| 1 |
On Monday 19 October 2009 16:59:55 Thomas Sachau wrote: |
| 2 |
> Mike Frysinger schrieb: |
| 3 |
> > the majority of the time, the compiler driver (i.e. `gcc`) should be used |
| 4 |
> > for linking. very few packages should invoke the linker directly. that |
| 5 |
> > is why currently the toolchain-func.eclass has tc-getLD return `ld` -- a |
| 6 |
> > few packages need it, but not most. if we're going to be forcing the |
| 7 |
> > setting of the LD env var all the time, then it needs to default to |
| 8 |
> > ${CC}. packages that need funky behavior should still work as they will |
| 9 |
> > be calling $(tc-getLD) anyways. |
| 10 |
> > |
| 11 |
> >>> - the -L paths to system dirs in LDFLAGS should not be there -- the |
| 12 |
> >>> toolchain can handle these just fine |
| 13 |
> >> |
| 14 |
> >> Last time i tried without, some packages failed to compile, will test it |
| 15 |
> >> again to check, if its still needed |
| 16 |
> > |
| 17 |
> > if things are failing, then we should look at gcc/binutils to make sure |
| 18 |
> > they use the right default search paths when given -m32/-m64 |
| 19 |
> |
| 20 |
> This is an example from configure failure with ABI=x86 for cvs-1.12.12-r6 |
| 21 |
> |
| 22 |
> last lines from configure: |
| 23 |
> |
| 24 |
> checking for ssh... ssh |
| 25 |
> checking for vim... /bin/nano |
| 26 |
> checking for temporary directory... /tmp |
| 27 |
> checking security/pam_appl.h usability... yes |
| 28 |
> checking security/pam_appl.h presence... yes |
| 29 |
> checking for security/pam_appl.h... yes |
| 30 |
> checking for pam_start in -lpam... no |
| 31 |
> configure: error: Could not find PAM libraries but the headers exist. |
| 32 |
> Give the --disable-pam option to compile without PAM support (or fix |
| 33 |
> your broken configuration) |
| 34 |
> |
| 35 |
> !!! Please attach the following file when seeking support: |
| 36 |
> !!! /var/tmp/portage/dev-util/cvs-1.12.12-r6/work/cvs-1.12.12/config.log |
| 37 |
> * ERROR: dev-util/cvs-1.12.12-r6 failed: |
| 38 |
> * econf failed |
| 39 |
> |
| 40 |
> relevant lines from config.log: |
| 41 |
> |
| 42 |
> configure:38697: checking for pam_start in -lpam |
| 43 |
> configure:38727: x86_64-pc-linux-gnu-gcc -o conftest -march=nocona -O2 |
| 44 |
> -pipe -m32 -L/emul/linux/x86/lib -L/emul/linux/x86/usr/lib -march=nocona |
| 45 |
> -O2 -pipe -Wl,--as-needed conftest.c -lpam -lnsl -lz >&5 |
| 46 |
> /usr/lib/gcc/x86_64-pc-linux-gnu/4.4.2/../../../../lib32/libpam.a(pam_dynam |
| 47 |
> ic.o): In function `_pam_dlerror': |
| 48 |
> (.text+0x1f): undefined reference to `dlerror' |
| 49 |
> /usr/lib/gcc/x86_64-pc-linux-gnu/4.4.2/../../../../lib32/libpam.a(pam_dynam |
| 50 |
> ic.o): In function `_pam_dlclose': |
| 51 |
> (.text+0x5f): undefined reference to `dlclose' |
| 52 |
> /usr/lib/gcc/x86_64-pc-linux-gnu/4.4.2/../../../../lib32/libpam.a(pam_dynam |
| 53 |
> ic.o): In function `_pam_dlsym': |
| 54 |
> (.text+0xa6): undefined reference to `dlsym' |
| 55 |
> /usr/lib/gcc/x86_64-pc-linux-gnu/4.4.2/../../../../lib32/libpam.a(pam_dynam |
| 56 |
> ic.o): In function `_pam_dlopen': |
| 57 |
> (.text+0xf2): undefined reference to `dlopen' |
| 58 |
> collect2: ld returned 1 exit status |
| 59 |
> configure:38733: $? = 1 |
| 60 |
> |
| 61 |
> If you need some more lines or complete build.log/config.log, feel free to |
| 62 |
> tell me and i will send them directly. |
| 63 |
|
| 64 |
please open a bug about this for the toolchain guys. i dont know when i'll |
| 65 |
get to researching this. |
| 66 |
-mike |
Archives