Again the doubt is on powerpc-eabi cross toolchain with newlib. Although I built the tool chain, I'm facing problem in using the cross built tools. I wrote a simple hello world test program and compiled. The following errors are issued from linker: powerpc-eabi-ld: warning: cannot find entry symbol _start; defaulting to 01800074 libgcc.a(eabi.o): In function `__eabi': eabi.S:232: undefined reference to `__init' libgcc.a(eabi.o):(.got2+0x8): undefined reference to `__SDATA_START__' libgcc.a(eabi.o):(.got2+0xc): undefined reference to `__SBSS_END__' libgcc.a(eabi.o):(.got2+0x14): undefined reference to `__SDATA2_START__' ... ... libgcc.a(eabi.o):(.got2+0x50): undefined reference to `__EXCEPT_END__' collect2: ld returned 1 exit status If I add -msim or -mads, it is compiled and linked succesfully. I understand that -msim should be added only if my application is going to run on any simulator, not on real hardware. Here startup module sim-crt0.o is called and the standard C libraries are ‘libsim.a’ and ‘libc.a’. I'm not able to understand what '-mads' is for. Could any one please beief this? Moreover, when I invoked the compiler and linker separately, linker issues error. powerpc-eabi-gcc -c test.c -> test.o is generated. I tried the following variant options for linker. powerpc-eabi-ld test.o powerpc-eabi-ld /path_to_startup/crt0.o test.o -lc All resulted in error as above. what are all the extra options I should supply to linker? Or is it the problem in building of my tool chain? Can't the linker automatically look for standard libraries like libc.a, libm.c and the startup module crt0.o? I used --target=powerpc-eabi to build my tool chain. Should I use specific target like powerpc-eabisim (for simulator)? My target is an embedded PowerPC processor with non-linux OS running in big-endian mode. It will be a great help if I you give solution for this. Thanks a lot. Jeffi -- View this message in context: http://www.nabble.com/Problem-in-using-powerpc-eabi-cross-linker...-tp26047249p26047249.html Sent from the gcc - Help mailing list archive at Nabble.com.
