned even when we need to consider it as signed,
+ we don't want to diagnose overflow here. */
+ (convert (view_convert:stype @1))))
/* (T)P - (T)(P + A) -> -(T) A */
(simplify
/* The second argument of pointer_plus must be interpreted as signed, and
thus sign-extended if necessary. */
(with { tree stype = signed_type_for (TREE_TYPE (@1)); }
- (negate (convert (convert:stype @1)))))
+ /* Use view_convert instead of convert here, as POINTER_PLUS_EXPR
+ second arg is unsigned even when we need to consider it as signed,
+ we don't want to diagnose overflow here. */
+ (negate (convert (view_convert:stype @1)))))
/* (T)(P + A) - (T)(P + B) -> (T)A - (T)B */
(simplify
/* The second argument of pointer_plus must be interpreted as signed, and
thus sign-extended if necessary. */
(with { tree stype = signed_type_for (TREE_TYPE (@1)); }
- (minus (convert (convert:stype @1)) (convert (convert:stype @2)))))))
-
+ /* Use view_convert instead of convert here, as POINTER_PLUS_EXPR
+ second arg is unsigned even when we need to consider it as signed,
+ we don't want to diagnose overflow here. */
+ (minus (convert (view_convert:stype @1))
+ (convert (view_convert:stype @2)))))))
+
+/* (A * C) +- (B * C) -> (A+-B) * C and (A * C) +- A -> A * (C+-1).
+ Modeled after fold_plusminus_mult_expr. */
+(if (!TYPE_SATURATING (type)
+ && (!FLOAT_TYPE_P (type) || flag_associative_math))
+ (for plusminus (plus minus)
+ (simplify
+ (plusminus (mult:cs@3 @0 @1) (mult:cs@4 @0 @2))
+ (if ((!ANY_INTEGRAL_TYPE_P (type)
+ || TYPE_OVERFLOW_WRAPS (type)
+ || (INTEGRAL_TYPE_P (type)
+ && tree_expr_nonzero_p (@0)
+ && expr_not_equal_to (@0, wi::minus_one (TYPE_PRECISION (type)))))
+ /* If @1 +- @2 is constant require a hard single-use on either
+ original operand (but not on both). */
+ && (single_use (@3) || single_use (@4)))
+ (mult (plusminus @1 @2) @0)))
+ /* We cannot generate constant 1 for fract. */
+ (if (!ALL_FRACT_MODE_P (TYPE_MODE (type)))
+ (simplify
+ (plusminus @0 (mult:c@3 @0 @2))
+ (if ((!ANY_INTEGRAL_TYPE_P (type)
+ || TYPE_OVERFLOW_WRAPS (type)
+ || (INTEGRAL_TYPE_P (type)
+ && tree_expr_nonzero_p (@0)
+ && expr_not_equal_to (@0, wi::minus_one (TYPE_PRECISION (type)))))
+ && single_use (@3))
+ (mult (plusminus { build_one_cst (type); } @2) @0)))
+ (simplify
+ (plusminus (mult:c@3 @0 @2) @0)
+ (if ((!ANY_INTEGRAL_TYPE_P (type)
+ || TYPE_OVERFLOW_WRAPS (type)
+ || (INTEGRAL_TYPE_P (type)
+ && tree_expr_nonzero_p (@0)
+ && expr_not_equal_to (@0, wi::minus_one (TYPE_PRECISION (type)))))
+ && single_use (@3))
+ (mult (plusminus @2 { build_one_cst (type); }) @0))))))
/* Simplifications of MIN_EXPR, MAX_EXPR, fmin() and fmax(). */
&& TYPE_UNSIGNED (TREE_TYPE (@1)))))
(view_convert @1)))
+/* Simplify a view-converted empty constructor. */
+(simplify
+ (view_convert CONSTRUCTOR@0)
+ (if (TREE_CODE (@0) != SSA_NAME
+ && CONSTRUCTOR_NELTS (@0) == 0)
+ { build_zero_cst (type); }))
+
/* Re-association barriers around constants and other re-association
barriers can be removed. */
(simplify
(mult (convert1? (exact_div @0 @@1)) (convert2? @1))
(convert @0))
+/* Simplify (A / B) * B + (A % B) -> A. */
+(for div (trunc_div ceil_div floor_div round_div)
+ mod (trunc_mod ceil_mod floor_mod round_mod)
+ (simplify
+ (plus:c (mult:c (div @0 @1) @1) (mod @0 @1))
+ @0))
+
+/* ((X /[ex] A) +- B) * A --> X +- A * B. */
+(for op (plus minus)
+ (simplify
+ (mult (convert1? (op (convert2? (exact_div @0 INTEGER_CST@@1)) INTEGER_CST@2)) @1)
+ (if (tree_nop_conversion_p (type, TREE_TYPE (@2))
+ && tree_nop_conversion_p (TREE_TYPE (@0), TREE_TYPE (@2)))
+ (with
+ {
+ wi::overflow_type overflow;
+ wide_int mul = wi::mul (wi::to_wide (@1), wi::to_wide (@2),
+ TYPE_SIGN (type), &overflow);
+ }
+ (if (types_match (type, TREE_TYPE (@2))
+ && types_match (TREE_TYPE (@0), TREE_TYPE (@2)) && !overflow)
+ (op @0 { wide_int_to_tree (type, mul); })
+ (with { tree utype = unsigned_type_for (type); }
+ (convert (op (convert:utype @0)
+ (mult (convert:utype @1) (convert:utype @2))))))))))
+
/* Canonicalization of binary operations. */
/* Convert X + -C into X - C. */
/* Reassociate (X * CST) * Y to (X * Y) * CST. This does not introduce
signed overflow for CST != 0 && CST != -1. */
(simplify
- (mult:c (mult:s @0 INTEGER_CST@1) @2)
+ (mult:c (mult:s@3 @0 INTEGER_CST@1) @2)
(if (TREE_CODE (@2) != INTEGER_CST
+ && single_use (@3)
&& !integer_zerop (@1) && !integer_minus_onep (@1))
(mult (mult @0 @2) @1)))
(if (integer_zerop (@0))
@2)))
+/* Sink unary operations to constant branches, but only if we do fold it to
+ constants. */
+(for op (negate bit_not abs absu)
+ (simplify
+ (op (vec_cond @0 VECTOR_CST@1 VECTOR_CST@2))
+ (with
+ {
+ tree cst1, cst2;
+ cst1 = const_unop (op, type, @1);
+ if (cst1)
+ cst2 = const_unop (op, type, @2);
+ }
+ (if (cst1 && cst2)
+ (vec_cond @0 { cst1; } { cst2; })))))
+
/* Simplification moved from fold_cond_expr_with_comparison. It may also
be extended. */
/* This pattern implements two kinds simplification:
&& (cmp == LT_EXPR || cmp == GE_EXPR)))
(with
{
- bool overflow = false;
+ wi::overflow_type overflow = wi::OVF_NONE;
enum tree_code code, cmp_code = cmp;
wide_int real_c1;
wide_int c1 = wi::to_wide (@1);
from if-conversion. */
(simplify
(cnd @0 @1 (cnd @2 @3 @4))
- (if (COMPARISON_CLASS_P (@0)
- && COMPARISON_CLASS_P (@2)
- && invert_tree_comparison
- (TREE_CODE (@0), HONOR_NANS (TREE_OPERAND (@0, 0))) == TREE_CODE (@2)
- && operand_equal_p (TREE_OPERAND (@0, 0), TREE_OPERAND (@2, 0), 0)
- && operand_equal_p (TREE_OPERAND (@0, 1), TREE_OPERAND (@2, 1), 0))
+ (if (inverse_conditions_p (@0, @2))
(cnd @0 @1 @3)))
(simplify
(cnd @0 (cnd @1 @2 @3) @4)
- (if (COMPARISON_CLASS_P (@0)
- && COMPARISON_CLASS_P (@1)
- && invert_tree_comparison
- (TREE_CODE (@0), HONOR_NANS (TREE_OPERAND (@0, 0))) == TREE_CODE (@1)
- && operand_equal_p (TREE_OPERAND (@0, 0), TREE_OPERAND (@1, 0), 0)
- && operand_equal_p (TREE_OPERAND (@0, 1), TREE_OPERAND (@1, 1), 0))
+ (if (inverse_conditions_p (@0, @1))
(cnd @0 @3 @4)))
/* A ? B : B -> B. */
(for cmp (le gt)
acmp (lt ge)
(simplify
- (cmp @0 INTEGER_CST@1)
- (if (tree_int_cst_sgn (@1) == -1)
- (acmp @0 { wide_int_to_tree (TREE_TYPE (@1), wi::to_wide (@1) + 1); }))))
+ (cmp @0 uniform_integer_cst_p@1)
+ (with { tree cst = uniform_integer_cst_p (@1); }
+ (if (tree_int_cst_sgn (cst) == -1)
+ (acmp @0 { build_uniform_cst (TREE_TYPE (@1),
