From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: by sourceware.org (Postfix, from userid 1726) id 095CB3858D32; Sun, 12 Feb 2023 06:21:23 +0000 (GMT) DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org 095CB3858D32 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=sourceware.org; s=default; t=1676182883; bh=WT4T+NszMJszp1Im/UivrY0czofFF8I0lcjlmutfEMs=; h=From:To:Subject:Date:From; b=XozvtnLK6+s/g+4iMNYRs3o15e18x8Wf2660NqVGJ6bCIQHZ5QLj4mPeQ2pj3U9bk 0gFNe1Dbw/91h3ZY2zvd66Raw5o2PZ5iFx82aXU7CPNYUCFzrMYHucuXb+uCTMQ0wi SMmImqskUc7aNyq0gORD9jn93NVEEaitRPWfiuGQ= Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable From: Andrew Burgess To: gdb-cvs@sourceware.org Subject: [binutils-gdb] gdb/c++: fix handling of breakpoints on @plt symbols X-Act-Checkin: binutils-gdb X-Git-Author: Andrew Burgess X-Git-Refname: refs/heads/master X-Git-Oldrev: 2ecee236752932672fb3d6cd63c6976927f747d8 X-Git-Newrev: f0bdf68d3fb6db1dd2b83e07062e2104cdb785c2 Message-Id: <20230212062123.095CB3858D32@sourceware.org> Date: Sun, 12 Feb 2023 06:21:23 +0000 (GMT) List-Id: https://sourceware.org/git/gitweb.cgi?p=3Dbinutils-gdb.git;h=3Df0bdf68d3fb6= db1dd2b83e07062e2104cdb785c2 commit f0bdf68d3fb6db1dd2b83e07062e2104cdb785c2 Author: Andrew Burgess Date: Fri Dec 16 15:15:42 2022 +0000 gdb/c++: fix handling of breakpoints on @plt symbols =20 This commit should fix PR gdb/20091, PR gdb/17201, and PR gdb/17071. Additionally, PR gdb/17199 relates to this area of code, but is more of a request to refactor some parts of GDB, this commit does not address that request, but it is probably worth reading that PR when looking at this commit. =20 When the current language is C++, and the user places a breakpoint on a function in a shared library, GDB will currently find two locations for the breakpoint, one location will be within the function itself as we would expect, but the other location will be within the PLT table for the call to the named function. Consider this session: =20 $ gdb -q /tmp/breakpoint-shlib-func Reading symbols from /tmp/breakpoint-shlib-func... (gdb) start Temporary breakpoint 1 at 0x40112e: file /tmp/breakpoint-shlib-func.c= c, line 20. Starting program: /tmp/breakpoint-shlib-func =20 Temporary breakpoint 1, main () at /tmp/breakpoint-shlib-func.cc:20 20 int answer =3D foo (); (gdb) break foo Breakpoint 2 at 0x401030 (2 locations) (gdb) info breakpoints Num Type Disp Enb Address What 2 breakpoint keep y 2.1 y 0x0000000000401030 2.2 y 0x00007ffff7fc50fd in foo() at /tmp/b= reakpoint-shlib-func-lib.cc:20 =20 This is not the expected behaviour. If we compile the same test using a C compiler then we see this: =20 (gdb) break foo Breakpoint 2 at 0x7ffff7fc50fd: file /tmp/breakpoint-shlib-func-c-lib= .c, line 20. (gdb) info breakpoints Num Type Disp Enb Address What 2 breakpoint keep y 0x00007ffff7fc50fd in foo at /tmp/bre= akpoint-shlib-func-c-lib.c:20 =20 Here's what's happening. When GDB parses the symbols in the main executable and the shared library we see a number of different symbols for foo, and use these to create entries in GDB's msymbol table: =20 - In the main executable we see a symbol 'foo@plt' that points at the plt entry for foo, from this we add two entries into GDB's msymbol table, one called 'foo@plt' which points at the plt entry and has type mst_text, then we create a second symbol, this time called 'foo' with type mst_solib_trampoline which also points at the plt entry, =20 - Then, when the shared library is loaded we see another symbol called 'foo', this one points at the actual implementation in the shared library. This time GDB creates a msymbol called 'foo' with type mst_text that points at the implementation. =20 This means that GDB creates 3 msymbols to represent the 2 symbols found in the executable and shared library. =20 When the user creates a breakpoint on 'foo' GDB eventually ends up in search_minsyms_for_name (linespec.c), this function then calls iterate_over_minimal_symbols passing in the name we are looking for wrapped in a lookup_name_info object. =20 In iterate_over_minimal_symbols we iterate over two hash tables (using the name we're looking for as the hash key), first we walk the hash table of symbol linkage names, then we walk the hash table of demangled symbol names. =20 When the language is C++ the symbols for 'foo' will all have been mangled, as a result, in this case, the iteration of the linkage name hash table will find no matching results. =20 However, when we walk the demangled hash table we do find some results. In order to match symbol names, GDB obtains a symbol name matching function by calling the get_symbol_name_matcher method on the language_defn class. For C++, in this case, the matching function we use is cp_fq_symbol_name_matches, which delegates the work to strncmp_iw_with_mode with mode strncmp_iw_mode::MATCH_PARAMS and language set to language_cplus. =20 The strncmp_iw_mode::MATCH_PARAMS mode means that strncmp_iw_mode will skip any parameters in the demangled symbol name when checking for a match, e.g. 