[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: [MoM] FAST test suite issues



On Thu, Sep 05, 2019 at 12:26:48PM +0100, Shayan Doust wrote:
> 
> > test cases: 221 |  32 passed | 189 failed
> > assertions: 385 | 196 passed | 189 failed
> 
> This is a good sign :). I've now managed to replicate this.

Yep! ;-)
 
> > what():  Unable to open directory /usr/share/fonts/truetype/ubuntu
> 
> Ahh. I installed ttf-ubuntu-font-family instead, which I guess is a
> subset of the package you installed. This is a depends of libfast-data,
> because you would need libfast-data to run the libfast-examples anyways
> (I forgot to depend libfast-examples to libfast-data so I will commit
> this shortly).

Ahhh, let me check this evening whether something might change.  BTW, if
libfast-data is just needed by libfast-examples (and nothing else) it
might make sense to simply put the data into the libfast-examples
package.
 
> > $ LC_ALL=C /usr/lib/fast/OpenIGTLinkServer
> 
> I can actually execute OpenIGTLinkServer just fine too and does not
> throw cl* issue, which makes me think this server binary does not use
> cl, or if it does, not to the length of functionality compared to the
> other binaries that are more "client side".
> 
> Apart from OpenIGTLinkServer executing just fine and not the test suite
> itself, it seems like you are getting the exact issues that I am
> getting. clGetPlatformIDs should ever throw an error only if it cannot
> find a GPU / rendering-capable device or probably experiences some
> trouble. Just to confirm, are you sure you have whatever graphical
> driver / CPU or GPU SDK installed? Note that I am on a VM that barely
> has any passthrough of any sort / limited SSE.

I remember you wrote before that you are using a VM.  I'm working right
on the hardware of my desktop.  However, I don't use this desktop for
nifty 3D things like games etc.  So it might be that despite its real
hardware it might not properly setup.
 
> Other than that, I think we are very close and just stuck with this one
> error. Was curious if my patching was faulty so I cloned upstream and
> built directly, but the error persists for me. Moreover, I will think of
> what to do next.

In any case you have picked quite a challenging task and your progress
is really good! :-)

> Kind regards & thanks for the logs,

You are welcome as always

      Andreas.

-- 
http://fam-tille.de


Reply to: