From: Arthur O. <art...@gm...> - 2003-04-20 17:59:53
|
Hello all, The said toolchain built seamlessly first time round. However, any attempts to rebuild result in the build failing somewhere around [gcc_initial] - I can't say for sure because at one point it failed at an earlier stage, somewhere at [binutils]. The thing is this phenomenon is purely random. Sometimes the toolchain builds without a hitch, most of the time it fails at the exact same point ([gcc_initial]). This line caught my attention though: /opt/dclinux/src/uClibc-builder-gcc-3.2.2/build/gcc-3.2.2/gcc/gthr-posix.h:37:21: pthread.h: No such file or directory And a load of compiler warnings after that. (Insists pthread.h is absent when actually it isn't, hrmm) Also, Christian Berger managed to compile (but not test natively on DC) the traditional hello.c and busybox-0.60.5 with this toolchain, but not after a simple hack to get it working. Basically what he did was (assuming the toolchain built and "installed" correctly): (CWD=$APPS/sh4/lib/gcc-lib/sh-linux/3.2.2) $ cp -R ../../../* . $ cp m4/libgcc.a . He suggested that the toolchain didn't install right, and that there was probably a better way on getting around this. Maybe you can comment more on this Christian? :) Adrian McMenamin has been so kind to host the error log at http://195.157.85.85/error.log Thanks Adrian ;) Hopefully someone will make something out of this, Regards, Arthur. -- Linux is a true multitasking system. Are you? |