[R] [SOLVED] Problem building R-2.15.3 from source
Prof Brian Ripley
ripley at stats.ox.ac.uk
Thu Jul 11 08:05:41 CEST 2013
On 11/07/2013 06:44, Pascal Oettli wrote:
> Hello,
>
> Maybe the reason why:
>
> http://www.cran.r-project.org/doc/manuals/R-admin.html#Compilation-flags
And since this is a ix86 system, you really should be using -mtune=native.
I do not believe the default in 2.15.3 was -O3: it is not in current R,
and it certainly included -g .
>
> Regards,
> Pascal
>
>
>
> 2013/7/11 Mark Dalphin <mark.dalphin at pacificedge.co.nz>
>
>> I have found a solution to the repeated seg-faults below.
>>
>> If I set environment variables:
>> setenv CFLAGS -O2
>> setenv FFLAGS -O2
>> rather than the default -O3, then R builds and "checks" successfully.
>>
>> A few more details about the Debian system on which I have been building:
>> gcc (Debian 4.7.2-5) 4.7.2
>>
>> My "configure" command is:
>> ./configure \
>> --prefix=$my_R_path \
>> --with-readline \
>> --without-x \
>> --enable-R-shlib \
>> --enable-BLAS-shlib \
>> --with-system-zlib \
>> --with-system-bzlib \
>> --with-system-pcre
>>
>> So, I'm good for the time being and hope this helps others who have
>> trouble building from source.
Others who do not follow the instructions? As the manual says at the end
'Beware of using high levels of optimization, at least initially. On
many compilers these reduce the degree of compliance to the IEEE model.'
>>
>> Cheers,
>> Mark
>>
>> Mark Dalphin wrote:
>>
>>> Hi,
>>>
>>> I have for many years build R from source for Linux. I have just run into
>>> my first problem with this in ... I don't know how long.
>>>
>>> uname -a
>>> Linux douglas 3.2.0-4-686-pae #1 SMP Debian 3.2.46-1 i686 GNU/Linux
>>>
>>> cat /etc/issue
>>> Debian GNU/Linux 7 \n \l
>>>
>>>
>>> The version of R is 2.15.3. I know it is old, but we are in a regulated
>>> environment and changes to R versions are painful. I have built R 2.15.3
>>> elsewhere and have it running on multiple Linux boxes around here, both
>>> 32-bit and 64-bit; Ubuntu distributions, however, not Debian.
>>>
>>> This build is on a virtual machine under OpenBox. The host is a 64-bit
>>> Debian; the guest is a 32-bit Debian installation.
>>>
>>> The symptoms are strange (to me). I get segfaults during the
>>> byte-compiling phase of libraries. If I re-run 'make', the make proceeds as
>>> if it finished the previous seg-faulted step, and then segfaults on the
>>> next byte-compile. The "permissions" makes me wonder about file
>>> permissions, but the whole 'make' is under my HOME. Furthermore, I have
>>> scanned the unpacked tar-gz package for something I don't "own" and it
>>> isn't there. I also think segfaults are usually in memory, though I don't
>>> know what "permission" I have there (don't I own the RAM I request?).
