bug in interred over integers? (segfault)

[kroeker@…]

Hello,

we are observing a segfault in 'interred()' over integers. Can someone reproduce this issue?

ring rng = integer,(x,y),dp;
ideal j= x, 5*y, -3*x-4*y+2;
interred(j); //segfault

also we observe a bug for

ring rng = integer,(x,y),dp;
ideal j= x, 3*y, x-2*y;
interred(j); //hangs

[gorzel]

Yes, it's true... "integer ring has limited functionality" (this message has been removed)

but certainly, there are commands which do not work over this basering.

[Oleksandr]

Hmm, i get

   ? not implemented for rings with rings as coeffients

after those interred calls... (i am on git commit id: 3b1cb6f)

[anne]

This is one of the things on Adi's and my TODO list and it is currently work in progress.

[Oleksandr]

Hmm, looks like ​https://github.com/Singular/Sources/pull/563 enables interred over arith. rings...

[Oleksandr]

Hm... Was this the only cause for "fix: disable interred for ring-cf" by Hans (4ffdb110bddefb15e3b1aad9f3a81cd8f82e5c02) ?

Or are there any deeper problems with interred?

[anne]

I am not 100% sure that this was the only issue. Hence, I scheduled a code review on these changes with adi for april 30th in KL (subject to confirmation of my business trip ...).

[Oleksandr]

Replying to anne:

I am not 100% sure that this was the only issue. Hence, I scheduled a code review on these changes with adi for april 30th in KL (subject to confirmation of my business trip ...).

What about accepting this Adi's changes for now but adding a warning about this feature being experimental ?

[anne]

Sounds fine to me.

[hannes]

interred is (currently) only usuable for field coefficients. To allow it for ring coeffs. it needs a major rewrite or (better) a separate routine. Closed for the time being.