On Tue, Mar 17, 2020 at 10:37 AM Jakub Jelinek via Gcc-patches
<gcc-patches@gcc.gnu.org> wrote:
>
> Hi!
>
> In the r10-7197-gbae7b38cf8a21e068ad5c0bab089dedb78af3346 commit I've
> noticed duplicated word in a message, which lead me to grep for those and
> we have a tons of them.
> I've used
> grep -v 'long long\|optab optab\|template template\|double double' *.[chS]
> */*.[chS] *.def config/*/* 2>/dev/null | grep ' \([a-zA-Z]\+\) \1 '
> Note, the command will not detect the doubled words at the start or end of
> line or when one of the words is at the end of line and the next one at the
> start of another one.
> Some of it is fairly obvious, e.g. all the "the the" cases which is
> something I've posted and committed patch for already e.g. in 2016,
> other cases are often valid, e.g. "that that" seems to look mostly ok to me.
> Some cases are quite hard to figure out, I've left out some of them from the
> patch (e.g. "and and" in some cases isn't talking about bitwise/logical and
> and so looks incorrect, but in other cases it is talking about those
> operations).
> In most cases the right solution seems to be to remove one of the duplicated
> words, but not always.
>
> I think most important are the ones with user visible messages (in the patch
> 3 of the first 4 hunks), the rest is just comments (and internal
> documentation; for that see the doc/tm.texi changes).
>
> Shall I split the patch somehow for easier review, or is it ok as is?
> Bootstrapped/regtested on x86_64-linux and i686-linux, ok for trunk?
OK.
> 2020-03-17 Jakub Jelinek <ja...@redhat.com>
>
> * lra-spills.c (remove_pseudos): Fix up duplicated word issue in
> a dump message.
> * tree-sra.c (create_access_replacement): Fix up duplicated word issue
> in a comment.
> * read-rtl-function.c (find_param_by_name,
> function_reader::parse_enum_value, function_reader::get_insn_by_uid):
> Likewise.
> * spellcheck.c (get_edit_distance_cutoff): Likewise.
> * tree-data-ref.c (create_ifn_alias_checks): Likewise.
> * tree.def (SWITCH_EXPR): Likewise.
> * selftest.c (assert_str_contains): Likewise.
> * ipa-param-manipulation.h (class ipa_param_body_adjustments):
> Likewise.
> * tree-ssa-math-opts.c (convert_expand_mult_copysign): Likewise.
> * tree-ssa-loop-split.c (find_vdef_in_loop): Likewise.
> * langhooks.h (struct lang_hooks_for_decls): Likewise.
> * ipa-prop.h (struct ipa_param_descriptor): Likewise.
> * tree-ssa-strlen.c (handle_builtin_string_cmp, handle_store):
> Likewise.
> * tree-ssa-dom.c (simplify_stmt_for_jump_threading): Likewise.
> * tree-ssa-reassoc.c (reassociate_bb): Likewise.
> * tree.c (component_ref_size): Likewise.
> * hsa-common.c (hsa_init_compilation_unit_data): Likewise.
> * gimple-ssa-sprintf.c (get_string_length, format_string,
> format_directive): Likewise.
> * omp-grid.c (grid_process_kernel_body_copy): Likewise.
> * input.c (string_concat_db::get_string_concatenation,
> test_lexer_string_locations_ucn4): Likewise.
> * cfgexpand.c (pass_expand::execute): Likewise.
> * gimple-ssa-warn-restrict.c (builtin_memref::offset_out_of_bounds,
> maybe_diag_overlap): Likewise.
> * rtl.c (RTX_CODE_HWINT_P_1): Likewise.
> * shrink-wrap.c (spread_components): Likewise.
> * tree-ssa-dse.c (initialize_ao_ref_for_dse, valid_ao_ref_for_dse):
> Likewise.
> * tree-call-cdce.c (shrink_wrap_one_built_in_call_with_conds):
> Likewise.
> * dwarf2out.c (dwarf2out_early_finish): Likewise.
> * gimple-ssa-store-merging.c: Likewise.
> * ira-costs.c (record_operand_costs): Likewise.
> * tree-vect-loop.c (vectorizable_reduction): Likewise.
> * target.def (dispatch): Likewise.
> (validate_dims, gen_ccmp_first): Fix up duplicated word issue
> in documentation text.
> * doc/tm.texi: Regenerated.
> * config/i386/x86-tune.def (X86_TUNE_PARTIAL_FLAG_REG_STALL): Fix up
> duplicated word issue in a comment.
> * config/i386/i386.c (ix86_test_loading_unspec): Likewise.
> * config/i386/i386-features.c (remove_partial_avx_dependency):
> Likewise.
> * config/msp430/msp430.c (msp430_select_section): Likewise.
> * config/gcn/gcn-run.c (load_image): Likewise.
> * config/aarch64/aarch64-sve.md (sve_ld1r<mode>): Likewise.
> * config/aarch64/aarch64.c (aarch64_gen_adjusted_ldpstp): Likewise.
> * config/aarch64/falkor-tag-collision-avoidance.c
> (single_dest_per_chain): Likewise.
> * config/nvptx/nvptx.c (nvptx_record_fndecl): Likewise.
> * config/fr30/fr30.c (fr30_arg_partial_bytes): Likewise.
> * config/rs6000/rs6000-string.c (expand_cmp_vec_sequence): Likewise.
> * config/rs6000/rs6000-p8swap.c (replace_swapped_load_constant):
> Likewise.
> * config/rs6000/rs6000-c.c (rs6000_target_modify_macros): Likewise.
> * config/rs6000/rs6000.c (rs6000_option_override_internal): Likewise.
> * config/rs6000/rs6000-logue.c
> (rs6000_emit_probe_stack_range_stack_clash): Likewise.
> * config/nds32/nds32-md-auxiliary.c (nds32_split_ashiftdi3): Likewise.
> Fix various other issues in the comment.
> c-family/
> * c-common.c (resolve_overloaded_builtin): Fix up duplicated word
> issue in a diagnostic message.
> cp/
> * pt.c (tsubst): Fix up duplicated word issue in a diagnostic message.
> (lookup_template_class_1, tsubst_expr): Fix up duplicated word issue
> in a comment.
> * parser.c (cp_parser_statement, cp_parser_linkage_specification,
> cp_parser_placeholder_type_specifier,
> cp_parser_constraint_requires_parens): Likewise.
> * name-lookup.c (suggest_alternative_in_explicit_scope): Likewise.
> fortran/
> * array.c (gfc_check_iter_variable): Fix up duplicated word issue
> in a comment.
