The LAPACK forum has moved to https://github.com/Reference-LAPACK/lapack/discussions.

LAPACK 3.1 for 3.2 GHz P4-64bit LINUX using ifort (8.1-9.0)

Open discussion regarding features, bugs, issues, vendors, etc.

LAPACK 3.1 for 3.2 GHz P4-64bit LINUX using ifort (8.1-9.0)

Postby khosrow06 » Thu Dec 21, 2006 4:51 am

Dear all lapack-users,

I have been trying to install lapack 3.1 on a 3.2 GHz P4-64bit LINUX machine. At some point during the TEST phase, the programs stops responding and TOP shows that a program XLINTSTD is taking %99.9 of one of the CPUs:

Code: Select all
31266 khosrow   25   0  4992 2276 1596 R 99.9  0.1  32:19.20 xlintstd


At this phase, I have successfully installed BLAS, and the upper LAPACK routine. I then tried to start the TEST phase by typing:

Code: Select all
ca:lapack-3.1.0>make
( cd INSTALL; make; ./testlsame; ./testslamch; \
  ./testdlamch; ./testsecond; ./testdsecnd; ./testversion )
make[1]: Entering directory `/u/khosrow/lib/lapack-3.1.0/INSTALL'
make[1]: Nothing to be done for `all'.


And this is where the code stops responding:

Code: Select all
ifort -O3  aladhd.o alaerh.o alaesm.o alahd.o alareq.o alasum.o alasvm.o chkxer.o icopy.o ilaenv.o xlaenv.o xerbla.o dlaord.o  dchkaa.o dchkeq.o dchkgb.o dchkge.o dchkgt.o dchklq.o dchkpb.o dchkpo.o dchkpp.o dchkpt.o dchkq3.o dchkql.o dchkqp.o dchkqr.o dchkrq.o dchksp.o dchksy.o dchktb.o dchktp.o dchktr.o dchktz.o ddrvgb.o ddrvge.o ddrvgt.o ddrvls.o ddrvpb.o ddrvpo.o ddrvpp.o ddrvpt.o ddrvsp.o ddrvsy.o derrge.o derrgt.o derrlq.o derrls.o derrpo.o derrql.o derrqp.o derrqr.o derrrq.o derrsy.o derrtr.o derrtz.o derrvx.o dgbt01.o dgbt02.o dgbt05.o dgelqs.o dgeqls.o dgeqrs.o dgerqs.o dget01.o dget02.o dget03.o dget04.o dget06.o dget07.o dgtt01.o dgtt02.o dgtt05.o dlaptm.o dlarhs.o dlatb4.o dlattb.o dlattp.o dlattr.o dlavsp.o dlavsy.o dlqt01.o dlqt02.o dlqt03.o dpbt01.o dpbt02.o dpbt05.o dpot01.o dpot02.o dpot03.o dpot05.o dppt01.o dppt02.o dppt03.o dppt05.o dptt01.o dptt02.o dptt05.o dqlt01.o dqlt02.o dqlt03.o dqpt01.o dqrt01.o dqrt02.o dqrt03.o dqrt11.o dqrt12.o dqrt13.o dqrt14.o dqrt15.o dqrt16.o dqrt17.o drqt01.o drqt02.o drqt03.o drzt01.o drzt02.o dspt01.o dsyt01.o dtbt02.o dtbt03.o dtbt05.o dtbt06.o dtpt01.o dtpt02.o dtpt03.o dtpt05.o dtpt06.o dtrt01.o dtrt02.o dtrt03.o dtrt05.o dtrt06.o dtzt01.o dtzt02.o \
        ../../tmglib_LINUX.a ../../lapack_LINUX.a ../../blas_LINUX.a  -o ../xlintstd
make[2]: Leaving directory `/u/khosrow/lib/lapack-3.1.0/TESTING/LIN'
Testing DOUBLE PRECISION LAPACK linear equation routines
./xlintstd < dtest.in > dtest.out 2>&1


I am using ifort 8.1 (but can switch to ifort 9.0) and my make.inc looks like the following. It is the same as the one suggested for Northwood Xeon structure by Julien on this forum.

