From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 16030 invoked by alias); 24 May 2011 21:21:36 -0000 Received: (qmail 16015 invoked by uid 22791); 24 May 2011 21:21:33 -0000 X-SWARE-Spam-Status: No, hits=-6.4 required=5.0 tests=AWL,BAYES_00,RCVD_IN_DNSWL_HI,SPF_HELO_PASS,T_RP_MATCHES_RCVD X-Spam-Check-By: sourceware.org Received: from mx1.redhat.com (HELO mx1.redhat.com) (209.132.183.28) by sourceware.org (qpsmtpd/0.43rc1) with ESMTP; Tue, 24 May 2011 21:21:14 +0000 Received: from int-mx02.intmail.prod.int.phx2.redhat.com (int-mx02.intmail.prod.int.phx2.redhat.com [10.5.11.12]) by mx1.redhat.com (8.14.4/8.14.4) with ESMTP id p4OLLEZ0027332 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=OK) for ; Tue, 24 May 2011 17:21:14 -0400 Received: from host1.jankratochvil.net ([10.3.113.11]) by int-mx02.intmail.prod.int.phx2.redhat.com (8.13.8/8.13.8) with ESMTP id p4OLLCuj006351 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Tue, 24 May 2011 17:21:14 -0400 Received: from host1.jankratochvil.net (localhost [127.0.0.1]) by host1.jankratochvil.net (8.14.4/8.14.4) with ESMTP id p4OLLBDX010289; Tue, 24 May 2011 23:21:11 +0200 Received: (from jkratoch@localhost) by host1.jankratochvil.net (8.14.4/8.14.4/Submit) id p4OLLBIL010288; Tue, 24 May 2011 23:21:11 +0200 Date: Tue, 24 May 2011 21:21:00 -0000 From: Jan Kratochvil To: Keith Seitz Cc: gdb-patches@sourceware.org Subject: [rfc 2/2] Follow DW_AT_linkage_name for methods [Re: The future of dwarf2_physname] Message-ID: <20110524212111.GB7747@host1.jankratochvil.net> References: <4DD44983.7060406@redhat.com> <20110523131659.GA30344@host1.jankratochvil.net> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20110523131659.GA30344@host1.jankratochvil.net> User-Agent: Mutt/1.5.21 (2010-09-15) X-IsSubscribed: yes Mailing-List: contact gdb-patches-help@sourceware.org; run by ezmlm Precedence: bulk List-Id: List-Subscribe: List-Archive: List-Post: List-Help: , Sender: gdb-patches-owner@sourceware.org X-SW-Source: 2011-05/txt/msg00579.txt.bz2 On Mon, 23 May 2011 15:16:59 +0200, Jan Kratochvil wrote: > problem #1 of post-physname GDB > > (gdb) p 'f()' > $1 = {long (void)} 0x4004e4 ()> > (gdb) p 'long f()' > $2 = {} 0x4004e4 ()> > (gdb) p 'f()' > No symbol "f()" in current context. > (gdb) p 'long f()' > $3 = {} 0x4004fb ()> > > ==> dis.h <== > template > long f () { return 1; } > ==> dis1.C <== > #include "dis.h" > int main () { f (); } > ==> dis2.C <== > #include "dis.h" > void g () { f (); } > > The two different aliases for 0x4004e4 where the $2 one does not have any type > is a bug, at least both should have the same type. It can be seen also on the > f() case ($3) where the return type missing alias is not present because > only ELF (and not DWARF) symbol is present. So far I have kept this bug of missing type as GDB currently cannot provide multiple search names for a single symbol. And IMO there should exist both `long f()' as it is its demangled ELF (minimal symbol) name and also the `f()' as otherwise user will never find such template function by -completion from CLI. > I haven't found how to reproduce the regression > "Cannot take address