Re: [PATCH] cse (canon_reg): Don't modify insns without using validate_change

[Bernd Schmidt <bernds_cb1 at t-online dot de>]

> > In your previous mail you mentioned the need to call apply_change_group 
> > after canon_reg; I see places in cse.c where that isn't done.  Why don't
> > we need to change them?
> The cases I was concerned about where cases where either INSN is NULL_RTX 
> or apply_change_group is called later on. According to the comment above 
> canon_reg the function must not call validate_change if INSN isn't set.
> That's why I've left the check for INSN != 0 in validate_canon_reg. It
> should be ok with this.

I'm not entirely sure what to make of the code that's processing 
CALL_INSN_FUNCTION_USAGE, but that function ends up in an unconditional 
call to apply_change_group so I guess it's ok.  The other callers of 
canon_reg look safe as well.

> 2006-05-30  Andreas Krebbel  <krebbel1@de.ibm.com>
>
> 	* cse.c (validate_canon_reg, cse_insn): Don't change insns without
> 	calling recog.

Ok.


Bernd