From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from simark.ca by simark.ca with LMTP id MZJ1BMjVdGr29w0AWB0awg (envelope-from ) for ; Thu, 06 Aug 2026 14:43:20 -0400 Authentication-Results: simark.ca; dkim=pass (1024-bit key; unprotected) header.d=redhat.com header.i=@redhat.com header.a=rsa-sha256 header.s=mimecast20190719 header.b=HP+mmyK6; dkim-atps=neutral Received: by simark.ca (Postfix, from userid 112) id 0CEDB1E09B; Thu, 06 Aug 2026 14:43:20 -0400 (EDT) X-Spam-Checker-Version: SpamAssassin 4.0.1 (2024-03-25) on simark.ca X-Spam-Level: X-Spam-Status: No, score=-6.4 required=5.0 tests=ARC_SIGNED,ARC_VALID,BAYES_00, DKIMWL_WL_HIGH,DKIM_SIGNED,DKIM_VALID,DKIM_VALID_AU,MAILING_LIST_MULTI, RCVD_IN_DNSWL_MED autolearn=ham autolearn_force=no version=4.0.1 Received: from vm01.sourceware.org (vm01.sourceware.org [IPv6:2620:52:6:3111::32]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange x25519 server-signature ECDSA (prime256v1) server-digest SHA256) (No client certificate requested) by simark.ca (Postfix) with ESMTPS id 29D6F1E09B for ; Thu, 06 Aug 2026 14:43:18 -0400 (EDT) Received: from vm01.sourceware.org (localhost [IPv6:::1]) by sourceware.org (Postfix) with ESMTP id AD3204BB5888 for ; Thu, 6 Aug 2026 18:43:16 +0000 (GMT) DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org AD3204BB5888 Authentication-Results: sourceware.org; dkim=pass (1024-bit key, unprotected) header.d=redhat.com header.i=@redhat.com header.a=rsa-sha256 header.s=mimecast20190719 header.b=HP+mmyK6 Received: from us-smtp-delivery-124.mimecast.com (us-smtp-delivery-124.mimecast.com [170.10.129.124]) by sourceware.org (Postfix) with ESMTP id E69634BA9020 for ; Thu, 6 Aug 2026 18:42:47 +0000 (GMT) DMARC-Filter: OpenDMARC Filter v1.4.2 sourceware.org E69634BA9020 Authentication-Results: sourceware.org; dmarc=pass (p=quarantine dis=none) header.from=redhat.com Authentication-Results: sourceware.org; spf=pass smtp.mailfrom=redhat.com ARC-Filter: OpenARC Filter v1.0.0 sourceware.org E69634BA9020 Authentication-Results: sourceware.org; arc=none smtp.remote-ip=170.10.129.124 ARC-Seal: i=1; a=rsa-sha256; d=sourceware.org; s=key; t=1786041768; cv=none; b=iN7meHAFA6I/BW/tRr0hkk4M2zTVDFG0irlaTi+d3RoX36qjl2jlo1m6tgsJfxQYvSbjLovphr3zTbCR/8fIHzwbWbF+s+x6ZhR7l4IHv0eVGdyobUA39aRMVHs5oGxFq978TjxxW0HoNDsS054wxngSs2hRduRybblGizNRVxU= ARC-Message-Signature: i=1; a=rsa-sha256; d=sourceware.org; s=key; t=1786041768; c=relaxed/simple; bh=v1YrwJHOicz6ESdaQUibvjnM156Qi3ctEaYqEvOeZ+g=; h=DKIM-Signature:Message-ID:Date:MIME-Version:Subject:To:From; b=nxcxa2ZKGVqYcPSE1DEtQNSps6GstHHrZk1O7Y7avngLPSHvRGAsD4XNNnfQ9nJBGzBQyf13BK6cqSNHCc27XvuGUu83gzdTQ8TBKTNHqyiD/lA4s0ysYHAXCNbvdrMkp1MS7t775q6XjrB7/AEYT3FyIEUwdl6tK0c4Gg9mOq0= ARC-Authentication-Results: i=1; sourceware.org; dkim=pass (1024-bit key, unprotected) header.d=redhat.com header.i=@redhat.com header.a=rsa-sha256 header.s=mimecast20190719 header.b=HP+mmyK6 DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org E69634BA9020 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1786041767; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:mime-version:mime-version:content-type:content-type: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=8fDtVSPgtrYaCw6owhetcM2v7cZ8ARHoa1gjjDjXWws=; b=HP+mmyK685S12AnP/gOnMHA/GoaHV1sG6O+iJYuI1jQocE7oQyiPHCT2VPShmCH6pOprZm