Using the 64-bit Solaris compilers make check dumps core in La_rs at 73 F77_CALL(dsyev)(jobv, uplo, &n, rx, &n, rvalues, work, &lwork, &info); I can't reproduce it easily, but example(eigen) occasionally stops with Error: abs(sm - V %*% diag(lam) %*% t(V)) < 60 * Meps is not TRUE which might be connected. The tiny rounding errors in example(eigen) aren't constant from time to time, but I don't know if they should be. -thomas Thomas Lumley Asst. Professor, Biostatistics tlumley@u.washington.edu University of Washington, Seattle ^^^^^^^^^^^^^^^^^^^^^^^^ NOTE NEW EMAIL ADDRESS -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- r-devel mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html Send "info", "help", or "[un]subscribe" (in the "body", not the subject !) To: r-devel-request@stat.math.ethz.ch _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
On Wed, 20 Jun 2001 tlumley@u.washington.edu wrote:> > Using the 64-bit Solaris compilers make check dumps core in La_rs at > 73 F77_CALL(dsyev)(jobv, uplo, &n, rx, &n, rvalues, work, &lwork, > &info); > > I can't reproduce it easily, but example(eigen) occasionally stops withI can now reproduce it -- I was looking in the wrong place: it's in example(cmdscale) in the mva package. -thomas -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- r-devel mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html Send "info", "help", or "[un]subscribe" (in the "body", not the subject !) To: r-devel-request@stat.math.ethz.ch _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
On Tue, 19 Jun 2001, Thomas Lumley wrote:> On Wed, 20 Jun 2001 tlumley@u.washington.edu wrote: > > > > > Using the 64-bit Solaris compilers make check dumps core in La_rs at > > 73 F77_CALL(dsyev)(jobv, uplo, &n, rx, &n, rvalues, work, &lwork, > > &info); > > > > I can't reproduce it easily, but example(eigen) occasionally stops with > > I can now reproduce it -- I was looking in the wrong place: it's in > example(cmdscale) in the mva package.Our Solaris compilers run this without problem: I've been running that test most days. We have markov% f77 -V f77: Sun WorkShop 6 update 1 FORTRAN 77 5.2 2000/09/11 Usage: f77 [ options ] files. Use 'f77 -flags' for details markov% cc -V cc: Sun WorkShop 6 update 1 C 5.2 2000/09/11 usage: cc [ options] files. Use 'cc -flags' for details and I am usually running agressive optimization options, e.g. CC="cc -xarch=v9" FC="f95 -xarch=v9" CFLAGS="-xO5 -xlibmil -dalign -g" FFLAGS="-xO5 -xlibmil -dalign -g" asn that by default uses the BLAS in -lsunperf. What is your setup? -- Brian D. Ripley, ripley@stats.ox.ac.uk Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/ University of Oxford, Tel: +44 1865 272861 (self) 1 South Parks Road, +44 1865 272860 (secr) Oxford OX1 3TG, UK Fax: +44 1865 272595 -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- r-devel mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html Send "info", "help", or "[un]subscribe" (in the "body", not the subject !) To: r-devel-request@stat.math.ethz.ch _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
On Wed, 20 Jun 2001, Prof Brian D Ripley wrote:> > Our Solaris compilers run this without problem: I've been running that test > most days. > > We have > > markov% f77 -V > f77: Sun WorkShop 6 update 1 FORTRAN 77 5.2 2000/09/11 > Usage: f77 [ options ] files. Use 'f77 -flags' for details > markov% cc -V > cc: Sun WorkShop 6 update 1 C 5.2 2000/09/11 > usage: cc [ options] files. Use 'cc -flags' for details > > and I am usually running agressive optimization options, e.g. > > CC="cc -xarch=v9" > FC="f95 -xarch=v9" > > CFLAGS="-xO5 -xlibmil -dalign -g" > FFLAGS="-xO5 -xlibmil -dalign -g" > > asn that by default uses the BLAS in -lsunperf. > > What is your setup?I have the 5.0 Sun Workshop compilers, and I wasn't using any of those optimisation flags, though Fortran is using -lsunmath. I just did configure, make, with CC="cc -xarch=v9" and FC="f77 -xarch=v9" I'll try again with the alignment flags specified. -thomas -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- r-devel mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html Send "info", "help", or "[un]subscribe" (in the "body", not the subject !) To: r-devel-request@stat.math.ethz.ch _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
The crash is caused by a bus error: address alignment. Recompiling without -lsunperf and with -dalign fixes it. I think there is a configuration issue here (not for 1.3.0, though). The documentation for the Sun BLAS says that -dalign is needed, but the default configuration is to use the Sun BLAS and not -dalign. -thomas -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- r-devel mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html Send "info", "help", or "[un]subscribe" (in the "body", not the subject !) To: r-devel-request@stat.math.ethz.ch _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
>>>>> tlumley writes:> The crash is caused by a bus error: address alignment. Recompiling without > -lsunperf and with -dalign fixes it.> I think there is a configuration issue here (not for 1.3.0, > though). The documentation for the Sun BLAS says that -dalign is > needed, but the default configuration is to use the Sun BLAS and not > -dalign.Ok. Note that this part of the configure code is taken from Octave, which used to be reliable. It would be great if we could pass info on problems and solutions on to John Eaton as well (in particular as I think that at some point of time there was talk about adding the code to the autoconf contrib archive). We had a related problem last week, resulting in dropping CXML/DXML support due to problems on 64-bit platforms with gcc (if I recall correctly). Hmm ... -k -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- r-devel mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html Send "info", "help", or "[un]subscribe" (in the "body", not the subject !) To: r-devel-request@stat.math.ethz.ch _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._