Mac silently crashing, redux

Support Client software
1

I’ve been seeing a lot of silent crashes on Mac recently. Sometimes there are crash logs in ~/Library/Logs/DiagnosticReports, sometimes not. It never creates an error dialog, though, nor have I seen a Crash Reporter window for any but one crash. Will probably reboot later tonight to see if that helps, but if it doesn’t what information can I provide to help debug this?

In the hope that they’ll be useful, I’m attaching the last 18 hours or so of diagnostic logs I could find. (Edit: Replaced attachment; WhatPulse crashed again while I was typing this.)

I’ve some motivation, as I’ve definitely lost many thousands of keys and lots of network traffic stats to this problem over the last few days.

OS X 10.10.5, MacBookPro8,1 (Early 2011, 13") Core i5 / 8GB RAM, WhatPulse 2.6.3

2

Fresh reboot, still crashing. The launchd job I’m using to keep WhatPulse running does not prevent the app from briefly stealing focus when it’s restarted, which oddly often also results in “ghost” clicks/keys being output to whatever application happened to be in the foreground—it’s always a snippet of the last maybe 10-15 or so input actions. Gotten a few questions wrong on my Japanese SRS because of poorly timed crashes.

Another log ZIP for you, this one containing all 59 reports in the above-mentioned directory. There may be some duplicates, but I will move the log files now so only new ones are submitted in the future.

I hope there’s at least a single cause for all this…

3

Over a week and multiple reboots later, still crashing frequently. More logs attached.

Could any of these Homebrew packages be causing conflicts?

[quote=brew list --versions]autoconf 2.69
automake 1.15
boost 1.59.0
brew-cask 0.59.0
cairo 1.14.4
cloog018 0.18.0
ffmpeg 2.8.1_1
flac 1.3.1
fontconfig 2.11.1
fortune 9708
freetype 2.6_1
fribidi 0.19.7
gcc49 4.9.3
gcc5 5.2.0
gdbm 1.11
geoip 1.6.7
gettext 0.19.6
git 2.6.3
git-flow 0.4.1
glib 2.46.2
gmp 6.0.0a
gmp4 4.3.2
go 1.5.1
gobject-introspection 1.46.0
harfbuzz 1.0.4
heroku-toolbelt 3.42.20
icu4c 55.1
isl011 0.11.1
isl014 0.14
jpeg 8d
lame 3.99.5
libass 0.13.0
libebml 1.3.3
libffi 3.0.13
libmagic 5.25
libmatroska 1.4.4
libmpc 1.0.3
libmpc08 0.8.1
libogg 1.3.2
libpng 1.6.18
libtiff 4.0.6
libvo-aacenc 0.1.3
libvorbis 1.3.5
little-cms2 2.7
lua 5.2.4_1
mad 0.15.1b
mkvtoolnix 8.5.2
mobile-shell 1.2.5
mpfr 3.1.3
mpfr2 2.4.2
mpg123 1.22.4
mpv 0.13.0
nmap 6.47
node 5.0.0
openssl 1.0.2d_1
pidof 0.1.4
pixman 0.32.8
pkg-config 0.29
protobuf 2.6.1
python 2.7.10_1 2.7.10_2
python3 3.5.0
readline 6.3.8
sox 14.4.2
speedtest_cli 0.3.2
sqlite 3.9.2
tree 1.7.0
wget 1.16.3
x264 r2601
xdelta 3.0.10
xvid 1.3.4
xz 5.2.2
youtube-dl 2015.11.13[/quote]

4

No ideas? No troubleshooting steps?

This is seriously affecting my ability to type, and has become frequent enough that the bug causes nonsense messages to be sent in IM/chat clients.

5

Hi,

It seems like it’s trying to open a icon file and gets forcibly shut down because of it…try running it from a Terminal with the trace option enabled to see if it logs a file name:

/Applications/WhatPulse.app/Contents/MacOS/whatpulse --verbose

6

All I got after the crash was this:

whatpulse(59622,0x7fff73cf9300) malloc: *** error for object 0x7fb109a62608: incorrect checksum for freed object - object was probably modified after being freed.
*** set a breakpoint in malloc_error_break to debug

Looks like the same error from the crash reports. :confused:

7

Look before the crash…

8

Last few lines were about 20 minutes before the crash:

18-12-2015 16:46:26.249 DEBUG WhatPulse  2.6.3  has started 
18-12-2015 16:46:28.084 DEBUG Starting timers and collector.. 
18-12-2015 16:46:28.085 DEBUG Starting PcapNetworkTrafficMonitor.. 
18-12-2015 16:46:28.097 DEBUG Found new interface:  4 "en0" "3C:07:54:11:ED:FA" "Ethernet" 
18-12-2015 16:46:28.097 DEBUG Found new interface:  5 "en1" "28:CF:DA:E2:94:EE" "Wi-Fi" 
18-12-2015 16:46:28.097 DEBUG PcapNetworkTrafficMonitor: Starting interface  5 
18-12-2015 16:46:28.098 DEBUG pcap thread started on interface  "en1"

That’s it up until the errors above.

