Re: bison-master

[Eelco Dolstra]

Hi,

On 13/03/12 10:23, Akim Demaille wrote:

> The bison-next jobset seems to be stuck.  I have pushed
> several times in it, but Hydra does not seem to be willing
> to restart it.  Is there something I am missing?

There was this error in the logs:

> considering jobset gnu:bison-next
> checking out Git input from git://git.sv.gnu.org/bison.git
> error evaluating jobset bison-next: Cannot check out Git repository branch 
> 'next' at `git://git.sv.gnu.org/bison.git':
> Initialized empty Git repository in 
> /tmp/git-checkout-tmp-23313/git-export/.git/
> From 
> /home/hydra/data/scm/20895af2e84b58011f1c3614c8162c3f494cfca808686a9e007f9d00343e4063
>  * [new branch]      master     -> origin/master
> From 
> /home/hydra/data/scm/20895af2e84b58011f1c3614c8162c3f494cfca808686a9e007f9d00343e4063
>  * [new tag]         BISON-1_28b -> BISON-1_28b
>  * [new tag]         BISON-1_28c -> BISON-1_28c
>  * [new tag]         BISON-1_49a -> BISON-1_49a
>  * [new tag]         BISON-1_49b -> BISON-1_49b
>  * [new tag]         BISON-1_49c -> BISON-1_49c
>  * [new tag]         BISON-1_75 -> BISON-1_75
>  * [new tag]         BISON-1_875 -> BISON-1_875
>  * [new tag]         BISON-2_0  -> BISON-2_0
>  * [new tag]         BISON-2_1  -> BISON-2_1
>  * [new tag]         GNU_1_26   -> GNU_1_26
>  * [new tag]         GNU_1_27   -> GNU_1_27
>  * [new tag]         GNU_1_28   -> GNU_1_28
>  * [new tag]         GNU_ORIG   -> GNU_ORIG
>  * [new tag]         after-merge-branch-1_29 -> after-merge-branch-1_29
>  * [new tag]         before-m4-back-end -> before-m4-back-end
>  * [new tag]         pre-merge-of-1-29-branch -> pre-merge-of-1-29-branch
>  * [new tag]         v2.3b      -> v2.3b
>  * [new tag]         v2.4       -> v2.4
> fatal: reference is not a tree: dac70b43a0fd90d2470873fb80b261c345944dd3
> Unable to checkout dac70b43a0fd90d2470873fb80b261c345944dd3 from 
> /home/hydra/data/scm/20895af2e84b58011f1c3614c8162c3f494cfca808686a9e007f9d00343e4063.

I've deleted the git  cache directory (/home/hydra/data/scm/2089...) which fixed
the problem (likewise for the master branch).  Not sure how we got into this
state, though.

-- 
Eelco Dolstra | http://www.st.ewi.tudelft.nl/~dolstra/