Re: regclass.c: scan_one_insn special casing

[Andrew Pinski <pinskia at physics dot uc dot edu>]

On Mar 29, 2004, at 00:59, Michael Matz wrote:

> Hi,
>
> On Sat, 27 Mar 2004, Eric Christopher wrote:
>
> > There's this code in scan_one_insn:
> >
> >
> > That appears to do something a port specific splitter would do.

I had looked into the history of this code and it was added before
define_splits were even in GCC.


> I disagree.  It's in the generic code specifically that _no_ port 
> specific
> splitters have to do this.  If there are splitters just doing this they
> should be removed.  And if they do something more it should be analyzed
> why they are doing this and it should be considered to add this to the
> generic code in order to enable the removing of the port specific 
> hacks.

The reason why it is generic code not specific for any port splitter is
because it was added before splitters were added so there was no way,
for port splitters to have been added.  It was in GCC 1.21 which
was before define_split was added (and before i386 was even added to
GCC, also).


> Well, the comment is quite clear.  For all two address machines.  I 
> would
> be more interested to know why port specifics doing the same (or 
> something
> similar) were added.

I would be interested too because this code has been there for a long 
time
before splitters were even thought of.  (The only thing I can think of 
is
they were added when the port was being designed as they did not know 
this
code existed).

Some times getting rid of code like this is the right thing to do as it
would make the compiler fast for target which will never invoke this
code at all.

Thanks,
Andrew Pinski