R devel - Nov 2012 - Crash - cause 'memory not mapped'

i'm using the following  c++ code
using namespace std;
#include <iostream>
#include <stdio.h>
#include <opencv/cv.h>
#include <opencv2/core/core.hpp>
#include <opencv2/highgui/highgui.hpp>
#include <opencv/highgui.h>
#include <opencv/cv.h> 
#include <R.h>
#include <Rinternals.h>
#include <Rmath.h>

extern "C" {
SEXP FiltroGaus(SEXP size1_r, SEXP size2_r, SEXP sigma_r) {

    int size1 = INTEGER(size1_r)[0];
    int size2 = INTEGER(size2_r)[0];
    double sigma = REAL(sigma_r)[0];
   
}
}

The compilation go without errors but when in R i type 
.C("FiltroGaus",3,3,2)
the message is:  Errore: INTEGER() can only be applied to a 'integer',
not a
'NULL'
if i type .C("FiltroGaus",as.integer(3),as.integer(3),as.double(2))
the message is Errore: INTEGER() can only be applied to a 'integer', not
a
'closure'

while if i type
.Call("FiltroGaus",as.integer(3),as.integer(3),as.double(2))
the message is 
NULL

 *** caught segfault ***
address 0x200000be, cause 'memory not mapped'

Possible actions:
1: abort (with core dump, if enabled)
2: normal R exit
3: exit R without saving workspace
4: exit R saving workspace


I now, i need to study and i know also for sure the answer is in some book
(i have  lot). I swear i read a lot and i'm still reading, but i really need
help :)



--
View this message in context:
http://r.789695.n4.nabble.com/Crash-cause-memory-not-mapped-tp4649141.html
Sent from the R devel mailing list archive at Nabble.com.

I'm not an expert, but from what I know you should certainly use the
.Call interface when you define a function as you did. The problem is
that your function must return a value (of type SEXP) which it does
not. If you don't return a value, all bets are off as to what will
happen, and if the internal R code uses a pointer to the result of
your function, you can easily get memory map errors.

HTH,

Peter

On Fri, Nov 9, 2012 at 2:31 PM, niandra <gianluca.mastrantonio at
yahoo.it> wrote:
> i'm using the following  c++ code
> using namespace std;
> #include <iostream>
> #include <stdio.h>
> #include <opencv/cv.h>
> #include <opencv2/core/core.hpp>
> #include <opencv2/highgui/highgui.hpp>
> #include <opencv/highgui.h>
> #include <opencv/cv.h>
> #include <R.h>
> #include <Rinternals.h>
> #include <Rmath.h>
>
> extern "C" {
> SEXP FiltroGaus(SEXP size1_r, SEXP size2_r, SEXP sigma_r) {
>
>     int size1 = INTEGER(size1_r)[0];
>     int size2 = INTEGER(size2_r)[0];
>     double sigma = REAL(sigma_r)[0];
>
> }
> }
>
> The compilation go without errors but when in R i type
> .C("FiltroGaus",3,3,2)
> the message is:  Errore: INTEGER() can only be applied to a
'integer', not a
> 'NULL'
> if i type
.C("FiltroGaus",as.integer(3),as.integer(3),as.double(2))
> the message is Errore: INTEGER() can only be applied to a
'integer', not a
> 'closure'
>
> while if i type
.Call("FiltroGaus",as.integer(3),as.integer(3),as.double(2))
> the message is
> NULL
>
>  *** caught segfault ***
> address 0x200000be, cause 'memory not mapped'
>
> Possible actions:
> 1: abort (with core dump, if enabled)
> 2: normal R exit
> 3: exit R without saving workspace
> 4: exit R saving workspace
>
>
> I now, i need to study and i know also for sure the answer is in some book
> (i have  lot). I swear i read a lot and i'm still reading, but i really
need
> help :)
>
>
>
> --
> View this message in context:
http://r.789695.n4.nabble.com/Crash-cause-memory-not-mapped-tp4649141.html
> Sent from the R devel mailing list archive at Nabble.com.
>
> ______________________________________________
> R-devel at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-devel

the error "memory non mapped" happen also if i use
void FiltroGaus(SEXP size1_r, SEXP size2_r, SEXP sigma_r) 
instead
SEXP FiltroGaus(SEXP size1_r, SEXP size2_r, SEXP sigma_r) {


