Skip to content

make check fails after base-Ex.R

3 messages · Brian Ripley, Charles C. Berry

#
Dear R-help,


I tried to build the R-release version of R-1.0.0.

./configure and make seemed to go just fine.

When I ran 'make check' from $RHOME, it dumped core apparently in the 

	quit('no') 

on the last line of  base-Ex.R

I had been having some problems with a package I am building in that R
would report with "Process R segmentation violation (core dumped)" upon
q() in this and R-0.90.1 even though it seemed to do calculations that
agree with an equivalent package compiled and run under S+3.4.

So I tried to mimic base-Ex.R, viz.
(gdb) run
Starting program: /usr/local/R/R-release/bin/R.X11
[snip banner]
Save workspace image? [y/n/c]: n

Program received signal SIGSEGV, Segmentation fault.
0xfefe333c in __do_global_dtors_aux ()
   from /usr/local/R/R-release/library/mva/libs/mva.so
(gdb) 

----------

FWIW, upon restarting R, loading mva, and setting a breakpoint at
__do_global_dtors_aux, I determined that the segfault occurs in the
third attempted 'step'

This also happens in R-0.90.1.

So what is wrong with my setup.

I have 

Solaris 7,
gcc version 2.95.1 19990816 (release) w/ g77 (binary distrn from
sunfreeware.com)
/usr/ccs/bin/make (Sun)

and can send my config.site, etc if it would help to sort things out.

Please advise.

TIA,

Chuck Berry

--please do not edit the information below--

Version:
 platform = sparc-sun-solaris2.7
 arch = sparc
 os = solaris2.7
 system = sparc, solaris2.7
 status = Patched
 major = 1
 minor = 0.0
 year = 2000
 month = February
 day = 29
 language = R

Search Path:
 .GlobalEnv, Autoloads, package:base
#
On Sat, 11 Mar 2000, Charles C. Berry wrote:

            
Do you mean the R-release version, period?  There have been problems
with the distribution of that.
I assume that is somewhere deep in the exit code (it is not a public
routine, as its name shows). My guess is that it cannot unload shared
libraries.
We have no problems with Solaris 2.7 aka 7, gcc 2.95.2 (never tried 2.95.1,
which was current for only a very short while) built ourselves (on 2.6) and
GNU or Sun make.  On what version was your binary compiler distribution
built? That site has 2.95.2 for Solaris 7 currently (and emacs-20.4 even
though .6 is out!)  I would try that for a start (although I would build my
own).
This indicates the R-release distribution during a phase when it was out of
date (and so actually unpatched).
#
Following the suggestion (below), I installed gcc-2.95.2.

This solved the segfault and make check hiccups.

Thanks so much,

Chuck Berry
On Sun, 12 Mar 2000, Prof Brian D Ripley wrote:

            
That's what I meant.
[delete description of segfault and make check problems]
Charles C. Berry                        (858) 534-2098 
                                         Dept of Family/Preventive Medicine
E mailto:cberry at tajo.ucsd.edu	         UC San Diego
http://hacuna.ucsd.edu/members/ccb.html  La Jolla, San Diego 92093-0645

 

-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-help 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-help-request at stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._