Some spurious printing in IsIsomorphic for codes has been fixed. Reported by A. Munemasa.

A hang in polynomial root finding over the complex field in a case where multiple complex roots had very small imaginary part (arising from computations in hyperelliptic curves) has been fixed. Reported by J. Voight.

A failure in EightDescent for a certain class of curves has been fixed. Reported by R. Rathbun.

A crash in TwoDescent for elliptic curves has been fixed. Reported by R. Rathbun.

The SelmerGroup algorithm has been improved so that it runs much faster on certain types of input. Example problem reported by J. Rouse.

For a finite field extension with the Zech representation but a non-primitive defining polynomial, the internal generator of the Zech-based field was not chosen uniquely; this been changed so sorting involving elements of such a field now always behaves the same, irrespective of the random seed. Issue reported by C. Rito.

p-adic rings constructed as a Completion of the integers were unintentionally returned as fields. Reported by E. Assaf.

A bad slowdown in resultant computation for some types of multivariate polynomials has been fixed. Reported by N. Katz.

Some crashes in the save/restore feature running under Mac OS X have been fixed. Reported by T. Uchiyama.

The AVX2/CUDA version is now built for CUDA version 11.x so please upgrade your system(s) to CUDA 11 if you have been using the CUDA version 10.x Magma executable.