On Fri, Nov 30, 2007 at 12:21:15PM -0800, jt wrote:
> On Fri, Nov 30, 2007 at 02:42:37PM +1100, Shaddy Baddah wrote:
> >
> > No, it does not seem to have. An iwlist eth2 scanning command still
> > lists a single access point, then fails with a bus error.
>
> Hmm... That's strange as I was seeing my code behaving as
> expected.
> I will ask you to compile a version of wireless tools with
> debugging enabled and run it with gdb. See my other e-mail.
Hi,
Sorry to bother you again. Next week I will go on vacation,
and may not be able to continue working on that bug. Your best bet for
a fix is this week.
Now, this is what you will do :
1) Get wireless_tools.30.pre3-debug2.tar.gz I sent you a while
ago. I know you have it, as you sent me debug log from it.
2) Go in the wireless_tools.30 directory.
3) edit Makefile. Go to line defining CFLAGS, replace "-Os"
with "-g". It should look like :
---------------------------------------
CFLAGS=-g -W -Wall -Wstrict-prototypes -Wmissing-prototypes -Wshadow \
-Wpointer-arith -Wcast-qual -Winline -I.
---------------------------------------
4) Recompile : make clean ; make
5) Start the debugger : gdb --args iwlist eth0 scan
6) Run the program : run
7) Get a stack trace : bt
If you have properly enabled debug, the stack trace will show
the precise line number where it crashes. That's mostly what I'm
after.
I would also like to have the debugging output of iwlist. The
function where it seems to crash is called many time, I want to know
in which context it is called. Feel free to send it private.
Thanks in advance...
Jean
-
To unsubscribe from this list: send the line "unsubscribe linux-wireless" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
