From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from us-smtp-delivery-124.mimecast.com (us-smtp-delivery-124.mimecast.com [170.10.133.124]) by sourceware.org (Postfix) with ESMTPS id B4B023858D35 for ; Mon, 9 Oct 2023 10:55:12 +0000 (GMT) DMARC-Filter: OpenDMARC Filter v1.4.2 sourceware.org B4B023858D35 Authentication-Results: sourceware.org; dmarc=pass (p=none dis=none) header.from=redhat.com Authentication-Results: sourceware.org; spf=pass smtp.mailfrom=redhat.com DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1696848912; h=from:from:reply-to:reply-to:subject:subject:date:date: message-id:message-id:to:to:cc:cc:mime-version:mime-version: content-type:content-type; bh=GPEIJTxl++vjZ4yLfO0Yj5cLpukYN3nlh7oSWe3T+6Q=; b=BJuOjkhu+DR9TdUP+Mj5Pm/ezRb8ppV2eW/tcylSaHr0YI0pSvQCZm6dwtRmnYyo4w2s4Q Bo/YT0U+xivVD7icbJOw0Wj9SJ0lGLTIDwhWLAEqMk9k2pIoxYY8aMqvDysHD7Hs+LanBi jrSideoJ7ueh0qwJZz51mx/FyZ1utlc= Received: from mimecast-mx02.redhat.com (mimecast-mx02.redhat.com [66.187.233.88]) by relay.mimecast.com with ESMTP with STARTTLS (version=TLSv1.2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id us-mta-170-gf1kzTKwOdiPyuhJkidXkA-1; Mon, 09 Oct 2023 06:55:08 -0400 X-MC-Unique: gf1kzTKwOdiPyuhJkidXkA-1 Received: from smtp.corp.redhat.com (int-mx07.intmail.prod.int.rdu2.redhat.com [10.11.54.7]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mimecast-mx02.redhat.com (Postfix) with ESMTPS id 1697281D9F5; Mon, 9 Oct 2023 10:55:08 +0000 (UTC) Received: from tucnak.zalov.cz (unknown [10.39.193.202]) by smtp.corp.redhat.com (Postfix) with ESMTPS id 2E933140E964; Mon, 9 Oct 2023 10:55:07 +0000 (UTC) Received: from tucnak.zalov.cz (localhost [127.0.0.1]) by tucnak.zalov.cz (8.17.1/8.17.1) with ESMTPS id 399At4MC786797 (version=TLSv1.3 cipher=TLS_AES_256_GCM_SHA384 bits=256 verify=NOT); Mon, 9 Oct 2023 12:55:04 +0200 Received: (from jakub@localhost) by tucnak.zalov.cz (8.17.1/8.17.1/Submit) id 399At3R0786795; Mon, 9 Oct 2023 12:55:03 +0200 Date: Mon, 9 Oct 2023 12:55:02 +0200 From: Jakub Jelinek To: Richard Biener , Richard Sandiford Cc: gcc-patches@gcc.gnu.org Subject: [PATCH] wide-int: Allow up to 16320 bits wide_int and change widest_int precision to 32640 bits [PR102989] Message-ID: Reply-To: Jakub Jelinek MIME-Version: 1.0 X-Scanned-By: MIMEDefang 3.1 on 10.11.54.7 X-Mimecast-Spam-Score: 0 X-Mimecast-Originator: redhat.com Content-Type: multipart/mixed; boundary="SaffQZ6I7XRJud0v" Content-Disposition: inline X-Spam-Status: No, score=-3.4 required=5.0 tests=BAYES_00,DKIMWL_WL_HIGH,DKIM_SIGNED,DKIM_VALID,DKIM_VALID_AU,DKIM_VALID_EF,RCVD_IN_DNSWL_NONE,RCVD_IN_MSPIKE_H3,RCVD_IN_MSPIKE_WL,SPF_HELO_NONE,SPF_NONE,TXREP autolearn=ham autolearn_force=no version=3.4.6 X-Spam-Checker-Version: SpamAssassin 3.4.6 (2021-04-09) on server2.sourceware.org List-Id: --SaffQZ6I7XRJud0v Content-Type: text/plain; charset=utf-8 Content-Disposition: inline Content-Transfer-Encoding: quoted-printable Hi! As mentioned in the _BitInt support thread, _BitInt(N) is currently limited by the wide_int/widest_int maximum precision limitation, which is depending on target 191, 319, 575 or 703 bits (one less than WIDE_INT_MAX_PRECISION). That is fairly low limit for _BitInt, especially on the targets with the 19= 1 bit limitation. The following patch bumps that limit to 16319 bits on all arches, which is the limit imposed by INTEGER_CST representation (unsigned char members holding number of HOST_WIDE_INT limbs). In order to achieve that, wide_int is changed from a trivially copyable typ= e which contained just an inline array of WIDE_INT_MAX_ELTS (3, 5, 9 or 11 limbs depending on target) limbs into a non-trivially copy constructible= , copy assignable and destructible type which for the usual small cases (up to WIDE_INT_MAX_INL_ELTS which is the former WIDE_INT_MAX_ELTS) still uses an inline array of limbs, but for larger precisions uses heap allocated limb array. This makes wide_int unusable in GC structures, so for dwarf2ou= t which was the only place which needed it there is a new rwide_int type (restricted wide_int) which supports only up to RWIDE_INT_MAX_ELTS limbs inline and is trivially copyable (dwarf2out should never deal with large _BitInt constants, those should have been lowered earlier). Similarly, widest_int has been changed from a trivially copyable type which contained also an inline array of WIDE_INT_MAX_ELTS limbs (but unlike wide_int didn't contain precision and assumed that to be WIDE_INT_MAX_PRECISION) into a non-trivially copy constructible, copy assignable and destructible type which has always WIDEST_INT_MAX_PRECISION precision (32640 bits currently, twice as much as INTEGER_CST limitation allows) and unlike wide_int decides depending on get_len () value whether it uses an inline array (again, up to WIDE_INT_MAX_INL_ELTS) or heap allocated one. In wide-int.h this means we need to estimate an upper bound on how many limbs will wide-int.cc (usually, sometimes wide-int.h) need to write, heap allocate if needed based on that estimation and upon set_len which is done at the end if we guessed over WIDE_INT_MAX_INL_ELTS and allocated dynamically, while we actually need less than that copy/deallocate. The unexact guesses are needed because the exact computation of the length in wide-int.cc is sometimes quite complex and especially canonicalize at the end can decrease it. widest_int is again because of this not usable in GC structures, so cfgloop.h has been changed to use fixed_wide_int_storage and punt if we'd have larger _BitInt based iterators, programs having more than 128-bit iterators will be hopefully rare and I think it is fine to treat loops with more than 2^127 iterations as effectively possibly infinite, omp-general.cc is changed to use fixed_wide_int_storage <1024>, as it better should suppor= t scores with the same precision on all arches. Code which used WIDE_INT_PRINT_BUFFER_SIZE sized buffers for printing wide_int/widest_int into buffer had to be changed to use XALLOCAVEC for larger lengths. On x86_64, the patch in --enable-checking=3Dyes,rtl,extra configured bootstrapped cc1plus enlarges the .text section by 1.01% - from 0x25725a5 to 0x25e5555 and similarly at least when compiling insn-recog.cc with the usual bootstrap option slows compilation down by 1.01%, user 4m22.046s and 4m22.384s on vanilla trunk vs. 4m25.947s and 4m25.581s on patched trunk. I'm afraid some code size growth and compile time slowdown is unavoidable in this case, we use wide_int and widest_int everywhere, and while the rare cases are marked with UNLIKELY macros, it still means extra checks for it. The patch also regresses +FAIL: gm2/pim/fail/largeconst.mod, -O =20 +FAIL: gm2/pim/fail/largeconst.mod, -O -g =20 +FAIL: gm2/pim/fail/largeconst.mod, -O3 -fomit-frame-pointer =20 +FAIL: gm2/pim/fail/largeconst.mod, -O3 -fomit-frame-pointer -finline-func= tions =20 +FAIL: gm2/pim/fail/largeconst.mod, -Os =20 +FAIL: gm2/pim/fail/largeconst.mod, -g =20 +FAIL: gm2/pim/fail/largeconst2.mod, -O =20 +FAIL: gm2/pim/fail/largeconst2.mod, -O -g =20 +FAIL: gm2/pim/fail/largeconst2.mod, -O3 -fomit-frame-pointer =20 +FAIL: gm2/pim/fail/largeconst2.mod, -O3 -fomit-frame-pointer -finline-fun= ctions =20 +FAIL: gm2/pim/fail/largeconst2.mod, -Os =20 +FAIL: gm2/pim/fail/largeconst2.mod, -g =20 tests, which previously were rejected with error: constant literal =E2=80=98123456789123456789123456791234567891234567= 891234567891234567891234567912345678912345678912345678912345678912345679123= 45678912345678912345678912345678912345679123456789123456789=E2=80=99 exceed= s internal ZTYPE range kind of errors, but now are accepted. Seems the FE tries to parse constant= s into widest_int in that case and only diagnoses if widest_int overflows, that seems wrong, it should at least punt if stuff doesn't fit into WIDE_INT_MAX_PRECISION, but perhaps far less than that, if it wants support for middle-end for precisions above 128-bit, it better should be using BITINT_TYPE. Will file a PR and defer to Modula2 maintainer. Bootstrapped/regtested on x86_64-linux and i686-linux, ok for trunk? I've additionally built it with the incremental attached patch and on make -C gcc check-gcc check-g++ -j32 -k it didn't show any wide_int/widest_int heap allocations unless a > 128-bit _BitInt or wb/uwb constant needing > 128-bit _BitInt was used in a testcase. 2023-10-09 Jakub Jelinek =09PR c/102989 =09* wide-int.h: Adjust file comment to mention 4 different kinds =09instead of 3 and how they behave. =09(WIDE_INT_MAX_INL_ELTS): Define to former value of WIDE_INT_MAX_ELTS. =09(WIDE_INT_MAX_INL_PRECISION): Define. =09(WIDE_INT_MAX_ELTS): Change to 255. Assert that WIDE_INT_MAX_INL_ELTS =09is smaller than WIDE_INT_MAX_ELTS. =09(RWIDE_INT_MAX_ELTS, RWIDE_INT_MAX_PRECISION, WIDEST_INT_MAX_ELTS, =09WIDEST_INT_MAX_PRECISION): Define. =09(WI_BINARY_RESULT_VAR, WI_UNARY_RESULT_VAR): Change write_val callers =09to pass 0 as a new argument. =09(class rwide_int_storage): Forward declare. =09(class widest_int_storage): Likewise. =09(rwide_int): New typedef. =09(widest_int, widest2_int): Change typedefs to use widest_int_storage =09rather than fixed_wide_int_storage. =09(enum wi::precision_type): Add WIDEST_CONST_PRECISION enumerator. =09(struct binary_traits): Add partial specializations for =09WIDEST_CONST_PRECISION. =09(generic_wide_int): Add needs_write_val_arg static data member. =09(int_traits): Likewise. =09(wide_int_storage): Replace val non-static data member with a union =09u of it and HOST_WIDE_INT *valp. Declare copy constructor, copy =09assignment operator and destructor. Add unsigned int argument to =09write_val. =09(wide_int_storage::wide_int_storage): Initialize precision to 0 =09in the default ctor. Remove unnecessary {}s around STATIC_ASSERTs. =09Assert in non-default ctor T's precision_type is not =09WIDEST_CONST_PRECISION and allocate u.valp for large precision. Add =09copy constructor. =09(wide_int_storage::~wide_int_storage): New. =09(wide_int_storage::operator=3D): Add copy assignment operator. In =09assignment operator remove unnecessary {}s around STATIC_ASSERTs, =09assert ctor T's precision_type is not WIDEST_CONST_PRECISION and =09if precision changes, deallocate and/or allocate u.valp. =09(wide_int_storage::get_val): Return u.valp rather than u.val for =09large precision. =09(wide_int_storage::write_val): Likewise. Add an unused unsigned int =09argument. =09(wide_int_storage::set_len): Use write_val instead of writing val =09directly. =09(wide_int_storage::from, wide_int_storage::from_array): Adjust =09write_val callers. =09(wide_int_storage::create): Allocate u.valp for large precisions. =09(wi::int_traits ::get_binary_precision): New. =09(class rwide_int_storage): New class, copied from old wide_int. =09(rwide_int_storage::write_val): New, add unused unsigned int argument. =09(wi::int_traits ): New. =09(wi::int_traits ::get_binary_precision): New. =09(fixed_wide_int_storage::fixed_wide_int_storage): Make default =09ctor defaulted. =09(fixed_wide_int_storage::write_val): Add unused unsigned int argument. =09(fixed_wide_int_storage::from, fixed_wide_int_storage::from_array): =09Adjust write_val callers. =09(wi::int_traits ::get_binary_precision): New. =09(WIDEST_INT): Define. =09(widest_int_storage): New template class. =09(wi::int_traits ): New. =09(trailing_wide_int_storage::write_val): Add unused unsigned int =09argument. =09(wi::get_binary_precision): Use =09wi::int_traits ::get_binary_precision =09rather than get_precision on get_binary_result. =09(wi::copy): Adjust write_val callers. Don't call set_len if =09needs_write_val_arg. =09(wi::bit_not): If result.needs_write_val_arg, call write_val =09again with upper bound estimate of len. =09(wi::sext, wi::zext, wi::set_bit): Likewise. =09(wi::bit_and, wi::bit_and_not, wi::bit_or, wi::bit_or_not, =09wi::bit_xor, wi::add, wi::sub, wi::mul, wi::mul_high, wi::div_trunc, =09wi::div_floor, wi::div_ceil, wi::div_round, wi::divmod_trunc, =09wi::mod_trunc, wi::mod_floor, wi::mod_ceil, wi::mod_round, =09wi::lshift, wi::lrshift, wi::arshift): Likewise. =09(wi::bswap, wi::bitreverse): Assert result.needs_write_val_arg =09is false. =09(gt_ggc_mx, gt_pch_nx): Remove generic template for all =09generic_wide_int, instead add functions and templates for each =09storage of generic_wide_int. Make functions for =09generic_wide_int and templates for =09generic_wide_int > deleted. =09(wi::mask, wi::shifted_mask): Adjust write_val calls. =09* wide-int.cc (zeros): Decrease array size to 1. =09(BLOCKS_NEEDED): Use CEIL. =09(canonize): Use HOST_WIDE_INT_M1. =09(wi::from_buffer): Pass 0 to write_val. =09(wi::to_mpz): Use CEIL. =09(wi::from_mpz): Likewise. Pass 0 to write_val. Use =09WIDE_INT_MAX_INL_ELTS instead of WIDE_INT_MAX_ELTS. =09(wi::mul_internal): Use WIDE_INT_MAX_INL_PRECISION instead of =09MAX_BITSIZE_MODE_ANY_INT in automatic array sizes, for prec =09above WIDE_INT_MAX_INL_PRECISION estimate precision from =09lengths of operands. Use XALLOCAVEC allocated buffers for =09prec above WIDE_INT_MAX_INL_PRECISION. =09(wi::divmod_internal): Likewise. =09(wi::lshift_large): For len > WIDE_INT_MAX_INL_ELTS estimate =09it from xlen and skip. =09(rshift_large_common): Remove xprecision argument, add len =09argument with len computed in caller. Don't return anything. =09(wi::lrshift_large, wi::arshift_large): Compute len here =09and pass it to rshift_large_common, for lengths above =09WIDE_INT_MAX_INL_ELTS using estimations from xlen if possible. =09(assert_deceq, assert_hexeq): For lengths above =09WIDE_INT_MAX_INL_ELTS use XALLOCAVEC allocated buffer. =09(test_printing): Use WIDE_INT_MAX_INL_PRECISION instead of =09WIDE_INT_MAX_PRECISION. =09* wide-int-print.h (WIDE_INT_PRINT_BUFFER_SIZE): Use =09WIDE_INT_MAX_INL_PRECISION instead of WIDE_INT_MAX_PRECISION. =09* wide-int-print.cc (print_decs, print_decu, print_hex): For =09lengths above WIDE_INT_MAX_INL_ELTS use XALLOCAVEC allocated buffer. =09* tree.h (wi::int_traits>): Change precision_type =09to WIDEST_CONST_PRECISION for N > ADDR_MAX_PRECISION. Add =09inl_precision static data member. =09(widest_extended_tree): Use WIDEST_INT_MAX_PRECISION instead of =09WIDE_INT_MAX_PRECISION. =09(wi::ints_for): Use int_traits >::precision_type =09instead of hard coded CONST_PRECISION. =09(widest2_int_cst): Use WIDEST_INT_MAX_PRECISION instead of =09WIDE_INT_MAX_PRECISION. =09(wi::extended_tree ::get_len): Use WIDEST_INT_MAX_PRECISION rather =09than WIDE_INT_MAX_PRECISION. =09(wi::ints_for::zero): Use =09wi::int_traits >::precision_type instead of =09wi::CONST_PRECISION. =09* tree.cc (build_replicated_int_cst): Formatting fix. Use =09WIDE_INT_MAX_INL_ELTS rather than WIDE_INT_MAX_ELTS. =09* print-tree.cc (print_node): Don't print TREE_UNAVAILABLE on =09INTEGER_CSTs, TREE_VECs or SSA_NAMEs. =09* poly-int.h (struct poly_coeff_traits): Add partial specialization =09for wi::WIDEST_CONST_PRECISION. =09* cfgloop.h (bound_wide_int): New typedef. =09(struct nb_iter_bound): Change bound type from widest_int to =09bound_wide_int. =09(struct loop): Change nb_iterations_upper_bound, =09nb_iterations_likely_upper_bound and nb_iterations_estimate type from =09widest_int to bound_wide_int. =09* cfgloop.cc (record_niter_bound): Return early if wi::min_precision =09of i_bound is too large for bound_wide_int. Adjustments for the =09widest_int to bound_wide_int type change in non-static data members. =09(get_estimated_loop_iterations, get_max_loop_iterations, =09get_likely_max_loop_iterations): Adjustments for the widest_int to =09bound_wide_int type change in non-static data members. =09* tree-vect-loop.cc (vect_transform_loop): Likewise. =09* tree-ssa-loop-niter.cc (do_warn_aggressive_loop_optimizations): Use =09XALLOCAVEC allocated buffer for i_bound len above =09WIDE_INT_MAX_INL_ELTS. =09(record_estimate): Return early if wi::min_precision of i_bound is too =09large for bound_wide_int. Adjustments for the widest_int to =09bound_wide_int type change in non-static data members. =09(wide_int_cmp): Use bound_wide_int instead of widest_int. =09(bound_index): Use bound_wide_int instead of widest_int. =09(discover_iteration_bound_by_body_walk): Likewise. Use =09widest_int::from to convert it to widest_int when passed to =09record_niter_bound. =09(maybe_lower_iteration_bound): Use widest_int::from to convert it to =09widest_int when passed to record_niter_bound. =09(estimate_numbers_of_iteration): Don't record upper bound if =09loop->nb_iterations has too large precision for bound_wide_int. =09(n_of_executions_at_most): Use widest_int::from. =09* tree-ssa-loop-ivcanon.cc (remove_redundant_iv_tests): Adjust for =09the widest_int to bound_wide_int changes. =09* match.pd (fold_sign_changed_comparison simplification): Use =09wide_int::from on wi::to_wide instead of wi::to_widest. =09* value-range.h (irange::maybe_resize): Avoid using memcpy on =09non-trivially copyable elements. =09* value-range.cc (irange_bitmask::dump): Use XALLOCAVEC allocated =09buffer for mask or value len above WIDE_INT_PRINT_BUFFER_SIZE. =09* fold-const.cc (fold_convert_const_int_from_int, fold_unary_loc): =09Use wide_int::from on wi::to_wide instead of wi::to_widest. =09* tree-ssa-ccp.cc (bit_value_binop): Zero extend r1max from width =09before calling wi::udiv_trunc. =09* dwarf2out.h (wide_int_ptr): Remove. =09(rwice_int_ptr): New typedef. =09(struct dw_val_node): Use rwide_int_ptr for val_wide rather than =09wide_int_ptr. =09* dwarf2out.cc (get_full_len): Use rwide_int instead of wide_int. =09(insert_wide_int, add_AT_wide, mem_loc_descriptor, loc_descriptor, =09add_const_value_attribute): Likewise. =09* lto-streamer-out.cc (output_cfg): Adjustments for the widest_int to =09bound_wide_int type change in non-static data members. =09* lto-streamer-in.cc (input_cfg): Likewise. =09(lto_input_tree_1): Use WIDE_INT_MAX_INL_ELTS rather than =09WIDE_INT_MAX_ELTS. For length above WIDE_INT_MAX_INL_ELTS use =09XALLOCAVEC allocated buffer. Formatting fix. =09* data-streamer-in.cc (streamer_read_wide_int, =09streamer_read_widest_int): Likewise. =09* tree-affine.cc (aff_combination_expand): Use placement new to =09construct name_expansion. =09(free_name_expansion): Destruct name_expansion. =09* gimple-ssa-strength-reduction.cc (struct slsr_cand_d): Change =09index type from widest_int to offset_int. =09(class incr_info_d): Change incr type from widest_int to offset_int. =09(alloc_cand_and_find_basis, backtrace_base_for_ref, =09restructure_reference, slsr_process_ref, create_mul_ssa_cand, =09create_mul_imm_cand, create_add_ssa_cand, create_add_imm_cand, =09slsr_process_add, cand_abs_increment, replace_mult_candidate, =09replace_unconditional_candidate, incr_vec_index, =09create_add_on_incoming_edge, create_phi_basis_1, =09replace_conditional_candidate, record_increment, =09record_phi_increments_1, phi_incr_cost_1, phi_incr_cost, =09lowest_cost_path, total_savings, ncd_with_phi, ncd_of_cand_and_phis, =09nearest_common_dominator_for_cands, insert_initializers, =09all_phi_incrs_profitable_1, replace_one_candidate, =09replace_profitable_candidates): Use offset_int rather than widest_int =09and wi::to_offset rather than wi::to_widest. =09* real.cc (real_to_integer): Use WIDE_INT_MAX_INL_ELTS rather than =092 * WIDE_INT_MAX_ELTS and for words above that use XALLOCAVEC =09allocated buffer. =09* tree-ssa-loop-ivopts.cc (niter_for_exit): Use placement new =09to construct tree_niter_desc and destruct it on failure. =09(free_tree_niter_desc): Destruct tree_niter_desc if value is non-NULL. =09* gengtype.cc (main): Remove widest_int handling. =09* graphite-isl-ast-to-gimple.cc (widest_int_from_isl_expr_int): Use =09WIDEST_INT_MAX_ELTS instead of WIDE_INT_MAX_ELTS. =09* gimple-ssa-warn-alloca.cc (pass_walloca::execute): Use =09WIDE_INT_MAX_INL_PRECISION instead of WIDE_INT_MAX_PRECISION and =09assert get_len () fits into it. =09* value-range-pretty-print.cc (vrange_printer::print_irange_bitmasks): =09For mask or value lengths above WIDE_INT_MAX_INL_ELTS use XALLOCAVEC =09allocated buffer. =09* gimple-ssa-sprintf.cc (adjust_range_for_overflow): Use =09wide_int::from on wi::to_wide instead of wi::to_widest. =09* omp-general.cc (score_wide_int): New typedef. =09(omp_context_compute_score): Use score_wide_int instead of widest_int =09and adjust for those changes. =09(struct omp_declare_variant_entry): Change score and =09score_in_declare_simd_clone non-static data member type from widest_int =09to score_wide_int. =09(omp_resolve_late_declare_variant, omp_resolve_declare_variant): Use =09score_wide_int instead of widest_int and adjust for those changes. =09(omp_lto_output_declare_variant_alt): Likewise. =09(omp_lto_input_declare_variant_alt): Likewise. =09* godump.cc (go_output_typedef): Assert get_len () is smaller than =09WIDE_INT_MAX_INL_ELTS. gcc/c-family/ =09* c-warn.cc (match_case_to_enum_1): Use wi::to_wide just once instead =09of 3 times, assert get_len () is smaller than WIDE_INT_MAX_INL_ELTS. gcc/testsuite/ =09* gcc.dg/bitint-38.c: New test. --- gcc/wide-int.h.jj=092023-10-08 16:37:32.095269217 +0200 +++ gcc/wide-int.h=092023-10-08 17:02:21.083935772 +0200 @@ -27,7 +27,7 @@ along with GCC; see the file COPYING3. other longer storage GCC representations (rtl and tree). =20 The actual precision of a wide_int depends on the flavor. There - are three predefined flavors: + are four predefined flavors: =20 1) wide_int (the default). This flavor does the math in the precision of its input arguments. It is assumed (and checked) @@ -53,6 +53,10 @@ along with GCC; see the file COPYING3. multiply, division, shifts, comparisons, and operations that need overflow detected), the signedness must be specified separately. =20 + For precisions up to WIDE_INT_MAX_INL_PRECISION, it uses an inline + buffer in the type, for larger precisions up to WIDEST_INT_MAX_PRECIS= ION + it uses a pointer to heap allocated buffer. + 2) offset_int. This is a fixed-precision integer that can hold any address offset, measured in either bits or bytes, with at least one extra sign bit. At the moment the maximum address @@ -76,11 +80,15 @@ along with GCC; see the file COPYING3. wi::leu_p (a, b) as a more efficient short-hand for "a >=3D 0 && a <=3D b". ] =20 - 3) widest_int. This representation is an approximation of + 3) rwide_int. Restricted wide_int. This is similar to + wide_int, but maximum possible precision is RWIDE_INT_MAX_PRECISION + and it always uses an inline buffer. offset_int and rwide_int are + GC-friendly, wide_int and widest_int are not. + + 4) widest_int. This representation is an approximation of infinite precision math. However, it is not really infinite precision math as in the GMP library. It is really finite - precision math where the precision is 4 times the size of the - largest integer that the target port can represent. + precision math where the precision is WIDEST_INT_MAX_PRECISION. =20 Like offset_int, widest_int is wider than all the values that it needs to represent, so the integers are logically signed. @@ -231,17 +239,34 @@ along with GCC; see the file COPYING3. can be arbitrarily different from X. */ =20 /* The MAX_BITSIZE_MODE_ANY_INT is automatically generated by a very - early examination of the target's mode file. The WIDE_INT_MAX_ELTS + early examination of the target's mode file. The WIDE_INT_MAX_INL_ELTS can accomodate at least 1 more bit so that unsigned numbers of that mode can be represented as a signed value. Note that it is still possible to create fixed_wide_ints that have precisions greater than MAX_BITSIZE_MODE_ANY_INT. This can be useful when representing a double-width multiplication result, for example. */ -#define WIDE_INT_MAX_ELTS \ - ((MAX_BITSIZE_MODE_ANY_INT + HOST_BITS_PER_WIDE_INT) / HOST_BITS_PER_WID= E_INT) - +#define WIDE_INT_MAX_INL_ELTS \ + ((MAX_BITSIZE_MODE_ANY_INT + HOST_BITS_PER_WIDE_INT) \ + / HOST_BITS_PER_WIDE_INT) + +#define WIDE_INT_MAX_INL_PRECISION \ + (WIDE_INT_MAX_INL_ELTS * HOST_BITS_PER_WIDE_INT) + +/* Precision of wide_int and largest _BitInt precision + 1 we can + support. */ +#define WIDE_INT_MAX_ELTS 255 #define WIDE_INT_MAX_PRECISION (WIDE_INT_MAX_ELTS * HOST_BITS_PER_WIDE_INT= ) =20 +#define RWIDE_INT_MAX_ELTS WIDE_INT_MAX_INL_ELTS +#define RWIDE_INT_MAX_PRECISION WIDE_INT_MAX_INL_PRECISION + +/* Precision of widest_int and largest _BitInt precision + 1 we can + support. */ +#define WIDEST_INT_MAX_ELTS 510 +#define WIDEST_INT_MAX_PRECISION (WIDEST_INT_MAX_ELTS * HOST_BITS_PER_WIDE= _INT) + +STATIC_ASSERT (WIDE_INT_MAX_INL_ELTS < WIDE_INT_MAX_ELTS); + /* This is the max size of any pointer on any machine. It does not seem to be as easy to sniff this out of the machine description as it is for MAX_BITSIZE_MODE_ANY_INT since targets may support @@ -307,17 +332,19 @@ along with GCC; see the file COPYING3. #define WI_BINARY_RESULT_VAR(RESULT, VAL, T1, X, T2, Y) \ WI_BINARY_RESULT (T1, T2) RESULT =3D \ wi::int_traits ::get_binary_result (X, Y); = \ - HOST_WIDE_INT *VAL =3D RESULT.write_val () + HOST_WIDE_INT *VAL =3D RESULT.write_val (0) =20 /* Similar for the result of a unary operation on X, which has type T. */ #define WI_UNARY_RESULT_VAR(RESULT, VAL, T, X) \ WI_UNARY_RESULT (T) RESULT =3D \ wi::int_traits ::get_binary_result (X, X); \ - HOST_WIDE_INT *VAL =3D RESULT.write_val () + HOST_WIDE_INT *VAL =3D RESULT.write_val (0) =20 template class generic_wide_int; template class fixed_wide_int_storage; class wide_int_storage; +class rwide_int_storage; +template class widest_int_storage; =20 /* An N-bit integer. Until we can use typedef templates, use this instead= . */ #define FIXED_WIDE_INT(N) \ @@ -325,10 +352,9 @@ class wide_int_storage; =20 typedef generic_wide_int wide_int; typedef FIXED_WIDE_INT (ADDR_MAX_PRECISION) offset_int; -typedef FIXED_WIDE_INT (WIDE_INT_MAX_PRECISION) widest_int; -/* Spelled out explicitly (rather than through FIXED_WIDE_INT) - so as not to confuse gengtype. */ -typedef generic_wide_int < fixed_wide_int_storage > widest2_int; +typedef generic_wide_int rwide_int; +typedef generic_wide_int = > widest_int; +typedef generic_wide_int > widest2_int; =20 /* wi::storage_ref can be a reference to a primitive type, so this is the conservatively-correct setting. */ @@ -380,7 +406,11 @@ namespace wi =20 /* The integer has a constant precision (known at GCC compile time) and is signed. */ - CONST_PRECISION + CONST_PRECISION, + + /* Like CONST_PRECISION, but with WIDEST_INT_MAX_PRECISION or larger + precision where not all elements of arrays are always present. */ + WIDEST_CONST_PRECISION }; =20 /* This class, which has no default implementation, is expected to @@ -390,9 +420,15 @@ namespace wi Classifies the type of T. =20 static const unsigned int precision; - Only defined if precision_type =3D=3D CONST_PRECISION. Specifies t= he + Only defined if precision_type =3D=3D CONST_PRECISION or + precision_type =3D=3D WIDEST_CONST_PRECISION. Specifies the precision of all integers of type T. =20 + static const unsigned int inl_precision; + Only defined if precision_type =3D=3D WIDEST_CONST_PRECISION. + Specifies precision which is represented in the inline + arrays. + static const bool host_dependent_precision; True if the precision of T depends (or can depend) on the host. =20 @@ -415,9 +451,10 @@ namespace wi struct binary_traits; =20 /* Specify the result type for each supported combination of binary - inputs. Note that CONST_PRECISION and VAR_PRECISION cannot be - mixed, in order to give stronger type checking. When both inputs - are CONST_PRECISION, they must have the same precision. */ + inputs. Note that CONST_PRECISION, WIDEST_CONST_PRECISION and + VAR_PRECISION cannot be mixed, in order to give stronger type + checking. When both inputs are CONST_PRECISION or both are + WIDEST_CONST_PRECISION, they must have the same precision. */ template struct binary_traits { @@ -447,6 +484,17 @@ namespace wi }; =20 template + struct binary_traits + { + typedef generic_wide_int < widest_int_storage +=09=09=09 ::inl_precision> > result_type; + typedef result_type operator_result; + typedef bool predicate_result; + typedef result_type signed_shift_result_type; + typedef bool signed_predicate_result; + }; + + template struct binary_traits { typedef wide_int result_type; @@ -468,6 +516,17 @@ namespace wi }; =20 template + struct binary_traits + { + typedef generic_wide_int < widest_int_storage +=09=09=09 ::inl_precision> > result_type; + typedef result_type operator_result; + typedef bool predicate_result; + typedef result_type signed_shift_result_type; + typedef bool signed_predicate_result; + }; + + template struct binary_traits { STATIC_ASSERT (int_traits ::precision =3D=3D int_traits ::prec= ision); @@ -482,6 +541,18 @@ namespace wi }; =20 template + struct binary_traits + { + STATIC_ASSERT (int_traits ::precision =3D=3D int_traits ::prec= ision); + typedef generic_wide_int < widest_int_storage +=09=09=09 ::inl_precision> > result_type; + typedef result_type operator_result; + typedef bool predicate_result; + typedef result_type signed_shift_result_type; + typedef bool signed_predicate_result; + }; + + template struct binary_traits { typedef wide_int result_type; @@ -709,8 +780,10 @@ wi::storage_ref::get_val () const Although not required by generic_wide_int itself, writable storage classes can also provide the following functions: =20 - HOST_WIDE_INT *write_val () - Get a modifiable version of get_val () + HOST_WIDE_INT *write_val (unsigned int) + Get a modifiable version of get_val (). The argument should be + upper estimation for LEN (ignored by all storages but + widest_int_storage). =20 unsigned int set_len (unsigned int len) Set the value returned by get_len () to LEN. */ @@ -777,6 +850,8 @@ public: =20 static const bool is_sign_extended =3D wi::int_traits >::is_sign_extended; + static const bool needs_write_val_arg + =3D wi::int_traits >::needs_write_val_arg; }; =20 template @@ -1049,6 +1124,7 @@ namespace wi static const enum precision_type precision_type =3D VAR_PRECISION; static const bool host_dependent_precision =3D HDP; static const bool is_sign_extended =3D SE; + static const bool needs_write_val_arg =3D false; }; } =20 @@ -1065,7 +1141,11 @@ namespace wi class GTY(()) wide_int_storage { private: - HOST_WIDE_INT val[WIDE_INT_MAX_ELTS]; + union + { + HOST_WIDE_INT val[WIDE_INT_MAX_INL_ELTS]; + HOST_WIDE_INT *valp; + } GTY((skip)) u; unsigned int len; unsigned int precision; =20 @@ -1073,14 +1153,17 @@ public: wide_int_storage (); template wide_int_storage (const T &); + wide_int_storage (const wide_int_storage &); + ~wide_int_storage (); =20 /* The standard generic_wide_int storage methods. */ unsigned int get_precision () const; const HOST_WIDE_INT *get_val () const; unsigned int get_len () const; - HOST_WIDE_INT *write_val (); + HOST_WIDE_INT *write_val (unsigned int); void set_len (unsigned int, bool =3D false); =20 + wide_int_storage &operator =3D (const wide_int_storage &); template wide_int_storage &operator =3D (const T &); =20 @@ -1099,12 +1182,15 @@ namespace wi /* Guaranteed by a static assert in the wide_int_storage constructor. = */ static const bool host_dependent_precision =3D false; static const bool is_sign_extended =3D true; + static const bool needs_write_val_arg =3D false; template static wide_int get_binary_result (const T1 &, const T2 &); + template + static unsigned int get_binary_precision (const T1 &, const T2 &); }; } =20 -inline wide_int_storage::wide_int_storage () {} +inline wide_int_storage::wide_int_storage () : precision (0) {} =20 /* Initialize the storage from integer X, in its natural precision. Note that we do not allow integers with host-dependent precision @@ -1113,21 +1199,75 @@ inline wide_int_storage::wide_int_storag template inline wide_int_storage::wide_int_storage (const T &x) { - { STATIC_ASSERT (!wi::int_traits::host_dependent_precision); } - { STATIC_ASSERT (wi::int_traits::precision_type !=3D wi::CONST_PRECIS= ION); } + STATIC_ASSERT (!wi::int_traits::host_dependent_precision); + STATIC_ASSERT (wi::int_traits::precision_type !=3D wi::CONST_PRECISIO= N); + STATIC_ASSERT (wi::int_traits::precision_type +=09=09 !=3D wi::WIDEST_CONST_PRECISION); WIDE_INT_REF_FOR (T) xi (x); precision =3D xi.precision; + if (UNLIKELY (precision > WIDE_INT_MAX_INL_PRECISION)) + u.valp =3D XNEWVEC (HOST_WIDE_INT, CEIL (precision, HOST_BITS_PER_WIDE= _INT)); wi::copy (*this, xi); } =20 +inline wide_int_storage::wide_int_storage (const wide_int_storage &x) +{ + len =3D x.len; + precision =3D x.precision; + if (UNLIKELY (precision > WIDE_INT_MAX_INL_PRECISION)) + { + u.valp =3D XNEWVEC (HOST_WIDE_INT, CEIL (precision, HOST_BITS_PER_WI= DE_INT)); + memcpy (u.valp, x.u.valp, len * sizeof (HOST_WIDE_INT)); + } + else if (LIKELY (precision)) + memcpy (u.val, x.u.val, len * sizeof (HOST_WIDE_INT)); +} + +inline wide_int_storage::~wide_int_storage () +{ + if (UNLIKELY (precision > WIDE_INT_MAX_INL_PRECISION)) + XDELETEVEC (u.valp); +} + +inline wide_int_storage& +wide_int_storage::operator =3D (const wide_int_storage &x) +{ + if (UNLIKELY (precision > WIDE_INT_MAX_INL_PRECISION)) + { + if (this =3D=3D &x) +=09return *this; + XDELETEVEC (u.valp); + } + len =3D x.len; + precision =3D x.precision; + if (UNLIKELY (precision > WIDE_INT_MAX_INL_PRECISION)) + { + u.valp =3D XNEWVEC (HOST_WIDE_INT, CEIL (precision, HOST_BITS_PER_WI= DE_INT)); + memcpy (u.valp, x.u.valp, len * sizeof (HOST_WIDE_INT)); + } + else if (LIKELY (precision)) + memcpy (u.val, x.u.val, len * sizeof (HOST_WIDE_INT)); + return *this; +} + template inline wide_int_storage& wide_int_storage::operator =3D (const T &x) { - { STATIC_ASSERT (!wi::int_traits::host_dependent_precision); } - { STATIC_ASSERT (wi::int_traits::precision_type !=3D wi::CONST_PRECIS= ION); } + STATIC_ASSERT (!wi::int_traits::host_dependent_precision); + STATIC_ASSERT (wi::int_traits::precision_type !=3D wi::CONST_PRECISIO= N); + STATIC_ASSERT (wi::int_traits::precision_type +=09=09 !=3D wi::WIDEST_CONST_PRECISION); WIDE_INT_REF_FOR (T) xi (x); - precision =3D xi.precision; + if (UNLIKELY (precision !=3D xi.precision)) + { + if (UNLIKELY (precision > WIDE_INT_MAX_INL_PRECISION)) +=09XDELETEVEC (u.valp); + precision =3D xi.precision; + if (UNLIKELY (precision > WIDE_INT_MAX_INL_PRECISION)) +=09u.valp =3D XNEWVEC (HOST_WIDE_INT, +=09=09=09 CEIL (precision, HOST_BITS_PER_WIDE_INT)); + } wi::copy (*this, xi); return *this; } @@ -1141,7 +1281,7 @@ wide_int_storage::get_precision () const inline const HOST_WIDE_INT * wide_int_storage::get_val () const { - return val; + return UNLIKELY (precision > WIDE_INT_MAX_INL_PRECISION) ? u.valp : u.va= l; } =20 inline unsigned int @@ -1151,9 +1291,9 @@ wide_int_storage::get_len () const } =20 inline HOST_WIDE_INT * -wide_int_storage::write_val () +wide_int_storage::write_val (unsigned int) { - return val; + return UNLIKELY (precision > WIDE_INT_MAX_INL_PRECISION) ? u.valp : u.va= l; } =20 inline void @@ -1161,8 +1301,10 @@ wide_int_storage::set_len (unsigned int { len =3D l; if (!is_sign_extended && len * HOST_BITS_PER_WIDE_INT > precision) - val[len - 1] =3D sext_hwi (val[len - 1], -=09=09=09 precision % HOST_BITS_PER_WIDE_INT); + { + HOST_WIDE_INT &v =3D write_val (len)[len - 1]; + v =3D sext_hwi (v, precision % HOST_BITS_PER_WIDE_INT); + } } =20 /* Treat X as having signedness SGN and convert it to a PRECISION-bit @@ -1172,7 +1314,7 @@ wide_int_storage::from (const wide_int_r =09=09=09signop sgn) { wide_int result =3D wide_int::create (precision); - result.set_len (wi::force_to_size (result.write_val (), x.val, x.len, + result.set_len (wi::force_to_size (result.write_val (x.len), x.val, x.le= n, =09=09=09=09 x.precision, precision, sgn)); return result; } @@ -1185,7 +1327,7 @@ wide_int_storage::from_array (const HOST =09=09=09 unsigned int precision, bool need_canon_p) { wide_int result =3D wide_int::create (precision); - result.set_len (wi::from_array (result.write_val (), val, len, precision= , + result.set_len (wi::from_array (result.write_val (len), val, len, precis= ion, =09=09=09=09 need_canon_p)); return result; } @@ -1196,6 +1338,9 @@ wide_int_storage::create (unsigned int p { wide_int x; x.precision =3D precision; + if (UNLIKELY (precision > WIDE_INT_MAX_INL_PRECISION)) + x.u.valp =3D XNEWVEC (HOST_WIDE_INT, +=09=09=09CEIL (precision, HOST_BITS_PER_WIDE_INT)); return x; } =20 @@ -1212,6 +1357,194 @@ wi::int_traits ::get_b return wide_int::create (wi::get_precision (x)); } =20 +template +inline unsigned int +wi::int_traits ::get_binary_precision (const T1 &x, +=09=09=09=09=09=09=09 const T2 &y) +{ + /* This shouldn't be used for two flexible-precision inputs. */ + STATIC_ASSERT (wi::int_traits ::precision_type !=3D FLEXIBLE_PRECISI= ON +=09=09 || wi::int_traits ::precision_type !=3D FLEXIBLE_PRECISION); + if (wi::int_traits ::precision_type =3D=3D FLEXIBLE_PRECISION) + return wi::get_precision (y); + else + return wi::get_precision (x); +} + +/* The storage used by rwide_int. */ +class GTY(()) rwide_int_storage +{ +private: + HOST_WIDE_INT val[RWIDE_INT_MAX_ELTS]; + unsigned int len; + unsigned int precision; + +public: + rwide_int_storage () =3D default; + template + rwide_int_storage (const T &); + + /* The standard generic_rwide_int storage methods. */ + unsigned int get_precision () const; + const HOST_WIDE_INT *get_val () const; + unsigned int get_len () const; + HOST_WIDE_INT *write_val (unsigned int); + void set_len (unsigned int, bool =3D false); + + template + rwide_int_storage &operator =3D (const T &); + + static rwide_int from (const wide_int_ref &, unsigned int, signop); + static rwide_int from_array (const HOST_WIDE_INT *, unsigned int, +=09=09=09 unsigned int, bool =3D true); + static rwide_int create (unsigned int); +}; + +namespace wi +{ + template <> + struct int_traits + { + static const enum precision_type precision_type =3D VAR_PRECISION; + /* Guaranteed by a static assert in the rwide_int_storage constructor.= */ + static const bool host_dependent_precision =3D false; + static const bool is_sign_extended =3D true; + static const bool needs_write_val_arg =3D false; + template + static rwide_int get_binary_result (const T1 &, const T2 &); + template + static unsigned int get_binary_precision (const T1 &, const T2 &); + }; +} + +/* Initialize the storage from integer X, in its natural precision. + Note that we do not allow integers with host-dependent precision + to become rwide_ints; rwide_ints must always be logically independent + of the host. */ +template +inline rwide_int_storage::rwide_int_storage (const T &x) +{ + STATIC_ASSERT (!wi::int_traits::host_dependent_precision); + STATIC_ASSERT (wi::int_traits::precision_type !=3D wi::CONST_PRECISIO= N); + STATIC_ASSERT (wi::int_traits::precision_type +=09=09 !=3D wi::WIDEST_CONST_PRECISION); + WIDE_INT_REF_FOR (T) xi (x); + precision =3D xi.precision; + gcc_assert (precision <=3D RWIDE_INT_MAX_PRECISION); + wi::copy (*this, xi); +} + +template +inline rwide_int_storage& +rwide_int_storage::operator =3D (const T &x) +{ + STATIC_ASSERT (!wi::int_traits::host_dependent_precision); + STATIC_ASSERT (wi::int_traits::precision_type !=3D wi::CONST_PRECISIO= N); + STATIC_ASSERT (wi::int_traits::precision_type +=09=09 !=3D wi::WIDEST_CONST_PRECISION); + WIDE_INT_REF_FOR (T) xi (x); + precision =3D xi.precision; + gcc_assert (precision <=3D RWIDE_INT_MAX_PRECISION); + wi::copy (*this, xi); + return *this; +} + +inline unsigned int +rwide_int_storage::get_precision () const +{ + return precision; +} + +inline const HOST_WIDE_INT * +rwide_int_storage::get_val () const +{ + return val; +} + +inline unsigned int +rwide_int_storage::get_len () const +{ + return len; +} + +inline HOST_WIDE_INT * +rwide_int_storage::write_val (unsigned int) +{ + return val; +} + +inline void +rwide_int_storage::set_len (unsigned int l, bool is_sign_extended) +{ + len =3D l; + if (!is_sign_extended && len * HOST_BITS_PER_WIDE_INT > precision) + val[len - 1] =3D sext_hwi (val[len - 1], +=09=09=09 precision % HOST_BITS_PER_WIDE_INT); +} + +/* Treat X as having signedness SGN and convert it to a PRECISION-bit + number. */ +inline rwide_int +rwide_int_storage::from (const wide_int_ref &x, unsigned int precision, +=09=09=09 signop sgn) +{ + rwide_int result =3D rwide_int::create (precision); + result.set_len (wi::force_to_size (result.write_val (x.len), x.val, x.le= n, +=09=09=09=09 x.precision, precision, sgn)); + return result; +} + +/* Create a rwide_int from the explicit block encoding given by VAL and + LEN. PRECISION is the precision of the integer. NEED_CANON_P is + true if the encoding may have redundant trailing blocks. */ +inline rwide_int +rwide_int_storage::from_array (const HOST_WIDE_INT *val, unsigned int len, +=09=09=09 unsigned int precision, bool need_canon_p) +{ + rwide_int result =3D rwide_int::create (precision); + result.set_len (wi::from_array (result.write_val (len), val, len, precis= ion, +=09=09=09=09 need_canon_p)); + return result; +} + +/* Return an uninitialized rwide_int with precision PRECISION. */ +inline rwide_int +rwide_int_storage::create (unsigned int precision) +{ + rwide_int x; + gcc_assert (precision <=3D RWIDE_INT_MAX_PRECISION); + x.precision =3D precision; + return x; +} + +template +inline rwide_int +wi::int_traits ::get_binary_result (const T1 &x, +=09=09=09=09=09=09 const T2 &y) +{ + /* This shouldn't be used for two flexible-precision inputs. */ + STATIC_ASSERT (wi::int_traits ::precision_type !=3D FLEXIBLE_PRECISI= ON +=09=09 || wi::int_traits ::precision_type !=3D FLEXIBLE_PRECISION); + if (wi::int_traits ::precision_type =3D=3D FLEXIBLE_PRECISION) + return rwide_int::create (wi::get_precision (y)); + else + return rwide_int::create (wi::get_precision (x)); +} + +template +inline unsigned int +wi::int_traits ::get_binary_precision (const T1 &x, +=09=09=09=09=09=09=09 const T2 &y) +{ + /* This shouldn't be used for two flexible-precision inputs. */ + STATIC_ASSERT (wi::int_traits ::precision_type !=3D FLEXIBLE_PRECISI= ON +=09=09 || wi::int_traits ::precision_type !=3D FLEXIBLE_PRECISION); + if (wi::int_traits ::precision_type =3D=3D FLEXIBLE_PRECISION) + return wi::get_precision (y); + else + return wi::get_precision (x); +} + /* The storage used by FIXED_WIDE_INT (N). */ template class GTY(()) fixed_wide_int_storage @@ -1221,7 +1554,7 @@ private: unsigned int len; =20 public: - fixed_wide_int_storage (); + fixed_wide_int_storage () =3D default; template fixed_wide_int_storage (const T &); =20 @@ -1229,7 +1562,7 @@ public: unsigned int get_precision () const; const HOST_WIDE_INT *get_val () const; unsigned int get_len () const; - HOST_WIDE_INT *write_val (); + HOST_WIDE_INT *write_val (unsigned int); void set_len (unsigned int, bool =3D false); =20 static FIXED_WIDE_INT (N) from (const wide_int_ref &, signop); @@ -1245,15 +1578,15 @@ namespace wi static const enum precision_type precision_type =3D CONST_PRECISION; static const bool host_dependent_precision =3D false; static const bool is_sign_extended =3D true; + static const bool needs_write_val_arg =3D false; static const unsigned int precision =3D N; template static FIXED_WIDE_INT (N) get_binary_result (const T1 &, const T2 &); + template + static unsigned int get_binary_precision (const T1 &, const T2 &); }; } =20 -template -inline fixed_wide_int_storage ::fixed_wide_int_storage () {} - /* Initialize the storage from integer X, in precision N. */ template template @@ -1288,7 +1621,7 @@ fixed_wide_int_storage ::get_len () c =20 template inline HOST_WIDE_INT * -fixed_wide_int_storage ::write_val () +fixed_wide_int_storage ::write_val (unsigned int) { return val; } @@ -1308,7 +1641,7 @@ inline FIXED_WIDE_INT (N) fixed_wide_int_storage ::from (const wide_int_ref &x, signop sgn) { FIXED_WIDE_INT (N) result; - result.set_len (wi::force_to_size (result.write_val (), x.val, x.len, + result.set_len (wi::force_to_size (result.write_val (x.len), x.val, x.le= n, =09=09=09=09 x.precision, N, sgn)); return result; } @@ -1323,7 +1656,7 @@ fixed_wide_int_storage ::from_array ( =09=09=09=09=09bool need_canon_p) { FIXED_WIDE_INT (N) result; - result.set_len (wi::from_array (result.write_val (), val, len, + result.set_len (wi::from_array (result.write_val (len), val, len, =09=09=09=09 N, need_canon_p)); return result; } @@ -1337,6 +1670,244 @@ get_binary_result (const T1 &, const T2 return FIXED_WIDE_INT (N) (); } =20 +template +template +inline unsigned int +wi::int_traits < fixed_wide_int_storage >:: +get_binary_precision (const T1 &, const T2 &) +{ + return N; +} + +#define WIDEST_INT(N) generic_wide_int < widest_int_storage > + +/* The storage used by widest_int. */ +template +class GTY(()) widest_int_storage +{ +private: + union + { + HOST_WIDE_INT val[WIDE_INT_MAX_HWIS (N)]; + HOST_WIDE_INT *valp; + } GTY((skip)) u; + unsigned int len; + +public: + widest_int_storage (); + widest_int_storage (const widest_int_storage &); + template + widest_int_storage (const T &); + ~widest_int_storage (); + widest_int_storage &operator =3D (const widest_int_storage &); + template + inline widest_int_storage& operator =3D (const T &); + + /* The standard generic_wide_int storage methods. */ + unsigned int get_precision () const; + const HOST_WIDE_INT *get_val () const; + unsigned int get_len () const; + HOST_WIDE_INT *write_val (unsigned int); + void set_len (unsigned int, bool =3D false); + + static WIDEST_INT (N) from (const wide_int_ref &, signop); + static WIDEST_INT (N) from_array (const HOST_WIDE_INT *, unsigned int, +=09=09=09=09 bool =3D true); +}; + +namespace wi +{ + template + struct int_traits < widest_int_storage > + { + static const enum precision_type precision_type =3D WIDEST_CONST_PRECI= SION; + static const bool host_dependent_precision =3D false; + static const bool is_sign_extended =3D true; + static const bool needs_write_val_arg =3D true; + static const unsigned int precision + =3D N / WIDE_INT_MAX_INL_PRECISION * WIDEST_INT_MAX_PRECISION; + static const unsigned int inl_precision =3D N; + template + static WIDEST_INT (N) get_binary_result (const T1 &, const T2 &); + template + static unsigned int get_binary_precision (const T1 &, const T2 &); + }; +} + +template +inline widest_int_storage ::widest_int_storage () : len (0) {} + +/* Initialize the storage from integer X, in precision N. */ +template +template +inline widest_int_storage ::widest_int_storage (const T &x) : len (0) +{ + /* Check for type compatibility. We don't want to initialize a + widest integer from something like a wide_int. */ + WI_BINARY_RESULT (T, WIDEST_INT (N)) *assertion ATTRIBUTE_UNUSED; + wi::copy (*this, WIDE_INT_REF_FOR (T) (x, N / WIDE_INT_MAX_INL_PRECISION +=09=09=09=09=09 * WIDEST_INT_MAX_PRECISION)); +} + +template +inline +widest_int_storage ::widest_int_storage (const widest_int_storage &x) +{ + len =3D x.len; + if (UNLIKELY (len > N / HOST_BITS_PER_WIDE_INT)) + { + u.valp =3D XNEWVEC (HOST_WIDE_INT, len); + memcpy (u.valp, x.u.valp, len * sizeof (HOST_WIDE_INT)); + } + else + memcpy (u.val, x.u.val, len * sizeof (HOST_WIDE_INT)); +} + +template +inline widest_int_storage ::~widest_int_storage () +{ + if (UNLIKELY (len > N / HOST_BITS_PER_WIDE_INT)) + XDELETEVEC (u.valp); +} + +template +inline widest_int_storage & +widest_int_storage ::operator =3D (const widest_int_storage &x) +{ + if (UNLIKELY (len > N / HOST_BITS_PER_WIDE_INT)) + { + if (this =3D=3D &x) +=09return *this; + XDELETEVEC (u.valp); + } + len =3D x.len; + if (UNLIKELY (len > N / HOST_BITS_PER_WIDE_INT)) + { + u.valp =3D XNEWVEC (HOST_WIDE_INT, len); + memcpy (u.valp, x.u.valp, len * sizeof (HOST_WIDE_INT)); + } + else + memcpy (u.val, x.u.val, len * sizeof (HOST_WIDE_INT)); + return *this; +} + +template +template +inline widest_int_storage & +widest_int_storage ::operator =3D (const T &x) +{ + /* Check for type compatibility. We don't want to assign a + widest integer from something like a wide_int. */ + WI_BINARY_RESULT (T, WIDEST_INT (N)) *assertion ATTRIBUTE_UNUSED; + if (UNLIKELY (len > N / HOST_BITS_PER_WIDE_INT)) + XDELETEVEC (u.valp); + len =3D 0; + wi::copy (*this, WIDE_INT_REF_FOR (T) (x, N / WIDE_INT_MAX_INL_PRECISION +=09=09=09=09=09 * WIDEST_INT_MAX_PRECISION)); + return *this; +} + +template +inline unsigned int +widest_int_storage ::get_precision () const +{ + return N / WIDE_INT_MAX_INL_PRECISION * WIDEST_INT_MAX_PRECISION; +} + +template +inline const HOST_WIDE_INT * +widest_int_storage ::get_val () const +{ + return UNLIKELY (len > N / HOST_BITS_PER_WIDE_INT) ? u.valp : u.val; +} + +template +inline unsigned int +widest_int_storage ::get_len () const +{ + return len; +} + +template +inline HOST_WIDE_INT * +widest_int_storage ::write_val (unsigned int l) +{ + if (UNLIKELY (len > N / HOST_BITS_PER_WIDE_INT)) + XDELETEVEC (u.valp); + len =3D l; + if (UNLIKELY (l > N / HOST_BITS_PER_WIDE_INT)) + { + u.valp =3D XNEWVEC (HOST_WIDE_INT, l); + return u.valp; + } + return u.val; +} + +template +inline void +widest_int_storage ::set_len (unsigned int l, bool) +{ + gcc_checking_assert (l <=3D len); + if (UNLIKELY (len > N / HOST_BITS_PER_WIDE_INT) + && l <=3D N / HOST_BITS_PER_WIDE_INT) + { + HOST_WIDE_INT *valp =3D u.valp; + memcpy (u.val, valp, l * sizeof (u.val[0])); + XDELETEVEC (valp); + } + len =3D l; + /* There are no excess bits in val[len - 1]. */ + STATIC_ASSERT (N % HOST_BITS_PER_WIDE_INT =3D=3D 0); +} + +/* Treat X as having signedness SGN and convert it to an N-bit number. */ +template +inline WIDEST_INT (N) +widest_int_storage ::from (const wide_int_ref &x, signop sgn) +{ + WIDEST_INT (N) result; + unsigned int exp_len =3D x.len; + unsigned int prec =3D result.get_precision (); + if (sgn =3D=3D UNSIGNED && prec > x.precision && x.val[x.len - 1] < 0) + exp_len =3D CEIL (x.precision, HOST_BITS_PER_WIDE_INT) + 1; + result.set_len (wi::force_to_size (result.write_val (exp_len), x.val, x.= len, +=09=09=09=09 x.precision, prec, sgn)); + return result; +} + +/* Create a WIDEST_INT (N) from the explicit block encoding given by + VAL and LEN. NEED_CANON_P is true if the encoding may have redundant + trailing blocks. */ +template +inline WIDEST_INT (N) +widest_int_storage ::from_array (const HOST_WIDE_INT *val, +=09=09=09=09 unsigned int len, +=09=09=09=09 bool need_canon_p) +{ + WIDEST_INT (N) result; + result.set_len (wi::from_array (result.write_val (len), val, len, +=09=09=09=09 result.get_precision (), need_canon_p)); + return result; +} + +template +template +inline WIDEST_INT (N) +wi::int_traits < widest_int_storage >:: +get_binary_result (const T1 &, const T2 &) +{ + return WIDEST_INT (N) (); +} + +template +template +inline unsigned int +wi::int_traits < widest_int_storage >:: +get_binary_precision (const T1 &, const T2 &) +{ + return N / WIDE_INT_MAX_INL_PRECISION * WIDEST_INT_MAX_PRECISION; +} + /* A reference to one element of a trailing_wide_ints structure. */ class trailing_wide_int_storage { @@ -1359,7 +1930,7 @@ public: unsigned int get_len () const; unsigned int get_precision () const; const HOST_WIDE_INT *get_val () const; - HOST_WIDE_INT *write_val (); + HOST_WIDE_INT *write_val (unsigned int); void set_len (unsigned int, bool =3D false); =20 template @@ -1445,7 +2016,7 @@ trailing_wide_int_storage::get_val () co } =20 inline HOST_WIDE_INT * -trailing_wide_int_storage::write_val () +trailing_wide_int_storage::write_val (unsigned int) { return m_val; } @@ -1528,6 +2099,7 @@ namespace wi static const enum precision_type precision_type =3D FLEXIBLE_PRECISION= ; static const bool host_dependent_precision =3D true; static const bool is_sign_extended =3D true; + static const bool needs_write_val_arg =3D false; static unsigned int get_precision (T); static wi::storage_ref decompose (HOST_WIDE_INT *, unsigned int, T); }; @@ -1699,6 +2271,7 @@ namespace wi precision of HOST_WIDE_INT. */ static const bool host_dependent_precision =3D false; static const bool is_sign_extended =3D true; + static const bool needs_write_val_arg =3D false; static unsigned int get_precision (const wi::hwi_with_prec &); static wi::storage_ref decompose (HOST_WIDE_INT *, unsigned int, =09=09=09=09 const wi::hwi_with_prec &); @@ -1804,8 +2377,8 @@ template inline unsigned int wi::get_binary_precision (const T1 &x, const T2 &y) { - return get_precision (wi::int_traits :: -=09=09=09get_binary_result (x, y)); + return wi::int_traits ::get_binary_precision = (x, +=09=09=09=09=09=09=09=09=09 y); } =20 /* Copy the contents of Y to X, but keeping X's current precision. */ @@ -1813,14 +2386,17 @@ template inline void wi::copy (T1 &x, const T2 &y) { - HOST_WIDE_INT *xval =3D x.write_val (); - const HOST_WIDE_INT *yval =3D y.get_val (); unsigned int len =3D y.get_len (); + HOST_WIDE_INT *xval =3D x.write_val (len); + const HOST_WIDE_INT *yval =3D y.get_val (); unsigned int i =3D 0; do xval[i] =3D yval[i]; while (++i < len); - x.set_len (len, y.is_sign_extended); + /* For widest_int write_val is called with an exact value, not + upper bound for len, so nothing is needed further. */ + if (!wi::int_traits ::needs_write_val_arg) + x.set_len (len, y.is_sign_extended); } =20 /* Return true if X fits in a HOST_WIDE_INT with no loss of precision. */ @@ -2162,6 +2738,8 @@ wi::bit_not (const T &x) { WI_UNARY_RESULT_VAR (result, val, T, x); WIDE_INT_REF_FOR (T) xi (x, get_precision (result)); + if (result.needs_write_val_arg) + val =3D result.write_val (xi.len); for (unsigned int i =3D 0; i < xi.len; ++i) val[i] =3D ~xi.val[i]; result.set_len (xi.len); @@ -2203,6 +2781,9 @@ wi::sext (const T &x, unsigned int offse unsigned int precision =3D get_precision (result); WIDE_INT_REF_FOR (T) xi (x, precision); =20 + if (result.needs_write_val_arg) + val =3D result.write_val (MAX (xi.len, +=09=09=09=09 CEIL (offset, HOST_BITS_PER_WIDE_INT))); if (offset <=3D HOST_BITS_PER_WIDE_INT) { val[0] =3D sext_hwi (xi.ulow (), offset); @@ -2230,6 +2811,9 @@ wi::zext (const T &x, unsigned int offse return result; } =20 + if (result.needs_write_val_arg) + val =3D result.write_val (MAX (xi.len, +=09=09=09=09 offset / HOST_BITS_PER_WIDE_INT + 1)); /* In these cases we know that at least the top bit will be clear, so no sign extension is necessary. */ if (offset < HOST_BITS_PER_WIDE_INT) @@ -2259,6 +2843,9 @@ wi::set_bit (const T &x, unsigned int bi WI_UNARY_RESULT_VAR (result, val, T, x); unsigned int precision =3D get_precision (result); WIDE_INT_REF_FOR (T) xi (x, precision); + if (result.needs_write_val_arg) + val =3D result.write_val (MAX (xi.len, +=09=09=09=09 bit / HOST_BITS_PER_WIDE_INT + 1)); if (precision <=3D HOST_BITS_PER_WIDE_INT) { val[0] =3D xi.ulow () | (HOST_WIDE_INT_1U << bit); @@ -2280,6 +2867,8 @@ wi::bswap (const T &x) WI_UNARY_RESULT_VAR (result, val, T, x); unsigned int precision =3D get_precision (result); WIDE_INT_REF_FOR (T) xi (x, precision); + if (result.needs_write_val_arg) + gcc_unreachable (); /* bswap on widest_int makes no sense. */ result.set_len (bswap_large (val, xi.val, xi.len, precision)); return result; } @@ -2292,6 +2881,8 @@ wi::bitreverse (const T &x) WI_UNARY_RESULT_VAR (result, val, T, x); unsigned int precision =3D get_precision (result); WIDE_INT_REF_FOR (T) xi (x, precision); + if (result.needs_write_val_arg) + gcc_unreachable (); /* bitreverse on widest_int makes no sense. */ result.set_len (bitreverse_large (val, xi.val, xi.len, precision)); return result; } @@ -2368,6 +2959,8 @@ wi::bit_and (const T1 &x, const T2 &y) WIDE_INT_REF_FOR (T1) xi (x, precision); WIDE_INT_REF_FOR (T2) yi (y, precision); bool is_sign_extended =3D xi.is_sign_extended && yi.is_sign_extended; + if (result.needs_write_val_arg) + val =3D result.write_val (MAX (xi.len, yi.len)); if (LIKELY (xi.len + yi.len =3D=3D 2)) { val[0] =3D xi.ulow () & yi.ulow (); @@ -2389,6 +2982,8 @@ wi::bit_and_not (const T1 &x, const T2 & WIDE_INT_REF_FOR (T1) xi (x, precision); WIDE_INT_REF_FOR (T2) yi (y, precision); bool is_sign_extended =3D xi.is_sign_extended && yi.is_sign_extended; + if (result.needs_write_val_arg) + val =3D result.write_val (MAX (xi.len, yi.len)); if (LIKELY (xi.len + yi.len =3D=3D 2)) { val[0] =3D xi.ulow () & ~yi.ulow (); @@ -2410,6 +3005,8 @@ wi::bit_or (const T1 &x, const T2 &y) WIDE_INT_REF_FOR (T1) xi (x, precision); WIDE_INT_REF_FOR (T2) yi (y, precision); bool is_sign_extended =3D xi.is_sign_extended && yi.is_sign_extended; + if (result.needs_write_val_arg) + val =3D result.write_val (MAX (xi.len, yi.len)); if (LIKELY (xi.len + yi.len =3D=3D 2)) { val[0] =3D xi.ulow () | yi.ulow (); @@ -2431,6 +3028,8 @@ wi::bit_or_not (const T1 &x, const T2 &y WIDE_INT_REF_FOR (T1) xi (x, precision); WIDE_INT_REF_FOR (T2) yi (y, precision); bool is_sign_extended =3D xi.is_sign_extended && yi.is_sign_extended; + if (result.needs_write_val_arg) + val =3D result.write_val (MAX (xi.len, yi.len)); if (LIKELY (xi.len + yi.len =3D=3D 2)) { val[0] =3D xi.ulow () | ~yi.ulow (); @@ -2452,6 +3051,8 @@ wi::bit_xor (const T1 &x, const T2 &y) WIDE_INT_REF_FOR (T1) xi (x, precision); WIDE_INT_REF_FOR (T2) yi (y, precision); bool is_sign_extended =3D xi.is_sign_extended && yi.is_sign_extended; + if (result.needs_write_val_arg) + val =3D result.write_val (MAX (xi.len, yi.len)); if (LIKELY (xi.len + yi.len =3D=3D 2)) { val[0] =3D xi.ulow () ^ yi.ulow (); @@ -2472,6 +3073,8 @@ wi::add (const T1 &x, const T2 &y) unsigned int precision =3D get_precision (result); WIDE_INT_REF_FOR (T1) xi (x, precision); WIDE_INT_REF_FOR (T2) yi (y, precision); + if (result.needs_write_val_arg) + val =3D result.write_val (MAX (xi.len, yi.len) + 1); if (precision <=3D HOST_BITS_PER_WIDE_INT) { val[0] =3D xi.ulow () + yi.ulow (); @@ -2515,6 +3118,8 @@ wi::add (const T1 &x, const T2 &y, signo unsigned int precision =3D get_precision (result); WIDE_INT_REF_FOR (T1) xi (x, precision); WIDE_INT_REF_FOR (T2) yi (y, precision); + if (result.needs_write_val_arg) + val =3D result.write_val (MAX (xi.len, yi.len) + 1); if (precision <=3D HOST_BITS_PER_WIDE_INT) { unsigned HOST_WIDE_INT xl =3D xi.ulow (); @@ -2558,6 +3163,8 @@ wi::sub (const T1 &x, const T2 &y) unsigned int precision =3D get_precision (result); WIDE_INT_REF_FOR (T1) xi (x, precision); WIDE_INT_REF_FOR (T2) yi (y, precision); + if (result.needs_write_val_arg) + val =3D result.write_val (MAX (xi.len, yi.len) + 1); if (precision <=3D HOST_BITS_PER_WIDE_INT) { val[0] =3D xi.ulow () - yi.ulow (); @@ -2601,6 +3208,8 @@ wi::sub (const T1 &x, const T2 &y, signo unsigned int precision =3D get_precision (result); WIDE_INT_REF_FOR (T1) xi (x, precision); WIDE_INT_REF_FOR (T2) yi (y, precision); + if (result.needs_write_val_arg) + val =3D result.write_val (MAX (xi.len, yi.len) + 1); if (precision <=3D HOST_BITS_PER_WIDE_INT) { unsigned HOST_WIDE_INT xl =3D xi.ulow (); @@ -2643,6 +3252,8 @@ wi::mul (const T1 &x, const T2 &y) unsigned int precision =3D get_precision (result); WIDE_INT_REF_FOR (T1) xi (x, precision); WIDE_INT_REF_FOR (T2) yi (y, precision); + if (result.needs_write_val_arg) + val =3D result.write_val (xi.len + yi.len + 2); if (precision <=3D HOST_BITS_PER_WIDE_INT) { val[0] =3D xi.ulow () * yi.ulow (); @@ -2664,6 +3275,8 @@ wi::mul (const T1 &x, const T2 &y, signo unsigned int precision =3D get_precision (result); WIDE_INT_REF_FOR (T1) xi (x, precision); WIDE_INT_REF_FOR (T2) yi (y, precision); + if (result.needs_write_val_arg) + val =3D result.write_val (xi.len + yi.len + 2); result.set_len (mul_internal (val, xi.val, xi.len, =09=09=09=09yi.val, yi.len, precision, =09=09=09=09sgn, overflow, false)); @@ -2698,6 +3311,8 @@ wi::mul_high (const T1 &x, const T2 &y, unsigned int precision =3D get_precision (result); WIDE_INT_REF_FOR (T1) xi (x, precision); WIDE_INT_REF_FOR (T2) yi (y, precision); + if (result.needs_write_val_arg) + gcc_unreachable (); /* mul_high on widest_int doesn't make sense. */ result.set_len (mul_internal (val, xi.val, xi.len, =09=09=09=09yi.val, yi.len, precision, =09=09=09=09sgn, 0, true)); @@ -2716,6 +3331,12 @@ wi::div_trunc (const T1 &x, const T2 &y, WIDE_INT_REF_FOR (T1) xi (x, precision); WIDE_INT_REF_FOR (T2) yi (y); =20 + if (quotient.needs_write_val_arg) + quotient_val =3D quotient.write_val ((sgn =3D=3D UNSIGNED +=09=09=09=09=09&& xi.val[xi.len - 1] < 0) +=09=09=09=09 ? CEIL (precision, +=09=09=09=09=09 HOST_BITS_PER_WIDE_INT) + 1 +=09=09=09=09 : xi.len + 1); quotient.set_len (divmod_internal (quotient_val, 0, 0, xi.val, xi.len, =09=09=09=09 precision, =09=09=09=09 yi.val, yi.len, yi.precision, @@ -2753,6 +3374,15 @@ wi::div_floor (const T1 &x, const T2 &y, WIDE_INT_REF_FOR (T2) yi (y); =20 unsigned int remainder_len; + if (quotient.needs_write_val_arg) + { + quotient_val =3D quotient.write_val ((sgn =3D=3D UNSIGNED +=09=09=09=09=09 && xi.val[xi.len - 1] < 0) +=09=09=09=09=09 ? CEIL (precision, +=09=09=09=09=09=09 HOST_BITS_PER_WIDE_INT) + 1 +=09=09=09=09=09 : xi.len + 1); + remainder_val =3D remainder.write_val (yi.len); + } quotient.set_len (divmod_internal (quotient_val, =09=09=09=09 &remainder_len, remainder_val, =09=09=09=09 xi.val, xi.len, precision, @@ -2795,6 +3425,15 @@ wi::div_ceil (const T1 &x, const T2 &y, WIDE_INT_REF_FOR (T2) yi (y); =20 unsigned int remainder_len; + if (quotient.needs_write_val_arg) + { + quotient_val =3D quotient.write_val ((sgn =3D=3D UNSIGNED +=09=09=09=09=09 && xi.val[xi.len - 1] < 0) +=09=09=09=09=09 ? CEIL (precision, +=09=09=09=09=09=09 HOST_BITS_PER_WIDE_INT) + 1 +=09=09=09=09=09 : xi.len + 1); + remainder_val =3D remainder.write_val (yi.len); + } quotient.set_len (divmod_internal (quotient_val, =09=09=09=09 &remainder_len, remainder_val, =09=09=09=09 xi.val, xi.len, precision, @@ -2828,6 +3467,15 @@ wi::div_round (const T1 &x, const T2 &y, WIDE_INT_REF_FOR (T2) yi (y); =20 unsigned int remainder_len; + if (quotient.needs_write_val_arg) + { + quotient_val =3D quotient.write_val ((sgn =3D=3D UNSIGNED +=09=09=09=09=09 && xi.val[xi.len - 1] < 0) +=09=09=09=09=09 ? CEIL (precision, +=09=09=09=09=09=09 HOST_BITS_PER_WIDE_INT) + 1 +=09=09=09=09=09 : xi.len + 1); + remainder_val =3D remainder.write_val (yi.len); + } quotient.set_len (divmod_internal (quotient_val, =09=09=09=09 &remainder_len, remainder_val, =09=09=09=09 xi.val, xi.len, precision, @@ -2871,6 +3519,15 @@ wi::divmod_trunc (const T1 &x, const T2 WIDE_INT_REF_FOR (T2) yi (y); =20 unsigned int remainder_len; + if (quotient.needs_write_val_arg) + { + quotient_val =3D quotient.write_val ((sgn =3D=3D UNSIGNED +=09=09=09=09=09 && xi.val[xi.len - 1] < 0) +=09=09=09=09=09 ? CEIL (precision, +=09=09=09=09=09=09 HOST_BITS_PER_WIDE_INT) + 1 +=09=09=09=09=09 : xi.len + 1); + remainder_val =3D remainder.write_val (yi.len); + } quotient.set_len (divmod_internal (quotient_val, =09=09=09=09 &remainder_len, remainder_val, =09=09=09=09 xi.val, xi.len, precision, @@ -2915,6 +3572,8 @@ wi::mod_trunc (const T1 &x, const T2 &y, WIDE_INT_REF_FOR (T2) yi (y); =20 unsigned int remainder_len; + if (remainder.needs_write_val_arg) + remainder_val =3D remainder.write_val (yi.len); divmod_internal (0, &remainder_len, remainder_val, =09=09 xi.val, xi.len, precision, =09=09 yi.val, yi.len, yi.precision, sgn, overflow); @@ -2955,6 +3614,15 @@ wi::mod_floor (const T1 &x, const T2 &y, WIDE_INT_REF_FOR (T2) yi (y); =20 unsigned int remainder_len; + if (quotient.needs_write_val_arg) + { + quotient_val =3D quotient.write_val ((sgn =3D=3D UNSIGNED +=09=09=09=09=09 && xi.val[xi.len - 1] < 0) +=09=09=09=09=09 ? CEIL (precision, +=09=09=09=09=09=09 HOST_BITS_PER_WIDE_INT) + 1 +=09=09=09=09=09 : xi.len + 1); + remainder_val =3D remainder.write_val (yi.len); + } quotient.set_len (divmod_internal (quotient_val, =09=09=09=09 &remainder_len, remainder_val, =09=09=09=09 xi.val, xi.len, precision, @@ -2991,6 +3659,15 @@ wi::mod_ceil (const T1 &x, const T2 &y, WIDE_INT_REF_FOR (T2) yi (y); =20 unsigned int remainder_len; + if (quotient.needs_write_val_arg) + { + quotient_val =3D quotient.write_val ((sgn =3D=3D UNSIGNED +=09=09=09=09=09 && xi.val[xi.len - 1] < 0) +=09=09=09=09=09 ? CEIL (precision, +=09=09=09=09=09=09 HOST_BITS_PER_WIDE_INT) + 1 +=09=09=09=09=09 : xi.len + 1); + remainder_val =3D remainder.write_val (yi.len); + } quotient.set_len (divmod_internal (quotient_val, =09=09=09=09 &remainder_len, remainder_val, =09=09=09=09 xi.val, xi.len, precision, @@ -3017,6 +3694,15 @@ wi::mod_round (const T1 &x, const T2 &y, WIDE_INT_REF_FOR (T2) yi (y); =20 unsigned int remainder_len; + if (quotient.needs_write_val_arg) + { + quotient_val =3D quotient.write_val ((sgn =3D=3D UNSIGNED +=09=09=09=09=09 && xi.val[xi.len - 1] < 0) +=09=09=09=09=09 ? CEIL (precision, +=09=09=09=09=09=09 HOST_BITS_PER_WIDE_INT) + 1 +=09=09=09=09=09 : xi.len + 1); + remainder_val =3D remainder.write_val (yi.len); + } quotient.set_len (divmod_internal (quotient_val, =09=09=09=09 &remainder_len, remainder_val, =09=09=09=09 xi.val, xi.len, precision, @@ -3086,12 +3772,16 @@ wi::lshift (const T1 &x, const T2 &y) /* Handle the simple cases quickly. */ if (geu_p (yi, precision)) { + if (result.needs_write_val_arg) +=09val =3D result.write_val (1); val[0] =3D 0; result.set_len (1); } else { unsigned int shift =3D yi.to_uhwi (); + if (result.needs_write_val_arg) +=09val =3D result.write_val (xi.len + shift / HOST_BITS_PER_WIDE_INT + 1); /* For fixed-precision integers like offset_int and widest_int, =09 handle the case where the shift value is constant and the =09 result is a single nonnegative HWI (meaning that we don't @@ -3130,12 +3820,23 @@ wi::lrshift (const T1 &x, const T2 &y) /* Handle the simple cases quickly. */ if (geu_p (yi, xi.precision)) { + if (result.needs_write_val_arg) +=09val =3D result.write_val (1); val[0] =3D 0; result.set_len (1); } else { unsigned int shift =3D yi.to_uhwi (); + if (result.needs_write_val_arg) +=09{ +=09 unsigned int est_len =3D xi.len; +=09 if (xi.val[xi.len - 1] < 0 && shift) +=09 /* Logical right shift of sign-extended value might need a very +=09 large precision e.g. for widest_int. */ +=09 est_len =3D CEIL (xi.precision - shift, HOST_BITS_PER_WIDE_INT) + 1= ; +=09 val =3D result.write_val (est_len); +=09} /* For fixed-precision integers like offset_int and widest_int, =09 handle the case where the shift value is constant and the =09 shifted value is a single nonnegative HWI (meaning that all @@ -3171,6 +3872,8 @@ wi::arshift (const T1 &x, const T2 &y) since the result can be no larger than that. */ WIDE_INT_REF_FOR (T1) xi (x); WIDE_INT_REF_FOR (T2) yi (y); + if (result.needs_write_val_arg) + val =3D result.write_val (xi.len); /* Handle the simple cases quickly. */ if (geu_p (yi, xi.precision)) { @@ -3374,25 +4077,56 @@ operator % (const T1 &x, const T2 &y) return wi::smod_trunc (x, y); } =20 -template +void gt_ggc_mx (generic_wide_int *) =3D delete; +void gt_pch_nx (generic_wide_int *) =3D delete; +void gt_pch_nx (generic_wide_int *, +=09=09gt_pointer_operator, void *) =3D delete; + +inline void +gt_ggc_mx (generic_wide_int *) +{ +} + +inline void +gt_pch_nx (generic_wide_int *) +{ +} + +inline void +gt_pch_nx (generic_wide_int *, gt_pointer_operator, vo= id *) +{ +} + +template void -gt_ggc_mx (generic_wide_int *) +gt_ggc_mx (generic_wide_int > *) { } =20 -template +template void -gt_pch_nx (generic_wide_int *) +gt_pch_nx (generic_wide_int > *) { } =20 -template +template void -gt_pch_nx (generic_wide_int *, gt_pointer_operator, void *) +gt_pch_nx (generic_wide_int > *, +=09 gt_pointer_operator, void *) { } =20 template +void gt_ggc_mx (generic_wide_int > *) =3D delete; + +template +void gt_pch_nx (generic_wide_int > *) =3D delete; + +template +void gt_pch_nx (generic_wide_int > *, +=09=09gt_pointer_operator, void *) =3D delete; + +template void gt_ggc_mx (trailing_wide_ints *) { @@ -3465,7 +4199,7 @@ inline wide_int wi::mask (unsigned int width, bool negate_p, unsigned int precision) { wide_int result =3D wide_int::create (precision); - result.set_len (mask (result.write_val (), width, negate_p, precision)); + result.set_len (mask (result.write_val (0), width, negate_p, precision))= ; return result; } =20 @@ -3477,7 +4211,7 @@ wi::shifted_mask (unsigned int start, un =09=09 unsigned int precision) { wide_int result =3D wide_int::create (precision); - result.set_len (shifted_mask (result.write_val (), start, width, negate_= p, + result.set_len (shifted_mask (result.write_val (0), start, width, negate= _p, =09=09=09=09precision)); return result; } @@ -3498,8 +4232,8 @@ wi::mask (unsigned int width, bool negat { STATIC_ASSERT (wi::int_traits::precision); T result; - result.set_len (mask (result.write_val (), width, negate_p, -=09=09=09wi::int_traits ::precision)); + result.set_len (mask (result.write_val (width / HOST_BITS_PER_WIDE_INT += 1), +=09=09=09width, negate_p, wi::int_traits ::precision)); return result; } =20 @@ -3512,9 +4246,13 @@ wi::shifted_mask (unsigned int start, un { STATIC_ASSERT (wi::int_traits::precision); T result; - result.set_len (shifted_mask (result.write_val (), start, width, -=09=09=09=09negate_p, -=09=09=09=09wi::int_traits ::precision)); + unsigned int prec =3D wi::int_traits ::precision; + unsigned int est_len + =3D result.needs_write_val_arg + ? ((start + (width > prec - start ? prec - start : width)) +=09 / HOST_BITS_PER_WIDE_INT + 1) : 0; + result.set_len (shifted_mask (result.write_val (est_len), start, width, +=09=09=09=09negate_p, prec)); return result; } =20 --- gcc/wide-int.cc.jj=092023-10-08 16:37:32.094269231 +0200 +++ gcc/wide-int.cc=092023-10-08 17:00:12.096716853 +0200 @@ -51,7 +51,7 @@ typedef unsigned int UDWtype __attribute #include "longlong.h" #endif =20 -static const HOST_WIDE_INT zeros[WIDE_INT_MAX_ELTS] =3D {}; +static const HOST_WIDE_INT zeros[1] =3D {}; =20 /* * Internal utilities. @@ -62,8 +62,7 @@ static const HOST_WIDE_INT zeros[WIDE_IN #define HALF_INT_MASK ((HOST_WIDE_INT_1 << HOST_BITS_PER_HALF_WIDE_INT) - = 1) =20 #define BLOCK_OF(TARGET) ((TARGET) / HOST_BITS_PER_WIDE_INT) -#define BLOCKS_NEEDED(PREC) \ - (PREC ? (((PREC) + HOST_BITS_PER_WIDE_INT - 1) / HOST_BITS_PER_WIDE_INT)= : 1) +#define BLOCKS_NEEDED(PREC) (PREC ? CEIL (PREC, HOST_BITS_PER_WIDE_INT) : = 1) #define SIGN_MASK(X) ((HOST_WIDE_INT) (X) < 0 ? -1 : 0) =20 /* Return the value a VAL[I] if I < LEN, otherwise, return 0 or -1 @@ -96,7 +95,7 @@ canonize (HOST_WIDE_INT *val, unsigned i top =3D val[len - 1]; if (len * HOST_BITS_PER_WIDE_INT > precision) val[len - 1] =3D top =3D sext_hwi (top, precision % HOST_BITS_PER_WIDE= _INT); - if (top !