Minor updates to MRF24J40 on SAME70 Xplained; Updated to README files.

This commit is contained in:
Gregory Nutt 2017-08-16 12:52:58 -06:00
parent 06473e89de
commit adef84b68a
4 changed files with 35 additions and 11 deletions

View File

@ -893,6 +893,11 @@ Click Shield
D12 PD20 (both) SPI-MISO PD20 SPI0_MISO GPIO_SPI0_MISO
D13 PD22 (both) SPI-SCK PD22 SPI0_SPCK GPIO_SPI0_SPCK
NOTE: The click shield fits close to the Arduino connect and cannot be
installed directly because it hits the RJ45 connector. You have to get
some extra Arduino connectors to raise the Click shield so that it clears
the RJ45.
Tickless OS
===========
@ -1289,10 +1294,11 @@ Configuration sub-directories
disable this feature and save 8KiB of RAM!
NOTE: There is an issue with capturing data in the RAMLOG: If
the system crashes, all of the crash dump information will go into
the RAMLOG and you will be unable to access it! You can tell that
the system has crashed because (a) it will be unresponsive and (b)
the LD2 will be blinking at about 2Hz.
the system crashes or hangs, then all of the crash dump/error
information will go into the RAMLOG and you will be unable to
access it! You can tell that the system has crashed because
(a) it will be unresponsive and (b) the LD2 may be blinking at
about 2Hz.
You can also reconfigure to use stdout for debug output be disabling
all of the CONFIG_RAMLOG* settings listed above and enabling the
@ -1324,8 +1330,16 @@ Configuration sub-directories
Where <server-ip> is the IP address of either the E1 or I2 endpoints.
STATUS:
2017-08-16: Configurations added. Currently hangs in mrf24j40_reset()
before the NSH appears on the serial console.
2017-08-16: Configurations added. Initially, I saw hangs i
mrf24j40_reset() before the NSH appears on the serial console.
Unlike the SAMV71-XULT, the SPI looks clean, but was hanging
nevertheless.
Then, on subsequent testing, it "magically" started behaving
properaly and seems quite stable now.. although I did nothing to
the the problem. Perhaps the radio was in a bad state for awhile;
perhaps something I did masked the problem. However, all is well
for the time being.
netnsh:

View File

@ -80,7 +80,6 @@ void sam_spidev_initialize(void)
/* Enable chip select for mikroBUS2 */
(void)sam_configgpio(CLICK_MB2_CS);
#endif
#endif

View File

@ -368,9 +368,6 @@
#define IRQ_MB1 SAM_IRQ_PA5
#define IRQ_MB2 SAM_IRQ_PA6
#define MB1_CSNO SPI0_CS0 /* REVISIT PA5 is not one of the NPCS pins */
#define MB2_CSNO SPI0_CS1 /* REVISIT PA6 is not one of the NPCS pins */
/* SP chip selects */
#define CLICK_MB1_CS (GPIO_OUTPUT | GPIO_CFG_DEFAULT | GPIO_OUTPUT_SET | \
@ -378,6 +375,9 @@
#define CLICK_MB2_CS (GPIO_OUTPUT | GPIO_CFG_DEFAULT | GPIO_OUTPUT_SET | \
GPIO_PORT_PIOC | GPIO_PIN9)
#define MB1_CSNO SPI0_CS1 /* GPIO_SPI0_NPCS1_1 */
#define MB2_CSNO SPI0_CS0 /* REVISIT PC9 is not one of the NPCS pins */
/************************************************************************************
* Public Types
************************************************************************************/

View File

@ -1937,14 +1937,25 @@ Configuration sub-directories
STATUS:
2017-07-02: Configurations added. Not yet tested.
2017-07-03: Initial testing, appears to be working, but endpoints
fail to associate; sniffer shows that nothing sent fro the star
hub. I am thinking that there is something wrong with the
GPIO interrupt configuration so that no MRF24J40 interrupt are
being received.
2017-08-15: I think the GPIO interrupts are fixed but there still
2017-08-15: I think the GPIO interrupts are fixed but there still
seems to be some issue with the SPI communications.
2017-08-16: I believe that there is something interfering with the
MRF24J40 on the SPI0. There are other things on the bus. The
MRF24J40 requires sole use of the SPI bus because it holds MISO
low when not selected.
I successfully brought the same logic up on the SAME70-Xplained.
The SPI signals look clean on the board and the MRF24J40 seems
fully functional.
mxtxplnd:
Configures the NuttShell (nsh) located at examples/nsh. There are five