CVSROOT: /cvs/purrs
Module name: purrs
Changes by: pescetti(a)cs.unipr.it 2004-05-11 00:26:04
Modified files:
tests : rrs_driver.cc
Log message:
rrs_driver in regression-testing mode can now solve every recurrence
multiple times. Pass a number to -r (e.g., rrs_driver -r 16 -v < heap),
go for a cup of coffee and see weird bugs popping.
Patches:
http://www.cs.unipr.it/cgi-bin/cvsweb.cgi/purrs/tests/rrs_driver.cc.diff?cv…
CVSROOT: /cvs/purrs
Module name: purrs
Changes by: pescetti(a)cs.unipr.it 2004-05-08 18:24:12
Modified files:
. : BUGS
Log message:
Is Bug #3 the Mother of All Bugs?
Solving and verifying the same recurrence fails after several
attempts.
Patches:
http://www.cs.unipr.it/cgi-bin/cvsweb.cgi/purrs/BUGS.diff?cvsroot=purrs&r1=…
Dear Richard,
apparently, CLN 1.1.5 does not compile with the
newly released GCC 3.4.0. My understanding is
that CLN 1.1.5 is relying on some non-standard
language constructs that are now rejected by g++.
Are there any plans to solve this problem?
All the best,
Roberto
--
Prof. Roberto Bagnara
Computer Science Group
Department of Mathematics, University of Parma, Italy
http://www.cs.unipr.it/~bagnara/
mailto:bagnara@cs.unipr.it
bernard.parisse wrote:
>
>>
>> template <class T>
>> monomial<T> Untrunc1(const T & t,int j=0){
>> index_m new_i;
>> new_i.i()->push_back(j);
>> // ^^ HERE!
>> return monomial<T>(t,new_i);
>> }
>>
> Ooops... Thanks! That's the kind of bug that are difficult
> to find. It's strange that earlier version of gcc did
> not complain... Does it compile after this fix?
No, unfortunately. The fixed code triggers a bug in GCC 3.4.0.
I have already submitted a bug report:
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15299
I will investigate it further in a couple of days.
Cheers,
Roberto
--
Prof. Roberto Bagnara
Computer Science Group
Department of Mathematics, University of Parma, Italy
http://www.cs.unipr.it/~bagnara/
mailto:bagnara@cs.unipr.it
Dear Sir,
I have downloaded Giac 0.4.0 from
ftp://ftp-fourier.ujf-grenoble.fr/xcas/giac_unstable.tgz
and tried to compile it with GCC 3.4.0. The compilation
aborts very soon as follows:
...
c++ -DHAVE_CONFIG_H -I. -I. -I.. -I. -I.. -g -O2 -c gausspol.cc -fPIC -DPIC -o
.libs/gausspol.lo
In file included from poly.h:26,
from sym2poly.h:23,
from sym2poly.cc:32:
monomial.h: In function `giac::monomial<T> giac::Untrunc1(const T&, int)':
monomial.h:151: error: insufficient contextual information to determine type
monomial.h:151: error: base operand of `->' is not a pointer
monomial.h:302: confused by earlier errors, bailing out
In file included from poly.h:26,
from gausspol.h:22,
from gausspol.cc:28:
monomial.h: In function `giac::monomial<T> giac::Untrunc1(const T&, int)':
monomial.h:151: error: insufficient contextual information to determine type
monomial.h:151: error: base operand of `->' is not a pointer
monomial.h:302: confused by earlier errors, bailing out
make[2]: *** [sym2poly.lo] Error 1
make[2]: *** Waiting for unfinished jobs....
make[2]: *** [gausspol.lo] Error 1
make[2]: Leaving directory `/usr/local/distrib/giac-0.4.0/src'
make[1]: *** [all-recursive] Error 1
make[1]: Leaving directory `/usr/local/distrib/giac-0.4.0'
make: *** [all-recursive-am] Error 2
I am unsure about whether this is an indication that Giac 0.4.0
uses some non-standard C++ features that are now rejected by
g++ or if we are presented with a bug in g++.
What do you think?
All the best, and thanks in advance for your time.
Roberto Bagnara
--
Prof. Roberto Bagnara
Computer Science Group
Department of Mathematics, University of Parma, Italy
http://www.cs.unipr.it/~bagnara/
mailto:bagnara@cs.unipr.it
