RFA: 4.2 reload infrastructure to fix PR target/21623

[Bernd Schmidt]

amylaar@spamcop.net wrote:
> Quoting Bernd Schmidt <bernds_cb1@t-online.de>:
> 
> 
> 
>>>>Not sure I agree with the comments about recursion. Isn't there a mismatch
>>>>between what targetm.secondary_reload_class wants (an rtx) and what we can
>>
>>give
>>
>>>>it (an intermediate class and mode)?
>>
>>Any comments?
> 
> 
> We still use the same rtx x.  The reg class of the original reload is replaced
> with the reg class of the intermediate register.

Wouldn't you need to have the hook decide how to move from a register in 
the intermediate class to a register in the reload reg class?  In which 
case you don't have an rtx to look at.  Am I being dense?

Not much of a point discussing the rest until one of us realizes he's 
missing something and we're on the same page again...


Bernd