Re: [ts-7000] 7260 SPI C Programming Example

["Breton M. Saunders" <>]

On 07/30/2010 01:41 PM, Peter Gammie wrote:

On Fri, Jul 30, 2010 at 8:11 PM, Breton M. Saunders <m("ntlworld.com","breton.saunders");">> wrote:

It isn't the code style that worries me.

It is the myth that one can write a reliable driver (real-time or not) as a user mode program that concerns me greatly.

There is a reason why real drivers run inside the kernel, and do not use peekpoke.h!

I'll mildly disagree with this - using Linux 2.6.33.x and the -RT patches and high-res timers, I have a rock-solid flicker-free multiplexed display hanging off the LCD port of a ts7250, being driven from entirely user space, using mmap'd I/O. By "rock solid" I mean while playing mp3s, dropbear sessions, etc. The scheduling overhead is a bit of a killer, I grant...

... but I have to wonder if the OP can use the kernel's SPI drivers... it is not hard to get a recent kernel working on the ts72xx boards, just consult the archives.

I don't know what the specific of your application are; sounds great that it works!

In my case I am running a 1khz interrupt servicing an IMU board that I designed.
Each cycle requires servicing of an A/D converter channel or SPI attached peripheral, and dumping of the results into a fifo, with time stamps.

I have yet to move it to hard realtime using the FIQ patches. The main reason to move to FIQ is to remove the latency introduced by servicing standard kernel operations.  There is absolutely no way I could do this as a user mode program. 

    -Brett

__._,_.___

Your email settings: Individual Email|Traditional
Change settings via the Web (Yahoo! ID required)
Change settings via email: =Email Delivery: Digest | m("yahoogroups.com?subject","ts-7000-fullfeatured");=Change Delivery Format: Fully Featured">Switch to Fully Featured
Visit Your Group | Yahoo! Groups Terms of Use | =Unsubscribe

__,_._,___