[PATCH] Xilinx MicroBlaze support - 4 of 4

Michael Eager eager@eagercon.com
Fri Jan 8 21:09:00 GMT 2010


Richard Henderson wrote:
> On 01/07/2010 01:01 PM, Michael Eager wrote:
>> Which of these changes are required for the patch to be accepted/
>> Xilinx would like to defer some of the cleanup -- such as
>> DImode arithmetic, new constraints, etc.
> 
> DImode arithmetic can definitely be deferred; that was more a suggestion 
> for improvement.  That includes the explicit representation of the carry 
> flag.
> 
> I guess I don't mind deferring conversion to constraints.md so long as 
> it gets done promptly.  We'd like to be able to remove the old method, 
> and I really didn't want any new users added.
> 
> Removing redundant attributes may be deferred.
> 
> Adding initial '*' to otherwise unused named patterns may be deferred, 
> but should be a priority.
> 
> I'd like to see the other bits cleaned up.
> 
> I do think we can make this easier for you by allowing you to post 
> cleanups for gcc/config/microblaze/* relative to your microblaze branch. 
>  Anyone who wants to examine the current state of the submission for 
> mainline can check out the branch.  That seems easier to me than dealing 
> with many large bziped patches.
> 
> Please do CC me on these cleanups.

Sounds like a good plan.  Creating large patches is a pain.
I'll post summaries of the changes to gcc-patches@.


-- 
Michael Eager	 eager@eagercon.com
1960 Park Blvd., Palo Alto, CA 94306  650-325-8077



More information about the Gcc-patches mailing list