> * arith.c (gfc_arith_concat): Likewise.
> * resolve.c (gfc_resolve_ref): Likewise.
> * frontend-passes.c (matmul_lhs_realloc): Likewise.
> * module.c (gfc_match_submodule, load_needed): Likewise.
> * trans-expr.c (gfc_init_se): Likewise.
>
> --- gcc/lra-spills.c.jj 2020-03-16 22:56:55.557043260 +0100
> +++ gcc/lra-spills.c 2020-03-16 23:14:52.458127503 +0100
> @@ -434,7 +434,7 @@ remove_pseudos (rtx *loc, rtx_insn *insn
> lra_get_insn_recog_data (insn)->used_insn_alternative = -1;
> if (lra_dump_file != NULL)
> fprintf (lra_dump_file,
> - "Memory subreg was simplified in in insn #%u\n",
> + "Memory subreg was simplified in insn #%u\n",
> INSN_UID (insn));
> }
> }
> --- gcc/c-family/c-common.c.jj 2020-03-05 07:57:41.650447053 +0100
> +++ gcc/c-family/c-common.c 2020-03-16 23:28:25.560109119 +0100
> @@ -7427,7 +7427,7 @@ resolve_overloaded_builtin (location_t l
> warning_at (input_location, 0,
> "this target does not define a speculation barrier;
> "
> "your program will still execute correctly, "
> - "but incorrect speculation may not be be "
> + "but incorrect speculation may not be "
> "restricted");
>
> /* If the optional second argument is present, handle any side
> --- gcc/cp/pt.c.jj 2020-03-15 01:25:58.369049305 +0100
> +++ gcc/cp/pt.c 2020-03-16 23:32:10.737777110 +0100
> @@ -9682,7 +9682,7 @@ lookup_template_class_1 (tree d1, tree a
> if (entry)
> return entry->spec;
>
> - /* If the the template's constraints are not satisfied,
> + /* If the template's constraints are not satisfied,
> then we cannot form a valid type.
>
> Note that the check is deferred until after the hash
> @@ -15688,7 +15688,7 @@ tsubst (tree t, tree args, tsubst_flags_
> else if (TYPENAME_IS_CLASS_P (t) && !CLASS_TYPE_P (f))
> {
> if (complain & tf_error)
> - error ("%qT resolves to %qT, which is is not a class type",
> + error ("%qT resolves to %qT, which is not a class type",
> t, f);
> else
> return error_mark_node;
> @@ -17855,7 +17855,7 @@ tsubst_expr (tree t, tree args, tsubst_f
> case RANGE_FOR_STMT:
> {
> /* Construct another range_for, if this is not a final
> - substitution (for inside inside a generic lambda of a
> + substitution (for inside a generic lambda of a
> template). Otherwise convert to a regular for. */
> tree decl, expr;
> stmt = (processing_template_decl
> --- gcc/tree-sra.c.jj 2020-02-21 21:54:56.752346267 +0100
> +++ gcc/tree-sra.c 2020-03-16 23:21:54.510889207 +0100
> @@ -2142,7 +2142,7 @@ sort_and_splice_var_accesses (tree var)
> /* Create a variable for the given ACCESS which determines the type, name
> and a
> few other properties. Return the variable declaration and store it also
> to
> ACCESS->replacement. REG_TREE is used when creating a declaration to
> base a
> - default-definition SSA name on on in order to facilitate an uninitialized
> + default-definition SSA name on in order to facilitate an uninitialized
> warning. It is used instead of the actual ACCESS type if that is not of a
> gimple register type. */
>
> --- gcc/read-rtl-function.c.jj 2020-01-12 11:54:36.917405805 +0100
> +++ gcc/read-rtl-function.c 2020-03-16 23:16:11.592957825 +0100
> @@ -532,7 +532,7 @@ function_reader::create_function ()
>
> }
>
> -/* Look within the the params of FNDECL for a param named NAME.
> +/* Look within the params of FNDECL for a param named NAME.
> Return NULL_TREE if one isn't found. */
>
> static tree
> @@ -969,7 +969,7 @@ function_reader::read_rtx_operand_u (rtx
>
> /* Read a name, looking for a match against a string found in array
> STRINGS of size NUM_VALUES.
> - Return the index of the the matched string, or emit an error. */
> + Return the index of the matched string, or emit an error. */
>
> int
> function_reader::parse_enum_value (int num_values, const char *const
> *strings)
> @@ -1611,7 +1611,7 @@ function_reader::apply_fixups ()
> }
>
> /* Given a UID value, try to locate a pointer to the corresponding
> - rtx_insn *, or NULL if if can't be found. */
> + rtx_insn *, or NULL if it can't be found. */
>
> rtx_insn **
> function_reader::get_insn_by_uid (int uid)
> --- gcc/spellcheck.c.jj 2020-01-12 11:54:36.933405563 +0100
> +++ gcc/spellcheck.c 2020-03-16 23:19:21.261154369 +0100
> @@ -166,7 +166,7 @@ find_closest_string (const char *target,
> to be meaningful, given a goal of length GOAL_LEN and a candidate of
> length CANDIDATE_LEN.
>
> - This is a third of the the length of the candidate or of the goal,
> + This is a third of the length of the candidate or of the goal,
> whichever is bigger. */
>
> edit_distance_t
> --- gcc/tree-data-ref.c.jj 2020-01-12 11:54:38.493382027 +0100
> +++ gcc/tree-data-ref.c 2020-03-16 23:21:08.920563064 +0100
> @@ -1751,7 +1751,7 @@ create_ifn_alias_checks (tree *cond_expr
> return false;
>
> /* Make sure that both DRs access the same pattern of bytes,
> - with a constant length and and step. */
> + with a constant length and step. */
> poly_uint64 seg_len;
> if (!operand_equal_p (dr_a.seg_len, dr_b.seg_len, 0)
> || !poly_int_tree_p (dr_a.seg_len, &seg_len)
> --- gcc/tree.def.jj 2020-01-12 11:54:38.527381514 +0100
> +++ gcc/tree.def 2020-03-16 23:40:07.734718877 +0100
> @@ -970,7 +970,7 @@ DEFTREECODE (SWITCH_EXPR, "switch_expr",
> is a 'default' label.
> Operand 1 is CASE_HIGH. If it is NULL_TREE, the label is a simple
> (one-value) case label. If it is non-NULL_TREE, the case is a range.
> - Operand 2 is CASE_LABEL, which is is the corresponding LABEL_DECL.
