jakobkroeker / test_m2 Goto Github PK

View Code? Open in Web Editor NEW

0.0 0.0 0.0 0 B

record M2 issues

test_m2's People

Watchers

test_m2's Issues

on fedora 32 bit I get an error

in package IntegralClosure.m2
-- -- M2-comint -- {* hash: -1496599229 *}

i1 : S = ZZ/32003[a,b,c];

i2 : F = a^2_b^2_c+a^3+b^3+c^3

  2 2     3    3    3

o2 = a b c + a + b + c

o2 : S

i3 : J = ideal jacobian ideal F

            2      2    2        2   2 2     2

o3 = ideal (2a_b c + 3a , 2a b_c + 3b , a b + 3c )

o3 : Ideal of S

i4 : time integralClosure J
stdio:4:6:(3): error: expected lists of the same length
-- used 0.175256 seconds

build error on fedora fc18 32 bit

configure option

i386-pc-linux-gnu

/home/jakob/newwork/Projects/M2-master/M2/BUILD/master/usr-host/lib/libgmpxx.so: undefined reference to `__atomic_fetch_add_4'

example times out / out of mem (from 00-rawCharSeries.m2"

kk = QQ -- times out/ out of mem
R = kk[x1,x2,x3,x4,x5,x6, MonomialOrder=>Lex] --
I = ideal(x1^2_x2^2_x5^8_x6^8-x3^4_x4^4,x2^8_x3^2_x4^8_x5^2-x1^4_x6^4,x1^8_x3^8_x4^2_x6^2-x2^4_x5^4)
debug Core
m = gens I
rawCharSeries raw m

this is not test M2, but an uncommited change (see if useful or not)

--- a/M2/libraries/gc/Makefile.in
+++ b/M2/libraries/gc/Makefile.in
@@ -41,6 +41,7 @@ LICENSEFILES = README.QUICK
 # That's why it's better for sharable libraries to be distributed as separate packages.

 # For gc "make check" requires far too much memory
+TLIMIT := 250
 VLIMIT := 2500000
 MLIMIT := $(VLIMIT)
 # oddly, unless the VLIMIT is bigger than 2GB, the test subthread_create fails with this message

--- a/M2/Macaulay2/e/dmat.cpp
+++ b/M2/Macaulay2/e/dmat.cpp
@@ -245,17 +245,17 @@ namespace ffpackInterface
   {
     DMatZZpFFPACK N(mat);

-    size_t * prof; // this is where the result will be placed
+    size_t * prof=NULL; // this is where the result will be placed
     size_t rk;
     if (row_profile)
-      rk = FFPACK::RowRankProfile(mat.ring().field(),
-                                  mat.numRows(),
-                                  mat.numColumns(),
-                                  N.array(),
-                                  mat.numColumns(),
-                                  prof);
+      rk = FFPACK::RowRankProfile (mat.ring().field(),
+                                     mat.numRows(),
+                                     mat.numColumns(),
+                                     N.array(),
+                                     mat.numColumns(),
+                                     prof);
     else
-      rk = FFPACK::ColumnRankProfile(mat.ring().field(),
+      rk = FFPACK::ColumnRankProfile  (mat.ring().field(),
                                      mat.numRows(),
                                      mat.numColumns(),
                                      N.array(),
@@ -274,33 +274,40 @@ namespace ffpackInterface
                    bool row_profile,
                    std::vector<size_t>& result_profile)

-  {
+  {               
     DMatZZpFFPACK N(mat);
-    
-    size_t * prof; // this is where the result will be placed
+
+    size_t * prof=NULL; // this is where the result will be placed
     size_t rk;
+
     if (row_profile)
-      rk = FFPACK::RowRankProfile(mat.ring().field(),
-                                  mat.numRows(),
-                                  mat.numColumns(),
-                                  N.array(),
-                                  mat.numColumns(),
-                                  prof);
+    {
+      rk = FFPACK::RowRankProfile (mat.ring().field(),
+                                     mat.numRows(),
+                                     mat.numColumns(),
+                                     N.array(),
+                                     mat.numColumns(),
+                                     prof);
+    }
     else
-      rk = FFPACK::ColumnRankProfile(mat.ring().field(),
+    {
+      rk = FFPACK::ColumnRankProfile  (mat.ring().field(),
                                      mat.numRows(),
                                      mat.numColumns(),
                                      N.array(),
                                      mat.numColumns(),
                                      prof);
-    
-    result_profile.resize(0);
-    for (size_t i=0; i<rk; i++)
+   }
+   result_profile.resize(0);
+   for (size_t i=0; i<rk; i++)
       result_profile.push_back(prof[i]);