'foo' will match the demangled name 'foo()'. The way this is done is that the strings are matched character by character, but, once the string we are looking for ('foo' here) is exhausted, if we are looking at '(' then we consider the match a success. =20 Lets consider the 3 symbols GDB created. If the function declaration is 'void foo ()' then from the main executable we added symbols '_Z3foov@plt' and '_Z3foov', while from the shared library we added another symbol call '_Z3foov'. When these are demangled they become 'foo()@plt', 'foo()', and 'foo()' respectively. =20 Now, the '_Z3foov' symbol from the main executable has the type mst_solib_trampoline, and in search_minsyms_for_name, we search for any symbols of type mst_solib_trampoline and filter these out of the results. =20 However, the '_Z3foov@plt' symbol (from the main executable), and the '_Z3foov' symbol (from the shared library) both have type mst_text. =20 During the demangled name matching, due to the use of MATCH_PARAMS mode, we stop the comparison as soon as we hit a '(' in the demangled name. And so, '_Z3foov@plt', which demangles to 'foo()@plt' matches 'foo', and '_Z3foov', which demangles to 'foo()' also matches 'foo'. =20 By contrast, for C, there are no demangled hash table entries to be iterated over (in iterate_over_minimal_symbols), we only consider the linkage name symbols which are 'foo@plt' and 'foo'. The plain 'foo' symbol obviously matches when we are looking for 'foo', but in this case the 'foo@plt' will not match due to the '@plt' suffix. =20 And so, when the user asks for a breakpoint in 'foo', and the language is C, search_minsyms_for_name, returns a single msymbol, the mst_text symbol for foo in the shared library, while, when the language is C++, we get two results, '_Z3foov' for the shared library function, and '_Z3foov@plt' for the plt entry in the main executable. =20 I propose to fix this in strncmp_iw_with_mode. When the mode is MATCH_PARAMS, instead of stopping at a '(' and assuming the match is a success, GDB will instead search forward for the matching, closing, ')', effectively skipping the parameter list, and then resume matching. Thus, when comparing 'foo' to 'foo()@plt' GDB will effectively compare against 'foo@plt' (skipping the parameter list), and the match will fail, just as it does when the language is C. =20 There is one slight complication, which is revealed by the test gdb.linespec/cpcompletion.exp, when searching for the symbol of a const member function, the demangled symbol will have 'const' at the end of its name, e.g.: =20 struct_with_const_overload::const_overload_fn() const =20 Previously, the matching would stop at the '(' character, but after my change the whole '()' is skipped, and the match resumes. As a result, the 'const' modifier results in a failure to match, when previously GDB would have found a match. =20 To work around this issue, in strncmp_iw_with_mode, when mode is MATCH_PARAMS, after skipping the parameter list, if the next character is '@' then we assume we are looking at something like '@plt' and return a value indicating the match failed, otherwise, we return a value indicating the match succeeded, this allows things like 'const' to be skipped. =20 With these changes in place I now see GDB correctly setting a breakpoint only at the implementation of 'foo' in the shared library. =20 Bug: https://sourceware.org/bugzilla/show_bug.cgi?id=3D20091 Bug: https://sourceware.org/bugzilla/show_bug.cgi?id=3D17201 Bug: https://sourceware.org/bugzilla/show_bug.cgi?id=3D17071 Bug: https://sourceware.org/bugzilla/show_bug.cgi?id=3D17199 =20 Tested-By: Bruno Larsen Approved-By: Simon Marchi Diff: --- gdb/testsuite/gdb.cp/breakpoint-shlib-func-lib.cc | 19 ++++++ gdb/testsuite/gdb.cp/breakpoint-shlib-func.cc | 22 ++++++ gdb/testsuite/gdb.cp/breakpoint-shlib-func.exp | 83 +++++++++++++++++++= ++++ gdb/utils.c | 31 ++++++++- 4 files changed, 154 insertions(+), 1 deletion(-) diff --git a/gdb/testsuite/gdb.cp/breakpoint-shlib-func-lib.cc b/gdb/testsu= ite/gdb.cp/breakpoint-shlib-func-lib.cc new file mode 100644 index 00000000000..2fd7e608df5 --- /dev/null +++ b/gdb/testsuite/gdb.cp/breakpoint-shlib-func-lib.cc @@ -0,0 +1,19 @@ +/* Copyright 2022-2023 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 +foo () +{ + return 0; /* Breakpoint in foo. */ +} diff --git a/gdb/testsuite/gdb.cp/breakpoint-shlib-func.cc b/gdb/testsuite/= gdb.cp/breakpoint-shlib-func.cc new file mode 100644 index 00000000000..a86d06560d4 --- /dev/null +++ b/gdb/testsuite/gdb.cp/breakpoint-shlib-func.cc @@ -0,0 +1,22 @@ +/* Copyright 2022-2023 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 . = */ + +extern int foo (); + +int +main () +{ + int answer =3D foo (); + return answer; +} diff --git a/gdb/testsuite/gdb.cp/breakpoint-shlib-func.exp b/gdb/testsuite= /gdb.cp/breakpoint-shlib-func.exp new file mode 100644 index 00000000000..dc7c4a8a8ea --- /dev/null +++ b/gdb/testsuite/gdb.cp/breakpoint-shlib-func.exp @@ -0,0 +1,83 @@ +# Copyright 2022-2023 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 . + +# Places a breakpoint on a function in a shared library before the +# inferior has started. GDB will place the breakpoint on the @plt +# symbol in the main executable. +# +# When the inferior is started GDB will re-evaluate the breakpoint +# location and move the breakpoint to the function implementation in +# the shared library. +# +# Then, with the inferior started, delete all breakpoints, and +# re-create the breakpoint on the shared library function, GDB should +# place a single breakpoint on the function implementation in the +# shared library. + +require allow_shlib_tests + +standard_testfile .cc -lib.cc + +set libobj [standard_output_file libfoo.so] +if {[build_executable "build shared library" $libobj $srcfile2 \ + {debug c++ shlib}] !=3D 0} { + return -1 +} + +if {[prepare_for_testing "failed to prepare" $testfile $srcfile \ + [list debug c++ shlib=3D$libobj]]} { + return -1 +} + +# Place the breakpoint before the shared library has been loaded, the +# breakpoint should be placed on the @plt symbol. +gdb_test "break foo" "Breakpoint $decimal at $hex" +gdb_test "info breakpoints" "" + +# This is used as an override for delete_breakpoints when we don't +# want functions in gdb.exp to delete breakpoints behind the scenes +# for us. +proc do_not_delete_breakpoints {} { + # Just do nothing. +} + +# Runto main, but don't delete all the breakpoints. +with_override delete_breakpoints do_not_delete_breakpoints { + if {![runto_main]} { + return -1 + } +} + +# The breakpoint should now be showing in `foo` for real. +gdb_test "info breakpoints" \ + "\r\n$decimal\\s+\[^\r\n\]+ in foo\\(\\) at \[^\r\n\]+\r\n.*" \ + "check breakpoints after starting the inferior" + +# Now we can delete the breakpoints. +delete_breakpoints + +# And recreate the foo breakpoint, we should only get one location, +# the actual location. +gdb_test "break foo" "Breakpoint $decimal at \[^\r\n\]+" \ + "recreate foo breakpoint" + +# Check the breakpoint was recreated correctly. +gdb_test "info breakpoints" \ + "\r\n$decimal\\s+\[^\r\n\]+ in foo\\(\\) at \[^\r\n\]+" \ + "check breakpoints after recreation" + +# Continue to breakpoint in foo and confirm we stop in the expected +# place. +gdb_continue_to_breakpoint "breakpoint in foo" \ + ".*/\\* Breakpoint in foo\\. \\*/.*" diff --git a/gdb/utils.c b/gdb/utils.c index d763e91d7e1..91e6974b976 100644 --- a/gdb/utils.c +++ b/gdb/utils.c @@ -2366,7 +2366,31 @@ strncmp_iw_with_mode (const char *string1, const cha= r *string2, return 0; } else - return (*string1 !=3D '\0' && *string1 !=3D '('); + { + if (*string1 =3D=3D '(') + { + int p_count =3D 0; + + do + { + if (*string1 =3D=3D '(') + ++p_count; + else if (*string1 =3D=3D ')') + --p_count; + ++string1; + } + while (*string1 !=3D '\0' && p_count > 0); + + /* There maybe things like 'const' after the parameters, + which we do want to ignore. However, if there's an '@' + then this likely indicates something like '@plt' which we + should not ignore. */ + return *string1 =3D=3D '@'; + } + + return *string1 =3D=3D '\0' ? 0 : 1; + } + } else return 1; @@ -2898,6 +2922,11 @@ strncmp_iw_with_mode_tests () CHECK_MATCH ("foo[abi:a][abi:b](bar[abi:c][abi:d])", "foo[abi:a][abi:b](= bar[abi:c][abi:d])", MATCH_PARAMS); CHECK_MATCH ("foo[abi:a][abi:b](bar[abi:c][abi:d])", "foo", MATCH_PARAMS= ); + CHECK_NO_MATCH ("foo(args)@plt", "foo", MATCH_PARAMS); + CHECK_NO_MATCH ("foo((())args(()))@plt", "foo", MATCH_PARAMS); + CHECK_MATCH ("foo((())args(()))", "foo", MATCH_PARAMS); + CHECK_MATCH ("foo(args) const", "foo", MATCH_PARAMS); + CHECK_MATCH ("foo(args)const", "foo", MATCH_PARAMS); =20 /* strncmp_iw_with_mode also supports case insensitivity. */ {