Re: When is the next upload of gcc-4.3 planned?
Aurelien Jarno <aurelien@aurel32.net> writes:
> | debian/acats-killer.sh -p /tmp/gcc/gnat-4.3-4.3-20080127/build/acats-killer.pid \
> | /tmp/gcc/gnat-4.3-4.3-20080127/build/gcc/testsuite/ada/acats/acats.log \
> | /tmp/gcc/gnat-4.3-4.3-20080127/build/gcc/testsuite/g++.log \
> | &
> | if [ ! -f /tmp/gcc/gnat-4.3-4.3-20080127/build/gcc/ada/rts/libgnat-4.3.so ]; then \
> | ln -sf libgnat-4.3.so.1 \
> | /tmp/gcc/gnat-4.3-4.3-20080127/build/gcc/ada/rts/libgnat-4.3.so; \
> | fi
> | ln: creating symbolic link `/tmp/gcc/gnat-4.3-4.3-20080127/build/gcc/ada/rts/libgnat-4.3.so': No such file or directory
> | make[1]: *** [stamps/06-check-stamp] Error 1
> | make[1]: Leaving directory `/tmp/gcc/gnat-4.3-4.3-20080127'
>
> The file libgnat-4.3.so is actually in rts-shared-zcx/ instead of rts/
> and I don't understand why.
Because we now provide both the ZCX and SJLJ exception mechanisms,
i.e. two different libgnats (actually three: rts-static-sjlj,
rts-static-zcx, and rts-shared-zcx).
Last I looked the test suite tests /usr/bin/gnatgcc, not the
just-built compiler, so is useless. I disabled the test suite
temporarily. That was one reason for the upload.
> Anyway I don't have time to look further at it today, so the best is
> that you don't wait for me, the fix will be for the upload after. Thanks
> already for having delayed the upload.
OK, I'll merge your patch in and upload that.
--
Ludovic Brenta.
Reply to: