Fix PR32540 exponential time behavior in PRE

["Sebastian Pop" <sebpop at gmail dot com>]

Hi,

The attached patch fixes the PR32540 and PR33922.  The patch is based
on the patch that Nick Clifton sent earlier, but adapted to stop the
computation before the phi translation.  I also changed the default
cutoff to 100 anticipable values, but I'm not sure about the best
value here.  Daniel what is your opinion on this cutoff value?

The patch regstrapped on amd64-linux.  Okay for trunk?

Sebastian
-- 
AMD - GNU Tools

---------- Forwarded message ----------
From:  <spop@gcc13.fsffrance.org>
Date: Nov 3, 2007 7:10 PM
Subject: [regtest] Results for 2007_11_03_14_25_26_pr32540.diff.asc on
x86_64-unknown-linux-gnu
To: sebpop@gmail.com


Checker: (2007_11_04_00_10_55): (cat
/home/spop/trunk-tester/state/testing/patched/report
there are no regressions with your patch.
Checker: (2007_11_04_00_10_55): tac)
Checker: (2007_11_04_00_10_55): FAILs with patched version:
Checker: (2007_11_04_00_10_55): (cat
/home/spop/trunk-tester/state/testing/patched/failed
gcc.sum gcc.c-torture/parameter execute/mayalias-2.c
gcc.sum gcc.c-torture/execute/mayalias-3.c
gcc.sum gcc.dg/torture/builtin-pow-mpfr-1.c
libmudflap.sum libmudflap.c++/pass41-frag.cxx
libstdc++.sum abi_check
Checker: (2007_11_04_00_10_55): tac)
Checker: (2007_11_04_00_10_55): FAILs with pristine version:
Checker: (2007_11_04_00_10_55): (cat
/home/spop/trunk-tester/state/trunk/129872/failed
gcc.sum gcc.c-torture/execute/mayalias-2.c
gcc.sum gcc.c-torture/execute/mayalias-3.c
gcc.sum gcc.dg/torture/builtin-pow-mpfr-1.c
libmudflap.sum libmudflap.c++/pass41-frag.cxx
libstdc++.sum abi_check
Checker: (2007_11_04_00_10_55): tac)
Checker: (2007_11_04_00_10_55): The files used for the validation of
your patch are stored in
/home/spop/trunk-tester/state/patched/2007_11_04_00_10_55 on the
tester machine.

Attachment: 2007_11_03_14_25_26_pr32540.diff.asc
Description: Text document