failure to configure atlas

[John Cremona]

I tried to build sage-5.3 on the new desktop my department has just
supplied me with (on which I do not have root access), and atlas
failed to configure.

uname -a
Linux host-56-160 3.0.38-0.5-default #1 SMP Fri Aug 3 09:02:17 UTC
2012 (358029e) x86_64 x86_64 x86_64 GNU/Linux

gcc -v
Using built-in specs.
Target: x86_64-suse-linux
Configured with: ../configure --prefix=/usr --infodir=/usr/share/info
--mandir=/usr/share/man --libdir=/usr/lib64 --libexecdir=/usr/lib64
--enable-languages=c,c++,objc,fortran,obj-c++,java,ada
--enable-checking=release --with-gxx-include-dir=/usr/include/c++/4.3
--enable-ssp --disable-libssp --with-bugurl=http://bugs.opensuse.org/
--with-pkgversion='SUSE Linux' --disable-libgcj --disable-libmudflap
--with-slibdir=/lib64 --with-system-zlib --enable-__cxa_atexit
--enable-libstdcxx-allocator=new --disable-libstdcxx-pch
--enable-version-specific-runtime-libs --program-suffix=-4.3
--enable-linux-futex --without-system-libunwind --with-cpu=generic
--build=x86_64-suse-linux
Thread model: posix
gcc version 4.3.4 [gcc-4_3-branch revision 152973] (SUSE Linux)

First attempt: automatic tuning.
Running configure with arch = None and isa extensions None
Running ../src/configure --prefix=/local/masgaj/sage-5.3/local
--with-netlib-lapack=/local/masgaj/sage-5.3/local/lib/liblapack.a
--cc="gcc" -Si cputhrchk 0 -Fa alg -fPIC -C if sage_fortran -C xc
"gcc" -C ic "gcc" -C dm "gcc" -C sm "gcc" -C dk "gcc" -C sk "gcc" -b
64 -O 1
Traceback (most recent call last):
File "./spkg-install", line 442, in <module>
rc = build()
File "./spkg-install", line 431, in build
assert rc==0, 'Configure failed!'
AssertionError: Configure failed!


John

[Volker Braun]

I think its a known issue that the old ATLAS won't work with the newest CPUs. Can you try the new ATLAS spkg:

http://trac.sagemath.org/10508

[John Cremona]

On 4 October 2012 12:01, Volker Braun <vbrau...@gmail.com> wrote:
> I think its a known issue that the old ATLAS won't work with the newest
> CPUs. Can you try the new ATLAS spkg:
>
> http://trac.sagemath.org/10508

OK, I will do.

For the record:

vendor_id : GenuineIntel
cpu family : 6
model : 42
model name : Intel(R) Core(TM) i3-2120 CPU @ 3.30GHz
stepping : 7
cpu MHz : 1600.000
cache size : 3072 KB
physical id : 0
siblings : 4
core id : 1
cpu cores : 2
apicid : 3
initial apicid : 3
fpu : yes
fpu_exception : yes
cpuid level : 13
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov
pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx
rdtscp lm constant_tsc arch_perfmon pebs bts rep_good nopl xtopology
nonstop_tsc aperfmperf pni pclmulqdq dtes64 monitor ds_cpl vmx est tm2
ssse3 cx16 xtpr pdcm sse4_1 sse4_2 popcnt xsave avx lahf_lm arat epb
xsaveopt pln pts dtherm tpr_shadow vnmi flexpriority ept vpid
bogomips : 6584.29
clflush size : 64
cache_alignment : 64
address sizes : 36 bits physical, 48 bits virtual

> --
> You received this message because you are subscribed to the Google Groups
> "sage-devel" group.
> To post to this group, send email to sage-...@googlegroups.com.
> To unsubscribe from this group, send email to
> sage-devel+...@googlegroups.com.
> Visit this group at http://groups.google.com/group/sage-devel?hl=en.
>
>

[John Cremona]

On 4 October 2012 12:01, Volker Braun <vbrau...@gmail.com> wrote:

> I think its a known issue that the old ATLAS won't work with the newest
> CPUs. Can you try the new ATLAS spkg:
>
> http://trac.sagemath.org/10508
>

To check I am doing the right thing:

1. #10508 had various dependencies now merged in 5.4 so I switched to
building 5.4.rc0 instead of 5.3. (This is a machine I will not be
doing development on so normally only build real releases).

