Home - Waterfall Grid T-Grid Console Builders Recent Builds Buildslaves Changesources - JSON API - About

Console View

Tags: MAIL TEST aarch64 aix cc-with-index centos7 debian fedora freebsd freebsd10 i686 index m32 m64 mingw32 native-extended-gdbserver native-extended-gdbserver-m32 native-extended-gdbserver-m64 native-gdbserver native-gdbserver-m32 native-gdbserver-m64 netbsd power8 ppc64le rhel s390x solaris solaris11 sparcv9 ubuntu x86_64
Legend:   Passed Failed Warnings Failed Again Running Exception Offline No data

MAIL TEST aarch64 aix cc-with-index centos7 debian fedora freebsd freebsd10 i686 index m32 m64 mingw32 native-extended-gdbserver native-extended-gdbserver-m32 native-extended-gdbserver-m64 native-gdbserver native-gdbserver-m32 native-gdbserver-m64 netbsd power8 ppc64le rhel s390x solaris solaris11 sparcv9 ubuntu x86_64
GDB Administrator
Automatic date update in version.in
GDB Administrator
Automatic date update in version.in
Tom Tromey
Fix some typos in ravenscar-thread.c
This fixes some typos I noticed in ravenscar-thread.c.

2019-02-15  Tom Tromey  <tromey@adacore.com>

* ravenscar-thread.c: Fix some typos.
Tom Tromey
Add push_target overload
This adds a push_target overload that takes a "target_ops_up &&".
This removes some calls to release a target_ops_up, and makes the
intent here clearer.

2019-02-15  Tom Tromey  <tromey@adacore.com>

* target.h (push_target): Declare new overload.
* target.c (push_target): New overload, taking an rvalue reference.
* remote.c (remote_target::open_1): Use push_target overload.
* corelow.c (core_target_open): Use push_target overload.
GDB Administrator
Automatic date update in version.in
Tom Tromey
Fix memory leak in create_ada_exception_catchpoint
Phillipe noticed that create_ada_exception_catchpoint was not freeing
the "addr_string" memory:

==14141== 114 bytes in 4 blocks are definitely lost in loss record 1,054 of 3,424
==14141==    at 0x4C2BE6D: malloc (vg_replace_malloc.c:309)
==14141==    by 0x405107: xmalloc (common-utils.c:44)
==14141==    by 0x7563F9: xstrdup (xstrdup.c:34)
==14141==    by 0x381B21: ada_exception_sal (ada-lang.c:13217)
==14141==    by 0x381B21: create_ada_exception_catchpoint(gdbarch*, ada_exception_catchpoint_kind, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, int, int, int) (ada-lang.c:13251)
==14141==    by 0x3820A8: catch_ada_exception_command(char const*, int, cmd_list_element*) (ada-lang.c:13285)
==14141==    by 0x3F4828: cmd_func(cmd_list_element*, char const*, int) (cli-decode.c:1892)

This patch fixes the problem by changing ada_exception_sal to return a
std::string via its out parameter.

2019-02-15  Philippe Waroquiers  <philippe.waroquiers@skynet.be>
    Tom Tromey  <tromey@adacore.com>

* ada-lang.c (ada_exception_sal): Change addr_string to a
(create_ada_exception_catchpoint): Update.
GDB Administrator
Automatic date update in version.in
Thomas Schwinge
[gdb, hurd] Repair build after "Use thread_info and inferior pointers more throughout"
..., that is commit 00431a78b28f913a9d5c912c49680e39cfd20847 causing:

    [...]/gdb/gnu-nat.c: In member function 'virtual void gnu_nat_target::detach(inferior*, int)':
    [...]/gdb/gnu-nat.c:2284:23: error: invalid conversion from 'int' to 'inferior*' [-fpermissive]
      detach_inferior (pid);
    In file included from [...]/gdb/gnu-nat.c:61:0:
    [...]/gdb/inferior.h:523:13: note:  initializing argument 1 of 'void detach_inferior(inferior*)'
    extern void detach_inferior (inferior *inf);

Fixed by inlining the removed code.

* gnu-nat.c (gnu_nat_target::detach): Instead of
'detach_inferior (pid)' call
'detach_inferior (find_inferior_pid (pid))'.
  • RHEL-s390x-m64: uploading gdb.log -  
  • RHEL-s390x-m64: uploading gdb.log -  
  • RHEL-s390x-m64: uploading gdb.log -  
  • RHEL-s390x-m64: uploading gdb.log -  
GDB Administrator
Automatic date update in version.in
Thomas Schwinge
[gdb, hurd] Avoid using 'PATH_MAX' in 'gdb/remote.c'
..., which is not defined in GNU/Hurd systems, and so commit
94585166dfea8232c248044f9f4b1c217dc4ac2e "Extended-remote follow-exec" caused:

    [...]/gdb/remote.c: In member function 'void remote_target::remote_parse_stop_reply(const char*, stop_reply*)':
    [...]/gdb/remote.c:7343:22: error: 'PATH_MAX' was not declared in this scope
            char pathname[PATH_MAX];

* remote.c (remote_target::remote_parse_stop_reply): Avoid using
Thomas Schwinge
[ld, hurd] Remove 'ld-elf/elf.exp' XFAILs
... as a follow-up to commit d98164028637041c5de99af0d057bde3f168a8a8 "Run more
ld tests when not native", which replaced by a proper solution the following
mess before present in 'ld/configure.host':

    -  # When creating static executables, we ought to use crt0.o instead of crt1.o,
    -  # <http://www.gnu.org/software/hurd/open_issues/binutils.html#static>,
    -  # but the testing infrastructure is not prepared for that.  This is not
    -  # relevant for most tests, and the few remaining ones have been XFAILed.
    -  HOSTING_CRT0='[...]'
    -  HOSTING_LIBS='[...]'

* testsuite/ld-elf/elf.exp: Remove Hurd XFAILs.
GDB Administrator
Automatic date update in version.in
Alan Modra
PR24225, nios2 buffer overflow
PR 24225
* elf32-nios2.c (nios2_elf32_relocate_section): Use asprintf and
PRIx64 to generate warning messages.  Print local sym names too.
Joel Brobecker
(Ada) fix GDB crash printing packed array
Trying to print a packed array sometimes leads to a crash (see
attached testcase for an example of when this happens):

  | (gdb) p bad
  | [1]    65571 segmentation fault  gdb -q foo

Variable "bad" is declared in the debug information as an array where
the array's type name has an XPnnn suffix:

  | .uleb128 0xc    # (DIE (0x566) DW_TAG_typedef)
  | .long  .LASF200        # DW_AT_name: "pck__t___XP1"
  | [loc info attributes snipped]
  | .long  0x550  # DW_AT_type
  | .byte  0x1    # DW_AT_alignment

The signals to GDB that the debugging information follows a GNAT encoding
used for packed arrays, and an in order to decode it, we need to find
the type whose name is the same minus the "___XPnnn" suffix: "pck__t".

For that, we make a call to ada-lang.c::standard_lookup, which is
a simple function which essentially does:

  | /* Return the result of a standard (literal, C-like) lookup of NAME in
  |    given DOMAIN, visible from lexical block BLOCK.  */
  |  [...]
  |  sym = lookup_symbol_in_language (name, block, domain, language_c, 0);

Unfortunately for us, while the intent of this call was to perform
an exact-match lookup, in our case, it returns ... type pck__t___XP1
instead! In other words, it finds itself back. The reason why it finds
this type is a confluence of two factors:

  (1) Forcing the lookup into language_c currently does not affect
      how symbol matching is done anymore, because we look at the symbol's
      language to determine which kind of matching should be done;

  (2) The lookup searches the local context (via block) first, beforei
      doing a more general lookup. And looking at the debug info for
      the main subprogram, we see that type "pck__t" is not declared
      there, only in the debug info for pck.ads. In other words,
      there is no way that we accidently find "pck__t" by random chance.

