From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: by sourceware.org (Postfix, from userid 2103) id A333D3857C50; Tue, 21 Jun 2022 18:30:15 +0000 (GMT) DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org A333D3857C50 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable From: Nick Alcock To: bfd-cvs@sourceware.org, gdb-cvs@sourceware.org Subject: [binutils-gdb] libctf: fix linking together multiple objects derived from the same source X-Act-Checkin: binutils-gdb X-Git-Author: Nick Alcock X-Git-Refname: refs/heads/master X-Git-Oldrev: bfda149dc1b0376f8cb4c423790246794c7ee873 X-Git-Newrev: 6bd2318f32842a27b03677b670421f93c14f9302 Message-Id: <20220621183015.A333D3857C50@sourceware.org> Date: Tue, 21 Jun 2022 18:30:15 +0000 (GMT) X-BeenThere: gdb-cvs@sourceware.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: Gdb-cvs mailing list List-Unsubscribe: , List-Archive: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 21 Jun 2022 18:30:15 -0000 https://sourceware.org/git/gitweb.cgi?p=3Dbinutils-gdb.git;h=3D6bd2318f3284= 2a27b03677b670421f93c14f9302 commit 6bd2318f32842a27b03677b670421f93c14f9302 Author: Nick Alcock Date: Fri Jun 10 17:05:50 2022 +0100 libctf: fix linking together multiple objects derived from the same sou= rce =20 Right now, if you compile the same .c input repeatedly with CTF enabled and different compilation flags, then arrange to link all of these together, then things misbehave in various ways. libctf may conflate either inputs (if the .o files have the same name, say if they are stored in different .a archives), or per-CU outputs when conflicting types are found: the latter can lead to entirely spurious errors when it tries to produce multiple per-CU outputs with the same name (discarding all but the last, but then looking for types in the earlier ones which have just been thrown away). =20 Fixing this is multi-pronged. Both inputs and outputs need to be differentiated in the hashtables libctf keeps them in: inputs with the same cuname and filename need to be considered distinct as long as they have different associated CTF dicts, and per-CU outputs need to be considered distinct as long as they have different associated input dicts. Right now there is nothing tying the two together other than the CU name: fix this by introducing a new field in the ctf_dict_t named ctf_link_in_out, which (for input dicts) points to the associated per-CU output dict (if any), and for output dicts points to the associated input dict. At creation time the name used is completely arbitrary: it's only important that it be distinct if CTF dicts are distinct. So, when a clash is found, adjust the CU name by sticking the number of elements in the input on the end. At output time, the CU name will appear in the linked object, so it matters a little more that it look slightly less ugly: in conflicting cases, append an incrementing integer, starting at 0. =20 This naming scheme is not very helpful, but it's hard to see what else we can do. The input .o name may be the same. The input .a name is not even visible to ctf_link, and even *that* might be the same, because .a's can contain many members with the same name, all of which participate in the link. All we really know is that the two have distinct dictionaries with distinct types in them, and at least this way they are all represented, any any symbols, variables etc referring to those types are accurately stored. =20 (As a side-effect this also fixes a use-after-free and double-free when errors are found during variable or symbol emission.) =20 Use the opportunity to prevent a couple of sources of problems, to wit