diff options
author | Pedro Alves <palves@redhat.com> | 2016-08-22 20:05:09 +0100 |
---|---|---|
committer | Pedro Alves <palves@redhat.com> | 2016-08-22 20:05:09 +0100 |
commit | 6bb90213cb7b8e2f3be20f2e46f11f57f0c9ce55 (patch) | |
tree | c4d9f2bf1199d01f59be4dc9b0c181432b008a49 | |
parent | 5fbe0d878a691b9be42bb2bdebd027ac3dfd38c2 (diff) | |
download | gdb-6bb90213cb7b8e2f3be20f2e46f11f57f0c9ce55.zip gdb-6bb90213cb7b8e2f3be20f2e46f11f57f0c9ce55.tar.gz gdb-6bb90213cb7b8e2f3be20f2e46f11f57f0c9ce55.tar.bz2 |
Fix PR gdb/20505 - Make vDSO detection work with core files
Loading a core dump that was either generated on a system running
pristine glibc master, or on a Fedora/RHEL system with LD_DEBUG=unused
set in the environment, solib-svr4.c:svr4_current_sos fails to filter
out the vDSO, resulting in:
(gdb) core-file corefile.core^M
[New LWP 2362]^M
warning: Could not load shared library symbols for linux-vdso.so.1.^M
Do you need "set solib-search-path" or "set sysroot"?^M
Core was generated by `build-gdb/gdb/testsuite/outputs/gdb.base/corefile/'.^M
...
The problem is that gdbarch_vsyscall_range does not support core
inferiors at all.
When live debugging, we're finding the vDSO's start address with
auxv/AT_SYSINFO_EHDR, and then we find the vDSO's size by look for the
corresponding mapping, by parsing /proc/PID/maps. When debugging a
core dump, we can also determine the starting address from
auxv/AT_SYSINFO_EHDR. However, we obviously can't read the core
mappings out of the host's /proc. But we can instead look for a
corresponding load segment in the core's bfd.
gdb/ChangeLog:
2016-08-22 Pedro Alves <palves@redhat.com>
PR gdb/20505
* linux-tdep.c (linux_vsyscall_range_raw): For core inferiors,
find the vDSO's start address with AT_SYSINFO_EHDR too, and
determine the vDSO's size by finding the PT_LOAD segment that
matches AT_SYSINFO_EHDR.
gdb/testsuite/ChangeLog:
2016-08-22 Pedro Alves <palves@redhat.com>
PR gdb/20505
* gdb.base/vdso-warning.exp: Test core dumps too. Use
with_test_prefix. Factor out bits to ...
(test_no_vdso): ... this new procedure.
-rw-r--r-- | gdb/ChangeLog | 8 | ||||
-rw-r--r-- | gdb/linux-tdep.c | 35 | ||||
-rw-r--r-- | gdb/testsuite/ChangeLog | 7 | ||||
-rw-r--r-- | gdb/testsuite/gdb.base/vdso-warning.exp | 76 |
4 files changed, 97 insertions, 29 deletions
diff --git a/gdb/ChangeLog b/gdb/ChangeLog index db3527b..8de59d0 100644 --- a/gdb/ChangeLog +++ b/gdb/ChangeLog @@ -1,3 +1,11 @@ +2016-08-22 Pedro Alves <palves@redhat.com> + + PR gdb/20505 + * linux-tdep.c (linux_vsyscall_range_raw): For core inferiors, + find the vDSO's start address with AT_SYSINFO_EHDR too, and + determine the vDSO's size by finding the PT_LOAD segment that + matches AT_SYSINFO_EHDR. + 2016-08-19 Yao Qi <yao.qi@linaro.org> * aarch64-tdep.c (aarch64_analyze_prologue): Handle register diff --git a/gdb/linux-tdep.c b/gdb/linux-tdep.c index ab110b0..718dc1a 100644 --- a/gdb/linux-tdep.c +++ b/gdb/linux-tdep.c @@ -2287,17 +2287,42 @@ linux_vsyscall_range_raw (struct gdbarch *gdbarch, struct mem_range *range) long pid; char *data; - /* Can't access /proc if debugging a core file. */ - if (!target_has_execution) + if (target_auxv_search (¤t_target, AT_SYSINFO_EHDR, &range->start) <= 0) return 0; + /* It doesn't make sense to access the host's /proc when debugging a + core file. Instead, look for the PT_LOAD segment that matches + the vDSO. */ + if (!target_has_execution) + { + Elf_Internal_Phdr *phdrs; + long phdrs_size; + int num_phdrs, i; + + phdrs_size = bfd_get_elf_phdr_upper_bound (core_bfd); + if (phdrs_size == -1) + return 0; + + phdrs = (Elf_Internal_Phdr *) alloca (phdrs_size); + num_phdrs = bfd_get_elf_phdrs (core_bfd, phdrs); + if (num_phdrs == -1) + return 0; + + for (i = 0; i < num_phdrs; i++) + if (phdrs[i].p_type == PT_LOAD + && phdrs[i].p_vaddr == range->start) + { + range->length = phdrs[i].p_memsz; + return 1; + } + + return 0; + } + /* We need to know the real target PID to access /proc. */ if (current_inferior ()->fake_pid_p) return 0; - if (target_auxv_search (¤t_target, AT_SYSINFO_EHDR, &range->start) <= 0) - return 0; - pid = current_inferior ()->pid; /* Note that reading /proc/PID/task/PID/maps (1) is much faster than diff --git a/gdb/testsuite/ChangeLog b/gdb/testsuite/ChangeLog index 9f45c39..7e5fce6 100644 --- a/gdb/testsuite/ChangeLog +++ b/gdb/testsuite/ChangeLog @@ -1,3 +1,10 @@ +2016-08-22 Pedro Alves <palves@redhat.com> + + PR gdb/20505 + * gdb.base/vdso-warning.exp: Test core dumps too. Use + with_test_prefix. Factor out bits to ... + (test_no_vdso): ... this new procedure. + 2016-08-19 Carl Love <cel@us.ibm.com> * gdb.arch/altivec-regs.exp: Use standard_testfile instead of diff --git a/gdb/testsuite/gdb.base/vdso-warning.exp b/gdb/testsuite/gdb.base/vdso-warning.exp index af2b2b0..aeb85a2 100644 --- a/gdb/testsuite/gdb.base/vdso-warning.exp +++ b/gdb/testsuite/gdb.base/vdso-warning.exp @@ -13,42 +13,70 @@ # You should have received a copy of the GNU General Public License # along with this program. If not, see <http://www.gnu.org/licenses/>. +# Test that on Linux, we don't warn about not finding the vDSO. E.g.: +# +# warning: Could not load shared library symbols for linux-vdso.so.1. + standard_testfile if { [prepare_for_testing "failed to prepare" ${testfile} $srcfile] } { return -1 } -gdb_breakpoint "main" +with_test_prefix "setup" { + gdb_breakpoint "main" -# At least some versions of Fedora/RHEL glibc have local patches that -# hide the vDSO. This lines re-exposes it. See PR libc/13097, -# comment 2. There's no support for passing environment variables in -# the remote protocol, but that's OK -- if we're testing against a -# glibc that doesn't list the vDSO without this, the test should still -# pass. -gdb_test_no_output "set environment LD_DEBUG=unused" + # At least some versions of Fedora/RHEL glibc have local patches that + # hide the vDSO. This lines re-exposes it. See PR libc/13097, + # comment 2. There's no support for passing environment variables in + # the remote protocol, but that's OK -- if we're testing against a + # glibc that doesn't list the vDSO without this, the test should still + # pass. + gdb_test_no_output "set environment LD_DEBUG=unused" +} -gdb_run_cmd +proc test_no_vdso {command} { + global srcfile + global gdb_prompt -set test "stop without warning" -gdb_test_multiple "" $test { - -re "Could not load shared library symbols .*\r\n$gdb_prompt $" { - fail $test + set message "startup" + gdb_test_multiple "$command" $message { + -re "Could not load shared library symbols .*\r\n$gdb_prompt $" { + fail $message + } + -re "main \\(\\) at .*$srcfile.*\r\n$gdb_prompt $" { + pass $message + } } - -re "\r\nBreakpoint \[0-9\]+, main .*\r\n$gdb_prompt $" { - pass $test + + # Extra testing in case the warning changes and we miss updating + # the above. + set test "no vdso without symbols is listed" + gdb_test_multiple "info shared" $test { + -re "No\[^\r\n\]+linux-(vdso|gate).*$gdb_prompt $" { + fail $test + } + -re "$gdb_prompt $" { + pass $test + } } } -# Extra testing in case the warning changes and we miss updating the -# above. -set test "no vdso without symbols is listed" -gdb_test_multiple "info shared" $test { - -re "No\[^\r\n\]+linux-(vdso|gate).*$gdb_prompt $" { - fail $test - } - -re "$gdb_prompt $" { - pass $test +# First, try a live process. +with_test_prefix "run" { + gdb_run_cmd + test_no_vdso "" +} + +# Now, dump a core, and reload it. +with_test_prefix "core" { + set corefile [standard_output_file $testfile.core] + set core_supported [gdb_gcore_cmd "$corefile" "save a corefile"] + if {!$core_supported} { + return -1 } + + clean_restart ${testfile} + + test_no_vdso "core-file $corefile" } |