Re: [RFC] Optimize swab operations on mips_r2 cpu

[Franck <vagabon.xyz@gmail.com>]

2006/1/26, Nigel Stephens <nigel@mips.com>:
>
> Then you'll have to have a look at the resulting disassembled code and
> figure what's changed. :)
>
> Thinking about this in more detail:
>
> 1) Using -march=4ksd reduces the cost of a multiply by 1 instruction
> (from 5 to 4 cycles), so a few more constant multiplications, previously
> expanded into a sequence of shifts, adds and subs, may now be replaced
> by a shorter sequence of "li" and "mul" instructions.
>

Is it really specific to 4ksd cpu ? Could this behaviour be triggered
by other options ?

> 2) Enabling branch-likely may allow some instructions to be moved into a
> branch delay slot which previously couldn't be -- but usually these are
> duplicates of the code at the original branch target, so have little
> effect on overall code size.
>
> 3) Using -march=mips32r2 with -O1 and above (but not -Os) enables 64-bit
> alignment of functions and frequently-used branch targets (e.g. loop
> headers); whereas -march=4ksc will not do that. This will add some
> additional "nops" to the code.
>

I noticed your last point when staring at the disassembled code. And
it seems to be ack by these figures:

   text    data     bss     dec     hex filename
2099642  110784   81956 2292382  22fa9e vmlinux-4ksd
2136269  110784   81956 2329009  2389b1 vmlinux-mips32r2
1953086  110784   81956 2145826  20be22 vmlinux-4ksd-Os
1954489  110784   81956 2147229  20c39d vmlinux-mips32r2-Os

I now have to check that your first and second points don't have too
much bad impact on the overall speed although I don't know how to
measure that...But if so, I could safely use -march=mips32r2 -Os
options.

Thanks
--
               Franck