Upgrade to Singular-3-1-7p1

jdemeyer commented 10 years ago

Singular 3-1-7p1 repackaged (see spkg-src):

http://boxen.math.washington.edu/home/jpflori/upstream/singular-3.1.7p1.tar.bz2

CC: @vbraun @nbruin @burcin

Component: packages: standard

Author: Jean-Pierre Flori

Branch/Commit: 4e287c4

Reviewer: Volker Braun

Issue created by migration from https://trac.sagemath.org/ticket/17184

jdemeyer commented 10 years ago

Branch: u/jdemeyer/ticket/17184

jdemeyer commented 10 years ago

Commit: 7c3f691

jdemeyer commented 10 years ago

comment:2

boxen is down, cannot test...

New commits:

7c3f691 Upgrade to Singular-3-1-7

jpflori commented 10 years ago

comment:3

One thing I actually did not test because i just forgot: check that it still builds with SAGE_DEBUG as it is quite involved for Singular.

jpflori commented 10 years ago

comment:4

Boxen is back up (from here at least)!

7ed8c4ca-6d56-4ae9-953a-41e42b4ed313 commented 10 years ago

Branch pushed to git repo; I updated commit sha1. New commits:

1027b8f Document templates.patch

7ed8c4ca-6d56-4ae9-953a-41e42b4ed313 commented 10 years ago

Changed commit from 7c3f691 to 1027b8f

kiwifb commented 10 years ago

comment:6

That's a lot of dropped patches. That, on its own, would justify the upgrade.

jpflori commented 10 years ago

comment:7

Needs some work for the debug version. I'm on it.

jpflori commented 10 years ago

Changed commit from 1027b8f to f70af8c

jpflori commented 10 years ago

Changed branch from u/jdemeyer/ticket/17184 to u/jpflori/ticket/17184

jpflori commented 10 years ago

comment:8

Should compile ok in debug mode now, though potentially not on top of the new NTL.

New commits:

`3968376`	`Update patches for Singular 3-1-7.`
`f70af8c`	`Add new alias for Singular debug memory allocation.`

7ed8c4ca-6d56-4ae9-953a-41e42b4ed313 commented 10 years ago

Branch pushed to git repo; I updated commit sha1. New commits:

cdedfc6 Correct fuzz for Singular patches.

7ed8c4ca-6d56-4ae9-953a-41e42b4ed313 commented 10 years ago

Changed commit from f70af8c to cdedfc6

jdemeyer commented 10 years ago

comment:11

Without the latest debug patches, without SAGE_DEBUG=yes, together with NTL-6.2.1:

sage -t --long src/sage/schemes/plane_curves/constructor.py
**********************************************************************
File "src/sage/schemes/plane_curves/constructor.py", line 69, in sage.schemes.plane_curves.constructor.Curve
Failed example:
    C.genus()
Expected:
    1
Got:
    // ** redefining my_awesome_sage_ring **
    1
**********************************************************************
1 item had failures:
   1 of  30 in sage.schemes.plane_curves.constructor.Curve
    [29 tests, 1 failure, 1.96 s]
sage -t --long src/sage/schemes/plane_curves/curve.py
**********************************************************************
File "src/sage/schemes/plane_curves/curve.py", line 116, in sage.schemes.plane_curves.curve.Curve_generic.geometric_genus
Failed example:
    C.geometric_genus()
Expected:
          1
Got:
    // ** redefining my_awesome_sage_ring **
    1
**********************************************************************
1 item had failures:
   1 of  16 in sage.schemes.plane_curves.curve.Curve_generic.geometric_genus
    [35 tests, 1 failure, 1.65 s]
sage -t --long src/sage/rings/polynomial/multi_polynomial_ideal.py
**********************************************************************
File "src/sage/rings/polynomial/multi_polynomial_ideal.py", line 1971, in sage.rings.polynomial.multi_polynomial_ideal.?.integral_closure
Failed example:
    I.integral_closure()
Expected:
    [x^2, y^5, -x*y^3]
Got:
    [x^2, x*y^4, y^5, x*y^3]
**********************************************************************
1 item had failures:
   1 of   4 in sage.rings.polynomial.multi_polynomial_ideal.?.integral_closure
    [708 tests, 1 failure, 11.48 s]
sage -t --long src/sage/schemes/plane_curves/projective_curve.py
**********************************************************************
File "src/sage/schemes/plane_curves/projective_curve.py", line 70, in sage.schemes.plane_curves.projective_curve.ProjectiveCurve_generic.arithmetic_genus
Failed example:
    C.genus()
Expected:
    4
Got:
    // ** redefining my_awesome_sage_ring **
    4
**********************************************************************