I believe Pedro added a new function called ada_lookup_encoded_symbol
for that specific purpose, so I started by replacing the lookup
by language above by this. Unfortunately, still no joy.

This was because, even though ada_lookup_encoded_symbol puts angle-
brackets around the search name to signal that we want a verbatim
search, we end up losing that information in the function called
to compare a symbol with the search name:

  | static bool
  | do_full_match (const char *symbol_search_name,
  |                const lookup_name_info &lookup_name,
  |                completion_match_result *comp_match_res)
  | {
  |  return full_match (symbol_search_name, ada_lookup_name (lookup_name));
                                    <=> lookup_name.m_ada.m_encoded_name
                                          (no angle brackets)

The way I fixed this was by introducing a new function called
do_exact_match, and then adjust ada_get_symbol_name_matcher to
return that function when seeing that we have a verbatim non-wild-match

As it happens, this fixes an incorrect test in gdb.ada/homony.exp,
where we were inserting a breakpoint on a symbol using the angle-brackets
notation, and got 2 locations for that breakpoint...

    (gdb) b <homonym__get_value>
    Breakpoint 1 at 0x4029fc: <homonym__get_value>. (2 locations)

...  each location being in a different function:

    (gdb) info break
    Num    Type          Disp Enb Address            What
    1      breakpoint    keep y  <MULTIPLE>
    1.1                        y  0x00000000004029fc in homonym.get_value
                                    at /[...]/homonym.adb:32
    1.2                        y  0x0000000000402a3a in homonym.get_value
                                    at /[...]/homonym.adb:50
    (gdb) x /i 0x00000000004029fc
      0x4029fc <homonym__get_value+8>:    movl  $0x1d,-0x4(%rbp)
    (gdb) x /i 0x0000000000402a3a
      0x402a3a <homonym__get_value__2+8>:  movl  $0x11,-0x4(%rbp)

Since we used angle-brackets, we shouldn't be matching the second one,
something this patch fixes.


        * ada-lang.c (standard_lookup): Use ada_lookup_encoded_symbol
        instead of lookup_symbol_in_language
        (do_exact_match): New function.
        (ada_get_symbol_name_matcher): Return do_exact_match when
        doing a verbatim match.


        * gdb.ada/big_packed_array: New testcase.
        * gdb.ada/homonym.exp: Fix incorrect expected output for
        "break <homonym__get_value>" test.

Tested on x86_64-linux.
Tom Tromey
Fix pager bugs with style output
I believe this fixes all the pager output problems with styling that
Philippe pointed out, plus at least one more.  The patch is somewhat
hard to reason about, so you may wish to give it a try.  Even writing
the tests was hard.

This removes the style caching, because it was difficult to keep the
style cache correct in all cases.  Since this would cause more style
escapes to be emitted, instead it changes fputs_styled to try to avoid
unnecessary changes.

Another bug was that the wrap buffer was not flushed in the case where
wrap_column==0.  In the old (pre-patch series) code, characters were
directly emitted in this case; so flushing the wrap buffer here
restores this behavior.

On error the wrap buffer must be emptied.  Otherwise, interrupting
output can leave characters in the buffer that will be emitted later.

As discussed on gdb-patches, this fixes the ada-lang.c problem where
filtered and unfiltered printing were mixed.  Now user_select_syms
uses filtered printing, which is what its callees were already doing.

Finally, it was possible for source line highlighting to be garbled
(and invalid escape sequences emitted) if the pager was invoked at the
wrong spot.  To fix this, the patch arranges for source line escapes
to always be emitted as a unit.