changing the active CU mappings when a link has already been done (no effect on ld, which doesn't use CU mappings at all), and causing multiple consecutive ctf_link's to have the same net effect as just doing the last one (no effect on ld, which only ever does one ctf_link) rather than having the links be a sort of half-incremental not-really-intended mess. =20 libctf/ChangeLog: =20 PR libctf/29242 * ctf-impl.h (struct ctf_dict) [ctf_link_in_out]: New. * ctf-dedup.c (ctf_dedup_emit_type): Set it. * ctf-link.c (ctf_link_add_ctf_internal): Set the input CU name uniquely when clashes are found. (ctf_link_add): Document what repeated additions do. (ctf_new_per_cu_name): New, come up with a consistent name for a new per-CU dict. (ctf_link_deduplicating): Use it. (ctf_create_per_cu): Use it, and ctf_link_in_out, and set ctf_link_in_out properly. Don't overwrite per-CU dicts with per-CU dicts relating to different inputs. (ctf_link_add_cu_mapping): Prevent per-CU mappings being set up if we already have per-CU outputs. (ctf_link_one_variable): Adjust ctf_link_per_cu call. (ctf_link_deduplicating_one_symtypetab): Likewise. (ctf_link_empty_outputs): New, delete all the ctf_link_outputs and blank out ctf_link_in_out on the corresponding inputs. (ctf_link): Clarify the effect of multiple ctf_link calls. Empty ctf_link_outputs if it already exists rather than having the old output leak into the new link. Fix a variable name. * testsuite/config/default.exp (AR): Add. (OBJDUMP): Likewise. * testsuite/libctf-regression/libctf-repeat-cu.exp: New test. * testsuite/libctf-regression/libctf-repeat-cu*: Main program, library, and expected results for the test. Diff: --- libctf/ctf-dedup.c | 2 + libctf/ctf-impl.h | 4 + libctf/ctf-link.c | 161 ++++++++++++++++-= ---- libctf/testsuite/config/default.exp | 7 + .../libctf-regression/libctf-repeat-cu-lib.c | 9 ++ .../libctf-regression/libctf-repeat-cu-main.c | 5 + .../testsuite/libctf-regression/libctf-repeat-cu.d | 7 + .../libctf-regression/libctf-repeat-cu.exp | 118 +++++++++++++++ 8 files changed, 280 insertions(+), 33 deletions(-) diff --git a/libctf/ctf-dedup.c b/libctf/ctf-dedup.c index cddf4376eae..dcde3e88c6f 100644 --- a/libctf/ctf-dedup.c +++ b/libctf/ctf-dedup.c @@ -2671,6 +2671,8 @@ ctf_dedup_emit_type (const char *hval, ctf_dict_t *ou= tput, ctf_dict_t **inputs, ctf_parent_name_set (target, _CTF_SECTION); =20 input->ctf_dedup.cd_output =3D target; + input->ctf_link_in_out =3D target; + target->ctf_link_in_out =3D input; } output_num =3D input_num; } diff --git a/libctf/ctf-impl.h b/libctf/ctf-impl.h index 6b6ec16291a..465f1c6c58e 100644 --- a/libctf/ctf-impl.h +++ b/libctf/ctf-impl.h @@ -449,6 +449,10 @@ struct ctf_dict ctf_dynhash_t *ctf_link_inputs; /* Inputs to this link. */ ctf_dynhash_t *ctf_link_outputs; /* Additional outputs from this link. = */ =20 + /* If a link input CU, points at the corresponding per-CU output (if any= ); + if an output, points at the input (if any). */ + ctf_dict_t *ctf_link_in_out; + /* Map input types to output types for ctf_add_type. Key is a ctf_link_type_key_t: value is a type ID. */ ctf_dynhash_t *ctf_link_type_mapping; diff --git a/libctf/ctf-link.c b/libctf/ctf-link.c index d92a6930dd0..f2317302e26 100644 --- a/libctf/ctf-link.c +++ b/libctf/ctf-link.c @@ -72,7 +72,7 @@ ctf_unnamed_cuname (ctf_dict_t *fp) never gets explicitly freed in the ctf_link_input. */ typedef struct ctf_link_input { - const char *clin_filename; + char *clin_filename; ctf_archive_t *clin_arc; ctf_dict_t *clin_fp; int n; @@ -84,6 +84,7 @@ ctf_link_input_close (void *input) ctf_link_input_t *i =3D (ctf_link_input_t *) input; if (i->clin_arc) ctf_arc_close (i->clin_arc); + free (i->clin_filename); free (i); } =20 @@ -93,27 +94,50 @@ static int ctf_link_add_ctf_internal (ctf_dict_t *fp, ctf_archive_t *ctf, ctf_dict_t *fp_input, const char *name) { - ctf_link_input_t *input =3D NULL; - char *dupname =3D NULL; + int existing =3D 0; + ctf_link_input_t *input; + char *filename, *keyname; =20 - if ((input =3D calloc (1, sizeof (ctf_link_input_t))) =3D=3D NULL) + /* Existing: return it, or (if a different dict with the same name + is already there) make up a new unique name. Always use the actual n= ame + for the filename, because that needs to be ctf_open()ed. */ + + if ((input =3D ctf_dynhash_lookup (fp->ctf_link_inputs, name)) !=3D NULL) + { + if ((fp_input !=3D NULL && (input->clin_fp =3D=3D fp_input)) + || (ctf !=3D NULL && (input->clin_arc =3D=3D ctf))) + return 0; + existing =3D 1; + } + + if ((filename =3D strdup (name)) =3D=3D NULL) goto oom; =20 - if ((dupname =3D strdup (name)) =3D=3D NULL) + if ((input =3D calloc (1, sizeof (ctf_link_input_t))) =3D=3D NULL) goto oom; =20 input->clin_arc =3D ctf; input->clin_fp =3D fp_input; - input->clin_filename =3D dupname; + input->clin_filename =3D filename; input->n =3D ctf_dynhash_elements (fp->ctf_link_inputs); =20 - if (ctf_dynhash_insert (fp->ctf_link_inputs, dupname, input) < 0) + if (existing) + { + if (asprintf (&keyname, "%s#%li", name, (long int) + ctf_dynhash_elements (fp->ctf_link_inputs)) < 0) + goto oom; + } + else if ((keyname =3D strdup (name)) =3D=3D NULL) + goto oom; + + if (ctf_dynhash_insert (fp->ctf_link_inputs, keyname, input) < 0) goto oom; =20 return 0; oom: free (input); - free (dupname); + free (filename); + free (keyname); return ctf_set_errno (fp, ENOMEM); } =20 @@ -133,6 +157,10 @@ ctf_link_add_ctf_internal (ctf_dict_t *fp, ctf_archive= _t *ctf, The order of calls to this function influences the order of types in t= he final link output, but otherwise is not important. =20 + Repeated additions of the same NAME have no effect; repeated additions= of + different dicts with the same NAME add all the dicts with unique NAMEs + derived from NAME. + Private for now, but may in time become public once support for BUF is implemented. */ =20 @@ -235,20 +263,50 @@ ctf_link_lazy_open (ctf_dict_t *fp, ctf_link_input_t = *input) return (ssize_t) count; } =20 -/* Return a per-CU output CTF dictionary suitable for the given CU, creati= ng and - interning it if need be. */ +/* Find a non-clashing unique name for a per-CU output dict, to prevent di= stinct + members corresponding to inputs with identical cunames from overwriting= each + other. The name should be something like NAME. */ + +static char * +ctf_new_per_cu_name (ctf_dict_t *fp, const char *name) +{ + char *dynname; + long int i =3D 0; + + if ((dynname =3D strdup (name)) =3D=3D NULL) + return NULL; + + while ((ctf_dynhash_lookup (fp->ctf_link_outputs, dynname)) !=3D NULL) + { + free (dynname); + if (asprintf (&dynname, "%s#%li", name, i++) < 0) + return NULL; + } + + return dynname; +} + +/* Return a per-CU output CTF dictionary suitable for the given INPUT or C= U, + creating and interning it if need be. */ =20 -_libctf_nonnull_((1,2)) static ctf_dict_t * -ctf_create_per_cu (ctf_dict_t *fp, const char *cu_name) +ctf_create_per_cu (ctf_dict_t *fp, ctf_dict_t *input, const char *cu_name) { ctf_dict_t *cu_fp; const char *ctf_name =3D NULL; char *dynname =3D NULL; =20 - /* First, check the mapping table and translate the per-CU name we use + /* Already has a per-CU mapping? Just return it. */ + + if (input && input->ctf_link_in_out) + return input->ctf_link_in_out; + + /* Check the mapping table