Re: armhf multiarch tuple
> On 17 April 2012 18:36, Wookey <firstname.lastname@example.org> wrote:
> > debian armel is currently
> > built for v4t but is likely to move forward to v5 at some point.
> -1 as we have and others have products that are arm4vt and that ship
> with Debian as its standard system. How big is the speed advantage for
> v5 users when optimizing for v5 instead of v4t? Enough to warrant the
> exclusion of our v4t parts?
ARMv5te introduces some benefits for code using doubleword (64-bit) values.
In practice most code doesn't use enough 64-bit values for this to be
The other benefit of v5 is improved interworking.
For ARM code the difference is probably not that great.
For Thumb code it's more significant, though I don't have real numbers for how
important it is.
However Debian builds everything as ARM, so this only really effects third
party binaries. The EABI means everything is binary compatible. i.e. anyone
who wants to build their code as ARMv5 Thumb (typically optimized for code
size) can already do so.
In short if we wanted a code-size optimized build then there's a good argument
for ARMv5te+Thumb. For general performance I don't expect moving from v4t to
v5 to give significant gains.