=3D 0 && top !=3D (HOST_WIDE_INT)-1) + if (top !=3D 0 && top !=3D HOST_WIDE_INT_M1) return len; =20 /* At this point we know that the top is either 0 or -1. Find the @@ -163,7 +162,7 @@ wi::from_buffer (const unsigned char *bu /* We have to clear all the bits ourself, as we merely or in values below. */ unsigned int len =3D BLOCKS_NEEDED (precision); - HOST_WIDE_INT *val =3D result.write_val (); + HOST_WIDE_INT *val =3D result.write_val (0); for (unsigned int i =3D 0; i < len; ++i) val[i] =3D 0; =20 @@ -232,8 +231,7 @@ wi::to_mpz (const wide_int_ref &x, mpz_t } else if (excess < 0 && wi::neg_p (x)) { - int extra -=09=3D (-excess + HOST_BITS_PER_WIDE_INT - 1) / HOST_BITS_PER_WIDE_INT; + int extra =3D CEIL (-excess, HOST_BITS_PER_WIDE_INT); HOST_WIDE_INT *t =3D XALLOCAVEC (HOST_WIDE_INT, len + extra); for (int i =3D 0; i < len; i++) =09t[i] =3D v[i]; @@ -280,8 +278,8 @@ wi::from_mpz (const_tree type, mpz_t x, extracted from the GMP manual, section "Integer Import and Export": http://gmplib.org/manual/Integer-Import-and-Export.html */ numb =3D CHAR_BIT * sizeof (HOST_WIDE_INT); - count =3D (mpz_sizeinbase (x, 2) + numb - 1) / numb; - HOST_WIDE_INT *val =3D res.write_val (); + count =3D CEIL (mpz_sizeinbase (x, 2), numb); + HOST_WIDE_INT *val =3D res.write_val (0); /* Read the absolute value. =20 Write directly to the wide_int storage if possible, otherwise leave @@ -289,7 +287,7 @@ wi::from_mpz (const_tree type, mpz_t x, to use mpz_tdiv_r_2exp for the latter case, but the situation is pathological and it seems safer to operate on the original mpz value in all cases. */ - void *valres =3D mpz_export (count <=3D WIDE_INT_MAX_ELTS ? val : 0, + void *valres =3D mpz_export (count <=3D WIDE_INT_MAX_INL_ELTS ? val : 0, =09=09=09 &count, -1, sizeof (HOST_WIDE_INT), 0, 0, x); if (count < 1) { @@ -1334,21 +1332,6 @@ wi::mul_internal (HOST_WIDE_INT *val, co unsigned HOST_WIDE_INT o0, o1, k, t; unsigned int i; unsigned int j; - unsigned int blocks_needed =3D BLOCKS_NEEDED (prec); - unsigned int half_blocks_needed =3D blocks_needed * 2; - /* The sizes here are scaled to support a 2x largest mode by 2x - largest mode yielding a 4x largest mode result. This is what is - needed by vpn. */ - - unsigned HOST_HALF_WIDE_INT - u[4 * MAX_BITSIZE_MODE_ANY_INT / HOST_BITS_PER_HALF_WIDE_INT]; - unsigned HOST_HALF_WIDE_INT - v[4 * MAX_BITSIZE_MODE_ANY_INT / HOST_BITS_PER_HALF_WIDE_INT]; - /* The '2' in 'R' is because we are internally doing a full - multiply. */ - unsigned HOST_HALF_WIDE_INT - r[2 * 4 * MAX_BITSIZE_MODE_ANY_INT / HOST_BITS_PER_HALF_WIDE_INT]; - HOST_WIDE_INT mask =3D ((HOST_WIDE_INT)1 << HOST_BITS_PER_HALF_WIDE_INT)= - 1; =20 /* If the top level routine did not really pass in an overflow, then just make sure that we never attempt to set it. */ @@ -1469,6 +1452,36 @@ wi::mul_internal (HOST_WIDE_INT *val, co return 1; } =20 + /* The sizes here are scaled to support a 2x WIDE_INT_MAX_INL_PRECISION = by 2x + WIDE_INT_MAX_INL_PRECISION yielding a 4x WIDE_INT_MAX_INL_PRECISION + result. */ + + unsigned HOST_HALF_WIDE_INT + ubuf[4 * WIDE_INT_MAX_INL_PRECISION / HOST_BITS_PER_HALF_WIDE_INT]; + unsigned HOST_HALF_WIDE_INT + vbuf[4 * WIDE_INT_MAX_INL_PRECISION / HOST_BITS_PER_HALF_WIDE_INT]; + /* The '2' in 'R' is because we are internally doing a full + multiply. */ + unsigned HOST_HALF_WIDE_INT + rbuf[2 * 4 * WIDE_INT_MAX_INL_PRECISION / HOST_BITS_PER_HALF_WIDE_INT]= ; + const HOST_WIDE_INT mask =3D ((HOST_WIDE_INT)1 << HOST_BITS_PER_HALF_WID= E_INT) - 1; + unsigned HOST_HALF_WIDE_INT *u =3D ubuf; + unsigned HOST_HALF_WIDE_INT *v =3D vbuf; + unsigned HOST_HALF_WIDE_INT *r =3D rbuf; + + if (prec > WIDE_INT_MAX_INL_PRECISION && !high) + prec =3D (op1len + op2len + 1) * HOST_BITS_PER_WIDE_INT; + unsigned int blocks_needed =3D BLOCKS_NEEDED (prec); + unsigned int half_blocks_needed =3D blocks_needed * 2; + if (UNLIKELY (prec > WIDE_INT_MAX_INL_PRECISION)) + { + unsigned HOST_HALF_WIDE_INT *buf +=09=3D XALLOCAVEC (unsigned HOST_HALF_WIDE_INT, 4 * 4 * blocks_needed); + u =3D buf; + v =3D u + 4 * blocks_needed; + r =3D v + 4 * blocks_needed; + } + /* We do unsigned mul and then correct it. */ wi_unpack (u, op1val, op1len, half_blocks_needed, prec, SIGNED); wi_unpack (v, op2val, op2len, half_blocks_needed, prec, SIGNED); @@ -1782,16 +1795,6 @@ wi::divmod_internal (HOST_WIDE_INT *quot =09=09 unsigned int divisor_prec, signop sgn, =09=09 wi::overflow_type *oflow) { - unsigned int dividend_blocks_needed =3D 2 * BLOCKS_NEEDED (dividend_prec= ); - unsigned int divisor_blocks_needed =3D 2 * BLOCKS_NEEDED (divisor_prec); - unsigned HOST_HALF_WIDE_INT - b_quotient[4 * MAX_BITSIZE_MODE_ANY_INT / HOST_BITS_PER_HALF_WIDE_INT]= ; - unsigned HOST_HALF_WIDE_INT - b_remainder[4 * MAX_BITSIZE_MODE_ANY_INT / HOST_BITS_PER_HALF_WIDE_INT= ]; - unsigned HOST_HALF_WIDE_INT - b_dividend[(4 * MAX_BITSIZE_MODE_ANY_INT / HOST_BITS_PER_HALF_WIDE_INT= ) + 1]; - unsigned HOST_HALF_WIDE_INT - b_divisor[4 * MAX_BITSIZE_MODE_ANY_INT / HOST_BITS_PER_HALF_WIDE_INT]; unsigned int m, n; bool dividend_neg =3D false; bool divisor_neg =3D false; @@ -1910,6 +1913,44 @@ wi::divmod_internal (HOST_WIDE_INT *quot =09} } =20 + unsigned HOST_HALF_WIDE_INT + b_quotient_buf[4 * WIDE_INT_MAX_INL_PRECISION +=09=09 / HOST_BITS_PER_HALF_WIDE_INT]; + unsigned HOST_HALF_WIDE_INT + b_remainder_buf[4 * WIDE_INT_MAX_INL_PRECISION +=09=09 / HOST_BITS_PER_HALF_WIDE_INT]; + unsigned HOST_HALF_WIDE_INT + b_dividend_buf[(4 * WIDE_INT_MAX_INL_PRECISION +=09=09 / HOST_BITS_PER_HALF_WIDE_INT) + 1]; + unsigned HOST_HALF_WIDE_INT + b_divisor_buf[4 * WIDE_INT_MAX_INL_PRECISION +=09=09 / HOST_BITS_PER_HALF_WIDE_INT]; + unsigned HOST_HALF_WIDE_INT *b_quotient =3D b_quotient_buf; + unsigned HOST_HALF_WIDE_INT *b_remainder =3D b_remainder_buf; + unsigned HOST_HALF_WIDE_INT *b_dividend =3D b_dividend_buf; + unsigned HOST_HALF_WIDE_INT *b_divisor =3D b_divisor_buf; + + if (dividend_prec > WIDE_INT_MAX_INL_PRECISION + && (sgn =3D=3D SIGNED || dividend_val[dividend_len - 1] >=3D 0)) + dividend_prec =3D (dividend_len + 1) * HOST_BITS_PER_WIDE_INT; + if (divisor_prec > WIDE_INT_MAX_INL_PRECISION) + divisor_prec =3D divisor_len * HOST_BITS_PER_WIDE_INT; + unsigned int dividend_blocks_needed =3D 2 * BLOCKS_NEEDED (dividend_prec= ); + unsigned int divisor_blocks_needed =3D 2 * BLOCKS_NEEDED (divisor_prec); + if (UNLIKELY (dividend_prec > WIDE_INT_MAX_INL_PRECISION) + || UNLIKELY (divisor_prec > WIDE_INT_MAX_INL_PRECISION)) + { + unsigned HOST_HALF_WIDE_INT *buf +=09=3D XALLOCAVEC (unsigned HOST_HALF_WIDE_INT, +=09=09 12 * dividend_blocks_needed +=09=09 + 4 * divisor_blocks_needed + 1); + b_quotient =3D buf; + b_remainder =3D b_quotient + 4 * dividend_blocks_needed; + b_dividend =3D b_remainder + 4 * dividend_blocks_needed; + b_divisor =3D b_dividend + 4 * dividend_blocks_needed + 1; + memset (b_quotient, 0, +=09 4 * dividend_blocks_needed * sizeof (HOST_HALF_WIDE_INT)); + } wi_unpack (b_dividend, dividend.get_val (), dividend.get_len (), =09 dividend_blocks_needed, dividend_prec, UNSIGNED); wi_unpack (b_divisor, divisor.get_val (), divisor.get_len (), @@ -1924,7 +1965,8 @@ wi::divmod_internal (HOST_WIDE_INT *quot while (n > 1 && b_divisor[n - 1] =3D=3D 0) n--; =20 - memset (b_quotient, 0, sizeof (b_quotient)); + if (b_quotient =3D=3D b_quotient_buf) + memset (b_quotient_buf, 0, sizeof (b_quotient_buf)); =20 divmod_internal_2 (b_quotient, b_remainder, b_dividend, b_divisor, m, n)= ; =20 @@ -1970,6 +2012,8 @@ wi::lshift_large (HOST_WIDE_INT *val, co =20 /* The whole-block shift fills with zeros. */ unsigned int len =3D BLOCKS_NEEDED (precision); + if (UNLIKELY (len > WIDE_INT_MAX_INL_ELTS)) + len =3D xlen + skip + 1; for (unsigned int i =3D 0; i < skip; ++i) val[i] =3D 0; =20 @@ -1993,22 +2037,17 @@ wi::lshift_large (HOST_WIDE_INT *val, co return canonize (val, len, precision); } =20 -/* Right shift XVAL by SHIFT and store the result in VAL. Return the +/* Right shift XVAL by SHIFT and store the result in VAL. LEN is the number of blocks in VAL. The input has XPRECISION bits and the output has XPRECISION - SHIFT bits. */ -static unsigned int +static void rshift_large_common (HOST_WIDE_INT *val, const HOST_WIDE_INT *xval, -=09=09 unsigned int xlen, unsigned int xprecision, -=09=09 unsigned int shift) +=09=09 unsigned int xlen, unsigned int shift, unsigned int len) { /* Split the shift into a whole-block shift and a subblock shift. */ unsigned int skip =3D shift / HOST_BITS_PER_WIDE_INT; unsigned int small_shift =3D shift % HOST_BITS_PER_WIDE_INT; =20 - /* Work out how many blocks are needed to store the significant bits - (excluding the upper zeros or signs). */ - unsigned int len =3D BLOCKS_NEEDED (xprecision - shift); - /* It's easier to handle the simple block case specially. */ if (small_shift =3D=3D 0) for (unsigned int i =3D 0; i < len; ++i) @@ -2025,7 +2064,6 @@ rshift_large_common (HOST_WIDE_INT *val, =09 val[i] |=3D curr << (-small_shift % HOST_BITS_PER_WIDE_INT); =09} } - return len; } =20 /* Logically right shift XVAL by SHIFT and store the result in VAL. @@ -2036,11 +2074,20 @@ wi::lrshift_large (HOST_WIDE_INT *val, c =09=09 unsigned int xlen, unsigned int xprecision, =09=09 unsigned int precision, unsigned int shift) { - unsigned int len =3D rshift_large_common (val, xval, xlen, xprecision, s= hift); + /* Work out how many blocks are needed to store the significant bits + (excluding the upper zeros or signs). */ + unsigned int blocks_needed =3D BLOCKS_NEEDED (xprecision - shift); + unsigned int len =3D blocks_needed; + if (UNLIKELY (len > WIDE_INT_MAX_INL_ELTS) + && len > xlen + && xval[xlen - 1] >=3D 0) + len =3D xlen; + + rshift_large_common (val, xval, xlen, shift, len); =20 /* The value we just created has precision XPRECISION - SHIFT. Zero-extend it to wider precisions. */ - if (precision > xprecision - shift) + if (precision > xprecision - shift && len =3D=3D blocks_needed) { unsigned int small_prec =3D (xprecision - shift) % HOST_BITS_PER_WID= E_INT; if (small_prec) @@ -2063,11 +2110,18 @@ wi::arshift_large (HOST_WIDE_INT *val, c =09=09 unsigned int xlen, unsigned int xprecision, =09=09 unsigned int precision, unsigned int shift) { - unsigned int len =3D rshift_large_common (val, xval, xlen, xprecision, s= hift); + /* Work out how many blocks are needed to store the significant bits + (excluding the upper zeros or signs). */ + unsigned int blocks_needed =3D BLOCKS_NEEDED (xprecision - shift); + unsigned int len =3D blocks_needed; + if (UNLIKELY (len > WIDE_INT_MAX_INL_ELTS) && len > xlen) + len =3D xlen; + + rshift_large_common (val, xval, xlen, shift, len); =20 /* The value we just created has precision XPRECISION - SHIFT. Sign-extend it to wider types. */ - if (precision > xprecision - shift) + if (precision > xprecision - shift && len =3D=3D blocks_needed) { unsigned int small_prec =3D (xprecision - shift) % HOST_BITS_PER_WID= E_INT; if (small_prec) @@ -2399,9 +2453,12 @@ from_int (int i) static void assert_deceq (const char *expected, const wide_int_ref &wi, signop sgn) { - char buf[WIDE_INT_PRINT_BUFFER_SIZE]; - print_dec (wi, buf, sgn); - ASSERT_STREQ (expected, buf); + char buf[WIDE_INT_PRINT_BUFFER_SIZE], *p =3D buf; + unsigned len =3D wi.get_len (); + if (UNLIKELY (len > WIDE_INT_MAX_INL_ELTS)) + p =3D XALLOCAVEC (char, len * HOST_BITS_PER_WIDE_INT / 4 + 4); + print_dec (wi, p, sgn); + ASSERT_STREQ (expected, p); } =20 /* Likewise for base 16. */ @@ -2409,9 +2466,12 @@ assert_deceq (const char *expected, cons static void assert_hexeq (const char *expected, const wide_int_ref &wi) { - char buf[WIDE_INT_PRINT_BUFFER_SIZE]; - print_hex (wi, buf); - ASSERT_STREQ (expected, buf); + char buf[WIDE_INT_PRINT_BUFFER_SIZE], *p =3D buf; + unsigned len =3D wi.get_len (); + if (UNLIKELY (len > WIDE_INT_MAX_INL_ELTS)) + p =3D XALLOCAVEC (char, len * HOST_BITS_PER_WIDE_INT / 4 + 4); + print_hex (wi, p); + ASSERT_STREQ (expected, p); } =20 /* Test cases. */ @@ -2428,7 +2488,7 @@ test_printing () assert_hexeq ("0x1fffffffffffffffff", wi::shwi (-1, 69)); assert_hexeq ("0xffffffffffffffff", wi::mask (64, false, 69)); assert_hexeq ("0xffffffffffffffff", wi::mask (64, false)); - if (WIDE_INT_MAX_PRECISION > 128) + if (WIDE_INT_MAX_INL_PRECISION > 128) { assert_hexeq ("0x20000000000000000fffffffffffffffe", =09=09 wi::lshift (1, 129) + wi::lshift (1, 64) - 2); --- gcc/wide-int-print.h.jj=092023-10-08 16:37:32.071269534 +0200 +++ gcc/wide-int-print.h=092023-10-08 17:00:12.098716826 +0200 @@ -22,7 +22,7 @@ along with GCC; see the file COPYING3. =20 #include =20 -#define WIDE_INT_PRINT_BUFFER_SIZE (WIDE_INT_MAX_PRECISION / 4 + 4) +#define WIDE_INT_PRINT_BUFFER_SIZE (WIDE_INT_MAX_INL_PRECISION / 4 + 4) =20 /* Printing functions. */ =20 --- gcc/wide-int-print.cc.jj=092023-10-08 16:37:32.052269785 +0200 +++ gcc/wide-int-print.cc=092023-10-08 17:00:12.082717047 +0200 @@ -74,9 +74,12 @@ print_decs (const wide_int_ref &wi, char void print_decs (const wide_int_ref &wi, FILE *file) { - char buf[WIDE_INT_PRINT_BUFFER_SIZE]; - print_decs (wi, buf); - fputs (buf, file); + char buf[WIDE_INT_PRINT_BUFFER_SIZE], *p =3D buf; + unsigned len =3D wi.get_len (); + if (UNLIKELY (len > WIDE_INT_MAX_INL_ELTS)) + p =3D XALLOCAVEC (char, len * HOST_BITS_PER_WIDE_INT / 4 + 4); + print_decs (wi, p); + fputs (p, file); } =20 /* Try to print the unsigned self in decimal to BUF if the number fits @@ -98,9 +101,12 @@ print_decu (const wide_int_ref &wi, char void print_decu (const wide_int_ref &wi, FILE *file) { - char buf[WIDE_INT_PRINT_BUFFER_SIZE]; - print_decu (wi, buf); - fputs (buf, file); + char buf[WIDE_INT_PRINT_BUFFER_SIZE], *p =3D buf; + unsigned len =3D wi.get_len (); + if (UNLIKELY (len > WIDE_INT_MAX_INL_ELTS)) + p =3D XALLOCAVEC (char, len * HOST_BITS_PER_WIDE_INT / 4 + 4); + print_decu (wi, p); + fputs (p, file); } =20 void @@ -134,9 +140,12 @@ print_hex (const wide_int_ref &val, char void print_hex (const wide_int_ref &wi, FILE *file) { - char buf[WIDE_INT_PRINT_BUFFER_SIZE]; - print_hex (wi, buf); - fputs (buf, file); + char buf[WIDE_INT_PRINT_BUFFER_SIZE], *p =3D buf; + unsigned len =3D wi.get_len (); + if (UNLIKELY (len > WIDE_INT_MAX_INL_ELTS)) + p =3D XALLOCAVEC (char, len * HOST_BITS_PER_WIDE_INT / 4 + 4); + print_hex (wi, p); + fputs (p, file); } =20 /* Print larger precision wide_int. Not defined as inline in a header --- gcc/tree.h.jj=092023-10-08 16:37:32.003270432 +0200 +++ gcc/tree.h=092023-10-08 17:00:12.086716992 +0200 @@ -6258,13 +6258,17 @@ namespace wi template struct int_traits > { - static const enum precision_type precision_type =3D CONST_PRECISION; + static const enum precision_type precision_type + =3D N =3D=3D ADDR_MAX_PRECISION ? CONST_PRECISION : WIDEST_CONST_PRE= CISION; static const bool host_dependent_precision =3D false; static const bool is_sign_extended =3D true; static const unsigned int precision =3D N; + static const unsigned int inl_precision + =3D N =3D=3D ADDR_MAX_PRECISION ? 0 +=09 : N / WIDEST_INT_MAX_PRECISION * WIDE_INT_MAX_INL_PRECISION; }; =20 - typedef extended_tree widest_extended_tree; + typedef extended_tree widest_extended_tree; typedef extended_tree offset_extended_tree; =20 typedef const generic_wide_int tree_to_widest_ref= ; @@ -6292,7 +6296,8 @@ namespace wi tree_to_poly_wide_ref to_poly_wide (const_tree); =20 template - struct ints_for >, CONST_PRECISION> + struct ints_for >, +=09=09 int_traits >::precision_type> { typedef generic_wide_int > extended; static extended zero (const extended &); @@ -6308,7 +6313,7 @@ namespace wi =20 /* Used to convert a tree to a widest2_int like this: widest2_int foo =3D widest2_int_cst (some_tree). */ -typedef generic_wide_int > +typedef generic_wide_int = > widest2_int_cst; =20 /* Refer to INTEGER_CST T as though it were a widest_int. @@ -6444,7 +6449,7 @@ wi::extended_tree ::get_len () const { if (N =3D=3D ADDR_MAX_PRECISION) return TREE_INT_CST_OFFSET_NUNITS (m_t); - else if (N >=3D WIDE_INT_MAX_PRECISION) + else if (N >=3D WIDEST_INT_MAX_PRECISION) return TREE_INT_CST_EXT_NUNITS (m_t); else /* This class is designed to be used for specific output precisions @@ -6530,7 +6535,8 @@ wi::to_poly_wide (const_tree t) template inline generic_wide_int > wi::ints_for >, -=09 wi::CONST_PRECISION>::zero (const extended &x) +=09 wi::int_traits >::precision_type +=09 >::zero (const extended &x) { return build_zero_cst (TREE_TYPE (x.get_tree ())); } --- gcc/tree.cc.jj=092023-10-08 16:37:32.000270471 +0200 +++ gcc/tree.cc=092023-10-08 17:00:12.091716923 +0200 @@ -2676,13 +2676,13 @@ build_zero_cst (tree type) tree build_replicated_int_cst (tree type, unsigned int width, HOST_WIDE_INT val= ue) { - int n =3D (TYPE_PRECISION (type) + HOST_BITS_PER_WIDE_INT - 1) - / HOST_BITS_PER_WIDE_INT; + int n =3D ((TYPE_PRECISION (type) + HOST_BITS_PER_WIDE_INT - 1) +=09 / HOST_BITS_PER_WIDE_INT); unsigned HOST_WIDE_INT low, mask; - HOST_WIDE_INT a[WIDE_INT_MAX_ELTS]; + HOST_WIDE_INT a[WIDE_INT_MAX_INL_ELTS]; int i; =20 - gcc_assert (n && n <=3D WIDE_INT_MAX_ELTS); + gcc_assert (n && n <=3D WIDE_INT_MAX_INL_ELTS); =20 if (width =3D=3D HOST_BITS_PER_WIDE_INT) low =3D value; @@ -2696,8 +2696,8 @@ build_replicated_int_cst (tree type, uns a[i] =3D low; =20 gcc_assert (TYPE_PRECISION (type) <=3D MAX_BITSIZE_MODE_ANY_INT); - return wide_int_to_tree - (type, wide_int::from_array (a, n, TYPE_PRECISION (type))); + return wide_int_to_tree (type, wide_int::from_array (a, n, +=09=09=09=09=09=09 TYPE_PRECISION (type))); } =20 /* If floating-point type TYPE has an IEEE-style sign bit, return an --- gcc/print-tree.cc.jj=092023-10-08 16:37:31.730274037 +0200 +++ gcc/print-tree.cc=092023-10-08 17:00:12.098716826 +0200 @@ -365,13 +365,13 @@ print_node (FILE *file, const char *pref fputs (code =3D=3D CALL_EXPR ? " must-tail-call" : " static", file); if (TREE_DEPRECATED (node)) fputs (" deprecated", file); - if (TREE_UNAVAILABLE (node)) - fputs (" unavailable", file); if (TREE_VISITED (node)) fputs (" visited", file); =20 if (code !=3D TREE_VEC && code !=3D INTEGER_CST && code !