Code: Select all
FORTRAN  = ifort
OPTS     = -O3
DRVOPTS  = $(OPTS)
NOOPT    =-O3 -fltconsistency -fp_port
LOADER   = ifort
LOADOPTS = -O3
#
#  The archiver and the flag(s) to use when building archive (library)
#  If you system has no ranlib, set RANLIB = echo.
#
ARCH     = ar
ARCHFLAGS= cr
RANLIB   = ranlib


I would appreciate if anyone could help me by providing the correct form of the make.inc file for the machine strcuture I am working with.

Cheers,
Khosrow
khosrow06
 
Posts: 3
Joined: Mon Dec 18, 2006 12:16 am
Location: Princeton, NJ

Postby Julien Langou » Fri Dec 22, 2006 3:47 pm

Hello Khosrow,
have a look at Julie's post at:
http://icl.cs.utk.edu/lapack-forum/viewtopic.php?t=295
The correct flags for NOOPT and the Intel Fortran compiler are -fltconsistency -fp_port.
(Do not forget the 'make clean' before recompiling.)
Let us know if the testing are still hanging after this,
Julien.
Julien Langou
 
Posts: 835
Joined: Thu Dec 09, 2004 12:32 pm
Location: Denver, CO, USA

similar hang on amd64

Postby steve » Tue Jul 10, 2007 8:52 pm

Seems like I'm having a similar issue on an amd64 box, at least the test script pegs the CPU at 99%, but stalls at a different test:

Code: Select all
...
Testing SINGLE-DOUBLE PRECISION LAPACK prototype linear equation routines
./xlintstds < dstest.in > dstest.out 2>&1
Testing COMPLEX-COMPLEX16 LAPACK protoype linear equation routines
./xlintstzc < zctest.in > zctest.out 2>&1


I ran it once last night and it sat at the above test for several hours, then again today, and it's been stuck there for over 10 hours. I can kill it, so it's not locked up, but it never seems to get any farther. I'm still researching it, so any hints would be appreciated.

Machine and toolchain info:

Code: Select all
Portage 2.1.3_rc6 (default-linux/amd64/2006.1/desktop, gcc-4.1.2, glibc-2.5-r3, 2.6.21.1 x86_64)
=================================================================
System uname: 2.6.21.1 x86_64 AMD Athlon(tm) 64 Processor 3000+
Gentoo Base System release 1.12.10
Timestamp of tree: Sun, 08 Jul 2007 09:20:01 +0000
ccache version 2.4 [enabled]
dev-java/java-config: 1.3.7, 2.0.33-r1
dev-lang/python:     2.4.4-r4
dev-python/pycrypto: 2.0.1-r5
dev-util/ccache:     2.4-r7
sys-apps/sandbox:    1.2.18.1
sys-devel/autoconf:  2.13, 2.61
sys-devel/automake:  1.4_p6, 1.5, 1.6.3, 1.7.9-r1, 1.8.5-r3, 1.9.6-r2, 1.10
sys-devel/binutils:  2.17
sys-devel/gcc-config: 1.3.16
sys-devel/libtool:   1.5.24
virtual/os-headers:  2.6.20-r2
ACCEPT_KEYWORDS="amd64"
AUTOCLEAN="yes"
CBUILD="x86_64-pc-linux-gnu"
CFLAGS="-march=k8 -O2 -pipe"
CHOST="x86_64-pc-linux-gnu"
CONFIG_PROTECT="/etc /usr/kde/3.5/env /usr/kde/3.5/share/config /usr/kde/3.5/shutdown /usr/share/X11/xkb /usr/share/config"
CONFIG_PROTECT_MASK="/etc/env.d /etc/env.d/java/ /etc/gconf /etc/revdep-rebuild /etc/terminfo /etc/texmf/web2c"
CXXFLAGS="-march=k8 -O2 -pipe"
DISTDIR="/usr/portage/distfiles"
FEATURES="autoaddcvs buildpkg ccache cvs digest distlocks metadata-transfer multilib-strict sandbox sfperms strict unmerge-orphans userpriv usersandbox"
GENTOO_MIRRORS="http://local/gentoo/"
LDFLAGS="-Wl,-O1"
LINGUAS="en"
MAKEOPTS="-j3"
PKGDIR="/usr/portage/packages"
PORTAGE_RSYNC_OPTS="--recursive --links --safe-links --perms --times --compress --force --whole-file --delete --delete-after --stats --timeout=180 --exclude=/distfiles --exclude=/local --exclude=/packages --filter=H_**/files/digest-*"
PORTAGE_TMPDIR="/var/tmp"
PORTDIR="/usr/portage"
PORTDIR_OVERLAY="/usr/local/portage"
SYNC="rsync://local/gentoo-portage"
Unset:  CTARGET, EMERGE_DEFAULT_OPTS, INSTALL_MASK, LANG, LC_ALL, PORTAGE_COMPRESS, PORTAGE_COMPRESS_FLAGS, PORTAGE_RSYNC_EXTRA_OPTS


