Re: Re: [f-cpu] Manual problems and strchr optimized routine

[whygee@club-internet.fr]

hi again,

>De: devik
>
>> > ** logic
>> > There might be explained what is meant by andn. These
>> > can be found at page 54 but is missing in OPs reference page.
>>
>> I don't understand what you mean by "missing in OPs reference page" (I fint it
>> page page 149).  But in fact if you ask to see something like this :
>> r1 = r2 andn r3, I am not sure it's needed.
>
>No I only wanted to see that andn is "r2 and not r3" because
>on i386 it is "not r2 and r3" for example.

huh i'll have to check that with the source.

>> > ** mux
>> > The description doesn't make sense as English sentence to me.
>> > Also from ROP2-YG-2001201.tgz sources it seems that mux is
>> > done bitwise, so that what is a size flag here for ?
>>
>> I understand mux as a SIMD conditionnal move, but it's perhaps an error of
>> interpretation (For the sense problem, I think I forgot to copy/past a part
>> of the original text).
>
>ahh probably. For me MUX makes more sense in bitwise meaning

this is how it is implemented. look at
http://f-cpu.seul.org/whygee/parinux/rop2.gif
and at the source.

> because it is superset of SIMD conditional
> move but more generic and SIMPLER to implement ..

yup. But conditional moves can have specific
conditions (zero, LSB etc)

>devik
YG

*************************************************************
To unsubscribe, send an e-mail to majordomo@seul.org with
unsubscribe f-cpu       in the body. http://f-cpu.seul.org/