Re: [PATCH 06/20] arch,avr32: Fold atomic_ops

[Hans-Christian Egtvedt <egtvedt@samfundet.no>]

Around Tue 13 May 2014 22:50:45 +0200 or thereabout, Peter Zijlstra wrote:
> On Tue, May 13, 2014 at 10:40:32PM +0200, Hans-Christian Egtvedt wrote:
>> Probably found the reason why we want to use sub with the signed 21-bit
>> limit, it uses one less register than the add instruction that can add up to
>> 32-bit values.
>> 
>> Both instructions are 32-bit, to use a 16-bit instruction the immediate is
>> very small; 4 bit.
>> 
>> sub 32-bit, type IV, takes a register and subtracts a 21-bit immediate.
>> add 32-bit, type II, adds two register values together.
>> 
>> So by simplifying you loose this optimization.
> 
> OK, let me try if I can come up with anything sane for this.
> 
> Re, adding atomic_{or,and,xor}() those should all use the same bits as
> add, right, except for the special case using sub.

OR and XOR (EOR in the manual) instruction will fall into type II or III,
both 32-bit, and (x)ors two registers together, difference is which way they
can do a shift on one of the registers.

So yes, identical as add.

-- 
Hans-Christian Egtvedt