Il giorno 09/nov/2012, alle ore 23:58, Peter Langfelder <peter.langfelder at
gmail.com> ha scritto:
> I'm not an expert, but from what I know you should certainly use the
> .Call interface when you define a function as you did. The problem is
> that your function must return a value (of type SEXP) which it does
> not. If you don't return a value, all bets are off as to what will
> happen, and if the internal R code uses a pointer to the result of
> your function, you can easily get memory map errors.
> 
> HTH,
> 
> Peter
> 
> On Fri, Nov 9, 2012 at 2:31 PM, niandra <gianluca.mastrantonio at
yahoo.it> wrote:
>> i'm using the following  c++ code
>> using namespace std;
>> #include <iostream>
>> #include <stdio.h>
>> #include <opencv/cv.h>
>> #include <opencv2/core/core.hpp>
>> #include <opencv2/highgui/highgui.hpp>
>> #include <opencv/highgui.h>
>> #include <opencv/cv.h>
>> #include <R.h>
>> #include <Rinternals.h>
>> #include <Rmath.h>
>> 
>> extern "C" {
>> SEXP FiltroGaus(SEXP size1_r, SEXP size2_r, SEXP sigma_r) {
>> 
>>    int size1 = INTEGER(size1_r)[0];
>>    int size2 = INTEGER(size2_r)[0];
>>    double sigma = REAL(sigma_r)[0];
>> 
>> }
>> }
>> 
>> The compilation go without errors but when in R i type
>> .C("FiltroGaus",3,3,2)
>> the message is:  Errore: INTEGER() can only be applied to a
'integer', not a
>> 'NULL'
>> if i type
.C("FiltroGaus",as.integer(3),as.integer(3),as.double(2))
>> the message is Errore: INTEGER() can only be applied to a
'integer', not a
>> 'closure'
>> 
>> while if i type
.Call("FiltroGaus",as.integer(3),as.integer(3),as.double(2))
>> the message is
>> NULL
>> 
>> *** caught segfault ***
>> address 0x200000be, cause 'memory not mapped'
>> 
>> Possible actions:
>> 1: abort (with core dump, if enabled)
>> 2: normal R exit
>> 3: exit R without saving workspace
>> 4: exit R saving workspace
>> 
>> 
>> I now, i need to study and i know also for sure the answer is in some
book
>> (i have  lot). I swear i read a lot and i'm still reading, but i
really need
>> help :)
>> 
>> 
>> 
>> --
>> View this message in context:
http://r.789695.n4.nabble.com/Crash-cause-memory-not-mapped-tp4649141.html
>> Sent from the R devel mailing list archive at Nabble.com.
>> 
>> ______________________________________________
>> R-devel at r-project.org mailing list
>> https://stat.ethz.ch/mailman/listinfo/r-devel

On 9 November 2012 at 14:31, niandra wrote:
| extern "C" {
| SEXP FiltroGaus(SEXP size1_r, SEXP size2_r, SEXP sigma_r) {

[...]
 
| The compilation go without errors but when in R i type 
| .C("FiltroGaus",3,3,2)

You missed the part in "Writing R Extension" about .C() and .Call()
arguments.  You are mixing both approaches here, which usually ends badly.

Do some more reading, and maybe study some existing (simpler) examples /
packages.  You really do not need to engage OpenCV headers and libraries
before you understand the basics better.  

Dirk

-- 
Dirk Eddelbuettel | edd at debian.org | http://dirk.eddelbuettel.com