Re: [R-pkg-devel] r-patched-solaris-sparc error

2015-06-23 Thread carlos cinelli
Thanks a lot, Martyn! On Tue, Jun 23, 2015 at 7:18 AM, Martyn Plummer wrote: > Hi Carlos, > > The underlying cause is probably some C code that makes assumptions > about byte order or alignment that are not true on Solaris SPARC. Since > your code contains no compiled code that points to one of

Re: [R-pkg-devel] debugging problems with valgrind

2015-06-23 Thread Dirk Eddelbuettel
My approach would be to write a small C++ test function and main program against the C++ library you use, and run it in both 32 and 64 bit. Without using R. You then know whether the difference you see is due to the library, or your R wrapping, which should give you an idea about where to make ch

[R-pkg-devel] debugging problems with valgrind

2015-06-23 Thread code M
i am developing a little package that basically is a simple wrapper around some C++ code, that is being called via the RCpp package. Basically the code works, but i did find out that the results are different for 32bit and 64bit environments, so this might point to problems in the C++ code. The 32b

Re: [R-pkg-devel] r-patched-solaris-sparc error

2015-06-23 Thread Martyn Plummer
Hi Carlos, The underlying cause is probably some C code that makes assumptions about byte order or alignment that are not true on Solaris SPARC. Since your code contains no compiled code that points to one of the package's dependencies. In fact the data.table package also fails on Solaris SPARC.