diff options
author | Martin Jambor <mjambor@suse.cz> | 2017-07-31 14:43:24 +0200 |
---|---|---|
committer | Martin Jambor <mjambor@suse.cz> | 2017-07-31 14:43:24 +0200 |
commit | b32f12dece884f1fa0f04c643a77105aff6ce8bc (patch) | |
tree | cdab5f10806561fc198f907299b0e55eb5701ef0 /gcc/tree-ssa-loop-niter.c | |
parent | 166bec868d991fdf71f9a66f994e5977fcab4aa2 (diff) | |
parent | a168a775e93ec31ae743ad282d8e60fa1c116891 (diff) | |
download | gcc-gcn.tar.gz |
Merge branch 'master' into gcngcn
Diffstat (limited to 'gcc/tree-ssa-loop-niter.c')
-rw-r--r-- | gcc/tree-ssa-loop-niter.c | 141 |
1 files changed, 103 insertions, 38 deletions
diff --git a/gcc/tree-ssa-loop-niter.c b/gcc/tree-ssa-loop-niter.c index 848e812cb96..e0107c28dfb 100644 --- a/gcc/tree-ssa-loop-niter.c +++ b/gcc/tree-ssa-loop-niter.c @@ -1142,8 +1142,12 @@ number_of_iterations_lt_to_ne (tree type, affine_iv *iv0, affine_iv *iv1, tree niter_type = TREE_TYPE (step); tree mod = fold_build2 (FLOOR_MOD_EXPR, niter_type, *delta, step); tree tmod; - tree assumption = boolean_true_node, bound; - tree type1 = (POINTER_TYPE_P (type)) ? sizetype : type; + mpz_t mmod; + tree assumption = boolean_true_node, bound, noloop; + bool ret = false, fv_comp_no_overflow; + tree type1 = type; + if (POINTER_TYPE_P (type)) + type1 = sizetype; if (TREE_CODE (mod) != INTEGER_CST) return false; @@ -1151,51 +1155,96 @@ number_of_iterations_lt_to_ne (tree type, affine_iv *iv0, affine_iv *iv1, mod = fold_build2 (MINUS_EXPR, niter_type, step, mod); tmod = fold_convert (type1, mod); + mpz_init (mmod); + wi::to_mpz (mod, mmod, UNSIGNED); + mpz_neg (mmod, mmod); + /* If the induction variable does not overflow and the exit is taken, - then the computation of the final value does not overflow. There - are three cases: - 1) The case if the new final value is equal to the current one. - 2) Induction varaible has pointer type, as the code cannot rely - on the object to that the pointer points being placed at the - end of the address space (and more pragmatically, - TYPE_{MIN,MAX}_VALUE is not defined for pointers). - 3) EXIT_MUST_BE_TAKEN is true, note it implies that the induction - variable does not overflow. */ - if (!integer_zerop (mod) && !POINTER_TYPE_P (type) && !exit_must_be_taken) + then the computation of the final value does not overflow. This is + also obviously the case if the new final value is equal to the + current one. Finally, we postulate this for pointer type variables, + as the code cannot rely on the object to that the pointer points being + placed at the end of the address space (and more pragmatically, + TYPE_{MIN,MAX}_VALUE is not defined for pointers). */ + if (integer_zerop (mod) || POINTER_TYPE_P (type)) + fv_comp_no_overflow = true; + else if (!exit_must_be_taken) + fv_comp_no_overflow = false; + else + fv_comp_no_overflow = + (iv0->no_overflow && integer_nonzerop (iv0->step)) + || (iv1->no_overflow && integer_nonzerop (iv1->step)); + + if (integer_nonzerop (iv0->step)) { - if (integer_nonzerop (iv0->step)) + /* The final value of the iv is iv1->base + MOD, assuming that this + computation does not overflow, and that + iv0->base <= iv1->base + MOD. */ + if (!fv_comp_no_overflow) { - /* The final value of the iv is iv1->base + MOD, assuming - that this computation does not overflow, and that - iv0->base <= iv1->base + MOD. */ bound = fold_build2 (MINUS_EXPR, type1, TYPE_MAX_VALUE (type1), tmod); assumption = fold_build2 (LE_EXPR, boolean_type_node, iv1->base, bound); + if (integer_zerop (assumption)) + goto end; } + if (mpz_cmp (mmod, bnds->below) < 0) + noloop = boolean_false_node; + else if (POINTER_TYPE_P (type)) + noloop = fold_build2 (GT_EXPR, boolean_type_node, + iv0->base, + fold_build_pointer_plus (iv1->base, tmod)); else + noloop = fold_build2 (GT_EXPR, boolean_type_node, + iv0->base, + fold_build2 (PLUS_EXPR, type1, + iv1->base, tmod)); + } + else + { + /* The final value of the iv is iv0->base - MOD, assuming that this + computation does not overflow, and that + iv0->base - MOD <= iv1->base. */ + if (!fv_comp_no_overflow) { - /* The final value of the iv is iv0->base - MOD, assuming - that this computation does not overflow, and that - iv0->base - MOD <= iv1->base. */ bound = fold_build2 (PLUS_EXPR, type1, TYPE_MIN_VALUE (type1), tmod); assumption = fold_build2 (GE_EXPR, boolean_type_node, iv0->base, bound); + if (integer_zerop (assumption)) + goto end; } - if (integer_zerop (assumption)) - return false; - else if (!integer_nonzerop (assumption)) - niter->assumptions = fold_build2 (TRUTH_AND_EXPR, boolean_type_node, - niter->assumptions, assumption); + if (mpz_cmp (mmod, bnds->below) < 0) + noloop = boolean_false_node; + else if (POINTER_TYPE_P (type)) + noloop = fold_build2 (GT_EXPR, boolean_type_node, + fold_build_pointer_plus (iv0->base, + fold_build1 (NEGATE_EXPR, + type1, tmod)), + iv1->base); + else + noloop = fold_build2 (GT_EXPR, boolean_type_node, + fold_build2 (MINUS_EXPR, type1, + iv0->base, tmod), + iv1->base); } - /* Since we are transforming LT to NE and DELTA is constant, there - is no need to compute may_be_zero because this loop must roll. */ - + if (!integer_nonzerop (assumption)) + niter->assumptions = fold_build2 (TRUTH_AND_EXPR, boolean_type_node, + niter->assumptions, + assumption); + if (!integer_zerop (noloop)) + niter->may_be_zero = fold_build2 (TRUTH_OR_EXPR, boolean_type_node, + niter->may_be_zero, + noloop); bounds_add (bnds, wi::to_widest (mod), type); *delta = fold_build2 (PLUS_EXPR, niter_type, *delta, mod); - return true; + + ret = true; +end: + mpz_clear (mmod); + return ret; } /* Add assertions to NITER that ensure that the control variable of the loop @@ -1668,18 +1717,34 @@ number_of_iterations_cond (struct loop *loop, exit_must_be_taken = true; } - /* We can handle the case when neither of the sides of the comparison is - invariant, provided that the test is NE_EXPR. This rarely occurs in - practice, but it is simple enough to manage. */ + /* We can handle cases which neither of the sides of the comparison is + invariant: + + {iv0.base, iv0.step} cmp_code {iv1.base, iv1.step} + as if: + {iv0.base, iv0.step - iv1.step} cmp_code {iv1.base, 0} + + provided that either below condition is satisfied: + + a) the test is NE_EXPR; + b) iv0.step - iv1.step is positive integer. + + This rarely occurs in practice, but it is simple enough to manage. */ if (!integer_zerop (iv0->step) && !integer_zerop (iv1->step)) { tree step_type = POINTER_TYPE_P (type) ? sizetype : type; - if (code != NE_EXPR) + tree step = fold_binary_to_constant (MINUS_EXPR, step_type, + iv0->step, iv1->step); + + /* No need to check sign of the new step since below code takes care + of this well. */ + if (code != NE_EXPR && TREE_CODE (step) != INTEGER_CST) return false; - iv0->step = fold_binary_to_constant (MINUS_EXPR, step_type, - iv0->step, iv1->step); - iv0->no_overflow = false; + iv0->step = step; + if (!POINTER_TYPE_P (type)) + iv0->no_overflow = false; + iv1->step = build_int_cst (step_type, 0); iv1->no_overflow = true; } @@ -2362,9 +2427,9 @@ number_of_iterations_exit (struct loop *loop, edge exit, return true; if (warn) - warning_at (gimple_location_safe (stmt), - OPT_Wunsafe_loop_optimizations, - "missed loop optimization, the loop counter may overflow"); + dump_printf_loc (MSG_MISSED_OPTIMIZATION, gimple_location_safe (stmt), + "missed loop optimization: niters analysis ends up " + "with assumptions.\n"); return false; } |