Summary: | sci-libs/blas-atlas-3.7.11 infinite loop during compilation | ||
---|---|---|---|
Product: | Gentoo Linux | Reporter: | Evil Compile Person <bugs> |
Component: | New packages | Assignee: | Gentoo Science Related Packages <sci> |
Status: | RESOLVED WONTFIX | ||
Severity: | normal | CC: | gentoo |
Priority: | High | ||
Version: | unspecified | ||
Hardware: | All | ||
OS: | Linux | ||
Whiteboard: | |||
Package list: | Runtime testing required: | --- |
Description
Evil Compile Person
2006-10-12 11:41:58 UTC
I see the same thing when I try to emerge blas-atlas-3.7.17 as recommended in bug 136003. I was wondering why after more than one hour it was still compiling but then I did $ grep smmcase0 sci-libs\:blas-atlas-3.7.17\:20061027-173247.log make[6]: *** [smmcase0] Error 1 make[6]: *** [smmcase0] Error 1 make[6]: *** [smmcase0] Error 1 make[6]: *** [smmcase0] Error 1 make[6]: *** [smmcase0] Error 1 make[6]: *** [smmcase0] Error 1 make[6]: *** [smmcase0] Error 139 make[6]: *** [smmcase0] Error 1 make[6]: *** [smmcase0] Error 1 make[6]: *** [smmcase0] Error 1 make[6]: *** [smmcase0] Error 1 make[6]: *** [smmcase0] Error 1 make[6]: *** [smmcase0] Error 1 make[6]: *** [smmcase0] Error 1 make[6]: *** [smmcase0] Error 139 make[6]: *** [smmcase0] Error 1 make[6]: *** [smmcase0] Error 1 Hmm, around that message I see smm.c:80:5: #error "This kernel requires x86-64 assembly!" make[6]: *** [smmcase0] Error 1 make[5]: *** [mmucase] Error 2 At other places in the log I see smm.c:88:5: #error KB must be multiple of 16! smm.c: In function `ATL_sJIK120x120x104TN104x104x0_a1_b1': smm.c:209: error: `vector' undeclared (first use in this function) smm.c:209: error: (Each undeclared identifier is reported only once smm.c:209: error: for each function it appears in.) smm.c:209: error: syntax error before "float" smm.c:212: error: syntax error before "float" smm.c:219: error: syntax error before "unsigned" smm.c:240: error: syntax error before "float" and lots of other error message, but not the same #error message as above. I don't know exactly where it cycles but I guess I could send the full log if that would help solving the problem. (I just need to build it because scipy obviously depends on it, if there would be a way to get scipy without blas-atlas I would be happy, too.) Hi Peter, Thanks for you report! This doesn't solve the blas-atlas problem, but scipy depends on virtual/blas/lapack and you should therefore be able to use the blas/lapack-reference ebuilds instead of the blas/lapack-atlas ebuilds. Thanks, Markus Markus, thanks for the hint emerge blas-reference lapack-reference scipy works beautifully. (Now I have to see how to actually use it...) Folks, is this still an issue with the latest version in portage (3.7.30)? Thanks, Markus Folks, I am closing this since bug since we now have blas-atlas-3.8.0 stable and this shouldn't be an issue any more. Thanks, Markus |