Re: New version of w3m 0.5.2 Crashes - GDB help needed

[Christopher Faylor <cgf-use-the-mailinglist-please at cygwin dot com>]

On Wed, Jan 05, 2011 at 04:16:37PM -0500, nyc4bos@aol.com wrote:
>nyc4bos@aol.com writes:
>
>I've been working with Bob on this but I think we've at
>an impass (see below):
>
>> "Robert S. Heckel Jr." <b.heckel@gmail.com> writes:
>>
>>> On Wed, Dec 15, 2010 at 21:07, Robert S. Heckel Jr. <b.heckel@gmail.com> wrote:
>>>> Will let you know if I find anything. ??Anyone else experience 0.5.2-1 crashes?
>>>
>>> Since I haven't heard of problems from other people (yet at least) and
>>> I can't replicate the crash, you may want to try either a reinstall
>>> of:
>>>
>>> w3m libgc1 libintl8 libncurses10 libopenssl098 zlib0
>>>
>>> or, better, a full reinstall of cygwin.
>>
>> I tried resinstalling those but it still got the same errors.
>>
>> I'm hesitant to reinstall a full cygwin...
>>>
>>> Bob
>
>Bob created a debug version of w3m for me.
>
>Here's what I see:
>
>The crash occurs when doin a `strncmp' in cygwin1.dll:
>
>      4 [main] w3m 3024 exception::handle: Exception: STATUS_ACCESS_VIOLATION
>   1564 [main] w3m 3024 open_stackdumpfile: Dumping stack trace to w3m.exe.stack
>dump
>
>Program received signal SIGSEGV, Segmentation fault.
>0x6110daf5 in strncmp () from /usr/bin/cygwin1.dll
>(gdb) bt full
>#0  0x6110daf5 in strncmp () from /usr/bin/cygwin1.dll
>No symbol table info available.
>#1  0x7c85ada0 in OutputDebugStringA ()
>   from /cygdrive/c/WINDOWS/system32/kernel32.dll
>No symbol table info available.
>#2  0x40010006 in ?? ()
>No symbol table info available.
>#3  0x00000000 in ?? ()
>No symbol table info available.
>(gdb)

Looks to me like you need a debugging version of w3m with symbols.
You need to find out what's being passed to strncmp, possibly by setting a
breakpoint at 0x40110006.

cgf

--
Problem reports:       http://cygwin.com/problems.html
FAQ:                   http://cygwin.com/faq/
Documentation:         http://cygwin.com/docs.html
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple