Flint-2.8.3 released (serious bug fixed)
20 views
Skip to first unread message
Bill Hart
Nov 3, 2021, 9:18:33 AM11/3/21
to flint-devel, sage-...@googlegroups.com, Brian Gladman, osca...@mathematik.uni-kl.de, Mike Stillman, Grayson, Daniel R.
Hello all,
I have tagged flint-2.8.3 which fixes a very serious bug in:
* fmpq_poly_xgcd
* fmpz_poly_xgcd
* nmod_poly_xgcd
which has been there for many years and which occurs for polynomials
of length at least 340.
The bug would result in incorrect results being returned. It was
caused by an overrun in a step that was not needed in a certain case
in the final iteration of a loop.
The website will be updated shortly, however we currently recommend
obtaining tarballs from GitHub [1], not our website, due to an
apparent bug in the web server which causes it to serve the wrong
tarball (with the correct name).
Bill.
[1] https://github.com/wbhart/flint2/tags
I have tagged flint-2.8.3 which fixes a very serious bug in:
* fmpq_poly_xgcd
* fmpz_poly_xgcd
* nmod_poly_xgcd
which has been there for many years and which occurs for polynomials
of length at least 340.
The bug would result in incorrect results being returned. It was
caused by an overrun in a step that was not needed in a certain case
in the final iteration of a loop.
The website will be updated shortly, however we currently recommend
obtaining tarballs from GitHub [1], not our website, due to an
apparent bug in the web server which causes it to serve the wrong
tarball (with the correct name).
Bill.
[1] https://github.com/wbhart/flint2/tags
Reply all
Reply to author
Forward
0 new messages