NoConvergence error for gammainc for certain arguments

[pontzm]

There is a NoConvergence error when computing mp.gammainc(-3200,3200). mp.gammainc(-3200,800) and mp.gammainc(-3200,3900) work and produce output similar to that of Wolfram alpha. It seems to occur for gammainc(-a,a+-x) with the smallest 'a' detected being 950.

Increasing dps doesnt help and increasing other parameters like maxterm or maxprec is not implemented for gammainc. Also the transfomation with mp.hyperu and increasing other parameters does yield an NoConvergence error.