Renaming over existing files needs additional care to restore
permissions and ownership, which may not always succeed.
Additionally, other properties of the file such as extended attributes
may be lost, making the operation flaky.
For predictable results, resort to rename() only if the file does not
exist, otherwise copy the file contents into the existing file. This
ensures that no additional tricks are needed to retain file
properties.
This also allows dropping of the redundant set_times on the tmpfile in
objcopy/strip since now we no longer rename over existing files.
binutils/
* ar.c (write_archive): Adjust call to SMART_RENAME.
* arsup.c (ar_save): Likewise.
* objcopy (strip_main): Don't set times on temporary file and
adjust call to SMART_RENAME.
(copy_main): Likewise.
* rename.c [!S_ISLNK]: Remove definitions.
(try_preserve_permissions): Remove function.
(smart_rename): Replace PRESERVE_DATES argument with
TARGET_STAT. Use rename system call only if TO does not exist.
* bucomm.h (smart_rename): Adjust declaration.
* config/tc-s390.c (s390_parse_cpu): New entry for arch14.
* doc/c-s390.texi: Document arch14 march option.
* testsuite/gas/s390/s390.exp: Run the arch14 related tests.
* testsuite/gas/s390/zarch-arch14.d: New test.
* testsuite/gas/s390/zarch-arch14.s: New test.
Nick Clifton [Fri, 5 Feb 2021 14:14:01 +0000 (14:14 +0000)]
Revert "binutils: Make smart_rename safe too"
This reverts commit 014cc7f849e8209623fc99264814bce7b3b6faf2.
Given the problems associated with this patch and the others intended to fix the smart_rename CVE, the decision has been taken to$
Nick Clifton [Fri, 5 Feb 2021 14:12:24 +0000 (14:12 +0000)]
Revert "Fix a build problem when using FreeBSD 12."
This reverts commit b143e2d506bee1020752597f979d5af174edc36d.
Given the problems associated with this patch and the others intended to fix the smart_rename CVE, the decision has been taken to$
Nick Clifton [Fri, 5 Feb 2021 14:10:21 +0000 (14:10 +0000)]
Revert "pr27270 and pr27284, ar segfaults and wrong file mode"
This reverts commit 95b91a043aeaeb546d2fea556d84a2de1e917770.
Given the problems associated with this patch and the others intended to fix the smart_rename CVE, the decision has been taken to$
Nick Clifton [Fri, 5 Feb 2021 14:08:16 +0000 (14:08 +0000)]
Revert "PR27345, binutils/arsup.c: lstat() not available on all targets"
This reverts commit c0034ac596db89dfb5fa007a63488ad7af555fe9.
Given the problems associated with this patch and the others intended to fix the smart_rename CVE, the decision has been taken to revert the patches for the 2.36.1 point release.
Alan Modra [Thu, 4 Feb 2021 03:26:34 +0000 (13:56 +1030)]
PR27311, (symbol from plugin): undefined reference, hidden sym
bfd/
PR 27311
* elflink.c (elf_link_add_object_symbols): Don't pull in as-needed
libraries for IR references on pass over libraries after LTO
recompilation.
ld/
* testsuite/ld-plugin/pr27311d.c: New test.
* testsuite/ld-plugin/lto.exp: Rename pr27311 to pr27311-1, compile
and link new test as pr27311-2.
Alan Modra [Sun, 31 Jan 2021 15:34:41 +0000 (02:04 +1030)]
pr27270 and pr27284, ar segfaults and wrong file mode
PR 27270
PR 27284
PR 26945
* ar.c: Don't include libbfd.h.
(write_archive): Replace xmalloc+strcpy with xstrdup. Use
bfd_stat rather than fstat on iostream. Move stat and fd tests
outside of _WIN32 ifdef. Delete skip_stat variable.
* arsup.c (temp_name, real_ofd): New static variables.
(ar_open): Use make_tempname and bfd_fdopenw.
(ar_save): Adjust to suit ar_open changes. Move stat output
of _WIN32 ifdef.
* objcopy.c: Don't include libbfd.h.
(copy_file): Use bfd_stat.
Alan Modra [Tue, 2 Feb 2021 04:35:14 +0000 (15:05 +1030)]
PR27311, ld.bfd (symbol from plugin): undefined reference
A default versioned symbol definition in a shared library is
overridden by an unversioned definition in a regular object file, and
thus should not be reason to make an as-needed library needed.
bfd/
PR 27311
* elflink.c (elf_link_add_object_symbols): Don't pull in as-needed
libraries when H is an indirect symbol after calling
_bfd_elf_add_default_symbol.
ld/
* testsuite/ld-ifunc/ifunc.exp (libpr16467b.so, libpr16467bn.so):
Link with --as-needed.
* testsuite/ld-plugin/pr27311.d,
* testsuite/ld-plugin/pr27311.ver,
* testsuite/ld-plugin/pr27311a.c,
* testsuite/ld-plugin/pr27311b.c,
* testsuite/ld-plugin/pr27311c.c: New testcase.
* testsuite/ld-plugin/lto.exp: Run it. Correct PR14918 and
PR12982 entries.
Alan Modra [Thu, 28 Jan 2021 00:00:36 +0000 (10:30 +1030)]
PR27259, SHF_LINK_ORDER self-link
This stops ld from endless looping on SHF_LINK_ORDER sh_link loops.
bfd/
PR 27259
* elflink.c (_bfd_elf_gc_mark_extra_sections): Use linker_mark to
prevent endless looping of linked-to sections.
ld/
PR 27259
* ldelf.c (ldelf_before_place_orphans): Use linker_mark to
prevent endless looping of linked-to sections.
bfd/
PR 27271
* elflink.c (bfd_elf_link_record_dynamic_symbol): Don't segfault
on symbols defined in absolute or other special sections.
ld/
* testsuite/ld-tic6x/tic6x.exp: Add pr27271 test.
Nick Alcock [Tue, 26 Jan 2021 16:05:17 +0000 (16:05 +0000)]
ld: depend on libctf
Since ld may depend on libctf (if present), and libctf may be relinked
by the installation process, libctf must be installed before ld is,
or the relink may fail if it calls on symbols or symbol versions that do
not exist in any libctf already present on the system. (If none is
present, the copy in the build tree will be automatically used, but
if one *is* present, it may take precedence and break things.)
(This is a maybe- dependency, so it will work even if libctf is
disabled.)
ChangeLog
2021-01-26 Nick Alcock <nick.alcock@oracle.com>
* Makefile.def: Add install-libctf dependency to install-ld.
* Makefile.in: Regenerated.
Alan Modra [Mon, 25 Jan 2021 04:47:15 +0000 (15:17 +1030)]
pr27228 testcase
This failed on ft32, hppa, and mips-irix targets. In the case of ft32
the problem was iterating over an array in reverse and not using the
proper condition, so BFD_RELOC_NONE was not recognised.
bfd/
* elf32-ft32.c (ft32_reloc_type_lookup): Don't miss ft32_reloc_map[0].
gas/
PR 27282
* testsuite/gas/all/none.d: Replace skip with xfail, don't xfail ft32.
* testsuite/gas/elf/pr27228.d: xfail hppa and allow OBJECT match.
H.J. Lu [Tue, 26 Jan 2021 16:18:39 +0000 (08:18 -0800)]
nios2: Don't disable relaxation with --gdwarf-N
GCC 11 passes --gdwarf-5 to assembler to enable DWARF5 debug info. Don't
disable relaxation when --gdwarf-N is specified. The assembler generated
debug information will treat the sequence of the relaxed instructions as
a single instruction.
PR gas/27243
* config/tc-nios2.c (md_begin): Don't disable relaxation with
--gdwarf-N.
* testsuite/gas/nios2/relax.d: New file.
* testsuite/gas/nios2/relax.s: Likewise.
H.J. Lu [Sun, 24 Jan 2021 15:00:34 +0000 (07:00 -0800)]
DWARF-5: Ignore empty range in DWARF-5 line number tables
The DWARF5 spec does indeed explicitly say: "A bounded range entry whose
beginning and ending address offsets are equal (including zero) indicates
an empty range and may be ignored."
Since arange_add already ignores empty ranges, remove the whole check
which is equivalent to the check plus explicit continue.
PR binutils/27231
* dwarf2.c (read_rnglists): Ignore empty range when parsing line
number tables.
Alan Modra [Sun, 24 Jan 2021 02:09:07 +0000 (12:39 +1030)]
PR27228, .reloc wrong symbol emitted for undefined local symbol
Local symbols are of course supposed to be defined by their object
file, but in other cases a local symbol is promoted to global by gas
if undefined and referenced. This patch stops gas wrongly replacing a
local undefined symbol with the undefined section symbol, resulting in
a .reloc undefined local symbol being emitted as global.
PR 27228
* write.c (resolve_reloc_expr_symbols): Don't assume local symbol
is defined.
