On Feb 4, 10:24 am, William Stein <wst...@gmail.com> wrote:
> On Wed, Feb 4, 2009 at 10:01 AM, Jaap Spies <j.sp...@hccnet.nl> wrote:
<SNIP>
> > On Fedora 9, 32 bits I had those known issues
> > plus 11 mysterious memory errors.
>
> If we can't fix this fedora bug before the release, we *have* to put a
> runtime check to see if Sage is
> being run on fedora, and if so print a huge warning message on
> startup. What a mess.
> I hope Michael's ideas to work around this problem in Singular work out.
So far the issue seems to be only at startup, i.e once Sage is up and
running I have not observed a libSingular related memory failure. And
the fix so far is only a hunch, I have not tested it. The allocation
failure also seems to be much, much more common on 64 bits than 32
bits since on FC 10 I see about 1 out of 5 Sage startups fail.
> William
Overall the long term fix is to use the system malloc, but that
exposes another bug when we call the wrong Rational dtor as well as
the minpoly in Sage ends up calling libSingular's minpoly when Givaro,
libSingular and Sage are involved. You can hit that bug by building
Singular with the system malloc - see recent singular.spkg and make it
use --with-malloc=system on all platforms. Getting this fixed is
fundamental in getting the 64 bit OSX port to pass 100% doctests, so
that should motivate someone to fix this.
Cheers,
Michael
--~--~---------~--~----~------------~-------~--~----~
To post to this group, send email to sage-devel@googlegroups.com
To unsubscribe from this group, send email to
sage-devel-unsubscr...@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---