jdemeyer commented 10 years ago

comment:12

I get these doctest failures also with the latest version of this ticket, without NTL.

jpflori commented 10 years ago

comment:14

I don't really remember who committed the "my_awesome_ring" code. Was it Volker, or Nils, or anyone else? I didn't have a look at the code yet, but if the author could help, that would... help!

jpflori commented 10 years ago

comment:15

Looking at the failing example in src/sage/schemes/plane_curves/constructor.py, which is a call to genus going through Singular, it seems that there are three creations of SingularFunction objects:

twice normal__lib;
then genus. Somehow the global currRingHdl is set back to zero when the second normal__lib object is created. This was not the case with 3.1.6. (I'm just discovering the way Singular internally works, so my description may seem naive and wrong.)

vbraun commented 10 years ago

comment:16

Wasn't me, but if there were any changes to the genus function then hopefully they fix #15477

jpflori commented 10 years ago

comment:17

When the normal lib is loaded, currRingHdl is set back to zero. Here is when:

#0  rSetHdl (h=0x0) at ipshell.cc:4592
#1  0x00003fffa487ce04 in iiPStart (pn=<optimized out>, v=<optimized out>) at iplib.cc:456
#2  0x00003fffa487d334 in iiMake_proc (pn=0x3fffa426dd10, pack=0x3fffa42584b8, sl=0x0) at iplib.cc:612
#3  0x00003fffa487f65c in iiRunInit (p=<error reading variable: value has been optimized out>) at iplib.cc:889
#4  iiLoadLIB (fp=0x0, libnamebuf=0x3fffffffa280 "/home/jpflori/sage.git/local/share/singular/modstd.lib", newlib=0x3fffa4270280 "modstd.lib", 
    pl=<optimized out>, autoexport=<optimized out>, tellerror=<optimized out>) at iplib.cc:941
#5  0x00003fffa487f840 in iiLibCmd (newlib=0x3fffa4270280 "modstd.lib", autoexport=<optimized out>, tellerror=<optimized out>, force=<optimized out>)
    at iplib.cc:842
#6  0x00003fffa487f3f8 in iiLoadLIB (fp=0x0, libnamebuf=0x3fffffffa4a0 "/home/jpflori/sage.git/local/share/singular/locnormal.lib", newlib=
    0x3fffa42701a0 "locnormal.lib", pl=<optimized out>, autoexport=<optimized out>, tellerror=<optimized out>) at iplib.cc:950
#7  0x00003fffa487f840 in iiLibCmd (newlib=0x3fffa42701a0 "locnormal.lib", autoexport=<optimized out>, tellerror=<optimized out>, 
    force=<optimized out>) at iplib.cc:842
#8  0x00003fffa487f3f8 in iiLoadLIB (fp=0x0, libnamebuf=0x3fffffffa6c0 "/home/jpflori/sage.git/local/share/singular/reesclos.lib", newlib=
    0x3fffa4254b20 "reesclos.lib", pl=<optimized out>, autoexport=<optimized out>, tellerror=<optimized out>) at iplib.cc:950
#9  0x00003fffa487f840 in iiLibCmd (newlib=0x3fffa4254b20 "reesclos.lib", autoexport=<optimized out>, tellerror=<optimized out>, 
    force=<optimized out>) at iplib.cc:842
#10 0x00003fffa487f3f8 in iiLoadLIB (fp=0x0, libnamebuf=0x3fffffffa8e0 "/home/jpflori/sage.git/local/share/singular/normal.lib", newlib=
    0x3fffa4254a60 "normal.lib", pl=<optimized out>, autoexport=<optimized out>, tellerror=<optimized out>) at iplib.cc:950
#11 0x00003fffa487f840 in iiLibCmd (newlib=0x3fffa4254a60 "normal.lib", autoexport=<optimized out>, tellerror=<optimized out>, force=<optimized out>)
    at iplib.cc:842
#12 0x00003fffa4082cd8 in __pyx_pf_4sage_4libs_8singular_8function_10lib (__pyx_self=<optimized out>, __pyx_v_name='normal.lib')
    at build/cythonized/sage/libs/singular/function.cpp:16752
#13 __pyx_pw_4sage_4libs_8singular_8function_11lib (__pyx_self=<optimized out>, __pyx_v_name='normal.lib')
    at build/cythonized/sage/libs/singular/function.cpp:16623
#14 0x00003fffb7ecbfb4 in call_function (oparg=<optimized out>, pp_stack=0x3fffffffab28) at Python/ceval.c:4021
#15 PyEval_EvalFrameEx (f=
    Frame 0x3fff9e5b0b60, for file /home/jpflori/sage.git/local/lib/python2.7/site-packages/sage/libs/singular/function_factory.py, line 44, in __getattr__ (self=<type unknown at remote 0x3fffa40dee90>, name='normal'), throwflag=<optimized out>) at Python/ceval.c:2679
#16 0x00003fffb7ecd7e0 in PyEval_EvalCodeEx (co=<code at remote 0x3fffa40f9430>, globals=<optimized out>, locals=<optimized out>, 
    args=<optimized out>, argcount=<optimized out>, kws=<optimized out>, kwcount=<optimized out>, defs=0x0, defcount=0, closure=0x0)
    at Python/ceval.c:3265
#17 0x00003fffb7e1d5d4 in function_call (func=<function at remote 0x3fffa40fc2a8>, arg=(<type unknown at remote 0x3fffa40dee90>, 'normal__lib'), kw=
    0x0) at Objects/funcobject.c:526

