Re: transpiling to low level C

On Mon, 23 Dec 2024 15:41:40 GMT
scott@slp53.sl.home (Scott Lurndal) wrote:

Michael S <already5chosen@yahoo.com> writes:

On Sun, 22 Dec 2024 20:41:44 +0100
Janis Papanagnou <janis_papanagnou+ng@hotmail.com> wrote:
 

 

 
Whether you have the test in an 'if', or in a ternary '?:', or
use it through a bool-int coercion as integer index to an indexed
function[-pointer] table; it's a conditional branch based on the
("Test Set") predicate i<n. You showed in your example how to get
rid of the 'if' symbol, but you could - as expected - not get rid
of the actual test that is the substance of a conditional branch.
 
I think that is what is to expect by the theory and the essence of
the point I tried to make.
 
Janis
 

>
>
You make no sense. I am starting to suspect that the reason for it
is ignorance rather than mere stubbornness.
>
https://godbolt.org/z/EKo5rrYce
Show me conditional branch in the right pane.
 

 
 The 'C' in 'CSET' is short for conditional.  Because
the branch is folded into the compare doesn't mean it
isn't there.

No, branch is not "folded". It is absent. CSET is an ALU operation.
The logical-arithmetic nature of comparison operator is even more
pronounced in code that gcc generates for POWER
https://godbolt.org/z/8Gs9s6nEo