yTxW24RKZMWs7cmDQ+W3E6lS042CqzndIZbkxmOsASH36GcPypj9ANwDF21JzLpDAxjTas aHYF+hZ93qjQDkxhwkiYajl6LJgdFDs= Received: from mail-pf1-f197.google.com (mail-pf1-f197.google.com [209.85.210.197]) by relay.mimecast.com with ESMTP with STARTTLS (version=TLSv1.3, cipher=TLS_AES_256_GCM_SHA384) id us-mta-629-Tf6CJfb6NGCFZosg2COfEQ-1; Thu, 06 Aug 2026 14:42:46 -0400 X-MC-Unique: Tf6CJfb6NGCFZosg2COfEQ-1 X-Mimecast-MFC-AGG-ID: Tf6CJfb6NGCFZosg2COfEQ_1786041765 Received: by mail-pf1-f197.google.com with SMTP id d2e1a72fcca58-84eccf9d899so3379602b3a.2 for ; Thu, 06 Aug 2026 11:42:45 -0700 (PDT) X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20251104; t=1786041765; x=1786646565; h=content-transfer-encoding:content-type:in-reply-to:from :content-language:references:to:subject:user-agent:mime-version:date :message-id:x-gm-gg:x-gm-message-state:from:to:cc:subject:date :message-id:reply-to:content-type; bh=8fDtVSPgtrYaCw6owhetcM2v7cZ8ARHoa1gjjDjXWws=; b=ESsvS4yT+HRHUGLxfWEC+m0+CjaBf9EO7fK1CSAxevxRewmZtnNuDaIKV4Tr5Xtgz5 Ya9EhOujqgHxMmnQuuqqS/ivLViuGme414htyqqa3B8birRMs3h45tAlNroqAFAeu4/k LOiavgm5BvajTqEqG+NHCcygboewLWPaxkRV7C/oizGWSRsgxABTsX51VrXq/D3282lb J/98ycPIPktYzj9PQoj43gjbYfT4lMNG15+sSrJQhDjv4Qq26qZbwS+1Nuw7pCRQBbYi kBXgAz++XEaVLT8E2h24ARxH6dhpmPV9YJAmNJrwDDI6kRZfRJHQMn3+4oQsuyoEFuo+ bQxQ== X-Gm-Message-State: AOJu0YxA6iMd9b3iA4qlSj161/uMKf2VRiHw5V64JWRyJvtmreYcngj7 aqNI1tN9C+n7wnN99QcnbGkUlyVQe3HoeXvG85Ro2gt4Y8XcTKHZn1zxulV70V8a9zFxqcyjQ3g qaPacpGR7Mjik+It9C6+12HC63m0AVPNUD6OuVJDB4FrzSi83Ga/zpxfOG9pSY31IQ9IgaukJOt MT6dfbDSauLNRRmFMG2R9mqz5P/ImaE/LHJE2cLPusgqisJQuY X-Gm-Gg: AR+sD10ZKp94OTSQh6rMc13bQmFi794qqGkstKnxbQxatbRe4TqUTYImp1zkoiRw6U6 U66lkRqOFmluTiV3Y3P1wlAh8Ze/7P9t6YcDfEWofUKnbUJxnbn3Jg7tTS4L5EDo2pklf9LoqoN to96qmtCHMhM6rGoFEWtaWWuf0R30CCpzBO8zfV6HGwwfEnyktcU35R5UJxjc4nPxwVYzwAIL4U 1QFBa9Kh6VKpBNDeUpl0HMJ/H8YJMsIxxtsg4tSkyx7vzTxAlncYnTHXULTDMmgvVsHXJCswSAf 5iyL5uV+cAu1lOS8DBpx4lrTL1S//uZ4sSPPA11kYrn7hAQi1Qdn2Ldrl7lw/gc5dYmaJx8exCh jacZS75VJzg== X-Received: by 2002:a05:6a21:3a82:b0:3c3:791e:5e18 with SMTP id adf61e73a8af0-3cb85df09cbmr20556077637.7.1786041764780; Thu, 06 Aug 2026 11:42:44 -0700 (PDT) X-Received: by 2002:a05:6a21:3a82:b0:3c3:791e:5e18 with SMTP id adf61e73a8af0-3cb85df09cbmr20555974637.7.1786041764020; Thu, 06 Aug 2026 11:42:44 -0700 (PDT) Received: from ?IPV6:2804:14d:8084:993e::75d? ([2804:14d:8084:993e::75d]) by smtp.gmail.com with ESMTPSA id 5a478bee46e88-315863b0c61sm32110872eec.6.2026.08.06.11.42.41 (version=TLS1_3 cipher=TLS_AES_128_GCM_SHA256 bits=128/128); Thu, 06 Aug 2026 11:42:43 -0700 (PDT) Message-ID: Date: Thu, 6 Aug 2026 15:42:38 -0300 MIME-Version: 1.0 User-Agent: Mozilla Thunderbird Subject: Re: [PING][PATCH v8] gdb: Print linker namespace when showing a frame To: gdb-patches@sourceware.org, Guinevere Larsen References: <20260507134822.701336-1-guinevere@redhat.com> <9bdde366-1efd-4802-8158-8be60bcd3614@redhat.com> From: Guinevere Larsen In-Reply-To: <9bdde366-1efd-4802-8158-8be60bcd3614@redhat.com> X-Mimecast-Spam-Score: 0 X-Mimecast-MFC-PROC-ID: RyBHcBK0AMyVX-oQYvXaWhCTzMwM_KKsgDLTeSS5X4Q_1786041765 