memleak in multi_modular_MultiModularBasis_base exposed by ModularSymbols(n,sign=1).decomposition()

[mabshoff]

Hello folks,

for n in range(10,100): a=ModularSymbols(n,sign=1).decomposition(); print n, get_memory_usage()

causes (among other things) the following:

==5107== 1,056 bytes in 132 blocks are definitely lost in loss record 2,208 of 2,944
==5107==    at 0x4A05809: malloc (vg_replace_malloc.c:149)
==5107==    by 0x94A2697: __gmpz_init (in /tmp/Work2/sage-2.8.3.rc3/local/lib/libgmp.so.3.4.1)
==5107==    by 0x1F4BD6EF: __pyx_tp_new_13multi_modular_MultiModularBasis_base (multi_modular.c:2324)
==5107==    by 0x1F4B92A0: __pyx_tp_new_13multi_modular_MultiModularBasis (multi_modular.c:5836)
==5107==    by 0x45A272: type_call (typeobject.c:422)
==5107==    by 0x4156A2: PyObject_Call (abstract.c:1860)
==5107==    by 0x47DB71: PyEval_CallObjectWithKeywords (ceval.c:3433)
==5107==    by 0x2041FF67: __pyx_f_20matrix_integer_dense_20Matrix_integer_dense__multiply_multi_modular (matrix_integer_den
se.c:7503)
==5107==    by 0x4156A2: PyObject_Call (abstract.c:1860)
==5107==    by 0x47DB71: PyEval_CallObjectWithKeywords (ceval.c:3433)
==5107==    by 0x204237E0: __pyx_f_20matrix_integer_dense_20Matrix_integer_dense__matrix_times_matrix_c_impl (matrix_integer
_dense.c:5487)
==5107==    by 0xDFD6878: __pyx_f_7element_6Matrix__matrix_times_matrix_c (element.c:11483)

Cheers,

Michael

[burcin]

The following code causes similar errors:

A = random_matrix(ZZ, 100, 200, density=.001)
B = random_matrix(ZZ, 200, 100, density=.001)
C = A*B
del A, B, C

The first chunk of the attached fixes these. The second chunk is useful for #561.

Unfortunately, this isn't enough to reduce all the noise MultiModularBasis? makes while running

m = ModularSymbols(501,2).decomposition(3); del m; ModularSymbols_clear_cache();

[mabshoff]

Patch will be submitted by me as a mercurial patch.

Cheers,

Michael

[mabshoff]

As it turns out the part of the patch the fixed this ticket was wrong and coincidentally I already fixed this with another patch. The part of the patch that fixes #561 is valid. I will post that patch with credit to burcin there.

Cheers,

Michael