[ell-i-developers] Re: ENCX24J600 driver status
- From: Jose Granados Vergara <jose.granados@xxxxxx>
- To: "ell-i-developers@xxxxxxxxxxxxx" <ell-i-developers@xxxxxxxxxxxxx>
- Date: Tue, 04 Nov 2014 15:23:13 +0000
Here is an initialization procedure I found for reference:
https://github.com/bootc/PolyController/blob/master/drivers/enc424j600.c#L73
-Jose
________________________________
De: ell-i-developers-bounce@xxxxxxxxxxxxx
[ell-i-developers-bounce@xxxxxxxxxxxxx] en nombre de Pekka Nikander
[pekka.nikander@xxxxxx]
Enviado: martes, 04 de noviembre de 2014 05:11 p.m.
Para: ell-i-developers@xxxxxxxxxxxxx
Asunto: [ell-i-developers] Re: ENCX24J600 driver status
Hi Asif,
Yes, you are on the right track. E_STAT low byte reading is indeed returning a
zero; I've verified that with SPI protocol analyser. I suspect there is some
initialisation problem. The next step would be to carefully read through what
we initialise in the registers, and in availablepackets probably read more
registers to see what is going on. ALso in receive to analyse the status bits.
--Pekka
On 2014–11–04, at 16:48 , Asif Sardar
<engr.asif.sardar@xxxxxxxxxxxxxx<mailto:engr.asif.sardar@xxxxxxxxxxxxxx>> wrote:
Hi all,
I have been reading the code for SPI and ENCX24J600 in the feature-runtime-wip.
To my understanding the ENCX24J600Class::availablePackets() in
encx24j600_packet.h is not returning correct packet count value in E_STAT
address read through SPI. That may be the only reason why its not reading
furthermore packets. Is it possible to test the value returned from lower
8-bits of E_STAT by writing it through serial port? There might be
configuration problems while initializing the registers for encx24j600.
On Thu, Oct 30, 2014 at 12:21 PM, Pekka Nikander
<pekka.nikander@xxxxxx<mailto:pekka.nikander@xxxxxx>> wrote:
We did some further joint work on the ENCX24J600 driver with Teemu and Otso:
https://github.com/Ell-i/Runtime/commits/feature-encX24J600-wip
The current status is as follows:
1. The test_ENCX24J600_send test case sends out packets correctly
2. The test_ENCX24J600_receive receives one packet, apparently correctly.
3. The test_CoAP_no_threads receives one packet, and apparently passes it
correctly to ARP, but ARP does not seem to respond for some reason.
The current blocking problem is that once the code has received one packet,
ENCX24J600.availablePackets no longer reports any more packets. In an earlier
version, before fixing the location where we were reading the packets from, the
code did apparently read a packet each and every time there was a packet:
https://github.com/Ell-i/Runtime/commit/3e58adbfdf2aeecb5e5b9b99c25c8000b8ad3813
So, somehow the later changes, while fixing the bug of reading wrong data,
caused a new bug where any packets beyond the first one are not read any more.
--Pekka
--
With Best Regards,
Asif Sardar.
+358 43 8265795
Other related posts:
