Re: backports to 2.25 branch

[Iain Buclaw <ibuclaw at gdcproject dot org>]

On 11 February 2015 at 02:41, Alan Modra <amodra@gmail.com> wrote:
> Hi Tristan,
> I have a large list of backports I'd like to apply to the branch.
> Most are PowerPC related, but you'll see an ARM and a SH fix too, as
> well as general fixes.  OK?
>
> 0cfd832f Recognize new DWARFv5 C11, C++11 and C++14 DW_LANG constants
> 8bc10620 dwarf.c handle new DWARFv5 C11, C++11 and C++14 DW_LANG constants
> 51aecdc5 ppc64_elf_edit_opd revamp
> 53df40a4 Sort relocs output by ld -r
> 31d999a5 ppc64_elf_edit_opd revamp, take 2
> 28dbcedc Don't sort ld -r relocs for mips
> e00e8198 Don't return DW_AT_name for function name in C++
> 4c219c2e Use autoconf check for long long in binutils
> 6cabe1ea Copy relocations against protected symbols
> de287215 Set bfd_error in _bfd_elf_adjust_dynamic_copy
> c322dea4 PowerPC register numbers in DWARF
> 6ddfe5b4 Display DW_LANG_C11 as (C11)
> 04c6a44c Set ppc COMMONPAGESIZE to 64k
> e30880c2 [GOLD] Fix powerpc overflow check
> 18d6a79d Move support code for linker script DEFINED to ldexp.c
> cd8e2bcf Correct logic for "defined by object"
> 4cc2bf08 Don't PROVIDE over top of common symbols
> 12b2843a Use a symbol flag bit to mark linker defined symbols
> ca0694ad Don't create .eh_frame_hdr on shared lib bfd
> 0abb10c8 Delete unnecessary code copying SHF_SH5_ISA32 flag
> b38ead21 Assorted compiler warning fixes
> bba33ab1 Fix build without makeinfo from release binutils tar
> 8539e4e8 Fix ARM fail of gap test
> ca55926c Omit section dynsyms for any linker created section
> c4621b33 Fix garbage collection of common symbols
> 1c9177d9 Fix garbage collection of common symbols for powerpc64
> 43d66c95 run_dump_test can now check linker mapfiles
> c05b575a Don't evaluate unneeded PROVIDE expressions
> 481765cd Update expected test results for 32-bit hosts
> b01a4b04 [GOLD] Correct powerpc64 ifunc plt entry test
> b86ac8e3 Correct PowerPC64 local-dynamic TLS linker optimization
> 0f81d3f0 Correct GOLD PowerPC64 local-dynamic TLS linker optimization
> 989f9879 Don't segfault or assert on NULL tls_sec
>

If you are going ahead with this, I'd like to also request backporting
the below too, if you recall me raising this just after the release
freeze.

5af04e20 Use strtod instead of strtold in libiberty/d-demangle.c


Regards
Iain