On 08/11/2011 07:32 AM, Mark Brown wrote:
On Thu, Aug 11, 2011 at 05:09:14AM +0200, Lars-Peter Clausen wrote:
On 08/11/2011 04:46 AM, Mark Brown wrote:
None of the current ASoC code will coalesce register writes at all, and in the case where you're doing writes to registers that aren't actually adjacent it's going to be marginal if it's better to transmit the intervening register or transmit another register address. That only really makes a difference during cache sync anyway.
I was think more in terms of in memory consumption and lookup time of the cache compared to a flat cache. If you have two blocks which have a gap of one register between them and that register gets inserted into the cache, ideally those two blocks would be merged, which doesn't seem to be the case currently. So instead of one rbnode with a block covering the whole register space you'll end up with a lot of smaller rbnodes.
My guess is that it's probably not worth worrying about, especially for performance where you mostly just need to be better than physical I/O. For small register maps the memory overhead is similarly probably not worth worrying about, and obviously there's also LZO.
I think I'll just test it with my codec and see how it turns out. If it there are dozens of small blocks, instead of a few larger blocks, I'll see if there is a way to easily improve the situation.
Yes, as I said in one of the earlier messages in this thread. It seems like a good combination of being writable/legible and compact.
Hm, ok I'll give it a try. Though I'm not sure yet how to efficiently implement the default register lookup when syncing the cache.
The caches can just unpack into their data, we need to take a copy anyway to allow the caches to be marked as __initdata and then the data will end up stored in a format that matches the method we're using to store the data.
Dimitris had done an initial version of the move of the cache over, though I didn't review it properly yet and he's on holiday now. I might repost it, there were a few issues but it's at least 90% of the way there IIRC from the time I had to look at it.
ok, that would be great, thanks.
- Lars