blas/netlib: run 'go generate'

[vladimir-ch]

Please take a look.

[vladimir-ch]

I have no idea what the segfaults mean and where they come from. I'll try to restart the build.

[kortschak]

Do you get the segfaults locally?

[vladimir-ch]

No.

…

On Fri, Jun 22, 2018, 00:27 Dan Kortschak ***@***.***> wrote: Do you get the segfaults locally? — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#36 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ABwJvvKo_L0bsCTFS5-yxNoP1xdNTh8Gks5t_B3OgaJpZM4UyKeO> .

[vladimir-ch]

I take it back, I do get the segfaults locally. I suspect it will be something in OpenBLAS. When the test fails, I see the following output from OpenBLAS:

OpenBLAS : Program will terminate because you tried to start too many threads.

I'll try to look into it.

[vladimir-ch]

I've submitted an issue to OpenBLAS: OpenMathLib/OpenBLAS#1641

[vladimir-ch]

Great, the segfaults have been fixed upstream but now Dgesvd fails due to numerical reasons. I will create our issue first to avoid discussing on a closed, unrelated PR.

[vladimir-ch]

And on top of that the segfaults don't seem to have been fixed.