+ wide_int_to_tree (TREE_TYPE (cst),
+ wi::to_wide (cst)
+ + 1)); })))))
(for cmp (ge lt)
acmp (gt le)
(simplify
- (cmp @0 INTEGER_CST@1)
- (if (tree_int_cst_sgn (@1) == 1)
- (acmp @0 { wide_int_to_tree (TREE_TYPE (@1), wi::to_wide (@1) - 1); }))))
-
+ (cmp @0 uniform_integer_cst_p@1)
+ (with { tree cst = uniform_integer_cst_p (@1); }
+ (if (tree_int_cst_sgn (cst) == 1)
+ (acmp @0 { build_uniform_cst (TREE_TYPE (@1),
+ wide_int_to_tree (TREE_TYPE (cst),
+ wi::to_wide (cst) - 1)); })))))
/* We can simplify a logical negation of a comparison to the
inverted comparison. As we cannot compute an expression
(if (tree_int_cst_sgn (@1) < 0)
(scmp @0 @2)
(cmp @0 @2))))))
-
+
/* Simplify comparison of something with itself. For IEEE
floating-point, we can only do some of these simplifications. */
(for cmp (eq ge le)
}
tree newtype
= (TYPE_PRECISION (TREE_TYPE (@0)) > TYPE_PRECISION (type1)
- ? TREE_TYPE (@0) : type1);
+ ? TREE_TYPE (@0) : type1);
}
(if (TYPE_PRECISION (TREE_TYPE (@2)) > TYPE_PRECISION (newtype))
(cmp (convert:newtype @0) (convert:newtype @1))))))
-
+
(simplify
(cmp @0 REAL_CST@1)
/* IEEE doesn't distinguish +0 and -0 in comparisons. */
code = swap_tree_comparison (code);
}
(switch
- /* x > +Inf is always false, if with ignore sNANs. */
+ /* x > +Inf is always false, if we ignore NaNs or exceptions. */
(if (code == GT_EXPR
- && ! HONOR_SNANS (@0))
+ && !(HONOR_NANS (@0) && flag_trapping_math))
{ constant_boolean_node (false, type); })
(if (code == LE_EXPR)
- /* x <= +Inf is always true, if we don't case about NaNs. */
+ /* x <= +Inf is always true, if we don't care about NaNs. */
(if (! HONOR_NANS (@0))
{ constant_boolean_node (true, type); }
- /* x <= +Inf is the same as x == x, i.e. !isnan(x). */
- (eq @0 @0)))
- /* x == +Inf and x >= +Inf are always equal to x > DBL_MAX. */
- (if (code == EQ_EXPR || code == GE_EXPR)
+ /* x <= +Inf is the same as x == x, i.e. !isnan(x), but this loses
+ an "invalid" exception. */
+ (if (!flag_trapping_math)
+ (eq @0 @0))))
+ /* x == +Inf and x >= +Inf are always equal to x > DBL_MAX, but
+ for == this introduces an exception for x a NaN. */
+ (if ((code == EQ_EXPR && !(HONOR_NANS (@0) && flag_trapping_math))
+ || code == GE_EXPR)
(with { real_maxval (&max, neg, TYPE_MODE (TREE_TYPE (@0))); }
(if (neg)
(lt @0 { build_real (TREE_TYPE (@0), max); })
(if (neg)
(ge @0 { build_real (TREE_TYPE (@0), max); })
(le @0 { build_real (TREE_TYPE (@0), max); }))))
- /* x != +Inf is always equal to !(x > DBL_MAX). */
+ /* x != +Inf is always equal to !(x > DBL_MAX), but this introduces
+ an exception for x a NaN so use an unordered comparison. */
(if (code == NE_EXPR)
(with { real_maxval (&max, neg, TYPE_MODE (TREE_TYPE (@0))); }
(if (! HONOR_NANS (@0))
(ge @0 { build_real (TREE_TYPE (@0), max); })
(le @0 { build_real (TREE_TYPE (@0), max); }))
(if (neg)
- (bit_xor (lt @0 { build_real (TREE_TYPE (@0), max); })
- { build_one_cst (type); })
- (bit_xor (gt @0 { build_real (TREE_TYPE (@0), max); })
- { build_one_cst (type); }))))))))))
+ (unge @0 { build_real (TREE_TYPE (@0), max); })
+ (unle @0 { build_real (TREE_TYPE (@0), max); }))))))))))
/* If this is a comparison of a real constant with a PLUS_EXPR
or a MINUS_EXPR of a real constant, we can convert it into a
(if (! HONOR_NANS (@0))
(cmp @0 @1))))))
+/* Optimize various special cases of (FTYPE) N CMP (FTYPE) M. */
+(for cmp (lt le eq ne ge gt unordered ordered unlt unle ungt unge uneq ltgt)
+ icmp (lt le eq ne ge gt unordered ordered lt le gt ge eq ne)
+ (simplify
+ (cmp (float@0 @1) (float @2))
+ (if (SCALAR_FLOAT_TYPE_P (TREE_TYPE (@0))
+ && ! DECIMAL_FLOAT_TYPE_P (TREE_TYPE (@0)))
+ (with
+ {
+ format_helper fmt (REAL_MODE_FORMAT (TYPE_MODE (TREE_TYPE (@0))));
+ tree type1 = TREE_TYPE (@1);
+ bool type1_signed_p = TYPE_SIGN (type1) == SIGNED;
+ tree type2 = TREE_TYPE (@2);
+ bool type2_signed_p = TYPE_SIGN (type2) == SIGNED;
+ }
+ (if (fmt.can_represent_integral_type_p (type1)
+ && fmt.can_represent_integral_type_p (type2))
+ (if (cmp == ORDERED_EXPR || cmp == UNORDERED_EXPR)
+ { constant_boolean_node (cmp == ORDERED_EXPR, type); }
+ (if (TYPE_PRECISION (type1) > TYPE_PRECISION (type2)
+ && type1_signed_p >= type2_signed_p)
+ (icmp @1 (convert @2))
+ (if (TYPE_PRECISION (type1) < TYPE_PRECISION (type2)
+ && type1_signed_p <= type2_signed_p)
+ (icmp (convert:type2 @1) @2)
+ (if (TYPE_PRECISION (type1) == TYPE_PRECISION (type2)
+ && type1_signed_p == type2_signed_p)
+ (icmp @1 @2))))))))))
+
/* Optimize various special cases of (FTYPE) N CMP CST. */
(for cmp (lt le eq ne ge gt)
icmp (le le eq ne ge ge)
(with
{
tree itype = TREE_TYPE (@0);
- signop isign = TYPE_SIGN (itype);
format_helper fmt (REAL_MODE_FORMAT (TYPE_MODE (TREE_TYPE (@1))));
const REAL_VALUE_TYPE *cst = TREE_REAL_CST_PTR (@1);
/* Be careful to preserve any potential exceptions due to
bool exception_p
= real_isnan (cst) && (cst->signalling
|| (cmp != EQ_EXPR && cmp != NE_EXPR));
- /* INT?_MIN is power-of-two so it takes
- only one mantissa bit. */
- bool signed_p = isign == SIGNED;
- bool itype_fits_ftype_p
- = TYPE_PRECISION (itype) - signed_p <= significand_size (fmt);
}
/* TODO: allow non-fitting itype and SNaNs when
-fno-trapping-math. */
- (if (itype_fits_ftype_p && ! exception_p)
+ (if (fmt.can_represent_integral_type_p (itype) && ! exception_p)
(with
{
+ signop isign = TYPE_SIGN (itype);
REAL_VALUE_TYPE imin, imax;
real_from_integer (&imin, fmt, wi::min_value (itype), isign);
real_from_integer (&imax, fmt, wi::max_value (itype), isign);
(FTYPE) N == CST -> 0
(FTYPE) N != CST -> 1. */
(if (cmp == EQ_EXPR || cmp == NE_EXPR)
- { constant_boolean_node (cmp == NE_EXPR, type); })
+ { constant_boolean_node (cmp == NE_EXPR, type); })
/* Otherwise replace with sensible integer constant. */
(with
{
(if (TREE_CODE (@1) == INTEGER_CST)
(with
{
- bool ovf;
+ wi::overflow_type ovf;
wide_int prod = wi::mul (wi::to_wide (@2), wi::to_wide (@1),
TYPE_SIGN (TREE_TYPE (@1)), &ovf);
}
(if (wi::gt_p (wi::to_wide (@1), 0, TYPE_SIGN (TREE_TYPE (@1))))
(with
{
- bool ovf;
+ wi::overflow_type ovf;
wide_int prod = wi::mul (wi::to_wide (@2), wi::to_wide (@1),
TYPE_SIGN (TREE_TYPE (@1)), &ovf);
}
!= (cmp == LT_EXPR || cmp == LE_EXPR), type); }
(cmp @0 { wide_int_to_tree (TREE_TYPE (@0), prod); }))))))
+/* Fold (size_t)(A /[ex] B) CMP C to (size_t)A CMP (size_t)B * C or A CMP' 0.
