[R] Help deciphering segfault in make check

Karl-Dieter Crisman kcrisman at gmail.com
Tue Jan 19 21:05:41 CET 2010


Dear R Help,

I work with the Sage project, and we are trying to improve the ability
to use R through Sage.  Most things work, but make check seems to
cause problems on certain platforms, and now that we want to upgrade
to 2.10.1 I thought we should ask for help!

R builds just fine on both Mac and Linux, but some things in make
check seem to break on certain Linux boxes that don't on Mac or other
machines.  Here is an example from (Red Hat?) in doing the basic
tests, in src/tests/Examples/, where base-Ex.R seems to have a
problem, at the end of base-Ex.Rout.fail:

> ### Name: groupGeneric
> ### Title: S3 Group Generic Functions
> ### Aliases: S3groupGeneric .Group Math Math.data.frame Ops Ops.data.frame
> ###   Summary Summary.data.frame Complex 'group generic'
> ### Keywords: methods
>
> ### ** Examples
>
> require(utils)
>
> d.fr <- data.frame(x=1:9, y=stats::rnorm(9))
> class(1 + d.fr) == "data.frame" ##-- add to d.f. ...
[1] TRUE
>
> methods("Math")

 *** caught segfault ***
address 0x4b8, cause 'memory not mapped'

Traceback:
 1: dyn.load(file, DLLpath = DLLpath, ...)
 2: library.dynam(lib, package, package.lib)
 3: loadNamespace(name)
 4: doTryCatch(return(expr), name, parentenv, handler)
 5: tryCatchOne(expr, names, parentenv, handlers[[1L]])
 6: tryCatchList(expr, classes, parentenv, handlers)
 7: tryCatch(loadNamespace(name), error = function(e) stop(e))
 8: getNamespace(ns)
 9: asNamespace(pkg)
10: get(name, envir = asNamespace(pkg), inherits = FALSE)
11: tools:::.make_S3_methods_stop_list
12: methods("Math")
aborting ...
Segmentation fault

Note that make check for R in Sage passes on Mac OSX, as far as I can
tell.  For this particular machine and build, earlier in the same fail
file the methods package was listed as

methods       /home/.../sage-4.3-linux-sage.math.washington.edu-x86_64-Linux/spkg/build/r-2.10.1/src/library/methods/libs/methods.so

so it seems that the recommended package methods did build and load properly.

Any help would be greatly appreciated!  Thank you.



More information about the R-help mailing list