Alan Modra [Thu, 21 Jan 2021 06:20:02 +0000 (16:50 +1030)]
PR27221, 058430b4a1 warnings while assembling the Linux kernel
PR 27221
* dwarf2dbg.c (dwarf2_gen_line_info_1): Don't warn about ignored
line number info when gas is generating it.
* testsuite/gas/elf/dwarf2-20.d: Adjust to not expect warnings.
* testsuite/gas/m68hc11/indexed12.d: Likewise.
* testsuite/gas/elf/elf.exp: Don't run warn-2.
* gas/testsuite/gas/elf/warn-2.s: Delete.
Alan Modra [Wed, 20 Jan 2021 05:30:58 +0000 (16:00 +1030)]
PowerPC64 synthetic symbols
If an assembler trims off unused section symbols there may be no
section symbols. This means that testing for the .opd section symbol
needs to test both the section name and symbol flags.
* elf64-ppc.c (ppc64_elf_get_synthetic_symtab): Don't assume
section symbols are present.
Nick Alcock [Tue, 19 Jan 2021 12:45:18 +0000 (12:45 +0000)]
libctf, create: fix ctf_type_add of structs with unnamed members
Our recent commit to support unnamed structure members better ditched
the old ctf_member_iter iterator body in favour of ctf_member_next.
However, these functions treat unnamed structure members differently:
ctf_member_iter just returned whatever the internal representation
contained, while ctf_member_next took care to always return "" rather
than sometimes returning "" and sometimes NULL depending on whether the
dict was dynamic (a product of ctf_create) or not (a product of
ctf_open). After this commit, ctf_member_iter did the same.
It was always a bug for external callers not to treat a "" return from
these functions as if it were NULL, so only buggy callers could be
affected -- but one of those buggy callers was ctf_add_type, which
assumed that it could just take whatever name was returned from
ctf_member_iter and slam it directly into the internal representation of
a dynamic dict -- which expects NULL for unnamed members, not "". The
net effect of all of this is that taking a struct containing unnamed
members and ctf_add_type'ing it into a dynamic dict produced a dict
whose unnamed members were inaccessible to ctf_member_info (though if
you wrote that dict out and then ctf_open'ed it, they would magically
reappear again).
Compensate for this by suitably transforming a "" name into NULL in the
internal representation, as should have been done all along.
libctf/ChangeLog
2021-01-19 Nick Alcock <nick.alcock@oracle.com>
* ctf-create.c (membadd): Transform ""-named members into
NULL-named ones.
* testsuite/libctf-regression/type-add-unnamed-struct*: New test.
Nick Alcock [Tue, 19 Jan 2021 12:45:18 +0000 (12:45 +0000)]
libctf: lookup_by_name: do not return success for nonexistent pointer types
The recent work allowing lookups of pointers in child dicts when the
pointed-to type is in the parent dict broke the case where a pointer
type that does not exist at all is looked up: we mistakenly return the
pointed-to type, which is likely not a pointer at all. This causes
considerable confusion.
Fixed, with a new testcase.
libctf/ChangeLog
2021-01-19 Nick Alcock <nick.alcock@oracle.com>
* ctf-lookup.c (ctf_lookup_by_name_internal): Do not return the
base type if looking up a nonexistent pointer type.
* testsuite/libctf-regression/pptrtab*: Test it.
Alan Modra [Tue, 19 Jan 2021 02:49:18 +0000 (13:19 +1030)]
[GOLD] powerpc assertion failure
A testcase with only ifuncs can result in no plt section (ifunc plt
entries might instead be in iplt), which means we can get to this code
without a static link.
Nelson Chu [Fri, 8 Jan 2021 12:04:13 +0000 (04:04 -0800)]
ld: Just xfail riscv little endian targets for compressed1d.d test.
The sizes of compressed and uncompressed .debug_aranges are the same
for the riscv big endian targets, but different for the little endian
targets. The compress rule is fine and isn't broken by riscv, just the
original compressed1d.d fails by accident. Ideally, we should fill the
R_RISCV_ADD/SUB relocations when disabling relaxations in the assembler.
But before that, Jim already had written an alternative test compressed1d-alt
to relpace compressed1d.d for riscv, so we can only xfail the riscv little
endian targets in the short-term.
ld/
* testsuite/ld-elf/elf.exp (riscv_little_endian): Added. Return true
if target is riscv little endian. Otherwise, return false.
* testsuite/ld-elf/compressed1d.d: Only xfail the riscv little endian
targets by [riscv_little_endian].
H.J. Lu [Mon, 18 Jan 2021 14:25:05 +0000 (06:25 -0800)]
as: Automatically enable DWARF5 support
Currently
$ as -o x.o x.s
fails when x.s contains DWARF5 ".file 0" or ".loc 0" directives. Update
assembler to automatically enable DWARF5 support so that
$ gcc -S -g -c x.c
$ gcc -c x.s
works.
