I can't give any specific advice about this, other than that your
system version of gcc seems to be an as-yet unreleased version that
might be buggier than the stable 4.7 or 4.8 series.

On Mon, Jul 22, 2013 at 11:24 PM, ASageWoodpecker <reply....@gmail.com> wrote:
> Hi again,
>
> I got this error when sage was building mathplotlib recursively as part of 
> its own build. It may be an error specific to my system of my version of gcc 
> (4.9.0 20130317 (experimental)). But if anyone has insights on this error, 
> I'd greatly appreciate it. I haven't run into internal compiler errors before 
> and don't have much experience troubleshooting them.
>
> -------
> In file included from agg24/include/agg_vertex_sequence.h:23:0,
>                  from agg24/include/agg_math_stroke.h:24,
>                  from agg24/include/agg_vcgen_contour.h:19,
>                  from agg24/src/agg_vcgen_contour.cpp:21:
> agg24/include/agg_array.h:332:29: internal compiler error: unexpected 
> expression ā€˜Sā€™ of kind template_parm_index
>              block_size  = 1 << block_shift,
>                              ^
>
> agg24/include/agg_array.h:332:29: internal compiler error: Abort trap: 6
> gcc: internal compiler error: Abort trap: 6 (program cc1plus)
> error: command 'gcc' terminated by signal 6
> Error building matplotlib package.
>
> ---
>
> Here's the snippet of C++ code that I plucked out of the file that triggered 
> the error in gcc.
>
> ---
>     template<class T, unsigned S=6> class pod_bvector
>     {                                                                         
>             public:
>         enum block_scale_e                                                    
>                 {
>             block_shift = S,                                                  
>                     block_size  = 1 << block_shift,
>             block_mask  = block_size - 1                                      
>                 };
>                                                                               
>                 typedef T value_type;
> ---
>
> Thanks,
> Anton
>
> On Tuesday, July 23, 2013, ASageWoodpecker wrote:
>> Thanks Robert. I did download the source as a fallback.
>>
>> But I also made a small change to my build process. I created my DOT_SAGE 
>> directory and removed the bottleneck to download the necessary packages from 
>> sagemath.org. I'd have imagined that with the user-level permissions, the 
>> install script would create the DOT_SAGE directory if it didn't exist. 
>> Instead a particular step exited with a non-zero status and blocked the 
>> entire install script.
>>
>> I'm now able to download packages automatically and compile the source that 
>> I checked out from the repo.
>>
>> -Anton
>>
>
> --
> You received this message because you are subscribed to the Google Groups 
> "sage-devel" group.
> To unsubscribe from this group and stop receiving emails from it, send an 
> email to sage-devel+unsubscr...@googlegroups.com.
> To post to this group, send email to sage-devel@googlegroups.com.
> Visit this group at http://groups.google.com/group/sage-devel.
> For more options, visit https://groups.google.com/groups/opt_out.
>
>

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at http://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/groups/opt_out.


Reply via email to