Seeking consensus on BZ 16734

[pinskia@gmail.com]

> On Feb 1, 2015, at 8:09 PM, H.J. Lu <hjl.tools@gmail.com> wrote:
> 
>> On Sun, Feb 1, 2015 at 6:13 PM, Paul Pluzhnikov <ppluzhnikov@google.com> wrote:
>> Greetings,
>> 
>> The arguments in
>>  https://sourceware.org/bugzilla/show_bug.cgi?id=16734
>> 
>> and in threads linked from it:
>>  https://sourceware.org/ml/libc-alpha/2006-11/msg00061.html
>>  https://sourceware.org/ml/libc-alpha/2006-11/msg00064.html
>> 
>> seem pretty convincing to me. It was probably a good idea to do it in
>> 2006, and it seems an even better idea now that everything is
>> multithreaded.
>> 
>> Can we just do it?
>> 
> 
> Do we have any current performance data on this?

Note I think not using mmap is even better for targets that use page sizes besides 4k. Most distros for AARCH64 are going to be using 64k pages.  So wasting 63k worth of address space is not good, especially for ilp32. 

Thanks,
Andrew

> 
> 
> -- 
> H.J.