[Bioc-devel] Trouble installing S4Vectors?

Hervé Pagès hpages at fredhutch.org
Thu Aug 27 23:14:07 CEST 2015


On 08/27/2015 10:21 AM, Charles Determan wrote:
> I am simply using gcc-4.8
>
> Here is the full output:
>
> BioC_mirror: http://bioconductor.org
> Using Bioconductor version 3.1 (BiocInstaller 1.18.4), R version 3.2.2.
> Installing package(s) ‘S4Vectors’
> trying URL
> 'http://bioconductor.org/packages/3.1/bioc/src/contrib/S4Vectors_0.6.3.tar.gz'
> Content type 'application/x-gzip' length 187491 bytes (183 KB)
> ==================================================
> downloaded 183 KB
>
> * installing *source* package ‘S4Vectors’ ...
> ** libs
> ccache gcc-4.8 -I/usr/share/R/include -DNDEBUG      -fpic  -Wall -O3
> -std=c99 -c AEbufs.c -o AEbufs.o
> ccache gcc-4.8 -I/usr/share/R/include -DNDEBUG      -fpic  -Wall -O3
> -std=c99 -c DataFrame_class.c -o DataFrame_class.o
> ccache gcc-4.8 -I/usr/share/R/include -DNDEBUG      -fpic  -Wall -O3
> -std=c99 -c Hits_class.c -o Hits_class.o
> Hits_class.c: In function ‘Hits_new’:
> Hits_class.c:157:13: warning: ‘q_hits2’ may be used uninitialized in
> this function [-Wmaybe-uninitialized]
>     tsort_hits(q_hits2, s_hits, qh_out, sh_out, nhit,
>               ^
> Hits_class.c:144:7: note: ‘q_hits2’ was declared here
>    int *q_hits2, *qh_out, *sh_out;
>         ^
> Hits_class.c:233:14: warning: ‘revmap’ may be used uninitialized in this
> function [-Wmaybe-uninitialized]
>    defineVar(install(translateChar(symbol)), revmap, revmap_envir);
>                ^
> ccache gcc-4.8 -I/usr/share/R/include -DNDEBUG      -fpic  -Wall -O3
> -std=c99 -c List_class.c -o List_class.o
> ccache gcc-4.8 -I/usr/share/R/include -DNDEBUG      -fpic  -Wall -O3
> -std=c99 -c R_init_S4Vectors.c -o R_init_S4Vectors.o
> ccache gcc-4.8 -I/usr/share/R/include -DNDEBUG      -fpic  -Wall -O3
> -std=c99 -c Rle_class.c -o Rle_class.o
> ccache gcc-4.8 -I/usr/share/R/include -DNDEBUG      -fpic  -Wall -O3
> -std=c99 -c Rle_utils.c -o Rle_utils.o
> ccache gcc-4.8 -I/usr/share/R/include -DNDEBUG      -fpic  -Wall -O3
> -std=c99 -c SEXP_utils.c -o SEXP_utils.o
> ccache gcc-4.8 -I/usr/share/R/include -DNDEBUG      -fpic  -Wall -O3
> -std=c99 -c SimpleList_class.c -o SimpleList_class.o
> ccache gcc-4.8 -I/usr/share/R/include -DNDEBUG      -fpic  -Wall -O3
> -std=c99 -c anyMissing.c -o anyMissing.o
> ccache gcc-4.8 -I/usr/share/R/include -DNDEBUG      -fpic  -Wall -O3
> -std=c99 -c eval_utils.c -o eval_utils.o
> ccache gcc-4.8 -I/usr/share/R/include -DNDEBUG      -fpic  -Wall -O3
> -std=c99 -c hash_utils.c -o hash_utils.o
> ccache gcc-4.8 -I/usr/share/R/include -DNDEBUG      -fpic  -Wall -O3
> -std=c99 -c int_utils.c -o int_utils.o
> ccache gcc-4.8 -I/usr/share/R/include -DNDEBUG      -fpic  -Wall -O3
> -std=c99 -c logical_utils.c -o logical_utils.o
> ccache gcc-4.8 -I/usr/share/R/include -DNDEBUG      -fpic  -Wall -O3
> -std=c99 -c safe_arithm.c -o safe_arithm.o
> ccache gcc-4.8 -I/usr/share/R/include -DNDEBUG      -fpic  -Wall -O3
> -std=c99 -c sort_utils.c -o sort_utils.o
> ccache gcc-4.8 -I/usr/share/R/include -DNDEBUG      -fpic  -Wall -O3
> -std=c99 -c str_utils.c -o str_utils.o
> str_utils.c: In function ‘get_svn_time’:
> str_utils.c:259:2: warning: implicit declaration of function ‘tzset’
> [-Wimplicit-function-declaration]
>    tzset();
>    ^
> str_utils.c:261:18: error: ‘timezone’ undeclared (first use in this
> function)
>    utc_offset = - (timezone / 3600);
>                    ^
> str_utils.c:261:18: note: each undeclared identifier is reported only
> once for each function it appears in
> make: *** [str_utils.o] Error 1
> ERROR: compilation failed for package ‘S4Vectors’
> * removing ‘/home/cdeterman/R/x86_64-pc-linux-gnu-library/3.2/S4Vectors’