9

It appears you did not supply the ‘–verbose’ parameter, otherwise it would give you a lot more.

10

I copied the command exactly as you posted a few comments above. That’s all it gave me.

Edit: You meant the app.log, didn’t you? One sec.
[hr]
Here’s the last minute or so of the last run. It takes 15-20 minutes to crash sometimes.

18-12-2015 17:52:10.246 TRACE Settings::get( "monochrome_icon" ) 
18-12-2015 17:52:10.255 TRACE Settings::get( "monochrome_icon" ) 
18-12-2015 17:52:10.413 TRACE Settings::get( "check_for_updates" ) 
18-12-2015 17:52:10.414 TRACE Settings::get( "online_send_heatmaps" ) 
18-12-2015 17:52:10.414 TRACE Settings::has( "online_last_heatmap_keyboard" ) 
18-12-2015 17:52:10.414 TRACE Settings::get( "online_last_heatmap_keyboard" ) 
18-12-2015 17:52:10.435 TRACE Query  "exec(...)"  took too much time! 
18-12-2015 17:52:10.435 TRACE   Time:  21 ms 
18-12-2015 17:52:10.435 TRACE   Query string:  "SELECT day, hour, key, count FROM keypress_frequency WHERE day BETWEEN '2015-12-17 0' AND '2015-12-17' ORDER BY day" 
18-12-2015 17:52:10.435 TRACE Settings::has( "online_last_heatmap_keyboard_apps" ) 
18-12-2015 17:52:10.435 TRACE Settings::get( "online_last_heatmap_keyboard_apps" ) 
18-12-2015 17:52:10.477 TRACE Query  "exec(...)"  took too much time! 
18-12-2015 17:52:10.477 TRACE   Time:  42 ms 
18-12-2015 17:52:10.477 TRACE   Query string:  "SELECT DISTINCT a.online_app_id, f.path FROM keypress_frequency_application f, applications a WHERE a.path = f.path AND a.online_app_id > 0 AND f.day BETWEEN '2015-12-17' AND '2015-12-17' ORDER BY a.online_app_id" 
18-12-2015 17:52:10.478 TRACE Settings::has( "online_last_heatmap_buttons" ) 
18-12-2015 17:52:10.478 TRACE Settings::get( "online_last_heatmap_buttons" ) 
18-12-2015 17:52:10.478 TRACE Settings::has( "online_last_heatmap_buttons_apps" ) 
18-12-2015 17:52:10.478 TRACE Settings::get( "online_last_heatmap_buttons_apps" ) 
18-12-2015 17:52:10.479 TRACE Settings::has( "online_last_heatmap_keycombinations" ) 
18-12-2015 17:52:10.480 TRACE Settings::get( "online_last_heatmap_keycombinations" ) 
18-12-2015 17:52:10.480 TRACE Settings::has( "online_last_heatmap_keycombinations_apps" ) 
18-12-2015 17:52:10.480 TRACE Settings::get( "online_last_heatmap_keycombinations_apps" ) 
18-12-2015 17:52:10.810 TRACE Settings::get( "monochrome_icon" ) 
18-12-2015 17:52:10.816 TRACE Settings::get( "monochrome_icon" ) 
18-12-2015 17:52:12.336 TRACE Settings::get( "monochrome_icon" ) 
18-12-2015 17:52:12.340 TRACE Settings::get( "monochrome_icon" ) 
18-12-2015 17:52:13.617 TRACE Settings::get( "monochrome_icon" ) 
18-12-2015 17:52:13.624 TRACE Settings::get( "monochrome_icon" ) 
18-12-2015 17:52:15.053 TRACE Settings::get( "monochrome_icon" ) 
18-12-2015 17:52:15.061 TRACE Settings::get( "monochrome_icon" ) 
18-12-2015 17:52:18.414 TRACE Settings::get( "stat_keyboard_enable" ) 
18-12-2015 17:52:18.414 TRACE Settings::get( "stat_mouse_enable" ) 
18-12-2015 17:52:18.414 TRACE Settings::get( "stat_uptime_enable" ) 
18-12-2015 17:52:18.414 TRACE Settings::get( "stats_input_enable" ) 
18-12-2015 17:52:18.415 TRACE Settings::get( "stats_net_enable" ) 
18-12-2015 17:52:18.496 TRACE Settings::get( "monochrome_icon" ) 
