Matching _sig_on and _sig_off

2 views
Skip to first unread message

Dr. David Kirkby

unread,
Jul 1, 2010, 6:18:42 PM7/1/10
to sage-...@googlegroups.com
When Sage starts on my OpenSolaris machine it crashes, with a segfault and a
message about "...or is not properly wrapped with _sig_on, _sig_off"

Having a look in the source code, I believe _sig_on and _sig_off should be in
pairs - is that correct? If so, should there not be an equal number of _sig_on's
and _sig_off's in any one file?

A quick check shows this is not the case - see below. One file in particular
(sage/libs/pari/gen.pyx) has 332 occurances of the text _sig_on, but only 33
occurrences of _sig_off.

My checking ignores comments, but a quick check of at least a few of the files
below shows that the differences are not due to comments (but in some cases it
might be, there are some lines with a comment which contain _sig_on or _sig_off).

In the list below, the filename is followed by the number of lines with _sig_on
or _sig_off. The first is the _sig_on and the second the _sig_off.

Does this indicate a problem, or am I worrying about nothing?

Dave


drkirkby@hawk:~/serial/sage-4.5.alpha1/spkg/standard/sage-4.5.alpha1$ ggrep -cR
_sig_on * > /tmp/on
drkirkby@hawk:~/serial/sage-4.5.alpha1/spkg/standard/sage-4.5.alpha1$ ggrep -cR
_sig_off * > /tmp/off
drkirkby@hawk:~/serial/sage-4.5.alpha1/spkg/standard/sage-4.5.alpha1$ diff
/tmp/on /tmp/off
133c133
< doc/en/developer/coding_in_other.rst:3
---
> doc/en/developer/coding_in_other.rst:2
1124c1124
< doc/output/html/en/developer/_sources/coding_in_other.txt:3
---
> doc/output/html/en/developer/_sources/coding_in_other.txt:2
3607c3607
< sage/plot/plot3d/index_face_set.pyx:5
---
> sage/plot/plot3d/index_face_set.pyx:6
3610c3610
< sage/plot/plot3d/parametric_surface.pyx:1
---
> sage/plot/plot3d/parametric_surface.pyx:2
3884c3884
< sage/matrix/matrix_sparse.pyx:2
---
> sage/matrix/matrix_sparse.pyx:3
3914c3914
< sage/matrix/matrix_rational_dense.pyx:25
---
> sage/matrix/matrix_rational_dense.pyx:26
3992c3992
< sage/rings/padics/padic_ZZ_pX_FM_element.pyx:6
---
> sage/rings/padics/padic_ZZ_pX_FM_element.pyx:5
4096c4096
< sage/rings/finite_rings/element_givaro.pyx:4
---
> sage/rings/finite_rings/element_givaro.pyx:5
4120c4120
< sage/rings/complex_double.pyx:6
---
> sage/rings/complex_double.pyx:5
4662c4662
< sage/libs/mpmath/ext_main.pyx:1
---
> sage/libs/mpmath/ext_main.pyx:0
4680c4680
< sage/libs/mwrank/mwrank.pyx:21
---
> sage/libs/mwrank/mwrank.pyx:16
4687c4687
< sage/libs/pari/gen.pyx:332
---
> sage/libs/pari/gen.pyx:33
4721c4721
< sage/libs/ntl/ntl_ZZ_pEX.pyx:29
---
> sage/libs/ntl/ntl_ZZ_pEX.pyx:28
4728c4728
< sage/libs/ntl/ntl_ZZ_pE.pyx:9
---
> sage/libs/ntl/ntl_ZZ_pE.pyx:8
4738c4738
< sage/libs/ntl/ntl_ZZ.pyx:7
---
> sage/libs/ntl/ntl_ZZ.pyx:8
4744,4745c4744,4745
< sage/libs/ntl/ntl_ZZX.pyx:23
< sage/libs/ntl/ntl_ZZ_pX.pyx:31
---
> sage/libs/ntl/ntl_ZZX.pyx:7
> sage/libs/ntl/ntl_ZZ_pX.pyx:32

William Stein

unread,
Jul 1, 2010, 6:24:57 PM7/1/10
to sage-...@googlegroups.com, Gonzalo Tornaria
On Thu, Jul 1, 2010 at 3:18 PM, Dr. David Kirkby
<david....@onetel.net> wrote:
> When Sage starts on my OpenSolaris machine it crashes, with a segfault and a
> message about "...or is not properly wrapped with _sig_on, _sig_off"
>
> Having a look in the source code, I believe _sig_on and _sig_off should be
> in pairs - is that correct?

They should be *called* in pairs.

> If so, should there not be an equal number of
> _sig_on's and _sig_off's in any one file?

No. Here is an example of _sig_on and _sig_off's, which are called
in pairs, but for which there is one _sig_off and two _sig_on's.

cdef foo():
_sig_off


def f1():
_sig_on
2+2
foo()

def f2():
_sig_on
2+3
foo()

> Does this indicate a problem, or am I worrying about nothing?

I'm sure we should worry more about _sig_on / _sig_off. However, the
worries are more subtle.

William

> --
> To post to this group, send an email to sage-...@googlegroups.com
> To unsubscribe from this group, send an email to
> sage-devel+...@googlegroups.com
> For more options, visit this group at
> http://groups.google.com/group/sage-devel
> URL: http://www.sagemath.org
>

--
William Stein
Professor of Mathematics
University of Washington
http://wstein.org

Reply all
Reply to author
Forward
0 new messages