Re: Help with reloading

[Jeff Law <law at redhat dot com>]

On 12/20/10 07:33, Mohamed Shafi wrote:
>     I didn't know that you could do that in TARGET_SECONDARY_RELOAD
> hook. Can you point me to some target that does this - figuring out
> what the destination is based on the source or vice versa. In my case
> only the address operand comes into TARGET_SECONDARY_RELOAD hook
> during the reload pass. I am not sure how to find out the source for
> the pattern which has this particular address as the destination.
The vast majority of targets define secondary reloads.  You'd have to 
scan through them to see if one does precisely what you want -- I don't 
know if one will do precisely what you want, but there should be enough 
examples that you can learn from to implement what your target needs.

The mn103 port is pretty simple and defines a few well understood 
secondary reloads.

The other, less desirable alternative would be to disallow constant 
stores to memory completely in the insn predicate, then have a peephole2 
to catch cases where you have a constant load into a reg followed by a 
store of that reg to a symbolic memory location and the reg dies.

jeff