Re: GCC and binutils updates for buster

Hi,

On 17/07/2018 17:56, Aurelien Jarno wrote:
> Hi,
>
> On 2018-07-16 17:59, Matthias Klose wrote:
>> The Debian release team lists toolchain support for our release architectures,
>> and according to [4], the amd64, i386, armel, armhf, arm64 architectures are
>> supported as primary architectures, and s390x is supported as a secondary
>> architectures.  Some notes on other candidates for release architectures:
>>
>>   - armel: The armv4t default isn't used very much anymore, and we had
>>     issues in the past.
>>
>>   - armhf: While arm-linux-gnueabihf is not explicitly listed as a primary
>>     architecture, I'm told that the arm-linux-armeabi triplet covers the
>>     hard float variants as well.
>>
>>   - ppc64el: Not documented as primary architecture, but according to the
>>     backend maintainers the powerpc64-linux-gnu triplet includes the le variant.
>>
>>   - mips*: There is no support for any mips-linux target either as a primary
>>     or secondary release architecture (only bare metal), which matches the
>>     experience with mips specific issues for the past Debian releases.
> Could we ask the backends maintainers of the mipsisa64-elf backend,
> similarly to what has been done for ppc64el? Last time I asked Matthew
> Fortune almost 2 years ago, he told me that the embedded mipsisa64-elf
> platform (a primary GCC platform) can be used to demonstrate pretty
> much any bug in mips therefore making every mips architecture a primary
> one.
>
> Let ask him his opinion publicly. I have Cced: him.
>
> Aurelien
>
>
>> [4] https://gcc.gnu.org/gcc-8/criteria.html
In terms of compiler support, i.e. codegen, then all MIPS compilers can 
be used (with appropriate options) to generate code suitable for Linux. 
There are a couple of minor things that may be configure time dependent 
but nothing that would mean a Linux triplet works sufficiently different 
to a bare metal to cause problems. The rest of the toolchain does 
diverge a bit but even that is relatively minor and the choice of 
targets for GCC does not necessarily mean that binutils or glibc is any 
less well tested for MIPS Linux triplets just because GCC doesn't list 
them as primary/secondary.

I'd suggest looking at whether there is something fundamentally 
problematic with MIPS Linux rather than just using the primary/secondary 
release target list.

Matthew