+
+ For small C (less than max/B), this is (size_t)A CMP (size_t)B * C.
+ For large C (more than min/B+2^size), this is also true, with the
+ multiplication computed modulo 2^size.
+ For intermediate C, this just tests the sign of A. */
+(for cmp (lt le gt ge)
+ cmp2 (ge ge lt lt)
+ (simplify
+ (cmp (convert (exact_div @0 INTEGER_CST@1)) INTEGER_CST@2)
+ (if (tree_nop_conversion_p (TREE_TYPE (@0), TREE_TYPE (@2))
+ && TYPE_UNSIGNED (TREE_TYPE (@2)) && !TYPE_UNSIGNED (TREE_TYPE (@0))
+ && wi::gt_p (wi::to_wide (@1), 0, TYPE_SIGN (TREE_TYPE (@1))))
+ (with
+ {
+ tree utype = TREE_TYPE (@2);
+ wide_int denom = wi::to_wide (@1);
+ wide_int right = wi::to_wide (@2);
+ wide_int smax = wi::sdiv_trunc (wi::max_value (TREE_TYPE (@0)), denom);
+ wide_int smin = wi::sdiv_trunc (wi::min_value (TREE_TYPE (@0)), denom);
+ bool small = wi::leu_p (right, smax);
+ bool large = wi::geu_p (right, smin);
+ }
+ (if (small || large)
+ (cmp (convert:utype @0) (mult @2 (convert @1)))
+ (cmp2 @0 { build_zero_cst (TREE_TYPE (@0)); }))))))
+
/* Unordered tests if either argument is a NaN. */
(simplify
(bit_ior (unordered @0 @0) (unordered @1 @1))
/* Disable this optimization if we're casting a function pointer
type on targets that require function pointer canonicalization. */
&& !(targetm.have_canonicalize_funcptr_for_compare ()
- && TREE_CODE (TREE_TYPE (@00)) == POINTER_TYPE
- && TREE_CODE (TREE_TYPE (TREE_TYPE (@00))) == FUNCTION_TYPE)
+ && ((POINTER_TYPE_P (TREE_TYPE (@00))
+ && FUNC_OR_METHOD_TYPE_P (TREE_TYPE (TREE_TYPE (@00))))
+ || (POINTER_TYPE_P (TREE_TYPE (@10))
+ && FUNC_OR_METHOD_TYPE_P (TREE_TYPE (TREE_TYPE (@10))))))
&& single_use (@0))
(if (TYPE_PRECISION (TREE_TYPE (@00)) == TYPE_PRECISION (TREE_TYPE (@0))
&& (TREE_CODE (@10) == INTEGER_CST
(simplify
(cmp (bit_and@2 @0 integer_pow2p@1) @1)
(icmp @2 { build_zero_cst (TREE_TYPE (@0)); })))
-
+
/* If we have (A & C) != 0 ? D : 0 where C and D are powers of 2,
convert this into a shift followed by ANDing with D. */
(simplify
(cond
(ne (bit_and @0 integer_pow2p@1) integer_zerop)
- integer_pow2p@2 integer_zerop)
- (with {
- int shift = (wi::exact_log2 (wi::to_wide (@2))
- - wi::exact_log2 (wi::to_wide (@1)));
- }
- (if (shift > 0)
- (bit_and
- (lshift (convert @0) { build_int_cst (integer_type_node, shift); }) @2)
- (bit_and
- (convert (rshift @0 { build_int_cst (integer_type_node, -shift); })) @2))))
+ INTEGER_CST@2 integer_zerop)
+ (if (integer_pow2p (@2))
+ (with {
+ int shift = (wi::exact_log2 (wi::to_wide (@2))
+ - wi::exact_log2 (wi::to_wide (@1)));
+ }
+ (if (shift > 0)
+ (bit_and
+ (lshift (convert @0) { build_int_cst (integer_type_node, shift); }) @2)
+ (bit_and
+ (convert (rshift @0 { build_int_cst (integer_type_node, -shift); }))
+ @2)))))
/* If we have (A & C) != 0 where C is the sign bit of A, convert
this into A < 0. Similarly for (A & C) == 0 into A >= 0. */
(simplify
(cond
(lt @0 integer_zerop)
- integer_pow2p@1 integer_zerop)
- (if (!TYPE_UNSIGNED (TREE_TYPE (@0)))
+ INTEGER_CST@1 integer_zerop)
+ (if (integer_pow2p (@1)
+ && !TYPE_UNSIGNED (TREE_TYPE (@0)))
(with {
int shift = element_precision (@0) - wi::exact_log2 (wi::to_wide (@1)) - 1;
}
|| TREE_CODE (base1) == SSA_NAME
|| TREE_CODE (base1) == STRING_CST))
equal = (base0 == base1);
+ if (equal == 0)
+ {
+ HOST_WIDE_INT ioff0 = -1, ioff1 = -1;
+ off0.is_constant (&ioff0);
+ off1.is_constant (&ioff1);
+ if ((DECL_P (base0) && TREE_CODE (base1) == STRING_CST)
+ || (TREE_CODE (base0) == STRING_CST && DECL_P (base1))
+ || (TREE_CODE (base0) == STRING_CST
+ && TREE_CODE (base1) == STRING_CST
+ && ioff0 >= 0 && ioff1 >= 0
+ && ioff0 < TREE_STRING_LENGTH (base0)
+ && ioff1 < TREE_STRING_LENGTH (base1)
+ /* This is a too conservative test that the STRING_CSTs
+ will not end up being string-merged. */
+ && strncmp (TREE_STRING_POINTER (base0) + ioff0,
+ TREE_STRING_POINTER (base1) + ioff1,
+ MIN (TREE_STRING_LENGTH (base0) - ioff0,
+ TREE_STRING_LENGTH (base1) - ioff1)) != 0))
+ ;
+ else if (!DECL_P (base0) || !DECL_P (base1))
+ equal = 2;
+ else if (cmp != EQ_EXPR && cmp != NE_EXPR)
+ equal = 2;
+ /* If this is a pointer comparison, ignore for now even
+ valid equalities where one pointer is the offset zero
+ of one object and the other to one past end of another one. */
+ else if (!INTEGRAL_TYPE_P (TREE_TYPE (@2)))
+ ;
+ /* Assume that automatic variables can't be adjacent to global
+ variables. */
+ else if (is_global_var (base0) != is_global_var (base1))
+ ;
+ else
+ {
+ tree sz0 = DECL_SIZE_UNIT (base0);
+ tree sz1 = DECL_SIZE_UNIT (base1);
+ /* If sizes are unknown, e.g. VLA or not representable,
+ punt. */
+ if (!tree_fits_poly_int64_p (sz0)
+ || !tree_fits_poly_int64_p (sz1))
+ equal = 2;
+ else
+ {
+ poly_int64 size0 = tree_to_poly_int64 (sz0);
+ poly_int64 size1 = tree_to_poly_int64 (sz1);
+ /* If one offset is pointing (or could be) to the beginning
+ of one object and the other is pointing to one past the
+ last byte of the other object, punt. */
+ if (maybe_eq (off0, 0) && maybe_eq (off1, size1))
+ equal = 2;
+ else if (maybe_eq (off1, 0) && maybe_eq (off0, size0))
+ equal = 2;
+ /* If both offsets are the same, there are some cases
+ we know that are ok. Either if we know they aren't
+ zero, or if we know both sizes are no zero. */
+ if (equal == 2
+ && known_eq (off0, off1)
+ && (known_ne (off0, 0)
+ || (known_ne (size0, 0) && known_ne (size1, 0))))
+ equal = 0;
+ }
+ }
+ }
}
- (if (equal == 1)
+ (if (equal == 1
+ && (cmp == EQ_EXPR || cmp == NE_EXPR
+ /* If the offsets are equal we can ignore overflow. */
+ || known_eq (off0, off1)
+ || TYPE_OVERFLOW_UNDEFINED (TREE_TYPE (@0))
+ /* Or if we compare using pointers to decls or strings. */
+ || (POINTER_TYPE_P (TREE_TYPE (@2))
+ && (DECL_P (base0) || TREE_CODE (base0) == STRING_CST))))
(switch
(if (cmp == EQ_EXPR && (known_eq (off0, off1) || known_ne (off0, off1)))
{ constant_boolean_node (known_eq (off0, off1), type); })
{ constant_boolean_node (known_ge (off0, off1), type); })
(if (cmp == GT_EXPR && (known_gt (off0, off1) || known_le (off0, off1)))
{ constant_boolean_node (known_gt (off0, off1), type); }))
- (if (equal == 0
- && DECL_P (base0) && DECL_P (base1)
- /* If we compare this as integers require equal offset. */
- && (!INTEGRAL_TYPE_P (TREE_TYPE (@2))
- || known_eq (off0, off1)))
- (switch
- (if (cmp == EQ_EXPR)
- { constant_boolean_node (false, type); })
- (if (cmp == NE_EXPR)
- { constant_boolean_node (true, type); })))))))))
+ (if (equal == 0)
+ (switch
+ (if (cmp == EQ_EXPR)
+ { constant_boolean_node (false, type); })
+ (if (cmp == NE_EXPR)
+ { constant_boolean_node (true, type); })))))))))
/* Simplify pointer equality compares using PTA. */
(for neeq (ne eq)
(neeq @0 @1)
(if (POINTER_TYPE_P (TREE_TYPE (@0))
&& ptrs_compare_unequal (@0, @1))
- { neeq == EQ_EXPR ? boolean_false_node : boolean_true_node; })))
+ { constant_boolean_node (neeq != EQ_EXPR, type); })))
/* PR70920: Transform (intptr_t)x eq/ne CST to x eq/ne (typeof x) CST.