-    delete [] prof;
+   delete [] prof;
+
   }
+

+
 }; // namespace ffpackInterface

radical: out of mem

ideal(13_x_1^2_x_2+8_x_2_x_3-1,(1/3)_x_1^3+(1/5)_x_2^3-7/2,2_x_1_x_2_x_3+6_x_1_x_3^2+4_x_2_x_3-(4/5)_x_3^2)

fc18: install-info: Command not found

caused by: install-info is not in PATH
(/sbin/install-info)
workaround: add to PATH :/sbin:

cannot download libfplll

http://perso.ens-lyon.fr/damien.stehle/fplll//libfplll-4.0.4.tar.gz

bug or feature? on ubuntu: test fails for 'monomialIdeals' but make continues.

make -f Makefile.ubuntu64

ask mike for idea testing gb

led I be an ideal and gI its groebner bases
my test may not detect the case when gI is (incorrectly) bigger than I
Question: how to test that?

setup Macaulay2 in a way that is possible to track down memory issues (not freeing memory)

gfortran link issue on ubuntu64 for lapack

`/home/jakob/Projects/M2-master/M2/BUILD/master/builds.tmp/ubuntu64.production/Macaulay2/e/unit-tests

/home/jakob/Projects/M2-master/M2/BUILD/master/builds.tmp/ubuntu64.production/usr-host/lib/liblapack.a(dhseqr.o): In function `dhseqr_':
/home/jakob/Projects/M2-master/M2/BUILD/master/builds.tmp/ubuntu64.production/libraries/lapack/build/lapack-3.5.0/SRC/dhseqr.f:454: undefined reference to `_gfortran_concat_string'
/home/jakob/Projects/M2-master/M2/BUILD/master/builds.tmp/ubuntu64.production/usr-host/lib/liblapack.a(dormbr.o): In function `dormbr_':
/home/jakob/Projects/M2-master/M2/BUILD/master/builds.tmp/ubuntu64.production/libraries/lapack/build/lapack-3.5.0/SRC/dormbr.f:274: undefined reference to `_gfortran_concat_string'
/home/jakob/Projects/M2-master/M2/BUILD/master/builds.tmp/ubuntu64.production/libraries/lapack/build/lapack-3.5.0/SRC/dormbr.f:282: undefined reference to `_gfortran_concat_string'
/home/jakob/Projects/M2-master/M2/BUILD/master/builds.tmp/ubuntu64.production/usr-host/lib/liblapack.a(dormbr.o):/home/jakob/Projects/M2-master/M2/BUILD/master/builds.tmp/ubuntu64.production/libraries/lapack/build/lapack-3.5.0/SRC/dormbr.f:277: more undefined references to `_gfortran_concat_string' follow
/home/jakob/Projects/M2-master/M2/BUILD/master/builds.tmp/ubuntu64.production/usr-host/lib/liblapack.a(dlasd0.o): In function `dlasd0_':
/home/jakob/Projects/M2-master/M2/BUILD/master/builds.tmp/ubuntu64.production/libraries/lapack/build/lapack-3.5.0/SRC/dlasd0.f:285: undefined reference to `_gfortran_pow_i4_i4'
/home/jakob/Projects/M2-master/M2/BUILD/master/builds.tmp/ubuntu64.production/usr-host/lib/liblapack.a(dlasda.o): In function `dlasda_':
/home/jakob/Projects/M2-master/M2/BUILD/master/builds.tmp/ubuntu64.production/libraries/lapack/build/lapack-3.5.0/SRC/dlasda.f:455: undefined reference to `_gfortran_pow_i4_i4'
/home/jakob/Projects/M2-master/M2/BUILD/master/builds.tmp/ubuntu64.production/libraries/lapack/build/lapack-3.5.0/SRC/dlasda.f:466: undefined reference to `_gfortran_pow_i4_i4'
/home/jakob/Projects/M2-master/M2/BUILD/master/builds.tmp/ubuntu64.production/usr-host/lib/liblapack.a(ilaenv.o): In function `ilaenv_':
/home/jakob/Projects/M2-master/M2/BUILD/master/builds.tmp/ubuntu64.production/libraries/lapack/build/lapack-3.5.0/SRC/ilaenv.f:272: undefined reference to `_gfortran_compare_string'
/home/jakob/Projects/M2-master/M2/BUILD/master/builds.tmp/ubuntu64.production/libraries/lapack/build/lapack-3.5.0/SRC/ilaenv.f:273: undefined reference to `_gfortran_compare_string'
/home/jakob/Projects/M2-master/M2/BUILD/master/builds.tmp/ubuntu64.production/libraries/lapack/build/lapack-3.5.0/SRC/ilaenv.f:279: undefined reference to `_gfortran_compare_string'
/home/jakob/Projects/M2-master/M2/BUILD/master/builds.tmp/ubuntu64.production/libraries/lapack/build/lapack-3.5.0/SRC/ilaenv.f:279: undefined reference to `_gfortran_compare_string'
/home/jakob/Projects/M2-master/M2/BUILD/master/builds.tmp/ubuntu64.production/libraries/lapack/build/lapack-3.5.0/SRC/ilaenv.f:279: undefined reference to `_gfortran_compare_string'
/home/jakob/Projects/M2-master/M2/BUILD/master/builds.tmp/ubuntu64.production/usr-host/lib/liblapack.a(ilaenv.o):/home/jakob/Projects/M2-master/M2/BUILD/master/builds.tmp/ubuntu64.production/libraries/lapack/build/lapack-3.5.0/SRC/ilaenv.f:279: more undefined references to `_gfortran_compare_string' follow
/home/jakob/Projects/M2-master/M2/BUILD/master/builds.tmp/ubuntu64.production/usr-host/lib/liblapack.a(xerbla.o): In function `xerbla_':
/home/jakob/Projects/M2-master/M2/BUILD/master/builds.tmp/ubuntu64.production/libraries/lapack/build/lapack-3.5.0/SRC/xerbla.f:90: undefined reference to `_gfortran_st_write'
/home/jakob/Projects/M2-master/M2/BUILD/master/builds.tmp/ubuntu64.production/libraries/lapack/build/lapack-3.5.0/SRC/xerbla.f:90: undefined reference to `_gfortran_string_len_trim'
/home/jakob/Projects/M2-master/M2/BUILD/master/builds.tmp/ubuntu64.production/libraries/lapack/build/lapack-3.5.0/SRC/xerbla.f:90: undefined reference to `_gfortran_transfer_character_write'
/home/jakob/Projects/M2-master/M2/BUILD/master/builds.tmp/ubuntu64.production/libraries/lapack/build/lapack-3.5.0/SRC/xerbla.f:90: undefined reference to `_gfortran_transfer_integer_write'
/home/jakob/Projects/M2-master/M2/BUILD/master/builds.tmp/ubuntu64.production/libraries/lapack/build/lapack-3.5.0/SRC/xerbla.f:90: undefined reference to `_gfortran_st_write_done'
/home/jakob/Projects/M2-master/M2/BUILD/master/builds.tmp/ubuntu64.production/libraries/lapack/build/lapack-3.5.0/SRC/xerbla.f:92: undefined reference to `_gfortran_stop_string'