PR gas/27195
* dwarf2dbg.c (dwarf2_gen_line_info): Set dwarf_level to 5 if
needed.
(dwarf2_directive_filename): Likewise.
(dwarf2_directive_loc): Likewise.
* testsuite/gas/elf/dwarf-5-file0.d: Pass --gdwarf-3.
* testsuite/gas/lns/lns-diag-1.l: Remove the
"Error: file number less than one" errors.
H.J. Lu [Mon, 18 Jan 2021 00:43:45 +0000 (16:43 -0800)]
gold: Remove the circular IFUNC dependency in ifuncmain6pie
On Fedora 33 x86-64 with glibc 2.32-3, ifuncmain6pie failed with:
./ifuncmain6pie: IFUNC symbol 'foo' referenced in './ifuncmod6.so' is defined in the executable and creates an unsatisfiable circular dependency.
FAIL ifuncmain6pie (exit status: 127)
Remove non-JUMP_SLOT relocations against foo in ifuncmod6.so, which
trigger the circular IFUNC dependency.
* testsuite/ifuncmain6pie.c: Remove non-JUMP_SLOT relocations
against foo in ifuncmod6.so.
* testsuite/ifuncmod6.c: Likewise.
H.J. Lu [Mon, 18 Jan 2021 04:01:16 +0000 (20:01 -0800)]
ld/elf: Ignore section symbols when matching linkonce with comdat
When deciding if a single member comdat group section in file FOO should
be discarded by a linkonce section in file BAR, we check if 2 sections
define the same set of local and global symbols. When only one of the
files doesn't contain the unused section symbols in the symbol table,
such as object files generated by clang or GNU assembler with
the check will fail since one file has the extra unused section symbols.
We should ignore both undefined and section symbols in the symbol table
when making such a decision.
bfd/
PR ld/27193
* elflink.c (elf_create_symbuf): Also ignore section symbols.
ld/
PR ld/27193
* testsuite/ld-i386/i386.exp: Run PR ld/27193 test.
* testsuite/ld-i386/pr27193.dd: New file.
* testsuite/ld-i386/pr27193a.o.bz2: Likewise.
* testsuite/ld-i386/pr27193b.s: Likewise.
H.J. Lu [Sat, 16 Jan 2021 15:00:09 +0000 (07:00 -0800)]
ld/elf/x86: Don't compare IFUNC address in the shared object
On x86, glibc 2.33 starts to issue a fatal error message when calling
IFUNC function defined in the unrelocated executable from a shared
library.
1. Update x86 ELF linker to always convert IFUNC function defined in
position-dependent executable (PDE) to the normal function. GOT in PDE
will be updated by R_*_IRELATIVE at run-time.
2. Update PR ld/23169 tests not to compare function address of external
IFUNC function in the shared object to avoid calling the IFUNC function
defined in the unrelocated executable.
3. Remove pr23169e tests which call the IFUNC function defined in the
unrelocated position-independent executable from a shared library.
H.J. Lu [Thu, 14 Jan 2021 13:23:58 +0000 (05:23 -0800)]
bfin: Skip non SEC_ALLOC section for R_BFIN_FUNCDESC
Linker should never generate dynamic relocations for relocations in
non-SEC_ALLOC section which has no impact on run-time behavior. Such
relocations should be resolved to 0.
PR ld/26688
* elf32-bfin.c (bfinfdpic_relocate_section): Skip non SEC_ALLOC
section for R_BFIN_FUNCDESC.
Kyrylo Tkachov [Tue, 5 Jan 2021 17:39:04 +0000 (17:39 +0000)]
aarch64: Remove support for CSRE
This patch removes support for the CSRE extension from aarch64
gas/objdump.
CSRE (FEAT_CSRE) is part of the Future Architecture Technologies program
and at this time Arm is withdrawing this particular feature.
The patch removes the system registers and the CSR PDEC instruction.
H.J. Lu [Sat, 9 Jan 2021 05:38:39 +0000 (21:38 -0800)]
elf: Verify section size for mixed ordered/unordered inputs
When fixing up SHF_LINK_ORDER, issue a fatal error if the output section
size is increased. Otherwise, bfd_set_section_contents will fail later
when attempting to write contents past the end of the output section.
PR ld/26256
PR ld/27160
* elflink.c (elf_fixup_link_order): Verify that fixing up
SHF_LINK_ORDER doesn't increase the output section size.