This is the mail archive of the
gcc-bugs@gcc.gnu.org
mailing list for the GCC project.
[Bug tree-optimization/19910] ICE with -ftree-loop-linear
- From: "dberlin at gcc dot gnu dot org" <gcc-bugzilla at gcc dot gnu dot org>
- To: gcc-bugs at gcc dot gnu dot org
- Date: 11 Feb 2005 21:09:54 -0000
- Subject: [Bug tree-optimization/19910] ICE with -ftree-loop-linear
- References: <20050211190826.19910.dalej@apple.com>
- Reply-to: gcc-bugzilla at gcc dot gnu dot org
------- Additional Comments From dberlin at gcc dot gnu dot org 2005-02-11 21:09 -------
This is caused by generating a complex linear offset in the transformed loop,
which the "back to gcc" converter doesn't handle ATM, since they were never
generated before.
(I didn't completly minimize linear offset generation. If i did, you'd notice
that this loop really goes from 5 to MAX (2, j) or something like that)
Loop w
step size = -1
linear offset:
Linear expression: -i + j constant: 5 invariants: denominator: 1
lower bound:
Linear expression: i - j constant: 0 invariants: denominator: 1
upper bound:
Linear expression: i - j constant: -3 invariants: denominator: 1
Linear expression: i constant: -5 invariants: denominator: 1
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=19910