Re: SPI rework

From: Aleksandr Rybalko <ray_at_dlink.ua>
Date: Tue, 18 Oct 2011 12:15:40 +0300
On Tue, 18 Oct 2011 06:17:46 +0800
Adrian Chadd <adrian_at_freebsd.org> wrote:

>> Hi,
>> 
>> That sounds logical to me. Maybe getting it done before 9.0-RELEASE
>> is a bit rushed?
>> 
>> I can help you test out the flash side of things on my atheros SoC
>> boards.
>> 
>> 
>> Adrian

More thinking give me a better way to fix that.

We leave same structure, but remove KASSERT(cmd->tx_cmd_sz ==
cmd->rx_cmd_sz,""); and fix SPI controllers drivers to care about
possible different sizes.

And of course will fix consumers to set exact what they need.
(dev/flash: cmd_tx_sz=1, data_rx_sz=3 for "device ID" call)

That way better because we will have ability to duplex SPI transfers on
controllers that able to do that, RT305x SPI will return error in case
when sizes specified for both directions.

Comments?

WBW
-- 
Alexandr Rybalko <ray_at_dlink.ua> 
aka Alex RAY <ray_at_ddteam.net>
Received on Tue Oct 18 2011 - 07:34:24 UTC

This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:40:19 UTC