The -std=c99 compilation flag seems to be causing this. On our build
machines that use gcc (i.e. Linux, Windows, and Snow Leopard), we use
-std=gnu99. Check what flag your R is using by running:

   R CMD config CC

in your terminal. My understanding is that gcc doesn't fully support
c99 yet, and that R packages are only required to support gnu99.
AFAIK the standard procedure for configuring/compiling/installing
R should set this to gnu99 when gcc is the default compiler.
So I'm not not sure why R is configured to use -std=c99 on your
system.

Anyway I just made a change to S4Vectors so now it should compile
with -std=c99. The change is in versions 0.6.4 (release) and 0.7.14
(devel) which should become available in the next 24 hours or less.
You might run into compilation problems with other packages that
don't support -std=c99 though so I would still encourage you to change
that setting (just edit R_HOME/etc/Makeconf if you know what you are
doing or ask a system admin to look into this).

Cheers,
H.


>
> On Thu, Aug 27, 2015 at 12:04 PM, Hervé Pagès <hpages at fredhutch.org
> <mailto:hpages at fredhutch.org>> wrote:
>
>     Hi Charles,
>
>     What compiler do you use? Please show the entire output of
>     biocLite("S4Vectors") plus your sessionInfo(). Thanks!
>
>     H.
>
>
>     On 08/27/2015 06:19 AM, Charles Determan wrote:
>
>         I am trying to run the basic install with biocLite.  I am
>         running R 3.2.2
>         on Ubuntu 14.04.
>
>         I install the BiocInstaller package with
>
>         source("http://bioconductor.org/biocLite.R")
>
>         I then try again to just install S4Vectors but it errors out on the
>         src_utils file with the following error:
>
>         str_utils.c: In function ‘get_svn_time’:
>         str_utils.c:259:2: warning: implicit declaration of function ‘tzset’
>         [-Wimplicit-function-declaration]
>             tzset();
>             ^
>         str_utils.c:261:18: error: ‘timezone’ undeclared (first use in this
>         function)
>             utc_offset = - (timezone / 3600);
>                             ^
>         str_utils.c:261:18: note: each undeclared identifier is reported
>         only once
>         for each function it appears in
>         make: *** [str_utils.o] Error 1
>         ERROR: compilation failed for package ‘S4Vectors’
>
>         Did I miss something?
>
>         Thanks,
>         Charles
>
>                  [[alternative HTML version deleted]]
>
>         _______________________________________________
>         Bioc-devel at r-project.org <mailto:Bioc-devel at r-project.org>
>         mailing list
>         https://stat.ethz.ch/mailman/listinfo/bioc-devel
>
>
>     --
>     Hervé Pagès
>
>     Program in Computational Biology
>     Division of Public Health Sciences
>     Fred Hutchinson Cancer Research Center
>     1100 Fairview Ave. N, M1-B514
>     P.O. Box 19024
>     Seattle, WA 98109-1024
>
>     E-mail: hpages at fredhutch.org <mailto:hpages at fredhutch.org>
>     Phone: (206) 667-5791 <tel:%28206%29%20667-5791>
>     Fax: (206) 667-1319 <tel:%28206%29%20667-1319>
>
>

-- 
Hervé Pagès

Program in Computational Biology
Division of Public Health Sciences
Fred Hutchinson Cancer Research Center
1100 Fairview Ave. N, M1-B514
P.O. Box 19024
Seattle, WA 98109-1024

E-mail: hpages at fredhutch.org
Phone:  (206) 667-5791
Fax:    (206) 667-1319



More information about the Bioc-devel mailing list