• State
  • Locked Locked
  • Replies 1 reply
  • Subscribers 99 subscribers
  • Views 1971 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

GPIO, IRQ & interrupt latency

NickAllen
Intellectual 525 points
Other Parts Discussed in Thread: CC1101

Dear all,

I am attempting to investigate the difficulty / feasibility of interfacing the DM365: 

  • directly (via SPI) to a CC1101  v.s. 
  • (via SPI) to a dedicated micro (eg MSP430 / PIC24) which in turn interfaces  (via SPI) to the CC1101. 

The ultimate would be the  former. In this regard however; my concerns (due to inexperience) are the following:  

  • How to configure a GPIO as an input with a neg interrupt on change.
  • How to access this IRQ from userland
  • The actual interrupt latency on a GPIO pin (is this sufficient / reliable enough to service the chipcon?!).
  • The difficulty in writing an SPI device driver for the CC1101 (rather vague question, apologies)

We are evaluating on the DM365 DVEVM with Davinci PSP 03.01 & 2.6.32-rc2.

I would really appreciate if  someone could point me in the right direction regarding recommended reading to get to grips with the above.

Many thanks,
Nick

  • 0
    Expert 2315 points

    Hi Nick,

    I can maybe help you with less than half of what I think you want to do by telling you I did on the EVMDM365 to get an GPIO going.  I used the gpio0_test evaluation code that came with our dvsdk_2_10_01_18 (you should see some thread on the forum where we discussed and solutions published).  I added to irqk.c in the linuxutils_2_24_02 my own functions (modified from gpio0_test) and added to the ioctl to support in userland.  This worked fine but I do not exactly know what the latency is but it is OK for me to sync to a 1Hz event.  I also added another isr to service a VDINT1 event which I use as a 50Hz event from PAL timing to have synchronisation in my system and also added ioctl to support this through to userland.  Seems like I am getting close to a few microseconds relative timing from experiments I am busy with.  Let me know if you need more info if you think I can be of help.

    Good luck, Jinh T.