> + Operand 2 is CASE_LABEL, which has the corresponding LABEL_DECL.
> Operand 3 is CASE_CHAIN. This operand is only used in tree-cfg.c to
> speed up the lookup of case labels which use a particular edge in
> the control flow graph. */
> --- gcc/cp/parser.c.jj 2020-03-16 09:01:27.723030059 +0100
> +++ gcc/cp/parser.c 2020-03-16 23:30:28.443290785 +0100
> @@ -11346,8 +11346,8 @@ cp_parser_statement (cp_parser* parser,
> /* This must be a namespace alias definition. */
> if (std_attrs != NULL_TREE)
> {
> - /* Attributes should be parsed as part of the the
> - declaration, so let's un-parse them. */
> + /* Attributes should be parsed as part of the
> + declaration, so let's un-parse them. */
> saved_tokens.rollback();
> std_attrs = NULL_TREE;
> }
> @@ -14554,7 +14554,7 @@ cp_parser_linkage_specification (cp_pars
> /* We're now using the new linkage. */
> push_lang_context (linkage);
>
> - /* Preserve the location of the the innermost linkage specification,
> + /* Preserve the location of the innermost linkage specification,
> tracking the locations of nested specifications via a local. */
> location_t saved_location
> = parser->innermost_linkage_specification_location;
> @@ -18316,7 +18316,7 @@ cp_parser_placeholder_type_specifier (cp
> }
>
> /* A type constraint constrains a contextually determined type or type
> - parameter pack. However, the the Concepts TS does allow concepts
> + parameter pack. However, the Concepts TS does allow concepts
> to introduce non-type and template template parameters. */
> if (TREE_CODE (proto) != TYPE_DECL)
> {
> @@ -27339,7 +27339,7 @@ cp_parser_constraint_requires_parens (cp
>
> case CPP_EQ:
> {
> - /* An equal sign may be part of the the definition of a function,
> + /* An equal sign may be part of the definition of a function,
> and not an assignment operator, when parsing the expression
> for a trailing requires-clause. For example:
>
> --- gcc/cp/name-lookup.c.jj 2020-02-14 00:03:21.022222232 +0100
> +++ gcc/cp/name-lookup.c 2020-03-16 23:29:19.366312932 +0100
> @@ -5938,7 +5938,7 @@ maybe_suggest_missing_header (location_t
> /* Generate a name_hint at LOCATION for NAME, an IDENTIFIER_NODE for which
> name
> lookup failed within the explicitly provided SCOPE.
>
> - Suggest the the best meaningful candidates (if any), otherwise
> + Suggest the best meaningful candidates (if any), otherwise
> an empty name_hint is returned. */
>
> name_hint
> --- gcc/selftest.c.jj 2020-01-12 11:54:36.931405594 +0100
> +++ gcc/selftest.c 2020-03-16 23:18:25.069984922 +0100
> @@ -95,7 +95,7 @@ assert_streq (const location &loc,
> }
>
> /* Implementation detail of ASSERT_STR_CONTAINS.
> - Use strstr to determine if val_needle is is within val_haystack.
> + Use strstr to determine if val_needle is within val_haystack.
> ::selftest::pass if it is found.
> ::selftest::fail if it is not found. */
>
> --- gcc/ipa-param-manipulation.h.jj 2020-01-12 11:54:36.653409788 +0100
> +++ gcc/ipa-param-manipulation.h 2020-03-16 23:11:09.232426980 +0100
> @@ -406,7 +406,7 @@ private:
>
> auto_vec<tree, 16> m_new_types;
>
> - /* Vector of structures telling how to replace old parameters in in the
> + /* Vector of structures telling how to replace old parameters in the
> function body. TODO: Even though there usually be only few, but should
> we
> use a hash? */
>
> --- gcc/tree-ssa-math-opts.c.jj 2020-01-12 11:54:38.508381801 +0100
> +++ gcc/tree-ssa-math-opts.c 2020-03-16 23:23:45.394250256 +0100
> @@ -2477,7 +2477,7 @@ is_copysign_call_with_1 (gimple *call)
> }
>
> /* Try to expand the pattern x * copysign (1, y) into xorsign (x, y).
> - This only happens when the the xorsign optab is defined, if the
> + This only happens when the xorsign optab is defined, if the
> pattern is not a xorsign pattern or if expansion fails FALSE is
> returned, otherwise TRUE is returned. */
> static bool
> --- gcc/tree-ssa-loop-split.c.jj 2020-01-12 11:54:38.507381816 +0100
> +++ gcc/tree-ssa-loop-split.c 2020-03-16 23:23:21.723600127 +0100
> @@ -770,7 +770,7 @@ find_vdef_in_loop (struct loop *loop)
>
> /* Non-pure call statement is conservatively assumed to impact all
> memory locations. So place call statements ahead of other memory
> - stores in the vector with an idea of of using them as shortcut
> + stores in the vector with an idea of using them as shortcut
> terminators to memory alias analysis. */
> if (gimple_code (stmt) == GIMPLE_CALL)
> info->memory_stores.safe_push (stmt);
> --- gcc/langhooks.h.jj 2020-01-12 11:54:36.670409531 +0100
> +++ gcc/langhooks.h 2020-03-16 23:14:00.503895437 +0100
> @@ -227,7 +227,7 @@ struct lang_hooks_for_decls
> bool (*ok_for_sibcall) (const_tree);
>
> /* Return a tree for the actual data of an array descriptor - or NULL_TREE
> - if original tree is not an array descriptor. If the the second argument
> + if original tree is not an array descriptor. If the second argument
> is true, only the TREE_TYPE is returned without generating a new tree.
