forked from poelzi/OpenChronos
-
Notifications
You must be signed in to change notification settings - Fork 0
/
NEWS
129 lines (94 loc) · 5.98 KB
/
NEWS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
2010-06-20 - OKAY TO FLASH THE WATCH WITH THIS PROJECT
It turns out there was never a problem with the build.
I discovered later that I had a Windows 7 problem that
prevented the BM Firmware Update Tool and also TI's
BSL_Scripter tool from accessing the the USB FET.
I've since rebooted my machine and have repeated flashed
the watch with mspgcc4 project with not problems.
Paul
============ FIXED 2010-06-20 ========================
2010-06-18 - DO NOT FLASH YOUR WATCH WITH THIS BUILD!!
Since flashing the watch with this build a couple of days ago
it has been working fine, until I tried to flash it again with
the USB FET and the BM Firmware Flash Tool , the same tool that
worked before I flashed the watch with this build.
Although I was not able to flash the device with the USB FET
I was able to flash it with the RFBSL. But even after reflashing
it with the factory TI firmware I still could not flash it with
the USB FET.
However, using the IAR Workbench debug process I was able to flash
it with the USB FET, and after flashing it this way with the original
TI firmware, the USB FET capability is back. I am able to flash it
once again with the BM Firmware Flash Tool via the USB FET.
So a little research is in order to determine how to incorporate the
USB BSL into this project.
2010-06-16 (2:30 AM CDT) - IT WORKS!!!
I just completed building the project. I then flashed my watch,
and all the functions I have tested work. Hour and Minutes,
Seconds, Month and Day, Day of the Week, and YES! even sync.
2010-06-15 - Project now compiles and links without errors. This
was accomplished by compiling and linking in separate steps.
Thanks go to Daniel Poelzleitherner for this contribution.
Daniel also contributed Python scripts that convert the build's elf
file to a TI txt file. This function is incorporated into the
makefile and is done automatically at the end of the build.
ITEMS NEEDING TESTING: (as of 2010-06-14
1) four intrinsic functions had to be written.
As of 2010-06-13 they are untested and couple of
them probably don't work yet.
2) bsp_msp430_defs.h - A compiler specific stanza for
MSPgcc was added near line #135. This stanza was copied
from the previous stanza specific to IAR and modified. The
only modifications were to the __bsp_ENABLE_INTERRUPTS__()
and __bsp_DISABLE_INTERRUPTS__() macros. The other macros
may or may not work with MSPgcc. They do compile without errors or
warnings. The one macro I am not sure about, haven't looked
into what it means, is __bsp_QUOTED_PRAGMA__(x)
3) FIXED: (see note on 2010-06-15)
With MSPgcc4 this project compiles cleanly (see minor warning
in "Before you build the first time:") but there are linking errors
"warning: internal error: unsupported relocation error"
BEFORE YOU BUILD THE FIRST TIME:
1) Many of the TI source code files #include <intrinsics.h> MSPgcc does
not come with an intrinsics.h file. Therefore the intrinsics.h included
in this project will need to be placed in a system directory; e.g., in
MSPGCC4\lib\gcc\msp430\4.4.3\include mspgcc4/msp430/include
2) I haven't figured out how to have an empty directory in git so that
clones will be complete. So, before the first build, it is necessary to
create the build directory specified in the makefile.
3) I am using Windows 7, and have experiemented with three different 'make'
utilities - CygWin, MinGW, and GNU make. The Unix based CygWin 'make'
had no problems with a blank in a directory name as long as the blank was
escaped, but one or both of the DOS versions - MinGW and/or GNU make - did
have a problem. To avoid this problem I have added an underscore to the
directory "simpliciti/Applications/application/End Device/". It is now
"simpliciti/Applications/application/End_Device/". And the reference in the
makefile has been adjusted accordingly. Other than this one thing the code
should still compile with either IAR, CCS, or MSPgcc4.
4) In general the source files have not been changed in such a way that the
code will not still compile with either IAR or CCS compilers (except for
the aforementioned adding of an underscore). However, I have not included
the precompiled library files that are need to copile with the code-size
limited versions of IAR and CCS.
5) MSPgcc4's signal.h issues an apparently harmless, but annoying warning,
at line 39. This warning can fill a page of the output from make, making it
difficult to identify the real errors. I commented this line out in my copy
of signal.h. You may want to consider doing the same.
HOW THE PORT WAS DONE:
1) All places where I modified the original code hopefully) are marked with my
initials PFS or pfs. The only actual code changes were to the two area
mentioned above in "Items needing testing:" and the #5 below.
2) bluerobin code had to be elimiated because mspgcc cannot link to the provided
bluerobin library and the source code is not available.
3) All compiler specific areas that accommodated IAR and CCS were enhanced to
accommodate MSPgcc4
4) The Simpliciti *.dat files were converted to *.h files
5) Since I never need to look at my watch to determine the year, but I often
look to my watch to determine the day of the week (I'm retired now), in date.c
I changed the date display to display the day of the week (as a number from 0 to 6).
The original code is still there, just commented out, just one line. BTW,
while doing this I noticed that TI did not use the RTC for the time and date
functions which might account for some of the inaccurate timekeeping that has
been reported.
Paul F. Sehorne
2010-06-14