setup Macaulay2 in a way that is possible to track down memory issues (not freeing memory)

00-charset-wrong either broken again or still not fixed.

incorrect report. The but IS fixed

probably a gb bug?

https://groups.google.com/forum/#!searchin/macaulay2/local$20ordering$20ZZ/macaulay2/-gbZJXBSkSM/GnPjEIR-X9MJ

build and sign rpm:

`
[ yes = yes ]
rm -rf files-common
umask 022 ; /usr/bin/install -c -d files-common/usr
umask 022 ; gtar --create --mode=a+rX,og-ws --exclude-from=/home/jakob/Projects/M2-master/M2/BUILD/dan/../../distributions/rpm/../tar-exclusions --file=- -C /home/jakob/Projects/M2-master/M2/BUILD/dan/builds.tmp/fedora64/usr-dist/common . | fakeroot gtar xfp - -C files-common/usr
: warning: fixed layout in use, just trying both common possibilities:
find
files-common/usr/man/man1
files-common/usr/share/man/man1
-name *.1 | xargs gzip
find: ‘files-common/usr/man/man1’: Datei oder Verzeichnis nicht gefunden
find files-common/usr -name .info | xargs gzip
sed -e 's/^Release: ./Release: 1/'
-e 's/%PKG_COM_RPM%/Macaulay2-1.7.0.1-1.common.rpm/g' <Macaulay2-common-body.spec >Macaulay2-common.spec
(cd files-common && find -type d -a -name Macaulay2 -prune -o -type f) | sed -e 's=^.="=' -e 's=$="=' >>Macaulay2-common.spec
fakeroot rpmbuild --sign --buildroot /home/jakob/Projects/M2-master/M2/BUILD/dan/builds.tmp/fedora64/distributions/rpm/files-common -bb Macaulay2-common.spec

