[Bug target/70341] [4.9/5/6/7 Regression] cost model for addresses is incorrect, slsr is using reg + reg + CST for arm

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70341

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
            Summary|[4.9/5/6/7 Regression] Code |[4.9/5/6/7 Regression] cost
                   |size increase due to        |model for addresses is
                   |earlier switch case         |incorrect, slsr is using
                   |lowering.                   |reg + reg + CST for arm

--- Comment #5 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
I am suspecting what is happening on arm is the cost model for addresses is
incorrect and allowing [r1 + r2 + const] so slsr is picking that.

This is why it works correctly on aarch64 because the cost model is correct and
rejects that mode.