and (typeof ptr_cst) x eq/ne ptr_cst to x eq/ne (typeof x) CST.
(for cmp (ne eq)
(simplify
(cmp (convert @0) INTEGER_CST@1)
- (if ((POINTER_TYPE_P (TREE_TYPE (@0)) && !FUNC_OR_METHOD_TYPE_P (TREE_TYPE (TREE_TYPE (@0)))
- && INTEGRAL_TYPE_P (TREE_TYPE (@1)))
- || (INTEGRAL_TYPE_P (TREE_TYPE (@0)) && POINTER_TYPE_P (TREE_TYPE (@1))
- && !FUNC_OR_METHOD_TYPE_P (TREE_TYPE (TREE_TYPE (@1)))))
+ (if (((POINTER_TYPE_P (TREE_TYPE (@0))
+ && !FUNC_OR_METHOD_TYPE_P (TREE_TYPE (TREE_TYPE (@0)))
+ && INTEGRAL_TYPE_P (TREE_TYPE (@1)))
+ || (INTEGRAL_TYPE_P (TREE_TYPE (@0))
+ && POINTER_TYPE_P (TREE_TYPE (@1))
+ && !FUNC_OR_METHOD_TYPE_P (TREE_TYPE (TREE_TYPE (@1)))))
+ && TYPE_PRECISION (TREE_TYPE (@0)) == TYPE_PRECISION (TREE_TYPE (@1)))
(cmp @0 (convert @1)))))
/* Non-equality compare simplifications from fold_binary */
/* Comparisons with the highest or lowest possible integer of
the specified precision will have known values. */
(simplify
- (cmp (convert?@2 @0) INTEGER_CST@1)
- (if ((INTEGRAL_TYPE_P (TREE_TYPE (@1)) || POINTER_TYPE_P (TREE_TYPE (@1)))
+ (cmp (convert?@2 @0) uniform_integer_cst_p@1)
+ (if ((INTEGRAL_TYPE_P (TREE_TYPE (@1))
+ || POINTER_TYPE_P (TREE_TYPE (@1))
+ || VECTOR_INTEGER_TYPE_P (TREE_TYPE (@1)))
&& tree_nop_conversion_p (TREE_TYPE (@2), TREE_TYPE (@0)))
(with
{
- tree arg1_type = TREE_TYPE (@1);
+ tree cst = uniform_integer_cst_p (@1);
+ tree arg1_type = TREE_TYPE (cst);
unsigned int prec = TYPE_PRECISION (arg1_type);
wide_int max = wi::max_value (arg1_type);
wide_int signed_max = wi::max_value (prec, SIGNED);
wide_int min = wi::min_value (arg1_type);
}
(switch
- (if (wi::to_wide (@1) == max)
+ (if (wi::to_wide (cst) == max)
(switch
(if (cmp == GT_EXPR)
{ constant_boolean_node (false, type); })
{ constant_boolean_node (true, type); })
(if (cmp == LT_EXPR)
(ne @2 @1))))
- (if (wi::to_wide (@1) == min)
+ (if (wi::to_wide (cst) == min)
(switch
(if (cmp == LT_EXPR)
{ constant_boolean_node (false, type); })
{ constant_boolean_node (true, type); })
(if (cmp == GT_EXPR)
(ne @2 @1))))
- (if (wi::to_wide (@1) == max - 1)
+ (if (wi::to_wide (cst) == max - 1)
(switch
(if (cmp == GT_EXPR)
- (eq @2 { wide_int_to_tree (TREE_TYPE (@1), wi::to_wide (@1) + 1); }))
+ (eq @2 { build_uniform_cst (TREE_TYPE (@1),
+ wide_int_to_tree (TREE_TYPE (cst),
+ wi::to_wide (cst)
+ + 1)); }))
(if (cmp == LE_EXPR)
- (ne @2 { wide_int_to_tree (TREE_TYPE (@1), wi::to_wide (@1) + 1); }))))
- (if (wi::to_wide (@1) == min + 1)
+ (ne @2 { build_uniform_cst (TREE_TYPE (@1),
+ wide_int_to_tree (TREE_TYPE (cst),
+ wi::to_wide (cst)
+ + 1)); }))))
+ (if (wi::to_wide (cst) == min + 1)
(switch
(if (cmp == GE_EXPR)
- (ne @2 { wide_int_to_tree (TREE_TYPE (@1), wi::to_wide (@1) - 1); }))
+ (ne @2 { build_uniform_cst (TREE_TYPE (@1),
+ wide_int_to_tree (TREE_TYPE (cst),
+ wi::to_wide (cst)
+ - 1)); }))
(if (cmp == LT_EXPR)
- (eq @2 { wide_int_to_tree (TREE_TYPE (@1), wi::to_wide (@1) - 1); }))))
- (if (wi::to_wide (@1) == signed_max
+ (eq @2 { build_uniform_cst (TREE_TYPE (@1),
+ wide_int_to_tree (TREE_TYPE (cst),
+ wi::to_wide (cst)
+ - 1)); }))))
+ (if (wi::to_wide (cst) == signed_max
&& TYPE_UNSIGNED (arg1_type)
/* We will flip the signedness of the comparison operator
associated with the mode of @1, so the sign bit is
/* The following case also applies to X < signed_max+1
and X >= signed_max+1 because previous transformations. */
(if (cmp == LE_EXPR || cmp == GT_EXPR)
- (with { tree st = signed_type_for (arg1_type); }
- (if (cmp == LE_EXPR)
- (ge (convert:st @0) { build_zero_cst (st); })
- (lt (convert:st @0) { build_zero_cst (st); }))))))))))
-
+ (with { tree st = signed_type_for (TREE_TYPE (@1)); }
+ (switch
+ (if (cst == @1 && cmp == LE_EXPR)
+ (ge (convert:st @0) { build_zero_cst (st); }))
+ (if (cst == @1 && cmp == GT_EXPR)
+ (lt (convert:st @0) { build_zero_cst (st); }))
+ (if (cmp == LE_EXPR)
+ (ge (view_convert:st @0) { build_zero_cst (st); }))
+ (if (cmp == GT_EXPR)
+ (lt (view_convert:st @0) { build_zero_cst (st); })))))))))))
+
(for cmp (unordered ordered unlt unle ungt unge uneq ltgt)
/* If the second operand is NaN, the result is constant. */
(simplify
(rdiv @2 @1))
(rdiv (op @0 @2) @1)))
+ (for cmp (lt le gt ge)