>>>
>>> I have attached a section of the 'make' output below, followed by a the
>>> next "make" output:
>>> ------------------------------**------------------------------**
>>> ------------------------------**----------------
>>> make[4]: Leaving directory `/home/mdalphin/src/R-2.15.3/**
>>> src/library/splines'
>>> make[4]: Entering directory `/home/mdalphin/src/R-2.15.3/**
>>> src/library/splines'
>>> byte-compiling package 'splines'
>>>
>>> *** caught segfault ***
>>> address 0x403ac3dc, cause 'invalid permissions'
>>>
>>> Traceback:
>>> 1: fun(libname, pkgname)
>>> 2: doTryCatch(return(expr), name, parentenv, handler)
>>> 3: tryCatchOne(expr, names, parentenv, handlers[[1L]])
>>> 4: tryCatchList(expr, classes, parentenv, handlers)
>>> 5: tryCatch(fun(libname, pkgname), error = identity)
>>> 6: runHook(".onLoad", env, package.lib, package)
>>> 7: loadNamespace(name)
>>> 8: doTryCatch(return(expr), name, parentenv, handler)
>>> 9: tryCatchOne(expr, names, parentenv, handlers[[1L]])
>>> 10: tryCatchList(expr, classes, parentenv, handlers)
>>> 11: tryCatch(loadNamespace(name), error = function(e) stop(e))
>>> 12: getNamespace(ns)
>>> 13: asNamespace(pkg)
>>> 14: get(name, envir = asNamespace(pkg), inherits = FALSE)
>>> 15: compiler:::tryCmpfun
>>> 16: .Call("R_**lazyLoadDBinsertValue", x[[1L]], file, ascii, compress,
>>> hook, PACKAGE = "base")
>>> 17: lazyLoadDBinsertVariable(vars[**i], from, datafile, ascii, compress,
>>> envhook)
>>> 18: makeLazyLoadDB(ns, dbbase, compress = compress)
>>> 19: code2LazyLoadDB(package, lib.loc = lib.loc, keep.source =
>>> keep.source, compress = compress)
>>> 20: tools:::makeLazyLoading("**splines")
>>> aborting ...
>>> /bin/bash: line 8: 18709 Done echo
>>> "tools:::makeLazyLoading(\"**splines\")"
>>> 18710 Segmentation fault | R_COMPILE_PKGS=1
>>> R_COMPILER_SUPPRESS_ALL=1 R_DEFAULT_PACKAGES=NULL LC_ALL=C ../../../bin/R
>>> --vanilla --slave > /dev/null
>>> make[4]: *** [../../../library/splines/R/**splines.rdb] Error 139
>>> make[4]: Leaving directory `/home/mdalphin/src/R-2.15.3/**
>>> src/library/splines'
>>> make[3]: *** [all] Error 2
>>> make[3]: Leaving directory `/home/mdalphin/src/R-2.15.3/**
>>> src/library/splines'
>>> make[2]: *** [R] Error 1
>>> make[2]: Leaving directory `/home/mdalphin/src/R-2.15.3/**src/library'
>>> make[1]: *** [R] Error 1
>>> make[1]: Leaving directory `/home/mdalphin/src/R-2.15.3/**src'
>>> make: *** [R] Error 1
>>> ------------------------------**------------------------------**
>>> ------------------------------**-
>>> make[4]: Entering directory `/home/mdalphin/src/R-2.15.3/**
>>> src/library/splines'
>>> make[4]: Nothing to be done for `mklazycomp'.
>>> make[4]: Leaving directory `/home/mdalphin/src/R-2.15.3/**
>>> src/library/splines'
>>> make[3]: Leaving directory `/home/mdalphin/src/R-2.15.3/**
>>> src/library/splines'
>>> make[3]: Entering directory `/home/mdalphin/src/R-2.15.3/**
>>> src/library/stats4'
>>> building package 'stats4'
>>> mkdir -p -- ../../../library/stats4
>>> make[4]: Entering directory `/home/mdalphin/src/R-2.15.3/**
>>> src/library/stats4'
>>> mkdir -p -- ../../../library/stats4/R
>>> mkdir -p -- ../../../library/stats4/po
>>> make[4]: Leaving directory `/home/mdalphin/src/R-2.15.3/**
>>> src/library/stats4'
>>> make[4]: Entering directory `/home/mdalphin/src/R-2.15.3/**
>>> src/library/stats4'
>>> byte-compiling package 'stats4'
>>>
>>> *** caught segfault ***
>>> address 0x403ac3dc, cause 'invalid permissions'
>>>
>>> Traceback:
>>> 1: fun(libname, pkgname)
>>> 2: doTryCatch(return(expr), name, parentenv, handler)
>>> 3: tryCatchOne(expr, names, parentenv, handlers[[1L]])
>>> 4: tryCatchList(expr, classes, parentenv, handlers)
>>> 5: tryCatch(fun(libname, pkgname), error = identity)
>>> 6: runHook(".onLoad", env, package.lib, package)
>>> 7: loadNamespace(name)
>>> 8: doTryCatch(return(expr), name, parentenv, handler)
>>> 9: tryCatchOne(expr, names, parentenv, handlers[[1L]])
>>> 10: tryCatchList(expr, classes, parentenv, handlers)
>>> 11: tryCatch(loadNamespace(name), error = function(e) stop(e))
>>> 12: getNamespace(ns)
>>> 13: asNamespace(pkg)
>>> 14: get(name, envir = asNamespace(pkg), inherits = FALSE)
>>> 15: compiler:::tryCmpfun
>>> 16: .Call("R_**lazyLoadDBinsertValue", x[[1L]], file, ascii, compress,
>>> hook, PACKAGE = "base")
>>> 17: lazyLoadDBinsertVariable(vars[**i], from, datafile, ascii, compress,
>>> envhook)
>>> 18: makeLazyLoadDB(ns, dbbase, compress = compress)
>>> 19: code2LazyLoadDB(package, lib.loc = lib.loc, keep.source =
>>> keep.source, compress = compress)
>>> 20: tools:::makeLazyLoading("**stats4")
>>> aborting ...
