view loadtools/scripts/pirelli.init @ 853:ae254ffeaec3

AT command interface works! The cause of the breakage was the same Nucleus API issue with NU_Create_Timer() which we encountered at the very beginning of this project with Riviera timers: the code in uartfax.c from TCS211 was passing 0 as the initial dummy value for the timer duration, and our FreeNucleus version doesn't like it. The fix is the same: pass 1 as the initial dummy value instead.
author Space Falcon <falcon@ivan.Harhan.ORG>
date Thu, 30 Apr 2015 01:46:26 +0000
parents 5b53cad88637
children
line wrap: on
line source

# This phone has 3 memory chip selects:
#
# nCS0: flash chip select 1
# nCS1: RAM chip select
# nCS3: flash chip select 2
#
# All 3 chip select lines go to the same physical IC, a RAM/flash MCP.
# We set WS=4 for all 3 here, copying what OsmocomBB does.  The access
# time listed in the datasheet is 70 ns for both RAM and flash, and per
# my math setting WS=3 *might* work, but it could be marginal, so let's
# play it safe for now.

w16 fffffb00 00A4
w16 fffffb02 00A4
w16 fffffb06 00A4

# We also need to switch the CS4/ADD22 pin from its default function
# of CS4 to the needed ADD22.

w16 fffef006 0008

# With this phone all Calypso serial access always goes through the
# CP2102 usb2serial IC inside the phone itself, which is programmed
# to support the high non-standard baud rates.  So we can safely
# switch to 812500 baud unconditionally.

baud 812500