[Bug target/54089] [SH] Refactor shift patterns

[olegendo at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54089

--- Comment #7 from Oleg Endo <olegendo at gcc dot gnu.org> 2012-08-09 23:36:10 UTC ---
Kaz, another thing I'm a bit unsure about ...

#define SH_DYNAMIC_SHIFT_COST \
  (TARGET_HARD_SH4 ? 1 : TARGET_DYNSHIFT ? (optimize_size ? 1 : 2) : 20)

Do you have any idea, why this is not simply
#define SH_DYNAMIC_SHIFT_COST (TARGET_DYNSHIFT) ?

I've checked the HW manuals for SH2A, SH3 and SH4 and all state that dynamic
shifts take one cycle, i.e. are as fast/slow as the other constant shift
insns...