[Rcpp-devel] why segmentation fault happens when I was trying to run complied examples in RInside "standard" folder
Cong Li
solvercorleone at gmail.com
Tue Dec 4 16:16:49 CET 2012
Hi Dirk,
The problem has been solved after I downloaded Rcpp from svn repository,
and build it on my computer.
Thank you for your help
Cong Li
On Tue, Dec 4, 2012 at 11:52 PM, Dirk Eddelbuettel <edd at debian.org> wrote:
>
> On 4 December 2012 at 23:33, Cong Li wrote:
> | Hi, Dirk
> |
> | Ok, I see. I will try it.
> | Thank you very much.
> | And I am sorry for relying on the wrong place.
>
> Thanks -- and here is a cleaner diff from rev4047.
>
> I found/fixed that very shortly after we rolled 0.10.2 (which fixed a real
> rather than cosmetic error and had to be released).
>
> As we said, this is purely cosmetic but a little irratice. Just rebuild
> Rcpp
> with thus small change and all is well.
>
> Dirk
>
>
> Modified: pkg/Rcpp/inst/include/Rcpp/iostream/Rostream.h
> ===================================================================
> --- pkg/Rcpp/inst/include/Rcpp/iostream/Rostream.h 2012-11-27
> 01:34:59 UTC (rev 4046)
> +++ pkg/Rcpp/inst/include/Rcpp/iostream/Rostream.h 2012-11-27
> 14:21:42 UTC (rev 4047)
> @@ -38,7 +38,12 @@
> std::ostream( new Rstreambuf(output) ),
> buf( static_cast<Rstreambuf*>( rdbuf() ) )
> {}
> - ~Rostream(){ delete buf ; }
> + ~Rostream() {
> + if (buf != NULL) {
> + delete buf;
> + buf = NULL;
> + }
> + }
> };
>
> // declare global variable
>
>
> |
> | Cong Li
> |
> | On Tue, Dec 4, 2012 at 9:51 PM, Dirk Eddelbuettel <edd at debian.org>
> wrote:
> |
> |
> | Please reply on-list.
> |
> | On 4 December 2012 at 21:37, Cong Li wrote:
> | | Hi Dirk
> | |
> | | Thank you very much for your answer. It really helps.
> | |
> | | I want to call R from C++ in my project. If this is only a problem
> of R
> | version
> | | conflict,
> | | could I just ignore this "segmentation fault" info for the time
> being,
> | since I
> | | found that it seemed to impose no negative effect on c++ part,
> besides R
> | | calling, of my program(based on my experiment)?
> |
> | Yes indeed, see my follow-up on-list. Or you could quickly rebuild
> Rcpp
> | using
> | this version of the Rostream destructor:
> |
> | ~Rostream() {
> | if (buf != NULL) {
> | delete buf;
> | buf = NULL;
> | }
> | }
> |
> | Dirk
> |
> | | I am looking forward to the new version releasing, and thank you
> for your
> | | excellent jobs.
> | |
> | | Best Regards
> | |
> | | Cong Li
> | |
> | | On Tue, Dec 4, 2012 at 9:12 PM, Dirk Eddelbuettel <edd at debian.org>
> wrote:
> | |
> | |
> | | Hi Cong,
> | |
> | | On 4 December 2012 at 19:15, Cong Li wrote:
> | | | Hi guys
> | | |
> | | | I am a novice to RInside, and I met a problem when trying to
> run
> | RInside
> | | | example in "standard" folder.
> | | |
> | | | It is that a segmentation fault error happened every time
> when I
> | was
> | | trying to
> | | | execute the compiled example program in the "standard"
> folder. My
> | OS is
> | | ubuntu
> | | | 12.04, and I am using R 2.15.2 and g++ 4.6.3.
> | | |
> | | | It seems that the executions of the examples were fine, but
> | somehow
> | | errors
> | | | happened when then main function was trying to return.
> | | |
> | | | Can someone help me with this issue.
> | |
> | | It is a change in R 2.15.2 which requires a change in RInside
> which
> | I have
> | | yet to release (as another, unrelated, error) is holding this
> back.
> | If
> | | you
> | | use R 2.15.1 it all works too.
> | |
> | | Sorry for this, and the new RInside should be there "soon", or
> I
> | should at
> | | least commit my fixes to SVN.
> | |
> | | Dirk
> | |
> | | --
> | | Dirk Eddelbuettel | edd at debian.org |
> http://dirk.eddelbuettel.com
> | |
> | |
> |
> | --
> | Dirk Eddelbuettel | edd at debian.org | http://dirk.eddelbuettel.com
> |
> |
>
> --
> Dirk Eddelbuettel | edd at debian.org | http://dirk.eddelbuettel.com
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.r-forge.r-project.org/pipermail/rcpp-devel/attachments/20121205/5e749cfb/attachment-0001.html>
More information about the Rcpp-devel
mailing list