This is the mail archive of the
fortran@gcc.gnu.org
mailing list for the GNU Fortran project.
[gfortran] patch for pr 21130 - 38822 lines of Fortran 90 takes more than 10 minutes to compile on a dual 3GHz P4 Linux box with lots of RAM
- From: Bud Davis <bdavis9659 at sbcglobal dot net>
- To: gcc-patches at gcc dot gnu dot org, fortran at gcc dot gnu dot org
- Date: Tue, 28 Mar 2006 17:40:09 -0800 (PST)
- Subject: [gfortran] patch for pr 21130 - 38822 lines of Fortran 90 takes more than 10 minutes to compile on a dual 3GHz P4 Linux box with lots of RAM
- Domainkey-signature: a=rsa-sha1; q=dns; c=nofws; s=s1024; d=sbcglobal.net; h=Message-ID:Received:Date:From:Subject:To:MIME-Version:Content-Type:Content-Transfer-Encoding; b=rApQ9jbvgjXLdLQNRB1QKn0RbJ8TdXQwm4XWX7zsV6SR0Z0erprXABUChxsYRwEgmPUB7g1DLug2+9eN19yshHoebKm/P6GQcGv1h9/XxqYTHP/z+MbKBESacDvusvTvqRxyPsoeqzSbAmCOE8G3YkkrjFMABFXQm6gLfHbFs24= ;
problem was the recursive routine was being called on
the whole tree for every UNUSED symbol in the tree.
when you get a lot of code (like the submitters
example), it takes a long time.
changed the recursive routine to keep track of how
many symbols that it dealt with, thus it knows that
the tree needs to be traversed at least one more time.
tested on i686 / gnu / linux with no new regressions
in the testsuite.
submitters example code compiles in ~45 seconds
(ArrayTest_ArrayOps_Impl.f90) vs. the ~ 8 minutes
time before .
no testsuite file provided; this is a performance, not
a correctness problem.
--bud
2006-03-28 Bud Davis <bdavis9659@sbcglobal.net>