On 29 March 2014 at 12:03, Ivo De Decker wrote:
| package: rmpi
| version: 0.6-3-2
| severity: serious
|
| Hi,
|
| The latest rmpi upload FTBFS on armel and armhf, but it built successfully
| before:
|
| https://buildd.debian.org/status/package.php?p=rmpi
Dang. The last change was a pretty innocent (and otherwise very useful) patch
by Don (CC'ed) -- see #741297 for more on that.
And the 'SHLIB_LIBADD' rings a bell, but I cannot recall where / how else we
fixed that :-/ I fear this is an arm-and-R issue, not a Rmpi issue. Eg on
amd64 box:
edd@max:~/src/debian/CRAN/Rmpi-0.6-3$ grep SHLIB_LIBADD /etc/R/Makeconf
SHLIB_LIBADD =
ALL_LIBS = $(PKG_LIBS) $(SHLIB_LIBADD) $(LIBR)# $(LIBINTL)
edd@max:~/src/debian/CRAN/Rmpi-0.6-3$
[ Ok, goes digging some more. ]
Bug #679180 was all about this already, so it is a regression.
HOWEVER this very issue is also said to be addressed in R 3.1.0. From its
NEWS file
* Some bugs in TRE have been corrected by updating from the git
repository. This allows R to be installed on some platforms for
which this was a blocker (PR#15087 suggests Linux on ARM and
HP-UX).
Here #15087 is a ticket in R's bugzilla.
Could you try with the r-base-core 3.0.3.20140328-1 build I uploaded
yesterday?
Dirk
|
| Cheers,
|
| Ivo
--
Dirk Eddelbuettel | [email protected] | http://dirk.eddelbuettel.com
--
To UNSUBSCRIBE, email to [email protected]
with a subject of "unsubscribe". Trouble? Contact [email protected]
