Date: Sat, 2 Jul 2016 17:21:08 +0000 From: Olivier Duchateau <olivierd@freebsd.org> To: Thomas Zander <riggs@freebsd.org> Cc: "ports-committers@FreeBSD.org" <ports-committers@freebsd.org>, svn-ports-all@freebsd.org, svn-ports-head@freebsd.org Subject: Re: svn commit: r417692 - in head/math/py-numpy: . files Message-ID: <20160702172108.dd0ed3c5cf80da6ea1a4325a@freebsd.org> In-Reply-To: <CAFU734xePTNQVfKzbDmDKYSxDvGt07Jk3FjeCBmYetEjcR%2BKMA@mail.gmail.com> References: <201606280334.u5S3YtD1058608@repo.freebsd.org> <CAFU734xePTNQVfKzbDmDKYSxDvGt07Jk3FjeCBmYetEjcR%2BKMA@mail.gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Thu, 30 Jun 2016 20:02:44 +0200 Thomas Zander <riggs@freebsd.org> wrote: > On 28 June 2016 at 05:34, Olivier Duchateau <olivierd@freebsd.org> wrote: > > > -NETLIB_LIB_DEPENDS= libcblas.so:math/cblas > > +NETLIB_LIB_DEPENDS= libcblas.so:math/cblas \ > > + libopenblas.so:math/openblas > > This is weird. Is this a bug in numpy's build system or is there > really some function in openblas that is not implemented by cblas? We > should not force the install of a 200 MiB math package just because of > a bug in a Makefile. No, it's not bug in Numpy, rather in QA script. Numpy detects perfectly Openblas during build process, and Openblas comes from math/suitesparse (which is set by default). > > Riggs -- Olivier Duchateau <olivierd@freebsd.org>
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20160702172108.dd0ed3c5cf80da6ea1a4325a>