Re: [sage-devel] Re: ecl build failure when building sage

2020-04-05 Thread Dima Pasechnik
Perhaps you can tell us what CYGWIN_NT-10.0 is, and how is it different from the usual 64-bit Cygwin? The most recent version 3.1.4, cf https://sourceware.org/legacy-ml/cygwin-announce/2020-02/msg6.html On Sun, Apr 5, 2020 at 4:01 PM Jakub Jabłoński wrote: > > Didn't help, started installing

Re: [sage-devel] Re: ecl build failure when building sage

2020-04-05 Thread Jakub Jabłoński
Didn't help, started installing ubuntu on virtual machine, will try here. Thanks for helping. On Saturday, April 4, 2020 at 5:04:57 AM UTC+2, Dima Pasechnik wrote: > > On Fri, Apr 3, 2020 at 11:04 PM Jakub Jabłoński > wrote: > > > > Sorry, didn't notice it's not rebuilding. Done it now and

Re: [sage-devel] Re: ecl build failure when building sage

2020-04-03 Thread Dima Pasechnik
On Fri, Apr 3, 2020 at 11:04 PM Jakub Jabłoński wrote: > > Sorry, didn't notice it's not rebuilding. Done it now and issue with omp > still occurs. hmm. Somewhere something instists on using openmp. Perhaps, do make distclean make > > On Friday, April 3, 2020 at 4:13:39 PM UTC+2, Dima

Re: [sage-devel] Re: ecl build failure when building sage

2020-04-03 Thread Jakub Jabłoński
Sorry, didn't notice it's not rebuilding. Done it now and issue with omp still occurs. On Friday, April 3, 2020 at 4:13:39 PM UTC+2, Dima Pasechnik wrote: > > have you rebuilt fflas-ffpack after your addition of flags? > > On Fri, Apr 3, 2020 at 10:00 PM Jakub Jabłoński > wrote: > > > >

Re: [sage-devel] Re: ecl build failure when building sage

2020-04-03 Thread Dima Pasechnik
have you rebuilt fflas-ffpack after your addition of flags? On Fri, Apr 3, 2020 at 10:00 PM Jakub Jabłoński wrote: > > Current version: > > # If SAGE_FAT_BINARY is set, disable all processor-specific optimizations > if [ "$SAGE_FAT_BINARY" = yes ]; then >

Re: [sage-devel] Re: ecl build failure when building sage

2020-04-03 Thread Dima Pasechnik
it seems that your addition is conditional on SAGE_FAT_BINARY being set to yes. It should not be so. On Fri, 3 Apr 2020, 19:56 Jakub Jabłoński, wrote: > I have added --disable-openmp here in spkg-install.in: > if [ "$SAGE_FAT_BINARY" = yes ]; then > FFLAS_FFPACK_CONFIGURE="--disable-openmp

Re: [sage-devel] Re: ecl build failure when building sage

2020-04-03 Thread Jakub Jabłoński
I have added --disable-openmp here in spkg-install.in: if [ "$SAGE_FAT_BINARY" = yes ]; then FFLAS_FFPACK_CONFIGURE="--disable-openmp --disable-sse --disable-sse2 --disable-sse3 --disable-ssse3 --disable-sse41 --disable-sse42 --disable-fma --disable-fma4 --disable-avx --disable-avx2

Re: [sage-devel] Re: ecl build failure when building sage

2020-04-03 Thread Dima Pasechnik
could you post the top-level config.log ? On Fri, 3 Apr 2020 at 16:59, Dima Pasechnik wrote: > you are building fflas-ffpack with OpenMP on, for some reason. > it should be possible to modify its spkg-install.in file so that it does > not happen, rebuild it, and sagelib itself. > > > On Fri, 3

Re: [sage-devel] Re: ecl build failure when building sage

2020-04-03 Thread Dima Pasechnik
you are building fflas-ffpack with OpenMP on, for some reason. it should be possible to modify its spkg-install.in file so that it does not happen, rebuild it, and sagelib itself. On Fri, 3 Apr 2020 at 16:40, Jakub Jabłoński wrote: > Okay, uploaded only log which seems to be relevant (it's

Re: [sage-devel] Re: ecl build failure when building sage

2020-04-03 Thread Dima Pasechnik
from this part of the log it is not possible to see much, except that Sage's Python library has failed to build. Please paste as many lines above these from the output as you can. On Fri, 3 Apr 2020, 15:09 Jakub Jabłoński, wrote: > I have removed whole cygwin and reinstalled everything, now

Re: [sage-devel] Re: ecl build failure when building sage

2020-04-01 Thread Dima Pasechnik
On Wed, Apr 1, 2020 at 11:53 PM Jakub Jabłoński wrote: > > And gcc -V is not working on my cygwin, "-V" is unrecognized command line > option. What is it supposed to do? this is normal - but it should not crash, it should exit gracefully. The configure script it trying to figure out how to talk