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
|
$NetBSD: patch-af,v 1.4 2013/03/16 18:52:58 gdt Exp $
From Giles Lean, who places it in the public domain. Workarounds for
problems with buggy USB serial adapators. Not applied by upstream
2009-02-21 because it's too dangerous to work around problems without
understanding them.
--- jeeps/gpsread.c.orig 2011-07-26 01:56:33.000000000 +0000
+++ jeeps/gpsread.c
@@ -98,6 +98,36 @@ int32 GPS_Serial_Packet_Read(gpsdevh* fd
GPS_Diag("%02x ", u);
if (!len) {
+ /*
+ * Missed DLE characters have been observed with Geko
+ * 201 and Legend GPSRs with Prolific USB-serial
+ * cables. The following kludge seems to help.
+ *
+ * It has been tested so far with the following
+ * combinations of software and hardware:
+ *
+ * OS X 10.4.x and 10.5.x:
+ *
+ * - a Geko 201 (firmware version 2.70) and a third
+ * party cable using a Prolific USB-serial converter
+ * - a Legend using a serial cable with a non-integral
+ * Prolific USB-serial converter.
+ *
+ * NetBSD-4.0/i386 with the Geko and cable as above.
+ *
+ * REVISIT GFL Should this be a switch in a .ini file?
+ *
+ * + it's a kludge, so leaving it always on is ugly
+ * - if it's harmless to properly working hardware,
+ * then it's a better user experience to leave it on
+ */
+ if (u == 0x06 || u == 0x15)
+ {
+ ++len;
+ (void) fprintf(stderr,"GPS_Packet_Read: inserted DLE due to 0x%02x.\n", u);
+ goto dle_missed;
+ }
+
if (u != DLE) {
(void) fprintf(stderr,"GPS_Packet_Read: No DLE. Data received, but probably not a garmin packet.\n");
(void) fflush(stderr);
@@ -107,6 +137,7 @@ int32 GPS_Serial_Packet_Read(gpsdevh* fd
continue;
}
+dle_missed:
if (len==1) {
(*packet)->type = u;
++len;
@@ -130,6 +161,20 @@ int32 GPS_Serial_Packet_Read(gpsdevh* fd
if (u == ETX)
if (isDLE) {
if (p-(*packet)->data-2 != (*packet)->n) {
+ /*
+ * When used with a buggy Prolific USB-serial converter the
+ * calling sequence GPS_A000() -> GPS_Get_Ack() sometimes
+ * returns data != type causing this routine to fail and the
+ * following error message to be emitted:
+ *
+ * GARMIN:Can't init /dev/cu.usbserial
+ *
+ * Manually retrying usually works, and subsequent
+ * investigation shows that the call that fails is made from
+ * GPS_Init() which in this case is GPS_Serial_Init(). Simply
+ * retrying the call a workaround: see note and retry loop in
+ * gpsapp.c:GPS_Init().
+ */
GPS_Error("GPS_Packet_Read: Bad count");
gps_errno = FRAMING_ERROR;
return 0;
|