[Bug tree-optimization/82627] [8 regression] r253809 breaks test cases gcc.dg/graphite/interchange-3.c -7.c, -9.c, and uns-interchange-9.c

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=82627

Richard Biener <rguenth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |WAITING
   Last reconfirmed|                            |2017-10-20
   Target Milestone|---                         |8.0
     Ever confirmed|0                           |1

--- Comment #1 from Richard Biener <rguenth at gcc dot gnu.org> ---
Can you provide full dumps before/after please?  The patch removed some "stray"
stmts that might have tricked ISL into doing some transform that ended up being
useful to enable tiling.

Note that on x86_64 the dump has

not tiled

so the dump scanning isn't exactly useful... (most of the testcases are
broken - at some point they all got adjusted to PASS without any sensible
look :/)

The test shouldn't have any target dependence whatsoever though :/