Re: [avr-gcc-list] Transfer bit conserved by gcc?

[Andy H]

Gcc only tests certain conditions and requires zero, carry and overflow 
sign flags to be set usefully (ie indicating result). The real AVR flag 
status is translated to a similar internal flag set within gcc.

It does not use H flag or T flag.  But certain sequences emitted by 
backend to synthesize gcc instruction might use any flag internally.

Each instruction template written defines what happens to CC_status - 
for example, "clobber" means instruction sequence leaves no useful 
information and "none" means sequence has no effect on flags. Gcc tracks 
this to see if explicit tests on results can be omitted.

You may note that gcc cant deal with sign flag only. So many instruction 
sequence results cant be tested for plus/minus - which causes otherwise 
redundant tests to be left in.

Andy





Ruud Vlaming wrote:
> On Sunday 21 September 2008 22:22, you wrote:
>
>   
> > T-bit is used by avr-gcc for internal purpose, the same way as temp reg (r0)
> > and cannot be used for storing value.
> >     
> I see. I looked at it and i cannot get clear in which circumstances,
> this is, but it does not really matter i suppose. Point is, you cannot destroy
> the value of the Tbit at your liking, although i never saw the Tbit being used 
> in my code by gcc.
>
> And how about the H flag? The Half carry is produced by a lot of instructions,
> but i never saw the compiler test for it,or make use of it. I looked the avr 
> directory over for H flag tests or direct H flag manipulating functions 
> (seh, clh, brhc, brhs) and found none.
>
> You would really help me if you know more on this.
>
> Ruud.
>
>
> _______________________________________________
> AVR-GCC-list mailing list
> address@hidden
> http://lists.nongnu.org/mailman/listinfo/avr-gcc-list
>