Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <7FF29DD088644C13860035D63055EC89@D9VGLK61>
Date: Wed, 21 Sep 2011 19:27:44 -0500
From: "JimF" <jfoug@....net>
To: <john-dev@...ts.openwall.com>
Subject: Re: Crash bug on win32-cygwin-x86-any builds

Try running the excutable (the x86-any), and run it doing 'real' work.

To test, simply make a few lines 32 bytes long of hex chars, and test it 
using raw-md5 or md5_gen(0).

The builds I was doing would crash in an instant, doing 'real' work.  But 
ONLY the x86-any build.

Jim.
  ----- Original Message ----- 
  From: Robert Harris
  To: john-dev@...ts.openwall.com
  Sent: Wednesday, September 21, 2011 7:22 PM
  Subject: RE: [john-dev] Crash bug on win32-cygwin-x86-any builds


  John-Dev,



  I successfully compiled JtR 1.7.8 jumbo-6-RC5 with patches 001-005 in 
Cygwin with win32-cygwin-x86-any and win32-cygwin-x86-sse.   (My gcc is 
v4.6.0)



  It required no modifications.



  -Robert B. Harris from VA



  From: jfoug [mailto:jfoug@....net]
  Sent: Wednesday, September 21, 2011 12:04 PM
  To: john-dev@...ts.openwall.com
  Subject: [john-dev] Crash bug on win32-cygwin-x86-any builds



  This may be 'similar' to what Alex is seeing on the Mac OSX (or may be 
totally different).  This crash ONLY happens on win32-cygwin-x86-any.  All 
other builds, work just fine.



  I have gotten it to work.  If I build john.o without -O2, things work. 
Otherwise, I get a core, with this type message:



  john 3692 exception::handle: Error while dumping state (probably corrupted 
stack)



  There is a john.exe.stackdump, but there is nothing in there, other than 
text like the line above (probably corrupted stack).





  NO other builds I have show this problem.  However, if I do a full build, 
then remove -O2 from the CFLAGS, then rm john.o so it rebuilds just that 
file, perform the make, and the file is happy as can be.



  When I run the 'broken' version under gdb (I know nothing about gdb, so 
this will be a trial and error thing), it simply bails out, listing:



  Program received signal SIGSEGV, Segmentation fault.

  0x0045bb27 in ?? ()



  So it appears GDB is not getting any form of stack frame.



  Jim.





Content of type "text/html" skipped

Powered by blists - more mailing lists

Confused about mailing lists and their use? Read about mailing lists on Wikipedia and check out these guidelines on proper formatting of your messages.