Il giorno Thu, 31 Jan 2008 15:38:58 +0200 Yavor Doganov <[EMAIL PROTECTED]> ha scritto:
> On Thu, Jan 31, 2008 at 12:33:32PM +0100, David Paleino wrote: > > Have you got any idea why configure fails? > > Try looking at config.log. > It doesn't fail for me (although I didn't try in a chroot). Here's what happens (relevant part): <http://pastebin.com/f3b94f132> Here's the relevant bit of the paste: configure:20303: checking for gsl_multifit_fdfsolver_iterate in -lgsl configure:20338: i486-linux-gnu-g++ -o conftest -O2 -g -Wall -Wl,--as-needed conftest.cpp -lgsl -lgslcblas -lz -lpthread -ldl -lm >&5 /usr/lib/gcc/i486-linux-gnu/4.2.3/../../../../lib/libgsl.so: undefined reference to `cblas_ctrmv' /usr/lib/gcc/i486-linux-gnu/4.2.3/../../../../lib/libgsl.so: undefined reference to `cblas_zswap' ... collect2: ld returned 1 exit status configure:20344: $? = 1 configure: failed program was: | /* confdefs.h. */ | #define PACKAGE_NAME "" | #define PACKAGE_TARNAME "" | #define PACKAGE_VERSION "" | #define PACKAGE_STRING "" | #define PACKAGE_BUGREPORT "" | #define PACKAGE "odin" | #define VERSION "1.6.5" | #define STDC_HEADERS 1 | #define HAVE_SYS_TYPES_H 1 | #define HAVE_SYS_STAT_H 1 | #define HAVE_STDLIB_H 1 | #define HAVE_STRING_H 1 | #define HAVE_MEMORY_H 1 | #define HAVE_STRINGS_H 1 | #define HAVE_INTTYPES_H 1 | #define HAVE_STDINT_H 1 | #define HAVE_UNISTD_H 1 | #define HAVE_DLFCN_H 1 | /* end confdefs.h. */ | | /* Override any GCC internal prototype to avoid an error. | Use char because int might match the return type of a GCC | builtin and then its argument prototype would still apply. */ | #ifdef __cplusplus | extern "C" | #endif | char gsl_multifit_fdfsolver_iterate (); | int | main () | { | return gsl_multifit_fdfsolver_iterate (); | ; | return 0; | } configure:20362: result: no ... It seems like the -lgslcblas flag doesn't get counted :s After some tests, here's the guilty: -Wl,--as-needed. (foo.cc is the same snippet of code in config.log) $ g++ foo.cc -Wall -Wl,-z,defs -lgsl -lgslcblas &>/dev/null $ echo $? 0 $ g++ foo.cc -Wall -Wl,-z,defs -Wl,--as-needed -lgsl -lgslcblas &>/dev/null $ echo $? 1 $ g++ foo.cc -Wall -Wl,-z,defs -Wl,--as-needed -lgsl -lgslcblas &>/dev/null $ echo $? 0 $ This is non-sense to me. Does the position of compiler/linker flags have any importance? Kindly, David P.S.: thanks for helping me in spotting the problem! :) -- . ''`. Debian maintainer | http://wiki.debian.org/DavidPaleino : :' : Linuxer #334216 --|-- http://www.hanskalabs.net/ `. `'` GPG: 1392B174 ----|---- http://snipr.com/qa_page `- 2BAB C625 4E66 E7B8 450A C3E1 E6AA 9017 1392 B174
signature.asc
Description: PGP signature
