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: Updated rwlock-in-C patchkit


On 04/13/2014 11:28 AM, Torvald Riegel wrote:
> On Sat, 2014-04-12 at 21:40 +0200, OndÅej BÃlka wrote:
>> On Sat, Apr 12, 2014 at 01:23:32PM -0400, Carlos O'Donell wrote:
>>> On 04/11/2014 04:36 PM, Torvald Riegel wrote:
>>>> On Tue, 2014-03-25 at 16:24 -0700, Andi Kleen wrote:
>>>>> Fixed the ChangeLog as requested. Otherwise no changes.
>>>>>
>>>>> Please consider merging.
>>>>
>>>> Based on a quick scan this looks good to me, and I definitely like the
>>>> direction.  However, I believe this won't get accepted unless you show
>>>> numbers by providing a benchtest for the uncontended case.
>>>
>>> Fully agreed.
>>>
>>> While I trust that you think it's faster I would like an objective way
>>> to measure this on the systems that I care about.
>>>
>> Carlos that was already addressed in this thread. See 
>>
>> https://www.sourceware.org/ml/libc-alpha/2014-03/msg00739.html
> 
> I didn't see a submitted microbenchmark in this case, which I believe is
> what Carlos is talking about.
> 

That's right. I'd like to see the whatever code was used to benchmark
the performance submitted as a microbenchmark so I can run it myself
and verify. Again, it's not that I don't trust Andi, but an objective
evaluation is always going to be the best defense for these changes.

Cheers,
Carlos.


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