2. I downloaded the new atlas and cvxopt spkgs from that ticket to
SAGE_ROOT/spkg/standard

3. Before starting to build, I see that tehre are two patches to be
applied. I normally do that using hg, but this machine has an old
version (1.0.2) and I cannot do hg import without getting the error

abort: requirement 'fncache' not supported!

So I applied the first patch (needed for building) using "patch"
instead of "hg" and have left the other two until later as they only
involve documentation and doctests.

I'll report back (here and/or on that ticket).

John

[Jeroen Demeyer]

On 2012-10-04 13:41, John Cremona wrote:
> 2. I downloaded the new atlas and cvxopt spkgs from that ticket to
> SAGE_ROOT/spkg/standard

Good.

> 3. Before starting to build, I see that tehre are two patches to be
> applied.

Not sure where you get the number "two" from, I assume you meant either
"one" or "three", depending on what you count.

> So I applied the first patch (needed for building) using "patch"
> instead of "hg" and have left the other two until later as they only
> involve documentation and doctests.

I don't know what will happen when the SAGE_ROOT spkg is installed. In
the worst case, it would fail because of uncommitted changes. Let's
hope not, and then this would indeed be the correct way.

[John Cremona]

On 4 October 2012 12:59, Jeroen Demeyer <jdem...@cage.ugent.be> wrote:
> On 2012-10-04 13:41, John Cremona wrote:
>> 2. I downloaded the new atlas and cvxopt spkgs from that ticket to
>> SAGE_ROOT/spkg/standard
> Good.
>
>> 3. Before starting to build, I see that tehre are two patches to be
>> applied.
> Not sure where you get the number "two" from, I assume you meant either
> "one" or "three", depending on what you count.
>

I had miscounted indeed!

>> So I applied the first patch (needed for building) using "patch"
>> instead of "hg" and have left the other two until later as they only
>> involve documentation and doctests.
> I don't know what will happen when the SAGE_ROOT spkg is installed. In
> the worst case, it would fail because of uncommitted changes. Let's
> hope not, and then this would indeed be the correct way.
>

We'll see. It has not failed yet (currently building gfan)...

[John Cremona]

Progress report: the build finished fine. Sage runs.

Next I started to apply the other two patches on #10508. Since I
don't have a working hg I used sage -hg but:

masgaj@host-56-160%pwd
/home/masgaj/local/sage-5.4.rc0/devel/sage

masgaj@host-56-160%../../sage -hg import
http://trac.sagemath.org/sage_trac/attachment/ticket/10508/trac_10508_doctest.patch
applying http://trac.sagemath.org/sage_trac/attachment/ticket/10508/trac_10508_doctest.patch
abort: failed to synchronize metadata for "sage/misc/package.py"

ideas? It's not serious, I am doing a full doctest anyway and will
see if anything appears not addressed by that (unapplied) patch.

John

[Volker Braun]

You are trying to import the html-formatted patch. The correct url would be 

sage -hg import http://trac.sagemath.org/sage_trac/raw-attachment/ticket/10508/trac_10508_doctest.patch

Note the /raw-attachment/ instead of /attachment/ in the url.

[John Cremona]

On 4 October 2012 16:35, Volker Braun <vbrau...@gmail.com> wrote:
> You are trying to import the html-formatted patch. The correct url would be
>
> sage -hg import
> http://trac.sagemath.org/sage_trac/raw-attachment/ticket/10508/trac_10508_doctest.patch
>
> Note the /raw-attachment/ instead of /attachment/ in the url.
>

Sorry, that was stupid of me. ANyway, all the doctests except the one
which that patch fixed did pass, and I now have a working 5.4.rc0 on
the new machine, built with the new atlas spkg, total build time 90m.

thanks for the help!

John