and translate the per-CU name we use accordingly. */ =20 + if (cu_name =3D=3D NULL) + cu_name =3D ctf_unnamed_cuname (input); + if (fp->ctf_link_in_cu_mapping) { if ((ctf_name =3D ctf_dynhash_lookup (fp->ctf_link_in_cu_mapping, @@ -259,7 +317,12 @@ ctf_create_per_cu (ctf_dict_t *fp, const char *cu_name) if (ctf_name =3D=3D NULL) ctf_name =3D cu_name; =20 - if ((cu_fp =3D ctf_dynhash_lookup (fp->ctf_link_outputs, ctf_name)) =3D= =3D NULL) + /* Look up the per-CU dict. If we don't know of one, or it is for + a different input CU which just happens to have the same name, + create a new one. */ + + if ((cu_fp =3D ctf_dynhash_lookup (fp->ctf_link_outputs, ctf_name)) =3D= =3D NULL + || cu_fp->ctf_link_in_out !=3D fp) { int err; =20 @@ -271,14 +334,19 @@ ctf_create_per_cu (ctf_dict_t *fp, const char *cu_nam= e) return NULL; } =20 - if ((dynname =3D strdup (ctf_name)) =3D=3D NULL) - goto oom; - if (ctf_dynhash_insert (fp->ctf_link_outputs, dynname, cu_fp) < 0) + ctf_import_unref (cu_fp, fp); + + if ((dynname =3D ctf_new_per_cu_name (fp, ctf_name)) =3D=3D NULL) goto oom; =20 - ctf_import_unref (cu_fp, fp); ctf_cuname_set (cu_fp, cu_name); + ctf_parent_name_set (cu_fp, _CTF_SECTION); + cu_fp->ctf_link_in_out =3D fp; + fp->ctf_link_in_out =3D cu_fp; + + if (ctf_dynhash_insert (fp->ctf_link_outputs, dynname, cu_fp) < 0) + goto oom; } return cu_fp; =20 @@ -304,6 +372,10 @@ ctf_link_add_cu_mapping (ctf_dict_t *fp, const char *f= rom, const char *to) char *f =3D NULL, *t =3D NULL; ctf_dynhash_t *one_out; =20 + /* Mappings cannot be set up if per-CU output dicts already exist. */ + if (fp->ctf_link_outputs && ctf_dynhash_elements (fp->ctf_link_outputs) = !=3D 0) + return (ctf_set_errno (fp, ECTF_LINKADDEDLATE)); + if (fp->ctf_link_in_cu_mapping =3D=3D NULL) fp->ctf_link_in_cu_mapping =3D ctf_dynhash_create (ctf_hash_string, ctf_hash_eq_string, free, @@ -481,7 +553,7 @@ ctf_link_one_variable (ctf_dict_t *fp, ctf_dict_t *in_f= p, const char *name, return 0; } =20 - if ((per_cu_out_fp =3D ctf_create_per_cu (fp, ctf_unnamed_cuname (in_fp)= )) =3D=3D NULL) + if ((per_cu_out_fp =3D ctf_create_per_cu (fp, in_fp, NULL)) =3D=3D NULL) return -1; /* errno is set for us. */ =20 /* If the type was not found, check for it in the child too. */ @@ -952,7 +1024,7 @@ ctf_link_deduplicating_one_symtypetab (ctf_dict_t *fp,= ctf_dict_t *input, continue; } =20 - if ((per_cu_out_fp =3D ctf_create_per_cu (fp, ctf_unnamed_cuname (in= put))) =3D=3D NULL) + if ((per_cu_out_fp =3D ctf_create_per_cu (fp, input, NULL)) =3D=3D N= ULL) return -1; /* errno is set for us. */ =20 /* If the type was not found, check for it in the child too. */ @@ -1257,6 +1329,31 @@ ctf_link_deduplicating_per_cu (ctf_dict_t *fp) return 0; } =20 +/* Empty all the ctf_link_outputs. */ +static int +ctf_link_empty_outputs (ctf_dict_t *fp) +{ + ctf_next_t *i =3D NULL; + void *v; + int err; + + ctf_dynhash_empty (fp->ctf_link_outputs); + + while ((err =3D ctf_dynhash_next (fp->ctf_link_inputs, &i, NULL, &v)) = =3D=3D 0) + { + ctf_dict_t *in =3D (ctf_dict_t *) v; + in->ctf_link_in_out =3D NULL; + } + if (err !=3D ECTF_NEXT_END) + { + fp->ctf_flags &=3D ~LCTF_LINKING; + ctf_err_warn (fp, 1, err, _("iteration error removing old outputs")); + ctf_set_errno (fp, err); + return -1; + } + return 0; +} + /* Do a deduplicating link using the ctf-dedup machinery. */ static void ctf_link_deduplicating (ctf_dict_t *fp) @@ -1320,7 +1417,7 @@ ctf_link_deduplicating (ctf_dict_t *fp) continue; } =20 - if ((dynname =3D strdup (ctf_cuname (outputs[i]))) =3D=3D NULL) + if ((dynname =3D ctf_new_per_cu_name (fp, ctf_cuname (outputs[i]))) = =3D=3D NULL) goto oom_one_output; =20 