Fix off-by-one error in ada_fold_name. (RHBZ 1905996, Kevin Buettner)

This commit is contained in:
Kevin Buettner 2020-12-09 14:42:26 -07:00
parent a8884a5be0
commit 61d59b7323
5 changed files with 91 additions and 8 deletions

View File

@ -131,10 +131,6 @@ Patch034: gdb-6.7-testsuite-stable-results.patch
#=fedoratest #=fedoratest
Patch035: gdb-6.5-ia64-libunwind-leak-test.patch Patch035: gdb-6.5-ia64-libunwind-leak-test.patch
# Test hiding unexpected breakpoints on intentional step commands.
#=fedoratest
Patch036: gdb-6.5-missed-trap-on-step-test.patch
# Test GCORE for shmid 0 shared memory mappings. # Test GCORE for shmid 0 shared memory mappings.
#=fedoratest: But it is broken anyway, sometimes the case being tested is not reproducible. #=fedoratest: But it is broken anyway, sometimes the case being tested is not reproducible.
Patch038: gdb-6.3-mapping-zero-inode-test.patch Patch038: gdb-6.3-mapping-zero-inode-test.patch
@ -382,3 +378,8 @@ Patch096: gdb-vla-intel-fix-print-char-array.patch
# =fedoratest # =fedoratest
Patch097: gdb-rhbz1553104-s390x-arch12-test.patch Patch097: gdb-rhbz1553104-s390x-arch12-test.patch
# Fix off-by-one error in ada_fold_name.patch (RH BZ 1905996)
# Upstream patch proposal: https://sourceware.org/pipermail/gdb-patches/2020-December/173935.html
# =fedoratest
Patch098: gdb-rhbz1905996-fix-off-by-one-error-in-ada_fold_name.patch

View File

@ -31,7 +31,6 @@
%patch033 -p1 %patch033 -p1
%patch034 -p1 %patch034 -p1
%patch035 -p1 %patch035 -p1
%patch036 -p1
%patch038 -p1 %patch038 -p1
%patch039 -p1 %patch039 -p1
%patch040 -p1 %patch040 -p1
@ -92,3 +91,4 @@
%patch095 -p1 %patch095 -p1
%patch096 -p1 %patch096 -p1
%patch097 -p1 %patch097 -p1
%patch098 -p1

View File

@ -95,3 +95,4 @@ gdb-rhbz1398387-tab-crash-test.patch
gdb-archer.patch gdb-archer.patch
gdb-vla-intel-fix-print-char-array.patch gdb-vla-intel-fix-print-char-array.patch
gdb-rhbz1553104-s390x-arch12-test.patch gdb-rhbz1553104-s390x-arch12-test.patch
gdb-rhbz1905996-fix-off-by-one-error-in-ada_fold_name.patch

View File

@ -0,0 +1,82 @@
From FEDORA_PATCHES Mon Sep 17 00:00:00 2001
From: Kevin Buettner <kevinb@redhat.com>
Date: Tue, 8 Dec 2020 14:07:45 -0700
Subject: gdb-rhbz1905996-fix-off-by-one-error-in-ada_fold_name.patch
;; Fix off-by-one error in ada_fold_name.patch (RH BZ 1905996)
;; Upstream patch proposal: https://sourceware.org/pipermail/gdb-patches/2020-December/173935.html
;; =fedoratest
Fix off-by-one error in ada_fold_name
I'm seeing a libstdc++ assertion failure when running GDB's "maint selftest"
command when GDB is configured with the following CFLAGS and CXXFLAGS as
part of the configure line:
CFLAGS='-D_GLIBCXX_DEBUG -g3 -O0' CXXFLAGS='-D_GLIBCXX_DEBUG -g3 -O0'
This is what I see when running the self tests:
(gdb) maint selftest
Running selftest aarch64-analyze-prologue.
Running selftest aarch64-process-record.
Running selftest arm-record.
Running selftest arm_analyze_prologue.
Running selftest array_view.
Running selftest child_path.
Running selftest cli_utils.
Running selftest command_structure_invariants.
Running selftest copy_bitwise.
Running selftest copy_integer_to_size.
Running selftest cp_remove_params.
Running selftest cp_symbol_name_matches.
Running selftest dw2_expand_symtabs_matching.
/usr/include/c++/11/string_view:211: constexpr const value_type& std::basic_string_view<_CharT, _Traits>::operator[](std::basic_string_view<_CharT, _Traits>::size_type) const [with _CharT = char; _Traits = std::char_traits<char>; std::basic_string_view<_CharT, _Traits>::const_reference = const char&; std::basic_string_view<_CharT, _Traits>::size_type = long unsigned int]: Assertion '__pos < this->_M_len' failed.
Aborted (core dumped)
Here's a partial stack trace:
#0 0x00007ffff6ef6262 in raise () from /lib64/libc.so.6
#1 0x00007ffff6edf8a4 in abort () from /lib64/libc.so.6
#2 0x00000000004249bf in std::__replacement_assert (
__file=0xef7480 "/usr/include/c++/11/string_view", __line=211,
__function=0xef7328 "constexpr const value_type& std::basic_string_view<_CharT, _Traits>::operator[](std::basic_string_view<_CharT, _Traits>::size_type) const [with _CharT = char; _Traits = std::char_traits<char>; std::ba"...,
__condition=0xef7311 "__pos < this->_M_len")
at /usr/include/c++/11/x86_64-redhat-linux/bits/c++config.h:2624
#3 0x0000000000451737 in std::basic_string_view<char, std::char_traits<char> >::operator[] (this=0x7fffffffc200, __pos=8)
at /usr/include/c++/11/string_view:211
#4 0x00000000004329f5 in ada_fold_name (name="function")
at /ironwood1/sourceware-git/rawhide-master/bld/../../worktree-master/gdb/ada-lang.c:988
And, looking at frame #4...
(top-gdb) up 4
at /ironwood1/sourceware-git/rawhide-master/bld/../../worktree-master/gdb/ada-lang.c:988
988 fold_buffer[i] = tolower (name[i]);
(top-gdb) p i
$1 = 8
(top-gdb) p name.size()
$2 = 8
My patch adjusts the comparison to only copy name.size() characters
from the string. I've added a separate statement for NUL character
termination of fold_buffer[].
gdb/ChangeLog:
* ada-lang.c (ada_fold_name): Fix off-by-one error.
diff --git a/gdb/ada-lang.c b/gdb/ada-lang.c
--- a/gdb/ada-lang.c
+++ b/gdb/ada-lang.c
@@ -1006,8 +1006,9 @@ ada_fold_name (gdb::string_view name)
{
int i;
- for (i = 0; i <= len; i += 1)
+ for (i = 0; i < len; i += 1)
fold_buffer[i] = tolower (name[i]);
+ fold_buffer[i] = '\0';
}
return fold_buffer;

View File

@ -1184,9 +1184,8 @@ fi
%endif %endif
%changelog %changelog
* Mon Dec 07 2020 Kevin Buettner <kevinb@redhat.com> - 10.1-2 * Wed Dec 09 2020 Kevin Buettner <kevinb@redhat.com> - 10.1-2
- Bump release number to avoid lower NVR (for rawhide) than that of F33. - Fix off-by-one error in ada_fold_name. (RHBZ 1905996, Kevin Buettner)
(RHBZ 1894912)
* Wed Nov 04 2020 Kevin Buettner <kevinb@redhat.com> - 10.1-1 * Wed Nov 04 2020 Kevin Buettner <kevinb@redhat.com> - 10.1-1
- Rebase to FSF GDB 10.1. - Rebase to FSF GDB 10.1.