This is the mail archive of the
gcc-regression@gcc.gnu.org
mailing list for the GCC project.
The gdb failures...
- From: Zack Weinberg <zack at codesourcery dot com>
- To: gcc-regression at gcc dot gnu dot org
- Date: Tue, 18 Dec 2001 09:34:39 -0800
- Subject: The gdb failures...
We've been seeing gdb failures on powerpc for at least a month now,
and they've just started happening on i386 too. Looking through the
logs, many of them seem to be because GCC now uses DW_FORM_strp which
the tester's version of gdb doesn't know about:
(gdb) file /maat/heart/tbox/native/build/test-gdb/gdb.base/run
Reading symbols from /maat/heart/tbox/native/build/test-gdb/gdb.base/run...
Dwarf Error: Cannot handle DW_FORM_strp in DWARF reader.
This patch for GDB added support for DW_FORM_strp, it went in a month
ago:
2001-11-16 Jakub Jelinek <jakub@redhat.com>
* dwarf2read.c (dwarf_str_buffer): New.
(struct dwarf2_pinfo): Add dwarf_str_buffer and dwarf_str_size.
(DWARF_STR_BUFFER, DWARF_STR_SIZE): Define.
(dwarf2_has_info): Clear dwarf_str_offset.
(dwarf2_build_psymtabs): Read .debug_str section if present.
(dwarf2_build_psymtabs_hard): Save DWARF_STR_BUFFER and
DWARF_STR_SIZE.
(psymtab_to_symtab_1): Restore DWARF_STR_BUFFER and DWARF_STR_SIZE.
(read_attribute): Handle DW_FORM_strp.
(read_n_bytes, read_string): Remove HOST_CHAR_BIT != 8
handling code.
(read_indirect_string): New.
(dump_die): Handle DW_FORM_strp.
I can only conclude that the regression tester's copy of gdb isn't
being updated automatically. It seems to me that either it should
be automatically updated, or it shouldn't be tested.
zw