X-Mimecast-Originator: redhat.com Content-Language: en-US Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: 8bit X-BeenThere: gdb-patches@sourceware.org X-Mailman-Version: 2.1.30 Precedence: list List-Id: Gdb-patches mailing list List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: gdb-patches-bounces~public-inbox=simark.ca@sourceware.org Ping :) On 6/22/26 5:34 PM, Guinevere Larsen wrote: > Ping :) > > On 5/7/26 10:48 AM, Guinevere Larsen wrote: >> When a user is stopped in a private linker namespace, the only way for >> them to realize that is using the _linker_namespace convenience >> variable. While serviceable, this is a sub-optimal solution, as most >> users are unaware of convenience variables. >> >> This commit introduces a new way for users to be informed of the linker >> namespace of a function, by printing it along with the function name. >> This is done by using the proposed syntax for symbols and locations, >> like so: >> >>    #0  [[0]]::main () >> >> This is done by introducing a new function on solib.h, >> linker_namespace_prefix, that returns either an empty string (if only >> one namespace is active, or we can't determine the PC for the frame), or >> the correct identifier. This should be printed manually every time GDB >> will print a frame, so that we don't confuse other systems that look at >> function names. >> >> A previous revision would break GDB's ability to strip the parameters of >> C++ symbols, so this commit also adds a C++ SO to the test. However, >> libstdc++ never gets unloaded, so many changes were required to make the >> rest of the test still valid. >> >> Reviewed-By: Eli Zaretskii >> --- >>   gdb/NEWS                                      |  4 + >>   gdb/doc/gdb.texinfo                           |  9 +++ >>   gdb/solib.c                                   | 43 ++++++++--- >>   gdb/solib.h                                   | 13 ++++ >>   gdb/stack.c                                   |  2 + >>   .../gdb.base/dlmopen-ns-ids-framefilter.py    | 31 ++++++++ >>   .../gdb.base/dlmopen-ns-ids-libc++.cc         | 36 +++++++++ >>   gdb/testsuite/gdb.base/dlmopen-ns-ids-main.c  | 19 ++++- >>   gdb/testsuite/gdb.base/dlmopen-ns-ids.exp     | 74 ++++++++++++++++--- >>   gdb/testsuite/gdb.mi/mi-dlmopen.exp           |  2 +- >>   10 files changed, 206 insertions(+), 27 deletions(-) >>   create mode 100644 >> gdb/testsuite/gdb.base/dlmopen-ns-ids-framefilter.py >>   create mode 100644 gdb/testsuite/gdb.base/dlmopen-ns-ids-libc++.cc >> >> diff --git a/gdb/NEWS b/gdb/NEWS >> index 480e1854002..941178c06a4 100644 >> --- a/gdb/NEWS >> +++ b/gdb/NEWS >> @@ -78,6 +78,10 @@ >>   * The Windows native target now supports non-stop mode.  This feature >>     requires Windows 10 or later. >>   +* When debugging an inferior with multiple linker namespaces, >> functions >> +  will be printed like so [[N]]::foo, where N is the namespace that >> +  contains foo.  For example [[0]]::main. >> + >>   * New targets >>     GNU/Linux/MicroBlaze (gdbserver) microblazeel-*linux* >> diff --git a/gdb/doc/gdb.texinfo b/gdb/doc/gdb.texinfo >> index ab0216ff477..98a43f3a57e 100644 >> --- a/gdb/doc/gdb.texinfo >> +++ b/gdb/doc/gdb.texinfo >> @@ -22431,6 +22431,15 @@ possibility that those objects may >> cross-talk.  