+ neg_cmp (gt ge lt le)
+ /* Simplify (x * C1) cmp C2 -> x cmp (C2 / C1), where C1 != 0. */
+ (simplify
+ (cmp (mult @0 REAL_CST@1) REAL_CST@2)
+ (with
+ { tree tem = const_binop (RDIV_EXPR, type, @2, @1); }
+ (if (tem
+ && !(REAL_VALUE_ISINF (TREE_REAL_CST (tem))
+ || (real_zerop (tem) && !real_zerop (@1))))
+ (switch
+ (if (real_less (&dconst0, TREE_REAL_CST_PTR (@1)))
+ (cmp @0 { tem; }))
+ (if (real_less (TREE_REAL_CST_PTR (@1), &dconst0))
+ (neg_cmp @0 { tem; })))))))
+
/* Simplify sqrt(x) * sqrt(y) -> sqrt(x*y). */
(for root (SQRT CBRT)
(simplify
(logs (pows @0 @1))
(mult @1 (logs @0))))
- /* pow(C,x) -> exp(log(C)*x) if C > 0. */
+ /* pow(C,x) -> exp(log(C)*x) if C > 0,
+ or if C is a positive power of 2,
+ pow(C,x) -> exp2(log2(C)*x). */
+#if GIMPLE
(for pows (POW)
exps (EXP)
logs (LOG)
+ exp2s (EXP2)
+ log2s (LOG2)
(simplify
(pows REAL_CST@0 @1)
- (if (real_compare (GT_EXPR, TREE_REAL_CST_PTR (@0), &dconst0)
- && real_isfinite (TREE_REAL_CST_PTR (@0)))
- (exps (mult (logs @0) @1)))))
+ (if (real_compare (GT_EXPR, TREE_REAL_CST_PTR (@0), &dconst0)
+ && real_isfinite (TREE_REAL_CST_PTR (@0))
+ /* As libmvec doesn't have a vectorized exp2, defer optimizing
+ the use_exp2 case until after vectorization. It seems actually
+ beneficial for all constants to postpone this until later,
+ because exp(log(C)*x), while faster, will have worse precision
+ and if x folds into a constant too, that is unnecessary
+ pessimization. */
+ && canonicalize_math_after_vectorization_p ())
+ (with {
+ const REAL_VALUE_TYPE *const value = TREE_REAL_CST_PTR (@0);
+ bool use_exp2 = false;
+ if (targetm.libc_has_function (function_c99_misc)
+ && value->cl == rvc_normal)
+ {
+ REAL_VALUE_TYPE frac_rvt = *value;
+ SET_REAL_EXP (&frac_rvt, 1);
+ if (real_equal (&frac_rvt, &dconst1))
+ use_exp2 = true;
+ }
+ }
+ (if (!use_exp2)
+ (if (optimize_pow_to_exp (@0, @1))
+ (exps (mult (logs @0) @1)))
+ (exp2s (mult (log2s @0) @1)))))))
+#endif
+
+ /* pow(C,x)*expN(y) -> expN(logN(C)*x+y) if C > 0. */
+ (for pows (POW)
+ exps (EXP EXP2 EXP10 POW10)
+ logs (LOG LOG2 LOG10 LOG10)
+ (simplify
+ (mult:c (pows:s REAL_CST@0 @1) (exps:s @2))
+ (if (real_compare (GT_EXPR, TREE_REAL_CST_PTR (@0), &dconst0)
+ && real_isfinite (TREE_REAL_CST_PTR (@0)))
+ (exps (plus (mult (logs @0) @1) @2)))))
(for sqrts (SQRT)
cbrts (CBRT)
(tans (atans @0))
@0)))
+ /* Simplify sin(atan(x)) -> x / sqrt(x*x + 1). */
+ (for sins (SIN)
+ atans (ATAN)
+ sqrts (SQRT)
+ copysigns (COPYSIGN)
+ (simplify
+ (sins (atans:s @0))
+ (with
+ {
+ REAL_VALUE_TYPE r_cst;
+ build_sinatan_real (&r_cst, type);
+ tree t_cst = build_real (type, r_cst);
+ tree t_one = build_one_cst (type);
+ }
+ (if (SCALAR_FLOAT_TYPE_P (type))
+ (cond (lt (abs @0) { t_cst; })
+ (rdiv @0 (sqrts (plus (mult @0 @0) { t_one; })))
+ (copysigns { t_one; } @0))))))
+
+/* Simplify cos(atan(x)) -> 1 / sqrt(x*x + 1). */
+ (for coss (COS)
+ atans (ATAN)
+ sqrts (SQRT)
+ copysigns (COPYSIGN)
+ (simplify
+ (coss (atans:s @0))
+ (with
+ {
+ REAL_VALUE_TYPE r_cst;
+ build_sinatan_real (&r_cst, type);
+ tree t_cst = build_real (type, r_cst);
+ tree t_one = build_one_cst (type);
+ tree t_zero = build_zero_cst (type);
+ }
+ (if (SCALAR_FLOAT_TYPE_P (type))
+ (cond (lt (abs @0) { t_cst; })
+ (rdiv { t_one; } (sqrts (plus (mult @0 @0) { t_one; })))
+ (copysigns { t_zero; } @0))))))
+
+ (if (!flag_errno_math)
+ /* Simplify sinh(atanh(x)) -> x / sqrt((1 - x)*(1 + x)). */
+ (for sinhs (SINH)
+ atanhs (ATANH)
+ sqrts (SQRT)
+ (simplify
+ (sinhs (atanhs:s @0))
+ (with { tree t_one = build_one_cst (type); }
+ (rdiv @0 (sqrts (mult (minus { t_one; } @0) (plus { t_one; } @0)))))))
+
+ /* Simplify cosh(atanh(x)) -> 1 / sqrt((1 - x)*(1 + x)) */
+ (for coshs (COSH)
+ atanhs (ATANH)
+ sqrts (SQRT)
+ (simplify
+ (coshs (atanhs:s @0))
+ (with { tree t_one = build_one_cst (type); }
+ (rdiv { t_one; } (sqrts (mult (minus { t_one; } @0) (plus { t_one; } @0))))))))
+
/* cabs(x+0i) or cabs(0+xi) -> abs(x). */
(simplify
(CABS (complex:C @0 real_zerop@1))
(if (wi::to_wide (@1) == -1)
(rdiv { build_real (type, dconst1); } @0))))
-/* Narrowing of arithmetic and logical operations.
+/* Narrowing of arithmetic and logical operations.
