[Bug rtl-optimization/33410] [4.2/4.3 regression] ICE in iv_analyze_expr, at loop-iv.c:934

[rth at gcc dot gnu dot org]

------- Comment #21 from rth at gcc dot gnu dot org  2008-02-06 17:38 -------
Created an attachment (id=15109)
 --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=15109&action=view)
fix, try 2

"I see," said the blind man.

It turns out that the emission of the libcall had been responsible for
copying the result from the hard register into the target pseudo.  Well,
it turns out there's an easy way to retain the libcall, but use a 
different expression; one that ought not confuse the loop optimizer.


-- 

rth at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
  Attachment #15066|0                           |1
        is obsolete|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=33410