=3D SSA_NAME) { + if (TREE_UNAVAILABLE (node)) +=09fputs (" unavailable", file); if (TREE_LANG_FLAG_0 (node)) =09fputs (" tree_0", file); if (TREE_LANG_FLAG_1 (node)) --- gcc/poly-int.h.jj=092023-10-08 16:37:31.703274393 +0200 +++ gcc/poly-int.h=092023-10-08 17:00:12.089716950 +0200 @@ -109,6 +109,21 @@ struct poly_coeff_traits +struct poly_coeff_traits +{ + typedef WI_UNARY_RESULT (T) result; + typedef int int_type; + /* These types are always signed. */ + static const int signedness =3D 1; + static const int precision =3D wi::int_traits::precision; + static const int inl_precision =3D wi::int_traits::inl_precision; + static const int rank =3D precision * 2 / CHAR_BIT; + + template + struct init_cast { using type =3D const Arg &; }; +}; + /* Information about a pair of coefficient types. */ template struct poly_coeff_pair_traits --- gcc/cfgloop.h.jj=092023-10-08 16:37:31.394278474 +0200 +++ gcc/cfgloop.h=092023-10-08 17:00:12.085717005 +0200 @@ -44,6 +44,9 @@ enum iv_extend_code IV_UNKNOWN_EXTEND }; =20 +typedef generic_wide_int > + bound_wide_int; + /* The structure describing a bound on number of iterations of a loop. */ =20 class GTY ((chain_next ("%h.next"))) nb_iter_bound { @@ -58,7 +61,7 @@ public: overflows (as MAX + 1 is sometimes produced as the estimate on num= ber =09of executions of STMT). b) it is consistent with the result of number_of_iterations_exit. */ - widest_int bound; + bound_wide_int bound; =20 /* True if, after executing the statement BOUND + 1 times, we will leave the loop; that is, all the statements after it are executed at = most @@ -161,14 +164,14 @@ public: =20 /* An integer guaranteed to be greater or equal to nb_iterations. Only valid if any_upper_bound is true. */ - widest_int nb_iterations_upper_bound; + bound_wide_int nb_iterations_upper_bound; =20 - widest_int nb_iterations_likely_upper_bound; + bound_wide_int nb_iterations_likely_upper_bound; =20 /* An integer giving an estimate on nb_iterations. Unlike nb_iterations_upper_bound, there is no guarantee that it is at least nb_iterations. */ - widest_int nb_iterations_estimate; + bound_wide_int nb_iterations_estimate; =20 /* If > 0, an integer, where the user asserted that for any I in [ 0, nb_iterations ) and for any J in --- gcc/cfgloop.cc.jj=092023-10-08 16:37:31.375278725 +0200 +++ gcc/cfgloop.cc=092023-10-08 17:00:12.087716978 +0200 @@ -1895,33 +1895,38 @@ void record_niter_bound (class loop *loop, const widest_int &i_bound, =09=09 bool realistic, bool upper) { + if (wi::min_precision (i_bound, SIGNED) > bound_wide_int ().get_precisio= n ()) + return; + + bound_wide_int bound =3D bound_wide_int::from (i_bound, SIGNED); + /* Update the bounds only when there is no previous estimation, or when = the current estimation is smaller. */ if (upper && (!loop->any_upper_bound -=09 || wi::ltu_p (i_bound, loop->nb_iterations_upper_bound))) +=09 || wi::ltu_p (bound, loop->nb_iterations_upper_bound))) { loop->any_upper_bound =3D true; - loop->nb_iterations_upper_bound =3D i_bound; + loop->nb_iterations_upper_bound =3D bound; if (!loop->any_likely_upper_bound) =09{ =09 loop->any_likely_upper_bound =3D true; -=09 loop->nb_iterations_likely_upper_bound =3D i_bound; +=09 loop->nb_iterations_likely_upper_bound =3D bound; =09} } if (realistic && (!loop->any_estimate -=09 || wi::ltu_p (i_bound, loop->nb_iterations_estimate))) +=09 || wi::ltu_p (bound, loop->nb_iterations_estimate))) { loop->any_estimate =3D true; - loop->nb_iterations_estimate =3D i_bound; + loop->nb_iterations_estimate =3D bound; } if (!realistic && (!loop->any_likely_upper_bound - || wi::ltu_p (i_bound, loop->nb_iterations_likely_upper_bound))) +=09 || wi::ltu_p (bound, loop->nb_iterations_likely_upper_bound))) { loop->any_likely_upper_bound =3D true; - loop->nb_iterations_likely_upper_bound =3D i_bound; + loop->nb_iterations_likely_upper_bound =3D bound; } =20 /* If an upper bound is smaller than the realistic estimate of the @@ -2018,7 +2023,7 @@ get_estimated_loop_iterations (class loo return false; } =20 - *nit =3D loop->nb_iterations_estimate; + *nit =3D widest_int::from (loop->nb_iterations_estimate, SIGNED); return true; } =20 @@ -2032,7 +2037,7 @@ get_max_loop_iterations (const class loo if (!loop->any_upper_bound) return false; =20 - *nit =3D loop->nb_iterations_upper_bound; + *nit =3D widest_int::from (loop->nb_iterations_upper_bound, SIGNED); return true; } =20 @@ -2066,7 +2071,7 @@ get_likely_max_loop_iterations (class lo if (!loop->any_likely_upper_bound) return false; =20 - *nit =3D loop->nb_iterations_likely_upper_bound; + *nit =3D widest_int::from (loop->nb_iterations_likely_upper_bound, SIGNE= D); return true; } =20 --- gcc/tree-vect-loop.cc.jj=092023-10-08 16:37:31.993270564 +0200 +++ gcc/tree-vect-loop.cc=092023-10-08 17:00:12.081717060 +0200 @@ -11681,7 +11681,7 @@ vect_transform_loop (loop_vec_info loop_ =09=09=09=09=09LOOP_VINFO_VECT_FACTOR (loop_vinfo), =09=09=09=09=09&bound)) =09 loop->nb_iterations_upper_bound -=09 =3D wi::umin ((widest_int) (bound - 1), +=09 =3D wi::umin ((bound_wide_int) (bound - 1), =09=09=09 loop->nb_iterations_upper_bound); } } --- gcc/tree-ssa-loop-niter.cc.jj=092023-10-08 16:37:31.968270894 +0200 +++ gcc/tree-ssa-loop-niter.cc=092023-10-08 17:00:12.085717005 +0200 @@ -3873,12 +3873,17 @@ do_warn_aggressive_loop_optimizations (c return; =20 gimple *estmt =3D last_nondebug_stmt (e->src); - char buf[WIDE_INT_PRINT_BUFFER_SIZE]; - print_dec (i_bound, buf, TYPE_UNSIGNED (TREE_TYPE (loop->nb_iterations)) + char buf[WIDE_INT_PRINT_BUFFER_SIZE], *p; + unsigned len =3D i_bound.get_len (); + if (len > WIDE_INT_MAX_INL_ELTS) + p =3D XALLOCAVEC (char, len * HOST_BITS_PER_WIDE_INT / 4 + 4); + else + p =3D buf; + print_dec (i_bound, p, TYPE_UNSIGNED (TREE_TYPE (loop->nb_iterations)) =09 ? UNSIGNED : SIGNED); auto_diagnostic_group d; if (warning_at (gimple_location (stmt), OPT_Waggressive_loop_optimizatio= ns, -=09=09 "iteration %s invokes undefined behavior", buf)) +=09=09 "iteration %s invokes undefined behavior", p)) inform (gimple_location (estmt), "within this loop"); loop->warned_aggressive_loop_optimizations =3D true; } @@ -3915,6 +3920,9 @@ record_estimate (class loop *loop, tree else gcc_checking_assert (i_bound =3D=3D wi::to_widest (bound)); =20 + if (wi::min_precision (i_bound, SIGNED) > bound_wide_int ().get_precisio= n ()) + return; + /* If we have a guaranteed upper bound, record it in the appropriate list, unless this is an !is_exit bound (i.e. undefined behavior in at_stmt) in a loop with known constant number of iterations. */ @@ -3925,7 +3933,7 @@ record_estimate (class loop *loop, tree { class nb_iter_bound *elt =3D ggc_alloc (); =20 - elt->bound =3D i_bound; + elt->bound =3D bound_wide_int::from (i_bound, SIGNED); elt->stmt =3D at_stmt; elt->is_exit =3D is_exit; elt->next =3D loop->bounds; @@ -4410,8 +4418,8 @@ infer_loop_bounds_from_undefined (class static int wide_int_cmp (const void *p1, const void *p2) { - const widest_int *d1 =3D (const widest_int *) p1; - const widest_int *d2 =3D (const widest_int *) p2; + const bound_wide_int *d1 =3D (const bound_wide_int *) p1; + const bound_wide_int *d2 =3D (const bound_wide_int *) p2; return wi::cmpu (*d1, *d2); } =20 @@ -4419,7 +4427,7 @@ wide_int_cmp (const void *p1, const void Lookup by binary search. */ =20 static int -bound_index (const vec &bounds, const widest_int &bound) +bound_index (const vec &bounds, const bound_wide_int &boun= d) { unsigned int end =3D bounds.length (); unsigned int begin =3D 0; @@ -4428,7 +4436,7 @@ bound_index (const vec &boun while (begin !=3D end) { unsigned int middle =3D (begin + end) / 2; - widest_int index =3D bounds[middle]; + bound_wide_int index =3D bounds[middle]; =20 if (index =3D=3D bound) =09return middle; @@ -4450,7 +4458,7 @@ static void discover_iteration_bound_by_body_walk (class loop *loop) { class nb_iter_bound *elt; - auto_vec bounds; + auto_vec bounds; vec > queues =3D vNULL; vec queue =3D vNULL; ptrdiff_t queue_index; @@ -4459,7 +4467,7 @@ discover_iteration_bound_by_body_walk (c /* Discover what bounds may interest us. */ for (elt =3D loop->bounds; elt; elt =3D elt->next) { - widest_int bound =3D elt->bound; + bound_wide_int bound =3D elt->bound; =20 /* Exit terminates loop at given iteration, while non-exits produce = undefined =09 effect on the next iteration. */ @@ -4492,7 +4500,7 @@ discover_iteration_bound_by_body_walk (c hash_map bb_bounds; for (elt =3D loop->bounds; elt; elt =3D elt->next) { - widest_int bound =3D elt->bound; + bound_wide_int bound =3D elt->bound; if (!elt->is_exit) =09{ =09 bound +=3D 1; @@ -4601,7 +4609,8 @@ discover_iteration_bound_by_body_walk (c =09 print_decu (bounds[latch_index], dump_file); =09 fprintf (dump_file, "\n"); =09} - record_niter_bound (loop, bounds[latch_index], false, true); + record_niter_bound (loop, widest_int::from (bounds[latch_index], +=09=09=09=09=09=09 SIGNED), false, true); } =20 queues.release (); @@ -4704,7 +4713,8 @@ maybe_lower_iteration_bound (class loop if (dump_file && (dump_flags & TDF_DETAILS)) =09fprintf (dump_file, "Reducing loop iteration estimate by 1; " =09=09 "undefined statement must be executed at the last iteration.\n"); - record_niter_bound (loop, loop->nb_iterations_upper_bound - 1, + record_niter_bound (loop, widest_int::from (loop->nb_iterations_uppe= r_bound, +=09=09=09=09=09=09 SIGNED) - 1, =09=09=09 false, true); } =20 @@ -4860,10 +4870,13 @@ estimate_numbers_of_iterations (class lo not break code with undefined behavior by not recording smaller maximum number of iterations. */ if (loop->nb_iterations - && TREE_CODE (loop->nb_iterations) =3D=3D INTEGER_CST) + && TREE_CODE (loop->nb_iterations) =3D=3D INTEGER_CST + && (wi::min_precision (wi::to_widest (loop->nb_iterations), SIGNED) +=09 <=3D bound_wide_int ().get_precision ())) { loop->any_upper_bound =3D true; - loop->nb_iterations_upper_bound =3D wi::to_widest (loop->nb_iteratio= ns); + loop->nb_iterations_upper_bound +=09=3D bound_wide_int::from (wi::to_widest (loop->nb_iterations), SIGNED); } } =20 @@ -5114,7 +5127,7 @@ n_of_executions_at_most (gimple *stmt, =09=09=09 class nb_iter_bound *niter_bound, =09=09=09 tree niter) { - widest_int bound =3D niter_bound->bound; + widest_int bound =3D widest_int::from (niter_bound->bound, SIGNED); tree nit_type =3D TREE_TYPE (niter), e; enum tree_code cmp; =20 --- gcc/tree-ssa-loop-ivcanon.cc.jj=092023-10-08 16:37:31.904271739 +0200 +++ gcc/tree-ssa-loop-ivcanon.cc=092023-10-08 17:00:12.098716826 +0200 @@ -622,10 +622,11 @@ remove_redundant_iv_tests (class loop *l =09 || !integer_zerop (niter.may_be_zero) =09 || !niter.niter =09 || TREE_CODE (niter.niter) !=3D INTEGER_CST -=09 || !wi::ltu_p (loop->nb_iterations_upper_bound, +=09 || !wi::ltu_p (widest_int::from (loop->nb_iterations_upper_bound, +=09=09=09=09=09 SIGNED), =09=09=09 wi::to_widest (niter.niter))) =09 continue; -=09 =20 + =09 if (dump_file && (dump_flags & TDF_DETAILS)) =09 { =09 fprintf (dump_file, "Removed pointless exit: "); --- gcc/match.pd.jj=092023-10-08 16:37:52.850995158 +0200 +++ gcc/match.pd=092023-10-08 17:02:04.986158055 +0200 @@ -6436,8 +6436,12 @@ (define_operator_list SYNC_FETCH_AND_AND code and here to avoid a spurious overflow flag on the resulting constant which fold_convert produces. */ (if (TREE_CODE (@1) =3D=3D INTEGER_CST) - (cmp @00 { force_fit_type (TREE_TYPE (@00), wi::to_widest (@1), 0, -=09=09=09=09TREE_OVERFLOW (@1)); }) + (cmp @00 { force_fit_type (TREE_TYPE (@00), +=09=09=09=09wide_int::from (wi::to_wide (@1), +=09=09=09=09=09=09MAX (TYPE_PRECISION (TREE_TYPE (@1)), +=09=09=09=09=09=09 TYPE_PRECISION (TREE_TYPE (@00))), +=09=09=09=09=09=09TYPE_SIGN (TREE_TYPE (@1))), +=09=09=09=090, TREE_OVERFLOW (@1)); }) (cmp @00 (convert @1))) =20 (if (TYPE_PRECISION (TREE_TYPE (@0)) > TYPE_PRECISION (TREE_TYPE (@00)= )) --- gcc/value-range.h.jj=092023-10-08 16:37:32.042269917 +0200 +++ gcc/value-range.h=092023-10-08 17:00:12.082717047 +0200 @@ -626,7 +626,9 @@ irange::maybe_resize (int needed) { m_max_ranges =3D HARD_MAX_RANGES; wide_int *newmem =3D new wide_int[m_max_ranges * 2]; - memcpy (newmem, m_base, sizeof (wide_int) * num_pairs () * 2); + unsigned n =3D num_pairs () * 2; + for (unsigned i =3D 0; i < n; ++i) +=09newmem[i] =3D m_base[i]; m_base =3D newmem; } } --- gcc/value-range.cc.jj=092023-10-08 16:37:32.022270181 +0200 +++ gcc/value-range.cc=092023-10-08 17:00:12.084717019 +0200 @@ -245,17 +245,24 @@ vrange::dump (FILE *file) const void irange_bitmask::dump (FILE *file) const { - char buf[WIDE_INT_PRINT_BUFFER_SIZE]; + char buf[WIDE_INT_PRINT_BUFFER_SIZE], *p; pretty_printer buffer; =20 pp_needs_newline (&buffer) =3D true; buffer.buffer->stream =3D file; pp_string (&buffer, "MASK "); - print_hex (m_mask, buf); - pp_string (&buffer, buf); + unsigned len_mask =3D m_mask.get_len (); + unsigned len_val =3D m_value.get_len (); + unsigned len =3D MAX (len_mask, len_val); + if (len > WIDE_INT_MAX_INL_ELTS) + p =3D XALLOCAVEC (char, len * HOST_BITS_PER_WIDE_INT / 4 + 4); + else + p =3D buf; + print_hex (m_mask, p); + pp_string (&buffer, p); pp_string (&buffer, " VALUE "); - print_hex (m_value, buf); - pp_string (&buffer, buf); + print_hex (m_value, p); + pp_string (&buffer, p); pp_flush (&buffer); } =20 --- gcc/fold-const.cc.jj=092023-10-08 16:37:31.488277232 +0200 +++ gcc/fold-const.cc=092023-10-08 17:02:04.989158013 +0200 @@ -2137,7 +2137,10 @@ fold_convert_const_int_from_int (tree ty /* Given an integer constant, make new constant with new type, appropriately sign-extended or truncated. Use widest_int so that any extension is done according ARG1's type. */ - return force_fit_type (type, wi::to_widest (arg1), + tree arg1_type =3D TREE_TYPE (arg1); + unsigned prec =3D MAX (TYPE_PRECISION (arg1_type), TYPE_PRECISION (type)= ); + return force_fit_type (type, wide_int::from (wi::to_wide (arg1), prec, +=09=09=09=09=09 TYPE_SIGN (arg1_type)), =09=09=09 !POINTER_TYPE_P (TREE_TYPE (arg1)), =09=09=09 TREE_OVERFLOW (arg1)); } @@ -9565,8 +9568,13 @@ fold_unary_loc (location_t loc, enum tre =09 } =09 if (change) =09 { -=09 tem =3D force_fit_type (type, wi::to_widest (and1), 0, -=09=09=09=09 TREE_OVERFLOW (and1)); +=09 tree and1_type =3D TREE_TYPE (and1); +=09 unsigned prec =3D MAX (TYPE_PRECISION (and1_type), +=09=09=09=09 TYPE_PRECISION (type)); +=09 tem =3D force_fit_type (type, +=09=09=09=09 wide_int::from (wi::to_wide (and1), prec, +=09=09=09=09=09=09 TYPE_SIGN (and1_type)), +=09=09=09=09 0, TREE_OVERFLOW (and1)); =09 return fold_build2_loc (loc, BIT_AND_EXPR, type, =09=09=09=09 fold_convert_loc (loc, type, and0), tem); =09 } --- gcc/tree-ssa-ccp.cc.jj=092023-10-08 16:37:31.865272254 +0200 +++ gcc/tree-ssa-ccp.cc=092023-10-08 17:02:04.984158082 +0200 @@ -1966,7 +1966,8 @@ bit_value_binop (enum tree_code code, si =09=09 } =09=09else =09=09 { -=09=09 widest_int upper =3D wi::udiv_trunc (r1max, r2min); +=09=09 widest_int upper +=09=09 =3D wi::udiv_trunc (wi::zext (r1max, width), r2min); =09=09 unsigned int lzcount =3D wi::clz (upper); =09=09 unsigned int bits =3D wi::get_precision (upper) - lzcount; =09=09 *mask =3D wi::mask (bits, false); --- gcc/dwarf2out.h.jj=092023-10-08 16:37:31.479277352 +0200 +++ gcc/dwarf2out.h=092023-10-08 17:00:12.098716826 +0200 @@ -30,7 +30,7 @@ typedef struct dw_cfi_node *dw_cfi_ref; typedef struct dw_loc_descr_node *dw_loc_descr_ref; typedef struct dw_loc_list_struct *dw_loc_list_ref; typedef struct dw_discr_list_node *dw_discr_list_ref; -typedef wide_int *wide_int_ptr; +typedef rwide_int *rwide_int_ptr; =20 =20 /* Call frames are described using a sequence of Call Frame @@ -252,7 +252,7 @@ struct GTY(()) dw_val_node { unsigned HOST_WIDE_INT =09GTY ((tag ("dw_val_class_unsigned_const"))) val_unsigned; double_int GTY ((tag ("dw_val_class_const_double"))) val_double; - wide_int_ptr GTY ((tag ("dw_val_class_wide_int"))) val_wide; + rwide_int_ptr GTY ((tag ("dw_val_class_wide_int"))) val_wide; dw_vec_const GTY ((tag ("dw_val_class_vec"))) val_vec; struct dw_val_die_union =09{ --- gcc/dwarf2out.cc.jj=092023-10-08 16:37:31.448277761 +0200 +++ gcc/dwarf2out.cc=092023-10-08 17:00:12.095716867 +0200 @@ -397,7 +397,7 @@ dump_struct_debug (tree type, enum debug of the number. */ =20 static unsigned int -get_full_len (const wide_int &op) +get_full_len (const rwide_int &op) { int prec =3D wi::get_precision (op); return ((prec + HOST_BITS_PER_WIDE_INT - 1) @@ -3900,7 +3900,7 @@ static void add_data_member_location_att =09=09=09=09=09=09struct vlr_context *); static bool add_const_value_attribute (dw_die_ref, machine_mode, rtx); static void insert_int (HOST_WIDE_INT, unsigned, unsigned char *); -static void insert_wide_int (const wide_int &, unsigned char *, int); +static void insert_wide_int (const rwide_int &, unsigned char *, int); static unsigned insert_float (const_rtx, unsigned char *); static rtx rtl_for_decl_location (tree); static bool add_location_or_const_value_attribute (dw_die_ref, tree, bool)= ; @@ -4598,14 +4598,14 @@ AT_unsigned (dw_attr_node *a) =20 static inline void add_AT_wide (dw_die_ref die, enum dwarf_attribute attr_kind, -=09 const wide_int& w) +=09 const rwide_int& w) { dw_attr_node attr; =20 attr.dw_attr =3D attr_kind; attr.dw_attr_val.val_class =3D dw_val_class_wide_int; attr.dw_attr_val.val_entry =3D NULL; - attr.dw_attr_val.v.val_wide =3D ggc_alloc (); + attr.dw_attr_val.v.val_wide =3D ggc_alloc (); *attr.dw_attr_val.v.val_wide =3D w; add_dwarf_attr (die, &attr); } @@ -16714,7 +16714,7 @@ mem_loc_descriptor (rtx rtl, machine_mod =09 mem_loc_result->dw_loc_oprnd1.v.val_die_ref.external =3D 0; =09 mem_loc_result->dw_loc_oprnd2.val_class =09 =3D dw_val_class_wide_int; -=09 mem_loc_result->dw_loc_oprnd2.v.val_wide =3D ggc_alloc (); +=09 mem_loc_result->dw_loc_oprnd2.v.val_wide =3D ggc_alloc (); =09 *mem_loc_result->dw_loc_oprnd2.v.val_wide =3D rtx_mode_t (rtl, mode); =09} break; @@ -17288,7 +17288,7 @@ loc_descriptor (rtx rtl, machine_mode mo =09 loc_result =3D new_loc_descr (DW_OP_implicit_value, =09=09=09=09 GET_MODE_SIZE (int_mode), 0); =09 loc_result->dw_loc_oprnd2.val_class =3D dw_val_class_wide_int; -=09 loc_result->dw_loc_oprnd2.v.val_wide =3D ggc_alloc (); +=09 loc_result->dw_loc_oprnd2.v.val_wide =3D ggc_alloc (); =09 *loc_result->dw_loc_oprnd2.v.val_wide =3D rtx_mode_t (rtl, int_mode); =09} break; @@ -20189,7 +20189,7 @@ extract_int (const unsigned char *src, u /* Writes wide_int values to dw_vec_const array. */ =20 static void -insert_wide_int (const wide_int &val, unsigned char *dest, int elt_size) +insert_wide_int (const rwide_int &val, unsigned char *dest, int elt_size) { int i; =20 @@ -20274,7 +20274,7 @@ add_const_value_attribute (dw_die_ref di =09 && (GET_MODE_PRECISION (int_mode) =09 & (HOST_BITS_PER_WIDE_INT - 1)) =3D=3D 0) =09{ -=09 wide_int w =3D rtx_mode_t (rtl, int_mode); +=09 rwide_int w =3D rtx_mode_t (rtl, int_mode); =09 add_AT_wide (die, DW_AT_const_value, w); =09 return true; =09} --- gcc/lto-streamer-out.cc.jj=092023-10-08 16:37:31.684274644 +0200 +++ gcc/lto-streamer-out.cc=092023-10-08 17:00:12.082717047 +0200 @@ -2173,13 +2173,26 @@ output_cfg (struct output_block *ob, str =09=09=09 loop_estimation, EST_LAST, loop->estimate_state); streamer_write_hwi (ob, loop->any_upper_bound); if (loop->any_upper_bound) -=09streamer_write_widest_int (ob, loop->nb_iterations_upper_bound); +=09{ +=09 widest_int w =3D widest_int::from (loop->nb_iterations_upper_bound, +=09=09=09=09=09 SIGNED); +=09 streamer_write_widest_int (ob, w); +=09} streamer_write_hwi (ob, loop->any_likely_upper_bound); if (loop->any_likely_upper_bound) -=09streamer_write_widest_int (ob, loop->nb_iterations_likely_upper_bound); +=09{ +=09 widest_int w +=09 =3D widest_int::from (loop->nb_iterations_likely_upper_bound, +=09=09=09=09SIGNED); +=09 streamer_write_widest_int (ob, w); +=09} streamer_write_hwi (ob, loop->any_estimate); if (loop->any_estimate) -=09streamer_write_widest_int (ob, loop->nb_iterations_estimate); +=09{ +=09 widest_int w =3D widest_int::from (loop->nb_iterations_estimate, +=09=09=09=09=09 SIGNED); +=09 streamer_write_widest_int (ob, w); +=09} =20 /* Write OMP SIMD related info. */ streamer_write_hwi (ob, loop->safelen); --- gcc/lto-streamer-in.cc.jj=092023-10-08 16:37:31.662274935 +0200 +++ gcc/lto-streamer-in.cc=092023-10-08 17:00:12.084717019 +0200 @@ -1122,13 +1122,16 @@ input_cfg (class lto_input_block *ib, cl loop->estimate_state =3D streamer_read_enum (ib, loop_estimation, ES= T_LAST); loop->any_upper_bound =3D streamer_read_hwi (ib); if (loop->any_upper_bound) -=09loop->nb_iterations_upper_bound =3D streamer_read_widest_int (ib); +=09loop->nb_iterations_upper_bound +=09 =3D bound_wide_int::from (streamer_read_widest_int (ib), SIGNED); loop->any_likely_upper_bound =3D streamer_read_hwi (ib); if (loop->any_likely_upper_bound) -=09loop->nb_iterations_likely_upper_bound =3D streamer_read_widest_int (ib= ); +=09loop->nb_iterations_likely_upper_bound +=09 =3D bound_wide_int::from (streamer_read_widest_int (ib), SIGNED); loop->any_estimate =3D streamer_read_hwi (ib); if (loop->any_estimate) -=09loop->nb_iterations_estimate =3D streamer_read_widest_int (ib); +=09loop->nb_iterations_estimate +=09 =3D bound_wide_int::from (streamer_read_widest_int (ib), SIGNED); =20 /* Read OMP SIMD related info. */ loop->safelen =3D streamer_read_hwi (ib); @@ -1888,13 +1891,17 @@ lto_input_tree_1 (class lto_input_block tree type =3D stream_read_tree_ref (ib, data_in); unsigned HOST_WIDE_INT len =3D streamer_read_uhwi (ib); unsigned HOST_WIDE_INT i; - HOST_WIDE_INT a[WIDE_INT_MAX_ELTS]; + HOST_WIDE_INT abuf[WIDE_INT_MAX_INL_ELTS], *a =3D abuf; =20 + if (UNLIKELY (len > WIDE_INT_MAX_INL_ELTS)) +=09a =3D XALLOCAVEC (HOST_WIDE_INT, len); for (i =3D 0; i < len; i++) =09a[i] =3D streamer_read_hwi (ib); gcc_assert (TYPE_PRECISION (type) <=3D WIDE_INT_MAX_PRECISION); - result =3D wide_int_to_tree (type, wide_int::from_array -=09=09=09=09 (a, len, TYPE_PRECISION (type))); + result +=09=3D wide_int_to_tree (type, +=09=09=09 wide_int::from_array (a, len, +=09=09=09=09=09=09 TYPE_PRECISION (type))); streamer_tree_cache_append (data_in->reader_cache, result, hash); } else if (tag =3D=3D LTO_tree_scc || tag =3D=3D LTO_trees) --- gcc/data-streamer-in.cc.jj=092023-10-08 16:37:31.409278276 +0200 +++ gcc/data-streamer-in.cc=092023-10-08 17:00:12.098716826 +0200 @@ -277,10 +277,12 @@ streamer_read_value_range (class lto_inp wide_int streamer_read_wide_int (class lto_input_block *ib) { - HOST_WIDE_INT a[WIDE_INT_MAX_ELTS]; + HOST_WIDE_INT abuf[WIDE_INT_MAX_INL_ELTS], *a =3D abuf; int i; int prec =3D streamer_read_uhwi (ib); int len =3D streamer_read_uhwi (ib); + if (UNLIKELY (len > WIDE_INT_MAX_INL_ELTS)) + a =3D XALLOCAVEC (HOST_WIDE_INT, len); for (i =3D 0; i < len; i++) a[i] =3D streamer_read_hwi (ib); return wide_int::from_array (a, len, prec); @@ -292,10 +294,12 @@ streamer_read_wide_int (class lto_input_ widest_int streamer_read_widest_int (class lto_input_block *ib) { - HOST_WIDE_INT a[WIDE_INT_MAX_ELTS]; + HOST_WIDE_INT abuf[WIDE_INT_MAX_INL_ELTS], *a =3D abuf; int i; int prec ATTRIBUTE_UNUSED =3D streamer_read_uhwi (ib); int len =3D streamer_read_uhwi (ib); + if (UNLIKELY (len > WIDE_INT_MAX_INL_ELTS)) + a =3D XALLOCAVEC (HOST_WIDE_INT, len); for (i =3D 0; i < len; i++) a[i] =3D streamer_read_hwi (ib); return widest_int::from_array (a, len); --- gcc/tree-affine.cc.jj=092023-10-08 16:37:31.845272518 +0200 +++ gcc/tree-affine.cc=092023-10-08 17:00:12.089716950 +0200 @@ -805,6 +805,7 @@ aff_combination_expand (aff_tree *comb A =09 continue; =09 } =09 exp =3D XNEW (class name_expansion); +=09 ::new (static_cast (exp)) name_expansion (); =09 exp->in_progress =3D 1; =09 if (!