Re: [gpsd-dev] Major progress!

[Frank Nicholas]

On Feb 10, 2015, at 6:03 AM, Eric S. Raymond <address@hidden> wrote:

Hal Murray <address@hidden>:

I'm having troubles on NetBSD and FreeBSD.

FreeBSD gets typical lost sentence problems.  So I changed WRITE_PAD from
0.001 to 0.01

Alas.  I was hoping we could zero out WRITE_PAD everywhere.  What is
sys.platform on that system?

Then it hung on the second or 3rd test.  It seems easy to reproduce.

top says:
 PID USERNAME  THR PRI NICE   SIZE    RES STATE   C   TIME   WCPU COMMAND
9021 murray      1 103    0 16116K  8216K CPU1    0   4:48 100.00% python

This morning, NetBSD also hung.  I haven't seen that this evening.

Those sound like they might be the same problem Greg Troxel reported
from NetBSD.  He was able to pin it to "CLOSE_DELAY is gone.  Test source
termination is now deterministic."


The only non-test-frame issue still pending is newdata.time not being
set in gpsmon.  I'll tackle that today.
--
<a href=""http://www.catb.org/~esr/" class="">http://www.catb.org/~esr/">Eric S. Raymond</a>

Same type of hang issue on Mac OS X 10.10 (darwin platform).  Hangs on first test:

./gpsctl -R 2>/dev/null; ./gpsfake -T; ./regress-driver test/daemon/*.log

sys darwin platform Darwin-14.1.0-x86_64-i386-64bit: WRITE_PAD = 0.03000

Testing the daemon...

Processing test/daemon/GPSmap-76S.log

^CTraceback (most recent call last):

  File "/Users/nicholfd/Desktop/Work/gpsd/gpsfake", line 211, in <module>

    test.run()

  File "/Users/nicholfd/Desktop/Work/gpsd/gps/fake.py", line 636, in run

    self.reporter(chosen.response)

IOError: [Errno 32] Broken pipe

scons: Build interrupted.

scons: building terminated because of errors.

scons: writing .sconsign file.

Processes: 286 total, 3 running, 16 stuck, 267 sleeping, 1544 threads                                            12:18:35

Load Avg: 1.70, 1.63, 1.54  CPU usage: 10.61% user, 4.95% sys, 84.43% idle

SharedLibs: 18M resident, 16M data, 0B linkedit. MemRegions: 433403 total, 4452M resident, 86M private, 634M shared.

PhysMem: 15G used (2038M wired), 897M unused.

VM: 811G vsize, 1062M framework vsize, 17075063(0) swapins, 19464701(0) swapouts.

Networks: packets: 28860815/24G in, 29379632/18G out. Disks: 7159161/178G read, 6122341/468G written.

PID    COMMAND      %CPU  TIME     #TH   #WQ  #PORT MEM    PURG   CMPRS  PGRP  PPID  STATE    BOOSTS            %CPU_ME

22577  Python       99.7  03:37.17 1/1   0    13    5836K  0B     0B     22164 22567 running  *0[1]             0.00000

And on the Gentoo system that fails completely, it is hanging on the first test, also:

./gpsctl -R 2>/dev/null; ./gpsfake -T; ./regress-driver test/daemon/*.log

sys linux2 platform address@hidden: WRITE_PAD = 0.00000

Testing the daemon...

Processing test/daemon/GPSmap-76S.log

gpsd:ERROR: can't create IPv6 socket

gpsd:ERROR: device open failed: Permission denied - retrying read-only

gpsd:ERROR: read-only device open failed: Permission denied

gpsd:ERROR: /dev/pts/2: device activation failed.

gpsd:ERROR: device open failed: Permission denied - retrying read-only

gpsd:ERROR: read-only device open failed: Permission denied

gpsd:ERROR: /dev/pts/2: device activation failed.

top - 12:25:10 up 6 days, 23:30,  2 users,  load average: 0.80, 0.29, 0.14

Tasks: 106 total,   3 running, 103 sleeping,   0 stopped,   0 zombie

%Cpu(s): 23.1 us,  2.2 sy,  0.0 ni, 74.6 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st

KiB Mem : 12310972 total, 11671152 free,   118100 used,   521720 buff/cache

KiB Swap:  1048572 total,  1048572 free,        0 used. 12126040 avail Mem 

  PID USER      PR  NI    VIRT    RES    SHR S  %CPU %MEM     TIME+ COMMAND                                              

30044 root      20   0   82672  11464   5868 R 100.0  0.1   1:18.03 gpsfake