2019-02-17  Tom Tromey  <tom@tromey.com>

* ada-lang.c (user_select_syms): Use filtered printing.
* utils.c (wrap_style): New global.
(desired_style): Remove.
(emit_style_escape): Add stream parameter.
(set_output_style, reset_terminal_style, prompt_for_continue):
(flush_wrap_buffer): Only flush gdb_stdout.
(wrap_here): Set wrap_style.
(fputs_maybe_filtered): Clear the wrap buffer on exception.  Don't
treat escape sequences as a character.  Change when wrap buffer is
(fputs_styled): Do not set the output style when the default is
* ui-style.h (struct ui_file_style) <is_default>: New method.
* source.c (print_source_lines_base): Emit escape sequences in one

2019-02-17  Tom Tromey  <tom@tromey.com>

* gdb.base/style.exp: Add line-wrapping tests.
* gdb.base/page.exp: Add test for quitting during pagination.
Weimin Pan
Updating test case
* Replaced "run" with "runto_main + continue".
* Replaced "gdb_compile + clean_restart" with "prepare_for_testing".
* Added comment for case "exited with code 01".

* Removed SET_WATCHPOINT marco.
* Removed redundent cleanup ().
* Cleaned up comment.
GDB Administrator
Automatic date update in version.in
Tom Tromey
Minor C++-ification in ravenscar-thread.c
This changes some functions in ravenscar-thread.c to return "bool"
rather than int, where appropriate, and also changes "(void)" to "()".

2019-02-15  Tom Tromey  <tromey@adacore.com>

* ravenscar-thread.c (is_ravenscar_task)
(ravenscar_task_is_currently_active): Return bool.
(ravenscar_update_inferior_ptid, get_running_thread_msymbol)
(_initialize_ravenscar): Remove "(void)".
(has_ravenscar_runtime, ravenscar_runtime_initialized): Likewise.
Return bool.
Tom Tromey
Minor Ada task cleanups
While working on the Ada task code, I noticed a few things that could
be cleaned up:

* task_list_valid_p was not set in all cases in ada_build_task_list.
  This causes many needless re-fetches of the task list.

* task_list_valid_p can be bool, and various functions can also return

* Nothing checks the return value of read_known_tasks, so it can be
  changed to return void.

* The call to ada_build_task_list in
  ravenscar_thread_target::update_thread_list is redundant, because
  this is the first thing done by iterate_over_live_ada_tasks.

Tested using the internal AdaCore test suite against a ravenscar

2019-02-19  Tom Tromey  <tromey@adacore.com>

* ravenscar-thread.c
(ravenscar_thread_target::update_thread_list): Don't call
* ada-lang.h (ada_build_task_list): Don't declare.
* ada-tasks.c (struct ada_tasks_inferior_data)
<task_list_valid_p>: Now bool.
(read_known_tasks, ada_task_list_changed)
(ada_tasks_invalidate_inferior_data): Update.
(read_known_tasks_array): Return bool.
(read_known_tasks_list): Likewise.
(read_known_tasks): Return void.
(ada_build_task_list): Now static.
Tom Tromey
Change macro_source_fullname to return a std::string
While working on the previous patch, I noticed that if
macro_source_fullname returned a std::string, then the callers would
be simplified.  This patch implements this idea.

2019-02-17  Tom Tromey  <tom@tromey.com>

* macrotab.h (macro_source_fullname): Return a std::string.
* macrotab.c (macro_include, check_for_redefinition)
(macro_undef, macro_lookup_definition, foreach_macro)
(foreach_macro_in_scope): Update.
(macro_source_fullname): Return a std::string.
* macrocmd.c (show_pp_source_pos): Update.
Thomas Schwinge
[gdb, hurd] Address "ISO C++ forbids converting a string constant to 'char*' [-Wwrite-strings]" diagnostics
... that appeared with 9bf2a700667c53003ece783c05e8b355801105f2
"-Wwrite-strings: Remove -Wno-write-strings".

* gnu-nat.c (gnu_write_inferior, parse_int_arg, _parse_bool_arg)
(check_empty): Use "const char *".
Saagar Jha
Use the correct name for various MACH-O based operating systems in comments.
include * mach-o/loader.h: Use new OS names in comments.
David Michael