ARMHF: pollard_brent_single....FAIL : Pollard Rho failed too many times (106 times)

[SnarkBoojum]

The armhf platform is broken because of the above message.

[thofma]

@SnarkBoojum Could you sponsor @tthsqe12 for a guest account on the arm hf server?

[SnarkBoojum]

I wanted to try a git bisect, but the old version tags are gone, and it looks like a few of the commits don't even build, so it ended fast.

[wbhart]

Were you able to get access for @tthsqe12 ?

[tthsqe12]

I have no access

[SnarkBoojum]

No access requested, no access granted... Here's the procedure ; I can sponsor but you have to provide some information...

[tthsqe12]

We are trying something a bit simpler first.

[SnarkBoojum]

Can I help?

[tthsqe12]

I don't think so. We will know in a few hours if we really need access to the machine that was failing. Did you observe this consistently failing at every try, or was it more a Heisenbug?

[SnarkBoojum]

I would say it fails consistently.

[wbhart]

@tthsqe12 it passes on x86 32 bit if that's what you are wondering. It really does seem to be ARM specific, and most likely something in longlong.h. I looked through the code for anything else it could be and came up blank.

[tthsqe12]

We will most likely be able to fix this ourselves by tomorrow or the day after. By the way, the other two failures on mips and something else should be fixed.

[wbhart]

Should be fixed in the flint-2.6 branch now. We'll issue 2.6.1 as soon as some of the other recent bug tickets are closed.

[SnarkBoojum]

I'll wait for 2.6.1 ; thanks!