> */
> tree (*omp_array_data) (tree, bool);
>
> --- gcc/ipa-prop.h.jj 2020-02-25 09:01:01.484608786 +0100
> +++ gcc/ipa-prop.h 2020-03-16 23:12:31.905204999 +0100
> @@ -438,7 +438,7 @@ ipa_get_jf_ancestor_type_preserved (stru
> struct GTY(()) ipa_param_descriptor
> {
> /* In analysis and modification phase, this is the PARAM_DECL of this
> - parameter, in IPA LTO phase, this is the type of the the described
> + parameter, in IPA LTO phase, this is the type of the described
> parameter or NULL if not known. Do not read this field directly but
> through ipa_get_param and ipa_get_type as appropriate. */
> tree decl_or_type;
> --- gcc/tree-ssa-strlen.c.jj 2020-03-16 22:56:46.799172692 +0100
> +++ gcc/tree-ssa-strlen.c 2020-03-16 23:26:57.195415265 +0100
> @@ -4381,7 +4381,7 @@ handle_builtin_string_cmp (gimple_stmt_i
> int idx1 = get_stridx (arg1);
> int idx2 = get_stridx (arg2);
>
> - /* For strncmp set to the the value of the third argument if known. */
> + /* For strncmp set to the value of the third argument if known. */
> HOST_WIDE_INT bound = -1;
> tree len = NULL_TREE;
> /* Extract the strncmp bound. */
> @@ -4937,7 +4937,7 @@ handle_store (gimple_stmt_iterator *gsi,
> tree ssaname = NULL_TREE, lhs = gimple_assign_lhs (stmt);
> tree rhs = gimple_assign_rhs1 (stmt);
>
> - /* The offset of the first byte in LHS modified by the the store. */
> + /* The offset of the first byte in LHS modified by the store. */
> unsigned HOST_WIDE_INT offset = 0;
>
> if (TREE_CODE (lhs) == MEM_REF
> --- gcc/tree-ssa-dom.c.jj 2020-03-03 11:04:46.367821907 +0100
> +++ gcc/tree-ssa-dom.c 2020-03-16 23:22:17.174554220 +0100
> @@ -875,7 +875,7 @@ simplify_stmt_for_jump_threading (gimple
> class avail_exprs_stack *avail_exprs_stack,
> basic_block bb ATTRIBUTE_UNUSED)
> {
> - /* First query our hash table to see if the the expression is available
> + /* First query our hash table to see if the expression is available
> there. A non-NULL return value will be either a constant or another
> SSA_NAME. */
> tree cached_lhs = avail_exprs_stack->lookup_avail_expr (stmt, false,
> true);
> --- gcc/tree-ssa-reassoc.c.jj 2020-03-16 09:01:52.436662813 +0100
> +++ gcc/tree-ssa-reassoc.c 2020-03-16 23:26:00.612251615 +0100
> @@ -6373,7 +6373,7 @@ reassociate_bb (basic_block bb)
> int width;
>
> /* For binary bit operations, if there are at least 3
> - operands and the last last operand in OPS is a constant,
> + operands and the last operand in OPS is a constant,
> move it to the front. This helps ensure that we generate
> (X & Y) & C rather than (X & C) & Y. The former will
> often match a canonical bit test when we get to RTL. */
> --- gcc/tree.c.jj 2020-03-11 13:25:17.852216259 +0100
> +++ gcc/tree.c 2020-03-16 23:20:14.445368262 +0100
> @@ -13663,7 +13663,7 @@ component_ref_size (tree ref, bool *inte
> }
>
> /* BASE is the declared object of which MEMBER is either a member
> - or that is is cast to REFTYPE (e.g., a char buffer used to store
> + or that is cast to REFTYPE (e.g., a char buffer used to store
> a REFTYPE object). */
> tree reftype = TREE_TYPE (TREE_OPERAND (ref, 0));
> tree basetype = TREE_TYPE (base);
> --- gcc/hsa-common.c.jj 2020-01-12 11:54:36.645409908 +0100
> +++ gcc/hsa-common.c 2020-03-16 23:09:15.367109770 +0100
> @@ -95,7 +95,7 @@ hsa_callable_function_p (tree fndecl)
> && !lookup_attribute ("oacc function", DECL_ATTRIBUTES (fndecl)));
> }
>
> -/* Allocate HSA structures that are are used when dealing with different
> +/* Allocate HSA structures that are used when dealing with different
> functions. */
>
> void
> --- gcc/gimple-ssa-sprintf.c.jj 2020-01-12 11:54:36.623410240 +0100
> +++ gcc/gimple-ssa-sprintf.c 2020-03-16 23:06:14.122788343 +0100
> @@ -2098,7 +2098,7 @@ get_string_length (tree str, unsigned el
> if (res.range.max < target_int_max ())
> {
> res.knownrange = true;
> - /* When the the length of the longest string is known and not
> + /* When the length of the longest string is known and not
> excessive use it as the likely length of the string(s). */
> res.range.likely = res.range.max;
> }
> @@ -2478,7 +2478,7 @@ format_string (const directive &dir, tre
> is bounded by MB_LEN_MAX * wcslen (S). */
> res.range.max *= target_mb_len_max ();
> res.range.unlikely = res.range.max;
> - /* It's likely that the the total length is not more that
> + /* It's likely that the total length is not more that
> 2 * wcslen (S).*/
> res.range.likely = res.range.min * 2;
>
> @@ -3337,7 +3337,7 @@ format_directive (const call_info &info,
> }
> else if (!info.is_string_func ())
> {
> - /* If the warning is for a file function function like fprintf
> + /* If the warning is for a file function like fprintf
> of printf with no destination size just print the computed
> result. */
> if (min == max)
> --- gcc/omp-grid.c.jj 2020-01-12 11:54:36.687409275 +0100
> +++ gcc/omp-grid.c 2020-03-16 23:15:17.390758982 +0100
> @@ -1186,7 +1186,7 @@ grid_mark_tiling_parallels_and_loops (gi
> /* Given freshly copied top level kernel SEQ, identify the individual OMP
> components, mark them as part of kernel, copy assignment leading to them
> just before DST, remapping them using WI and adding new temporaries to
> - TGT_BIND, and and return the loop that will be used for kernel dispatch.
> */
> + TGT_BIND, and return the loop that will be used for kernel dispatch. */
>
> static gomp_for *
> grid_process_kernel_body_copy (grid_prop *grid, gimple_seq seq,
> --- gcc/input.c.jj 2020-01-12 11:54:36.648409863 +0100
> +++ gcc/input.c 2020-03-16 23:10:20.585145928 +0100
> @@ -1296,7 +1296,7 @@ string_concat_db::record_string_concaten
> m_table->put (key_loc, concat);
> }
>
> -/* Determine if LOC was the location of the the initial token of a
> +/* Determine if LOC was the location of the initial token of a
> concatenation of string literal tokens.
> If so, *OUT_NUM is written to with the number of tokens, and
> *OUT_LOCS with the location of an array of locations of the
> @@ -2701,7 +2701,7 @@ test_lexer_string_locations_ucn4 (const
>
> /* Verify that cpp_interpret_string works.
