On Mon, Feb 3, 2020 at 9:22 AM Dima Pasechnik <[email protected]> wrote:
>
> On Mon, Feb 3, 2020 at 2:01 AM Matthias Köppe <[email protected]> 
> wrote:
> >
> > On Sunday, February 2, 2020 at 5:52:32 PM UTC-5, Volker Braun wrote:
> >>
> >> As always, you can get the latest beta version from the "develop" git 
> >> branch. Alternatively, the self-contained source tarball is at 
> >> http://www.sagemath.org/download-latest.html
> >>
> >> 1d465c7e3c (tag: 9.1.beta3, trac/develop) Updated SageMath version to 
> >> 9.1.beta3 [...]
> >
> >
> > A partial build of 9.1.beta3 (packages cbc csdp fflas_ffpack gsl iml numpy 
> > r suitesparse cvxopt rpy2 ecl maxima) on a number of platforms had the 
> > following results.
> >
> > - ubuntu-trusty minimal/standard ... build error for fflas_ffpack
> > - ubuntu-trusty minimal/standard OK
> > - ubuntu-bionic/eoan/focal minimal and debian-sid minimal ... build error 
> > for R
> > - ubuntu-bionic/eoan standard OK
> > - ubuntu-focal standard and debian-sid/bullseye standard ... build error 
> > for ecl
> > - debian-jessie minimal ... build error for gfortran
> > - debian-jessie/buster standard ... build error for fflas_ffpack
> > - debian-buster minimal ... build error for R and fflas_ffpack
> > - debian-bullseye minimal ... build error for R
> > - fedora-26/31/32 minimal and centos-7 minimal ... build error for gfortran
> > - fedora-26/28/31 standard OK
> > - fedora-32 standard ... build error for openblas
> > - centos-7 standard ... build error for fflas_ffpack
> > - centos-8 minimal ... build error for pcre, yasm
> > - centos-8 standard ... build error for pcre, gsl
> > - archlinux-latest minimal/standard ... build error for R
> > - conda-forge minimal/standard ... build error in python3
> > - ubuntu-bionic-i386 standard OK
> > - homebrew minimal/standard OK
> >
> > Logs are at https://github.com/mkoeppe/sage/actions/runs/34012261
> >
> > These platforms are defined in SAGE_ROOT/tox.ini (see 
> > https://github.com/sagemath/sage/blob/develop/tox.ini; added in 
> > https://trac.sagemath.org/ticket/29053, merged in 9.1.beta3). See also 
> > https://trac.sagemath.org/ticket/29087 (GitHub actions workflow).
> >
> > A ticket fixing the gfortran failures is at 
> > https://trac.sagemath.org/ticket/29089 .
> >
> > R, fflas_ffpack, and ecl seem to require attention.
>
> fflas_ffpack problems are manifesting themselves on old gcc, which do
> not know about vx512, see
> https://trac.sagemath.org/ticket/29102
>
> symptoms:
> 2020-02-02T23:41:08.9253838Z libtool: compile:  g++ -std=gnu++11
> -DHAVE_CONFIG_H -I. -I../../.. -I../../.. -O2 -Wall -g
> -I/sage/local/include -msse -msse2 -msse3 -mssse3 -msse4.1 -msse4.2
> -mavx -mavx2 -mavx512f -mavx512vl -mavx512dq -mfma -fabi-version=6
> -I/sage/local/include -I/sage/local/include -fopenmp -c ffpack.C
> -fPIC -DPIC -o .libs/ffpack.o
> 2020-02-02T23:41:08.9254140Z g++: error: unrecognized command line
> option '-mavx512f'
> 2020-02-02T23:41:08.9254453Z g++: error: unrecognized command line
> option '-mavx512vl'
> 2020-02-02T23:41:08.9254720Z g++: error: unrecognized command line
> option '-mavx512dq'
>
>
> ecl fails on new libffi, see https://trac.sagemath.org/ticket/29128

I've fixed the ECL/libffi  problem on
https://trac.sagemath.org/ticket/21811 (needs review)

>
> the R problem is due to a Fortan problem, already discussed here at length.
> symptoms:
> 2020-02-03T00:13:19.7256378Z gfortran -fno-optimize-sibling-calls
> -fpic  -g -O2   -c dpbfa.f -o dpbfa.o
> 2020-02-03T00:13:19.7256491Z In file included from
> ../../src/include/R_ext/Applic.h:41,
> 2020-02-03T00:13:19.7256578Z                  from maxcol.c:37:
> 2020-02-03T00:13:19.7256895Z ../../src/include/R_ext/BLAS.h:103:55:
> error: expected declaration specifiers or '...' before ')' token
> 2020-02-03T00:13:19.7257012Z    const double *beta, double *y, const
> int *incy FCLEN);
> 2020-02-03T00:13:19.7257118Z
>              ^
> 2020-02-03T00:13:19.7257434Z ../../src/include/R_ext/BLAS.h:110:35:
> error: expected declaration specifiers or '...' before ')' token
> 2020-02-03T00:13:19.7257533Z    double *y, const int *incy FCLEN);
> 2020-02-03T00:13:19.7257629Z                                    ^

-- 
You received this message because you are subscribed to the Google Groups 
"sage-release" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-release/CAAWYfq1Hs1OjudvR%2Bb%2BGbQMDMwvPj9awHici-AXWkVnqvZvGbg%40mail.gmail.com.

Reply via email to