Each set of isolated >>   shared objects is said to belong to a ``namespace'', and linker >> related >>   actions such as relocations do not cross namespace boundaries. >>   +When multiple linker namespaces are active, @value{GDBN} will >> print the >> +identifier of the namespace that contains the function, when a function >> +is printed.  It will look like this: >> + >> +@smallexample >> +#0  [[1]]::inc (n=0) at dlmopen-ns-ids-lib.c >> +#1  0x0000000000400591 in [[0]]::main () at dlmopen-ns-ids-main.c >> +@end smallexample >> + >>   @kindex info dll >>   @item info dll @var{regex} >>   This is an alias of @code{info sharedlibrary}. >> diff --git a/gdb/solib.c b/gdb/solib.c >> index 782b844aa2f..add842b5771 100644 >> --- a/gdb/solib.c >> +++ b/gdb/solib.c >> @@ -1796,6 +1796,36 @@ solib_linker_namespace_count (program_space >> *pspace) >>     return 0; >>   } >>   +/* See solib.h.  */ >> + >> +int >> +linker_namespace_for_addr (CORE_ADDR addr, program_space *pspace) >> +{ >> +  for (const solib &so : pspace->solibs ()) >> +    if (solib_contains_address_p (so, addr)) >> +      { >> +    if (so.ops ().supports_namespaces ()) >> +      return so.ops ().find_solib_ns (so); >> +    break; >> +      } >> + >> +  return 0; >> +} >> + >> +/* See solib.h.  */ >> + >> +std::string >> +linker_namespace_prefix (const frame_info_ptr &frame) >> +{ >> +  std::optional pc = get_frame_pc_if_available (frame); >> +  program_space *pspace = get_frame_program_space (frame); >> +  if (!pc.has_value () || solib_linker_namespace_count (pspace) <= 1) >> +    return std::string (""); >> + >> +  return string_printf ("[[%d]]::", linker_namespace_for_addr >> (pc.value (), >> +                                   pspace)); >> +} >> + >>   /* Implementation of the linker_namespace convenience variable. >>        This returns the GDB internal identifier of the linker namespace, >> @@ -1804,19 +1834,10 @@ solib_linker_namespace_count (program_space >> *pspace) >>     static value * >>   linker_namespace_make_value (gdbarch *gdbarch, internalvar *var, >> -                     void *ignore) >> +                 void *ignore) >>   { >> -  int nsid = 0; >>     CORE_ADDR curr_pc = get_frame_pc (get_selected_frame ()); >> - >> -  for (const solib &so : current_program_space->solibs ()) >> -    if (solib_contains_address_p (so, curr_pc)) >> -      { >> -    if (so.ops ().supports_namespaces ()) >> -      nsid = so.ops ().find_solib_ns (so); >> - >> -    break; >> -      } >> +  int nsid = linker_namespace_for_addr (curr_pc, >> current_program_space); >>       /* If the PC is not in an SO, or the solib_ops doesn't support >>        linker namespaces, the inferior is in the default namespace.  */ >> diff --git a/gdb/solib.h b/gdb/solib.h >> index 564655c4a57..39931bb1b93 100644 >> --- a/gdb/solib.h >> +++ b/gdb/solib.h >> @@ -334,6 +334,19 @@ extern const char *solib_name_from_address >> (struct program_space *, CORE_ADDR); >>     extern bool solib_contains_address_p (const solib &, CORE_ADDR); >>   +/* Given the address ADDR, return which linker namespace contains >> +   this address in PSPACE.  