jpflori commented 10 years ago

comment:18

And this is the new piece of code in 3-1-7 in iplib.cc around line 453:

+    if ((currRing!=NULL) &&
+      ((currRingHdl==NULL)||(IDRING(currRingHdl)!=currRing)
+       ||(IDLEV(currRingHdl)>=myynest-1)))
+    {
+      rSetHdl(rFindHdl(currRing,NULL));
+      iiLocalRing[myynest-1]=NULL;
+    }

In fact the whole block (not only the one included above) used to be around line 556 in 3-1-6.

jpflori commented 10 years ago

comment:19

In that very piece of code, Singular 4-0-1 and 3-1-7 look similar, so we might face this issue even if going for 4-0-1 directly...

jpflori commented 10 years ago

comment:20

Groumpf, I don't remmeber if anyone from the Singualr team has an account here on trac...

ea1d0bf8-c27a-4548-8cb7-de0b1d02441a commented 10 years ago

comment:21

Groumpf, I don't remmeber if anyone from the Singualr team has an account here on trac...

I do, you are welcome to ask questions if you have any.

Jakob

jpflori commented 10 years ago

comment:22

The following segfaults with Singular 3-1-7.

3-1-6:

sage: from sage.libs.singular import ff
sage: ff.modstd__lib
<sage.libs.singular.function_factory.SingularFunctionFactory object at 0x3fff9e6c5450>
sage: ff.modstd__lib
<sage.libs.singular.function_factory.SingularFunctionFactory object at 0x3fff9e6c5590>
sage: ff.modstd__lib
<sage.libs.singular.function_factory.SingularFunctionFactory object at 0x3fff9e6c5690>

3-1-7:

sage: from sage.libs.singular import ff
sage: ff.modstd__lib
<sage.libs.singular.function_factory.SingularFunctionFactory object at 0x3fff9e695450>
sage: ff.modstd__lib
// ** redefining root my_awesome_sage_ring **
Program received signal SIGSEGV, Segmentation fault.
nSetChar (r=0x3fffa42365f0) at numbers.cc:205

with backtrace:

#0  nSetChar (r=0x3fffa42365f0) at numbers.cc:205
#1  0x00003fffa49aef34 in rChangeCurrRing (r=<error reading variable: value has been optimized out>) at ring.cc:112
#2  0x00003fffa4864b54 in killlocals (v=<optimized out>) at ipshell.cc:386
#3  0x00003fffa484ce24 in iiPStart (pn=<optimized out>, v=<optimized out>) at iplib.cc:499
#4  0x00003fffa484d334 in iiMake_proc (pn=0x3fffa424fc80, pack=0x3fffa4228260, sl=0x0) at iplib.cc:612
#5  0x00003fffa484f65c in iiRunInit (p=<error reading variable: value has been optimized out>) at iplib.cc:889
#6  iiLoadLIB (fp=0x0, libnamebuf=0x3fffffffb980 "/home/jpflori/sage.git/local/share/singular/modstd.lib", newlib=0x3fffa42521c0 "modstd.lib", 
    pl=<optimized out>, autoexport=<optimized out>, tellerror=<optimized out>) at iplib.cc:941
#7  0x00003fffa484f840 in iiLibCmd (newlib=0x3fffa42521c0 "modstd.lib", autoexport=<optimized out>, tellerror=<optimized out>, force=<optimized out>)
    at iplib.cc:842
#8  0x00003fffa4052cd8 in __pyx_pf_4sage_4libs_8singular_8function_10lib (__pyx_self=<optimized out>, __pyx_v_name='modstd.lib')
    at build/cythonized/sage/libs/singular/function.cpp:16752
...

jpflori commented 10 years ago

comment:23

No issue using ./sage -singular:

> option(loadLib);
> LIB "modstd.lib";
// ** loaded /home/jpflori/sage.git/local/share/singular/modstd.lib (3-1-7-0,May_2014)
// ** loaded /home/jpflori/sage.git/local/share/singular/parallel.lib (3-1-7-0,Dec_2013)
// ** loaded /home/jpflori/sage.git/local/share/singular/tasks.lib (3-1-7-0,Dec_2013)
// ** loaded /home/jpflori/sage.git/local/share/singular/resources.lib (3-1-7-0,Dec_2013)
// ** loaded /home/jpflori/sage.git/local/share/singular/modular.lib (3-1-7-0,Dec_2013)
// ** loaded /home/jpflori/sage.git/local/share/singular/poly.lib (3-1-7-0,Sep_2013)
// ** loaded /home/jpflori/sage.git/local/share/singular/ring.lib (4.0.0.0,Jun_2013)
// ** loaded /home/jpflori/sage.git/local/share/singular/primdec.lib (4.0.0.0,Jun_2013)
// ** loaded /home/jpflori/sage.git/local/share/singular/absfact.lib (3-1-7-0,Sep_2013)
// ** loaded /home/jpflori/sage.git/local/share/singular/triang.lib (3-1-7-0,Sep_2013)
// ** loaded /home/jpflori/sage.git/local/share/singular/matrix.lib (3-1-7-0,Sep_2013)
// ** loaded /home/jpflori/sage.git/local/share/singular/nctools.lib (3-1-7-0,Sep_2013)
// ** loaded /home/jpflori/sage.git/local/share/singular/random.lib (3-1-7-0,Sep_2013)
// ** loaded /home/jpflori/sage.git/local/share/singular/elim.lib (3-1-7-0,Sep_2013)
// ** loaded /home/jpflori/sage.git/local/share/singular/inout.lib (3-1-7-0,Sep_2013)
// ** loaded /home/jpflori/sage.git/local/share/singular/general.lib (3-1-7-0,Sep_2013)
> LIB "modstd.lib";
// ** redefining modStd **
...

jpflori commented 10 years ago

comment:25

This looks related... #13447 comment:63

7ed8c4ca-6d56-4ae9-953a-41e42b4ed313 commented 10 years ago

Branch pushed to git repo; I updated commit sha1. New commits:

32f3be6 Avoid redefining Singular mock ring.

7ed8c4ca-6d56-4ae9-953a-41e42b4ed313 commented 10 years ago

Changed commit from cdedfc6 to 32f3be6

jpflori commented 10 years ago

comment:27

Fix extracted from the discussion at #13447.