> The string should be encoded in the execution character
> - set. Assuming that that is UTF-8, we should have the following:
> + set. Assuming that is UTF-8, we should have the following:
> ----------- ---- ----- ------- ----------------
> Byte offset Byte Octal Unicode Source Column(s)
> ----------- ---- ----- ------- ----------------
> --- gcc/cfgexpand.c.jj 2020-02-24 15:22:17.303787284 +0100
> +++ gcc/cfgexpand.c 2020-03-16 23:03:24.427296232 +0100
> @@ -6656,7 +6656,7 @@ pass_expand::execute (function *fun)
> if (crtl->tail_call_emit)
> fixup_tail_calls ();
>
> - /* BB subdivision may have created basic blocks that are are only reachable
> + /* BB subdivision may have created basic blocks that are only reachable
> from unlikely bbs but not marked as such in the profile. */
> if (optimize)
> propagate_unlikely_bbs_forward ();
> --- gcc/gimple-ssa-warn-restrict.c.jj 2020-03-06 11:35:46.287074813 +0100
> +++ gcc/gimple-ssa-warn-restrict.c 2020-03-16 23:08:59.298347249 +0100
> @@ -577,7 +577,7 @@ builtin_memref::offset_out_of_bounds (in
> bool hib = wi::les_p (offrng[0], offrng[1]);
> bool lob = !hib;
>
> - /* Set to the size remaining in the object object after subtracting
> + /* Set to the size remaining in the object after subtracting
> REFOFF. It may become negative as a result of negative indices
> into the enclosing object, such as in:
> extern struct S { char a[4], b[3], c[1]; } *p;
> @@ -1430,7 +1430,7 @@ builtin_access::overlap ()
> }
>
> /* Attempt to detect and diagnose an overlapping copy in a call expression
> - EXPR involving an an access ACS to a built-in memory or string function.
> + EXPR involving an access ACS to a built-in memory or string function.
> Return true when one has been detected, false otherwise. */
>
> static bool
> --- gcc/rtl.c.jj 2020-01-12 11:54:36.924405699 +0100
> +++ gcc/rtl.c 2020-03-16 23:17:55.958415219 +0100
> @@ -106,7 +106,7 @@ const enum rtx_class rtx_class[NUM_RTX_C
> #undef DEF_RTL_EXPR
> };
>
> -/* Whether rtxs with the given code code store data in the hwint field. */
> +/* Whether rtxs with the given code store data in the hwint field. */
>
> #define RTX_CODE_HWINT_P_1(ENUM) \
> ((ENUM) == CONST_INT || (ENUM) == CONST_DOUBLE \
> --- gcc/shrink-wrap.c.jj 2020-01-12 11:54:36.932405578 +0100
> +++ gcc/shrink-wrap.c 2020-03-16 23:18:51.106600073 +0100
> @@ -1380,7 +1380,7 @@ spread_components (sbitmap components)
>
> todo.release ();
>
> - /* Finally, mark everything not not needed both forwards and backwards. */
> + /* Finally, mark everything not needed both forwards and backwards. */
>
> bool did_changes = false;
>
> --- gcc/fortran/array.c.jj 2020-02-10 22:43:56.574481441 +0100
> +++ gcc/fortran/array.c 2020-03-16 23:34:09.482020015 +0100
> @@ -1475,7 +1475,7 @@ static cons_stack *base;
> static bool check_constructor (gfc_constructor_base, bool (*) (gfc_expr *));
>
> /* Check an EXPR_VARIABLE expression in a constructor to make sure
> - that that variable is an iteration variables. */
> + that that variable is an iteration variable. */
>
> bool
> gfc_check_iter_variable (gfc_expr *expr)
> --- gcc/fortran/arith.c.jj 2020-01-12 11:54:36.582410859 +0100
> +++ gcc/fortran/arith.c 2020-03-16 23:33:25.265674305 +0100
> @@ -994,7 +994,7 @@ gfc_arith_concat (gfc_expr *op1, gfc_exp
> gfc_expr *result;
> size_t len;
>
> - /* By cleverly playing around with constructors, is is possible
> + /* By cleverly playing around with constructors, it is possible
> to get mismaching types here. */
> if (op1->ts.type != BT_CHARACTER || op2->ts.type != BT_CHARACTER
> || op1->ts.kind != op2->ts.kind)
> --- gcc/fortran/resolve.c.jj 2020-03-09 12:43:00.858036331 +0100
> +++ gcc/fortran/resolve.c 2020-03-16 23:35:40.031680138 +0100
> @@ -5318,7 +5318,7 @@ gfc_resolve_ref (gfc_expr *expr)
> {
> array_ref->u.ar.type = AR_ELEMENT;
> expr->rank = 0;
> - /* INQUIRY_LEN is not evaluated from the the rest of the expr
> + /* INQUIRY_LEN is not evaluated from the rest of the expr
> but directly from the string length. This means that setting
> the array indices to one does not matter but might trigger
> a runtime bounds error. Suppress the check. */
> --- gcc/fortran/frontend-passes.c.jj 2020-01-31 19:18:02.419904105 +0100
> +++ gcc/fortran/frontend-passes.c 2020-03-16 23:34:43.627514760 +0100
> @@ -3190,7 +3190,7 @@ matmul_lhs_realloc (gfc_expr *c, gfc_exp
> gcc_assert (ar && ar->type == AR_FULL);
>
> /* c comes in as a full ref. Change it into a copy and make it into an
> - element ref so it has the right form for for ALLOCATE. In the same
> + element ref so it has the right form for ALLOCATE. In the same
> switch statement, also generate the size comparison for the secod IF
> statement. */
>
> --- gcc/fortran/module.c.jj 2020-03-03 07:57:42.422826729 +0100
> +++ gcc/fortran/module.c 2020-03-16 23:35:22.646937383 +0100
> @@ -743,7 +743,7 @@ cleanup:
> ordered pair whose first element is the ancestor module name and
> whose second element is the submodule name. 'Submodule_name' is
> used for the submodule filename and uses '@' as a separator, whilst
> - the name of the symbol for the module uses '.' as a a separator.
> + the name of the symbol for the module uses '.' as a separator.
> The reasons for these choices are:
> (i) To follow another leading brand in the submodule filenames;
> (ii) Since '.' is not particularly visible in the filenames; and
> @@ -5044,7 +5044,7 @@ load_needed (pointer_info *p)
> sym->attr.use_assoc = 1;
>
> /* Unliked derived types, a STRUCTURE may share names with other symbols.