If ADDR is present in multiple namespaces, >> +   an arbitrary one is returned.  If ADDR is not present in any >> +   namespace, or if the target doesn't support linker namespaces, >> +   returns 0.  */ >> + >> +extern int linker_namespace_for_addr (CORE_ADDR addr, program_space >> *pspace); >> + >> +/* Calculate the linker namespace identifier, and turn it into a prefix >> +   to be used to print a frame.  */ >> + >> +extern std::string linker_namespace_prefix (const frame_info_ptr >> &frame); >> + >>   /* Return whether the data starting at VADDR, size SIZE, must be kept >>      in a core file for shared libraries loaded before "gcore" is used >>      to be handled correctly when the core file is loaded.  This only >> diff --git a/gdb/stack.c b/gdb/stack.c >> index 7329430adab..ac270718a91 100644 >> --- a/gdb/stack.c >> +++ b/gdb/stack.c >> @@ -1306,6 +1306,7 @@ print_frame (struct ui_out *uiout, >>       annotate_frame_function_name (); >>         string_file stb; >> +    gdb_puts (linker_namespace_prefix (frame).c_str (), &stb); >>       gdb_puts (funname ? funname.get () : "??", &stb); >>       uiout->field_stream ("func", stb, function_name_style.style ()); >>       uiout->wrap_hint (3); >> @@ -1466,6 +1467,7 @@ info_frame_command_core (const frame_info_ptr >> &fi, bool selected_frame_p) >>       fputs_styled ("", metadata_style.style (), >> gdb_stdout); >>       gdb_stdout->wrap_here (3); >> +  gdb_puts (linker_namespace_prefix (fi).c_str ()); >>     if (funname.get () != nullptr) >>       { >>         gdb_puts (" in "); >> diff --git a/gdb/testsuite/gdb.base/dlmopen-ns-ids-framefilter.py >> b/gdb/testsuite/gdb.base/dlmopen-ns-ids-framefilter.py >> new file mode 100644 >> index 00000000000..11d96fcfbc4 >> --- /dev/null >> +++ b/gdb/testsuite/gdb.base/dlmopen-ns-ids-framefilter.py >> @@ -0,0 +1,31 @@ >> +# Copyright (C) 2026 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 . >> + >> + >> +# This file is part of the GDB testsuite.  It tests if printing the >> +# linker namespace id is affected by a null frame filter >> + >> +import gdb >> + >> + >> +class NullFilter: >> +    def __init__(self): >> +        self.name = "null filter" >> +        self.priority = 100 >> +        self.enabled = True >> +        gdb.frame_filters[self.name] = self >> + >> +    def filter(self, iterator): >> +        return iterator >> diff --git a/gdb/testsuite/gdb.base/dlmopen-ns-ids-libc++.cc >> b/gdb/testsuite/gdb.base/dlmopen-ns-ids-libc++.cc >> new file mode 100644 >> index 00000000000..9abf3f25096 >> --- /dev/null >> +++ b/gdb/testsuite/gdb.base/dlmopen-ns-ids-libc++.cc >> @@ -0,0 +1,36 @@ >> +/* This testcase is part of GDB, the GNU debugger. >> + >> +   Copyright 2025-2026 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 >> . >> + >> +*/ >> + >> +int gdb_dlmopen_glob = 0; >> + >> +int >> +cpp_inc (int n) >> +{ >> +    int amount = gdb_dlmopen_glob; >> +    return n + amount;  /* bp.inc.  */ >> +} >> + >> +extern "C" { >> +  __attribute__((visibility ("default"))) >> +  int >> +  inc (int n) >> +  { >> +    return cpp_inc (n); >> +  } >> +} >> diff --git a/gdb/testsuite/gdb.base/dlmopen-ns-ids-main.c >> b/gdb/testsuite/gdb.base/dlmopen-ns-ids-main.c >> index 4f08a5135b2..f13eb9168de 100644 >> --- a/gdb/testsuite/gdb.base/dlmopen-ns-ids-main.c >> +++ b/gdb/testsuite/gdb.base/dlmopen-ns-ids-main.c >> @@ -27,7 +27,7 @@ >>   int >>   main (void) >>   { >> -  void *handle[4]; >> +  void *handle[5]; >>     int (*fun) (int); >>     Lmid_t lmid; >>     int dl; >> @@ -41,20 +41,31 @@ main (void) >>     handle[2] = dlmopen (LM_ID_NEWLM, DSO_NAME, RTLD_LAZY | RTLD_LOCAL); >>     assert (handle[2] != NULL); >>   +  /* We open the C++ SO on the main namespace because libstdc++ >> marks a >> +     lot of objects as GNU_UNIQUE meaning that they effectively >> can't be >> +     unloaded, so the final namespace would never be fully >> unloaded.  */ >> +  handle[3] = dlmopen (LM_ID_BASE, CPP_DSO_NAME, RTLD_LAZY | >> RTLD_LOCAL); >> +  assert (handle[3] != NULL); >> + >>     for (dl = 2; dl >= 0; dl--) >>       { >>         fun = dlsym (handle[dl], "inc"); >>         fun (dl); >>       } >> +  /* C++ call, to make sure the linker namespace doesn't break anything >> +     of the C++ handling of symbols.  */ >> +  fun = dlsym (handle[3], "inc"); >> +  fun (-1); >>       dlclose (handle[0]); /* TAG: first dlclose */ >>     dlclose (handle[1]); /* TAG: second dlclose */ >>     dlclose (handle[2]); /* TAG: third dlclose */ >> +  dlclose (handle[3]); /* TAG: fourth dlclose */ >>   -  handle[3] = dlmopen (LM_ID_NEWLM, DSO_NAME, RTLD_LAZY | >> RTLD_LOCAL); >> -  dlinfo (handle[3], RTLD_DI_LMID, &lmid); >> +  handle[4] = dlmopen (LM_ID_NEWLM, DSO_NAME, RTLD_LAZY | RTLD_LOCAL); >> +  dlinfo (handle[4], RTLD_DI_LMID, &lmid); >>   -  dlclose (handle[3]); /* TAG: fourth dlclose */ >> +  dlclose (handle[4]); /* TAG: final dlclose */ >>       return 0; >>   } >> diff --git a/gdb/testsuite/gdb.base/dlmopen-ns-ids.exp >> b/gdb/testsuite/gdb.base/dlmopen-ns-ids.exp >> index e40bc213454..cc1696f4229 100644 >> --- a/gdb/testsuite/gdb.base/dlmopen-ns-ids.exp >> +++ b/gdb/testsuite/gdb.base/dlmopen-ns-ids.exp >> @@ -21,20 +21,29 @@ >>     require allow_dlmopen_tests >>   -standard_testfile -main.c -lib.c >> +standard_testfile -main.c -lib.c -framefilter.py -libc++.cc >>     set srcfile_lib $srcfile2 >> +set srcfile_libcpp $srcfile4 >>   set so_name dlmopen-lib.so >> +set cpp_so_name dlmopen-libc++.so >>   set binfile_lib [standard_output_file $so_name] >> +set binfile_libcpp [standard_output_file $cpp_so_name] >>     if { [build_executable "build shlib" $binfile_lib $srcfile_lib \ >>         [list debug shlib]] == -1 } { >>       return >>   } >>   +if { [build_executable "build shlib" $binfile_libcpp >> $srcfile_libcpp \ >> +      [list debug shlib c++]] == -1 } { >> +    return >> +} >> + >>   if { [build_executable "failed to build" $testfile $srcfile \ >>         [list additional_flags=-DDSO_NAME=\"$binfile_lib\" \ >> -           shlib_load debug]] } { >> +        additional_flags=-DCPP_DSO_NAME=\"$binfile_libcpp\" \ >> +        shlib_load debug]] } { >>       return >>   } >>   @@ -107,11 +116,13 @@ proc check_no_duplicates {} { >>   proc get_first_so_ns {} { >>       set ns -1 >>       set lib_regexp [string_to_regexp ${::binfile_lib}] >> +    set libcpp_regexp [string_to_regexp ${::binfile_libcpp}] >> +    set both_regexp "($lib_regexp|$libcpp_regexp)" >>       gdb_test_multiple "info sharedlibrary $::so_name" "get SO >> namespace" -lbl { >>       -re "\r\nFrom\\s+To\\s+\(Linker NS\\s+\)?Syms\\s+Read\\s+Shared >> Object Library(?