7ed8c4ca-6d56-4ae9-953a-41e42b4ed313 commented 10 years ago

Changed commit from 32f3be6 to e2d2920

7ed8c4ca-6d56-4ae9-953a-41e42b4ed313 commented 10 years ago

Branch pushed to git repo; I updated commit sha1. New commits:

e2d2920 Remove debug stuff.

jpflori commented 10 years ago

comment:29

There is still one doctest to fix.

7ed8c4ca-6d56-4ae9-953a-41e42b4ed313 commented 10 years ago

Changed commit from e2d2920 to cde2930

7ed8c4ca-6d56-4ae9-953a-41e42b4ed313 commented 10 years ago

Branch pushed to git repo; I updated commit sha1. New commits:

cde2930 Update doctest for Singular new behavior.

nbruin commented 10 years ago

comment:32

Replying to @jpflori:

I don't really remember who committed the "my_awesome_ring" code.

It looks like it was there from the start: #6628 introduced this in 2009. It also looks like you've resolved your issue with it already.

jpflori commented 10 years ago

comment:33

I hope so, I just picked up a minimal change from #13447.

vbraun commented 9 years ago

comment:34

lgtm

vbraun commented 9 years ago

Reviewer: Volker Braun

vbraun commented 9 years ago

comment:35

A build with SAGE_DEBUG=yes fails in:

sage -t --long src/sage/schemes/plane_curves/curve.py
**********************************************************************
File "src/sage/schemes/plane_curves/curve.py", line 122, in sage.schemes.plane_curves.curve.Curve_generic.geometric_genus
Failed example:
    C.geometric_genus()
Expected:
          3
Got:
    <BLANKLINE>
    // ***dError: T[3] pLength error: has 1, specified to have 2 occured at: 
    <BLANKLINE>
    // ***dError: T[3] pLength error: has 1, specified to have 2 occured at: 
    3

there are more failures, always in geometric genus. If somebody can extract a singular-only testcase from that and report it on libsingular-devel then Hans is usually quick to patch that...

vbraun commented 9 years ago

comment:36

Log: http://build.sagedev.org/release/builders/%20%20fast%20Volker%20Desktop%20%28Fedora%2021%20x86_64%29%20incremental/builds/43/steps/shell_4/logs/stdio

jpflori commented 9 years ago

comment:37

Ok it seems to be upstream and not because of our broken reference handling. I'm just discovering singular syntax and might be wrong, but I get the following:

$ ./sage -singular
                     SINGULAR                                 /  Development
 A Computer Algebra System for Polynomial Computations       /   version 3-1-7
                                                           0<
 by: W. Decker, G.-M. Greuel, G. Pfister, H. Schoenemann     \   Aug 2013
FB Mathematik der Universitaet, D-67653 Kaiserslautern        \
> LIB "normal.lib";
// ** loaded /home/jpflori/sage.git/local/share/singular/normal.lib (3-1-7-0,Sep_2013)
// ** loaded /home/jpflori/sage.git/local/share/singular/algebra.lib (3-1-7-0,Sep_2013)
// ** loaded /home/jpflori/sage.git/local/share/singular/matrix.lib (3-1-7-0,Sep_2013)
// ** loaded /home/jpflori/sage.git/local/share/singular/nctools.lib (3-1-7-0,Sep_2013)
// ** loaded /home/jpflori/sage.git/local/share/singular/random.lib (3-1-7-0,Sep_2013)
// ** loaded /home/jpflori/sage.git/local/share/singular/reesclos.lib (3-1-7-1,Jan_2014)
// ** loaded /home/jpflori/sage.git/local/share/singular/locnormal.lib (3-1-7-1,Jan_2014)
// ** loaded /home/jpflori/sage.git/local/share/singular/modstd.lib (3-1-7-0,May_2014)
// ** loaded /home/jpflori/sage.git/local/share/singular/parallel.lib (3-1-7-0,Dec_2013)
// ** loaded /home/jpflori/sage.git/local/share/singular/tasks.lib (3-1-7-0,Dec_2013)
// ** loaded /home/jpflori/sage.git/local/share/singular/resources.lib (3-1-7-0,Dec_2013)
// ** loaded /home/jpflori/sage.git/local/share/singular/modular.lib (3-1-7-0,Dec_2013)
// ** loaded /home/jpflori/sage.git/local/share/singular/hnoether.lib (3-1-7-0,Sep_2013)
// ** loaded /home/jpflori/sage.git/local/share/singular/primitiv.lib (3-1-7-0,Sep_2013)
// ** loaded /home/jpflori/sage.git/local/share/singular/ring.lib (4.0.0.0,Jun_2013)
// ** loaded /home/jpflori/sage.git/local/share/singular/inout.lib (3-1-7-0,Sep_2013)
// ** loaded /home/jpflori/sage.git/local/share/singular/presolve.lib (3-1-7-0,Sep_2013)
// ** loaded /home/jpflori/sage.git/local/share/singular/elim.lib (3-1-7-0,Sep_2013)
// ** loaded /home/jpflori/sage.git/local/share/singular/primdec.lib (4.0.0.0,Jun_2013)
// ** loaded /home/jpflori/sage.git/local/share/singular/absfact.lib (3-1-7-0,Sep_2013)
// ** loaded /home/jpflori/sage.git/local/share/singular/triang.lib (3-1-7-0,Sep_2013)
// ** loaded /home/jpflori/sage.git/local/share/singular/sing.lib (3-1-7-0,Sep_2013)
// ** loaded /home/jpflori/sage.git/local/share/singular/poly.lib (3-1-7-0,Sep_2013)
// ** loaded /home/jpflori/sage.git/local/share/singular/general.lib (3-1-7-0,Sep_2013)
> ring R=5,(x,y,z),dp;
// ** Could not open dynamic library: /home/jpflori/sage.git/local/bin/p_Procs_FieldIndep.sog
// ** Error message from system: /home/jpflori/sage.git/local/bin/p_Procs_FieldIndep.sog: cannot open shared object file: No such file or directory
// ** Singular will work properly, but much slower.
// ** See the INSTALL section in the Singular manual for details.
> ideal I = x^10 + y^7*z^3 + z^10;
> genus(I);

// ***dError: T[3] pLength error: has 1, specified to have 2 occured at: 

// ***dError: T[3] pLength error: has 1, specified to have 2 occured at: 
3

Also note the Singular's warning, so it seems our way of launching singular might not be the best.

vbraun commented 9 years ago

comment:38

Looks good, can you post to libsingular-devel?

jpflori commented 9 years ago

comment:39

I'm not subscribed there and I just applied. In the meantime I posted on the forum on their homepage:

http://www.singular.uni-kl.de/forum/viewtopic.php?f=10&t=2470 (And I don't have access to the bugtracker because of the unusual port it uses.)

jpflori commented 9 years ago

comment:40

Great a fix was posted upstream and a 3-1-7p1 version released. I'm in the process of testing it.

7ed8c4ca-6d56-4ae9-953a-41e42b4ed313 commented 9 years ago

Branch pushed to git repo; I updated commit sha1. New commits:

`badc6f7`	`Merge remote-tracking branch 'trac/develop' into ticket/17184`
`ac60a76`	`Update Singular to 3-1-7p1.`

7ed8c4ca-6d56-4ae9-953a-41e42b4ed313 commented 9 years ago

Changed commit from cde2930 to ac60a76

jpflori commented 9 years ago

Description changed:

--- 
+++ 
@@ -1,2 +1,2 @@
-Singular 3-1-7 repackaged (see spkg-src):
+Singular 3-1-7p1 repackaged (see spkg-src):
 * http://boxen.math.washington.edu/home/jpflori/upstream/singular-3.1.7.tar.bz2

7ed8c4ca-6d56-4ae9-953a-41e42b4ed313 commented 9 years ago

Changed commit from ac60a76 to 9b3b933

7ed8c4ca-6d56-4ae9-953a-41e42b4ed313 commented 9 years ago

Branch pushed to git repo; I updated commit sha1. New commits:

9b3b933 Use actual upstream version number for Singular.

sagemath / sage

Upgrade to Singular-3-1-7p1 #17184