]> www.ginac.de Git - cln.git/commit
git://www.ginac.de / cln.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 8e19c1a) | patch
Fix double_approx(const cl_RA& x) on 64-Bit platforms.
authorRichard Kreckel <kreckel@ginac.de>
Fri, 11 Sep 2009 20:54:27 +0000 (22:54 +0200)
committerRichard Kreckel <kreckel@ginac.de>
Fri, 11 Sep 2009 20:54:27 +0000 (22:54 +0200)
commitbc3e15e46f6dee959bbeb68dd266d4c69543e1f8
treea918d198bb1a694dcd4d3970f2372c8046d47522tree | snapshot
parent8e19c1a9d9ea4116112490d6f43904258b928649commit | diff
Fix double_approx(const cl_RA& x) on 64-Bit platforms.

When we extended immediate types on 64-bit platforms back for CLN
1.2.0, we broke the assumption that integers 2^53..2^55-1 are
bignums. This lead to segfaults when approximating a non-integer cl_RA
to double, as reported by Igor Khavkine <igor.kh@gmail.com>.
src/float/dfloat/conv/cl_RA_to_double.cc diff | blob | history
CLN - Class Library for Numbers