*cache) =09 *cache =3D new hash_map; @@ -860,6 +861,7 @@ tree_to_aff_combination_expand (tree exp bool free_name_expansion (tree const &, name_expansion **value, void *) { + (*value)->~name_expansion (); free (*value); return true; } --- gcc/gimple-ssa-strength-reduction.cc.jj=092023-10-08 16:37:31.534276625= +0200 +++ gcc/gimple-ssa-strength-reduction.cc=092023-10-08 17:00:12.087716978 +0= 200 @@ -238,7 +238,7 @@ public: tree stride; =20 /* The index constant i. */ - widest_int index; + offset_int index; =20 /* The type of the candidate. This is normally the type of base_expr, but casts may have occurred when combining feeding instructions. @@ -333,7 +333,7 @@ class incr_info_d { public: /* The increment that relates a candidate to its basis. */ - widest_int incr; + offset_int incr; =20 /* How many times the increment occurs in the candidate tree. */ unsigned count; @@ -677,7 +677,7 @@ record_potential_basis (slsr_cand_t c, t =20 static slsr_cand_t alloc_cand_and_find_basis (enum cand_kind kind, gimple *gs, tree base, -=09=09=09 const widest_int &index, tree stride, tree ctype, +=09=09=09 const offset_int &index, tree stride, tree ctype, =09=09=09 tree stype, unsigned savings) { slsr_cand_t c =3D (slsr_cand_t) obstack_alloc (&cand_obstack, @@ -893,7 +893,7 @@ slsr_process_phi (gphi *phi, bool speed) int (i * S). Otherwise, just return double int zero. */ =20 -static widest_int +static offset_int backtrace_base_for_ref (tree *pbase) { tree base_in =3D *pbase; @@ -922,7 +922,7 @@ backtrace_base_for_ref (tree *pbase) =09{ =09 /* X =3D B + (1 * S), S is integer constant. */ =09 *pbase =3D base_cand->base_expr; -=09 return wi::to_widest (base_cand->stride); +=09 return wi::to_offset (base_cand->stride); =09} else if (base_cand->kind =3D=3D CAND_ADD =09 && TREE_CODE (base_cand->stride) =3D=3D INTEGER_CST @@ -966,13 +966,13 @@ backtrace_base_for_ref (tree *pbase) *PINDEX: C1 + (C2 * C3) + C4 + (C5 * C3) */ =20 static bool -restructure_reference (tree *pbase, tree *poffset, widest_int *pindex, +restructure_reference (tree *pbase, tree *poffset, offset_int *pindex, =09=09 tree *ptype) { tree base =3D *pbase, offset =3D *poffset; - widest_int index =3D *pindex; + offset_int index =3D *pindex; tree mult_op0, t1, t2, type; - widest_int c1, c2, c3, c4, c5; + offset_int c1, c2, c3, c4, c5; offset_int mem_offset; =20 if (!base @@ -985,18 +985,18 @@ restructure_reference (tree *pbase, tree return false; =20 t1 =3D TREE_OPERAND (base, 0); - c1 =3D widest_int::from (mem_offset, SIGNED); + c1 =3D offset_int::from (mem_offset, SIGNED); type =3D TREE_TYPE (TREE_OPERAND (base, 1)); =20 mult_op0 =3D TREE_OPERAND (offset, 0); - c3 =3D wi::to_widest (TREE_OPERAND (offset, 1)); + c3 =3D wi::to_offset (TREE_OPERAND (offset, 1)); =20 if (TREE_CODE (mult_op0) =3D=3D PLUS_EXPR) =20 if (TREE_CODE (TREE_OPERAND (mult_op0, 1)) =3D=3D INTEGER_CST) { =09t2 =3D TREE_OPERAND (mult_op0, 0); -=09c2 =3D wi::to_widest (TREE_OPERAND (mult_op0, 1)); +=09c2 =3D wi::to_offset (TREE_OPERAND (mult_op0, 1)); } else return false; @@ -1006,7 +1006,7 @@ restructure_reference (tree *pbase, tree if (TREE_CODE (TREE_OPERAND (mult_op0, 1)) =3D=3D INTEGER_CST) { =09t2 =3D TREE_OPERAND (mult_op0, 0); -=09c2 =3D -wi::to_widest (TREE_OPERAND (mult_op0, 1)); +=09c2 =3D -wi::to_offset (TREE_OPERAND (mult_op0, 1)); } else return false; @@ -1057,7 +1057,7 @@ slsr_process_ref (gimple *gs) HOST_WIDE_INT cbitpos; if (reversep || !bitpos.is_constant (&cbitpos)) return; - widest_int index =3D cbitpos; + offset_int index =3D cbitpos; =20 if (!restructure_reference (&base, &offset, &index, &type)) return; @@ -1079,7 +1079,7 @@ create_mul_ssa_cand (gimple *gs, tree ba { tree base =3D NULL_TREE, stride =3D NULL_TREE, ctype =3D NULL_TREE; tree stype =3D NULL_TREE; - widest_int index; + offset_int index; unsigned savings =3D 0; slsr_cand_t c; slsr_cand_t base_cand =3D base_cand_from_table (base_in); @@ -1112,7 +1112,7 @@ create_mul_ssa_cand (gimple *gs, tree ba =09 =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D =09 X =3D B + ((i' * S) * Z) */ =09 base =3D base_cand->base_expr; -=09 index =3D base_cand->index * wi::to_widest (base_cand->stride); +=09 index =3D base_cand->index * wi::to_offset (base_cand->stride); =09 stride =3D stride_in; =09 ctype =3D base_cand->cand_type; =09 stype =3D TREE_TYPE (stride_in); @@ -1149,7 +1149,7 @@ static slsr_cand_t create_mul_imm_cand (gimple *gs, tree base_in, tree stride_in, bool speed) { tree base =3D NULL_TREE, stride =3D NULL_TREE, ctype =3D NULL_TREE; - widest_int index, temp; + offset_int index, temp; unsigned savings =3D 0; slsr_cand_t c; slsr_cand_t base_cand =3D base_cand_from_table (base_in); @@ -1165,7 +1165,7 @@ create_mul_imm_cand (gimple *gs, tree ba =09 X =3D Y * c =09 =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D =09 X =3D (B + i') * (S * c) */ -=09 temp =3D wi::to_widest (base_cand->stride) * wi::to_widest (stride_in= ); +=09 temp =3D wi::to_offset (base_cand->stride) * wi::to_offset (stride_in= ); =09 if (wi::fits_to_tree_p (temp, TREE_TYPE (stride_in))) =09 { =09 base =3D base_cand->base_expr; @@ -1200,7 +1200,7 @@ create_mul_imm_cand (gimple *gs, tree ba =09 =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D =09 X =3D (B + S) * c */ =09 base =3D base_cand->base_expr; -=09 index =3D wi::to_widest (base_cand->stride); +=09 index =3D wi::to_offset (base_cand->stride); =09 stride =3D stride_in; =09 ctype =3D base_cand->cand_type; =09 if (has_single_use (base_in)) @@ -1281,7 +1281,7 @@ create_add_ssa_cand (gimple *gs, tree ba { tree base =3D NULL_TREE, stride =3D NULL_TREE, ctype =3D NULL_TREE; tree stype =3D NULL_TREE; - widest_int index; + offset_int index; unsigned savings =3D 0; slsr_cand_t c; slsr_cand_t base_cand =3D base_cand_from_table (base_in); @@ -1300,7 +1300,7 @@ create_add_ssa_cand (gimple *gs, tree ba =09 =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D =09 X =3D Y + ((+/-1 * S) * B) */ =09 base =3D base_in; -=09 index =3D wi::to_widest (addend_cand->stride); +=09 index =3D wi::to_offset (addend_cand->stride); =09 if (subtract_p) =09 index =3D -index; =09 stride =3D addend_cand->base_expr; @@ -1350,7 +1350,7 @@ create_add_ssa_cand (gimple *gs, tree ba =09=09 =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D =09=09 Value: X =3D Y + ((-1 * S) * B) */ =09=09 base =3D base_in; -=09=09 index =3D wi::to_widest (subtrahend_cand->stride); +=09=09 index =3D wi::to_offset (subtrahend_cand->stride); =09=09 index =3D -index; =09=09 stride =3D subtrahend_cand->base_expr; =09=09 ctype =3D TREE_TYPE (base_in); @@ -1389,13 +1389,13 @@ create_add_ssa_cand (gimple *gs, tree ba about BASE_IN into the new candidate. Return the new candidate. */ =20 static slsr_cand_t -create_add_imm_cand (gimple *gs, tree base_in, const widest_int &index_in, +create_add_imm_cand (gimple *gs, tree base_in, const offset_int &index_in, =09=09 bool speed) { enum cand_kind kind =3D CAND_ADD; tree base =3D NULL_TREE, stride =3D NULL_TREE, ctype =3D NULL_TREE; tree stype =3D NULL_TREE; - widest_int index, multiple; + offset_int index, multiple; unsigned savings =3D 0; slsr_cand_t c; slsr_cand_t base_cand =3D base_cand_from_table (base_in); @@ -1405,7 +1405,7 @@ create_add_imm_cand (gimple *gs, tree ba signop sign =3D TYPE_SIGN (TREE_TYPE (base_cand->stride)); =20 if (TREE_CODE (base_cand->stride) =3D=3D INTEGER_CST -=09 && wi::multiple_of_p (index_in, wi::to_widest (base_cand->stride), +=09 && wi::multiple_of_p (index_in, wi::to_offset (base_cand->stride), =09=09=09=09sign, &multiple)) =09{ =09 /* Y =3D (B + i') * S, S constant, c =3D kS for some integer k @@ -1494,7 +1494,7 @@ slsr_process_add (gimple *gs, tree rhs1, else if (TREE_CODE (rhs2) =3D=3D INTEGER_CST) { /* Record an interpretation for the add-immediate. */ - widest_int index =3D wi::to_widest (rhs2); + offset_int index =3D wi::to_offset (rhs2); if (subtract_p) =09index =3D -index; =20 @@ -2079,7 +2079,7 @@ phi_dependent_cand_p (slsr_cand_t c) /* Calculate the increment required for candidate C relative to=20 its basis. */ =20 -static widest_int +static offset_int cand_increment (slsr_cand_t c) { slsr_cand_t basis; @@ -2102,10 +2102,10 @@ cand_increment (slsr_cand_t c) for this candidate, return the absolute value of that increment instead. */ =20 -static inline widest_int +static inline offset_int cand_abs_increment (slsr_cand_t c) { - widest_int increment =3D cand_increment (c); + offset_int increment =3D cand_increment (c); =20 if (!address_arithmetic_p && wi::neg_p (increment)) increment =3D -increment; @@ -2126,7 +2126,7 @@ cand_already_replaced (slsr_cand_t c) replace_conditional_candidate. */ =20 static void -replace_mult_candidate (slsr_cand_t c, tree basis_name, widest_int bump) +replace_mult_candidate (slsr_cand_t c, tree basis_name, offset_int bump) { tree target_type =3D TREE_TYPE (gimple_assign_lhs (c->cand_stmt)); enum tree_code cand_code =3D gimple_assign_rhs_code (c->cand_stmt); @@ -2245,7 +2245,7 @@ replace_unconditional_candidate (slsr_ca return; =20 basis =3D lookup_cand (c->basis); - widest_int bump =3D cand_increment (c) * wi::to_widest (c->stride); + offset_int bump =3D cand_increment (c) * wi::to_offset (c->stride); =20 replace_mult_candidate (c, gimple_assign_lhs (basis->cand_stmt), bump); } @@ -2255,7 +2255,7 @@ replace_unconditional_candidate (slsr_ca MAX_INCR_VEC_LEN increments have been found. */ =20 static inline int -incr_vec_index (const widest_int &increment) +incr_vec_index (const offset_int &increment) { unsigned i; =20 @@ -2275,7 +2275,7 @@ incr_vec_index (const widest_int &increm =20 static tree create_add_on_incoming_edge (slsr_cand_t c, tree basis_name, -=09=09=09 widest_int increment, edge e, location_t loc, +=09=09=09 offset_int increment, edge e, location_t loc, =09=09=09 bool known_stride) { tree lhs, basis_type; @@ -2299,7 +2299,7 @@ create_add_on_incoming_edge (slsr_cand_t { tree bump_tree; enum tree_code code =3D plus_code; - widest_int bump =3D increment * wi::to_widest (c->stride); + offset_int bump =3D increment * wi::to_offset (c->stride); if (wi::neg_p (bump) && !POINTER_TYPE_P (basis_type)) =09{ =09 code =3D MINUS_EXPR; @@ -2427,7 +2427,7 @@ create_phi_basis_1 (slsr_cand_t c, gimpl =09 feeding_def =3D gimple_assign_lhs (basis->cand_stmt); =09else =09 { -=09 widest_int incr =3D -basis->index; +=09 offset_int incr =3D -basis->index; =09 feeding_def =3D create_add_on_incoming_edge (c, basis_name, incr, =09=09=09=09=09=09 e, loc, known_stride); =09 } @@ -2444,7 +2444,7 @@ create_phi_basis_1 (slsr_cand_t c, gimpl =09 else =09 { =09 slsr_cand_t arg_cand =3D base_cand_from_table (arg); -=09 widest_int diff =3D arg_cand->index - basis->index; +=09 offset_int diff =3D arg_cand->index - basis->index; =09 feeding_def =3D create_add_on_incoming_edge (c, basis_name, diff, =09=09=09=09=09=09=09 e, loc, known_stride); =09 } @@ -2525,7 +2525,7 @@ replace_conditional_candidate (slsr_cand =09=09=09 basis_name, loc, KNOWN_STRIDE); =20 /* Replace C with an add of the new basis phi and a constant. */ - widest_int bump =3D c->index * wi::to_widest (c->stride); + offset_int bump =3D c->index * wi::to_offset (c->stride); =20 replace_mult_candidate (c, name, bump); } @@ -2614,7 +2614,7 @@ replace_uncond_cands_and_profitable_phis { /* A multiply candidate with a stride of 1 is just an artifice =09 of a copy or cast; there is no value in replacing it. */ - if (c->kind =3D=3D CAND_MULT && wi::to_widest (c->stride) !=3D 1) + if (c->kind =3D=3D CAND_MULT && wi::to_offset (c->stride) !=3D 1) =09{ =09 /* A candidate dependent upon a phi will replace a multiply by=20 =09 a constant with an add, and will insert at most one add for @@ -2681,7 +2681,7 @@ count_candidates (slsr_cand_t c) candidates with the same increment, also record T_0 for subsequent use.= */ =20 static void -record_increment (slsr_cand_t c, widest_int increment, bool is_phi_adjust) +record_increment (slsr_cand_t c, offset_int increment, bool is_phi_adjust) { bool found =3D false; unsigned i; @@ -2786,7 +2786,7 @@ record_phi_increments_1 (slsr_cand_t bas =09record_phi_increments_1 (basis, arg_def); else =09{ -=09 widest_int diff; +=09 offset_int diff; =20 =09 if (operand_equal_p (arg, phi_cand->base_expr, 0)) =09 { @@ -2856,7 +2856,7 @@ record_increments (slsr_cand_t c) /* Recursive helper function for phi_incr_cost. */ =20 static int -phi_incr_cost_1 (slsr_cand_t c, const widest_int &incr, gimple *phi, +phi_incr_cost_1 (slsr_cand_t c, const offset_int &incr, gimple *phi, =09=09 int *savings) { unsigned i; @@ -2883,7 +2883,7 @@ phi_incr_cost_1 (slsr_cand_t c, const wi =09} else =09{ -=09 widest_int diff; +=09 offset_int diff; =09 slsr_cand_t arg_cand; =20 =09 /* When the PHI argument is just a pass-through to the base @@ -2925,7 +2925,7 @@ phi_incr_cost_1 (slsr_cand_t c, const wi uses. */ =20 static int -phi_incr_cost (slsr_cand_t c, const widest_int &incr, gimple *phi, +phi_incr_cost (slsr_cand_t c, const offset_int &incr, gimple *phi, =09 int *savings) { int retval =3D phi_incr_cost_1 (c, incr, phi, savings); @@ -2981,10 +2981,10 @@ optimize_cands_for_speed_p (slsr_cand_t =20 static int lowest_cost_path (int cost_in, int repl_savings, slsr_cand_t c, -=09=09 const widest_int &incr, bool count_phis) +=09=09 const offset_int &incr, bool count_phis) { int local_cost, sib_cost, savings =3D 0; - widest_int cand_incr =3D cand_abs_increment (c); + offset_int cand_incr =3D cand_abs_increment (c); =20 if (cand_already_replaced (c)) local_cost =3D cost_in; @@ -3027,11 +3027,11 @@ lowest_cost_path (int cost_in, int repl_ would go dead. */ =20 static int -total_savings (int repl_savings, slsr_cand_t c, const widest_int &incr, +total_savings (int repl_savings, slsr_cand_t c, const offset_int &incr, =09 bool count_phis) { int savings =3D 0; - widest_int cand_incr =3D cand_abs_increment (c); + offset_int cand_incr =3D cand_abs_increment (c); =20 if (incr =3D=3D cand_incr && !cand_already_replaced (c)) savings +=3D repl_savings + c->dead_savings; @@ -3239,7 +3239,7 @@ ncd_for_two_cands (basic_block bb1, basi candidates, return the earliest candidate in the block in *WHERE. */ =20 static basic_block -ncd_with_phi (slsr_cand_t c, const widest_int &incr, gphi *phi, +ncd_with_phi (slsr_cand_t c, const offset_int &incr, gphi *phi, =09 basic_block ncd, slsr_cand_t *where) { unsigned i; @@ -3255,7 +3255,7 @@ ncd_with_phi (slsr_cand_t c, const wides =09ncd =3D ncd_with_phi (c, incr, as_a (arg_def), ncd, where); else=20 =09{ -=09 widest_int diff; +=09 offset_int diff; =20 =09 if (operand_equal_p (arg, phi_cand->base_expr, 0)) =09 diff =3D -basis->index; @@ -3282,7 +3282,7 @@ ncd_with_phi (slsr_cand_t c, const wides return the earliest candidate in the block in *WHERE. */ =20 static basic_block -ncd_of_cand_and_phis (slsr_cand_t c, const widest_int &incr, slsr_cand_t *= where) +ncd_of_cand_and_phis (slsr_cand_t c, const offset_int &incr, slsr_cand_t *= where) { basic_block ncd =3D NULL; =20 @@ -3308,7 +3308,7 @@ ncd_of_cand_and_phis (slsr_cand_t c, con *WHERE. */ =20 static basic_block -nearest_common_dominator_for_cands (slsr_cand_t c, const widest_int &incr, +nearest_common_dominator_for_cands (slsr_cand_t c, const offset_int &incr, =09=09=09=09 slsr_cand_t *where) { basic_block sib_ncd =3D NULL, dep_ncd =3D NULL, this_ncd =3D NULL, ncd; @@ -3385,7 +3385,7 @@ insert_initializers (slsr_cand_t c) gassign *init_stmt; gassign *cast_stmt =3D NULL; tree new_name, incr_tree, init_stride; - widest_int incr =3D incr_vec[i].incr; + offset_int incr =3D incr_vec[i].incr; =20 if (!profitable_increment_p (i) =09 || incr =3D=3D 1 @@ -3550,7 +3550,7 @@ all_phi_incrs_profitable_1 (slsr_cand_t else =09{ =09 int j; -=09 widest_int increment; +=09 offset_int increment; =20 =09 if (operand_equal_p (arg, phi_cand->base_expr, 0)) =09 increment =3D -basis->index; @@ -3681,7 +3681,7 @@ replace_one_candidate (slsr_cand_t c, un tree orig_rhs1, orig_rhs2; tree rhs2; enum tree_code orig_code, repl_code; - widest_int cand_incr; + offset_int cand_incr; =20 orig_code =3D gimple_assign_rhs_code (c->cand_stmt); orig_rhs1 =3D gimple_assign_rhs1 (c->cand_stmt); @@ -3839,7 +3839,7 @@ replace_profitable_candidates (slsr_cand { if (!cand_already_replaced (c)) { - widest_int increment =3D cand_abs_increment (c); + offset_int increment =3D cand_abs_increment (c); enum tree_code orig_code =3D gimple_assign_rhs_code (c->cand_stmt); int i; =20 --- gcc/real.cc.jj=092023-10-08 16:37:31.735273971 +0200 +++ gcc/real.cc=092023-10-08 17:00:12.088716964 +0200 @@ -1477,7 +1477,7 @@ real_to_integer (const REAL_VALUE_TYPE * wide_int real_to_integer (const REAL_VALUE_TYPE *r, bool *fail, int precision) { - HOST_WIDE_INT val[2 * WIDE_INT_MAX_ELTS]; + HOST_WIDE_INT valb[WIDE_INT_MAX_INL_ELTS], *val; int exp; int words, w; wide_int result; @@ -1516,7 +1516,11 @@ real_to_integer (const REAL_VALUE_TYPE * =09 is the smallest HWI-multiple that has at least PRECISION bits. =09 This ensures that the top bit of the significand is in the =09 top bit of the wide_int. */ - words =3D (precision + HOST_BITS_PER_WIDE_INT - 1) / HOST_BITS_PER_W= IDE_INT; + words =3D ((precision + HOST_BITS_PER_WIDE_INT - 1) +=09 / HOST_BITS_PER_WIDE_INT); + val =3D valb; + if (UNLIKELY (words > WIDE_INT_MAX_INL_ELTS)) +=09val =3D XALLOCAVEC (HOST_WIDE_INT, words); w =3D words * HOST_BITS_PER_WIDE_INT; =20 #if (HOST_BITS_PER_WIDE_INT =3D=3D HOST_BITS_PER_LONG) --- gcc/tree-ssa-loop-ivopts.cc.jj=092023-10-08 16:37:31.945271198 +0200 +++ gcc/tree-ssa-loop-ivopts.cc=092023-10-08 17:00:12.083717033 +0200 @@ -1036,10 +1036,12 @@ niter_for_exit (struct ivopts_data *data =09 names that appear in phi nodes on abnormal edges, so that we do not =09 create overlapping life ranges for them (PR 27283). */ desc =3D XNEW (class tree_niter_desc); + ::new (static_cast (desc)) tree_niter_desc (); if (!number_of_iterations_exit (data->current_loop, =09=09=09=09 exit, desc, true) =09 || contains_abnormal_ssa_name_p (desc->niter)) =09{ +=09 desc->~tree_niter_desc (); =09 XDELETE (desc); =09 desc =3D NULL; =09} @@ -7894,7 +7896,11 @@ remove_unused_ivs (struct ivopts_data *d bool free_tree_niter_desc (edge const &, tree_niter_desc *const &value, void *) { - free (value); + if (value) + { + value->~tree_niter_desc (); + free (value); + } return true; } =20 --- gcc/gengtype.cc.jj=092023-10-08 16:37:31.491277193 +0200 +++ gcc/gengtype.cc=092023-10-08 17:00:12.092716909 +0200 @@ -5235,7 +5235,6 @@ main (int argc, char **argv) POS_HERE (do_scalar_typedef ("FIXED_VALUE_TYPE", &pos)); POS_HERE (do_scalar_typedef ("double_int", &pos)); POS_HERE (do_scalar_typedef ("offset_int", &pos)); - POS_HERE (do_scalar_typedef ("widest_int", &pos)); POS_HERE (do_scalar_typedef ("int64_t", &pos)); POS_HERE (do_scalar_typedef ("poly_int64", &pos)); POS_HERE (do_scalar_typedef ("poly_uint64", &pos)); --- gcc/graphite-isl-ast-to-gimple.cc.jj=092023-10-08 16:37:31.548276440 +0= 200 +++ gcc/graphite-isl-ast-to-gimple.cc=092023-10-08 17:00:12.089716950 +0200 @@ -274,7 +274,7 @@ widest_int_from_isl_expr_int (__isl_keep isl_val *val =3D isl_ast_expr_get_val (expr); size_t n =3D isl_val_n_abs_num_chunks (val, sizeof (HOST_WIDE_INT)); HOST_WIDE_INT *chunks =3D XALLOCAVEC (HOST_WIDE_INT, n); - if (n > WIDE_INT_MAX_ELTS + if (n > WIDEST_INT_MAX_ELTS || isl_val_get_abs_num_chunks (val, sizeof (HOST_WIDE_INT), chunks) = =3D=3D -1) { isl_val_free (val); --- gcc/gimple-ssa-warn-alloca.cc.jj=092023-10-08 16:37:31.534276625 +0200 +++ gcc/gimple-ssa-warn-alloca.cc=092023-10-08 17:00:12.089716950 +0200 @@ -310,7 +310,7 @@ pass_walloca::execute (function *fun) =20 =09 enum opt_code wcode =09 =3D is_vla ? OPT_Wvla_larger_than_ : OPT_Walloca_larger_than_; -=09 char buff[WIDE_INT_MAX_PRECISION / 4 + 4]; +=09 char buff[WIDE_INT_MAX_INL_PRECISION / 4 + 4]; =09 switch (t.type) =09 { =09 case ALLOCA_OK: @@ -329,6 +329,7 @@ pass_walloca::execute (function *fun) =09=09=09=09 "large"))) =09=09 && t.limit !=3D 0) =09=09 { +=09=09 gcc_assert (t.limit.get_len () < WIDE_INT_MAX_INL_ELTS); =09=09 print_decu (t.limit, buff); =09=09 inform (loc, "limit is %wu bytes, but argument " =09=09=09=09 "may be as large as %s", @@ -347,6 +348,7 @@ pass_walloca::execute (function *fun) =09=09=09=09 : G_("argument to % is too large"))) =09=09 && t.limit !=3D 0) =09=09 { +=09=09 gcc_assert (t.limit.get_len () < WIDE_INT_MAX_INL_ELTS); =09=09 print_decu (t.limit, buff); =09=09 inform (loc, "limit is %wu bytes, but argument is %s", =09=09=09 is_vla ? warn_vla_limit : adjusted_alloca_limit, --- gcc/value-range-pretty-print.cc.jj=092023-10-08 16:37:32.021270194 +020= 0 +++ gcc/value-range-pretty-print.cc=092023-10-08 17:00:12.098716826 +0200 @@ -99,12 +99,19 @@ vrange_printer::print_irange_bitmasks (c return; =20 pp_string (pp, " MASK "); - char buf[WIDE_INT_PRINT_BUFFER_SIZE]; - print_hex (bm.mask (), buf); - pp_string (pp, buf); + char buf[WIDE_INT_PRINT_BUFFER_SIZE], *p; + unsigned len_mask =3D bm.mask ().get_len (); + unsigned len_val =3D bm.value ().get_len (); + unsigned len =3D MAX (len_mask, len_val); + if (len > WIDE_INT_MAX_INL_ELTS) + p =3D XALLOCAVEC (char, len * HOST_BITS_PER_WIDE_INT / 4 + 4); + else + p =3D buf; + print_hex (bm.mask (), p); + pp_string (pp, p); pp_string (pp, " VALUE "); - print_hex (bm.value (), buf); - pp_string (pp, buf); + print_hex (bm.value (), p); + pp_string (pp, p); } =20 void --- gcc/gimple-ssa-sprintf.cc.jj=092023-10-08 16:37:31.532276652 +0200 +++ gcc/gimple-ssa-sprintf.cc=092023-10-08 17:02:04.985158068 +0200 @@ -1181,8 +1181,15 @@ adjust_range_for_overflow (tree dirtype, =09=09=09=09=09=09=09 *argmin), =09=09=09=09=09 size_int (dirprec))))) { - *argmin =3D force_fit_type (dirtype, wi::to_widest (*argmin), 0, fal= se); - *argmax =3D force_fit_type (dirtype, wi::to_widest (*argmax), 0, fal= se); + unsigned int maxprec =3D MAX (argprec, dirprec); + *argmin =3D force_fit_type (dirtype, +=09=09=09=09wide_int::from (wi::to_wide (*argmin), maxprec, +=09=09=09=09=09=09TYPE_SIGN (argtype)), +=09=09=09=090, false); + *argmax =3D force_fit_type (dirtype, +=09=09=09=09wide_int::from (wi::to_wide (*argmax), maxprec, +=09=09=09=09=09=09TYPE_SIGN (argtype)), +=09=09=09=090, false); =20 /* If *ARGMIN is still less than *ARGMAX the conversion above =09 is safe. Otherwise, it has overflowed and would be unsafe. */ --- gcc/omp-general.cc.jj=092023-10-08 16:37:31.701274420 +0200 +++ gcc/omp-general.cc=092023-10-08 17:00:12.088716964 +0200 @@ -1986,13 +1986,17 @@ omp_get_context_selector (tree ctx, cons return NULL_TREE; } =20 +/* Needs to be a GC-friendly widest_int variant, but precision is + desirable to be the same on all targets. */ +typedef generic_wide_int > score_wide_int; + /* Compute *SCORE for context selector CTX. Return true if the score would be different depending on whether it is a declare simd clone or not. DECLARE_SIMD should be true for the case when it would be a declare simd clone. */ =20 static bool -omp_context_compute_score (tree ctx, widest_int *score, bool declare_simd) +omp_context_compute_score (tree ctx, score_wide_int *score, bool declare_s= imd) { tree construct =3D omp_get_context_selector (ctx, "construct", NULL); bool has_kind =3D omp_get_context_selector (ctx, "device", "kind"); @@ -2007,7 +2011,11 @@ omp_context_compute_score (tree ctx, wid =09 if (TREE_PURPOSE (t3) =09 && strcmp (IDENTIFIER_POINTER (TREE_PURPOSE (t3)), " score") =3D= =3D 0 =09 && TREE_CODE (TREE_VALUE (t3)) =3D=3D INTEGER_CST) -=09 *score +=3D wi::to_widest (TREE_VALUE (t3)); +=09 { +=09 tree t4 =3D TREE_VALUE (t3); +=09 *score +=3D score_wide_int::from (wi::to_wide (t4), +=09=09=09=09=09 TYPE_SIGN (TREE_TYPE (t4))); +=09 } if (construct || has_kind || has_arch || has_isa) { int scores[12]; @@ -2028,16 +2036,16 @@ omp_context_compute_score (tree ctx, wid =09=09 *score =3D -1; =09=09 return ret; =09=09} -=09 *score +=3D wi::shifted_mask (scores[b + n], 1, fals= e); +=09 *score +=3D wi::shifted_mask (scores[b + n], 1, = false); =09 } =09 if (has_kind) -=09 *score +=3D wi::shifted_mask (scores[b + nconstructs], +=09 *score +=3D wi::shifted_mask (scores[b + nconstruc= ts], =09=09=09=09=09=09 1, false); =09 if (has_arch) -=09 *score +=3D wi::shifted_mask (scores[b + nconstructs] = + 1, +=09 *score +=3D wi::shifted_mask (scores[b + nconstruc= ts] + 1, =09=09=09=09=09=09 1, false); =09 if (has_isa) -=09 *score +=3D wi::shifted_mask (scores[b + nconstructs] = + 2, +=09 *score +=3D wi::shifted_mask (scores[b + nconstruc= ts] + 2, =09=09=09=09=09=09 1, false); =09} else /* FIXME: Implement this. */ @@ -2051,9 +2059,9 @@ struct GTY(()) omp_declare_variant_entry /* NODE of the variant. */ cgraph_node *variant; /* Score if not in declare simd clone. */ - widest_int score; + score_wide_int score; /* Score if in declare simd clone. */ - widest_int score_in_declare_simd_clone; + score_wide_int score_in_declare_simd_clone; /* Context selector for the variant. */ tree ctx; /* True if the context selector is known to match already. */ @@ -2214,12 +2222,12 @@ omp_resolve_late_declare_variant (tree a =09 } } =20 - widest_int max_score =3D -1; + score_wide_int max_score =3D -1; varentry2 =3D NULL; FOR_EACH_VEC_SAFE_ELT (entryp->variants, i, varentry1) if (matches[i]) { -=09widest_int score +=09score_wide_int score =09 =3D (cur_node->simdclone ? varentry1->score_in_declare_simd_clone =09 : varentry1->score); =09if (score > max_score) @@ -2300,8 +2308,8 @@ omp_resolve_declare_variant (tree base) =20 if (any_deferred) { - widest_int max_score1 =3D 0; - widest_int max_score2 =3D 0; + score_wide_int max_score1 =3D 0; + score_wide_int max_score2 =3D 0; bool first =3D true; unsigned int i; tree attr1, attr2; @@ -2311,8 +2319,8 @@ omp_resolve_declare_variant (tree base) vec_alloc (entry.variants, variants.length ()); FOR_EACH_VEC_ELT (variants, i, attr1) =09{ -=09 widest_int score1; -=09 widest_int score2; +=09 score_wide_int score1; +=09 score_wide_int score2; =09 bool need_two; =09 tree ctx =3D TREE_VALUE (TREE_VALUE (attr1)); =09 need_two =3D omp_context_compute_score (ctx, &score1, false); @@ -2471,16 +2479,16 @@ omp_resolve_declare_variant (tree base) =09=09variants[j] =3D NULL_TREE; =09 } } - widest_int max_score1 =3D 0; - widest_int max_score2 =3D 0; + score_wide_int max_score1 =3D 0; + score_wide_int max_score2 =3D 0; bool first =3D true; FOR_EACH_VEC_ELT (variants, i, attr1) if (attr1) { =09if (variant1) =09 { -=09 widest_int score1; -=09 widest_int score2; +=09 score_wide_int score1; +=09 score_wide_int score2; =09 bool need_two; =09 tree ctx; =09 if (first) @@ -2552,7 +2560,7 @@ omp_lto_output_declare_variant_alt (lto_ gcc_assert (nvar !