BSP Updates – NET+OS 6.3

For FSF CC9P9360

Last Updated: 07/12/07 Fix Count: 6

 
Title
I2C errors using one wire I2C device on the bus
 
Case: 22063
 
Date Fixed: 06/18/07




Description
One wire bus buster on I2C bus can lose data.
 
Solution
 I2C interrupt is disabled until ISR is installed completely
 
 Default spike filter value is set to 7 
 
 Write queue count and I2C status are changed after updating the register (initially it was done before) to
 prevent possible race condition. The reason is, with the old code, I2C ISR can read these values before register
 update and could mess up the state machine 
 
 Use of volatile data to prevent possible code omission due to optimization 
 ________________________________________________________________________________________________________________________
 
Title
I2C error when more than one device is on the bus
 
Case: 1231908
 
Date Fixed: 06/14/07




Description
When using more than one device on the i2c bus, and NET+OS does a read from the first device, followed by a write to the other device,
then a read from the first device again, the second read has the incorrect i2c device address placed on the bus.
 
Solution
Fixed problem in NAI2CKickStartRx().
 

 
Title
Update for DST
 
Case: 22262
 
Date Fixed: 06/05/07
 
Description
DST Update
 
Summary 
Made modification to select DST by Time Zone.
 

 
Title
Serial Channel C transmitter DMA stalls
 
Case: 1230281
 
Date Fixed: 05/1407




Description
DMA stall causes problems with IM display
 
Solution
The DMA CE bit for TX should always be re-set to a “1” regardless of whether it was a "1" already.
__________________________________________________________________________________________________________________________
 
Title
Data corruption while accessing NVRAM via I2C from multiple threads
 
Case: 1224835
 
Date Fixed: 04/13/07




Description
Data corruption was possible if accessing NVRAM from multiple threads.
 
Solution
Replaced 2 calls to NAI2COperation() with one in NANVRead() to make it atomic to avoid data corruption in multiple thread access condition.
__________________________________________________________________________________________________________________________
 
Title
CC9P9360 hangs on reboot while connected to 10BaseT hub, or with cable unattached
 
Case: 21982
 
Date Fixed: 04/02/07




Description
When rebooting the CC9P9360 modules on a 10BaseT network, or with no Ethernet cable attached, the module could hang up.
 
Solution
eth_init_isr was moved from within eth_init to within eth_enable, which succeeds in enabling the interrupts only after the PHY chip has been set up.
 

Files:

netos\h\gpiomux_def.h

netos\src\bsp\devices\common\ethernet\eth_isr.c
netos\src\bsp\devices\common\ethernet\eth_init.c

netos\src\bsp\platforms\fsf_cc9p9360\bsp.c

netos\src\bsp\platforms\fsf_cc9p9360\boardparams.c

netos\src\bsp\devices\common\serial\mercury_serl.c

netos\src\bsp\devices\ns9xxx\common\i2c\i2c_api.c

netos\src\bsp\devices\ns9xxx\common\i2c\i2c_driver.c

netos\src\bsp\devices\ns9xxx\common\i2c\i2c_driver.h

netos\src\bsp\devices\ns9xxx\ns9360\gpio\gpio.c

Special Instructions

• Unzip the patch(es) to the root of your NET+OS installation, for example C:\netos63_gnu\.
• Be sure to install any patches listed under Dependencies below
• Clean and rebuild your BSP.
• Rebuild your application.

Patch Link: FSF_BSPUpdates_63

Dependencies

This patch also requires the installation of the following patch(es):

NA