Re: [PATCH 4/4] arm: Enable ARM mode for armv6 strlen

[Wilco Dijkstra <Wilco dot Dijkstra at arm dot com>]

Phil Blundell wrote:

> But here we are, two decades later, and the landscape is a bit
> different.  It's not entirely clear that -marm/-mthumb are very useful
> options in an ARMv7 world because the user shouldn't need to care.  In
> an ideal world the compiler would be able to predict for any given
> function whether T32 or A32 encoding would give the best results
> (either speed or space according to the selected optimisation settings)
> and proceed accordingly.  Unfortunately in practice I don't think
> we're quite there yet and for critical code there's still an element of
> "try building it both ways round and see which one runs quickest" which
> means the compiler flags probably are still necessary.  Users might
> even legitimately wish to try that experiment with glibc, so forbidding
> them to compile it under -marm would not obviously be the right thing
> to do.

If there are still cases where GCC generates worse code for Thumb-2,
then those can be addressed easily. Our very first Thumb-2 compiler 
had performance within 2% on ARM1156T2 more than 2 decades ago!

Things have improved a lot since then, and system wide effects of code
footprint have become more important, so using only Arm on a modern
core just doesn't make any sense today.

Wilco