7.0R panic in nfs_readdirrpc()

[Hiroki Sato]

This panic seems reproducible when NFS over TCP is used (RELENG_7_0 on
HP rx1600, btw):

----
fatal kernel trap (cpu 0):

    trap vector = 0x1e (Unaligned Reference)
    cr.iip      = 0xe000000004814a60
    cr.ipsr     = 0x1210080a6010 (mfl,ic,i,dt,dfh,rt,cpl=0,it,ri=1,bn)
    cr.isr      = 0xa0400000000 (code=0,vector=0,r,ei=1,ed)
    cr.ifa      = 0xe00000000e745335
    curthread   = 0xe0000000119366c0
        pid = 13638, comm = cvs

[thread pid 13638 tid 100134 ]
Stopped at      nfs_readdirrpc+0xd11:   [M1] (p17) ld4 r49=[r8],0x4
db> bt
Tracing pid 13638 tid 100134 td 0xe0000000119366c0
nfs_readdirrpc(0xe0000000279db228, 0xa00000001b2d3178, 0xe00000001078fd00, 0xd, 0x10) at nfs_readdirrpc+0xd11
nfs_doio(0xe000000011596f80, 0xa00000000f5eb4a0, 0xe00000001078fd00, 0xe0000000119366c0) at nfs_doio+0x670
nfs_bioread(0xe000000011596f80, 0xa00000001b2d3378, 0x0, 0xe00000001078fd00) at nfs_bioread+0x10a0
nfs_readdir(0xa00000001b2d33c8, 0xe000000011596f80, 0xa00000001b2d3390) at nfs_readdir+0x290
VOP_READDIR_APV(0xe000000004b16120, 0xa00000001b2d33c8, 0xe0000000119366c0, 0xe000000004637390, 0x13ac) at VOP_READDIR_APV+0x1c0
getdirentries(0xe0000000119366c0, 0xa00000001b2d34e8, 0x0, 0xe000000011596f80) at getdirentries+0x410
syscall(0xa00000001b2d3400, 0xc4, 0x2000, 0xe0000000119366c0, 0xe00000001130ed38, 0xe000000004af3140, 0xc4, 0xa00000001b2d34e8) at syscall+0x410
epc_syscall_return() at epc_syscall_return
db>
----

--
| Hiroki SATO
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 195 bytes
Desc: not available
Url : http://lists.freebsd.org/pipermail/freebsd-ia64/attachments/20080725/8f3de6f2/attachment.pgp