This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
Re: [libc/string] State of PAGE_COPY_FWD / PAGE_COPY_THRESHOLD
On Thursday 10 November 2016 02:59 PM, Florian Weimer wrote:
> Another option would be to get the complete implementation of memcmp
> etc. fom the vDSO, I think.
Hmm, that is an interesting thought.
> In the ARM context, there were also suggestions that we should support
> big/little asymmetric multi-processing. I hope we don't have to do
> this. It would seriously limit IFUNC selection and other
> microarchitecture-based optimizations.
Even if we don't actually have optimal routines for big.LITTLE, we still
have to account for the possibility of their existence, hence the need
to peek into all online processors in the hope (because some could be
offline, giving an incorrect view of the system) of guessing the correct
execution environment.
Siddhesh