of method m." for function templates (which use the > return type in their linkage) as I haven't found how to reference function > template externally from a class - this field is missing there in such case: Therefore I do not find this double-entry for template function such a real problem, just the return type featured name is missing its real type. > problem #2 of post-physname GDB > > (gdb) p 'int f()' > $1 = {} 0x40050f ()> > (gdb) p 'short f()' > $2 = {} 0x400531 ()> > (gdb) p 'f()' > $3 = {int (void)} 0x40050f ()> > > ==> manya.C <== > template > int f () { return 4; } > int a () { return f () == 4; } > extern int b (); > int main () { return a () && b () ? 0 : 1; } > ==> manyb.C <== > template > short f () { return 2; } > int b () { return f () == 2; } > > As the return type is stripped the name can become ambiguous. Different > template functions with same name, same parameter types but different return > type can be linked in a single program. It is true one cannot declare both > such template functions in a single CU (Compilation Unit) so I do not expect > such case is found in real world. I do not think such symbols exist in real world cases - when they are incompatible in a single CU anyway. Therefore the ambiguity is not a real world problem. > I was trying to propose DMGL_RET_POSTFIX but it has other drawbacks Also after a chat with Tom Tromey I no longer use DMGL_RET_POSTFIX. That is for ELF symbol `long m()' GDB with proposed patch uses `m()' (and not `m()long'). This is compatible with gdb-7.2 (physname) template functions. But I have found GDB requires full DWARF symbols for methods anyway so some of my examples could be a bit unreal, I have to reconsider. Anyway DW_AT_linkage_name following patch with no regressions on {x86_64,x86_64-m32,i686}-fedora15-linux-gnu is attached. It would need GCC push of the libiberty/ change. Thanks, Jan gdb/ 2011-05-24 Jan Kratochvil Tom Tromey * dwarf2read.c (check_physname): New variable. (dwarf2_physname): Prefer DW_AT_linkage_name over dwarf2_compute_name. (show_check_physname): New function. (_initialize_dwarf2_read): Add `check-physname' for check_physname. gdb/testsuite/ 2011-05-24 Jan Kratochvil * gdb.base/break-interp.exp (reach_1, test_ld): Allow also the prefix __GI_. * gdb.cp/psymtab-parameter.cc (func): Make it a template function. (f): New function. * gdb.cp/psymtab-parameter.exp (complete break 'func(): Rename to ... (complete p 'func(): ... here. * gdb.dwarf2/dw2-linkage-name-trust-main.cc: New file. * gdb.dwarf2/dw2-linkage-name-trust.S: New file. * gdb.dwarf2/dw2-linkage-name-trust.exp: New file. include/ 2011-05-24 Jan Kratochvil * demangle.h (DMGL_RET_POSTFIX): Extend the comment. (DMGL_RET_DROP): New. * cp-demangle.c (d_print_comp) : Do not pass DMGL_RET_POSTFIX or DMGL_RET_DROP. Suppress d_print_mod for DMGL_RET_POSTFIX. Support DMGL_RET_DROP. * testsuite/demangle-expected: New testcases for --ret-postfix and --ret-drop. * testsuite/test-demangle.c: Document --ret-drop in a comment. (main): New variable ret_drop, fill it, call cplus_demangle with it. --- a/gdb/dwarf2read.c +++ b/gdb/dwarf2read.c @@ -119,6 +119,9 @@ _STATEMENT_PROLOGUE; /* When non-zero, dump DIEs after they are read in. */ static int dwarf2_die_debug = 0; +/* When non-zero, cross-check physname against demangler. */ +static int check_physname = 0; + static int pagesize; /* When set, the file that we're processing is known to have debugging @@ -5147,7 +5150,93 @@ dwarf2_full_name (char *name, struct die_info *die, struct dwarf2_cu *cu) static const char * dwarf2_physname (char *name, struct die_info *die, struct dwarf2_cu *cu) { - return dwarf2_compute_name (name, die, cu, 1); + struct attribute *attr; + const char *retval, *mangled = NULL; + char *canon = NULL; + struct cleanup *back_to; + int need_copy = 1; + + /* In this case dwarf2_compute_name is just a shortcut not building anything + on its own. */ + if (!die_needs_namespace (die, cu)) + return dwarf2_compute_name (name, die, cu, 1); + + back_to = make_cleanup (null_cleanup, NULL); + + attr = dwarf2_attr (die, DW_AT_linkage_name, cu); + if (!attr) + attr = dwarf2_attr (die, DW_AT_MIPS_linkage_name, cu); + + /* DW_AT_linkage_name is missing in some cases - depend on what GDB + has computed. */ + if (attr && DW_STRING (attr)) + { + char *demangled; + + mangled = DW_STRING (attr); + + /* Use DMGL_RET_DROP for C++ template functions to suppress their return + type. It is easier for GDB users to search for such functions as + `name(params)' than `long name(params)'. In such case the minimal + symbol names do not match the full symbol names but for template + functions there is never a need to look up their definition from their + declaration so the only disadvantage remains the minimal symbol + variant `long name(params)' does not have the proper inferior type. + */ + + demangled = cplus_demangle (mangled, (DMGL_PARAMS | DMGL_ANSI + | DMGL_VERBOSE + | (cu->language == language_java + ? DMGL_JAVA | DMGL_RET_POSTFIX + : DMGL_RET_DROP))); + if (demangled) + { + make_cleanup (xfree, demangled); + canon = demangled; + } + else + { + canon = (char *) mangled; + need_copy = 0; + } + } + + if (canon == NULL || check_physname) + { + const char *physname = dwarf2_compute_name (name, die, cu, 1); + + if (canon != NULL && strcmp (physname, canon) != 0) + { + /* It may not mean a bug in GDB. The compiler could also + compute DW_AT_linkage_name incorrectly. But in such case + GDB would need to be bug-to-bug compatible. */ + + complaint (&symfile_complaints, + _("Computed physname <%s> does not match demangled <%s> " + "(from linkage <%s>) - DIE at 0x%x [in module %s]"), + physname, canon, mangled, die->offset, cu->objfile->name); + + /* Prefer DW_AT_linkage_name (in the CANON form) - when it + is available here - over computed PHYSNAME. It is safer + against both buggy GDB and buggy compilers. */ + + retval = canon; + } + else + { + retval = physname; + need_copy = 0; + } + } + else + retval = canon; + + if (need_copy) + retval = obsavestring (retval, strlen (retval), + &cu->objfile->objfile_obstack); + + do_cleanups (back_to); + return retval; } /* Read the import statement specified by the given die and record it. */ @@ -16165,6 +16254,15 @@ show_dwarf2_always_disassemble (struct ui_file *file, int from_tty, value); } +static void +show_check_physname (struct ui_file *file, int from_tty, + struct cmd_list_element *c, const char *value) +{ + fprintf_filtered (file, + _("Whether to check \"physname\" is %s.