umask 022
cd /home/jakob/rpmbuild/BUILD
/usr/bin/rm -rf /home/jakob/Projects/M2-master/M2/BUILD/dan/builds.tmp/fedora64/distributions/rpm/files-common
exit 0
"%_gpg_name" muss in der Makro-Datei gesetzt sein
Makefile:115: recipe for target '../../Macaulay2-1.7.0.1-1.common.rpm' failed
make[3]: *** [../../Macaulay2-1.7.0.1-1.common.rpm] Error 1
make[3]: *** Deleting file '../../Macaulay2-1.7.0.1-1.common.rpm'
make[3]: Leaving directory '/home/jakob/Projects/M2-master/M2/BUILD/dan/builds.tmp/fedora64/distributions/rpm'
Makefile:6: recipe for target 'all-in-rpm' failed
make[2]: *** [all-in-rpm] Error 2
make[2]: Leaving directory '/home/jakob/Projects/M2-master/M2/BUILD/dan/builds.tmp/fedora64/distributions'
GNUmakefile:144: recipe for target 'all-in-distributions' failed
make[1]: *** [all-in-distributions] Error 2
make[1]: Leaving directory '/home/jakob/Projects/M2-master/M2/BUILD/dan/builds.tmp/fedora64'
Makefile.include:19: recipe for target 'all' failed
make: *** [all] Error 2
`

/*
#ifndef _4ti2_HAVE_MPZ_INT64_CONVERSION
template <>
inline
void
VectorArrayAPI::convert(const mpz_class& v1, _4ti2_int64_t& v2)
{
  std::cerr << "UNIMPLEMENTED: Need to convert from mpz to _4ti2_int64_t" << std::endl;
  exit(1);
}
#endif
*/

similar issue in src/zsolve/VectorArrayAPI.h

failing test on fc18

i1 : S = ZZ/32003[a,b,c];

i2 : F = a^2*b^2*c+a^3+b^3+c^3

      2 2     3    3    3
o2 = a b c + a  + b  + c

o2 : S

i3 : J = ideal jacobian ideal F

                2      2    2        2   2 2     2
o3 = ideal (2a*b c + 3a , 2a b*c + 3b , a b  + 3c )

o3 : Ideal of S

i4 : time integralClosure J
stdio:4:6:(3): error: expected lists of the same length

testing primary decomposition: out of mem

R = QQ[x,y,z]
I = ideal((1/5)*y^2*z+11*z^2-1,-5*x^3+8*z^2-(8/3)*x,x^2*y-(7/6)*y^2*z-y*z^2-3/2)
pd := primaryDecomposition I; 
pEHV := primaryDecomposition ( I, Strategy=>EisenbudHunekeVasconcelos);
pSY  :=  primaryDecomposition ( I, Strategy=>ShimoyamaYokoyama);

possible solution for the hashTable issue (if it is an issue)

never allow to use a symbol as a hash table key? using #
allow hashTable dot symbil
forbid hashtable # symbol ?

but, is it really an issue?
szenarios to check where no warnings occur

-- with dot syntax

library uses symbol and user overwrites or shadows it or introduces a shadowed one
user uses a symbol and a package overwrites or shadows it or introduces a shadowed one
-a package uses a symbol and another package overwrites or shadows it or introduces a shadowed one

-- with hash syntax:

a package exports a symbol with value which is mutable (but that is a package fault)

silently failing examle (but that is probably ok)

restart
loadPackage "BlackBoxIdeals"
keys User#"private dictionary"
fragileHT = new HashTable from {
getGlobalSymbol(User#"private dictionary", "valuesAt") => 5
}
keys User#"private dictionary"
fragileHT.valuesAt -- fails silently
fragileHT#(local valuesAt)
keys User#"private dictionary"
kk = ZZ/7;
R  = ZZ[x,y]
I = ideal x;
bb = blackBoxIdeal(I)
bb.valuesAt

getGlobalSymbol(User#"private dictionary", "valuesAt")===valuesAt

rng = ZZ[x_1..x_7]

I = ideal(-2*x_1*x_2*x_7-x_1*x_3*x_7-4,2*x_1*x_2*x_3-5*x_2*x_3*x_5-x_1*x_7,4*x_1*x_3*x_7,-4*x_1*x_2-x_6^2);
gI = ideal gens gb I;
assert( ( (gens I)%gI) == 0 );
ggI = ideal gens gb gI;
assert( gens ggI == gens gI ) ; --fails: #gI = 27 #ggI = 28

what if a package modifies kernel objects like ZZ by accident?

for example ZZ#0 or ZZ#isPrime
package Evil :

ZZ#isPrime= ()->true

loadPackage "Evil"
isprime 6
-- true

jakobkroeker / test_m2 Goto Github PK

test_m2's People

Watchers

test_m2's Issues

Recommend Projects

Recommend Topics

Recommend Org