=3D LCC_NOT_FOUND); streamer_write_hwi_stream (ob->main_stream, nvar); =20 - for (widest_int *w =3D &varentry->score; ; + for (score_wide_int *w =3D &varentry->score; ; =09 w =3D &varentry->score_in_declare_simd_clone) =09{ =09 unsigned len =3D w->get_len (); @@ -2602,15 +2610,15 @@ omp_lto_input_declare_variant_alt (lto_i omp_declare_variant_entry varentry; varentry.variant =09=3D dyn_cast (nodes[streamer_read_hwi (ib)]); - for (widest_int *w =3D &varentry.score; ; + for (score_wide_int *w =3D &varentry.score; ; =09 w =3D &varentry.score_in_declare_simd_clone) =09{ =09 unsigned len2 =3D streamer_read_hwi (ib); -=09 HOST_WIDE_INT arr[WIDE_INT_MAX_ELTS]; -=09 gcc_assert (len2 <=3D WIDE_INT_MAX_ELTS); +=09 HOST_WIDE_INT arr[WIDE_INT_MAX_HWIS (1024)]; +=09 gcc_assert (len2 <=3D WIDE_INT_MAX_HWIS (1024)); =09 for (unsigned int j =3D 0; j < len2; j++) =09 arr[j] =3D streamer_read_hwi (ib); -=09 *w =3D widest_int::from_array (arr, len2, true); +=09 *w =3D score_wide_int::from_array (arr, len2, true); =09 if (w =3D=3D &varentry.score_in_declare_simd_clone) =09 break; =09} --- gcc/godump.cc.jj=092023-10-08 16:37:31.535276612 +0200 +++ gcc/godump.cc=092023-10-08 17:00:12.097716840 +0200 @@ -1154,7 +1154,11 @@ go_output_typedef (class godump_containe =09 snprintf (buf, sizeof buf, HOST_WIDE_INT_PRINT_UNSIGNED, =09=09 tree_to_uhwi (value)); =09 else -=09 print_hex (wi::to_wide (element), buf); +=09 { +=09 wide_int w =3D wi::to_wide (element); +=09 gcc_assert (w.get_len () <=3D WIDE_INT_MAX_INL_ELTS); +=09 print_hex (w, buf); +=09 } =20 =09 mhval->value =3D xstrdup (buf); =09 *slot =3D mhval; --- gcc/c-family/c-warn.cc.jj=092023-10-08 16:37:31.319279464 +0200 +++ gcc/c-family/c-warn.cc=092023-10-08 17:00:12.096716853 +0200 @@ -1517,13 +1517,15 @@ match_case_to_enum_1 (tree key, tree typ return; =20 char buf[WIDE_INT_PRINT_BUFFER_SIZE]; + wide_int w =3D wi::to_wide (key); =20 + gcc_assert (w.get_len () <=3D WIDE_INT_MAX_INL_ELTS); if (tree_fits_uhwi_p (key)) - print_dec (wi::to_wide (key), buf, UNSIGNED); + print_dec (w, buf, UNSIGNED); else if (tree_fits_shwi_p (key)) - print_dec (wi::to_wide (key), buf, SIGNED); + print_dec (w, buf, SIGNED); else - print_hex (wi::to_wide (key), buf); + print_hex (w, buf); =20 if (TYPE_NAME (type) =3D=3D NULL_TREE) warning_at (DECL_SOURCE_LOCATION (CASE_LABEL (label)), --- gcc/testsuite/gcc.dg/bitint-38.c.jj=092023-10-08 17:00:12.084717019 +02= 00 +++ gcc/testsuite/gcc.dg/bitint-38.c=092023-10-08 17:00:12.084717019 +0200 @@ -0,0 +1,18 @@ +/* PR c/102989 */ +/* { dg-do compile { target { bitint } } } */ +/* { dg-options "-std=3Dc2x" } */ + +#if __BITINT_MAXWIDTH__ >=3D 16319 +constexpr unsigned _BitInt(16319) a + =3D 46809856770167726127621548193677044225438364376699537824160022717939= 628343291686588133221586710648915925157749537208566348709231774324477059728= 763319900537499845533358728035749014999310181139205148376149598710826496473= 833711815515586271543891072166123033253318533558175760051184685411593263726= 196963313436586869536391457057811006447186847584134858936693364541098769997= 908014021284990908118817091046496748623135893521289709626062603305553614183= 559928449847473785848765847011514477192311482631228386303550370060014144072= 442636469963633024041427127562602129493942248325061962900595999224341866123= 012213266776978118379033875934588490382169559099157722852052372530204821544= 784157311384081159363841342505493821326296144831789857414053309000499273268= 852511500478297389324409142700039689042715222530866107895467106606923445375= 759318175390086520343903540248030641357223961046714259192080918736743807117= 010096956744004469142748795978563733838165130991678206367028604654758524083= 789230717092884948587718679328070760084086678347179914817925081838771618312= 732334619953338746336344235621880377969700575932441037647685522242087626242= 598557198281818035387041014982421454431301328519954419349662422321998640294= 484962248942200767856494617479789279508933089953562472777752533078949270357= 456411225295514777094292976154560435086940424655827475235351037015722948500= 440213104315345429039792938727637405493857897687860646721735939868427505051= 910441391428602410680811634071227305942736229370315135549833621317069889444= 840536939875718852316046029271487585787996817357832819135821597249351327129= 787563440079330192925005282225863601565085768302390070984541083848793677853= 325040788618095457604634069790858402095129504884493804786565702907285079744= 297614689529418499373699950548566574281131379540553067419984805580275990178= 637682206952934297126196311933247650406428586936204966208340578982843313215= 493324281743280941581054818065875039369227272958623284206565849097120192778= 001425881533311545969511794227355187664684482107672366404028277283451141989= 135127816901710398709480382959428635234046834661872608878149262681618865733= 135910417181982267380585631782849903908808822313725829737392904330767357009= 039694778959879992292864384353261701216481107461888177462262894353903797488= 381268913080186091509003587024406100581941813006839098647031467785360508010= 331341183790435828783740154625741324046693989352750893154106524192987230720= 387644388210619326254465229013236469167191033200612786414699140401536668356= 931724805794959607035492936115832695555160023607526843504410588016279838079= 916160798736528245866203159909692182517620270789073002369870685576293269168= 825936535896407659582457777527599118314911837204720605511846311286460406385= 389482040724983787136893494143811968060552854688725693433424607559674641029= 795445863235817142871414182091818338443568133237931754104825239171071219662= 340633870206119521372456930328540224285367138611314821153569168546183645829= 503753803437831805510824008241444120530040152673239995922834692652858685274= 338949097873478792672199985538879471183716442300771962610917900546611370645= 076526968758081982277218930108450362729738967513422822233728686764111051106= 198023124788453349244289893674342964195831413532907340649577636920815803211= 588385069101056904898394112677147799097609225239197281269166984744679850724= 410612166788542302561376925810277385553750973329580501331393740228280489721= 384722107264711160517234946456408991490649350813385538962717766342605776325= 208628632534381125475768180306827627804875799742528433471319022681846302307= 446190017695801005557243498313517114536524233927332698446518106428726464547= 083209111510064058410437557730405695196945620013848531356000927233822810363= 776386328926167325872673675340704414366407947949697258056053449480617081046= 930477300587359062628007238799966852254674798570159961397510118854385785214= 155925163405867671830800032486980962819944268156561566291262602279606441449= 610634423643128569768835770799298996656155717172997209353300747694786221592= 258320481118901555050564208247540064763952078218777682539559825742171410647= 386979764267826638075587335674781227397769160414784274115172291946473489032= 677259497902240322819107558691046420487025467429043766886117763971311276299= 639024610203099491718695782698208419415687039831233605910052156603409274069= 464261319290985064400393374512929106257634121387481551009983570872335543297= 009013967112023291074766590619136016025951219816084978419759730010622394596= 088660312713603712000086496866865145241104837289560738290749427881097147566= 394494879145861866225023837516652348484750734204006680185622232898866204957= 929960054568249041275448362105119023162319626554939196425978017807049564253= 888378950337940653127933886695515764665491340518187925418918590429832586550= 339568878631106766927360967060307658260725352708497774453318714564268623635= 016559398042857511932991192138224078050452742263065408694106024275713131318= 4709635181001199631726283364158943337968797uwb + + 99354435180574564299271266552222578172075113116713358325600655730552= 766787479906529073488397418185627579390846490733481721083971838270203779417= 259831075136362874065305263582535084372902419372769083862829043530791029045= 356756086045764861629983194277028512784082136414548372230796164016158756724= 532501484216792238294178342275181330910551802702492661616766771761496751642= 576408123442979356507296298018787580599440901688627305198172033523414583103= 638114823180832702324343293173238228189911345006016698689223960135129694778= 394564723458123123219242152418497721476874557602245592409527373190093485408= 949663635681583495013552292646467700180715905024417027872690979739798998376= 831221941031100897284256766902460911469939550379184257728400222882228329325= 425160915011494771608565644643769102932300919635731192306480266678963993527= 909826119575699789720381785195702784475407075028616785026579051927432258932= 256639948075689186448982737022854836763857176511040420021053529931765121664= 200850644524317531813658058335489226767488904124203326946090968197797656003= 452163903943072575567782237434439589839621137231935512478979954237623480921= 038936837113738971391682894202676606114099476445487150077878329592511675531= 750966391476747761179731004479032436269028923822637675913280382357085934015= 637930194181244531663864717924684210038558942065843547314893636681340779462= 035460672372356577464802968316517917903859813975584589059046413942462797827= 467360091018623668680683634119763885576979219143171793712064440853907796348= 313697233700507646788528467793694972323747806919052809923680797627473522455= 196072641541971489588969556619042149091849522899961420506048216087499004178= 451377275969031004523500675513058409982804827752098832788730718955887518114= 623425178257534938149979184184374554749924222439195499673719644234574402872= 962708556058509546859126443033540190587169167355225330653230577554798036687= 825302503819882110750346557601232502494414406843384509538232903469096898225= 276526987235028723125703052611967684774988980207930718087589033817968738686= 823788509252116293927606286852227450735441166156355579108053576235902180237= 158327163725325193728620938285457973255678036919980517851560658615668888714= 611301335220393218434390179643820300807524767093987313411730624302750031119= 549076278372084883486866669047657106569177064709243184321601554507260076680= 354945717797931292122421012932748532378508488061527744636892434266832958846= 486807902403630970152183479663991663800903706285912887123051331718696396799= 228540664930767731669701904829888280170310168915619719862796753719630209324= 693372640613177863305668393839893847609355902992879635468638481199994517395= 484051240015140330966956055807661216114406385499888959702624251332181598480= 617272171634871318064816867668437899714652479035348538379514138457866671224= 271826489891565995296474394195537851585616131140232673038699275651705077817= 823664470113408512581785341015859500814234377037784923474482304738976435057= 739573855041121824466905850338237471759669290912936932010618586701412091290= 914528612922762760129106240712411654020891616069444238262454616085949357324= 819001982408622934094423088006900195508316304798830005798846146019069617230= 113544498045767943398260569869576800909160468486734197235296943846538094003= 772185450752691487661291946370394082255156780133321880749972176678354949400= 430149178774383549026731074531642752800102510403600409373087389256894757251= 316390320119790096427135422928942190593529729331511123761973838149253632886= 709955562694478049949250867917281369066932495071150978070603658721109982107= 683360783895087241848635972859877369120730719801371625907796646750334291193= 278553078271746737492574629830542216317975270099875957324602221973676084409= 734882118984714393020513888068185216596858736723838280213298481534102049266= 077109716782685416775844216952380117843513860478691587871566346306938724280= 678649803200632934358875747458590670249884857423532785487044675442987935115= 835876597137116770657923711993294193723927203219818622698900248323489998654= 493398563392203868531626419844449349981762488217031547747940268634238466653= 611479125803101793332398493141451581038137243712771560318260702136561892184= 285511714925793677366526502405108405244792806619221493703814048636680382299= 221050646583350833149468425459780504970217952171249479595750654717498722788= 027563713908714410042326332526118257486585935406678310988740272233275415237= 428577509541196157085415141451108639250492045175740008247979008175853769614= 627545214951001988296751009580666395319581067041597172650352055971610478795= 108499005875657466032257631298774343179498421057423869658861371177986421681= 907333674141267979294346275323078554488410354337952290312755458858728768488= 466666664754658669053322930953814940967023286499207405066589305030531627779= 448214333834072831551787079709064580238271416811403729683560846170010538704= 990798843840198208755858431290828946877405339467637568469249528252513830263= 64635539377880784234770789463152435704464616uwb; +constexpr unsigned _BitInt(16319) b + =3D 20129744567093027275741005070628998262449166046517026903695683755854= 448756834360166513132405078796314602781998330705368407367482030156637206994= 877425582250124595106718397028199112773892105727478029626122540718672466812= 244172521968825004812596684190534400169291245019886664334632347203172906471= 830047918779870667296830826108769036384267604969509336398421516482677170697= 323144807237345130767733861415665037591249948490085867356183319101541167176= 586195051721766552194667530417142250556133895688441663400613014781276825394= 358975458967475147806589013506569415945496841131100738180426238464950629268= 379774013285627049621529192047736803089092751891513992605419086502588233332= 057296638567290306093910878742093500873864277174719410183640765821580587831= 967716708363976225535905317908137780497267444416760176647705834046996010820= 212494244083222254037700699529789991033448979912128507710343500466786839351= 071045788239200231971288879352062329627654083430317549832483148696514166354= 870702716570783257707960927427529476249626444239951812293100465038963807939= 297639901456086408459677292249078230581624034160083198437374539728677906306= 289960873601083706201882999243554025429957091619812945018432503309674349427= 513057767160754691227365332241845175797106713295593063635202655344273695438= 810685712451003351469460085582752740414723264094665962205140763820691773090= 780866423727990711323748512766522537850976590598658397979845215595029782750= 537140603588592215363608992433922289542233458102634259275757690440754308009= 593855238137227351798446486981151672766513716998027602215751256719370429397= 129549459120277202327118788743080998483470436192625398340057850391478909668= 185290635380423955404607217710958636050373730838469336370845039431945543326= 700579270919052885975364141422331087288874462285858637176621255141698264412= 903522678033317989170115880081516284097559300133507799471895326457336815172= 421155995525168781635131143991136416642016744949082321204689839861376266795= 485532171923826942486502913400286963940309484507484129423576156798044985198= 780159055788525538310878089397895175129162099671894337526801235280427428321= 205321530735108239848594278720839317921782831352363541199919557577597546876= 704462612904924694431903072332864341465745291866718067601041404212430941956= 177407763481845568339170224196193106463030409080073136605433869775860974939= 991008596874978506245689726966715206639438259724689301019692258116991317695= 012205036157177039536905494005833948384397446492918129185274359806145454148= 241131925838562069991934872329314452016900728948186477387223161994145551216= 156032211038319475270853818660079065895119923373317496777184177315345923787= 700803986965175033224375435249224949151191006574511519055220741174631165879= 299688118138728380219550143006894817522270338472413899079751917314505754802= 052988622174392135207139715960212346858882422543222621408433817817181595201= 086403368301839080592455115463829425708132345811270911456928961301265223101= 989524481521721969838980208647528038509328501705428950749820080720418776718= 084142086501267418284241370398868561282277848391673847937247873117719906103= 441015578245152673184719538896073697272475250261227685660058944107087333786= 104761624391816175414338999215260190162551489343436332492645887029551964578= 826432156700872459216605843463884228343167159924792752429816064841479438134= 662749621639560203443871326810129872763539114284811330805213188716333471069= 710270583945841626338361700846410927750916663908367683188084193258384935122= 236639934335284160522042065088923421928660724095726039642836343542211473282= 392554371973074108770797447448654428325845253304889062021031599531436606775= 029315849674756213988932349651640552571880780461452187094400408403309806507= 698230071584809861634596000425300485805174853406774961321055086995665513868= 382285048348264250174388793184093524675621762558537763747237314473883173686= 633576273836946507237880619627632543093619281096675643877749217588495383292= 078713230253993525326209732859301842016440189010027733234997657748351253359= 664018894197346327201303258090754079801393874104215986193719394144148559622= 409051961205332355846077533183278890738832391535561074612724819789952480872= 328880408266970201766239451001690274739141595541572957753788951050043026811= 943691163688663710637928472363177936029259448725818579129920714382357882142= 208643606823754520733994646572586821541644398149238544337745998203264678454= 665487925173493921777764033537269522992103115842823750405588538846833724101= 543165897489915300004787110814394934465518176677482202804123781727309993329= 004830726928892557850582806559007396866888620985629055058474721708813614135= 721948922060211334334572381348586196886746758900465692833094336637178459072= 850215866106799456460266354416689624866015411034238864944123721969568161372= 557215009049887790769403406590484422511214573790761107726077762451440539965= 975955360773797196902546431341823788555069435728043202455375041817472821677= 779625286961992491729576392881089462100341878uwb + / 42uwb; +constexpr unsigned _BitInt(16319) c + =3D 26277232382028447345935282100364413976442241120491848683780108318345= 774920397366452596924421335605374686659278612312801604887370376076386444511= 450318895545695570784577285598906650901929444302296033412199632594998376064= 124714220414913923213779444306833277388995703552219430575080927111195417046= 911177019070713847128826447830096432003962403463656558600431115273248877177= 875063381111477888059798858016050213420475851620413016793445517539227019973= 682699447952322388748860981947593432985730684746088183583225184347825110697= 327973294826205227564425769950503423435597165969299975681406974619941538502= 827193742760455245269483134360940023933986344217577102114800134253879530890= 064362520368475535738854741806292542624386473461274620987891355541987873664= 157022522167908591164654787501854546457737341526763516705032705254046172926= 268968997302379261582933264475402063191548343982201230445504659038868786347= 667710658240088825869575188227013335559298579845948690316856611693386990691= 782821847535492639223427223360712994033576990398197160051785889033125034223= 732954451076425681456628201904077784454089380196178912326887148822779198657= 689238010492393879170486604804437202791286852035982584159978541711417080787= 022338893101116171974852272032081114570327098305927880933671644227124990161= 298341841320653588271798586647749346370617067175316167393884414111921877638= 201303618067479025167446526964230732790261566590993315887290551248612349150= 417516918700813876388862131622594037955509016393068514645257179527317715173= 019090736514553638608004576856188118523434383702648256819068546345047653068= 719910165573154521302405552789235554333112380164692074092017083602440917300= 094238211450798274305773890594242881597233221582216100516212402569681571888= 843321851284369613879319709906369098535804168065394213774970627125064665536= 078444150533436796088491087726051879648804306086489894004214709726215682689= 504951069889191755818331155532574370572928592103344141366890552816031266922= 028893616252999452323417869066941579667306347161357254079241809644500681547= 267163742601555111699376923690500014172294337681007418735910341792131377741= 308586228268385825579773985382339854821729670313925456724869607910114957040= 810377671394779834675225181536565444830551924417794139736686594557660483813= 045525089850285373756403594900392226296617656189774567019900237644329891280= 192776067340109751100025818473155267503490628146429306493520953677660612094= 758307190480072039980575323428994009982415676875786338343681850769724258724= 