These are conceptually similar to the transformations performed for
the C/C++ front-ends by shorten_binary_op and shorten_compare. Long
term we want to move all that code out of the front-ends into here. */
-/* If we have a narrowing conversion of an arithmetic operation where
- both operands are widening conversions from the same type as the outer
- narrowing conversion. Then convert the innermost operands to a suitable
- unsigned type (to avoid introducing undefined behavior), perform the
- operation and convert the result to the desired type. */
-(for op (plus minus)
- (simplify
- (convert (op:s (convert@2 @0) (convert?@3 @1)))
- (if (INTEGRAL_TYPE_P (type)
- /* We check for type compatibility between @0 and @1 below,
- so there's no need to check that @1/@3 are integral types. */
- && INTEGRAL_TYPE_P (TREE_TYPE (@0))
- && INTEGRAL_TYPE_P (TREE_TYPE (@2))
- /* The precision of the type of each operand must match the
- precision of the mode of each operand, similarly for the
- result. */
- && type_has_mode_precision_p (TREE_TYPE (@0))
- && type_has_mode_precision_p (TREE_TYPE (@1))
- && type_has_mode_precision_p (type)
- /* The inner conversion must be a widening conversion. */
- && TYPE_PRECISION (TREE_TYPE (@2)) > TYPE_PRECISION (TREE_TYPE (@0))
- && types_match (@0, type)
- && (types_match (@0, @1)
- /* Or the second operand is const integer or converted const
- integer from valueize. */
- || TREE_CODE (@1) == INTEGER_CST))
- (if (TYPE_OVERFLOW_WRAPS (TREE_TYPE (@0)))
- (op @0 (convert @1))
- (with { tree utype = unsigned_type_for (TREE_TYPE (@0)); }
- (convert (op (convert:utype @0)
- (convert:utype @1))))))))
+/* Convert (outertype)((innertype0)a+(innertype1)b)
+ into ((newtype)a+(newtype)b) where newtype
+ is the widest mode from all of these. */
+(for op (plus minus mult rdiv)
+ (simplify
+ (convert (op:s@0 (convert1?@3 @1) (convert2?@4 @2)))
+ /* If we have a narrowing conversion of an arithmetic operation where
+ both operands are widening conversions from the same type as the outer
+ narrowing conversion. Then convert the innermost operands to a
+ suitable unsigned type (to avoid introducing undefined behavior),
+ perform the operation and convert the result to the desired type. */
+ (if (INTEGRAL_TYPE_P (type)
+ && op != MULT_EXPR
+ && op != RDIV_EXPR
+ /* We check for type compatibility between @0 and @1 below,
+ so there's no need to check that @2/@4 are integral types. */
+ && INTEGRAL_TYPE_P (TREE_TYPE (@1))
+ && INTEGRAL_TYPE_P (TREE_TYPE (@3))
+ /* The precision of the type of each operand must match the
+ precision of the mode of each operand, similarly for the
+ result. */
+ && type_has_mode_precision_p (TREE_TYPE (@1))
+ && type_has_mode_precision_p (TREE_TYPE (@2))
+ && type_has_mode_precision_p (type)
+ /* The inner conversion must be a widening conversion. */
+ && TYPE_PRECISION (TREE_TYPE (@3)) > TYPE_PRECISION (TREE_TYPE (@1))
+ && types_match (@1, type)
+ && (types_match (@1, @2)
+ /* Or the second operand is const integer or converted const
+ integer from valueize. */
+ || TREE_CODE (@2) == INTEGER_CST))
+ (if (TYPE_OVERFLOW_WRAPS (TREE_TYPE (@1)))
+ (op @1 (convert @2))
+ (with { tree utype = unsigned_type_for (TREE_TYPE (@1)); }
+ (convert (op (convert:utype @1)
+ (convert:utype @2)))))
+ (if (FLOAT_TYPE_P (type)
+ && DECIMAL_FLOAT_TYPE_P (TREE_TYPE (@0))
+ == DECIMAL_FLOAT_TYPE_P (type))
+ (with { tree arg0 = strip_float_extensions (@1);
+ tree arg1 = strip_float_extensions (@2);
+ tree itype = TREE_TYPE (@0);
+ tree ty1 = TREE_TYPE (arg0);
+ tree ty2 = TREE_TYPE (arg1);
+ enum tree_code code = TREE_CODE (itype); }
+ (if (FLOAT_TYPE_P (ty1)
+ && FLOAT_TYPE_P (ty2))
+ (with { tree newtype = type;
+ if (TYPE_MODE (ty1) == SDmode
+ || TYPE_MODE (ty2) == SDmode
+ || TYPE_MODE (type) == SDmode)
+ newtype = dfloat32_type_node;
+ if (TYPE_MODE (ty1) == DDmode
+ || TYPE_MODE (ty2) == DDmode
+ || TYPE_MODE (type) == DDmode)
+ newtype = dfloat64_type_node;
+ if (TYPE_MODE (ty1) == TDmode
+ || TYPE_MODE (ty2) == TDmode
+ || TYPE_MODE (type) == TDmode)
+ newtype = dfloat128_type_node; }
+ (if ((newtype == dfloat32_type_node
+ || newtype == dfloat64_type_node
+ || newtype == dfloat128_type_node)
+ && newtype == type
+ && types_match (newtype, type))
+ (op (convert:newtype @1) (convert:newtype @2))
+ (with { if (TYPE_PRECISION (ty1) > TYPE_PRECISION (newtype))
+ newtype = ty1;
+ if (TYPE_PRECISION (ty2) > TYPE_PRECISION (newtype))
+ newtype = ty2; }
+ /* Sometimes this transformation is safe (cannot
+ change results through affecting double rounding
+ cases) and sometimes it is not. If NEWTYPE is
+ wider than TYPE, e.g. (float)((long double)double
+ + (long double)double) converted to
+ (float)(double + double), the transformation is
+ unsafe regardless of the details of the types
+ involved; double rounding can arise if the result
+ of NEWTYPE arithmetic is a NEWTYPE value half way
+ between two representable TYPE values but the
+ exact value is sufficiently different (in the
+ right direction) for this difference to be
+ visible in ITYPE arithmetic. If NEWTYPE is the
+ same as TYPE, however, the transformation may be
+ safe depending on the types involved: it is safe
+ if the ITYPE has strictly more than twice as many
+ mantissa bits as TYPE, can represent infinities
+ and NaNs if the TYPE can, and has sufficient
+ exponent range for the product or ratio of two
+ values representable in the TYPE to be within the
+ range of normal values of ITYPE. */
+ (if (TYPE_PRECISION (newtype) < TYPE_PRECISION (itype)
+ && (flag_unsafe_math_optimizations
+ || (TYPE_PRECISION (newtype) == TYPE_PRECISION (type)
+ && real_can_shorten_arithmetic (TYPE_MODE (itype),
+ TYPE_MODE (type))
+ && !excess_precision_type (newtype)))
+ && !types_match (itype, newtype))
+ (convert:type (op (convert:newtype @1)
+ (convert:newtype @2)))
+ )))) )
+ ))
+)))
/* This is another case of narrowing, specifically when there's an outer
BIT_AND_EXPR which masks off bits outside the type of the innermost
(convert (bit_and (op (convert:utype @0) (convert:utype @1))
(convert:utype @4))))))))
-/* Transform (@0 < @1 and @0 < @2) to use min,
+/* Transform (@0 < @1 and @0 < @2) to use min,
(@0 > @1 and @0 > @2) to use max */
-(for op (lt le gt ge)
- ext (min min max max)
+(for logic (bit_and bit_and bit_and bit_and bit_ior bit_ior bit_ior bit_ior)
+ op (lt le gt ge lt le gt ge )
+ ext (min min max max max max min min )
(simplify
- (bit_and (op:cs @0 @1) (op:cs @0 @2))
+ (logic (op:cs @0 @1) (op:cs @0 @2))
(if (INTEGRAL_TYPE_P (TREE_TYPE (@0))
&& TREE_CODE (@0) != INTEGER_CST)
(op @0 (ext @1 @2)))))
/* Canonicalizations of BIT_FIELD_REFs. */
+(simplify
+ (BIT_FIELD_REF (BIT_FIELD_REF @0 @1 @2) @3 @4)
+ (BIT_FIELD_REF @0 @3 { const_binop (PLUS_EXPR, bitsizetype, @2, @4); }))
+
+(simplify
+ (BIT_FIELD_REF (view_convert @0) @1 @2)
+ (BIT_FIELD_REF @0 @1 @2))