> - We greedily converted the the symbol name to lowercase before we knew
> its
> + We greedily converted the symbol name to lowercase before we knew its
> type, so now we must fix it. */
> if (sym->attr.flavor == FL_STRUCT)
> sym->name = gfc_dt_upper_string (sym->name);
> --- gcc/fortran/trans-expr.c.jj 2020-03-02 13:33:10.911495134 +0100
> +++ gcc/fortran/trans-expr.c 2020-03-16 23:36:14.245173870 +0100
> @@ -1639,7 +1639,7 @@ gfc_copy_se_loopvars (gfc_se * dest, gfc
>
> Care must be taken when multiple se are created with the same parent.
> The child se must be kept in sync. The easiest way is to delay creation
> - of a child se until after after the previous se has been translated. */
> + of a child se until after the previous se has been translated. */
>
> void
> gfc_init_se (gfc_se * se, gfc_se * parent)
> --- gcc/tree-ssa-dse.c.jj 2020-03-12 09:33:28.981544120 +0100
> +++ gcc/tree-ssa-dse.c 2020-03-16 23:22:56.060979442 +0100
> @@ -88,7 +88,7 @@ static bitmap need_eh_cleanup;
> /* STMT is a statement that may write into memory. Analyze it and
> initialize WRITE to describe how STMT affects memory.
>
> - Return TRUE if the the statement was analyzed, FALSE otherwise.
> + Return TRUE if the statement was analyzed, FALSE otherwise.
>
> It is always safe to return FALSE. But typically better optimziation
> can be achieved by analyzing more statements. */
> @@ -147,7 +147,7 @@ initialize_ao_ref_for_dse (gimple *stmt,
> return false;
> }
>
> -/* Given REF from the the alias oracle, return TRUE if it is a valid
> +/* Given REF from the alias oracle, return TRUE if it is a valid
> memory reference for dead store elimination, false otherwise.
>
> In particular, the reference must have a known base, known maximum
> --- gcc/tree-call-cdce.c.jj 2020-01-12 11:54:38.490382073 +0100
> +++ gcc/tree-call-cdce.c 2020-03-16 23:20:47.849874512 +0100
> @@ -832,7 +832,7 @@ shrink_wrap_one_built_in_call_with_conds
> 4. [guard m]: [guard m+1] for 0 <= m <= n-2
> 5. [join]: [guard n-1]
>
> - We punt for the more complex case case of [join] being old and
> + We punt for the more complex case of [join] being old and
> simply free the dominance info. We also punt on postdominators,
> which aren't expected to be available at this point anyway. */
> bi_call_bb = gimple_bb (bi_call);
> --- gcc/dwarf2out.c.jj 2020-02-28 09:33:40.641403823 +0100
> +++ gcc/dwarf2out.c 2020-03-16 23:04:46.253086946 +0100
> @@ -32152,7 +32152,7 @@ dwarf2out_early_finish (const char *file
> location related output removed and some LTO specific changes.
> Some refactoring might make both smaller and easier to match up. */
>
> - /* Traverse the DIE's and add add sibling attributes to those DIE's
> + /* Traverse the DIE's and add sibling attributes to those DIE's
> that have children. */
> add_sibling_attributes (comp_unit_die ());
> for (limbo_die_node *node = limbo_die_list; node; node = node->next)
> --- gcc/gimple-ssa-store-merging.c.jj 2020-03-03 19:47:57.314470589 +0100
> +++ gcc/gimple-ssa-store-merging.c 2020-03-16 23:06:37.057449396 +0100
> @@ -61,7 +61,7 @@
> record the surrounding bit region, i.e. bits that could be stored in
> a read-modify-write operation when storing the bit-field. Record store
> chains to different bases in a hash_map (m_stores) and make sure to
> - terminate such chains when appropriate (for example when when the stored
> + terminate such chains when appropriate (for example when the stored
> values get used subsequently).
> These stores can be a result of structure element initializers, array
> stores
> etc. A store_immediate_info object is recorded for every such store.
> --- gcc/ira-costs.c.jj 2020-01-12 11:54:36.658409712 +0100
> +++ gcc/ira-costs.c 2020-03-16 23:13:38.714217509 +0100
> @@ -1319,7 +1319,7 @@ record_operand_costs (rtx_insn *insn, en
> hard_reg_class = REGNO_REG_CLASS (other_regno);
> bigger_hard_reg_class =
> ira_pressure_class_translate[hard_reg_class];
> /* Target code may return any cost for mode which does not
> - fit the the hard reg class (e.g. DImode for AREG on
> + fit the hard reg class (e.g. DImode for AREG on
> i386). Check this and use a bigger class to get the
> right cost. */
> if (bigger_hard_reg_class != NO_REGS
> --- gcc/target.def.jj 2020-01-24 22:34:36.097644949 +0100
> +++ gcc/target.def 2020-03-16 23:37:38.791922814 +0100
> @@ -1519,7 +1519,7 @@ in its second parameter.",
> void, (rtx_insn *insn, int x),
> hook_void_rtx_insn_int)
>
> -/* The following member value is a a function that returns true is
> +/* The following member value is a function that returns true is
> dispatch schedling is supported in hardware and condition passed
> as the second parameter is true. */
> DEFHOOK
> @@ -1694,7 +1694,7 @@ DEFHOOK
> "This hook should check the launch dimensions provided for an OpenACC\n\
> compute region, or routine. Defaulted values are represented as -1\n\
> and non-constant values as 0. The @var{fn_level} is negative for the\n\
> -function corresponding to the compute region. For a routine is is the\n\
> +function corresponding to the compute region. For a routine it is the\n\
> outermost level at which partitioned execution may be spawned. The hook\n\
> should verify non-default values. If DECL is NULL, global defaults\n\
> are being validated and unspecified defaults should be filled in.\n\
> @@ -2675,7 +2675,7 @@ DEFHOOK
> with @code{CC} for passing to @code{gen_ccmp_next} or
> @code{cbranch_optab}.\n\
> The insns to prepare the compare are saved in @var{prep_seq} and the
> compare\n\
> insns are saved in @var{gen_seq}. They will be emitted when all the\n\
> - compares in the the conditional comparision are generated without error.\n\
> + compares in the conditional comparision are generated without error.\n\
> @var{code} is the @code{rtx_code} of the compare for @var{op0} and
> @var{op1}.",
> rtx, (rtx_insn **prep_seq, rtx_insn **gen_seq, int code, tree op0, tree
> op1),
> NULL)
> --- gcc/tree-vect-loop.c.jj 2020-01-28 14:08:08.126108002 +0100
> +++ gcc/tree-vect-loop.c 2020-03-16 23:27:28.831947649 +0100
> @@ -6541,7 +6541,7 @@ vectorizable_reduction (stmt_vec_info st
> }
> /* The epilogue code relies on the number of elements being a multiple
> of the group size. The duplicate-and-interleave approach to setting
> - up the the initial vector does too. */
> + up the initial vector does too. */
> if (!multiple_p (nunits_out, group_size))
> {
> if (dump_enabled_p ())
> --- gcc/doc/tm.texi.jj 2020-03-12 09:35:12.413024024 +0100
> +++ gcc/doc/tm.texi 2020-03-16 23:38:31.310145690 +0100
> @@ -6157,7 +6157,7 @@ but might be present in another OpenMP c
> This hook should check the launch dimensions provided for an OpenACC
> compute region, or routine. Defaulted values are represented as -1
> and non-constant values as 0. The @var{fn_level} is negative for the
> -function corresponding to the compute region. For a routine is is the
> +function corresponding to the compute region. For a routine it is the
> outermost level at which partitioned execution may be spawned. The hook
> should verify non-default values. If DECL is NULL, global defaults
> are being validated and unspecified defaults should be filled in.
> @@ -11858,7 +11858,7 @@ This function prepares to emit a compari
> with @code{CC} for passing to @code{gen_ccmp_next} or @code{cbranch_optab}.
> The insns to prepare the compare are saved in @var{prep_seq} and the compare
> insns are saved in @var{gen_seq}. They will be emitted when all the
> - compares in the the conditional comparision are generated without error.
> + compares in the conditional comparision are generated without error.
> @var{code} is the @code{rtx_code} of the compare for @var{op0} and
> @var{op1}.
> @end deftypefn
>
> --- gcc/config/i386/x86-tune.def.jj 2020-01-31 19:18:02.396904445 +0100
> +++ gcc/config/i386/x86-tune.def 2020-03-17 09:13:58.411694557 +0100
> @@ -70,7 +70,7 @@ DEF_TUNE (X86_TUNE_SSE_PARTIAL_REG_DEPEN
> upper part undefined. */
> DEF_TUNE (X86_TUNE_SSE_SPLIT_REGS, "sse_split_regs", m_ATHLON_K8)
>
> -/* X86_TUNE_PARTIAL_FLAG_REG_STALL: this flag disables use of of flags
> +/* X86_TUNE_PARTIAL_FLAG_REG_STALL: this flag disables use of flags
> set by instructions affecting just some flags (in particular shifts).
> This is because Core2 resolves dependencies on whole flags register
> and such sequences introduce false dependency on previous instruction
> --- gcc/config/i386/i386.c.jj 2020-03-15 22:35:30.342490275 +0100
> +++ gcc/config/i386/i386.c 2020-03-17 09:12:10.761275054 +0100
> @@ -22928,7 +22928,7 @@ ix86_test_loading_unspec ()
> /* Verify that the two mems are thus treated as equal. */
> ASSERT_TRUE (rtx_equal_p (dst, v0));
>
> - /* Verify the the insn is recognized. */
> + /* Verify that the insn is recognized. */
> ASSERT_NE(-1, recog_memoized (insn));
>
> /* Test of an UNSPEC_VOLATILE, which has its own enum values. */
> --- gcc/config/i386/i386-features.c.jj 2020-02-05 15:38:06.594293102 +0100
> +++ gcc/config/i386/i386-features.c 2020-03-17 09:12:38.016874905 +0100
> @@ -2224,7 +2224,7 @@ remove_partial_avx_dependency (void)
> loop_optimizer_init (AVOID_CFG_MODIFICATIONS);
>
> /* Generate a vxorps at entry of the nearest dominator for basic
> - blocks with conversions, which is in the the fake loop that
> + blocks with conversions, which is in the fake loop that
> contains the whole function, so that there is only a single
> vxorps in the whole function. */
> bb = nearest_common_dominator_for_set (CDI_DOMINATORS,
> --- gcc/config/msp430/msp430.c.jj 2020-01-12 11:54:36.368414088 +0100
> +++ gcc/config/msp430/msp430.c 2020-03-17 09:14:48.143964304 +0100
> @@ -1916,7 +1916,7 @@ msp430_select_section (tree decl, int re
> case SECCAT_RODATA_MERGE_CONST:
> return default_elf_select_section (decl, reloc, align);
>
> - /* The sections listed below are are not supported for MSP430.
> + /* The sections listed below are not supported for MSP430.
> They should not be generated, but in case they are, we use
> default_select_section so they get placed in sections
> the msp430 assembler and linker understand. */
> --- gcc/config/gcn/gcn-run.c.jj 2020-01-12 11:54:36.306415023 +0100
> +++ gcc/config/gcn/gcn-run.c 2020-03-17 09:11:05.385234861 +0100
> @@ -575,7 +575,7 @@ found_main:;
> break;
> case R_AMDGPU_REL64:
> /* FIXME
> - LLD seems to emit REL64 where the the assembler has ABS64.
> + LLD seems to emit REL64 where the assembler has ABS64.
> This is clearly wrong because it's not what the compiler
> is expecting. Let's assume, for now, that it's a bug.
> In any case, GCN kernels are always self contained and
> --- gcc/config/aarch64/aarch64-sve.md.jj 2020-02-21 21:54:56.302353043
> +0100
> +++ gcc/config/aarch64/aarch64-sve.md 2020-03-17 09:07:22.207511407 +0100
> @@ -2477,7 +2477,7 @@ (define_insn "@aarch64_vec_duplicate_vq<
> )
>
> ;; This is used for vec_duplicate<mode>s from memory, but can also
> -;; be used by combine to optimize selects of a a vec_duplicate<mode>
> +;; be used by combine to optimize selects of a vec_duplicate<mode>
> ;; with zero.