712947129844865182522700509869810541147515988955709784790248266593581532414= 091983670376426534289079098742549505127694160521110700035496658932724007621= 759500091227595477831200325335242614162624218010753586306794482732500765136= 299548052958345872488446969032973871418565484570096440609125401439516349061= 951073344772753817168731533186740449206533184858409824331269879276752302819= 075938894191764603880669059804914705202932220114574769307945938446355744093= 058483466098741029671133305308451601510124097336668044362140994842230895354= 232007936193610666215236351383330719496758577095102466235782700820575938453= 736277546445932135116947993404356975890051717304128693125699951445791328843= 668647245439797933691355015781238038148597339831348341049751957204680813855= 138272253234219030458164179195368888878989362640509486440530112337687890165= 646824152338885218611665567933423652236621168833497594762922586523151554244= 316284075364923316223457798336995440229801638249044555841786652868778333857= 626201712694823945146208412572567947403078655159448178467488335673853886982= 143607843369103504905837049147006413324087204923968347162406372146304110247= 436210704329838033967549296094708909042352807942165389054391217609084676765= 464997803900415653278041220586434133698802658726748950122980183615091029049= 242919298428066745937148593879994539254240070220900694662200741796632687373= 414952817000938093930497338259168439649970963774406833411431113922194082765= 390241161715106142638681072839764035976877223152727829248475639970029777900= 589595383604989099084081251802305001465530685587689066710306032849298712531= 664047230963409638484129598076118133347670029704549206295184751171783054889= 490211218045322681317529569999778899567668829982207035948032411418382057247= 326141072264502161892285323531743728756335449414720326329614400327415751813= 608405440522389476951223717685562226240221655814783640319063683104993438443= 847695342093582440489676230855515734722099028773790309518629302472390856918= 840009781940193713784596688294176313226823907143925396584175086934911386332= 502448539920116580493698106175151294846382915609543814748269873022997601962= 804377576934064368480060369871027634248583037300264157126892396407333810094= 970488786868749240778818119777818968060847669660858189435863648299750130319= 878885182309492320093569553086644726783916663680961005542160003603514646606= 310756647257217877792590840884087816175376150368236330721380807047180835128= 240716072193739218623529235235449408073833764uwb + >> 171; +static_assert (a =3D=3D 10403542085759133691203342137159028259461894955438= 331210801665800234672962180907518788681055608925051917190662144445433835595= 489501570265148539013616306519011285861864113638610998587283343748668959870= 044400340187367869274012726759732348878437230149364081610941398977036594823= 591463255731808309715219781556045092524781748798096243155527048746090614751= 043610821560662864236720952557147844731917800712343725546175449104075627616= 077829385396994452199410766816558008090921987787438967590914249326913953731= 957899714113110918563882837045448642562338486517475793442626878243475178869= 958697311252767202125088496235928130685145568023992654921893286093433280015= 789621699281948053130963767216950901322064090115301029360256916486236324346= 980555378227825665231041206505932451054100655891377307183657244188881780309= 602697733965633806548575793711470844175477213922050584861112947113328821094= 578714380110663964395764964375008963336325761662071121014767368961020824065= 775639039724097407257977371623360602667242992626829630277589757195892131842= 788347638167481783472539736593840645020141666099662762763659119482517961624= 374850646183224354529879255694192077493038699570091875155722960929748259201= 284457182471153956119946261637096783796538046622701136421992223281799392319= 105563566498086105138357131671079600937329401554014025354725298453142629483= 842874038291307431207948198280389112036878226218928165845324560374437065373= 122000792930554833265840423016148390974876479752688661617125284208020330726= 704780298561478529279775092768807953202013307072084373090254748865483609183= 726295735240865516817482898554990450888147008484162850924835809973020042760= 450232447237837196378388135483084055028396408249214425019231777824054821326= 738728924661602608905318664721047678808734917923923121217803736039325080641= 571812479260200189082647677675380297657174607422686495562781202604884582727= 406463545308236800937463493199421020490845203940782000643133713413924683795= 888948837880891750307666957538835987772265423203470320354145742841869795472= 799186154631385288573730129094228733379855432514817031425884584962254283999= 586850250406406681047191820544352342046667950146374296364655891915135310082= 529994904874562441551527081311638121766367661807914647092917287784017613115= 795691373814041086838720316968010349263776702775009771662737124600992709418= 630470128579612748138807983617697487500079502839532266478317788699680283395= 230308668613168191852557234122469290277763000256531531071762280960597416576= 452124575885006363492171314551026369237325119844147154972582617127637240421= 323781252125819313268498872048683068789228870983086306586111793007178693570= 562554975762384431236664489360478109692520183356042112794589756922036102025= 380888246082763911915622037570736969677850621708281909652070776450422110772= 285659921383413532725137107621514770958361581240471968542997294446402584844= 918179956881219978405772785713402046471903103404871352324277109089891640558= 983922159359479964068994923538490500501798825116238188381267330618026093160= 290205596669795981834842352271011063939632623926629960113926326029952143452= 354640614061049438932665467928443113232214498101774523178129020155017228802= 221901469548072234073334681052461327832268955923701109732874360984002493130= 025470753861967432493102395766279717815113135763810886216491770265724160887= 688887515282293447287121039545323777928286876711267049135547760773655845950= 622676327972280622345486253084626121247885891757458308974259466441284967765= 824561478351421051923081842594791616249682768594796413184742007504540382141= 773556098929461233842797978566466734240436032269122908057438314319410489575= 244845739320693764798687398942275314333361838560358278583766983210126081046= 020231469705836544611252075187733112560778125560225565803349953151880800601= 890382648216375737077015744684142132303864494083237680306898134033570758401= 131735819237730280209424231954121970154195575070728876653187928423918894211= 617093567094857926079694003950142962763480728907322409338954277493711834363= 423032309296862081371923061150409402403668284066920335645815769603890931600= 189625120845560771835017710222988445713995722670892970377791415975424998772= 977793133120924108755323766471601770964843725827421304729349535336212587039= 242582503381150992918495310760366078232133800372960134691178665615437284018= 675587037783965019497398984583781291648236566997741116811234934754542646608= 973862932050896956712947890625239848619289180051302224085308716715734850608= 995498117691600907423641124622236235949675965926735290984369155077055324647= 942699875972019355174794849379024365265476001505043957802797349447782453767= 742359446787304217770032967959809288342189111153359045680464231699344620995= 535326063943372491385550455978845273436611631962336651743357242055102619760= 848116407351488643448217122169718350824452317641509534606434395208225350712= 889271762643740106849245478364448395994915755050465135468245061369394410933= 866013068008514339549345174558881983866497072827311379042433413uwb); +static_assert (b =3D=3D 47927963254983398275573821596735710148688490586945= 302151656389894891544659129428967888410488282848368101861900787393734303255= 909595611040969035422441862500296655015996183400474078033076442208281022919= 339382663505873362486125052306726201934754009977462857545931535761634915082= 457969531364063028166780758999692064992454347878021515200637154689307943805= 765515434945644517436059064850821739923175860513488184741071305928775874657= 579331194145636134729035837432777505225398881945576787038414037353432531906= 221463764944822367521370140398750351920450032158498609394040097931192233762= 919615392739593496142319079251492975289355219161278102593077980694080934774= 807348815686269838231658663255453109747406854147841668747295898035046214722= 954204337096637695161230258009467203656917423590804239279208200992286134875= 490081064276216238601176771626719652470715951261404740555830904552686923119= 865477301873427026359632829140952933264973522266815070542033531976946547220= 197973086938491321120720753739960137399906970065546372022920105333218600697= 858250092770971284041999765371634305856374505354948168051485795619245710565= 177475544471205491166573508574008824290197617246557203404659741951935583377= 220245975415117684554899445620844502922298410099631370945492174513316818179= 053941295889751044787346934407150836832047822816077953368388724034918957631= 287532906408983549478253389828549312675591697063148899645182358568234280904= 393370464356625596517001437115695750865735696271243546529127281196748236370= 851643906557876213318702947945779409043920207097980173253604088090541910037= 502992188977212850308451093143517148397901877959716663055881990934822377000= 137739027330737305203072941381961798582398137407044371548508882948736515168= 678665314156055539763283978378697347560390812910312112592558243537758659944= 336321765948248602151244471507899974214561619241705438327522143175018570171= 179348707944798029574180941726592337226502723788420039623849392735910288582= 594856812800635227346505171247207005920245031905445152238832105970200308151= 371801900107107616143235847115536995978281165233083750307528808742605565540= 002941143874829336203146501750257713925224473144855518861387693696103669523= 617994232375111611201101459297439748647388267459200813013679266349328732383= 431914791502242752803351817813918019855167200467126443959596212095412230012= 937785180621368904740496659226139300584975540396940968189138713630212621475= 457757421407899273838583419421850094135489271442461781867612967840281259964= 938951919393938448193171251996576357123654457926939171468811259400443993779= 102766652727502895609600502472189226835366234904950156893142674698374992326= 628993607966485208811438064202797698153274845831487974169502396605979807274= 335098034836109236427828852711258048141786054778320994100643663029556902570= 837898367870844766792830052796171750493189799905267492521148625102911003353= 413851945670464764491436591194854953791559798723403394543172251931597408230= 783241193488626433308391622670766594854714782494114377403163099298640358928= 143049334330420757343195444050636710200574691425877526862566305694461542707= 733031232666443103430989472012268269487427473562080231601131548241018299190= 616533588303175681201813391409086131938902379083952833720360688912943648792= 014016737028487092443886087383029664801442484437819591293255142678077981975= 752535336855805082530356241998952865342550778119356839913188367344788882869= 555211229365407308833977580823432443662765954396216494645039675972304007590= 676650615202226481515809367464962286957243012116484337925382676418395332482= 943675100503507815220367552316843116120946303449177210299631555487831100050= 075236979610968511974561546844657652354600832503906077552097096336790921653= 334305722166205970710071599011452051510942858155477347155178222397083241240= 607349989679794924719726305591105357558068555200222677799099434663185151779= 136463033055175444365657794849872636280668141970553674032426859753989628280= 355279972608055457330269595842841726967166030617385338134381402404827936273= 803947019883936570628616414755586493336436328787509713812842557390990443318= 379509867020380053354885621917457990109708412341140216044839027465621606220= 773380452267811600783048591111833813729141550004024463664622846527554661318= 545121547721492409389740865925389787233163029436137942926808211251948997928= 382653291328290814782484778151796477938082491839492432242010471783901296042= 252376674439710606346399821841652194708961984612546483314531228197199405727= 591759159127914527483728327356941190487588359081892701108376611136862387628= 866146969785698402392454111735458471072816206092874754444972907108640607282= 082670735270509846957021243000503176987077098449014754492254187858251649602= 605563421853473982976704443111427277286348462896880059204798597700568726057= 437433260876574696564797640594970930403341444263058148836225175692288351728= 756577265334618966609417525651898087863205788909104258464451037447721910608= 0358138511257658994752983022904583136418485544787844335722425uwb); +static_assert (c =3D=3D 87791074236971898375693906050841211797859249085219= 857442103255912236679245196526258183737200195092459037070061326325721733862= 550642013557351987594406882625147809841117910427395663017848973163739949221= 920509632722884340603422885119715696976800265237608112255164300526997540446= 828188926798191319956002162809660627367323847324113616574443996958838650961= 034287596228138677355472599785293194368898640136872193905676042833180111007= 999534515209684412648660318139544886280584751143487292754141431589178747095= 995562471836958538385523210889734458760880425568104799106614493746619996750= 828111038144533532941948866129614927372632772715518890386107307604784595692= 561493219983504140230663638149893111097283117129890229962472801825879214491= 853539228859378776045004007387742400087099452897916050111777396577201816014= 535122598820045644624158286527149042897272352105372777213898166876433661452= 000011777121121975156955788874837929887554354013884561458544888805370883603= 979946432160148284956624602056864485481132298410979556139584409013754162565= 328645118522986963276115172332413247990709194912864261597887926317238337174= 515380434373640171852377431824028356700876831256026403188874515966503235287= 201281881985472704629716121576034879585267050059555804094416707718493880164= 380358501945858703270134092362367309142177220256553194722311416667902879556= 857136362746535655774542758385903508061686391652646764704409303516129925189= 046646477158058659410384237683768466978175431224095175917172922387459403459= 005304585514685192457678645317421021786288543765245133679832091869745757657= 072739737753868400812388038803350957408363865272082673118089735224503911890= 557398289369373596931672405246606249458569070420412573471920869840096409845= 093226225038902560463247683416326435464557790353760020616911131212342731649= 379841717742423277699156887425640494541631583181218185827647752680912924708= 890884455751080226880692716971982831514696454008705070066637993306617027027= 474432542204783110564072207496481031234354733815835208730552187341151209786= 784404558964588524975699899667232359656087068265936071288476301376185091512= 558347426364387962855698738699677293418712135210300114273729873885726742284= 413334588575122260492832433475214578049120087810369667863747603253414920332= 978483681609032604700190675353306116459095608887974519070883897641904030079= 983051686730294469340122451388381805960985594425706961500112962181441863870= 246158853022907449053406669059217439700137798133324937711920480432972814232= 484890568414170138076703081910957324642214513769972707454684597021527968182= 227457305657212026631030431211601014598336832495586844591088625369619943085= 350399708145578212681703887459419803788389699105928956705542918117397687718= 299410438578196037512469579622360911547558939620383631206904838624230010389= 486206816112538671492964636904178280343035479227922490985224047514289607138= 750504639061341508460897057144703039182990126916002853558594129248477604970= 769784327224466025218250890974545423433548473473960450795877572106353569992= 687064654257888333111905176230618606752300109941271964590303221667515716566= 423216907874719066094734960347896437104781622556640929912514467878876353518= 529338268207197817337545781610734013626681098191139242522911257413952714743= 423055745369749182739385135974189637873089945934341918906877303024959106860= 723388364131591622810722635427582576995880898386774693974678993480652935817= 510358443898483871618474351603272760666036831317032464104091228327933767515= 126887451955640216460692459923633964681005135362116514506105233152116971257= 746388453132439730835364176920759624869188446674321443530197229596536386329= 482940499842668618701512553150233467246714304992579939580490880661608705450= 252765979751548555376202656903540410287427427550743965976319653203807825009= 445684240534200383575249171250992413349900321895264658381929721109708613800= 609868020819480443455264148571585699390058952366723063443482128058512699207= 110438913068758730163306016739732493270725035718735183667505750700910512885= 907647886301909667768540315789393826907090226674217344428417846808264941466= 205898628296127042795216377404216941950514000952780847169746246152083925855= 732001826641570668138493460583217631565239656984659013960251521596421935629= 007438127158858110572125790178604885399603344067027526885952173602194709687= 380097740679150371570274922091088013377075625712668977239114012033743084907= 932262009743533568353117563848956929098027209489681315046048554669619873147= 018464603421352019143561525916848109246883509291401201876930893242559246345= 785764270044263392994938334345029515939025514510022928396350009042532500218= 846254176287564398629643255627207095287849648686873308478944769995773265823= 323502131488612054136523374993834165315457072729079947556383396302215767079= 549642362109626938046397147546686798411349283930812842091580982026837446505= 139189201683305984323623897774718706310394884087693548630019675317294156866= 31571754649uwb); +#endif =09Jakub --SaffQZ6I7XRJud0v Content-Type: text/plain; charset=us-ascii Content-Disposition: attachment; filename=Q012sa --- gcc/wide-int.cc.jj 2023-10-06 12:31:56.841517949 +0200 +++ gcc/wide-int.cc 2023-10-06 17:21:59.930022075 +0200 @@ -2406,6 +2406,17 @@ debug (const widest_int *ptr) fprintf (stderr, "\n"); } +bool wide_int_bitint_seen = false; + +void +wide_int_log (const char *p, int n) +{ + extern const char *current_function_name (void); + FILE *f = fopen ("/tmp/wis", "a"); + fprintf (f, "%d %s %s %s %d %c\n", (int) BITS_PER_WORD, main_input_filename ? main_input_filename : "-", current_function_name (), p, n, wide_int_bitint_seen ? 'y' : 'n'); + fclose (f); +} + #if CHECKING_P namespace selftest { --- gcc/tree.cc.jj 2023-10-05 11:36:54.618251787 +0200 +++ gcc/tree.cc 2023-10-06 17:23:07.321118844 +0200 @@ -7178,6 +7178,8 @@ tree build_bitint_type (unsigned HOST_WIDE_INT precision, int unsignedp) { tree itype, ret; +extern bool wide_int_bitint_seen; +if (precision > 128) wide_int_bitint_seen = true; gcc_checking_assert (precision >= 1 + !unsignedp); --- gcc/wide-int.h.jj 2023-10-06 13:12:05.720338130 +0200 +++ gcc/wide-int.h 2023-10-06 17:42:59.980139497 +0200 @@ -1206,7 +1206,11 @@ inline wide_int_storage::wide_int_storag WIDE_INT_REF_FOR (T) xi (x); precision = xi.precision; if (UNLIKELY (precision > WIDE_INT_MAX_INL_PRECISION)) +{ +extern void wide_int_log (const char *, int); +wide_int_log ("ctor", precision); u.valp = XNEWVEC (HOST_WIDE_INT, CEIL (precision, HOST_BITS_PER_WIDE_INT)); +} wi::copy (*this, xi); } @@ -1216,6 +1220,8 @@ inline wide_int_storage::wide_int_storag precision = x.precision; if (UNLIKELY (precision > WIDE_INT_MAX_INL_PRECISION)) { +extern void wide_int_log (const char *, int); +wide_int_log ("copy ctor", precision); u.valp = XNEWVEC (HOST_WIDE_INT, CEIL (precision, HOST_BITS_PER_WIDE_INT)); memcpy (u.valp, x.u.valp, len * sizeof (HOST_WIDE_INT)); } @@ -1242,6 +1248,8 @@ wide_int_storage::operator = (const wide precision = x.precision; if (UNLIKELY (precision > WIDE_INT_MAX_INL_PRECISION)) { +extern void wide_int_log (const char *, int); +wide_int_log ("operator=1", precision); u.valp = XNEWVEC (HOST_WIDE_INT, CEIL (precision, HOST_BITS_PER_WIDE_INT)); memcpy (u.valp, x.u.valp, len * sizeof (HOST_WIDE_INT)); } @@ -1265,8 +1273,12 @@ wide_int_storage::operator = (const T &x XDELETEVEC (u.valp); precision = xi.precision; if (UNLIKELY (precision > WIDE_INT_MAX_INL_PRECISION)) +{ +extern void wide_int_log (const char *, int); +wide_int_log ("operator=2", precision); u.valp = XNEWVEC (HOST_WIDE_INT, CEIL (precision, HOST_BITS_PER_WIDE_INT)); +} } wi::copy (*this, xi); return *this; @@ -1339,8 +1351,12 @@ wide_int_storage::create (unsigned int p wide_int x; x.precision = precision; if (UNLIKELY (precision > WIDE_INT_MAX_INL_PRECISION)) +{ +extern void wide_int_log (const char *, int); +wide_int_log ("create", precision); x.u.valp = XNEWVEC (HOST_WIDE_INT, CEIL (precision, HOST_BITS_PER_WIDE_INT)); +} return x; } @@ -1756,6 +1772,8 @@ widest_int_storage ::widest_int_stora len = x.len; if (UNLIKELY (len > N / HOST_BITS_PER_WIDE_INT)) { +extern void wide_int_log (const char *, int); +wide_int_log ("wi copy ctor", len); u.valp = XNEWVEC (HOST_WIDE_INT, len); memcpy (u.valp, x.u.valp, len * sizeof (HOST_WIDE_INT)); } @@ -1783,6 +1801,8 @@ widest_int_storage ::operator = (cons len = x.len; if (UNLIKELY (len > N / HOST_BITS_PER_WIDE_INT)) { +extern void wide_int_log (const char *, int); +wide_int_log ("wi operator=1", len); u.valp = XNEWVEC (HOST_WIDE_INT, len); memcpy (u.valp, x.u.valp, len * sizeof (HOST_WIDE_INT)); } @@ -1837,6 +1857,8 @@ widest_int_storage ::write_val (unsig len = l; if (UNLIKELY (l > N / HOST_BITS_PER_WIDE_INT)) { +extern void wide_int_log (const char *, int); +wide_int_log ("wi write_val", l); u.valp = XNEWVEC (HOST_WIDE_INT, l); return u.valp; } --- gcc/tree-streamer-in.cc.jj 2023-07-11 13:40:40.255431914 +0200 +++ gcc/tree-streamer-in.cc 2023-10-09 09:40:15.294088267 +0200 @@ -391,6 +391,11 @@ unpack_ts_type_common_value_fields (stru if (RECORD_OR_UNION_TYPE_P (expr)) TYPE_INCLUDES_FLEXARRAY (expr) = (unsigned) bp_unpack_value (bp, 1); TYPE_PRECISION_RAW (expr) = bp_unpack_var_len_unsigned (bp); +if (TREE_CODE (expr) == BITINT_TYPE && TYPE_PRECISION (expr) > 128) +{ +extern bool wide_int_bitint_seen; +wide_int_bitint_seen = true; +} SET_TYPE_ALIGN (expr, bp_unpack_var_len_unsigned (bp)); #ifdef ACCEL_COMPILER if (TYPE_ALIGN (expr) > targetm.absolute_biggest_alignment) --- gcc/c-family/c-lex.cc.jj 2023-10-08 16:37:31.301279702 +0200 +++ gcc/c-family/c-lex.cc 2023-10-09 12:11:31.673687246 +0200 @@ -885,6 +885,8 @@ interpret_integer (const cpp_token *toke { unsigned char buf[3 + max_buf]; memcpy (buf, token->val.str.text, prefix_len); +extern bool wide_int_bitint_seen; +wide_int_bitint_seen = true; wval = 0U; const unsigned char *p = token->val.str.text + prefix_len; cpp_token tok = *token; --SaffQZ6I7XRJud0v--