Quantcast

running tests

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
2 messages Options
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

running tests

Rajeev Singh
Hi,

I was wondering if it is possible to test julia on a machine with
moderately sized ram (say 4gb). The usual "make test" launches all the
threads and takes up full memory freezing the machine. I tried the
following too -

julia> Base.runtests(["all"], 1)

     * linalg/triangular     maxrss  639.86 MB
     * linalg/qr             maxrss  731.17 MB
     * linalg/dense          maxrss  766.48 MB
     * linalg/matmul         maxrss  819.12 MB
...

     * abstractarray         maxrss 3048.26 MB
     * intfuncs              maxrss 3048.26 MB
     * simdloop              maxrss 3048.26 MB
     * vecelement            maxrss 3048.26 MB
     * blas                 ^C

I had to stop here as my ram was almost full. I could not find this
issue in the mailing list or github issues. (Perhaps I did not look
thoroughly.) I think this should be considered a bug. Also there
should be a way to run the tests with a single thread using make.

With best regards,
Rajeev
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: running tests

Mauro
Note, discussion moved to https://discourse.julialang.org/, better
re-post there (with a link to this).

On Fri, 2016-12-02 at 11:38, Rajeev Singh <[hidden email]> wrote:

> Hi,
>
> I was wondering if it is possible to test julia on a machine with
> moderately sized ram (say 4gb). The usual "make test" launches all the
> threads and takes up full memory freezing the machine. I tried the
> following too -
>
> julia> Base.runtests(["all"], 1)
>
>      * linalg/triangular     maxrss  639.86 MB
>      * linalg/qr             maxrss  731.17 MB
>      * linalg/dense          maxrss  766.48 MB
>      * linalg/matmul         maxrss  819.12 MB
> ...
>
>      * abstractarray         maxrss 3048.26 MB
>      * intfuncs              maxrss 3048.26 MB
>      * simdloop              maxrss 3048.26 MB
>      * vecelement            maxrss 3048.26 MB
>      * blas                 ^C
>
> I had to stop here as my ram was almost full. I could not find this
> issue in the mailing list or github issues. (Perhaps I did not look
> thoroughly.) I think this should be considered a bug. Also there
> should be a way to run the tests with a single thread using make.
>
> With best regards,
> Rajeev
Loading...