prpr
Well-Known Member
You've missed the point, which was in the latter part of the post.the Fan package may be scripts rather than MIPS code.
You've missed the point, which was in the latter part of the post.the Fan package may be scripts rather than MIPS code.
I saw that too, different version of MIPS processor like I said. I thought you were pointing out a script based package might work ported from one box to the other.You've missed the point, which was in the latter part of the post.
It's a bit more than that. One is big-endian, the other is little-endian. It's effectively a complete architecture change.different version of MIPS processor like I said.
It might, all other things being equal. But all other things aren't equal. If they were, don't folks think the 'port' would've been done and the package put in the repository?I thought you were pointing out a script based package might work ported from one box to the other.
Oh wow. Well I wrote code that got compiled for MIPS Magnum workstations back in the late 1990s among many other unix variants, and when I asked "why is the MIPS binary twice the size of the binary for any other processor?" work stopped and we spent the afternoon on a deep dive into the stupidities (as I saw it) of the MIPS architecture at the time. The main issue being up to half the binary ends up being NOPs that the compiler has to put in because the processor had no register score boarding. MIPS thought that would be a performance enhancement, and pretty much proved the opposite. If half the code is NOPs, half of the cache is full of NOPs for a start.It's a bit more than that. One is big-endian, the other is little-endian. It's effectively a complete architecture change.