On Thu, Jul 13, 2017 at 01:44:10PM +0100, Richard Fitzgerald wrote:
On Thu, 2017-07-13 at 11:05 +0100, Mark Brown wrote:
On Thu, Jul 13, 2017 at 09:02:10AM +0100, Lee Jones wrote:
This patch has been rejected by Linus.
Hrm, when I used to push the register definition patches I did elide all the obviously repeated register banks like the write sequencer one that Linus is calling out there. I'm surprised by the "every single line" bit though...
Linux doesn't say what he wants us to do about it. I could manually strip out a few more definitions but seriously, it makes the code a lot harder to maintain if we can't grep it for use of registers and register fields. They are big chips, they have a lot of stuff in the registers.
I'd take it up with him, if you can explain why it looks very repetitive but isn't actually that'd help...
Building up copies of the repeated IPs with macros would help too (the AIF and especially frame control registers stick out like a sore thumb here), as would removing the individual registers for the write sequencer block.