Sage 4.1.2.alpha2 doctest failures on Debian

4 views
Skip to first unread message

Minh Nguyen

unread,
Sep 22, 2009, 7:32:24 PM9/22/09
to sage-...@googlegroups.com
Hi folks,

Sage 4.1.2.alpha2 compiles fine on x86 Debian 5. The following doctests failed:

{{{
sage -t -long "devel/sage/sage/crypto/boolean_function.pyx"
**********************************************************************
File "/space/wstein/farm/sage-4.1.2.alpha2/devel/sage/sage/crypto/boolean_function.pyx",
line 1013:
sage: B.nonlinearity()
Expected:
222
Got:
217
**********************************************************************
1 items had failures:
1 of 6 in __main__.example_36
***Test Failed*** 1 failures.
For whitespace errors, see the file
/space/wstein/farm/sage-4.1.2.alpha2/tmp/.doctest_boolean_function.py
[4.1 s]

<SNIP>

sage -t -long "devel/sage/sage/rings/polynomial/pbori.pyx"
**********************************************************************
File "/space/wstein/farm/sage-4.1.2.alpha2/devel/sage/sage/rings/polynomial/pbori.pyx",
line 3940:
sage: x.stable_hash()
Expected:
173100285919
Got:
-845955105
**********************************************************************
File "/space/wstein/farm/sage-4.1.2.alpha2/devel/sage/sage/rings/polynomial/pbori.pyx",
line 4849:
sage: s.stable_hash()
Expected:
173100285919
Got:
-845955105
**********************************************************************
File "/space/wstein/farm/sage-4.1.2.alpha2/devel/sage/sage/rings/polynomial/pbori.pyx",
line 1976:
sage: m.stable_hash()
Expected:
173100285919
Got:
-845955105
**********************************************************************
3 items had failures:
1 of 4 in __main__.example_128
1 of 5 in __main__.example_165
1 of 5 in __main__.example_48
***Test Failed*** 3 failures.
For whitespace errors, see the file
/space/wstein/farm/sage-4.1.2.alpha2/tmp/.doctest_pbori.py
[5.3 s]

<SNIP>

sage -t -long "devel/sage/sage/interfaces/expect.py"
**********************************************************************
File "/space/wstein/farm/sage-4.1.2.alpha2/devel/sage/sage/interfaces/expect.py",
line 1599:
sage: gp(10.^80)._sage_repr()
Expected nothing
Got:
'1.000000000000000000000000000e80'
**********************************************************************
1 items had failures:
1 of 3 in __main__.example_45
***Test Failed*** 1 failures.
For whitespace errors, see the file
/space/wstein/farm/sage-4.1.2.alpha2/tmp/.doctest_expect.py
[17.4 s]
sage -t -long "devel/sage/sage/interfaces/all.py"
[0.1 s]
sage -t -long "devel/sage/sage/interfaces/rubik.py"
[37.8 s]
sage -t -long "devel/sage/sage/interfaces/gp.py"
**********************************************************************
File "/space/wstein/farm/sage-4.1.2.alpha2/devel/sage/sage/interfaces/gp.py",
line 567:
sage: repr(gp(10.^80)).replace(gp._exponent_symbol(), 'e')
Expected nothing
Got:
'1.000000000000000000000000000e80'
**********************************************************************
1 items had failures:
1 of 4 in __main__.example_26
***Test Failed*** 1 failures.
For whitespace errors, see the file
/space/wstein/farm/sage-4.1.2.alpha2/tmp/.doctest_gp.py
[3.5 s]

----------------------------------------------------------------------
The following tests failed:


sage -t -long "devel/sage/sage/crypto/boolean_function.pyx"
sage -t -long "devel/sage/sage/rings/polynomial/pbori.pyx"
sage -t -long "devel/sage/sage/interfaces/expect.py"
sage -t -long "devel/sage/sage/interfaces/gp.py"
Total time for all tests: 6990.6 seconds

}}}

--
Regards
Minh Van Nguyen

Minh Nguyen

unread,
Sep 22, 2009, 7:35:34 PM9/22/09
to sage-...@googlegroups.com
On Wed, Sep 23, 2009 at 9:32 AM, Minh Nguyen <nguye...@gmail.com> wrote:
> Hi folks,
>
> Sage 4.1.2.alpha2 compiles fine on x86 Debian 5. The following doctests failed:

The build machine is debian5-32, a VM guest on boxen.math.

Sage 4.1.2.alpha2 also builds on debian5-64, an x86_64 Debian 5 VM
guest on boxen.math. All doctests pass.

Felix Lawrence

unread,
Sep 22, 2009, 10:10:46 PM9/22/09
to sage-devel
Hi Minh,

The doctest failures in
sage -t -long "devel/sage/sage/interfaces/expect.py"
sage -t -long "devel/sage/sage/interfaces/gp.py"
are my fault - I forgot to include a 32-bit answer as well as the 64-
bit answer, e.g.
sage: gp(10.^80)._sage_repr()
'1.0000000000000000000000000000000000000e80' # 64-bit
needs to be changed to
sage: gp(10.^80)._sage_repr()
'1.0000000000000000000000000000000000000e80' # 64-bit
'1.000000000000000000000000000e80' # 32-bit
I've created a patch to apply these two trivial changes, but was
unsure what the etiquette is for fixing these doctest failures - I
imagine creating a new ticket would be overkill. Should I add the
patch to the original (closed) ticket?

Cheers,
Felix


On Sep 23, 9:35 am, Minh Nguyen <nguyenmi...@gmail.com> wrote:

John H Palmieri

unread,
Sep 22, 2009, 10:52:35 PM9/22/09
to sage-devel
On Sep 22, 7:10 pm, Felix Lawrence <fe...@physics.usyd.edu.au> wrote:

> I've created a patch to apply these two trivial changes, but was
> unsure what the etiquette is for fixing these doctest failures - I
> imagine creating a new ticket would be overkill. Should I add the
> patch to the original (closed) ticket?

If a ticket is closed, you almost certainly shouldn't reopen it. Open
a new ticket and put pointers from the old ticket to this one, and
vice versa. (We don't tend to worry about overkill: a ticket from
earlier today changed a single word in the documentation.)

Minh Nguyen

unread,
Sep 22, 2009, 10:58:32 PM9/22/09
to sage-...@googlegroups.com
Hi Felix,

On Wed, Sep 23, 2009 at 12:10 PM, Felix Lawrence
<fe...@physics.usyd.edu.au> wrote:
>
> Hi Minh,
>
> The doctest failures in
> sage -t -long "devel/sage/sage/interfaces/expect.py"
> sage -t -long "devel/sage/sage/interfaces/gp.py"
> are my fault - I forgot to include a 32-bit answer as well as the 64-
> bit answer

Oh, I see. That explains why I consistently get those doctest failures
on 32-bit platforms. Thank you for making this clear.


> I've created a patch to apply these two trivial changes, but was
> unsure what the etiquette is for fixing these doctest failures - I
> imagine creating a new ticket would be overkill. Should I add the
> patch to the original (closed) ticket?

I would agree with what John said: open another ticket to address the issue.

Felix Lawrence

unread,
Sep 23, 2009, 12:22:58 AM9/23/09
to sage-devel
Done: http://trac.sagemath.org/sage_trac/ticket/6999
The patch is attached to that ticket.

While I've got your attention, I believe that the patches for #4948
also fixed the doctest failure that was #6331 : http://trac.sagemath.org/sage_trac/ticket/6331
.

Cheers,
Felix
Reply all
Reply to author
Forward
0 new messages