+
+(simplify
+ (BIT_FIELD_REF @0 @1 integer_zerop)
+ (if (tree_int_cst_equal (@1, TYPE_SIZE (TREE_TYPE (@0))))
+ (view_convert @0)))
+
(simplify
(BIT_FIELD_REF @0 @1 @2)
(switch
{ build_constructor (type, NULL); }
(if (count == 1)
(if (elt < CONSTRUCTOR_NELTS (ctor))
- { CONSTRUCTOR_ELT (ctor, elt)->value; }
+ (view_convert { CONSTRUCTOR_ELT (ctor, elt)->value; })
{ build_zero_cst (type); })
{
vec<constructor_elt, va_gc> *vals;
(if (CONSTRUCTOR_NELTS (ctor) <= idx / const_k)
{ build_zero_cst (type); })
(if (n == const_k)
- { CONSTRUCTOR_ELT (ctor, idx / const_k)->value; })
+ (view_convert { CONSTRUCTOR_ELT (ctor, idx / const_k)->value; }))
(BIT_FIELD_REF { CONSTRUCTOR_ELT (ctor, idx / const_k)->value; }
@1 { bitsize_int ((idx % const_k) * width); })))))))))
|| wi::geu_p (wi::to_wide (@rpos),
wi::to_wide (@ipos) + isize))
(BIT_FIELD_REF @0 @rsize @rpos)))))
+
+(if (canonicalize_math_after_vectorization_p ())
+ (for fmas (FMA)
+ (simplify
+ (fmas:c (negate @0) @1 @2)
+ (IFN_FNMA @0 @1 @2))
+ (simplify
+ (fmas @0 @1 (negate @2))
+ (IFN_FMS @0 @1 @2))
+ (simplify
+ (fmas:c (negate @0) @1 (negate @2))
+ (IFN_FNMS @0 @1 @2))
+ (simplify
+ (negate (fmas@3 @0 @1 @2))
+ (if (single_use (@3))
+ (IFN_FNMS @0 @1 @2))))
+
+ (simplify
+ (IFN_FMS:c (negate @0) @1 @2)
+ (IFN_FNMS @0 @1 @2))
+ (simplify
+ (IFN_FMS @0 @1 (negate @2))
+ (IFN_FMA @0 @1 @2))
+ (simplify
+ (IFN_FMS:c (negate @0) @1 (negate @2))
+ (IFN_FNMA @0 @1 @2))
+ (simplify
+ (negate (IFN_FMS@3 @0 @1 @2))
+ (if (single_use (@3))
+ (IFN_FNMA @0 @1 @2)))
+
+ (simplify
+ (IFN_FNMA:c (negate @0) @1 @2)
+ (IFN_FMA @0 @1 @2))
+ (simplify
+ (IFN_FNMA @0 @1 (negate @2))
+ (IFN_FNMS @0 @1 @2))
+ (simplify
+ (IFN_FNMA:c (negate @0) @1 (negate @2))
+ (IFN_FMS @0 @1 @2))
+ (simplify
+ (negate (IFN_FNMA@3 @0 @1 @2))
+ (if (single_use (@3))
+ (IFN_FMS @0 @1 @2)))
+
+ (simplify
+ (IFN_FNMS:c (negate @0) @1 @2)
+ (IFN_FMS @0 @1 @2))
+ (simplify
+ (IFN_FNMS @0 @1 (negate @2))
+ (IFN_FNMA @0 @1 @2))
+ (simplify
+ (IFN_FNMS:c (negate @0) @1 (negate @2))
+ (IFN_FMA @0 @1 @2))
+ (simplify
+ (negate (IFN_FNMS@3 @0 @1 @2))
+ (if (single_use (@3))
+ (IFN_FMA @0 @1 @2))))
+
+/* POPCOUNT simplifications. */
+(for popcount (BUILT_IN_POPCOUNT BUILT_IN_POPCOUNTL BUILT_IN_POPCOUNTLL
+ BUILT_IN_POPCOUNTIMAX)
+ /* popcount(X&1) is nop_expr(X&1). */
+ (simplify
+ (popcount @0)
+ (if (tree_nonzero_bits (@0) == 1)
+ (convert @0)))
+ /* popcount(X) + popcount(Y) is popcount(X|Y) when X&Y must be zero. */
+ (simplify
+ (plus (popcount:s @0) (popcount:s @1))
+ (if (wi::bit_and (tree_nonzero_bits (@0), tree_nonzero_bits (@1)) == 0)
+ (popcount (bit_ior @0 @1))))
+ /* popcount(X) == 0 is X == 0, and related (in)equalities. */
+ (for cmp (le eq ne gt)
+ rep (eq eq ne ne)
+ (simplify
+ (cmp (popcount @0) integer_zerop)
+ (rep @0 { build_zero_cst (TREE_TYPE (@0)); }))))
+
+/* Simplify:
+
+ a = a1 op a2
+ r = c ? a : b;
+
+ to:
+
+ r = c ? a1 op a2 : b;
+
+ if the target can do it in one go. This makes the operation conditional
+ on c, so could drop potentially-trapping arithmetic, but that's a valid
+ simplification if the result of the operation isn't needed.
+
+ Avoid speculatively generating a stand-alone vector comparison
+ on targets that might not support them. Any target implementing
+ conditional internal functions must support the same comparisons
+ inside and outside a VEC_COND_EXPR. */
+
+#if GIMPLE
+(for uncond_op (UNCOND_BINARY)
+ cond_op (COND_BINARY)
+ (simplify
+ (vec_cond @0 (view_convert? (uncond_op@4 @1 @2)) @3)
+ (with { tree op_type = TREE_TYPE (@4); }
+ (if (vectorized_internal_fn_supported_p (as_internal_fn (cond_op), op_type)
+ && element_precision (type) == element_precision (op_type))
+ (view_convert (cond_op @0 @1 @2 (view_convert:op_type @3))))))
+ (simplify
+ (vec_cond @0 @1 (view_convert? (uncond_op@4 @2 @3)))
+ (with { tree op_type = TREE_TYPE (@4); }
+ (if (vectorized_internal_fn_supported_p (as_internal_fn (cond_op), op_type)
+ && element_precision (type) == element_precision (op_type))
+ (view_convert (cond_op (bit_not @0) @2 @3 (view_convert:op_type @1)))))))
+
+/* Same for ternary operations. */
+(for uncond_op (UNCOND_TERNARY)
+ cond_op (COND_TERNARY)
+ (simplify
+ (vec_cond @0 (view_convert? (uncond_op@5 @1 @2 @3)) @4)
+ (with { tree op_type = TREE_TYPE (@5); }
+ (if (vectorized_internal_fn_supported_p (as_internal_fn (cond_op), op_type)
+ && element_precision (type) == element_precision (op_type))
+ (view_convert (cond_op @0 @1 @2 @3 (view_convert:op_type @4))))))
+ (simplify
+ (vec_cond @0 @1 (view_convert? (uncond_op@5 @2 @3 @4)))
+ (with { tree op_type = TREE_TYPE (@5); }
+ (if (vectorized_internal_fn_supported_p (as_internal_fn (cond_op), op_type)
+ && element_precision (type) == element_precision (op_type))
+ (view_convert (cond_op (bit_not @0) @2 @3 @4
+ (view_convert:op_type @1)))))))
+#endif
+
+/* Detect cases in which a VEC_COND_EXPR effectively replaces the
+ "else" value of an IFN_COND_*. */
+(for cond_op (COND_BINARY)
+ (simplify
+ (vec_cond @0 (view_convert? (cond_op @0 @1 @2 @3)) @4)
+ (with { tree op_type = TREE_TYPE (@3); }
+ (if (element_precision (type) == element_precision (op_type))
+ (view_convert (cond_op @0 @1 @2 (view_convert:op_type @4))))))
+ (simplify
+ (vec_cond @0 @1 (view_convert? (cond_op @2 @3 @4 @5)))
+ (with { tree op_type = TREE_TYPE (@5); }
+ (if (inverse_conditions_p (@0, @2)
+ && element_precision (type) == element_precision (op_type))
+ (view_convert (cond_op @2 @3 @4 (view_convert:op_type @1)))))))
+
+/* Same for ternary operations. */
+(for cond_op (COND_TERNARY)
+ (simplify
+ (vec_cond @0 (view_convert? (cond_op @0 @1 @2 @3 @4)) @5)
+ (with { tree op_type = TREE_TYPE (@4); }
+ (if (element_precision (type) == element_precision (op_type))
+ (view_convert (cond_op @0 @1 @2 @3 (view_convert:op_type @5))))))
+ (simplify
+ (vec_cond @0 @1 (view_convert? (cond_op @2 @3 @4 @5 @6)))
+ (with { tree op_type = TREE_TYPE (@6); }
+ (if (inverse_conditions_p (@0, @2)
+ && element_precision (type) == element_precision (op_type))
+ (view_convert (cond_op @2 @3 @4 @5 (view_convert:op_type @1)))))))
+
+/* For pointers @0 and @2 and nonnegative constant offset @1, look for
+ expressions like:
+
+ A: (@0 + @1 < @2) | (@2 + @1 < @0)
+ B: (@0 + @1 <= @2) | (@2 + @1 <= @0)
+
+ If pointers are known not to wrap, B checks whether @1 bytes starting
+ at @0 and @2 do not overlap, while A tests the same thing for @1 + 1
+ bytes. A is more efficiently tested as:
+
+ A: (sizetype) (@0 + @1 - @2) > @1 * 2
+
+ The equivalent expression for B is given by replacing @1 with @1 - 1:
+
+ B: (sizetype) (@0 + (@1 - 1) - @2) > (@1 - 1) * 2
+
+ @0 and @2 can be swapped in both expressions without changing the result.
+
+ The folds rely on sizetype's being unsigned (which is always true)
+ and on its being the same width as the pointer (which we have to check).