> (define_insn "sve_ld1r<mode>"
> [(set (match_operand:SVE_ALL 0 "register_operand" "=w")
> --- gcc/config/aarch64/aarch64.c.jj 2020-03-13 11:33:00.736261599 +0100
> +++ gcc/config/aarch64/aarch64.c 2020-03-17 09:06:48.201010668 +0100
> @@ -21296,7 +21296,7 @@ aarch64_gen_adjusted_ldpstp (rtx *operan
> {
> base_off = 0x1000 - 1;
> /* We must still make sure that the base offset is aligned with respect
> - to the address. But it may may not be made any bigger. */
> + to the address. But it may not be made any bigger. */
> base_off -= (((base_off % msize) - (off_val_1 % msize)) + msize) %
> msize;
> }
>
> --- gcc/config/aarch64/falkor-tag-collision-avoidance.c.jj 2020-01-12
> 11:54:36.255415792 +0100
> +++ gcc/config/aarch64/falkor-tag-collision-avoidance.c 2020-03-17
> 09:08:57.457113016 +0100
> @@ -699,7 +699,7 @@ in_same_chain (rtx_insn *insn, rtx_insn
>
>
> /* Callback function to traverse the tag map and drop loads that have the
> same
> - destination and and in the same chain of occurrence. Routine always
> returns
> + destination and are in the same chain of occurrence. Routine always
> returns
> true to allow traversal through all of TAG_MAP. */
> bool
> single_dest_per_chain (const rtx &t ATTRIBUTE_UNUSED, insn_info_list_t *v,
> --- gcc/config/nvptx/nvptx.c.jj 2020-01-12 11:54:36.376413967 +0100
> +++ gcc/config/nvptx/nvptx.c 2020-03-17 09:17:49.513301135 +0100
> @@ -975,7 +975,7 @@ write_fn_proto_from_insn (std::stringstr
> }
>
> /* DECL is an external FUNCTION_DECL, make sure its in the fndecl hash
> - table and and write a ptx prototype. These are emitted at end of
> + table and write a ptx prototype. These are emitted at end of
> compilation. */
>
> static void
> --- gcc/config/fr30/fr30.c.jj 2020-01-12 11:54:36.299415129 +0100
> +++ gcc/config/fr30/fr30.c 2020-03-17 09:10:32.509717517 +0100
> @@ -781,7 +781,7 @@ fr30_arg_partial_bytes (cumulative_args_
> are sufficient argument registers available (or if no registers
> are needed because the parameter must be passed on the stack)
> then return zero, as this parameter does not require partial
> - register, partial stack stack space. */
> + register, partial stack space. */
> if (*cum + fr30_num_arg_regs (arg) <= FR30_NUM_ARG_REGS)
> return 0;
>
> --- gcc/config/rs6000/rs6000-string.c.jj 2020-01-12 11:54:36.396413665
> +0100
> +++ gcc/config/rs6000/rs6000-string.c 2020-03-17 09:21:59.035637230 +0100
> @@ -679,7 +679,7 @@ expand_cmp_vec_sequence (unsigned HOST_W
> bnl 6,.Lmismatch
>
> For the P8 LE case, we use lxvd2x and compare full 16 bytes
> - but then use use vgbbd and a shift to get two bytes with the
> + but then use vgbbd and a shift to get two bytes with the
> information we need in the correct order.
>
> VEC/VSX compare sequence if TARGET_P9_VECTOR:
> --- gcc/config/rs6000/rs6000-p8swap.c.jj 2020-01-12 11:54:36.396413665
> +0100
> +++ gcc/config/rs6000/rs6000-p8swap.c 2020-03-17 09:21:02.554466588 +0100
> @@ -1922,7 +1922,7 @@ replace_swapped_load_constant (swap_web_
> XEXP (new_mem, 0) = base_reg;
>
> /* Move the newly created insn ahead of the load insn. */
> - /* The last insn is the the insn that forced new_mem into a register. */
> + /* The last insn is the insn that forced new_mem into a register. */
> rtx_insn *force_insn = get_last_insn ();
> /* Remove this insn from the end of the instruction sequence. */
> remove_insn (force_insn);
> --- gcc/config/rs6000/rs6000-c.c.jj 2020-03-11 09:28:53.956214091 +0100
> +++ gcc/config/rs6000/rs6000-c.c 2020-03-17 09:20:06.880284089 +0100
> @@ -397,7 +397,7 @@ rs6000_target_modify_macros (bool define
> (e.g. ISA_2_1_MASKS, ISA_3_0_MASKS_SERVER) and for a list of
> the specific flags that are associated with each of the cpu
> choices that can be specified as the target of a -mcpu=target
> - compile option, or as the the target of a --with-cpu=target
> + compile option, or as the target of a --with-cpu=target
> configure option. Target flags that are specified in either
> of these two ways are considered "implicit" since the flags
> are not mentioned specifically by name.
> --- gcc/config/rs6000/rs6000.c.jj 2020-03-14 08:11:44.999452074 +0100
> +++ gcc/config/rs6000/rs6000.c 2020-03-17 09:19:23.889915341 +0100
> @@ -3956,7 +3956,7 @@ rs6000_option_override_internal (bool gl
> }
>
> /* Enable the default support for IEEE 128-bit floating point on Linux VSX
> - sytems. In GCC 7, we would enable the the IEEE 128-bit floating point
> + sytems. In GCC 7, we would enable the IEEE 128-bit floating point
> infrastructure (-mfloat128-type) but not enable the actual __float128
> type
> unless the user used the explicit -mfloat128. In GCC 8, we enable both
> the keyword as well as the type. */
> --- gcc/config/rs6000/rs6000-logue.c.jj 2020-02-07 11:12:22.370903871 +0100
> +++ gcc/config/rs6000/rs6000-logue.c 2020-03-17 09:20:39.004812386 +0100
> @@ -1547,7 +1547,7 @@ rs6000_emit_probe_stack_range_stack_clas
>
> /* If explicitly requested,
> or the rounded size is not the same as the original size
> - or the the rounded size is greater than a page,
> + or the rounded size is greater than a page,
> then we will need a copy of the original stack pointer. */
> if (rounded_size != orig_size
> || rounded_size > probe_interval
> --- gcc/config/nds32/nds32-md-auxiliary.c.jj 2020-01-12 11:54:36.370414057
> +0100
> +++ gcc/config/nds32/nds32-md-auxiliary.c 2020-03-17 09:17:09.905882720
> +0100
> @@ -3304,12 +3304,12 @@ nds32_split_ashiftdi3 (rtx dst, rtx src,
> ext_start = gen_reg_rtx (SImode);
>
> /*
> - # In fact, we want to check shift amonut is great than or equal 32,
> - # but in some corner case, the shift amount might be very large
> value,
> - # however we've defined SHIFT_COUNT_TRUNCATED, so GCC think we've
> - # handle that correctly without any truncate.
> - # so check the the condition of (shiftamount & 32) is most
> - # safe way to do.
> + # In fact, we want to check shift amount is greater than or equal to
> + # 32, but in some corner case, the shift amount might be very large
> + # value, however we've defined SHIFT_COUNT_TRUNCATED, so GCC thinks
> + # we've handled that correctly without any truncate.
> + # So checking the condition of (shiftamount & 32) is the safest
> + # way to do it.
> if (shiftamount & 32)
> dst_low_part = 0
> dst_high_part = src_low_part << shiftamount & 0x1f
>
> Jakub
>