>>> /bin/bash: line 8: 19554 Done echo
>>> "tools:::makeLazyLoading(\"**stats4\")"
>>> 19555 Segmentation fault | R_COMPILE_PKGS=1
>>> R_COMPILER_SUPPRESS_ALL=1 R_DEFAULT_PACKAGES="methods,**graphics,stats"
>>> LC_ALL=C ../../../bin/R --vanilla --slave > /dev/null
>>> make[4]: *** [../../../library/stats4/R/**stats4.rdb] Error 139
>>> make[4]: Leaving directory `/home/mdalphin/src/R-2.15.3/**
>>> src/library/stats4'
>>> make[3]: *** [all] Error 2
>>> make[3]: Leaving directory `/home/mdalphin/src/R-2.15.3/**
>>> src/library/stats4'
>>> make[2]: *** [R] Error 1
>>> make[2]: Leaving directory `/home/mdalphin/src/R-2.15.3/**src/library'
>>> make[1]: *** [R] Error 1
>>> make[1]: Leaving directory `/home/mdalphin/src/R-2.15.3/**src'
>>> make: *** [R] Error 1
>>>
>>>
>>> ------------------------------**------------------------------**
>>> ------------
>>>
>>> ______________________________**________________
>>> R-help at r-project.org mailing list
>>> https://stat.ethz.ch/mailman/**listinfo/r-help<https://stat.ethz.ch/mailman/listinfo/r-help>
>>> PLEASE do read the posting guide http://www.R-project.org/**
>>> posting-guide.html <http://www.R-project.org/posting-guide.html>
>>> and provide commented, minimal, self-contained, reproducible code.
>>>
>>>
>>
>> --
>>
>>
>>
>> Mark Dalphin Ph.D.
>>
>> Director of Bioinformatics
>>
>> mark.dalphin at pacificedge.co.nz <mailto:mark.dalphin@**pacificedge.co.nz<mark.dalphin at pacificedge.co.nz>
>>>
>> *Ph:* +64-3-479-5805
>> *Cell:* +64-21-156-7625
>> *Skype:* mark.dalphin.pel
>> <http://www.facebook.com/**pages/Pacific-Edge/**111356775582456<http://www.facebook.com/pages/Pacific-Edge/111356775582456>>
>> <http://twitter.com/#%21/**pacificEdgeLtd<http://twitter.com/#%21/pacificEdgeLtd>>
>> <http://www.youtube.com/**PacificEdgeLtd<http://www.youtube.com/PacificEdgeLtd>
>>>
>>
>> 87 St David St, PO Box 56, Dunedin, New Zealand 9016www.pacificedge.co.nz
>>
>>
>> ______________________________________________
>> R-help at r-project.org mailing list
>> https://stat.ethz.ch/mailman/listinfo/r-help
>> PLEASE do read the posting guide
>> http://www.R-project.org/posting-guide.html
>> and provide commented, minimal, self-contained, reproducible code.
>>
>>
>
> [[alternative HTML version deleted]]
>
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>
--
Brian D. Ripley, ripley at 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 272866 (PA)
Oxford OX1 3TG, UK Fax: +44 1865 272595
More information about the R-help
mailing list