=\r\n)" { >>           exp_continue >>       } >> -    -re >> "\r\n$::hex\\s+$::hex\\s+($::decimal)\\s+\[^\r\n]+${lib_regexp}(?=\r\n)" >> { >> +    -re >> "\r\n$::hex\\s+$::hex\\s+($::decimal)\\s+\[^\r\n]+${both_regexp}(?=\r\n)" >> { >>           if {$ns == -1} { >>           set ns $expect_out(1,string) >>           } >> @@ -155,8 +166,8 @@ proc test_info_shared {} { >>       gdb_test "next" ".*third dlclose.*" "close second library" >>       gdb_assert {[get_first_so_ns] == 3} "before closing two libraries" >>   -    gdb_breakpoint [gdb_get_line_number "TAG: fourth dlclose"] >> -    gdb_continue_to_breakpoint "TAG: fourth dlclose" >> +    gdb_breakpoint [gdb_get_line_number "TAG: final dlclose"] >> +    gdb_continue_to_breakpoint "TAG: final dlclose" >>       # As of writing this test, glibc's LMID is just an index on an >> array of >>       # namespaces.  After closing a namespace, requesting a new one >> will >>       # return the index of the lowest-closed namespace, so this will >> likely >> @@ -192,6 +203,11 @@ proc_with_prefix test_conv_vars {} { >>       gdb_test "print \$_linker_namespace" ".* = 0" \ >>       "Still in the default namespace" >>   +    # There should be no namespace ID visible, since there's >> +    # only one namespace loaded. >> +    gdb_test "backtrace" "\#0\\s+main .*" \ >> +    "No namespace ID in backtrace" >> + >>       gdb_breakpoint "inc" allow-pending >>       gdb_breakpoint [gdb_get_line_number "TAG: first dlclose"] >>   @@ -200,14 +216,41 @@ proc_with_prefix test_conv_vars {} { >>         gdb_test "print \$_linker_namespace" ".* = $dl" \ >>           "Verify we're in namespace $dl" >> + >> +    gdb_test "frame" "\#0\\s+\\\[\\\[$dl\\\]\\\]::inc.*" \ >> +        "Namespace ID in the frame" >>       } >>         # Check that we display the namespace of the selected >>       # frame, not the lowermost one. >> -    gdb_test "up" "\#1.*in main.*" >> +    gdb_test "up" "\#1.*in \\\[\\\[0\\\]\\\]::main.*" >>       gdb_test "print \$_linker_namespace" ".* = 0" \ >>       "print namespace of selected frame" >>   +    gdb_test "backtrace" [multi_line \ >> +        "\#0\\s+\\\[\\\[1\\\]\\\]::inc \[^\r\n\]*" \ >> +        "\#1\\s+$::hex in \\\[\\\[0\\\]\\\]::main \[^\r\n\]*" ] \ >> +    "raw backtrace" >> + >> +    set remote_python_file [gdb_remote_download host \ >> +                $::srcdir/$::subdir/$::srcfile3] >> +    gdb_test_no_output "source $remote_python_file" "load frame filter" >> +    gdb_test "backtrace" [multi_line \ >> +        "\#0\\s+\\\[\\\[1\\\]\\\]::inc \[^\r\n\]*" \ >> +        "\#1\\s+$::hex in \\\[\\\[0\\\]\\\]::main \[^\r\n\]*" ] \ >> +    "With null frame filter" >> + >> +    # Ensure that the code for identifying and removing parameters for >> +    # C++ symbols still works with linker namespace IDs. >> +    gdb_continue_to_breakpoint "inc on c++ SO" >> +    gdb_test "step" "\\\[\\\[0\\\]\\\]::cpp_inc \\\(n=-1\\\).