I have. I can tell you it required a lot of hacking, most of which I haven't written down.
I can send you my sources if you'd like. They're hardwired for Au1500/Little Endian, but they build ;) Shoudn't take long to change the actual build type.
Gilad.
Clem Taylor wrote:
This isn't completely on topic, but it is a step on the way. I'm getting ready for my Au1550 based hardware that will be back from assembly soon. I'm trying to get the AMD provided yamon source to compile. I'm using gcc 3.4.3 and bintools 2.15.94.0.2. After a few tweeks to the makefile, yamon compiles but fails to link:
mips-ld -G 0 -T ./../link/link_el.xn -o ./yamon-02.23DB1550_el.elf -Map ... mips-ld: section .data [000000009fc3d650 -> 000000009fc40faf] overlaps section .rodata.str1.4 [000000009fc3d650 -> 000000009fc47197] mips-ld: ./yamon-02.23DB1550_el.elf: section .rodata.str1.4 lma 0x9fc3d650 overlaps previous sections mips-ld: ./yamon-02.23DB1550_el.elf: section .rodata.cst4 lma 0x9fc47198 overlaps previous sections
The linker command file (bin/link/link_el.xn) puts .data and .rodata in _etext. I changed the *(.rodata) to *(.rodata*) in the link_el.xn, but that didn't help.
Any ideas what might be going on? Has anyone tried compiling this yamon with a recent gcc/bintools?
Thanks, Clem
