bug#17753: Cygwin emacs-X11 core dump

[markus.hoenicka@mhoenicka.de]

Ken Brown writes:
 > A bug in the Cygwin DLL (64-bit only) was just discovered and fixed. 
 > The bug could have prevented stack space from being committed after the 
 > initial commit.  Could you try the latest snapshot from 
 > https://cygwin.com/snapshots/ and see if these strange crashes stop?  I 
 > realize that you may have to run emacs for several weeks before you can 
 > be confident.
 > 

Hi,

I don't know if the latest official Cygwin DLL contains any additional
bugfixes that are relevant for the Emacs crashes, but I've upgraded my
system anyway, using the latest Cygwin DLL and Emacs as of Aug
4th. Emacs crashed on the very same day, but not once during the next
two days. The core dumps look as useless as ever, although slightly
different compared to before the upgrade:

(gdb) thread apply all bt

Thread 6 (Thread 0x14dc):
#0  0x0004003300001f80 in ?? ()
#1  0x0000000000000000 in ?? ()

Thread 5 (Thread 0x57c):
#0  0x0000003300001f80 in ?? ()
#1  0x000007fefda010dc in ?? ()
#2  0x0000000000000000 in ?? ()

Thread 4 (Thread 0x1300):
#0  0x0000003300001f80 in ?? ()
#1  0x000007fefda010dc in ?? ()
#2  0x0000000000000000 in ?? ()

Thread 3 (Thread 0x1080):
#0  0x0000003300001f80 in ?? ()
#1  0x00000000777eb037 in ?? ()
#2  0x0000000000000002 in ?? ()
#3  0x0000000000000000 in ?? ()

Thread 2 (Thread 0xf38):
#0  0x0000003300001f80 in ?? ()
#1  0x000007fefda01a7a in ?? ()
#2  0x000000000042ce00 in ?? ()
#3  0x0000000100582620 in schedule_atimer ()
    at /usr/src/debug/emacs-24.3-7/src/atimer.c:343
#4  0x00000001802de9f8 in ?? ()
#5  0x0000000000000000 in ?? ()

Thread 1 (Thread 0x11b4):
#0  0x0000003300001fa4 in ?? ()
#1  0x0000000100010000 in ?? ()
#2  0x0000000000000000 in ?? ()

$ uname -a
CYGWIN_NT-6.1 SBHC123 1.7.31(0.272/5/3) 2014-07-25 11:26 x86_64 Cygwin
$ cygcheck -f /usr/bin/emacs-X11.exe
emacs-X11-24.3-7

regards,
Markus

-- 
Markus Hoenicka
http://www.mhoenicka.de
AQ score 38