*" >> "step into c++ symbol" >> +    gdb_test "backtrace" [multi_line \ >> +        "\#0\\s+\\\[\\\[0\\\]\\\]::cpp_inc \\\(n=-1\\\)\[^\r\n\]*" \ >> +        "\#1\\s+$::hex in \\\[\\\[0\\\]\\\]::inc >> \\\(n=-1\\\)\[^\r\n\]*" \ >> +        "\#2\\s+$::hex in \\\[\\\[0\\\]\\\]::main \[^\r\n\]*" ] \ >> +    "bt through c++ symbols" >> + >> +    # Check that everything still works as we close namespaces. >>       gdb_continue_to_breakpoint "first dlclose" >>       gdb_test "print \$_linker_namespace_count" "4" "all SOs loaded" >>   @@ -264,15 +307,23 @@ proc test_info_linker_namespaces {} { >>         # Some systems may add libc and libm to every loaded namespace, >>       # others may load only one or neither, because the SO doesn't >> -    # actually use either library.  The best we can do is check if >> -    # we found the dynamic linker, and up to 2 more libraries. >> -    gdb_assert {$n_libraries <= 3}  "the correct number of libraries >> was reported" >> +    # actually use either library.  Similarly, because of the c++ >> +    # SO, libstdc++ and libgcc_s may be included.  The best we can >> +    # do is check if  we found the dynamic linker, the SO, and up to >> +    # 4 more libraries.  Ensuring that only the right one shows up is >> +    # in the foreach block. >> +    gdb_assert {$n_libraries <= 6}  "the correct number of libraries >> was reported" >>         set binfile_lib_re [string_to_regexp $::binfile_lib] >>   -    foreach_with_prefix ns {1 2 3} { >> +    foreach_with_prefix ns {1 2 3 0} { >>       set found_test_so false >>       set n_libraries 999 >> +    set expected_libs 4 >> +    if {$ns == 0} { >> +        set expected_libs 6 >> +        set binfile_lib_re [string_to_regexp $::binfile_libcpp] >> +    } >>         gdb_test_multiple "info linker-namespaces $ns" "print >> namespace $ns" { >>           -re ".*($::decimal) librar(?:y|ies) loaded in linker >> namespace $ns:\r\n" { >> @@ -290,7 +341,8 @@ proc test_info_linker_namespaces {} { >>       # actually use either library.  The best we can do is check if >>       # we found the dynamic linker, the test SO, and maybe up to 2 >>       # more libraries. >> -    gdb_assert {$n_libraries <= 4} "the correct number of libraries >> was reported" >> +    gdb_assert {$n_libraries <= $expected_libs} \ >> +        "the correct number of libraries was reported" >>       gdb_assert {$found_test_so} "this testfile's SO was reported" >>       } >>   diff --git a/gdb/testsuite/gdb.mi/mi-dlmopen.exp >> b/gdb/testsuite/gdb.mi/mi-dlmopen.exp >> index ff854ac7dd4..832c6ea6fe6 100644 >> --- a/gdb/testsuite/gdb.mi/mi-dlmopen.exp >> +++ b/gdb/testsuite/gdb.mi/mi-dlmopen.exp >> @@ -156,7 +156,7 @@ proc check_solib_unload_events {} { >>       -disp keep -func main -file ".*$::srcfile" -line $::bp_main >>         # Run past all the dlopen and dlmopen calls. >> -    mi_execute_to "exec-continue" "breakpoint-hit" main "" ".*" >> $::bp_loaded \ >> +    mi_execute_to "exec-continue" "breakpoint-hit" {\[\[0\]\]::main} >> "" ".*" $::bp_loaded \ >>       {"" "disp=\"keep\""} "continue until all libraries are loaded" >>         # Check that the dynamic linker has now been loaded multiple >> times. >> >> base-commit: 9e6e0c3cd9e6d5ee4953f72a805cff5f150aa2fd > > -- Cheers, Guinevere Larsen it/its she/her (deprecated)