In fact I don't understand alarm(...) at all. Maybe it works better on
linux or older OS X releases, but on the OS X machines to which I have
access, it feels broken. If I do:
sage: alarm(0.3)
then after 0.3 seconds I see a message
KeyboardInterrupt escaped interact()
That's fine. But then
Perhaps related, and I also see this with 10.3.beta1: after using
alarm(...), Sage does not exit gracefully.
% ./sage
┌┐
│ SageMath version 10.3.beta2, Release Date: 2023-12-13 │
│ Using Python 3.11.6. Type
On Thursday, December 14, 2023 at 4:05:31 PM UTC John H Palmieri wrote:
On OS X (both Intel and Apple Silicon) I'm getting a lot of new failures,
many of the form "Killed due to alarm":
sage -t --long --random-seed=144845266083009760424007645487960076680
On Debian testing running on, core i7 + 16 GB RAM, upgrading 10.3.beta1 to
10.3.beta2 and running ptestlong passes without any failure.
HTH,
Le jeudi 14 décembre 2023 à 01:42:36 UTC+1, Volker Braun a écrit :
> As always, you can get the latest beta version from the "develop" git
> branch.
On OS X (both Intel and Apple Silicon) I'm getting a lot of new failures,
many of the form "Killed due to alarm":
sage -t --long --random-seed=144845266083009760424007645487960076680
src/sage/libs/flint/nmod_poly_linkage.pxi # Killed due to alarm
sage -t --long