relevant processes:

Code: Select all
 7468 ?        Ss     0:00 SCREEN
 7469 pts/0    Ss     0:00 -/bin/bash
 7533 pts/0    S+     0:06 /usr/bin/python -O /usr/bin/emerge --oneshot lapack-reference -v
 7874 ?        Z      0:00 [ruby] <defunct>
14224 pts/0    S+     0:00 tee -i -a /var/log/portage/sci-libs:lapack-reference-3.1.0:20070710-141438.log
14225 pts/0    S+     0:00 [lapack-reference-3.1.0] sandbox /usr/lib/portage/bin/ebuild.sh test
14226 pts/0    S+     0:00 /bin/bash /usr/lib/portage/bin/ebuild.sh test
14503 pts/0    S+     0:00 make -j3
16827 pts/0    S+     0:00 /bin/sh -c ./xlintsts < stest.in > stest.out 2>&1
16828 pts/0    R+   548:48 ./xlintsts
steve
 
Posts: 3
Joined: Tue Jul 10, 2007 8:32 pm
Location: Santa Maria, CA

Postby Julie » Wed Jul 11, 2007 11:23 am

Khosrow
Could you provide the make.inc you are using
Thanks
Julie
Julie
 
Posts: 299
Joined: Wed Feb 23, 2005 12:32 am
Location: ICL, Denver. Colorado

test issue

Postby steve » Fri Jul 13, 2007 8:12 pm

No hints for me?
steve
 
Posts: 3
Joined: Tue Jul 10, 2007 8:32 pm
Location: Santa Maria, CA

weird update...

Postby steve » Fri Jul 13, 2007 10:58 pm

Here's the latest run; it hasn't moved in over an hour, and it looks like there are no current *.out files:

