gpsd-dev
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Lexer issue? was Re: Problems with miss-behaving uBlox Module


From: Gary E. Miller
Subject: Re: Lexer issue? was Re: Problems with miss-behaving uBlox Module
Date: Tue, 30 Mar 2021 11:10:15 -0700

Yo James!

On Tue, 30 Mar 2021 08:00:00 -0700
James Browning <jamesb.fe80@gmail.com> wrote:

> > Nothing verified until I have a data sample that used to fail.  
> 
> 
> Attached is a sample where a malformed ublox packet prevents the
> detection of several later ublox packets. Removing the first byte via
> dd allows the construed test (gpscat -t) to detect the packets. This
> was extracted from the first email in the chain and reformated for
> gpscat.

Thanks. That shows a problem.

This is what ubxtool thinks:

----
# ubxtool -f  construed.log 
gps/ubx: ERROR checksum failed,was (243,187) s/b (8, 25)

UBX-NAV-PVT:
  Poll request

UBX-NAV-STATUS:
  iTOW 134433600 gpsFix 3 flags 0xdf fixStat 0x0 flags2 0xc
  ttff 20362, msss 335956

UBX-NAV-POSECEF:
  iTOW 134433600 ecefX 407326387 Y 116232443 Z 475298407 pAcc 50

UBX-NAV-DOP:
  iTOW 134433600 gDOP 124 pDOP 111 tDOP 55 vDOP 88
  hDOP 67 nDOP 48 eDOP 47

UBX-NAV-VELECEF:
  iTOW 134433600 ecef: VX -893 VY -930 VZ 1088 vAcc:11

UBX-NAV-TIMEGPS:
  iTOW 134433600 fTOW 411147 week 2151 leapS 18 valid x7 tAcc 48
------

Bad checksums are most often caused by bad serial cables or serial
voltage mis-match.

I fed the file to the regression-driver.  Without the patch it just fails

# ./regress-driver -b /tmp/construed.log 
Rebuilding daemon regressions...
Processing /tmp/construed.log
gpsfake: zero-length packet from /tmp/construed.log
Elapsed time: 1.90


With the patch it looks better:

# ./regress-driver -b /tmp/construed.log 
Rebuilding daemon regressions...
Processing /tmp/construed.log
Elapsed time: 2.53

But /tmp/construed.log.chk is empty. I need to look more closely.  The
fragment is too short to be fully helpful.  The packet engine has not
even identified the device as u-blox before the failure point.  I'll run
some more tests with it.

RGDS
GARY
---------------------------------------------------------------------------
Gary E. Miller Rellim 109 NW Wilmington Ave., Suite E, Bend, OR 97703
        gem@rellim.com  Tel:+1 541 382 8588

            Veritas liberabit vos. -- Quid est veritas?
    "If you can't measure it, you can't improve it." - Lord Kelvin

Attachment: pgpLmqiHGpR7W.pgp
Description: OpenPGP digital signature


reply via email to

[Prev in Thread] Current Thread [Next in Thread]