Netgraph/mpd5 stability issues

[Gleb Smirnoff]

On Mon, Jan 31, 2011 at 02:40:55PM +0100, Przemyslaw Frasunek wrote:
P> > In this dump, can we seek for where did 0x74 came from? Can you look at
P> > ng_name_hash[hash]?
P> 
P> (kgdb) print hash
P> No symbol "hash" in current context.
P> (kgdb) info all
P> eax            0xffffff9a       -102
P> ecx            0xe7ce6895       -405903211
P> edx            0xffffff9a       -102
P> ebx            0x74     116
P> esp            0xcc28c380       0xcc28c380
P> ebp            0xe7ce6878       0xe7ce6878
P> esi            0xe7ce6894       -405903212
P> edi            0xe7ce6ac8       -405902648
P> [...]
P> (kgdb) print ng_name_hash[116]
P> $4 = {lh_first = 0xcbd6fb00}
P> (kgdb) print *ng_name_hash[116].lh_first
P> $5 = {nd_name = "ng357", '\0' <repeats 26 times>, nd_type = 0xc61871a0,
P>   nd_flags = 0, nd_refs = 1, nd_numhooks = 0, nd_private = 0xcc3b5e80,
P>   nd_ID = 454681, nd_hooks = {lh_first = 0x0}, nd_nodes = {
P>     le_next = 0xcbbf7900, le_prev = 0xc5f4fad0}, nd_idnodes = {
P>     le_next = 0xcc061680, le_prev = 0xc5f4f744}, nd_work = {tqe_next = 0x0,
P>     tqe_prev = 0x0}, nd_input_queue = {q_flags = 0, q_mtx = {lock_object = {
P>         lo_name = 0xc5f4e986 "ng_node", lo_type = 0xc5f4e986 "ng_node",
P>         lo_flags = 16973824, lo_witness_data = {lod_list = {stqe_next = 0x0},
P>           lod_witness = 0x0}}, mtx_lock = 4, mtx_recurse = 0}, queue = 0x0,
P>     last = 0xcbd6fb70, q_node = 0xcbd6fb00}}

Thanks. Yep, "ng357" also has hash of 116.

Can we look at ng_name_hash[116].lh_first->nd_nodes.le_next and further? Which
one does point to 0x74?

-- 
Totus tuus, Glebius.