if (ctf_dynhash_insert (fp->ctf_link_outputs, dynname, outputs[i]) <= 0) @@ -1374,20 +1471,15 @@ ctf_link_deduplicating (ctf_dict_t *fp) return; =20 err_clean_outputs: - for (i =3D 1; i < noutputs; i++) - { - ctf_dynhash_remove (fp->ctf_link_outputs, ctf_cuname (outputs[i])); - ctf_dict_close (outputs[i]); - } + ctf_link_empty_outputs (fp); goto err; } =20 /* Merge types and variable sections in all dicts added to the link togeth= er. - All the added dicts are closed. */ + The result of any previous link is discarded. */ int ctf_link (ctf_dict_t *fp, int flags) { - ctf_next_t *i =3D NULL; int err; =20 fp->ctf_link_flags =3D flags; @@ -1395,7 +1487,9 @@ ctf_link (ctf_dict_t *fp, int flags) if (fp->ctf_link_inputs =3D=3D NULL) return 0; /* Nothing to do. */ =20 - if (fp->ctf_link_outputs =3D=3D NULL) + if (fp->ctf_link_outputs !=3D NULL) + ctf_link_empty_outputs (fp); + else fp->ctf_link_outputs =3D ctf_dynhash_create (ctf_hash_string, ctf_hash_eq_string, free, (ctf_hash_free_fun) @@ -1411,13 +1505,14 @@ ctf_link (ctf_dict_t *fp, int flags) fp->ctf_flags |=3D LCTF_LINKING; if (fp->ctf_link_out_cu_mapping && (flags & CTF_LINK_EMPTY_CU_MAPPINGS)) { - void *v; + ctf_next_t *i =3D NULL; + void *k; =20 - while ((err =3D ctf_dynhash_next (fp->ctf_link_out_cu_mapping, &i, &= v, + while ((err =3D ctf_dynhash_next (fp->ctf_link_out_cu_mapping, &i, &= k, NULL)) =3D=3D 0) { - const char *to =3D (const char *) v; - if (ctf_create_per_cu (fp, to) =3D=3D NULL) + const char *to =3D (const char *) k; + if (ctf_create_per_cu (fp, NULL, to) =3D=3D NULL) { fp->ctf_flags &=3D ~LCTF_LINKING; ctf_next_destroy (i); diff --git a/libctf/testsuite/config/default.exp b/libctf/testsuite/config/= default.exp index f3239468ac8..29d31363071 100644 --- a/libctf/testsuite/config/default.exp +++ b/libctf/testsuite/config/default.exp @@ -55,6 +55,13 @@ if {![info exists CC_FOR_TARGET]} { if {![info exists CFLAGS_FOR_TARGET]} { set CFLAGS_FOR_TARGET $CFLAGS } +if ![info exists AR] then { + set AR [findfile $base_dir/../binutils/ar] +} + +if {![info exists OBJDUMP]} { + set OBJDUMP [findfile $base_dir/../binutils/objdump] +} =20 # load the utility procedures load_lib ctf-lib.exp diff --git a/libctf/testsuite/libctf-regression/libctf-repeat-cu-lib.c b/li= bctf/testsuite/libctf-regression/libctf-repeat-cu-lib.c new file mode 100644 index 00000000000..7ebdb09b84c --- /dev/null +++ b/libctf/testsuite/libctf-regression/libctf-repeat-cu-lib.c @@ -0,0 +1,9 @@ +#ifdef INT +typedef int ret_t; +#elif CHAR +typedef char *ret_t; +#else +typedef short *ret_t; +#endif + +ret_t FUN (void) { return 0; } diff --git a/libctf/testsuite/libctf-regression/libctf-repeat-cu-main.c b/l= ibctf/testsuite/libctf-regression/libctf-repeat-cu-main.c new file mode 100644 index 00000000000..bfbaf0cc8dc --- /dev/null +++ b/libctf/testsuite/libctf-regression/libctf-repeat-cu-main.c @@ -0,0 +1,5 @@ +typedef short ret_t; +int a (void); +int b (void); +int c (void); +int blah (void) { a(); b(); c(); } diff --git a/libctf/testsuite/libctf-regression/libctf-repeat-cu.d b/libctf= /testsuite/libctf-regression/libctf-repeat-cu.d new file mode 100644 index 00000000000..81df80442ad --- /dev/null +++ b/libctf/testsuite/libctf-regression/libctf-repeat-cu.d @@ -0,0 +1,7 @@ +#... +CTF archive member: .*/libctf-repeat-cu-lib.c: +#... +CTF archive member: .*/libctf-repeat-cu-lib.c#0: +#... +CTF archive member: .