[Bug tree-optimization/68644] [6/7 Regression] FAIL: gcc.dg/tree-ssa/ivopts-lt-2.c scan-tree-dump-times ivopts "PHI

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68644

--- Comment #15 from Richard Biener <rguenth at gcc dot gnu.org> ---
For PA it selects i as IV (reasonable).  For power it looks like it tries to
use
sth with auto-inc (pre-inc).  Not sure if that's reasonable cost-wise.  You
decide (and then skip, like for PA, if so).