\n"), + value); +} + void _initialize_dwarf2_read (void); void @@ -16220,6 +16318,14 @@ The value is the maximum depth to print."), NULL, &setdebuglist, &showdebuglist); + add_setshow_boolean_cmd ("check-physname", no_class, &check_physname, _("\ +Set cross-checking of \"physname\" code against demangler."), _("\ +Show cross-checking of \"physname\" code against demangler."), _("\ +When enabled, GDB's internal \"physname\" code is checked against\n\ +the demangler."), + NULL, show_check_physname, + &setdebuglist, &showdebuglist); + c = add_cmd ("gdb-index", class_files, save_gdb_index_command, _("\ Save a gdb-index file.\n\ --- a/gdb/testsuite/gdb.base/break-interp.exp +++ b/gdb/testsuite/gdb.base/break-interp.exp @@ -140,14 +140,14 @@ proc reach_1 {func command displacement} { } exp_continue } - -re "Breakpoint \[0-9\]+, \\.?$func \\(.*\\) at .*:\[0-9\]+\r\n.*$gdb_prompt $" { + -re "Breakpoint \[0-9\]+, \\.?(__GI_)?$func \\(.*\\) at .*:\[0-9\]+\r\n.*$gdb_prompt $" { if {$func == "_dl_debug_state"} { fail $test } else { pass $test } } - -re "Breakpoint \[0-9\]+, \[0-9xa-f\]+ in \\.?$func \\(\\).*\r\n$gdb_prompt $" { + -re "Breakpoint \[0-9\]+, \[0-9xa-f\]+ in \\.?(__GI_)?$func \\(\\).*\r\n$gdb_prompt $" { if {$func == "_dl_debug_state"} { fail $test } else { @@ -403,7 +403,7 @@ proc test_ld {file ifmain trynosym displacement} { reach "_dl_debug_state" "run" $displacement - gdb_test "bt" "#0 +\[^\r\n\]*\\m_dl_debug_state\\M.*" "dl bt" + gdb_test "bt" "#0 +\[^\r\n\]*\\m(__GI_)?_dl_debug_state\\M.*" "dl bt" if $ifmain { reach "main" continue "NONE" --- a/gdb/testsuite/gdb.cp/psymtab-parameter.cc +++ b/gdb/testsuite/gdb.cp/psymtab-parameter.cc @@ -16,7 +16,14 @@ along with this program. If not, see . */ -typedef int typedefed; -void func (typedefed param) +template +long +func () { } + +void +f () +{ + func (); +} --- a/gdb/testsuite/gdb.cp/psymtab-parameter.exp +++ b/gdb/testsuite/gdb.cp/psymtab-parameter.exp @@ -35,5 +35,6 @@ gdb_test_no_output "set language c++" # XFAIL than FAIL here. For example -readnow breaks it. gdb_test_no_output "maintenance info symtabs" -# GDB has shown only the `func(int)' entry before. -gdb_test "complete break 'func(" "break 'func\\(int\\)\r\nbreak 'func\\(typedefed\\)" +# GDB has shown only the `long func()' ELF symbol before, not the DWARF +# symbol +gdb_test "complete p 'func(" "p 'func\\(\\)" --- /dev/null +++ b/gdb/testsuite/gdb.dwarf2/dw2-linkage-name-trust-main.cc @@ -0,0 +1,41 @@ +/* This testcase is part of GDB, the GNU debugger. + + Copyright 2011 Free Software Foundation, Inc. + + 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, see . */ + +/* class C + { + public: + static int f (); + }; */ + +asm (".globl cu_text_start"); +asm ("cu_text_start:"); + +int +f (void) +{ + return 31173; +} + +int +main (void) +{ + f (); + return 0; +} + +asm (".globl cu_text_end"); +asm ("cu_text_end:"); --- /dev/null +++ b/gdb/testsuite/gdb.dwarf2/dw2-linkage-name-trust.S @@ -0,0 +1,134 @@ +/* This testcase is part of GDB, the GNU debugger. + + Copyright 2011 Free Software Foundation, Inc. + + 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, see . */ + +/* Debug information */ + + .data + .globl c +c: .4byte 0 + + .section .debug_info +.Lcu1_begin: + /* CU header */ + .4byte .Lcu1_end - .Lcu1_start /* Length of Compilation Unit */ +.Lcu1_start: + .2byte 2 /* DWARF Version */ + .4byte .Labbrev1_begin /* Offset into abbrev section */ + .byte 4 /* Pointer size */ + + /* CU die */ + .uleb128 1 /* Abbrev: DW_TAG_compile_unit */ + .4byte cu_text_start /* DW_AT_low_pc */ + .4byte cu_text_end /* DW_AT_high_pc */ + .ascii "file1.txt\0" /* DW_AT_name */ + .ascii "GNU C 3.3.3\0" /* DW_AT_producer */ + .byte 4 /* DW_AT_language (DW_LANG_C_plus_plus) */ + +.Ltype_class: + .uleb128 3 /* Abbrev: DW_TAG_class_type */ + .ascii "C\0" /* DW_AT_name */ + + .uleb128 5 /* Abbrev: DW_TAG_subprogram */ + .byte 1 /* DW_AT_external */ + .ascii "membername\0" /* DW_AT_name */ + .ascii "f\0" /* DW_AT_MIPS_linkage_name */ + .4byte .Ltype_int-.Lcu1_begin /* DW_AT_type */ + + .byte 0 /* End of children of DW_TAG_class_type */ + + .uleb128 4 /* Abbrev: DW_TAG_variable */ + .ascii "c\0" /* DW_AT_name */ + .4byte .Ltype_class-.Lcu1_begin /* DW_AT_type */ + .byte 1 /* DW_AT_external */ + +.Ltype_int: + .uleb128 2 /* Abbrev: DW_TAG_base_type */ + .ascii "int\0" /* DW_AT_name */ + .byte 4 /* DW_AT_byte_size */ + .byte 5 /* DW_AT_encoding */ + + .byte 0 /* End of children of CU */ + +.Lcu1_end: + +/* Abbrev table */ + .section .debug_abbrev +.Labbrev1_begin: + .uleb128 1 /* Abbrev code */ + .uleb128 0x11 /* DW_TAG_compile_unit */ + .byte 1 /* has_children */ + .uleb128 0x11 /* DW_AT_low_pc */ + .uleb128 0x1 /* DW_FORM_addr */ + .uleb128 0x12 /* DW_AT_high_pc */ + .uleb128 0x1 /* DW_FORM_addr */ + .uleb128 0x3 /* DW_AT_name */ + .uleb128 0x8 /* DW_FORM_string */ + .uleb128 0x25 /* DW_AT_producer */ + .uleb128 0x8 /* DW_FORM_string */ + .uleb128 0x13 /* DW_AT_language */ + .uleb128 0xb /* DW_FORM_data1 */ + .byte 0x0 /* Terminator */ + .byte 0x0 /* Terminator */ + + .uleb128 2 /* Abbrev code */ + .uleb128 0x24 /* DW_TAG_base_type */ + .byte 0 /* has_children */ + .uleb128 0x3 /* DW_AT_name */ + .uleb128 0x8 /* DW_FORM_string */ + .uleb128 0xb /* DW_AT_byte_size */ + .uleb128 0xb /* DW_FORM_data1 */ + .uleb128 0x3e /* DW_AT_encoding */ + .uleb128 0xb /* DW_FORM_data1 */ + .byte 0x0 /* Terminator */ + .byte 0x0 /* Terminator */ + + .uleb128 3 /* Abbrev code */ + .uleb128 0x2 /* DW_TAG_class_type */ + .byte 1 /* has_children */ + .uleb128 0x3 /* DW_AT_name */ + .uleb128 0x8 /* DW_FORM_string */ + .byte 0x0 /* Terminator */ + .byte 0x0 /* Terminator */ + + .uleb128 4 /* Abbrev code */ + .uleb128 0x34 /* DW_TAG_variable */ + .byte 0 /* has_children */ + .uleb128 0x3 /* DW_AT_name */ + .uleb128 0x8 /* DW_FORM_string */ + .uleb128 0x49 /* DW_AT_type */ + .uleb128 0x13 /* DW_FORM_ref4 */ + .uleb128 0x3f /* DW_AT_external */ + .uleb128 0xc /* DW_FORM_flag */ + .byte 0x0 /* Terminator */ + .byte 0x0 /* Terminator */ + + .uleb128 5 /* Abbrev code */ + .uleb128 0x2e /* DW_TAG_subprogram */ + .byte 0 /* has_children */ + .uleb128 0x3f /* DW_AT_external */ + .uleb128 0xc /* DW_FORM_flag */ + .uleb128 0x3 /* DW_AT_name */ + .uleb128 0x8 /* DW_FORM_string */ + .uleb128 0x2007 /* DW_AT_MIPS_linkage_name */ + .uleb128 0x8 /* DW_FORM_string */ + .uleb128 0x49 /* DW_AT_type */ + .uleb128 0x13 /* DW_FORM_ref4 */ + .byte 0x0 /* Terminator */ + .byte 0x0 /* Terminator */ + + .byte 0x0 /* Terminator */ + .byte 0x0 /* Terminator */ --- /dev/null +++ b/gdb/testsuite/gdb.dwarf2/dw2-linkage-name-trust.exp @@ -0,0 +1,55 @@ +# Copyright 2011 Free Software Foundation, Inc. + +# 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, see . + +# Check that GDB can call C++ functions whose parameters or return values have +# type containing a static member of the same type. + +# Still no C++ compiler is used. +if { [skip_cplus_tests] } { continue } + +load_lib dwarf.exp +# This test can only be run on targets which support DWARF-2 and use gas. +if {![dwarf2_support]} { + return 0 +} + +set testfile "dw2-linkage-name-trust" +set executable ${testfile} +set binfile ${objdir}/${subdir}/${executable} +if { [gdb_compile ${srcdir}/${subdir}/${testfile}-main.cc "${objdir}/${subdir}/${testfile}-main.o" object {c++ debug}] != "" + || [gdb_compile "${srcdir}/${subdir}/${testfile}.S" "${objdir}/${subdir}/${testfile}.o" object {}] != "" + || [gdb_compile "${objdir}/${subdir}/${testfile}-main.o ${objdir}/${subdir}/${testfile}.o" "${binfile}" executable {c++}] != "" } { + return -1 +} + +clean_restart $executable + +if ![runto_main] then { + return -1 +} + +# main is not provided by DWARF. +gdb_test_no_output "set language c++" + +# There are no mangled names in DWARF to suggest the v3 ABI. +gdb_test_no_output "set cp-abi gnu-v3" + +# GDB cannot resolve external member function for which only ELF (and not +# DWARF) symbol is available. Therefore the function `f' must have DWARF which +# confuses it a bit. + +gdb_test "p c.membername" " = {.*} 0x\[0-9a-f\]+ " +gdb_breakpoint "C::membername" +gdb_test "p c.membername ()" "\r\nBreakpoint \[0-9\]+, .*" --- a/include/demangle.h +++ b/include/demangle.h @@ -45,7 +45,12 @@ extern "C" { #define DMGL_VERBOSE (1 << 3) /* Include implementation details. */ #define DMGL_TYPES (1 << 4) /* Also try to demangle type encodings. */ #define DMGL_RET_POSTFIX (1 << 5) /* Print function return types (when - present) after function signature */ + present) after function signature. + It applies only for the toplevel + function type. */ +#define DMGL_RET_DROP (1 << 6) /* Suppress function return types, even + if present. It applies only for the + toplevel function type. */ #define DMGL_AUTO (1 << 8) #define DMGL_GNU (1 << 9) --- a/libiberty/cp-demangle.c +++ b/libiberty/cp-demangle.c @@ -3917,10 +3917,14 @@ d_print_comp (struct d_print_info *dpi, const struct demangle_component *dc, case DEMANGLE_COMPONENT_FUNCTION_TYPE: { if ((options & DMGL_RET_POSTFIX) != 0) - d_print_function_type (dpi, dc, dpi->modifiers, options); + d_print_function_type (dpi, dc, dpi->modifiers, + options & ~(DMGL_RET_POSTFIX | DMGL_RET_DROP)); /* Print return type if present */ - if (d_left (dc) != NULL) + if (d_left (dc) != NULL && (options & DMGL_RET_POSTFIX) != 0) + d_print_comp (dpi, d_left (dc), + options & ~(DMGL_RET_POSTFIX | DMGL_RET_DROP)); + else if (d_left (dc) != NULL && (options & DMGL_RET_DROP) == 0) { struct d_print_mod dpm; @@ -3932,7 +3936,8 @@ d_print_comp (struct d_print_info *dpi, const struct demangle_component *dc, dpm.printed = 0; dpm.templates = dpi->templates; - d_print_comp (dpi, d_left (dc), options); + d_print_comp (dpi, d_left (dc), + options & ~(DMGL_RET_POSTFIX | DMGL_RET_DROP)); dpi->modifiers = dpm.next; @@ -3946,7 +3951,8 @@ d_print_comp (struct d_print_info *dpi, const struct demangle_component *dc, } if ((options & DMGL_RET_POSTFIX) == 0) - d_print_function_type (dpi, dc, dpi->modifiers, options); + d_print_function_type (dpi, dc, dpi->modifiers, + options & ~(DMGL_RET_POSTFIX | DMGL_RET_DROP)); return; } --- a/libiberty/testsuite/demangle-expected +++ b/libiberty/testsuite/demangle-expected @@ -3959,6 +3959,33 @@ f(decltype(nullptr)) --format=gnu-v3 _ZN5aaaaa6bbbbbb5cccccIN23ddddddddddddddddddddddd3eeeENS2_4ffff16ggggggggggggggggENS0_9hhhhhhhhhES6_S6_S6_S6_S6_S6_S6_EE aaaaa::bbbbbb::ccccc +--format=gnu-v3 +_Z5outerIsEcPFilE +char outer(int (*)(long)) +--format=gnu-v3 +_Z5outerPFsiEl +outer(short (*)(int), long) +--format=gnu-v3 +_Z6outer2IsEPFilES1_ +int (*outer2(int (*)(long)))(long) +--format=gnu-v3 --ret-postfix +_Z5outerIsEcPFilE +outer(int (*)(long))char +--format=gnu-v3 --ret-postfix +_Z5outerPFsiEl +outer(short (*)(int), long) +--format=gnu-v3 --ret-postfix +_Z6outer2IsEPFilES1_ +outer2(int (*)(long))int (*)(long) +--format=gnu-v3 --ret-drop +_Z5outerIsEcPFilE +outer(int (*)(long)) +--format=gnu-v3 --ret-drop +_Z5outerPFsiEl +outer(short (*)(int), long) +--format=gnu-v3 --ret-drop +_Z6outer2IsEPFilES1_ +outer2(int (*)(long)) # # Ada (GNAT) tests. # --- a/libiberty/testsuite/test-demangle.c +++ b/libiberty/testsuite/test-demangle.c @@ -159,6 +159,7 @@ exp: %s\n", output is an integer representing ctor_kind. --is-v3-dtor Likewise, but for dtors. --ret-postfix Passes the DMGL_RET_POSTFIX option + --ret-drop Passes the DMGL_RET_DROP option For compatibility, just in case it matters, the options line may be empty, to mean --format=auto. If it doesn't start with --, then it @@ -174,7 +175,7 @@ main(argc, argv) int no_params; int is_v3_ctor; int is_v3_dtor; - int ret_postfix; + int ret_postfix, ret_drop; struct line format; struct line input; struct line expect; @@ -209,6 +210,7 @@ main(argc, argv) no_params = 0; ret_postfix = 0; + ret_drop = 0; is_v3_ctor = 0; is_v3_dtor = 0; if (format.data[0] == '\0') @@ -265,6 +267,8 @@ main(argc, argv) is_v3_dtor = 1; else if (strcmp (opt, "--ret-postfix") == 0) ret_postfix = 1; + else if (strcmp (opt, "--ret-drop") == 0) + ret_drop = 1; else { printf ("FAIL at line %d: unrecognized option %s\n", @@ -307,9 +311,9 @@ main(argc, argv) cplus_demangle_set_style (style); - result = cplus_demangle (inp, - DMGL_PARAMS|DMGL_ANSI|DMGL_TYPES - |(ret_postfix ? DMGL_RET_POSTFIX : 0)); + result = cplus_demangle (inp, (DMGL_PARAMS | DMGL_ANSI | DMGL_TYPES + | (ret_postfix ? DMGL_RET_POSTFIX : 0) + | (ret_drop ? DMGL_RET_DROP : 0))); if (result ? strcmp (result, expect.data)