crash after updating iw and libnl to latest.

[Ben Greear]

I updated everything yesterday, and now iw crashes every time.
The biggest change was that I updated libnl to the latest....

I'll start digging, but will not be offended if someone
finds it first!


[root at lec2010-ath9k-1 lanforge]# ldd ./local/sbin/iw
	linux-gate.so.1 =>  (0xb78bb000)
	libnl-genl.so.3 => /home/lanforge/local/lib/libnl-genl.so.3 (0xb78b6000)
	libnl-genl-3.so.200 => /home/lanforge/local/lib/libnl-genl-3.so.200 (0xb78b2000)
	libnl-3.so.200 => /home/lanforge/local/lib/libnl-3.so.200 (0xb789f000)
	libc.so.6 => /lib/libc.so.6 (0x42b28000)
	libnl.so.3 => /home/lanforge/local/lib/libnl.so.3 (0xb7878000)
	libm.so.6 => /lib/libm.so.6 (0x42cfa000)
	libpthread.so.0 => /lib/libpthread.so.0 (0x42cb4000)
	/lib/ld-linux.so.2 (0x42b07000)
[root at lec2010-ath9k-1 lanforge]# ./local/sbin/iw phy
nl80211 not found.
Segmentation fault

Core was generated by `./local/sbin/iw phy'.
Program terminated with signal 11, Segmentation fault.
#0  nl_list_del (ops=0xb780d2c0) at ../include/netlink/list.h:51
51	../include/netlink/list.h: No such file or directory.
	in ../include/netlink/list.h
(gdb) bt
#0  nl_list_del (ops=0xb780d2c0) at ../include/netlink/list.h:51
#1  genl_unregister (ops=0xb780d2c0) at genl/mngt.c:212
#2  0xb780b440 in ctrl_exit () at genl/ctrl.c:377
#3  0xb780b3b8 in __do_global_dtors_aux () from /home/lanforge/local/lib/libnl-genl-3.so.200
#4  0xb780cb50 in _fini () from /home/lanforge/local/lib/libnl-genl-3.so.200
#5  0x42b15f2d in _dl_fini () at dl-fini.c:249
#6  0x42b56cdf in __run_exit_handlers (status=1) at exit.c:78
#7  exit (status=1) at exit.c:100
#8  0x42b3ee3e in __libc_start_main (main=0x8049c60 <main>, argc=2, ubp_av=0xbfe93c74, init=0x80553b0 <__libc_csu_init>,
     fini=0x80553a0 <__libc_csu_fini>, rtld_fini=0x42b15d50 <_dl_fini>, stack_end=0xbfe93c6c) at libc-start.c:258
#9  0x08049261 in _start ()
(gdb)

[root at lec2010-ath9k-1 lanforge]# valgrind ./local/sbin/iw phy
==25107== Memcheck, a memory error detector
==25107== Copyright (C) 2002-2010, and GNU GPL'd, by Julian Seward et al.
==25107== Using Valgrind-3.7.0.SVN and LibVEX; rerun with -h for copyright info
==25107== Command: ./local/sbin/iw phy
==25107==
nl80211 not found.
==25107== Invalid read of size 4
==25107==    at 0x400F7C5: genl_unregister (list.h:51)
==25107==    by 0x401343F: ctrl_exit (ctrl.c:377)
==25107==    by 0x40133B7: ??? (in /home/lanforge/local/lib/libnl-genl-3.so.200.3.0)
==25107==    by 0x4014B4F: ??? (in /home/lanforge/local/lib/libnl-genl-3.so.200.3.0)
==25107==    by 0x42B15F2C: _dl_fini (dl-fini.c:249)
==25107==    by 0x42B56CDE: exit (exit.c:78)
==25107==    by 0x42B3EE3D: (below main) (libc-start.c:258)
==25107==  Address 0x18 is not stack'd, malloc'd or (recently) free'd
==25107==
==25107==
==25107== Process terminating with default action of signal 11 (SIGSEGV): dumping core
==25107==  Access not within mapped region at address 0x18
==25107==    at 0x400F7C5: genl_unregister (list.h:51)
==25107==    by 0x401343F: ctrl_exit (ctrl.c:377)
==25107==    by 0x40133B7: ??? (in /home/lanforge/local/lib/libnl-genl-3.so.200.3.0)
==25107==    by 0x4014B4F: ??? (in /home/lanforge/local/lib/libnl-genl-3.so.200.3.0)
==25107==    by 0x42B15F2C: _dl_fini (dl-fini.c:249)
==25107==    by 0x42B56CDE: exit (exit.c:78)
==25107==    by 0x42B3EE3D: (below main) (libc-start.c:258)
==25107==  If you believe this happened as a result of a stack
==25107==  overflow in your program's main thread (unlikely but
==25107==  possible), you can try to increase the size of the
==25107==  main thread stack using the --main-stacksize= flag.
==25107==  The main thread stack size used in this run was 8388608.
==25107==
==25107== HEAP SUMMARY:
==25107==     in use at exit: 0 bytes in 0 blocks
==25107==   total heap usage: 11 allocs, 11 frees, 9,420 bytes allocated
==25107==
==25107== All heap blocks were freed -- no leaks are possible
==25107==
==25107== For counts of detected and suppressed errors, rerun with: -v
==25107== ERROR SUMMARY: 1 errors from 1 contexts (suppressed: 22 from 8)
Segmentation fault

-- 
Ben Greear <greearb at candelatech.com>
Candela Technologies Inc  http://www.candelatech.com