18-12-2015 17:52:18.503 TRACE Settings::get( "monochrome_icon" ) 
18-12-2015 17:52:20.731 TRACE Settings::get( "monochrome_icon" ) 
18-12-2015 17:52:20.738 TRACE Settings::get( "monochrome_icon" ) 
18-12-2015 17:52:21.460 TRACE Settings::get( "monochrome_icon" ) 
18-12-2015 17:52:21.468 TRACE Settings::get( "monochrome_icon" ) 
18-12-2015 17:52:21.686 TRACE Settings::get( "monochrome_icon" ) 
18-12-2015 17:52:21.692 TRACE Settings::get( "monochrome_icon" ) 
18-12-2015 17:52:28.413 TRACE Settings::get( "stat_keyboard_enable" ) 
18-12-2015 17:52:28.413 TRACE Settings::get( "stat_mouse_enable" ) 
18-12-2015 17:52:28.414 TRACE Settings::get( "stat_uptime_enable" ) 
18-12-2015 17:52:28.414 TRACE Settings::get( "stats_input_enable" ) 
18-12-2015 17:52:28.414 TRACE Settings::get( "stats_net_enable" ) 
18-12-2015 17:52:38.413 TRACE Settings::get( "stat_keyboard_enable" ) 
18-12-2015 17:52:38.414 TRACE Settings::get( "stat_mouse_enable" ) 
18-12-2015 17:52:38.414 TRACE Settings::get( "stat_uptime_enable" ) 
18-12-2015 17:52:38.414 TRACE Settings::get( "stats_input_enable" ) 
18-12-2015 17:52:38.414 TRACE Settings::get( "stats_net_enable" ) 
18-12-2015 17:52:48.413 TRACE Settings::get( "stat_keyboard_enable" ) 
18-12-2015 17:52:48.414 TRACE Settings::get( "stat_mouse_enable" ) 
18-12-2015 17:52:48.414 TRACE Settings::get( "stat_uptime_enable" ) 
18-12-2015 17:52:48.414 TRACE Settings::get( "stats_input_enable" ) 
18-12-2015 17:52:48.414 TRACE Settings::get( "stats_net_enable" ) 
18-12-2015 17:52:58.413 TRACE Settings::get( "stat_keyboard_enable" ) 
18-12-2015 17:52:58.413 TRACE Settings::get( "stat_mouse_enable" ) 
18-12-2015 17:52:58.413 TRACE Settings::get( "stat_uptime_enable" ) 
18-12-2015 17:52:58.413 TRACE Settings::get( "stats_input_enable" ) 
18-12-2015 17:52:58.413 TRACE Settings::get( "stats_net_enable" ) 
18-12-2015 17:53:08.413 TRACE Settings::get( "stat_keyboard_enable" ) 
18-12-2015 17:53:08.413 TRACE Settings::get( "stat_mouse_enable" ) 
18-12-2015 17:53:08.413 TRACE Settings::get( "stat_uptime_enable" ) 
18-12-2015 17:53:08.413 TRACE Settings::get( "stats_input_enable" ) 
18-12-2015 17:53:08.413 TRACE Settings::get( "stats_net_enable" ) 
18-12-2015 17:53:10.412 TRACE Settings::get( "online_send_applications" ) 
18-12-2015 17:53:10.412 TRACE Settings::get( "work_offline" ) 
Segmentation fault: 11

Sadly, still no filename.

11

This is still occurring, despite redownloading the WhatPulse installer and quitting WhatPulse, running the installation again, and starting it once more.

What other things can I try, if debug logging isn’t showing a filename?

12

Given the continued silence, I finally had to try the nuclear option. Quit WhatPulse, renamed ~/Library/WhatPulse to ~/Library/WhatPulse.crashy and started afresh. Might be overly optimistic, but 45 minutes later WhatPulse hasn’t crashed (previously the crash rate was at least once every 15-20 min).

My guess is that something in one of the preference files was corrupted. Any interest in getting the files that caused WP to be crashy, @smitmartijin? I’d proactively post them here but for all I know there are things in there that shouldn’t be made available on a public forum.