Code: Select all
lorentz TESTING # ls -l *.out
-rw-r--r-- 1 portage portage    335 Jul 13 18:48 cbak.out
-rw-r--r-- 1 portage portage    377 Jul 13 18:48 cbal.out
-rw-r--r-- 1 portage portage    985 Jul 13 18:48 cbb.out
-rw-r--r-- 1 portage portage    544 Jul 13 18:47 cec.out
-rw-r--r-- 1 portage portage   3529 Jul 13 18:47 ced.out
-rw-r--r-- 1 portage portage    459 Jul 13 18:48 cgbak.out
-rw-r--r-- 1 portage portage    389 Jul 13 18:48 cgbal.out
-rw-r--r-- 1 portage portage   4599 Jul 13 18:48 cgd.out
-rw-r--r-- 1 portage portage  47549 Jul 13 18:48 cgg.out
-rw-r--r-- 1 portage portage    738 Jul 13 18:48 cglm.out
-rw-r--r-- 1 portage portage    670 Jul 13 18:48 cgqr.out
-rw-r--r-- 1 portage portage    779 Jul 13 18:48 cgsv.out
-rw-r--r-- 1 portage portage    723 Jul 13 18:48 clse.out
-rw-r--r-- 1 portage portage   8445 Jul 13 18:47 cnep.out
-rw-r--r-- 1 portage portage    696 Jul 13 18:48 csb.out
-rw-r--r-- 1 portage portage  33481 Jul 13 18:47 csep.out
-rw-r--r-- 1 portage portage    945 Jul 13 18:48 csg.out
-rw-r--r-- 1 portage portage   2034 Jul 13 18:49 csvd.out
-rw-r--r-- 1 portage portage 964053 Jul 13 18:49 ctest.out
-rw-r--r-- 1 portage portage    335 Jul 13 18:46 dbak.out
-rw-r--r-- 1 portage portage    377 Jul 13 18:46 dbal.out
-rw-r--r-- 1 portage portage    985 Jul 13 18:46 dbb.out
-rw-r--r-- 1 portage portage    515 Jul 13 18:46 dec.out
-rw-r--r-- 1 portage portage   3529 Jul 13 18:46 ded.out
-rw-r--r-- 1 portage portage    459 Jul 13 18:46 dgbak.out
-rw-r--r-- 1 portage portage    377 Jul 13 18:46 dgbal.out
-rw-r--r-- 1 portage portage  19655 Jul 13 18:46 dgd.out
-rw-r--r-- 1 portage portage   1593 Jul 13 18:46 dgg.out
-rw-r--r-- 1 portage portage    738 Jul 13 18:46 dglm.out
-rw-r--r-- 1 portage portage    670 Jul 13 18:46 dgqr.out
-rw-r--r-- 1 portage portage    779 Jul 13 18:46 dgsv.out
-rw-r--r-- 1 portage portage    723 Jul 13 18:46 dlse.out
-rw-r--r-- 1 portage portage   1787 Jul 13 18:46 dnep.out
-rw-r--r-- 1 portage portage    691 Jul 13 18:46 dsb.out
-rw-r--r-- 1 portage portage   3540 Jul 13 18:46 dsep.out
-rw-r--r-- 1 portage portage    945 Jul 13 18:46 dsg.out
-rw-r--r-- 1 portage portage   2034 Jul 13 18:47 dsvd.out
-rw-r--r-- 1 portage portage   4138 Jul 13 18:48 dtest.out
-rw-r--r-- 1 portage portage    335 Jul 13 18:47 sbak.out
-rw-r--r-- 1 portage portage    377 Jul 13 18:47 sbal.out
-rw-r--r-- 1 portage portage    985 Jul 13 18:47 sbb.out
-rw-r--r-- 1 portage portage    722 Jul 13 18:47 sec.out
-rw-r--r-- 1 portage portage  11154 Jul 13 18:47 sed.out
-rw-r--r-- 1 portage portage    459 Jul 13 18:47 sgbak.out
-rw-r--r-- 1 portage portage    377 Jul 13 18:47 sgbal.out
-rw-r--r-- 1 portage portage   8934 Jul 13 18:47 sgd.out
-rw-r--r-- 1 portage portage  45437 Jul 13 18:47 sgg.out
-rw-r--r-- 1 portage portage    738 Jul 13 18:47 sglm.out
-rw-r--r-- 1 portage portage   2825 Jul 13 18:47 sgqr.out
-rw-r--r-- 1 portage portage    779 Jul 13 18:47 sgsv.out
-rw-r--r-- 1 portage portage    723 Jul 13 18:47 slse.out
-rw-r--r-- 1 portage portage  18735 Jul 13 18:46 snep.out
-rw-r--r-- 1 portage portage    691 Jul 13 18:47 ssb.out
-rw-r--r-- 1 portage portage   8675 Jul 13 18:47 ssep.out
-rw-r--r-- 1 portage portage    945 Jul 13 18:47 ssg.out
-rw-r--r-- 1 portage portage   3543 Jul 13 18:47 ssvd.out
-rw-r--r-- 1 portage portage 262029 Jul 13 18:46 stest.out
lorentz TESTING # date
Fri Jul 13 19:55:08 PDT 2007

Weird...
steve
 
Posts: 3
Joined: Tue Jul 10, 2007 8:32 pm
Location: Santa Maria, CA

Postby Julie » Sat Jul 14, 2007 9:29 am

Steve,

Could I get the output of dstest.out and zctest.out?

Usually, if a test stalls,
Code: Select all
lower the optimization
of the compiler by changing the compiler's flag.do a
Code: Select all
make cleanall
then a
Code: Select all
make
.
If you do no remove the TESTING/*.out then it will not run again the testings.

Julie
Julie
 
Posts: 299
Joined: Wed Feb 23, 2005 12:32 am
Location: ICL, Denver. Colorado


Return to User Discussion

Who is online

Users browsing this forum: No registered users and 4 guests