Re: Release 2.21 - Pre tests

[Tristan Gingold <gingold at adacore dot com>]

On Dec 1, 2010, at 1:14 AM, Ian Lance Taylor wrote:

> Tristan Gingold <gingold@adacore.com> writes:
> 
>> On Nov 23, 2010, at 10:03 PM, Ian Lance Taylor wrote:
>> 
>>> Tristan Gingold <gingold@adacore.com> writes:
>>> 
>>>> 2) I wasn't able to build gold with the compilers I used (gcc 4.1.2 from RH or gcc 4.1.0 from SUSE)
>>> 
>>> What failed?
>> 
>> ../../binutils-2.20.90/gold/arm.cc: In instantiation of 'const size_t <unnamed>::Target_arm<false>::fake_relnum_for_stubs':
>> ../../binutils-2.20.90/gold/arm.cc:8627:   instantiated from 'bool<unnamed>::Target_arm<big_endian>::Relocate::relocate(const gold::Relocate_info<32, big_endian>*, <unnamed>::Target_arm<big_endian>*, gold::Output_section*, size_t, const elfcpp::Rel<32, big_endian>&, unsigned int, const gold::Sized_symbol<32>*, const gold::Symbol_value<32>*, unsigned char*, <unnamed>::Arm_address, gold::section_size_type) [with bool big_endian = false]'
>> ../../binutils-2.20.90/gold/target-reloc.h:334:   instantiated from 'void gold::relocate_section(const gold::Relocate_info<size, big_endian>*, Target_type*, const unsigned char*, size_t, gold::Output_section*, bool, unsigned char*, typename elfcpp::Elf_types<size>::Elf_Addr, gold::section_size_type, const gold::Reloc_symbol_changes*) [with int size = 32, bool big_endian = false, Target_type = <unnamed>::Target_arm<false>, int sh_type = 9, Relocate = <unnamed>::Target_arm<false>::Relocate]'
>> ../../binutils-2.20.90/gold/arm.cc:9263:   instantiated from 'void<unnamed>::Target_arm<big_endian>::relocate_section(const gold::Relocate_info<32, big_endian>*, unsigned int, const unsigned char*, size_t, gold::Output_section*, bool, unsigned char*, <unnamed>::Arm_address, gold::section_size_type, const gold::Reloc_symbol_changes*) [with bool big_endian = false]'
>> ../../binutils-2.20.90/gold/arm.cc:11770:   instantiated from here
>> ../../binutils-2.20.90/gold/arm.cc:2140: internal compiler error: in make_rtl_for_nonlocal_decl, at cp/decl.c:5067
>> Please submit a full bug report,
> 
> I just tried building gold with gcc 4.1.3 20080704 from the 4.1 release
> branch and it succeeded.
> 
> I'm not sure whether to do anything about this.

Not sure too.  Maybe we should document that gcc 4.1.3 is know to work too, while gcc 4.1.0 - 4.1.2 are known to fail.

Anyway, we will see if we have complaints from users.

Tristan.