Opened 5 years ago

Last modified 5 years ago

#20421 new defect

libgap workspace() doctest failures

Reported by: jdemeyer Owned by:
Priority: major Milestone: sage-7.2
Component: interfaces Keywords: random_fail
Cc: vbraun Merged in:
Authors: Reviewers:
Report Upstream: N/A Work issues:
Branch: Commit:
Dependencies: Stopgaps:

Status badges

Description

I regularly get this. I'm not sure whether it is a random failure or not, it might be because of doing make distclean first: 

sage -t src/sage/libs/gap/assigned_names.py
**********************************************************************
File "src/sage/libs/gap/assigned_names.py", line 59, in sage.libs.gap.assigned_names.load_or_compute
Failed example:
    workspace(name='globals')
Expected:
    ('...', True)
Got:
    ('/home/jdemeyer/.sage/gap/libgap-globals-4034600654683281042', False)
**********************************************************************

Change History (11)

comment:1 Changed 5 years ago by jdemeyer

Happened again. The scenario was:

  1. make distclean
  2. build Sage but don't run it
  3. make ptestlong -> got this failure

comment:2 Changed 5 years ago by fbissey

For what it's worth I had some doctest failures when the gap upgrade landed ​https://github.com/cschwan/sage-on-gentoo/issues/412. They did go away as mysteriously as they came after a couple of beta releases. They were persistent across rebuilding gap/libgap and sage.

comment:3 Changed 5 years ago by vbraun

Thats probably a race with building the workspace caches...

comment:5 Changed 5 years ago by embray

I'm getting this pretty regularly in the Docker container. This is probably because the container always starts with a fresh environment, and doesn't have timings for the tests yet, so runs them in a different order. At what point are the "workspace caches" built, and is there some way we can make sure that is done as a prerequisite to this test?

Last edited 5 years ago by embray (previous) (diff)

comment:6 follow-up: Changed 5 years ago by nthiery

Just for info: I just got this while upgrading to (lib)gap 4.9 (dev)

comment:7 in reply to: ↑ 6 Changed 5 years ago by dimpase

Replying to nthiery:

Just for info: I just got this while upgrading to (lib)gap 4.9 (dev)

with #22570, or without?

comment:8 follow-up: Changed 5 years ago by nthiery

That was with 7.6.beta6 (and some of #22626). So without #22570 I believe.

Last edited 5 years ago by nthiery (previous) (diff)

comment:9 in reply to: ↑ 8 Changed 5 years ago by fbissey

Replying to nthiery:

That was with 7.6.beta6 (and some of #22626). So without #22570 I believe.

If you pull Volker's current development tree at ​https://github.com/vbraun/sage you'll get #22570. I personnaly follow Volker's dev branch so I can anticipate or discover what's going to land for sage-on-gentoo. If you get an angry message from me on a positively reviewed ticket that's because I discovered you made my life harder from pulling on Volker's tree :P

comment:10 follow-up: Changed 5 years ago by vbraun

Don't do that. You can look at the development branch but don't touch it.... commits may be rewritting / removed until it is published.

comment:11 in reply to: ↑ 10 Changed 5 years ago by fbissey

Replying to vbraun:

Don't do that. You can look at the development branch but don't touch it.... commits may be rewritting / removed until it is published.

Not the right place to discuss it. But yes I know, and I have been burned a couple of times. But it is very handy for what I do. However it happens on a special branch of the sage-on-gentoo overlay which normal user are not supposed to touch and very few know about.

Note: See TracTickets for help on using tickets.