*/libctf-repeat-cu-lib.c#1: +#... diff --git a/libctf/testsuite/libctf-regression/libctf-repeat-cu.exp b/libc= tf/testsuite/libctf-regression/libctf-repeat-cu.exp new file mode 100644 index 00000000000..becee958ca3 --- /dev/null +++ b/libctf/testsuite/libctf-regression/libctf-repeat-cu.exp @@ -0,0 +1,118 @@ +# Copyright (C) 2021-2022 Free Software Foundation, Inc. +# +# This file is part of the GNU Binutils. +# +# This program is free software; you can redistribute it and/or modify +# it under the terms of the GNU General Public License as published by +# the Free Software Foundation; either version 3 of the License, or +# (at your option) any later version. +# +# This program is distributed in the hope that it will be useful, +# but WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +# GNU General Public License for more details. +# +# You should have received a copy of the GNU General Public License +# along with this program; if not, write to the Free Software +# Foundation, Inc., 51 Franklin Street - Fifth Floor, Boston, +# MA 02110-1301, USA. +# + +load_file $srcdir/../../ld/testsuite/lib/ld-lib.exp + +global srcdir subdir OBJDUMP +global testname +global subsrcdir + +set subsrcdir "$srcdir/$subdir/" +set testname "$dir/libctf-repeat-cu.exp" + +if ![is_elf_format] { + unsupported "CTF needs bfd changes to be emitted on non-ELF" + return 0 +} + +if {![check_ctf_available]} { + unsupported "no CTF format support in the compiler" + return 0 +} + +if {[info exists env(LC_ALL)]} { + set old_lc_all $env(LC_ALL) +} +set env(LC_ALL) "C" + +# Compile one SRC to OBJ and put it into ARCHIVE. +proc one_lib_compile { src flags obj archive } { + global CC_FOR_TARGET CFLAGS_FOR_TARGET AR subsrcdir + + if [is_remote host] { + set src [remote_download host [file join $subsrcdir $src]] + } else { + set src [file join $subsrcdir $src] + } + + set comp_output [run_host_cmd "$CC_FOR_TARGET" "$CFLAGS_FOR_TARGET $fl= ags -gctf -fPIC -c -o $obj $src"] + if { $comp_output !=3D "" } { + return $comp_output + } + + set ar_output [run_host_cmd "$AR" "rc $archive $obj"] + return $comp_output +} + +# Compile one SRC to OBJ and put it into ARCHIVE: error-check the result. +proc one_lib_compile_check { src flags obj archive } { + global testname + + set comp_output [one_lib_compile $src $flags $obj $archive] + + if { $comp_output !=3D ""} { + send_log "compilation of $src with $flags failed with <$comp_output>" + perror "compilation of $src with $flags failed" + fail $testname + return 0 + } + return 1 +} + +if { ! [one_lib_compile_check libctf-repeat-cu-lib.c "-DINT -DFUN=3Da" tmp= dir/libctf-repeat-cu-lib.o tmpdir/a.a] } { + return 0 +} + +if { ! [one_lib_compile_check libctf-repeat-cu-lib.c "-DCHAR -DFUN=3Db" tm= pdir/libctf-repeat-cu-lib.o tmpdir/b.a] } { + return 0 +} + +if { ! [one_lib_compile_check libctf-repeat-cu-lib.c "-DFUN=3Dc" tmpdir/li= bctf-repeat-cu-lib.o tmpdir/c.a] } { + return 0 +} + +if [is_remote host] { + set src [remote_download host [file join $subsrcdir libctf-repeat-cu-m= ain.c]] +} else { + set src [file join $subsrcdir libctf-repeat-cu-main.c] +} + +set comp_output [run_host_cmd "$CC_FOR_TARGET" "$CFLAGS_FOR_TARGET -gctf -= fPIC -shared -o tmpdir/libctf-repeat-cu-main.so $src tmpdir/a.a tmpdir/b.a = tmpdir/c.a"] +if { $comp_output !=3D "" } { + send_log "compilation of tmpdir/libctf-repeat-cu-main.so failed" + perror "compilation of tmpdir/libctf-repeat-cu-main.so failed" + fail $testname + return $comp_output +} + +set comp_output [run_host_cmd "$OBJDUMP" "--ctf tmpdir/libctf-repeat-cu-ma= in.so > tmpdir/dump.out"] + +if { [regexp_diff "tmpdir/dump.out" [file join $subsrcdir libctf-repeat-c= u.d] ] } { + fail $testname + if { $verbose =3D=3D 2 } then { verbose "output is [file_contents tmpd= ir/dump.out]" 2 } +} + +pass $testname + +if {[info exists old_lc_all]} { + set env(LC_ALL) $old_lc_all +} else { + unset env(LC_ALL) +}