Re: Phoronix benchmarks DF 3.2 vs Ubuntu - question

[Alex Hornung <alex@xxxxxxxxxxxxxxx>]

On 2012-11-28 09:13, elekktretterr@exemail.com.au wrote:
>>  From what I could tell from a casual read, it looks like this range
>> of benchmarks was more compute related - so this particular test
>> would probably be more related to libc / compiler & math lib items
>> than any of the IO path / threading optimizations of late.
>>
>> Unfortunately the 'test platform' details on page 1 don't list
>> anything about the ubuntu compiler, which makes it hard to
>> determine how much is compiler vs libc. Yes I could look this up..
>>
>> A comparison of the actual generated code by someone well versed
>> in these things might be useful here.
>
> Ive seen some benchmarks with kFreeBSD/GNU and it seems to perform 
> almost
> as well as Linux, so I think you are correct that this has very 
> little to
> do with kernel but rather libc.

I didn't really want to comment on this, but well - here I go anyway. 
Most (all?) the benchmarks mentioned in the article are 
hardware/compiler benchmarks. Whoever tells you that Dhrystone is a 
benchmark for anything but the compiler (and the underlying hardware) 
doesn't know what he's talking about. At a decent optimization level it 
won't even touch libc functions other than for reporting (it'll use the 
compiler's optimized strcpy, etc).

Cheers,
Alex