+
+ The fold replaces two pointer_plus expressions, two comparisons and
+ an IOR with a pointer_plus, a pointer_diff, and a comparison, so in
+ the best case it's a saving of two operations. The A fold retains one
+ of the original pointer_pluses, so is a win even if both pointer_pluses
+ are used elsewhere. The B fold is a wash if both pointer_pluses are
+ used elsewhere, since all we end up doing is replacing a comparison with
+ a pointer_plus. We do still apply the fold under those circumstances
+ though, in case applying it to other conditions eventually makes one of the
+ pointer_pluses dead. */
+(for ior (truth_orif truth_or bit_ior)
+ (for cmp (le lt)
+ (simplify
+ (ior (cmp:cs (pointer_plus@3 @0 INTEGER_CST@1) @2)
+ (cmp:cs (pointer_plus@4 @2 @1) @0))
+ (if (TYPE_OVERFLOW_UNDEFINED (TREE_TYPE (@0))
+ && TYPE_OVERFLOW_WRAPS (sizetype)
+ && TYPE_PRECISION (TREE_TYPE (@0)) == TYPE_PRECISION (sizetype))
+ /* Calculate the rhs constant. */
+ (with { offset_int off = wi::to_offset (@1) - (cmp == LE_EXPR ? 1 : 0);
+ offset_int rhs = off * 2; }
+ /* Always fails for negative values. */
+ (if (wi::min_precision (rhs, UNSIGNED) <= TYPE_PRECISION (sizetype))
+ /* Since the order of @0 and @2 doesn't matter, let tree_swap_operands_p
+ pick a canonical order. This increases the chances of using the
+ same pointer_plus in multiple checks. */
+ (with { bool swap_p = tree_swap_operands_p (@0, @2);
+ tree rhs_tree = wide_int_to_tree (sizetype, rhs); }
+ (if (cmp == LT_EXPR)
+ (gt (convert:sizetype
+ (pointer_diff:ssizetype { swap_p ? @4 : @3; }
+ { swap_p ? @0 : @2; }))
+ { rhs_tree; })
+ (gt (convert:sizetype
+ (pointer_diff:ssizetype
+ (pointer_plus { swap_p ? @2 : @0; }
+ { wide_int_to_tree (sizetype, off); })
+ { swap_p ? @0 : @2; }))
+ { rhs_tree; })))))))))
+
+/* Fold REDUC (@0 & @1) -> @0[I] & @1[I] if element I is the only nonzero
+ element of @1. */
+(for reduc (IFN_REDUC_PLUS IFN_REDUC_IOR IFN_REDUC_XOR)
+ (simplify (reduc (view_convert? (bit_and @0 VECTOR_CST@1)))
+ (with { int i = single_nonzero_element (@1); }
+ (if (i >= 0)
+ (with { tree elt = vector_cst_elt (@1, i);
+ tree elt_type = TREE_TYPE (elt);
+ unsigned int elt_bits = tree_to_uhwi (TYPE_SIZE (elt_type));
+ tree size = bitsize_int (elt_bits);
+ tree pos = bitsize_int (elt_bits * i); }
+ (view_convert
+ (bit_and:elt_type
+ (BIT_FIELD_REF:elt_type @0 { size; } { pos; })
+ { elt; })))))))
+
+(simplify
+ (vec_perm @0 @1 VECTOR_CST@2)
+ (with
+ {
+ tree op0 = @0, op1 = @1, op2 = @2;
+
+ /* Build a vector of integers from the tree mask. */
+ vec_perm_builder builder;
+ if (!tree_to_vec_perm_builder (&builder, op2))
+ return NULL_TREE;
+
+ /* Create a vec_perm_indices for the integer vector. */
+ poly_uint64 nelts = TYPE_VECTOR_SUBPARTS (type);
+ bool single_arg = (op0 == op1);
+ vec_perm_indices sel (builder, single_arg ? 1 : 2, nelts);
+ }
+ (if (sel.series_p (0, 1, 0, 1))
+ { op0; }
+ (if (sel.series_p (0, 1, nelts, 1))
+ { op1; }
+ (with
+ {
+ if (!single_arg)
+ {
+ if (sel.all_from_input_p (0))
+ op1 = op0;
+ else if (sel.all_from_input_p (1))
+ {
+ op0 = op1;
+ sel.rotate_inputs (1);
+ }
+ else if (known_ge (poly_uint64 (sel[0]), nelts))
+ {
+ std::swap (op0, op1);
+ sel.rotate_inputs (1);
+ }
+ }
+ gassign *def;
+ tree cop0 = op0, cop1 = op1;
+ if (TREE_CODE (op0) == SSA_NAME
+ && (def = dyn_cast <gassign *> (SSA_NAME_DEF_STMT (op0)))
+ && gimple_assign_rhs_code (def) == CONSTRUCTOR)
+ cop0 = gimple_assign_rhs1 (def);
+ if (TREE_CODE (op1) == SSA_NAME
+ && (def = dyn_cast <gassign *> (SSA_NAME_DEF_STMT (op1)))
+ && gimple_assign_rhs_code (def) == CONSTRUCTOR)
+ cop1 = gimple_assign_rhs1 (def);
+
+ tree t;
+ }
+ (if ((TREE_CODE (cop0) == VECTOR_CST
+ || TREE_CODE (cop0) == CONSTRUCTOR)
+ && (TREE_CODE (cop1) == VECTOR_CST
+ || TREE_CODE (cop1) == CONSTRUCTOR)
+ && (t = fold_vec_perm (type, cop0, cop1, sel)))
+ { t; }
+ (with
+ {
+ bool changed = (op0 == op1 && !single_arg);
+ tree ins = NULL_TREE;
+ unsigned at = 0;
+
+ /* See if the permutation is performing a single element
+ insert from a CONSTRUCTOR or constant and use a BIT_INSERT_EXPR
+ in that case. But only if the vector mode is supported,
+ otherwise this is invalid GIMPLE. */
+ if (TYPE_MODE (type) != BLKmode
+ && (TREE_CODE (cop0) == VECTOR_CST
+ || TREE_CODE (cop0) == CONSTRUCTOR
+ || TREE_CODE (cop1) == VECTOR_CST
+ || TREE_CODE (cop1) == CONSTRUCTOR))
+ {
+ if (sel.series_p (1, 1, nelts + 1, 1))
+ {
+ /* After canonicalizing the first elt to come from the
+ first vector we only can insert the first elt from
+ the first vector. */
+ at = 0;
+ if ((ins = fold_read_from_vector (cop0, sel[0])))
+ op0 = op1;
+ }
+ else
+ {
+ unsigned int encoded_nelts = sel.encoding ().encoded_nelts ();
+ for (at = 0; at < encoded_nelts; ++at)
+ if (maybe_ne (sel[at], at))
+ break;
+ if (at < encoded_nelts && sel.series_p (at + 1, 1, at + 1, 1))
+ {
+ if (known_lt (at, nelts))
+ ins = fold_read_from_vector (cop0, sel[at]);
+ else
+ ins = fold_read_from_vector (cop1, sel[at] - nelts);
+ }
+ }
+ }
+
+ /* Generate a canonical form of the selector. */
+ if (!ins && sel.encoding () != builder)
+ {
+ /* Some targets are deficient and fail to expand a single
+ argument permutation while still allowing an equivalent
+ 2-argument version. */
+ tree oldop2 = op2;
+ if (sel.ninputs () == 2
+ || can_vec_perm_const_p (TYPE_MODE (type), sel, false))
+ op2 = vec_perm_indices_to_tree (TREE_TYPE (op2), sel);
+ else
+ {
+ vec_perm_indices sel2 (builder, 2, nelts);
+ if (can_vec_perm_const_p (TYPE_MODE (type), sel2, false))
+ op2 = vec_perm_indices_to_tree (TREE_TYPE (op2), sel2);
+ else
+ /* Not directly supported with either encoding,
+ so use the preferred form. */
+ op2 = vec_perm_indices_to_tree (TREE_TYPE (op2), sel);
+ }
+ if (!operand_equal_p (op2, oldop2, 0))
+ changed = true;
+ }
+ }
+ (if (ins)
+ (bit_insert { op0; } { ins; }
+ { bitsize_int (at * tree_to_uhwi (TYPE_SIZE (TREE_TYPE (type)))); })
+ (if (changed)
+ (vec_perm { op0; } { op1; } { op2; }))))))))))
+
+/* VEC_PERM_EXPR (v, v, mask) -> v where v contains same element. */
+
+(match vec_same_elem_p
+ @0
+ (if (uniform_vector_p (@0))))
+
+(match vec_same_elem_p
+ (vec_duplicate @0))
+
+(simplify
+ (vec_perm vec_same_elem_p@0 @0 @1)
+ @0)
projects / thirdparty / gcc.git / blobdiff