This is the mail archive of the libc-alpha@sourceware.org mailing list for the glibc project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: 2.26 release blockers?



On 05/07/2017 17:16, Carlos O'Donell wrote:
> On 07/05/2017 02:19 PM, Adhemerval Zanella wrote:
>>
>>
>> On 05/07/2017 10:51, Joseph Myers wrote:
>>> On Wed, 5 Jul 2017, Adhemerval Zanella wrote:
>>>
>>>>> The thread types headers refactoring, to make them usable for both C11 
>>>>> threads and pthreads, went in some time ago.
>>>>>
>>>>> All that architecture-specific header content will get used in C11 threads 
>>>>> and it's far from obvious that any issues that appear with C11 threads 
>>>>> would also appear with pthreads.
>>>>
>>>> It was not clear to me if you still consider C11 threads patches a
>>>> disruptive for arch-testing that can't not get validated in current
>>>> cross-compiling testing.  
>>>
>>> I think the ABI baselines are adequately validated by cross-compilation 
>>> testing, but there's a risk of architecture-specific problems that only 
>>> show up in C11 threads execution testing being missed.
>>>
>> I do see your rationale, but we are already pushing changes without much
>> testing in master that are much more disruptive IMHO (maybe not for
>> architecture cases, but on performance for instance) such as single thread
>> FILE optimization and thread local cache.  
> 
> Each issue needs to be discussed independently.
> 
> Each issue has their own risks and rewards.

I understand that and I give you that both examples I gave had a more 
thoughtfully review before inclusion. And I also understand C11 threads
seems to not be an urgent upcoming new feature for 2.26 (given relative
lack of interest and review).

> 
>> But I think we can postpone C11 threads for 2.27, although for the specific
>> architecture where you already posted some testing results (arm hard-float,
>> powerpc soft-float, and mips), I think it would be feasible to run the C11
>> threads tests on emulated qemu system to validate them.
>  
> Can we commit to checking in C11 as soon as 2.27 opens?
> 
> What else do we need to do to get there?

Mostly some reviews of the patch themselves. Mostly are straightforward wrapper
and code examples, maybe expect the thrd_* one that changes some internal
pthread code for c11 call format.  The documentation got some review, but I 
am not sure if it would be better to glue it to threads.texi or let in its
own file.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]